From d16e6d19ccc6d3aa6b96d6a8fdb9e04fb9dffdbd Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Tue, 5 Oct 2021 18:09:33 -0700 Subject: [PATCH 0001/2048] Bluetooth: hci_vhci: Fix calling hci_{suspend,resume}_dev Defer calls to hci_{suspend,resume}_dev to work so it doesn't block the processing of the events. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_vhci.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index b45db0db347c..20f8ce995555 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -38,6 +38,7 @@ struct vhci_data { struct mutex open_mutex; struct delayed_work open_timeout; + struct work_struct suspend_work; bool suspended; bool wakeup; @@ -114,6 +115,17 @@ static ssize_t force_suspend_read(struct file *file, char __user *user_buf, return simple_read_from_buffer(user_buf, count, ppos, buf, 2); } +static void vhci_suspend_work(struct work_struct *work) +{ + struct vhci_data *data = container_of(work, struct vhci_data, + suspend_work); + + if (data->suspended) + hci_suspend_dev(data->hdev); + else + hci_resume_dev(data->hdev); +} + static ssize_t force_suspend_write(struct file *file, const char __user *user_buf, size_t count, loff_t *ppos) @@ -129,16 +141,10 @@ static ssize_t force_suspend_write(struct file *file, if (data->suspended == enable) return -EALREADY; - if (enable) - err = hci_suspend_dev(data->hdev); - else - err = hci_resume_dev(data->hdev); - - if (err) - return err; - data->suspended = enable; + schedule_work(&data->suspend_work); + return count; } @@ -440,6 +446,7 @@ static int vhci_open(struct inode *inode, struct file *file) mutex_init(&data->open_mutex); INIT_DELAYED_WORK(&data->open_timeout, vhci_open_timeout); + INIT_WORK(&data->suspend_work, vhci_suspend_work); file->private_data = data; nonseekable_open(inode, file); @@ -455,6 +462,7 @@ static int vhci_release(struct inode *inode, struct file *file) struct hci_dev *hdev; cancel_delayed_work_sync(&data->open_timeout); + flush_work(&data->suspend_work); hdev = data->hdev; From 83775456504c251e6e4e1ee50c470e77202b8d21 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Tue, 5 Oct 2021 18:09:34 -0700 Subject: [PATCH 0002/2048] Bluetooth: Fix handling of SUSPEND_DISCONNECTING When SUSPEND_DISCONNECTING bit is set that means Disconnect is pending but the code was evaluating if the list is empty before calling hci_conn_del which does the actual cleanup and remove the connection from the list thus the bit is never cleared causing the suspend procedure to always timeout when there are connections to be disconnected: Suspend/Resume - Success 5 (Pairing - Legacy) - waiting done Set the system into Suspend via force_suspend = mgmt-tester: Suspend/Resume - Success 5 (Pairing -.. 17:03:13.200458 = mgmt-tester: Set the system into Suspend via force_suspend 17:03:13.205812 < HCI Command: Write Scan E.. (0x03|0x001a) plen 1 #122 [hci0] 17:03:13.213561 Scan enable: No Scans (0x00) > HCI Event: Command Complete (0x0e) plen 4 #123 [hci0] 17:03:13.214710 Write Scan Enable (0x03|0x001a) ncmd 1 Status: Success (0x00) < HCI Command: Disconnect (0x01|0x0006) plen 3 #124 [hci0] 17:03:13.215830 Handle: 42 Reason: Remote Device Terminated due to Power Off (0x15) > HCI Event: Command Status (0x0f) plen 4 #125 [hci0] 17:03:13.216602 Disconnect (0x01|0x0006) ncmd 1 Status: Success (0x00) > HCI Event: Disconnect Complete (0x05) plen 4 #126 [hci0] 17:03:13.217342 Status: Success (0x00) Handle: 42 Reason: Remote Device Terminated due to Power Off (0x15) @ MGMT Event: Device Disconn.. (0x000c) plen 8 {0x0002} [hci0] 17:03:13.217688 BR/EDR Address: 00:AA:01:01:00:00 (Intel Corporation) Reason: Connection terminated by local host for suspend (0x05) @ MGMT Event: Device Disconn.. (0x000c) plen 8 {0x0001} [hci0] 17:03:13.217688 BR/EDR Address: 00:AA:01:01:00:00 (Intel Corporation) Reason: Connection terminated by local host for suspend (0x05) Suspend/Resume - Success 5 (Pairing - Legacy) - test timed out = mgmt-tester: Suspend/Resume - Success 5 (Pairing -.. 17:03:13.939317 Suspend/Resume - Success 5 (Pairing - Legacy) - teardown = mgmt-tester: Suspend/Resume - Success 5 (Pairing -.. 17:03:13.947267 [ 13.284291] Bluetooth: hci0: Timed out waiting for suspend events [ 13.287324] Bluetooth: hci0: Suspend timeout bit: 6 Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_event.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 7d0db1ca1248..3cba2bbefcd6 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -2987,14 +2987,6 @@ static void hci_disconn_complete_evt(struct hci_dev *hdev, struct sk_buff *skb) hci_disconn_cfm(conn, ev->reason); - /* The suspend notifier is waiting for all devices to disconnect so - * clear the bit from pending tasks and inform the wait queue. - */ - if (list_empty(&hdev->conn_hash.list) && - test_and_clear_bit(SUSPEND_DISCONNECTING, hdev->suspend_tasks)) { - wake_up(&hdev->suspend_wait_q); - } - /* Re-enable advertising if necessary, since it might * have been disabled by the connection. From the * HCI_LE_Set_Advertise_Enable command description in @@ -3012,6 +3004,14 @@ static void hci_disconn_complete_evt(struct hci_dev *hdev, struct sk_buff *skb) hci_conn_del(conn); + /* The suspend notifier is waiting for all devices to disconnect so + * clear the bit from pending tasks and inform the wait queue. + */ + if (list_empty(&hdev->conn_hash.list) && + test_and_clear_bit(SUSPEND_DISCONNECTING, hdev->suspend_tasks)) { + wake_up(&hdev->suspend_wait_q); + } + unlock: hci_dev_unlock(hdev); } From 107fe0482b549a0e43a971e5fd104719c6e495ef Mon Sep 17 00:00:00 2001 From: Kiran K Date: Tue, 5 Oct 2021 20:15:56 +0530 Subject: [PATCH 0003/2048] Bluetooth: Read codec capabilities only if supported Read codec capabilities only if HCI_READ_LOCAL_CODEC_CAPABILITIES command is supported. If capablities are not supported, then cache codec data without caps. Signed-off-by: Kiran K Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_codec.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/hci_codec.c b/net/bluetooth/hci_codec.c index f0421d0edaa3..38201532f58e 100644 --- a/net/bluetooth/hci_codec.c +++ b/net/bluetooth/hci_codec.c @@ -25,9 +25,11 @@ static int hci_codec_list_add(struct list_head *list, } entry->transport = sent->transport; entry->len = len; - entry->num_caps = rp->num_caps; - if (rp->num_caps) + entry->num_caps = 0; + if (rp) { + entry->num_caps = rp->num_caps; memcpy(entry->caps, caps, len); + } list_add(&entry->list, list); return 0; @@ -58,6 +60,18 @@ static void hci_read_codec_capabilities(struct hci_dev *hdev, __u8 transport, __u32 len; cmd->transport = i; + + /* If Read_Codec_Capabilities command is not supported + * then just add codec to the list without caps + */ + if (!(hdev->commands[45] & 0x08)) { + hci_dev_lock(hdev); + hci_codec_list_add(&hdev->local_codecs, cmd, + NULL, NULL, 0); + hci_dev_unlock(hdev); + continue; + } + skb = __hci_cmd_sync(hdev, HCI_OP_READ_LOCAL_CODEC_CAPS, sizeof(*cmd), cmd, HCI_CMD_TIMEOUT); From 8b89637dbac2d73d9f3dadf91b4a7dcdb1fc23af Mon Sep 17 00:00:00 2001 From: Tedd Ho-Jeong An Date: Mon, 4 Oct 2021 10:01:26 -0700 Subject: [PATCH 0004/2048] Bluetooth: hci_vhci: Fix to set the force_wakeup value This patch sets the wakeup state of the vhci driver when the force_wakeup is updated. Fixes: 60edfad4fd0b6 ("Bluetooth: hci_vhci: Add force_prevent_wake entry") Signed-off-by: Tedd Ho-Jeong An Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_vhci.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index 20f8ce995555..56c6b22be10b 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -182,6 +182,8 @@ static ssize_t force_wakeup_write(struct file *file, if (data->wakeup == enable) return -EALREADY; + data->wakeup = enable; + return count; } From b15bfa4df63529150df9ff0585675f728436e0c1 Mon Sep 17 00:00:00 2001 From: Tedd Ho-Jeong An Date: Wed, 6 Oct 2021 09:32:28 -0700 Subject: [PATCH 0005/2048] Bluetooth: mgmt: Fix Experimental Feature Changed event This patch fixes the controller index in the Experimental Features Changed event for the offload_codec and the quality_report features to use the actual hdev index instead of non-controller index(0xffff) so the client can receive the event and know which controller the event is for. Fixes: ad93315183285 ("Bluetooth: Add offload feature under experimental flag") Fixes: ae7d925b5c043 ("Bluetooth: Support the quality report events") Signed-off-by: Tedd Ho-Jeong An Signed-off-by: Marcel Holtmann --- net/bluetooth/mgmt.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 3e5283607b97..44683443300c 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -3927,7 +3927,9 @@ static int exp_debug_feature_changed(bool enabled, struct sock *skip) } #endif -static int exp_quality_report_feature_changed(bool enabled, struct sock *skip) +static int exp_quality_report_feature_changed(bool enabled, + struct hci_dev *hdev, + struct sock *skip) { struct mgmt_ev_exp_feature_changed ev; @@ -3935,7 +3937,7 @@ static int exp_quality_report_feature_changed(bool enabled, struct sock *skip) memcpy(ev.uuid, quality_report_uuid, 16); ev.flags = cpu_to_le32(enabled ? BIT(0) : 0); - return mgmt_limited_event(MGMT_EV_EXP_FEATURE_CHANGED, NULL, + return mgmt_limited_event(MGMT_EV_EXP_FEATURE_CHANGED, hdev, &ev, sizeof(ev), HCI_MGMT_EXP_FEATURE_EVENTS, skip); } @@ -4156,14 +4158,15 @@ static int set_quality_report_func(struct sock *sk, struct hci_dev *hdev, &rp, sizeof(rp)); if (changed) - exp_quality_report_feature_changed(val, sk); + exp_quality_report_feature_changed(val, hdev, sk); unlock_quality_report: hci_req_sync_unlock(hdev); return err; } -static int exp_offload_codec_feature_changed(bool enabled, struct sock *skip) +static int exp_offload_codec_feature_changed(bool enabled, struct hci_dev *hdev, + struct sock *skip) { struct mgmt_ev_exp_feature_changed ev; @@ -4171,7 +4174,7 @@ static int exp_offload_codec_feature_changed(bool enabled, struct sock *skip) memcpy(ev.uuid, offload_codecs_uuid, 16); ev.flags = cpu_to_le32(enabled ? BIT(0) : 0); - return mgmt_limited_event(MGMT_EV_EXP_FEATURE_CHANGED, NULL, + return mgmt_limited_event(MGMT_EV_EXP_FEATURE_CHANGED, hdev, &ev, sizeof(ev), HCI_MGMT_EXP_FEATURE_EVENTS, skip); } @@ -4229,7 +4232,7 @@ static int set_offload_codec_func(struct sock *sk, struct hci_dev *hdev, &rp, sizeof(rp)); if (changed) - exp_offload_codec_feature_changed(val, sk); + exp_offload_codec_feature_changed(val, hdev, sk); return err; } From 709fca500067524381e28a5f481882930eebac88 Mon Sep 17 00:00:00 2001 From: Nguyen Dinh Phi Date: Fri, 8 Oct 2021 03:04:24 +0800 Subject: [PATCH 0006/2048] Bluetooth: hci_sock: purge socket queues in the destruct() callback The receive path may take the socket right before hci_sock_release(), but it may enqueue the packets to the socket queues after the call to skb_queue_purge(), therefore the socket can be destroyed without clear its queues completely. Moving these skb_queue_purge() to the hci_sock_destruct() will fix this issue, because nothing is referencing the socket at this point. Signed-off-by: Nguyen Dinh Phi Reported-by: syzbot+4c4ffd1e1094dae61035@syzkaller.appspotmail.com Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_sock.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index d0dad1fafe07..446573a12571 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -889,10 +889,6 @@ static int hci_sock_release(struct socket *sock) } sock_orphan(sk); - - skb_queue_purge(&sk->sk_receive_queue); - skb_queue_purge(&sk->sk_write_queue); - release_sock(sk); sock_put(sk); return 0; @@ -2058,6 +2054,12 @@ static int hci_sock_getsockopt(struct socket *sock, int level, int optname, return err; } +static void hci_sock_destruct(struct sock *sk) +{ + skb_queue_purge(&sk->sk_receive_queue); + skb_queue_purge(&sk->sk_write_queue); +} + static const struct proto_ops hci_sock_ops = { .family = PF_BLUETOOTH, .owner = THIS_MODULE, @@ -2111,6 +2113,7 @@ static int hci_sock_create(struct net *net, struct socket *sock, int protocol, sock->state = SS_UNCONNECTED; sk->sk_state = BT_OPEN; + sk->sk_destruct = hci_sock_destruct; bt_sock_link(&hci_sk_list, sk); return 0; From f5ff291098f70a70b344df1e388596755c3c8315 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Thu, 7 Oct 2021 13:26:25 -0700 Subject: [PATCH 0007/2048] Bluetooth: L2CAP: Fix not initializing sk_peer_pid In order to group sockets being connected using L2CAP_MODE_EXT_FLOWCTL the pid is used but sk_peer_pid was not being initialized as it is currently only done for af_unix. Fixes: b48596d1dc25 ("Bluetooth: L2CAP: Add get_peer_pid callback") Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- net/bluetooth/l2cap_sock.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index 160c016a5dfb..4574c5cb1b59 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -172,6 +172,21 @@ done: return err; } +static void l2cap_sock_init_pid(struct sock *sk) +{ + struct l2cap_chan *chan = l2cap_pi(sk)->chan; + + /* Only L2CAP_MODE_EXT_FLOWCTL ever need to access the PID in order to + * group the channels being requested. + */ + if (chan->mode != L2CAP_MODE_EXT_FLOWCTL) + return; + + spin_lock(&sk->sk_peer_lock); + sk->sk_peer_pid = get_pid(task_tgid(current)); + spin_unlock(&sk->sk_peer_lock); +} + static int l2cap_sock_connect(struct socket *sock, struct sockaddr *addr, int alen, int flags) { @@ -243,6 +258,8 @@ static int l2cap_sock_connect(struct socket *sock, struct sockaddr *addr, if (chan->psm && bdaddr_type_is_le(chan->src_type) && !chan->mode) chan->mode = L2CAP_MODE_LE_FLOWCTL; + l2cap_sock_init_pid(sk); + err = l2cap_chan_connect(chan, la.l2_psm, __le16_to_cpu(la.l2_cid), &la.l2_bdaddr, la.l2_bdaddr_type); if (err) @@ -298,6 +315,8 @@ static int l2cap_sock_listen(struct socket *sock, int backlog) goto done; } + l2cap_sock_init_pid(sk); + sk->sk_max_ack_backlog = backlog; sk->sk_ack_backlog = 0; From dc1650fc94a8566fb89f3fd14a26d1cec7865f16 Mon Sep 17 00:00:00 2001 From: David Yang Date: Wed, 13 Oct 2021 08:56:33 +0800 Subject: [PATCH 0008/2048] Bluetooth: btusb: Fix application of sizeof to pointer The coccinelle check report: "./drivers/bluetooth/btusb.c:2239:36-42: ERROR: application of sizeof to pointer". Using the real size to fix it. Fixes: 5a87679ffd443 ("Bluetooth: btusb: Support public address configuration for MediaTek Chip.") Reported-by: Zeal Robot Signed-off-by: David Yang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btusb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index 75c83768c257..3e26f60b457e 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2236,7 +2236,7 @@ static int btusb_set_bdaddr_mtk(struct hci_dev *hdev, const bdaddr_t *bdaddr) struct sk_buff *skb; long ret; - skb = __hci_cmd_sync(hdev, 0xfc1a, sizeof(bdaddr), bdaddr, HCI_INIT_TIMEOUT); + skb = __hci_cmd_sync(hdev, 0xfc1a, 6, bdaddr, HCI_INIT_TIMEOUT); if (IS_ERR(skb)) { ret = PTR_ERR(skb); bt_dev_err(hdev, "changing Mediatek device address failed (%ld)", From 5a4bb6a8e981d3d0d492aa38412ee80b21033177 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Wed, 13 Oct 2021 16:55:46 +0800 Subject: [PATCH 0009/2048] Bluetooth: Fix debugfs entry leak in hci_register_dev() Fault injection test report debugfs entry leak as follows: debugfs: Directory 'hci0' with parent 'bluetooth' already present! When register_pm_notifier() failed in hci_register_dev(), the debugfs create by debugfs_create_dir() do not removed in the error handing path. Add the remove debugfs code to fix it. Signed-off-by: Wei Yongjun Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 8d33aa64846b..98533def61a3 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3882,6 +3882,7 @@ int hci_register_dev(struct hci_dev *hdev) return id; err_wqueue: + debugfs_remove_recursive(hdev->debugfs); destroy_workqueue(hdev->workqueue); destroy_workqueue(hdev->req_workqueue); err: From 893505319c74cf3faa45a5ed9d5338ff03b66949 Mon Sep 17 00:00:00 2001 From: Kiran K Date: Wed, 13 Oct 2021 13:35:11 +0530 Subject: [PATCH 0010/2048] Bluetooth: btintel: Fix bdaddress comparison with garbage value Intel Read Verision(TLV) data is parsed into a local structure variable and it contains a field for bd address. Bd address is returned only in bootloader mode and hence bd address in TLV structure needs to be validated only if controller is present in boot loader mode. Signed-off-by: Kiran K Reviewed-by: Tedd Ho-Jeong An Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btintel.c | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/drivers/bluetooth/btintel.c b/drivers/bluetooth/btintel.c index 9359bff47296..8f9109b40961 100644 --- a/drivers/bluetooth/btintel.c +++ b/drivers/bluetooth/btintel.c @@ -2081,14 +2081,16 @@ static int btintel_prepare_fw_download_tlv(struct hci_dev *hdev, if (ver->img_type == 0x03) { btintel_clear_flag(hdev, INTEL_BOOTLOADER); btintel_check_bdaddr(hdev); - } - - /* If the OTP has no valid Bluetooth device address, then there will - * also be no valid address for the operational firmware. - */ - if (!bacmp(&ver->otp_bd_addr, BDADDR_ANY)) { - bt_dev_info(hdev, "No device address configured"); - set_bit(HCI_QUIRK_INVALID_BDADDR, &hdev->quirks); + } else { + /* + * Check for valid bd address in boot loader mode. Device + * will be marked as unconfigured if empty bd address is + * found. + */ + if (!bacmp(&ver->otp_bd_addr, BDADDR_ANY)) { + bt_dev_info(hdev, "No device address configured"); + set_bit(HCI_QUIRK_INVALID_BDADDR, &hdev->quirks); + } } btintel_get_fw_name_tlv(ver, fwname, sizeof(fwname), "sfi"); @@ -2466,6 +2468,10 @@ static int btintel_setup_combined(struct hci_dev *hdev) goto exit_error; } + /* memset ver_tlv to start with clean state as few fields are exclusive + * to bootloader mode and are not populated in operational mode + */ + memset(&ver_tlv, 0, sizeof(ver_tlv)); /* For TLV type device, parse the tlv data */ err = btintel_parse_version_tlv(hdev, &ver_tlv, skb); if (err) { From 75d9b8559ac36e059238ee4f8e33cd86086586ba Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Wed, 13 Oct 2021 16:55:01 +0800 Subject: [PATCH 0011/2048] Bluetooth: Fix memory leak of hci device Fault injection test reported memory leak of hci device as follows: unreferenced object 0xffff88800b858000 (size 8192): comm "kworker/0:2", pid 167, jiffies 4294955747 (age 557.148s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 ad 4e ad de .............N.. backtrace: [<0000000070eb1059>] kmem_cache_alloc_trace mm/slub.c:3208 [<00000000015eb521>] hci_alloc_dev_priv include/linux/slab.h:591 [<00000000dcfc1e21>] bpa10x_probe include/net/bluetooth/hci_core.h:1240 [<000000005d3028c7>] usb_probe_interface drivers/usb/core/driver.c:397 [<00000000cbac9243>] really_probe drivers/base/dd.c:517 [<0000000024cab3f0>] __driver_probe_device drivers/base/dd.c:751 [<00000000202135cb>] driver_probe_device drivers/base/dd.c:782 [<000000000761f2bc>] __device_attach_driver drivers/base/dd.c:899 [<00000000f7d63134>] bus_for_each_drv drivers/base/bus.c:427 [<00000000c9551f0b>] __device_attach drivers/base/dd.c:971 [<000000007f79bd16>] bus_probe_device drivers/base/bus.c:487 [<000000007bb8b95a>] device_add drivers/base/core.c:3364 [<000000009564d9ea>] usb_set_configuration drivers/usb/core/message.c:2171 [<00000000e4657087>] usb_generic_driver_probe drivers/usb/core/generic.c:239 [<0000000071ede518>] usb_probe_device drivers/usb/core/driver.c:294 [<00000000cbac9243>] really_probe drivers/base/dd.c:517 hci_alloc_dev() do not init the device's flag. And hci_free_dev() using put_device() to free the memory allocated for this device, but it calls just put_device(dev) only in case of HCI_UNREGISTER flag is set, So any error handing before hci_register_dev() success will cause memory leak. To avoid this behaviour we can using kfree() to release dev before hci_register_dev() success. Signed-off-by: Wei Yongjun Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_sysfs.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/bluetooth/hci_sysfs.c b/net/bluetooth/hci_sysfs.c index 7827639ecf5c..4e3e0451b08c 100644 --- a/net/bluetooth/hci_sysfs.c +++ b/net/bluetooth/hci_sysfs.c @@ -86,6 +86,8 @@ static void bt_host_release(struct device *dev) if (hci_dev_test_flag(hdev, HCI_UNREGISTER)) hci_release_dev(hdev); + else + kfree(hdev); module_put(THIS_MODULE); } From 60c6a63a3d3080a62f3e0e20084f58dbeff16748 Mon Sep 17 00:00:00 2001 From: "Mark-YW.Chen" Date: Thu, 14 Oct 2021 00:22:04 +0800 Subject: [PATCH 0012/2048] Bluetooth: btusb: fix memory leak in btusb_mtk_submit_wmt_recv_urb() Driver should free `usb->setup_packet` to avoid the leak. $ cat /sys/kernel/debug/kmemleak unreferenced object 0xffffffa564a58080 (size 128): backtrace: [<000000007eb8dd70>] kmem_cache_alloc_trace+0x22c/0x384 [<000000008a44191d>] btusb_mtk_hci_wmt_sync+0x1ec/0x994 [btusb] [<00000000ca7189a3>] btusb_mtk_setup+0x6b8/0x13cc [btusb] [<00000000c6105069>] hci_dev_do_open+0x290/0x974 [bluetooth] [<00000000a583f8b8>] hci_power_on+0xdc/0x3cc [bluetooth] [<000000005d80e687>] process_one_work+0x514/0xc80 [<00000000f4d57637>] worker_thread+0x818/0xd0c [<00000000dc7bdb55>] kthread+0x2f8/0x3b8 [<00000000f9999513>] ret_from_fork+0x10/0x30 Fixes: a1c49c434e150 ("Bluetooth: btusb: Add protocol support for MediaTek MT7668U USB devices") Signed-off-by: Mark-YW.Chen Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btusb.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index 3e26f60b457e..87b71740fad8 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2265,6 +2265,7 @@ static void btusb_mtk_wmt_recv(struct urb *urb) skb = bt_skb_alloc(HCI_WMT_MAX_EVENT_SIZE, GFP_ATOMIC); if (!skb) { hdev->stat.err_rx++; + kfree(urb->setup_packet); return; } @@ -2285,6 +2286,7 @@ static void btusb_mtk_wmt_recv(struct urb *urb) data->evt_skb = skb_clone(skb, GFP_ATOMIC); if (!data->evt_skb) { kfree_skb(skb); + kfree(urb->setup_packet); return; } } @@ -2293,6 +2295,7 @@ static void btusb_mtk_wmt_recv(struct urb *urb) if (err < 0) { kfree_skb(data->evt_skb); data->evt_skb = NULL; + kfree(urb->setup_packet); return; } @@ -2303,6 +2306,7 @@ static void btusb_mtk_wmt_recv(struct urb *urb) wake_up_bit(&data->flags, BTUSB_TX_WAIT_VND_EVT); } + kfree(urb->setup_packet); return; } else if (urb->status == -ENOENT) { /* Avoid suspend failed when usb_kill_urb */ @@ -2323,6 +2327,7 @@ static void btusb_mtk_wmt_recv(struct urb *urb) usb_anchor_urb(urb, &data->ctrl_anchor); err = usb_submit_urb(urb, GFP_ATOMIC); if (err < 0) { + kfree(urb->setup_packet); /* -EPERM: urb is being killed; * -ENODEV: device got disconnected */ From b8f5482c9638915c19963f06d2f4094e276aaab6 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 13 Oct 2021 15:17:01 -0700 Subject: [PATCH 0013/2048] Bluetooth: vhci: Add support for setting msft_opcode and aosp_capable This adds a debugfs entries to set msft_opcode and aosp_capable enabling vhci to emulate controllers with MSFT/AOSP extension support. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_vhci.c | 94 ++++++++++++++++++++++++++++++++++++ 1 file changed, 94 insertions(+) diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index 56c6b22be10b..9cb7c8fafbf9 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -42,6 +42,8 @@ struct vhci_data { bool suspended; bool wakeup; + __u16 msft_opcode; + bool aosp_capable; }; static int vhci_open_dev(struct hci_dev *hdev) @@ -194,6 +196,88 @@ static const struct file_operations force_wakeup_fops = { .llseek = default_llseek, }; +static int msft_opcode_set(void *data, u64 val) +{ + struct vhci_data *vhci = data; + + if (val > 0xffff || (val & 0xffff >> 10) != 0x3f) + return -EINVAL; + + if (vhci->msft_opcode) + return -EALREADY; + + vhci->msft_opcode = val; + + return 0; +} + +static int msft_opcode_get(void *data, u64 *val) +{ + struct vhci_data *vhci = data; + + *val = vhci->msft_opcode; + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(msft_opcode_fops, msft_opcode_get, msft_opcode_set, + "%llu\n"); + +static ssize_t aosp_capable_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + struct vhci_data *vhci = file->private_data; + char buf[3]; + + buf[0] = vhci->aosp_capable ? 'Y' : 'N'; + buf[1] = '\n'; + buf[2] = '\0'; + return simple_read_from_buffer(user_buf, count, ppos, buf, 2); +} + +static ssize_t aosp_capable_write(struct file *file, + const char __user *user_buf, size_t count, + loff_t *ppos) +{ + struct vhci_data *vhci = file->private_data; + bool enable; + int err; + + err = kstrtobool_from_user(user_buf, count, &enable); + if (err) + return err; + + if (!enable) + return -EINVAL; + + if (vhci->aosp_capable) + return -EALREADY; + + vhci->aosp_capable = enable; + + return count; +} + +static const struct file_operations aosp_capable_fops = { + .open = simple_open, + .read = aosp_capable_read, + .write = aosp_capable_write, + .llseek = default_llseek, +}; + +static int vhci_setup(struct hci_dev *hdev) +{ + struct vhci_data *vhci = hci_get_drvdata(hdev); + + if (vhci->msft_opcode) + hci_set_msft_opcode(hdev, vhci->msft_opcode); + + if (vhci->aosp_capable) + hci_set_aosp_capable(hdev); + + return 0; +} + static int __vhci_create_device(struct vhci_data *data, __u8 opcode) { struct hci_dev *hdev; @@ -236,6 +320,8 @@ static int __vhci_create_device(struct vhci_data *data, __u8 opcode) hdev->get_data_path_id = vhci_get_data_path_id; hdev->get_codec_config_data = vhci_get_codec_config_data; hdev->wakeup = vhci_wakeup; + hdev->setup = vhci_setup; + set_bit(HCI_QUIRK_NON_PERSISTENT_SETUP, &hdev->quirks); /* bit 6 is for external configuration */ if (opcode & 0x40) @@ -259,6 +345,14 @@ static int __vhci_create_device(struct vhci_data *data, __u8 opcode) debugfs_create_file("force_wakeup", 0644, hdev->debugfs, data, &force_wakeup_fops); + if (IS_ENABLED(CONFIG_BT_MSFTEXT)) + debugfs_create_file("msft_opcode", 0644, hdev->debugfs, data, + &msft_opcode_fops); + + if (IS_ENABLED(CONFIG_BT_AOSPEXT)) + debugfs_create_file("aosp_capable", 0644, hdev->debugfs, data, + &aosp_capable_fops); + hci_skb_pkt_type(skb) = HCI_VENDOR_PKT; skb_put_u8(skb, 0xff); From 1d0688421449718c6c5f46e458a378c9b530ba18 Mon Sep 17 00:00:00 2001 From: Soenke Huster Date: Wed, 20 Oct 2021 10:14:44 +0200 Subject: [PATCH 0014/2048] Bluetooth: virtio_bt: fix memory leak in virtbt_rx_handle() On the reception of packets with an invalid packet type, the memory of the allocated socket buffers is never freed. Add a default case that frees these to avoid a memory leak. Fixes: afd2daa26c7a ("Bluetooth: Add support for virtio transport driver") Signed-off-by: Soenke Huster Signed-off-by: Marcel Holtmann --- drivers/bluetooth/virtio_bt.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/bluetooth/virtio_bt.c b/drivers/bluetooth/virtio_bt.c index 57908ce4fae8..076e4942a3f0 100644 --- a/drivers/bluetooth/virtio_bt.c +++ b/drivers/bluetooth/virtio_bt.c @@ -202,6 +202,9 @@ static void virtbt_rx_handle(struct virtio_bluetooth *vbt, struct sk_buff *skb) hci_skb_pkt_type(skb) = pkt_type; hci_recv_frame(vbt->hdev, skb); break; + default: + kfree_skb(skb); + break; } } From 7a517ac9c00bab519770c0790b47845b8f5ecebb Mon Sep 17 00:00:00 2001 From: Kyle Copperfield Date: Wed, 20 Oct 2021 13:00:22 +0000 Subject: [PATCH 0015/2048] Bluetooth: btsdio: Do not bind to non-removable BCM4345 and BCM43455 BCM4345 and BCM43455 devices soldered onto the PCB (non-removable), use an UART connection for Bluetooth, such as in the Pinebook Pro, and the advertised btsdio support as an SDIO function is ignored. Reviewed-by: Dragan Simic Signed-off-by: Kyle Copperfield Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btsdio.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/bluetooth/btsdio.c b/drivers/bluetooth/btsdio.c index 199e8f7d426d..795be33f2892 100644 --- a/drivers/bluetooth/btsdio.c +++ b/drivers/bluetooth/btsdio.c @@ -295,6 +295,8 @@ static int btsdio_probe(struct sdio_func *func, switch (func->device) { case SDIO_DEVICE_ID_BROADCOM_43341: case SDIO_DEVICE_ID_BROADCOM_43430: + case SDIO_DEVICE_ID_BROADCOM_4345: + case SDIO_DEVICE_ID_BROADCOM_43455: case SDIO_DEVICE_ID_BROADCOM_4356: return -ENODEV; } From f33b0068cdaf2b9998fa3662585858ef30bc4b9e Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Thu, 21 Oct 2021 11:44:35 -0700 Subject: [PATCH 0016/2048] Bluetooth: vhci: Fix checking of msft_opcode msft_opcode shall be use a vendor ogf (0x3f) but the check was swifting the bits in the wrong order due to a missing parantesis over val & 0xffff, but since the code already checks for values over 0xffff it shall not be necessary to perform that operation it now just removes which makes it work properly when setting opcodes like 0xfce1. Fixes: b8f5482c9638 ("Bluetooth: vhci: Add support for setting msft_opcode and aosp_capable") Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_vhci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index 9cb7c8fafbf9..49ac884d996e 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -200,7 +200,7 @@ static int msft_opcode_set(void *data, u64 val) { struct vhci_data *vhci = data; - if (val > 0xffff || (val & 0xffff >> 10) != 0x3f) + if (val > 0xffff || hci_opcode_ogf(val) != 0x3f) return -EINVAL; if (vhci->msft_opcode) From 2a7ca7459d905febf519163bd9e3eed894de6bb7 Mon Sep 17 00:00:00 2001 From: Wang Hai Date: Mon, 25 Oct 2021 21:10:12 +0800 Subject: [PATCH 0017/2048] Bluetooth: cmtp: fix possible panic when cmtp_init_sockets() fails I got a kernel BUG report when doing fault injection test: ------------[ cut here ]------------ kernel BUG at lib/list_debug.c:45! ... RIP: 0010:__list_del_entry_valid.cold+0x12/0x4d ... Call Trace: proto_unregister+0x83/0x220 cmtp_cleanup_sockets+0x37/0x40 [cmtp] cmtp_exit+0xe/0x1f [cmtp] do_syscall_64+0x35/0xb0 entry_SYSCALL_64_after_hwframe+0x44/0xae If cmtp_init_sockets() in cmtp_init() fails, cmtp_init() still returns success. This will cause a kernel bug when accessing uncreated ctmp related data when the module exits. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Hulk Robot Signed-off-by: Wang Hai Signed-off-by: Marcel Holtmann --- net/bluetooth/cmtp/core.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/net/bluetooth/cmtp/core.c b/net/bluetooth/cmtp/core.c index 0a2d78e811cf..83eb84e8e688 100644 --- a/net/bluetooth/cmtp/core.c +++ b/net/bluetooth/cmtp/core.c @@ -501,9 +501,7 @@ static int __init cmtp_init(void) { BT_INFO("CMTP (CAPI Emulation) ver %s", VERSION); - cmtp_init_sockets(); - - return 0; + return cmtp_init_sockets(); } static void __exit cmtp_exit(void) From b5e6fa7a12572c82f1e7f2f51fbb02a322291291 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Mon, 25 Oct 2021 13:39:44 +0200 Subject: [PATCH 0018/2048] Bluetooth: bfusb: fix division by zero in send path Add the missing bulk-out endpoint sanity check to probe() to avoid division by zero in bfusb_send_frame() in case a malicious device has broken descriptors (or when doing descriptor fuzz testing). Note that USB core will reject URBs submitted for endpoints with zero wMaxPacketSize but that drivers doing packet-size calculations still need to handle this (cf. commit 2548288b4fb0 ("USB: Fix: Don't skip endpoint descriptors with maxpacket=0")). Cc: stable@vger.kernel.org Signed-off-by: Johan Hovold Signed-off-by: Marcel Holtmann --- drivers/bluetooth/bfusb.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/bluetooth/bfusb.c b/drivers/bluetooth/bfusb.c index 5a321b4076aa..cab93935cc7f 100644 --- a/drivers/bluetooth/bfusb.c +++ b/drivers/bluetooth/bfusb.c @@ -628,6 +628,9 @@ static int bfusb_probe(struct usb_interface *intf, const struct usb_device_id *i data->bulk_out_ep = bulk_out_ep->desc.bEndpointAddress; data->bulk_pkt_size = le16_to_cpu(bulk_out_ep->desc.wMaxPacketSize); + if (!data->bulk_pkt_size) + goto done; + rwlock_init(&data->lock); data->reassembly = NULL; From 2d6d4089ea894cf8eb481b10769160a35c181360 Mon Sep 17 00:00:00 2001 From: Paul Cercueil Date: Sat, 23 Oct 2021 11:10:27 +0100 Subject: [PATCH 0019/2048] Bluetooth: hci_bcm: Remove duplicated entry in OF table The entry "brcm,bcm4330-bt" was listed twice in the table. Signed-off-by: Paul Cercueil Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_bcm.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/bluetooth/hci_bcm.c b/drivers/bluetooth/hci_bcm.c index ef54afa29357..7852abf15ddf 100644 --- a/drivers/bluetooth/hci_bcm.c +++ b/drivers/bluetooth/hci_bcm.c @@ -1508,7 +1508,6 @@ static const struct of_device_id bcm_bluetooth_of_match[] = { { .compatible = "brcm,bcm4330-bt" }, { .compatible = "brcm,bcm4334-bt" }, { .compatible = "brcm,bcm4345c5" }, - { .compatible = "brcm,bcm4330-bt" }, { .compatible = "brcm,bcm43438-bt", .data = &bcm43438_device_data }, { .compatible = "brcm,bcm43540-bt", .data = &bcm4354_device_data }, { .compatible = "brcm,bcm4335a0" }, From 8c0d17b6b06c5bef45de7e08c5c3cab8367f6cbc Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:12 +0800 Subject: [PATCH 0020/2048] Bluetooth: mediatek: add BT_MTK module Add BT_MTK module that is a preliminary patch to introduce mt7921s support to share the logic betweem btusb and btmtksdio. Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/Kconfig | 5 + drivers/bluetooth/Makefile | 1 + drivers/bluetooth/btmtk.c | 289 ++++++++++++++++++++++++++++++++ drivers/bluetooth/btmtk.h | 97 +++++++++++ drivers/bluetooth/btusb.c | 331 +------------------------------------ 5 files changed, 398 insertions(+), 325 deletions(-) create mode 100644 drivers/bluetooth/btmtk.c create mode 100644 drivers/bluetooth/btmtk.h diff --git a/drivers/bluetooth/Kconfig b/drivers/bluetooth/Kconfig index 851842372c9b..24c3aea4a7d6 100644 --- a/drivers/bluetooth/Kconfig +++ b/drivers/bluetooth/Kconfig @@ -19,6 +19,10 @@ config BT_QCA tristate select FW_LOADER +config BT_MTK + tristate + select FW_LOADER + config BT_HCIBTUSB tristate "HCI USB driver" depends on USB @@ -55,6 +59,7 @@ config BT_HCIBTUSB_BCM config BT_HCIBTUSB_MTK bool "MediaTek protocol support" depends on BT_HCIBTUSB + select BT_MTK default n help The MediaTek protocol support enables firmware download diff --git a/drivers/bluetooth/Makefile b/drivers/bluetooth/Makefile index 16286ea2655d..3321a8aea4a0 100644 --- a/drivers/bluetooth/Makefile +++ b/drivers/bluetooth/Makefile @@ -25,6 +25,7 @@ obj-$(CONFIG_BT_QCOMSMD) += btqcomsmd.o obj-$(CONFIG_BT_BCM) += btbcm.o obj-$(CONFIG_BT_RTL) += btrtl.o obj-$(CONFIG_BT_QCA) += btqca.o +obj-$(CONFIG_BT_MTK) += btmtk.o obj-$(CONFIG_BT_VIRTIO) += virtio_bt.o diff --git a/drivers/bluetooth/btmtk.c b/drivers/bluetooth/btmtk.c new file mode 100644 index 000000000000..c2ee5c4b975a --- /dev/null +++ b/drivers/bluetooth/btmtk.c @@ -0,0 +1,289 @@ +// SPDX-License-Identifier: ISC +/* Copyright (C) 2021 MediaTek Inc. + * + */ +#include +#include + +#include +#include + +#include "btmtk.h" + +#define VERSION "0.1" + +/* It is for mt79xx download rom patch*/ +#define MTK_FW_ROM_PATCH_HEADER_SIZE 32 +#define MTK_FW_ROM_PATCH_GD_SIZE 64 +#define MTK_FW_ROM_PATCH_SEC_MAP_SIZE 64 +#define MTK_SEC_MAP_COMMON_SIZE 12 +#define MTK_SEC_MAP_NEED_SEND_SIZE 52 + +struct btmtk_patch_header { + u8 datetime[16]; + u8 platform[4]; + __le16 hwver; + __le16 swver; + __le32 magicnum; +} __packed; + +struct btmtk_global_desc { + __le32 patch_ver; + __le32 sub_sys; + __le32 feature_opt; + __le32 section_num; +} __packed; + +struct btmtk_section_map { + __le32 sectype; + __le32 secoffset; + __le32 secsize; + union { + __le32 u4SecSpec[13]; + struct { + __le32 dlAddr; + __le32 dlsize; + __le32 seckeyidx; + __le32 alignlen; + __le32 sectype; + __le32 dlmodecrctype; + __le32 crc; + __le32 reserved[6]; + } bin_info_spec; + }; +} __packed; + +int btmtk_setup_firmware_79xx(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync) +{ + struct btmtk_hci_wmt_params wmt_params; + struct btmtk_global_desc *globaldesc = NULL; + struct btmtk_section_map *sectionmap; + const struct firmware *fw; + const u8 *fw_ptr; + const u8 *fw_bin_ptr; + int err, dlen, i, status; + u8 flag, first_block, retry; + u32 section_num, dl_size, section_offset; + u8 cmd[64]; + + err = request_firmware(&fw, fwname, &hdev->dev); + if (err < 0) { + bt_dev_err(hdev, "Failed to load firmware file (%d)", err); + return err; + } + + fw_ptr = fw->data; + fw_bin_ptr = fw_ptr; + globaldesc = (struct btmtk_global_desc *)(fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE); + section_num = le32_to_cpu(globaldesc->section_num); + + for (i = 0; i < section_num; i++) { + first_block = 1; + fw_ptr = fw_bin_ptr; + sectionmap = (struct btmtk_section_map *)(fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE + + MTK_FW_ROM_PATCH_GD_SIZE + MTK_FW_ROM_PATCH_SEC_MAP_SIZE * i); + + section_offset = le32_to_cpu(sectionmap->secoffset); + dl_size = le32_to_cpu(sectionmap->bin_info_spec.dlsize); + + if (dl_size > 0) { + retry = 20; + while (retry > 0) { + cmd[0] = 0; /* 0 means legacy dl mode. */ + memcpy(cmd + 1, + fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE + + MTK_FW_ROM_PATCH_GD_SIZE + + MTK_FW_ROM_PATCH_SEC_MAP_SIZE * i + + MTK_SEC_MAP_COMMON_SIZE, + MTK_SEC_MAP_NEED_SEND_SIZE + 1); + + wmt_params.op = BTMTK_WMT_PATCH_DWNLD; + wmt_params.status = &status; + wmt_params.flag = 0; + wmt_params.dlen = MTK_SEC_MAP_NEED_SEND_SIZE + 1; + wmt_params.data = &cmd; + + err = wmt_cmd_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", + err); + goto err_release_fw; + } + + if (status == BTMTK_WMT_PATCH_UNDONE) { + break; + } else if (status == BTMTK_WMT_PATCH_PROGRESS) { + msleep(100); + retry--; + } else if (status == BTMTK_WMT_PATCH_DONE) { + goto next_section; + } else { + bt_dev_err(hdev, "Failed wmt patch dwnld status (%d)", + status); + goto err_release_fw; + } + } + + fw_ptr += section_offset; + wmt_params.op = BTMTK_WMT_PATCH_DWNLD; + wmt_params.status = NULL; + + while (dl_size > 0) { + dlen = min_t(int, 250, dl_size); + if (first_block == 1) { + flag = 1; + first_block = 0; + } else if (dl_size - dlen <= 0) { + flag = 3; + } else { + flag = 2; + } + + wmt_params.flag = flag; + wmt_params.dlen = dlen; + wmt_params.data = fw_ptr; + + err = wmt_cmd_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", + err); + goto err_release_fw; + } + + dl_size -= dlen; + fw_ptr += dlen; + } + } +next_section: + continue; + } + /* Wait a few moments for firmware activation done */ + usleep_range(100000, 120000); + +err_release_fw: + release_firmware(fw); + + return err; +} +EXPORT_SYMBOL_GPL(btmtk_setup_firmware_79xx); + +int btmtk_setup_firmware(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync) +{ + struct btmtk_hci_wmt_params wmt_params; + const struct firmware *fw; + const u8 *fw_ptr; + size_t fw_size; + int err, dlen; + u8 flag, param; + + err = request_firmware(&fw, fwname, &hdev->dev); + if (err < 0) { + bt_dev_err(hdev, "Failed to load firmware file (%d)", err); + return err; + } + + /* Power on data RAM the firmware relies on. */ + param = 1; + wmt_params.op = BTMTK_WMT_FUNC_CTRL; + wmt_params.flag = 3; + wmt_params.dlen = sizeof(param); + wmt_params.data = ¶m; + wmt_params.status = NULL; + + err = wmt_cmd_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to power on data RAM (%d)", err); + goto err_release_fw; + } + + fw_ptr = fw->data; + fw_size = fw->size; + + /* The size of patch header is 30 bytes, should be skip */ + if (fw_size < 30) { + err = -EINVAL; + goto err_release_fw; + } + + fw_size -= 30; + fw_ptr += 30; + flag = 1; + + wmt_params.op = BTMTK_WMT_PATCH_DWNLD; + wmt_params.status = NULL; + + while (fw_size > 0) { + dlen = min_t(int, 250, fw_size); + + /* Tell device the position in sequence */ + if (fw_size - dlen <= 0) + flag = 3; + else if (fw_size < fw->size - 30) + flag = 2; + + wmt_params.flag = flag; + wmt_params.dlen = dlen; + wmt_params.data = fw_ptr; + + err = wmt_cmd_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", + err); + goto err_release_fw; + } + + fw_size -= dlen; + fw_ptr += dlen; + } + + wmt_params.op = BTMTK_WMT_RST; + wmt_params.flag = 4; + wmt_params.dlen = 0; + wmt_params.data = NULL; + wmt_params.status = NULL; + + /* Activate funciton the firmware providing to */ + err = wmt_cmd_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to send wmt rst (%d)", err); + goto err_release_fw; + } + + /* Wait a few moments for firmware activation done */ + usleep_range(10000, 12000); + +err_release_fw: + release_firmware(fw); + + return err; +} +EXPORT_SYMBOL_GPL(btmtk_setup_firmware); + +int btmtk_set_bdaddr(struct hci_dev *hdev, const bdaddr_t *bdaddr) +{ + struct sk_buff *skb; + long ret; + + skb = __hci_cmd_sync(hdev, 0xfc1a, 6, bdaddr, HCI_INIT_TIMEOUT); + if (IS_ERR(skb)) { + ret = PTR_ERR(skb); + bt_dev_err(hdev, "changing Mediatek device address failed (%ld)", + ret); + return ret; + } + kfree_skb(skb); + + return 0; +} +EXPORT_SYMBOL_GPL(btmtk_set_bdaddr); + +MODULE_AUTHOR("Sean Wang "); +MODULE_AUTHOR("Mark Chen "); +MODULE_DESCRIPTION("Bluetooth support for MediaTek devices ver " VERSION); +MODULE_VERSION(VERSION); +MODULE_LICENSE("GPL"); +MODULE_FIRMWARE(FIRMWARE_MT7663); +MODULE_FIRMWARE(FIRMWARE_MT7668); +MODULE_FIRMWARE(FIRMWARE_MT7961); diff --git a/drivers/bluetooth/btmtk.h b/drivers/bluetooth/btmtk.h new file mode 100644 index 000000000000..87c25c6cf969 --- /dev/null +++ b/drivers/bluetooth/btmtk.h @@ -0,0 +1,97 @@ +/* SPDX-License-Identifier: ISC */ +/* Copyright (C) 2021 MediaTek Inc. */ + +#define FIRMWARE_MT7663 "mediatek/mt7663pr2h.bin" +#define FIRMWARE_MT7668 "mediatek/mt7668pr2h.bin" +#define FIRMWARE_MT7961 "mediatek/BT_RAM_CODE_MT7961_1_2_hdr.bin" + +#define HCI_WMT_MAX_EVENT_SIZE 64 + +enum { + BTMTK_WMT_PATCH_DWNLD = 0x1, + BTMTK_WMT_FUNC_CTRL = 0x6, + BTMTK_WMT_RST = 0x7, + BTMTK_WMT_SEMAPHORE = 0x17, +}; + +enum { + BTMTK_WMT_INVALID, + BTMTK_WMT_PATCH_UNDONE, + BTMTK_WMT_PATCH_PROGRESS, + BTMTK_WMT_PATCH_DONE, + BTMTK_WMT_ON_UNDONE, + BTMTK_WMT_ON_DONE, + BTMTK_WMT_ON_PROGRESS, +}; + +struct btmtk_wmt_hdr { + u8 dir; + u8 op; + __le16 dlen; + u8 flag; +} __packed; + +struct btmtk_hci_wmt_cmd { + struct btmtk_wmt_hdr hdr; + u8 data[]; +} __packed; + +struct btmtk_hci_wmt_evt { + struct hci_event_hdr hhdr; + struct btmtk_wmt_hdr whdr; +} __packed; + +struct btmtk_hci_wmt_evt_funcc { + struct btmtk_hci_wmt_evt hwhdr; + __be16 status; +} __packed; + +struct btmtk_tci_sleep { + u8 mode; + __le16 duration; + __le16 host_duration; + u8 host_wakeup_pin; + u8 time_compensation; +} __packed; + +struct btmtk_hci_wmt_params { + u8 op; + u8 flag; + u16 dlen; + const void *data; + u32 *status; +}; + +typedef int (*wmt_cmd_sync_func_t)(struct hci_dev *, + struct btmtk_hci_wmt_params *); + +#if IS_ENABLED(CONFIG_BT_MTK) + +int btmtk_set_bdaddr(struct hci_dev *hdev, const bdaddr_t *bdaddr); + +int btmtk_setup_firmware_79xx(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync); + +int btmtk_setup_firmware(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync); +#else + +static inline int btmtk_set_bdaddr(struct hci_dev *hdev, + const bdaddr_t *bdaddr) +{ + return -EOPNOTSUPP; +} + +static int btmtk_setup_firmware_79xx(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync) +{ + return -EOPNOTSUPP; +} + +static int btmtk_setup_firmware(struct hci_dev *hdev, const char *fwname, + wmt_cmd_sync_func_t wmt_cmd_sync) +{ + return -EOPNOTSUPP; +} + +#endif diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index 87b71740fad8..88bfdf1347a7 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -24,6 +24,7 @@ #include "btintel.h" #include "btbcm.h" #include "btrtl.h" +#include "btmtk.h" #define VERSION "0.8" @@ -2131,122 +2132,6 @@ static int btusb_send_frame_intel(struct hci_dev *hdev, struct sk_buff *skb) #define MTK_BT_RST_DONE 0x00000100 #define MTK_BT_RESET_WAIT_MS 100 #define MTK_BT_RESET_NUM_TRIES 10 -#define FIRMWARE_MT7663 "mediatek/mt7663pr2h.bin" -#define FIRMWARE_MT7668 "mediatek/mt7668pr2h.bin" - -#define HCI_WMT_MAX_EVENT_SIZE 64 -/* It is for mt79xx download rom patch*/ -#define MTK_FW_ROM_PATCH_HEADER_SIZE 32 -#define MTK_FW_ROM_PATCH_GD_SIZE 64 -#define MTK_FW_ROM_PATCH_SEC_MAP_SIZE 64 -#define MTK_SEC_MAP_COMMON_SIZE 12 -#define MTK_SEC_MAP_NEED_SEND_SIZE 52 - -enum { - BTMTK_WMT_PATCH_DWNLD = 0x1, - BTMTK_WMT_FUNC_CTRL = 0x6, - BTMTK_WMT_RST = 0x7, - BTMTK_WMT_SEMAPHORE = 0x17, -}; - -enum { - BTMTK_WMT_INVALID, - BTMTK_WMT_PATCH_UNDONE, - BTMTK_WMT_PATCH_PROGRESS, - BTMTK_WMT_PATCH_DONE, - BTMTK_WMT_ON_UNDONE, - BTMTK_WMT_ON_DONE, - BTMTK_WMT_ON_PROGRESS, -}; - -struct btmtk_wmt_hdr { - u8 dir; - u8 op; - __le16 dlen; - u8 flag; -} __packed; - -struct btmtk_hci_wmt_cmd { - struct btmtk_wmt_hdr hdr; - u8 data[]; -} __packed; - -struct btmtk_hci_wmt_evt { - struct hci_event_hdr hhdr; - struct btmtk_wmt_hdr whdr; -} __packed; - -struct btmtk_hci_wmt_evt_funcc { - struct btmtk_hci_wmt_evt hwhdr; - __be16 status; -} __packed; - -struct btmtk_tci_sleep { - u8 mode; - __le16 duration; - __le16 host_duration; - u8 host_wakeup_pin; - u8 time_compensation; -} __packed; - -struct btmtk_hci_wmt_params { - u8 op; - u8 flag; - u16 dlen; - const void *data; - u32 *status; -}; - -struct btmtk_patch_header { - u8 datetime[16]; - u8 platform[4]; - __le16 hwver; - __le16 swver; - __le32 magicnum; -} __packed; - -struct btmtk_global_desc { - __le32 patch_ver; - __le32 sub_sys; - __le32 feature_opt; - __le32 section_num; -} __packed; - -struct btmtk_section_map { - __le32 sectype; - __le32 secoffset; - __le32 secsize; - union { - __le32 u4SecSpec[13]; - struct { - __le32 dlAddr; - __le32 dlsize; - __le32 seckeyidx; - __le32 alignlen; - __le32 sectype; - __le32 dlmodecrctype; - __le32 crc; - __le32 reserved[6]; - } bin_info_spec; - }; -} __packed; - -static int btusb_set_bdaddr_mtk(struct hci_dev *hdev, const bdaddr_t *bdaddr) -{ - struct sk_buff *skb; - long ret; - - skb = __hci_cmd_sync(hdev, 0xfc1a, 6, bdaddr, HCI_INIT_TIMEOUT); - if (IS_ERR(skb)) { - ret = PTR_ERR(skb); - bt_dev_err(hdev, "changing Mediatek device address failed (%ld)", - ret); - return ret; - } - kfree_skb(skb); - - return 0; -} static void btusb_mtk_wmt_recv(struct urb *urb) { @@ -2502,209 +2387,6 @@ err_free_wc: return err; } -static int btusb_mtk_setup_firmware_79xx(struct hci_dev *hdev, const char *fwname) -{ - struct btmtk_hci_wmt_params wmt_params; - struct btmtk_global_desc *globaldesc = NULL; - struct btmtk_section_map *sectionmap; - const struct firmware *fw; - const u8 *fw_ptr; - const u8 *fw_bin_ptr; - int err, dlen, i, status; - u8 flag, first_block, retry; - u32 section_num, dl_size, section_offset; - u8 cmd[64]; - - err = request_firmware(&fw, fwname, &hdev->dev); - if (err < 0) { - bt_dev_err(hdev, "Failed to load firmware file (%d)", err); - return err; - } - - fw_ptr = fw->data; - fw_bin_ptr = fw_ptr; - globaldesc = (struct btmtk_global_desc *)(fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE); - section_num = le32_to_cpu(globaldesc->section_num); - - for (i = 0; i < section_num; i++) { - first_block = 1; - fw_ptr = fw_bin_ptr; - sectionmap = (struct btmtk_section_map *)(fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE + - MTK_FW_ROM_PATCH_GD_SIZE + MTK_FW_ROM_PATCH_SEC_MAP_SIZE * i); - - section_offset = le32_to_cpu(sectionmap->secoffset); - dl_size = le32_to_cpu(sectionmap->bin_info_spec.dlsize); - - if (dl_size > 0) { - retry = 20; - while (retry > 0) { - cmd[0] = 0; /* 0 means legacy dl mode. */ - memcpy(cmd + 1, - fw_ptr + MTK_FW_ROM_PATCH_HEADER_SIZE + - MTK_FW_ROM_PATCH_GD_SIZE + MTK_FW_ROM_PATCH_SEC_MAP_SIZE * i + - MTK_SEC_MAP_COMMON_SIZE, - MTK_SEC_MAP_NEED_SEND_SIZE + 1); - - wmt_params.op = BTMTK_WMT_PATCH_DWNLD; - wmt_params.status = &status; - wmt_params.flag = 0; - wmt_params.dlen = MTK_SEC_MAP_NEED_SEND_SIZE + 1; - wmt_params.data = &cmd; - - err = btusb_mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", - err); - goto err_release_fw; - } - - if (status == BTMTK_WMT_PATCH_UNDONE) { - break; - } else if (status == BTMTK_WMT_PATCH_PROGRESS) { - msleep(100); - retry--; - } else if (status == BTMTK_WMT_PATCH_DONE) { - goto next_section; - } else { - bt_dev_err(hdev, "Failed wmt patch dwnld status (%d)", - status); - goto err_release_fw; - } - } - - fw_ptr += section_offset; - wmt_params.op = BTMTK_WMT_PATCH_DWNLD; - wmt_params.status = NULL; - - while (dl_size > 0) { - dlen = min_t(int, 250, dl_size); - if (first_block == 1) { - flag = 1; - first_block = 0; - } else if (dl_size - dlen <= 0) { - flag = 3; - } else { - flag = 2; - } - - wmt_params.flag = flag; - wmt_params.dlen = dlen; - wmt_params.data = fw_ptr; - - err = btusb_mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", - err); - goto err_release_fw; - } - - dl_size -= dlen; - fw_ptr += dlen; - } - } -next_section: - continue; - } - /* Wait a few moments for firmware activation done */ - usleep_range(100000, 120000); - -err_release_fw: - release_firmware(fw); - - return err; -} - -static int btusb_mtk_setup_firmware(struct hci_dev *hdev, const char *fwname) -{ - struct btmtk_hci_wmt_params wmt_params; - const struct firmware *fw; - const u8 *fw_ptr; - size_t fw_size; - int err, dlen; - u8 flag, param; - - err = request_firmware(&fw, fwname, &hdev->dev); - if (err < 0) { - bt_dev_err(hdev, "Failed to load firmware file (%d)", err); - return err; - } - - /* Power on data RAM the firmware relies on. */ - param = 1; - wmt_params.op = BTMTK_WMT_FUNC_CTRL; - wmt_params.flag = 3; - wmt_params.dlen = sizeof(param); - wmt_params.data = ¶m; - wmt_params.status = NULL; - - err = btusb_mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to power on data RAM (%d)", err); - goto err_release_fw; - } - - fw_ptr = fw->data; - fw_size = fw->size; - - /* The size of patch header is 30 bytes, should be skip */ - if (fw_size < 30) { - err = -EINVAL; - goto err_release_fw; - } - - fw_size -= 30; - fw_ptr += 30; - flag = 1; - - wmt_params.op = BTMTK_WMT_PATCH_DWNLD; - wmt_params.status = NULL; - - while (fw_size > 0) { - dlen = min_t(int, 250, fw_size); - - /* Tell device the position in sequence */ - if (fw_size - dlen <= 0) - flag = 3; - else if (fw_size < fw->size - 30) - flag = 2; - - wmt_params.flag = flag; - wmt_params.dlen = dlen; - wmt_params.data = fw_ptr; - - err = btusb_mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", - err); - goto err_release_fw; - } - - fw_size -= dlen; - fw_ptr += dlen; - } - - wmt_params.op = BTMTK_WMT_RST; - wmt_params.flag = 4; - wmt_params.dlen = 0; - wmt_params.data = NULL; - wmt_params.status = NULL; - - /* Activate funciton the firmware providing to */ - err = btusb_mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt rst (%d)", err); - goto err_release_fw; - } - - /* Wait a few moments for firmware activation done */ - usleep_range(10000, 12000); - -err_release_fw: - release_firmware(fw); - - return err; -} - static int btusb_mtk_func_query(struct hci_dev *hdev) { struct btmtk_hci_wmt_params wmt_params; @@ -2862,7 +2544,8 @@ static int btusb_mtk_setup(struct hci_dev *hdev) snprintf(fw_bin_name, sizeof(fw_bin_name), "mediatek/BT_RAM_CODE_MT%04x_1_%x_hdr.bin", dev_id & 0xffff, (fw_version & 0xff) + 1); - err = btusb_mtk_setup_firmware_79xx(hdev, fw_bin_name); + err = btmtk_setup_firmware_79xx(hdev, fw_bin_name, + btusb_mtk_hci_wmt_sync); /* It's Device EndPoint Reset Option Register */ btusb_mtk_uhw_reg_write(data, MTK_EP_RST_OPT, MTK_EP_RST_IN_OUT_OPT); @@ -2908,7 +2591,8 @@ static int btusb_mtk_setup(struct hci_dev *hdev) } /* Setup a firmware which the device definitely requires */ - err = btusb_mtk_setup_firmware(hdev, fwname); + err = btmtk_setup_firmware(hdev, fwname, + btusb_mtk_hci_wmt_sync); if (err < 0) return err; @@ -3069,9 +2753,6 @@ static int btusb_recv_acl_mtk(struct hci_dev *hdev, struct sk_buff *skb) return hci_recv_frame(hdev, skb); } -MODULE_FIRMWARE(FIRMWARE_MT7663); -MODULE_FIRMWARE(FIRMWARE_MT7668); - #ifdef CONFIG_PM /* Configure an out-of-band gpio as wake-up pin, if specified in device tree */ static int marvell_config_oob_wake(struct hci_dev *hdev) @@ -3873,7 +3554,7 @@ static int btusb_probe(struct usb_interface *intf, hdev->shutdown = btusb_mtk_shutdown; hdev->manufacturer = 70; hdev->cmd_timeout = btusb_mtk_cmd_timeout; - hdev->set_bdaddr = btusb_set_bdaddr_mtk; + hdev->set_bdaddr = btmtk_set_bdaddr; set_bit(HCI_QUIRK_NON_PERSISTENT_SETUP, &hdev->quirks); data->recv_acl = btusb_recv_acl_mtk; } From 3a722044aacf7f8524226951515dbc0390e030e0 Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:13 +0800 Subject: [PATCH 0021/2048] Bluetooth: btmtksido: rely on BT_MTK module Rely on btmtk module to reduce duplicated code Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/Kconfig | 1 + drivers/bluetooth/btmtk.h | 3 + drivers/bluetooth/btmtksdio.c | 192 +++++----------------------------- 3 files changed, 28 insertions(+), 168 deletions(-) diff --git a/drivers/bluetooth/Kconfig b/drivers/bluetooth/Kconfig index 24c3aea4a7d6..36380e618ba4 100644 --- a/drivers/bluetooth/Kconfig +++ b/drivers/bluetooth/Kconfig @@ -388,6 +388,7 @@ config BT_ATH3K config BT_MTKSDIO tristate "MediaTek HCI SDIO driver" depends on MMC + select BT_MTK help MediaTek Bluetooth HCI SDIO driver. This driver is required if you want to use MediaTek Bluetooth diff --git a/drivers/bluetooth/btmtk.h b/drivers/bluetooth/btmtk.h index 87c25c6cf969..18f43c641b3b 100644 --- a/drivers/bluetooth/btmtk.h +++ b/drivers/bluetooth/btmtk.h @@ -9,6 +9,9 @@ enum { BTMTK_WMT_PATCH_DWNLD = 0x1, + BTMTK_WMT_TEST = 0x2, + BTMTK_WMT_WAKEUP = 0x3, + BTMTK_WMT_HIF = 0x4, BTMTK_WMT_FUNC_CTRL = 0x6, BTMTK_WMT_RST = 0x7, BTMTK_WMT_SEMAPHORE = 0x17, diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 9872ef18f9fe..53682a4c2ced 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -12,7 +12,6 @@ #include #include -#include #include #include #include @@ -28,12 +27,10 @@ #include #include "h4_recv.h" +#include "btmtk.h" #define VERSION "0.1" -#define FIRMWARE_MT7663 "mediatek/mt7663pr2h.bin" -#define FIRMWARE_MT7668 "mediatek/mt7668pr2h.bin" - #define MTKBTSDIO_AUTOSUSPEND_DELAY 8000 static bool enable_autosuspend; @@ -90,69 +87,12 @@ MODULE_DEVICE_TABLE(sdio, btmtksdio_table); #define BTMTKSDIO_TX_WAIT_VND_EVT 1 -enum { - MTK_WMT_PATCH_DWNLD = 0x1, - MTK_WMT_TEST = 0x2, - MTK_WMT_WAKEUP = 0x3, - MTK_WMT_HIF = 0x4, - MTK_WMT_FUNC_CTRL = 0x6, - MTK_WMT_RST = 0x7, - MTK_WMT_SEMAPHORE = 0x17, -}; - -enum { - BTMTK_WMT_INVALID, - BTMTK_WMT_PATCH_UNDONE, - BTMTK_WMT_PATCH_DONE, - BTMTK_WMT_ON_UNDONE, - BTMTK_WMT_ON_DONE, - BTMTK_WMT_ON_PROGRESS, -}; - struct mtkbtsdio_hdr { __le16 len; __le16 reserved; u8 bt_type; } __packed; -struct mtk_wmt_hdr { - u8 dir; - u8 op; - __le16 dlen; - u8 flag; -} __packed; - -struct mtk_hci_wmt_cmd { - struct mtk_wmt_hdr hdr; - u8 data[256]; -} __packed; - -struct btmtk_hci_wmt_evt { - struct hci_event_hdr hhdr; - struct mtk_wmt_hdr whdr; -} __packed; - -struct btmtk_hci_wmt_evt_funcc { - struct btmtk_hci_wmt_evt hwhdr; - __be16 status; -} __packed; - -struct btmtk_tci_sleep { - u8 mode; - __le16 duration; - __le16 host_duration; - u8 host_wakeup_pin; - u8 time_compensation; -} __packed; - -struct btmtk_hci_wmt_params { - u8 op; - u8 flag; - u16 dlen; - const void *data; - u32 *status; -}; - struct btmtksdio_dev { struct hci_dev *hdev; struct sdio_func *func; @@ -174,27 +114,32 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, struct btmtk_hci_wmt_evt_funcc *wmt_evt_funcc; u32 hlen, status = BTMTK_WMT_INVALID; struct btmtk_hci_wmt_evt *wmt_evt; - struct mtk_hci_wmt_cmd wc; - struct mtk_wmt_hdr *hdr; + struct btmtk_hci_wmt_cmd *wc; + struct btmtk_wmt_hdr *hdr; int err; + /* Send the WMT command and wait until the WMT event returns */ hlen = sizeof(*hdr) + wmt_params->dlen; if (hlen > 255) return -EINVAL; - hdr = (struct mtk_wmt_hdr *)&wc; + wc = kzalloc(hlen, GFP_KERNEL); + if (!wc) + return -ENOMEM; + + hdr = &wc->hdr; hdr->dir = 1; hdr->op = wmt_params->op; hdr->dlen = cpu_to_le16(wmt_params->dlen + 1); hdr->flag = wmt_params->flag; - memcpy(wc.data, wmt_params->data, wmt_params->dlen); + memcpy(wc->data, wmt_params->data, wmt_params->dlen); set_bit(BTMTKSDIO_TX_WAIT_VND_EVT, &bdev->tx_state); - err = __hci_cmd_send(hdev, 0xfc6f, hlen, &wc); + err = __hci_cmd_send(hdev, 0xfc6f, hlen, wc); if (err < 0) { clear_bit(BTMTKSDIO_TX_WAIT_VND_EVT, &bdev->tx_state); - return err; + goto err_free_wc; } /* The vendor specific WMT commands are all answered by a vendor @@ -211,13 +156,14 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, if (err == -EINTR) { bt_dev_err(hdev, "Execution of wmt command interrupted"); clear_bit(BTMTKSDIO_TX_WAIT_VND_EVT, &bdev->tx_state); - return err; + goto err_free_wc; } if (err) { bt_dev_err(hdev, "Execution of wmt command timed out"); clear_bit(BTMTKSDIO_TX_WAIT_VND_EVT, &bdev->tx_state); - return -ETIMEDOUT; + err = -ETIMEDOUT; + goto err_free_wc; } /* Parse and handle the return WMT event */ @@ -230,13 +176,13 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, } switch (wmt_evt->whdr.op) { - case MTK_WMT_SEMAPHORE: + case BTMTK_WMT_SEMAPHORE: if (wmt_evt->whdr.flag == 2) status = BTMTK_WMT_PATCH_UNDONE; else status = BTMTK_WMT_PATCH_DONE; break; - case MTK_WMT_FUNC_CTRL: + case BTMTK_WMT_FUNC_CTRL: wmt_evt_funcc = (struct btmtk_hci_wmt_evt_funcc *)wmt_evt; if (be16_to_cpu(wmt_evt_funcc->status) == 0x404) status = BTMTK_WMT_ON_DONE; @@ -253,6 +199,8 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, err_free_skb: kfree_skb(bdev->evt_skb); bdev->evt_skb = NULL; +err_free_wc: + kfree(wc); return err; } @@ -663,7 +611,7 @@ static int btmtksdio_func_query(struct hci_dev *hdev) u8 param = 0; /* Query whether the function is enabled */ - wmt_params.op = MTK_WMT_FUNC_CTRL; + wmt_params.op = BTMTK_WMT_FUNC_CTRL; wmt_params.flag = 4; wmt_params.dlen = sizeof(param); wmt_params.data = ¶m; @@ -678,96 +626,6 @@ static int btmtksdio_func_query(struct hci_dev *hdev) return status; } -static int mtk_setup_firmware(struct hci_dev *hdev, const char *fwname) -{ - struct btmtk_hci_wmt_params wmt_params; - const struct firmware *fw; - const u8 *fw_ptr; - size_t fw_size; - int err, dlen; - u8 flag, param; - - err = request_firmware(&fw, fwname, &hdev->dev); - if (err < 0) { - bt_dev_err(hdev, "Failed to load firmware file (%d)", err); - return err; - } - - /* Power on data RAM the firmware relies on. */ - param = 1; - wmt_params.op = MTK_WMT_FUNC_CTRL; - wmt_params.flag = 3; - wmt_params.dlen = sizeof(param); - wmt_params.data = ¶m; - wmt_params.status = NULL; - - err = mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to power on data RAM (%d)", err); - goto free_fw; - } - - fw_ptr = fw->data; - fw_size = fw->size; - - /* The size of patch header is 30 bytes, should be skip */ - if (fw_size < 30) { - err = -EINVAL; - goto free_fw; - } - - fw_size -= 30; - fw_ptr += 30; - flag = 1; - - wmt_params.op = MTK_WMT_PATCH_DWNLD; - wmt_params.status = NULL; - - while (fw_size > 0) { - dlen = min_t(int, 250, fw_size); - - /* Tell device the position in sequence */ - if (fw_size - dlen <= 0) - flag = 3; - else if (fw_size < fw->size - 30) - flag = 2; - - wmt_params.flag = flag; - wmt_params.dlen = dlen; - wmt_params.data = fw_ptr; - - err = mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt patch dwnld (%d)", - err); - goto free_fw; - } - - fw_size -= dlen; - fw_ptr += dlen; - } - - wmt_params.op = MTK_WMT_RST; - wmt_params.flag = 4; - wmt_params.dlen = 0; - wmt_params.data = NULL; - wmt_params.status = NULL; - - /* Activate funciton the firmware providing to */ - err = mtk_hci_wmt_sync(hdev, &wmt_params); - if (err < 0) { - bt_dev_err(hdev, "Failed to send wmt rst (%d)", err); - goto free_fw; - } - - /* Wait a few moments for firmware activation done */ - usleep_range(10000, 12000); - -free_fw: - release_firmware(fw); - return err; -} - static int btmtksdio_setup(struct hci_dev *hdev) { struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); @@ -782,7 +640,7 @@ static int btmtksdio_setup(struct hci_dev *hdev) calltime = ktime_get(); /* Query whether the firmware is already download */ - wmt_params.op = MTK_WMT_SEMAPHORE; + wmt_params.op = BTMTK_WMT_SEMAPHORE; wmt_params.flag = 1; wmt_params.dlen = 0; wmt_params.data = NULL; @@ -800,7 +658,7 @@ static int btmtksdio_setup(struct hci_dev *hdev) } /* Setup a firmware which the device definitely requires */ - err = mtk_setup_firmware(hdev, bdev->data->fwname); + err = btmtk_setup_firmware(hdev, bdev->data->fwname, mtk_hci_wmt_sync); if (err < 0) return err; @@ -823,7 +681,7 @@ ignore_setup_fw: } /* Enable Bluetooth protocol */ - wmt_params.op = MTK_WMT_FUNC_CTRL; + wmt_params.op = BTMTK_WMT_FUNC_CTRL; wmt_params.flag = 0; wmt_params.dlen = sizeof(param); wmt_params.data = ¶m; @@ -891,7 +749,7 @@ static int btmtksdio_shutdown(struct hci_dev *hdev) pm_runtime_get_sync(bdev->dev); /* Disable the device */ - wmt_params.op = MTK_WMT_FUNC_CTRL; + wmt_params.op = BTMTK_WMT_FUNC_CTRL; wmt_params.flag = 0; wmt_params.dlen = sizeof(param); wmt_params.data = ¶m; @@ -1112,5 +970,3 @@ MODULE_AUTHOR("Sean Wang "); MODULE_DESCRIPTION("MediaTek Bluetooth SDIO driver ver " VERSION); MODULE_VERSION(VERSION); MODULE_LICENSE("GPL"); -MODULE_FIRMWARE(FIRMWARE_MT7663); -MODULE_FIRMWARE(FIRMWARE_MT7668); From 877ec9e1d07b631587dbaf205ed929fa08579adf Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:14 +0800 Subject: [PATCH 0022/2048] Bluetooth: btmtksdio: add .set_bdaddr support add .set_bdaddr support Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 53682a4c2ced..77d1c4ac582d 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -834,6 +834,8 @@ static int btmtksdio_probe(struct sdio_func *func, hdev->setup = btmtksdio_setup; hdev->shutdown = btmtksdio_shutdown; hdev->send = btmtksdio_send_frame; + hdev->set_bdaddr = btmtk_set_bdaddr; + SET_HCIDEV_DEV(hdev, &func->dev); hdev->manufacturer = 70; From 77b210d1ae7870159f4ec9ab10b0c29d661b7f99 Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:15 +0800 Subject: [PATCH 0023/2048] Bluetooth: btmtksdio: explicitly set WHISR as write-1-clear That is a preliminary patch to introduce mt7921s support. Explicitly set WHISR as write-1-clear method to all devices that is the expected behavior the driver rely on. Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 77d1c4ac582d..57126a95e292 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -487,8 +487,8 @@ static void btmtksdio_interrupt(struct sdio_func *func) static int btmtksdio_open(struct hci_dev *hdev) { struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); + u32 status, val; int err; - u32 status; sdio_claim_host(bdev->func); @@ -533,8 +533,13 @@ static int btmtksdio_open(struct hci_dev *hdev) if (err < 0) goto err_release_irq; - /* Setup write-1-clear for CHISR register */ - sdio_writel(bdev->func, C_INT_CLR_CTRL, MTK_REG_CHCR, &err); + /* Explitly set write-1-clear method */ + val = sdio_readl(bdev->func, MTK_REG_CHCR, &err); + if (err < 0) + goto err_release_irq; + + val |= C_INT_CLR_CTRL; + sdio_writel(bdev->func, val, MTK_REG_CHCR, &err); if (err < 0) goto err_release_irq; From 26270bc189ea4b5a8356ec99561357fc87f00b32 Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:16 +0800 Subject: [PATCH 0024/2048] Bluetooth: btmtksdio: move interrupt service to work btmtksdio belongs to WIFI/BT combo chip that would serve two radios in one sdio_irq so that we have to move interrupt service to worker to ensure ISR as short as possible. The worker would serve the both Tx and Rx in a batch to effectively reduce many interrupts to the host and to avoid excessive sdio lock contention between various context (even from WiFi driver) and help to be more efficient to complete command/event transation. Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 136 ++++++++++++++++------------------ 1 file changed, 65 insertions(+), 71 deletions(-) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 57126a95e292..f712b52e5797 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -98,7 +98,7 @@ struct btmtksdio_dev { struct sdio_func *func; struct device *dev; - struct work_struct tx_work; + struct work_struct txrx_work; unsigned long tx_state; struct sk_buff_head txq; @@ -249,32 +249,6 @@ static u32 btmtksdio_drv_own_query(struct btmtksdio_dev *bdev) return sdio_readl(bdev->func, MTK_REG_CHLPCR, NULL); } -static void btmtksdio_tx_work(struct work_struct *work) -{ - struct btmtksdio_dev *bdev = container_of(work, struct btmtksdio_dev, - tx_work); - struct sk_buff *skb; - int err; - - pm_runtime_get_sync(bdev->dev); - - sdio_claim_host(bdev->func); - - while ((skb = skb_dequeue(&bdev->txq))) { - err = btmtksdio_tx_packet(bdev, skb); - if (err < 0) { - bdev->hdev->stat.err_tx++; - skb_queue_head(&bdev->txq, skb); - break; - } - } - - sdio_release_host(bdev->func); - - pm_runtime_mark_last_busy(bdev->dev); - pm_runtime_put_autosuspend(bdev->dev); -} - static int btmtksdio_recv_event(struct hci_dev *hdev, struct sk_buff *skb) { struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); @@ -425,65 +399,83 @@ err_kfree_skb: return err; } -static void btmtksdio_interrupt(struct sdio_func *func) +static void btmtksdio_txrx_work(struct work_struct *work) { - struct btmtksdio_dev *bdev = sdio_get_drvdata(func); + struct btmtksdio_dev *bdev = container_of(work, struct btmtksdio_dev, + txrx_work); + unsigned long txrx_timeout; + struct sk_buff *skb; u32 int_status; u16 rx_size; - - /* It is required that the host gets ownership from the device before - * accessing any register, however, if SDIO host is not being released, - * a potential deadlock probably happens in a circular wait between SDIO - * IRQ work and PM runtime work. So, we have to explicitly release SDIO - * host here and claim again after the PM runtime work is all done. - */ - sdio_release_host(bdev->func); + int err; pm_runtime_get_sync(bdev->dev); sdio_claim_host(bdev->func); /* Disable interrupt */ - sdio_writel(func, C_INT_EN_CLR, MTK_REG_CHLPCR, NULL); + sdio_writel(bdev->func, C_INT_EN_CLR, MTK_REG_CHLPCR, 0); - int_status = sdio_readl(func, MTK_REG_CHISR, NULL); - - /* Ack an interrupt as soon as possible before any operation on - * hardware. - * - * Note that we don't ack any status during operations to avoid race - * condition between the host and the device such as it's possible to - * mistakenly ack RX_DONE for the next packet and then cause interrupts - * not be raised again but there is still pending data in the hardware - * FIFO. - */ - sdio_writel(func, int_status, MTK_REG_CHISR, NULL); - - if (unlikely(!int_status)) - bt_dev_err(bdev->hdev, "CHISR is 0"); - - if (int_status & FW_OWN_BACK_INT) - bt_dev_dbg(bdev->hdev, "Get fw own back"); - - if (int_status & TX_EMPTY) - schedule_work(&bdev->tx_work); - else if (unlikely(int_status & TX_FIFO_OVERFLOW)) - bt_dev_warn(bdev->hdev, "Tx fifo overflow"); - - if (int_status & RX_DONE_INT) { - rx_size = (int_status & RX_PKT_LEN) >> 16; - - if (btmtksdio_rx_packet(bdev, rx_size) < 0) - bdev->hdev->stat.err_rx++; + while ((skb = skb_dequeue(&bdev->txq))) { + err = btmtksdio_tx_packet(bdev, skb); + if (err < 0) { + bdev->hdev->stat.err_tx++; + skb_queue_head(&bdev->txq, skb); + break; + } } + txrx_timeout = jiffies + 5 * HZ; + + do { + int_status = sdio_readl(bdev->func, MTK_REG_CHISR, NULL); + + /* Ack an interrupt as soon as possible before any operation on + * hardware. + * + * Note that we don't ack any status during operations to avoid race + * condition between the host and the device such as it's possible to + * mistakenly ack RX_DONE for the next packet and then cause interrupts + * not be raised again but there is still pending data in the hardware + * FIFO. + */ + sdio_writel(bdev->func, int_status, MTK_REG_CHISR, NULL); + + if (int_status & FW_OWN_BACK_INT) + bt_dev_dbg(bdev->hdev, "Get fw own back"); + + if (int_status & TX_EMPTY) + schedule_work(&bdev->txrx_work); + else if (unlikely(int_status & TX_FIFO_OVERFLOW)) + bt_dev_warn(bdev->hdev, "Tx fifo overflow"); + + if (int_status & RX_DONE_INT) { + rx_size = (int_status & RX_PKT_LEN) >> 16; + if (btmtksdio_rx_packet(bdev, rx_size) < 0) + bdev->hdev->stat.err_rx++; + } + + } while (int_status || time_is_before_jiffies(txrx_timeout)); + /* Enable interrupt */ - sdio_writel(func, C_INT_EN_SET, MTK_REG_CHLPCR, NULL); + sdio_writel(bdev->func, C_INT_EN_SET, MTK_REG_CHLPCR, 0); + + sdio_release_host(bdev->func); pm_runtime_mark_last_busy(bdev->dev); pm_runtime_put_autosuspend(bdev->dev); } +static void btmtksdio_interrupt(struct sdio_func *func) +{ + struct btmtksdio_dev *bdev = sdio_get_drvdata(func); + + /* Disable interrupt */ + sdio_writel(bdev->func, C_INT_EN_CLR, MTK_REG_CHLPCR, 0); + + schedule_work(&bdev->txrx_work); +} + static int btmtksdio_open(struct hci_dev *hdev) { struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); @@ -583,6 +575,8 @@ static int btmtksdio_close(struct hci_dev *hdev) sdio_release_irq(bdev->func); + cancel_work_sync(&bdev->txrx_work); + /* Return ownership to the device */ sdio_writel(bdev->func, C_FW_OWN_REQ_SET, MTK_REG_CHLPCR, NULL); @@ -604,7 +598,7 @@ static int btmtksdio_flush(struct hci_dev *hdev) skb_queue_purge(&bdev->txq); - cancel_work_sync(&bdev->tx_work); + cancel_work_sync(&bdev->txrx_work); return 0; } @@ -795,7 +789,7 @@ static int btmtksdio_send_frame(struct hci_dev *hdev, struct sk_buff *skb) skb_queue_tail(&bdev->txq, skb); - schedule_work(&bdev->tx_work); + schedule_work(&bdev->txrx_work); return 0; } @@ -818,7 +812,7 @@ static int btmtksdio_probe(struct sdio_func *func, bdev->dev = &func->dev; bdev->func = func; - INIT_WORK(&bdev->tx_work, btmtksdio_tx_work); + INIT_WORK(&bdev->txrx_work, btmtksdio_txrx_work); skb_queue_head_init(&bdev->txq); /* Initialize and register HCI device */ From 5b23ac1adbc55fc0cda3553a12a6ec4785af748a Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:17 +0800 Subject: [PATCH 0025/2048] Bluetooth: btmtksdio: update register CSDIOCSR operation That is a preliminary patch to introduce mt7921s support. Update register CSDIOCSR operation that is suggested from the vendor driver and is compatible to the all devices. Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index f712b52e5797..3381c8844902 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -520,8 +520,12 @@ static int btmtksdio_open(struct hci_dev *hdev) /* SDIO CMD 5 allows the SDIO device back to idle state an * synchronous interrupt is supported in SDIO 4-bit mode */ - sdio_writel(bdev->func, SDIO_INT_CTL | SDIO_RE_INIT_EN, - MTK_REG_CSDIOCSR, &err); + val = sdio_readl(bdev->func, MTK_REG_CSDIOCSR, &err); + if (err < 0) + goto err_release_irq; + + val |= SDIO_INT_CTL; + sdio_writel(bdev->func, val, MTK_REG_CSDIOCSR, &err); if (err < 0) goto err_release_irq; From 184ea403ccfc1ba04b75908de4eace979ce2ce4e Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:18 +0800 Subject: [PATCH 0026/2048] Bluetooth: btmtksdio: use register CRPLR to read packet length That is a preliminary patch to introduce mt7921s support. Use the register CRPLR to read packet length to make all the devices share the common logic. Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 3381c8844902..f816a7cb0a74 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -83,6 +83,8 @@ MODULE_DEVICE_TABLE(sdio, btmtksdio_table); #define MTK_REG_CRDR 0x1c +#define MTK_REG_CRPLR 0x24 + #define MTK_SDIO_BLOCK_SIZE 256 #define BTMTKSDIO_TX_WAIT_VND_EVT 1 @@ -404,9 +406,8 @@ static void btmtksdio_txrx_work(struct work_struct *work) struct btmtksdio_dev *bdev = container_of(work, struct btmtksdio_dev, txrx_work); unsigned long txrx_timeout; + u32 int_status, rx_size; struct sk_buff *skb; - u32 int_status; - u16 rx_size; int err; pm_runtime_get_sync(bdev->dev); @@ -450,11 +451,11 @@ static void btmtksdio_txrx_work(struct work_struct *work) bt_dev_warn(bdev->hdev, "Tx fifo overflow"); if (int_status & RX_DONE_INT) { - rx_size = (int_status & RX_PKT_LEN) >> 16; + rx_size = sdio_readl(bdev->func, MTK_REG_CRPLR, NULL); + rx_size = (rx_size & RX_PKT_LEN) >> 16; if (btmtksdio_rx_packet(bdev, rx_size) < 0) bdev->hdev->stat.err_rx++; } - } while (int_status || time_is_before_jiffies(txrx_timeout)); /* Enable interrupt */ From 10fe40e1d70a6a6e5210281cebb124d87c41581b Mon Sep 17 00:00:00 2001 From: Mark-yw Chen Date: Tue, 19 Oct 2021 05:30:19 +0800 Subject: [PATCH 0027/2048] Bluetooth: btmtksdio: transmit packet according to status TX_EMPTY Each packet should be sent out until the device emits the TX_EMPTY signal to the host, that is firmware people suggested to add to meet the actual firmware behavior and that is compatible among all the devices. Co-developed-by: Sean Wang Signed-off-by: Sean Wang Signed-off-by: Mark-yw Chen Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtksdio.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index f816a7cb0a74..3266c5d83cae 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -103,6 +103,7 @@ struct btmtksdio_dev { struct work_struct txrx_work; unsigned long tx_state; struct sk_buff_head txq; + bool hw_tx_ready; struct sk_buff *evt_skb; @@ -229,6 +230,7 @@ static int btmtksdio_tx_packet(struct btmtksdio_dev *bdev, sdio_hdr->reserved = cpu_to_le16(0); sdio_hdr->bt_type = hci_skb_pkt_type(skb); + bdev->hw_tx_ready = false; err = sdio_writesb(bdev->func, MTK_REG_CTDR, skb->data, round_up(skb->len, MTK_SDIO_BLOCK_SIZE)); if (err < 0) @@ -417,15 +419,6 @@ static void btmtksdio_txrx_work(struct work_struct *work) /* Disable interrupt */ sdio_writel(bdev->func, C_INT_EN_CLR, MTK_REG_CHLPCR, 0); - while ((skb = skb_dequeue(&bdev->txq))) { - err = btmtksdio_tx_packet(bdev, skb); - if (err < 0) { - bdev->hdev->stat.err_tx++; - skb_queue_head(&bdev->txq, skb); - break; - } - } - txrx_timeout = jiffies + 5 * HZ; do { @@ -446,10 +439,21 @@ static void btmtksdio_txrx_work(struct work_struct *work) bt_dev_dbg(bdev->hdev, "Get fw own back"); if (int_status & TX_EMPTY) - schedule_work(&bdev->txrx_work); + bdev->hw_tx_ready = true; else if (unlikely(int_status & TX_FIFO_OVERFLOW)) bt_dev_warn(bdev->hdev, "Tx fifo overflow"); + if (bdev->hw_tx_ready) { + skb = skb_dequeue(&bdev->txq); + if (skb) { + err = btmtksdio_tx_packet(bdev, skb); + if (err < 0) { + bdev->hdev->stat.err_tx++; + skb_queue_head(&bdev->txq, skb); + } + } + } + if (int_status & RX_DONE_INT) { rx_size = sdio_readl(bdev->func, MTK_REG_CRPLR, NULL); rx_size = (rx_size & RX_PKT_LEN) >> 16; @@ -642,6 +646,7 @@ static int btmtksdio_setup(struct hci_dev *hdev) u8 param = 0x1; calltime = ktime_get(); + bdev->hw_tx_ready = true; /* Query whether the firmware is already download */ wmt_params.op = BTMTK_WMT_SEMAPHORE; From 1705643faecde95bdeb11bea5ab5baed084e9f91 Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:20 +0800 Subject: [PATCH 0028/2048] mmc: add MT7921 SDIO identifiers for MediaTek Bluetooth devices The MT7961 SDIO identifier for MediaTek Bluetooth devices were being referred in the MediaTek Bluetooth driver. Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- include/linux/mmc/sdio_ids.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/mmc/sdio_ids.h b/include/linux/mmc/sdio_ids.h index a85c9f0bd470..53f0efa0bccf 100644 --- a/include/linux/mmc/sdio_ids.h +++ b/include/linux/mmc/sdio_ids.h @@ -105,6 +105,7 @@ #define SDIO_VENDOR_ID_MEDIATEK 0x037a #define SDIO_DEVICE_ID_MEDIATEK_MT7663 0x7663 #define SDIO_DEVICE_ID_MEDIATEK_MT7668 0x7668 +#define SDIO_DEVICE_ID_MEDIATEK_MT7961 0x7961 #define SDIO_VENDOR_ID_MICROCHIP_WILC 0x0296 #define SDIO_DEVICE_ID_MICROCHIP_WILC1000 0x5347 From c603bf1f94d0ef8ce3ef081112eae13cd744ef17 Mon Sep 17 00:00:00 2001 From: Sean Wang Date: Tue, 19 Oct 2021 05:30:21 +0800 Subject: [PATCH 0029/2048] Bluetooth: btmtksdio: add MT7921s Bluetooth support add MT7921s Bluetooth support Co-developed-by: Mark-yw Chen Signed-off-by: Mark-yw Chen Signed-off-by: Sean Wang Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmtk.h | 11 +++ drivers/bluetooth/btmtksdio.c | 141 ++++++++++++++++++++++++++++++++-- 2 files changed, 144 insertions(+), 8 deletions(-) diff --git a/drivers/bluetooth/btmtk.h b/drivers/bluetooth/btmtk.h index 18f43c641b3b..6e7b0c7567c0 100644 --- a/drivers/bluetooth/btmtk.h +++ b/drivers/bluetooth/btmtk.h @@ -7,6 +7,8 @@ #define HCI_WMT_MAX_EVENT_SIZE 64 +#define BTMTK_WMT_REG_READ 0x2 + enum { BTMTK_WMT_PATCH_DWNLD = 0x1, BTMTK_WMT_TEST = 0x2, @@ -14,6 +16,7 @@ enum { BTMTK_WMT_HIF = 0x4, BTMTK_WMT_FUNC_CTRL = 0x6, BTMTK_WMT_RST = 0x7, + BTMTK_WMT_REGISTER = 0x8, BTMTK_WMT_SEMAPHORE = 0x17, }; @@ -49,6 +52,14 @@ struct btmtk_hci_wmt_evt_funcc { __be16 status; } __packed; +struct btmtk_hci_wmt_evt_reg { + struct btmtk_hci_wmt_evt hwhdr; + u8 rsv[2]; + u8 num; + __le32 addr; + __le32 val; +} __packed; + struct btmtk_tci_sleep { u8 mode; __le16 duration; diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c index 3266c5d83cae..ce6a6c00ff98 100644 --- a/drivers/bluetooth/btmtksdio.c +++ b/drivers/bluetooth/btmtksdio.c @@ -37,14 +37,22 @@ static bool enable_autosuspend; struct btmtksdio_data { const char *fwname; + u16 chipid; }; static const struct btmtksdio_data mt7663_data = { .fwname = FIRMWARE_MT7663, + .chipid = 0x7663, }; static const struct btmtksdio_data mt7668_data = { .fwname = FIRMWARE_MT7668, + .chipid = 0x7668, +}; + +static const struct btmtksdio_data mt7921_data = { + .fwname = FIRMWARE_MT7961, + .chipid = 0x7921, }; static const struct sdio_device_id btmtksdio_table[] = { @@ -52,6 +60,8 @@ static const struct sdio_device_id btmtksdio_table[] = { .driver_data = (kernel_ulong_t)&mt7663_data }, {SDIO_DEVICE(SDIO_VENDOR_ID_MEDIATEK, SDIO_DEVICE_ID_MEDIATEK_MT7668), .driver_data = (kernel_ulong_t)&mt7668_data }, + {SDIO_DEVICE(SDIO_VENDOR_ID_MEDIATEK, SDIO_DEVICE_ID_MEDIATEK_MT7961), + .driver_data = (kernel_ulong_t)&mt7921_data }, { } /* Terminating entry */ }; MODULE_DEVICE_TABLE(sdio, btmtksdio_table); @@ -115,6 +125,7 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, { struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); struct btmtk_hci_wmt_evt_funcc *wmt_evt_funcc; + struct btmtk_hci_wmt_evt_reg *wmt_evt_reg; u32 hlen, status = BTMTK_WMT_INVALID; struct btmtk_hci_wmt_evt *wmt_evt; struct btmtk_hci_wmt_cmd *wc; @@ -194,6 +205,19 @@ static int mtk_hci_wmt_sync(struct hci_dev *hdev, else status = BTMTK_WMT_ON_UNDONE; break; + case BTMTK_WMT_PATCH_DWNLD: + if (wmt_evt->whdr.flag == 2) + status = BTMTK_WMT_PATCH_DONE; + else if (wmt_evt->whdr.flag == 1) + status = BTMTK_WMT_PATCH_PROGRESS; + else + status = BTMTK_WMT_PATCH_UNDONE; + break; + case BTMTK_WMT_REGISTER: + wmt_evt_reg = (struct btmtk_hci_wmt_evt_reg *)wmt_evt; + if (le16_to_cpu(wmt_evt->whdr.dlen) == 12) + status = le32_to_cpu(wmt_evt_reg->val); + break; } if (wmt_params->status) @@ -634,20 +658,14 @@ static int btmtksdio_func_query(struct hci_dev *hdev) return status; } -static int btmtksdio_setup(struct hci_dev *hdev) +static int mt76xx_setup(struct hci_dev *hdev, const char *fwname) { - struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); struct btmtk_hci_wmt_params wmt_params; - ktime_t calltime, delta, rettime; struct btmtk_tci_sleep tci_sleep; - unsigned long long duration; struct sk_buff *skb; int err, status; u8 param = 0x1; - calltime = ktime_get(); - bdev->hw_tx_ready = true; - /* Query whether the firmware is already download */ wmt_params.op = BTMTK_WMT_SEMAPHORE; wmt_params.flag = 1; @@ -667,7 +685,7 @@ static int btmtksdio_setup(struct hci_dev *hdev) } /* Setup a firmware which the device definitely requires */ - err = btmtk_setup_firmware(hdev, bdev->data->fwname, mtk_hci_wmt_sync); + err = btmtk_setup_firmware(hdev, fwname, mtk_hci_wmt_sync); if (err < 0) return err; @@ -719,6 +737,113 @@ ignore_func_on: } kfree_skb(skb); + return 0; +} + +static int mt79xx_setup(struct hci_dev *hdev, const char *fwname) +{ + struct btmtk_hci_wmt_params wmt_params; + u8 param = 0x1; + int err; + + err = btmtk_setup_firmware_79xx(hdev, fwname, mtk_hci_wmt_sync); + if (err < 0) { + bt_dev_err(hdev, "Failed to setup 79xx firmware (%d)", err); + return err; + } + + /* Enable Bluetooth protocol */ + wmt_params.op = BTMTK_WMT_FUNC_CTRL; + wmt_params.flag = 0; + wmt_params.dlen = sizeof(param); + wmt_params.data = ¶m; + wmt_params.status = NULL; + + err = mtk_hci_wmt_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to send wmt func ctrl (%d)", err); + return err; + } + + return err; +} + +static int btsdio_mtk_reg_read(struct hci_dev *hdev, u32 reg, u32 *val) +{ + struct btmtk_hci_wmt_params wmt_params; + struct reg_read_cmd { + u8 type; + u8 rsv; + u8 num; + __le32 addr; + } __packed reg_read = { + .type = 1, + .num = 1, + }; + u32 status; + int err; + + reg_read.addr = cpu_to_le32(reg); + wmt_params.op = BTMTK_WMT_REGISTER; + wmt_params.flag = BTMTK_WMT_REG_READ; + wmt_params.dlen = sizeof(reg_read); + wmt_params.data = ®_read; + wmt_params.status = &status; + + err = mtk_hci_wmt_sync(hdev, &wmt_params); + if (err < 0) { + bt_dev_err(hdev, "Failed to read reg(%d)", err); + return err; + } + + *val = status; + + return err; +} + +static int btmtksdio_setup(struct hci_dev *hdev) +{ + struct btmtksdio_dev *bdev = hci_get_drvdata(hdev); + ktime_t calltime, delta, rettime; + unsigned long long duration; + char fwname[64]; + int err, dev_id; + u32 fw_version = 0; + + calltime = ktime_get(); + bdev->hw_tx_ready = true; + + switch (bdev->data->chipid) { + case 0x7921: + err = btsdio_mtk_reg_read(hdev, 0x70010200, &dev_id); + if (err < 0) { + bt_dev_err(hdev, "Failed to get device id (%d)", err); + return err; + } + + err = btsdio_mtk_reg_read(hdev, 0x80021004, &fw_version); + if (err < 0) { + bt_dev_err(hdev, "Failed to get fw version (%d)", err); + return err; + } + + snprintf(fwname, sizeof(fwname), + "mediatek/BT_RAM_CODE_MT%04x_1_%x_hdr.bin", + dev_id & 0xffff, (fw_version & 0xff) + 1); + err = mt79xx_setup(hdev, fwname); + if (err < 0) + return err; + break; + case 0x7663: + case 0x7668: + err = mt76xx_setup(hdev, bdev->data->fwname); + if (err < 0) + return err; + break; + default: + return -ENODEV; + } + rettime = ktime_get(); delta = ktime_sub(rettime, calltime); duration = (unsigned long long)ktime_to_ns(delta) >> 10; From 2128939fe2e771645dd88e1938c27fdf96bd1cd0 Mon Sep 17 00:00:00 2001 From: Archie Pusaka Date: Thu, 28 Oct 2021 19:17:25 +0800 Subject: [PATCH 0030/2048] Bluetooth: Fix removing adv when processing cmd complete If we remove one instance of adv using Set Extended Adv Enable, there is a possibility of issue occurs when processing the Command Complete event. Especially, the adv_info might not be found since we already remove it in hci_req_clear_adv_instance() -> hci_remove_adv_instance(). If that's the case, we will mistakenly proceed to remove all adv instances instead of just one single instance. This patch fixes the issue by checking the content of the HCI command instead of checking whether the adv_info is found. Signed-off-by: Archie Pusaka Reviewed-by: Sonny Sasaka Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_event.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 3cba2bbefcd6..894670419a27 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -1326,8 +1326,10 @@ static void hci_cc_le_set_ext_adv_enable(struct hci_dev *hdev, &conn->le_conn_timeout, conn->conn_timeout); } else { - if (adv) { - adv->enabled = false; + if (cp->num_of_sets) { + if (adv) + adv->enabled = false; + /* If just one instance was disabled check if there are * any other instance enabled before clearing HCI_LE_ADV */ From 6a98e3836fa2077b169f10a35c2ca9952d53f987 Mon Sep 17 00:00:00 2001 From: Marcel Holtmann Date: Wed, 27 Oct 2021 16:58:38 -0700 Subject: [PATCH 0031/2048] Bluetooth: Add helper for serialized HCI command execution The usage of __hci_cmd_sync() within the hdev->setup() callback allows for a nice and simple serialized execution of HCI commands. More importantly it allows for result processing before issueing the next command. With the current usage of hci_req_run() it is possible to batch up commands and execute them, but it is impossible to react to their results or errors. This is an attempt to generalize the hdev->setup() handling and provide a simple way of running multiple HCI commands from a single function context. There are multiple struct work that are decdicated to certain tasks already used right now. It is add a lot of bloat to hci_dev struct and extra handling code. So it might be possible to put all of these behind a common HCI command infrastructure and just execute the HCI commands from the same work context in a serialized fashion. For example updating the white list and resolving list can be done now without having to know the list size ahead of time. Also preparing for suspend or resume shouldn't require a state machine anymore. There are other tasks that should be simplified as well. Signed-off-by: Marcel Holtmann Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 11 +- include/net/bluetooth/hci_sync.h | 42 ++++ net/bluetooth/Makefile | 2 +- net/bluetooth/hci_core.c | 23 +-- net/bluetooth/hci_request.c | 68 ------- net/bluetooth/hci_request.h | 4 + net/bluetooth/hci_sync.c | 330 +++++++++++++++++++++++++++++++ 7 files changed, 385 insertions(+), 95 deletions(-) create mode 100644 include/net/bluetooth/hci_sync.h create mode 100644 net/bluetooth/hci_sync.c diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index dd8840e70e25..17ddea51d161 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -30,6 +30,7 @@ #include #include +#include #include /* HCI priority */ @@ -475,6 +476,9 @@ struct hci_dev { struct work_struct power_on; struct delayed_work power_off; struct work_struct error_reset; + struct work_struct cmd_sync_work; + struct list_head cmd_sync_work_list; + struct mutex cmd_sync_work_lock; __u16 discov_timeout; struct delayed_work discov_off; @@ -1690,10 +1694,6 @@ static inline int hci_check_conn_params(u16 min, u16 max, u16 latency, int hci_register_cb(struct hci_cb *hcb); int hci_unregister_cb(struct hci_cb *hcb); -struct sk_buff *__hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u32 timeout); -struct sk_buff *__hci_cmd_sync_ev(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u8 event, u32 timeout); int __hci_cmd_send(struct hci_dev *hdev, u16 opcode, u32 plen, const void *param); @@ -1704,9 +1704,6 @@ void hci_send_sco(struct hci_conn *conn, struct sk_buff *skb); void *hci_sent_cmd_data(struct hci_dev *hdev, __u16 opcode); -struct sk_buff *hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u32 timeout); - u32 hci_conn_get_phy(struct hci_conn *conn); /* ----- HCI Sockets ----- */ diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h new file mode 100644 index 000000000000..fcfdeb3cbd7c --- /dev/null +++ b/include/net/bluetooth/hci_sync.h @@ -0,0 +1,42 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * BlueZ - Bluetooth protocol stack for Linux + * + * Copyright (C) 2021 Intel Corporation + */ + +typedef int (*hci_cmd_sync_work_func_t)(struct hci_dev *hdev, void *data); +typedef void (*hci_cmd_sync_work_destroy_t)(struct hci_dev *hdev, void *data, + int err); + +struct hci_cmd_sync_work_entry { + struct list_head list; + hci_cmd_sync_work_func_t func; + void *data; + hci_cmd_sync_work_destroy_t destroy; +}; + +/* Function with sync suffix shall not be called with hdev->lock held as they + * wait the command to complete and in the meantime an event could be received + * which could attempt to acquire hdev->lock causing a deadlock. + */ +struct sk_buff *__hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout); +struct sk_buff *hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout); +struct sk_buff *__hci_cmd_sync_ev(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout); +struct sk_buff *__hci_cmd_sync_sk(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout, + struct sock *sk); +int __hci_cmd_sync_status(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout); +int __hci_cmd_sync_status_sk(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout, + struct sock *sk); + +void hci_cmd_sync_init(struct hci_dev *hdev); +void hci_cmd_sync_clear(struct hci_dev *hdev); + +int hci_cmd_sync_queue(struct hci_dev *hdev, hci_cmd_sync_work_func_t func, + void *data, hci_cmd_sync_work_destroy_t destroy); diff --git a/net/bluetooth/Makefile b/net/bluetooth/Makefile index 291770fc9551..a52bba8500e1 100644 --- a/net/bluetooth/Makefile +++ b/net/bluetooth/Makefile @@ -15,7 +15,7 @@ bluetooth_6lowpan-y := 6lowpan.o bluetooth-y := af_bluetooth.o hci_core.o hci_conn.o hci_event.o mgmt.o \ hci_sock.o hci_sysfs.o l2cap_core.o l2cap_sock.o smp.o lib.o \ ecdh_helper.o hci_request.o mgmt_util.o mgmt_config.o hci_codec.o \ - eir.o + eir.o hci_sync.o bluetooth-$(CONFIG_BT_BREDR) += sco.o bluetooth-$(CONFIG_BT_HS) += a2mp.o amp.o diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 98533def61a3..fdc1cce5eb5b 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3747,6 +3747,8 @@ struct hci_dev *hci_alloc_dev_priv(int sizeof_priv) INIT_WORK(&hdev->error_reset, hci_error_reset); INIT_WORK(&hdev->suspend_prepare, hci_prepare_suspend); + hci_cmd_sync_init(hdev); + INIT_DELAYED_WORK(&hdev->power_off, hci_power_off); skb_queue_head_init(&hdev->rx_q); @@ -3905,6 +3907,8 @@ void hci_unregister_dev(struct hci_dev *hdev) cancel_work_sync(&hdev->power_on); + hci_cmd_sync_clear(hdev); + if (!test_bit(HCI_QUIRK_NO_SUSPEND_NOTIFIER, &hdev->quirks)) { hci_suspend_clear_tasks(hdev); unregister_pm_notifier(&hdev->suspend_notifier); @@ -4271,25 +4275,6 @@ void *hci_sent_cmd_data(struct hci_dev *hdev, __u16 opcode) return hdev->sent_cmd->data + HCI_COMMAND_HDR_SIZE; } -/* Send HCI command and wait for command complete event */ -struct sk_buff *hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u32 timeout) -{ - struct sk_buff *skb; - - if (!test_bit(HCI_UP, &hdev->flags)) - return ERR_PTR(-ENETDOWN); - - bt_dev_dbg(hdev, "opcode 0x%4.4x plen %d", opcode, plen); - - hci_req_sync_lock(hdev); - skb = __hci_cmd_sync(hdev, opcode, plen, param, timeout); - hci_req_sync_unlock(hdev); - - return skb; -} -EXPORT_SYMBOL(hci_cmd_sync); - /* Send ACL data */ static void hci_add_acl_hdr(struct sk_buff *skb, __u16 handle, __u16 flags) { diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 92611bfc0b9e..c98340bba5c7 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -32,10 +32,6 @@ #include "msft.h" #include "eir.h" -#define HCI_REQ_DONE 0 -#define HCI_REQ_PEND 1 -#define HCI_REQ_CANCELED 2 - void hci_req_init(struct hci_request *req, struct hci_dev *hdev) { skb_queue_head_init(&req->cmd_q); @@ -126,70 +122,6 @@ void hci_req_sync_cancel(struct hci_dev *hdev, int err) } } -struct sk_buff *__hci_cmd_sync_ev(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u8 event, u32 timeout) -{ - struct hci_request req; - struct sk_buff *skb; - int err = 0; - - bt_dev_dbg(hdev, ""); - - hci_req_init(&req, hdev); - - hci_req_add_ev(&req, opcode, plen, param, event); - - hdev->req_status = HCI_REQ_PEND; - - err = hci_req_run_skb(&req, hci_req_sync_complete); - if (err < 0) - return ERR_PTR(err); - - err = wait_event_interruptible_timeout(hdev->req_wait_q, - hdev->req_status != HCI_REQ_PEND, timeout); - - if (err == -ERESTARTSYS) - return ERR_PTR(-EINTR); - - switch (hdev->req_status) { - case HCI_REQ_DONE: - err = -bt_to_errno(hdev->req_result); - break; - - case HCI_REQ_CANCELED: - err = -hdev->req_result; - break; - - default: - err = -ETIMEDOUT; - break; - } - - hdev->req_status = hdev->req_result = 0; - skb = hdev->req_skb; - hdev->req_skb = NULL; - - bt_dev_dbg(hdev, "end: err %d", err); - - if (err < 0) { - kfree_skb(skb); - return ERR_PTR(err); - } - - if (!skb) - return ERR_PTR(-ENODATA); - - return skb; -} -EXPORT_SYMBOL(__hci_cmd_sync_ev); - -struct sk_buff *__hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, - const void *param, u32 timeout) -{ - return __hci_cmd_sync_ev(hdev, opcode, plen, param, 0, timeout); -} -EXPORT_SYMBOL(__hci_cmd_sync); - /* Execute request and wait for completion. */ int __hci_req_sync(struct hci_dev *hdev, int (*func)(struct hci_request *req, unsigned long opt), diff --git a/net/bluetooth/hci_request.h b/net/bluetooth/hci_request.h index f31420f58525..74c5de5970a7 100644 --- a/net/bluetooth/hci_request.h +++ b/net/bluetooth/hci_request.h @@ -25,6 +25,10 @@ #define hci_req_sync_lock(hdev) mutex_lock(&hdev->req_lock) #define hci_req_sync_unlock(hdev) mutex_unlock(&hdev->req_lock) +#define HCI_REQ_DONE 0 +#define HCI_REQ_PEND 1 +#define HCI_REQ_CANCELED 2 + struct hci_request { struct hci_dev *hdev; struct sk_buff_head cmd_q; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c new file mode 100644 index 000000000000..b2048287fe90 --- /dev/null +++ b/net/bluetooth/hci_sync.c @@ -0,0 +1,330 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * BlueZ - Bluetooth protocol stack for Linux + * + * Copyright (C) 2021 Intel Corporation + */ + +#include +#include +#include + +#include "hci_request.h" +#include "smp.h" + +static void hci_cmd_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, + struct sk_buff *skb) +{ + bt_dev_dbg(hdev, "result 0x%2.2x", result); + + if (hdev->req_status != HCI_REQ_PEND) + return; + + hdev->req_result = result; + hdev->req_status = HCI_REQ_DONE; + + wake_up_interruptible(&hdev->req_wait_q); +} + +static struct sk_buff *hci_cmd_sync_alloc(struct hci_dev *hdev, u16 opcode, + u32 plen, const void *param, + struct sock *sk) +{ + int len = HCI_COMMAND_HDR_SIZE + plen; + struct hci_command_hdr *hdr; + struct sk_buff *skb; + + skb = bt_skb_alloc(len, GFP_ATOMIC); + if (!skb) + return NULL; + + hdr = skb_put(skb, HCI_COMMAND_HDR_SIZE); + hdr->opcode = cpu_to_le16(opcode); + hdr->plen = plen; + + if (plen) + skb_put_data(skb, param, plen); + + bt_dev_dbg(hdev, "skb len %d", skb->len); + + hci_skb_pkt_type(skb) = HCI_COMMAND_PKT; + hci_skb_opcode(skb) = opcode; + + return skb; +} + +static void hci_cmd_sync_add(struct hci_request *req, u16 opcode, u32 plen, + const void *param, u8 event, struct sock *sk) +{ + struct hci_dev *hdev = req->hdev; + struct sk_buff *skb; + + bt_dev_dbg(hdev, "opcode 0x%4.4x plen %d", opcode, plen); + + /* If an error occurred during request building, there is no point in + * queueing the HCI command. We can simply return. + */ + if (req->err) + return; + + skb = hci_cmd_sync_alloc(hdev, opcode, plen, param, sk); + if (!skb) { + bt_dev_err(hdev, "no memory for command (opcode 0x%4.4x)", + opcode); + req->err = -ENOMEM; + return; + } + + if (skb_queue_empty(&req->cmd_q)) + bt_cb(skb)->hci.req_flags |= HCI_REQ_START; + + bt_cb(skb)->hci.req_event = event; + + skb_queue_tail(&req->cmd_q, skb); +} + +static int hci_cmd_sync_run(struct hci_request *req) +{ + struct hci_dev *hdev = req->hdev; + struct sk_buff *skb; + unsigned long flags; + + bt_dev_dbg(hdev, "length %u", skb_queue_len(&req->cmd_q)); + + /* If an error occurred during request building, remove all HCI + * commands queued on the HCI request queue. + */ + if (req->err) { + skb_queue_purge(&req->cmd_q); + return req->err; + } + + /* Do not allow empty requests */ + if (skb_queue_empty(&req->cmd_q)) + return -ENODATA; + + skb = skb_peek_tail(&req->cmd_q); + bt_cb(skb)->hci.req_complete_skb = hci_cmd_sync_complete; + bt_cb(skb)->hci.req_flags |= HCI_REQ_SKB; + + spin_lock_irqsave(&hdev->cmd_q.lock, flags); + skb_queue_splice_tail(&req->cmd_q, &hdev->cmd_q); + spin_unlock_irqrestore(&hdev->cmd_q.lock, flags); + + queue_work(hdev->workqueue, &hdev->cmd_work); + + return 0; +} + +/* This function requires the caller holds hdev->req_lock. */ +struct sk_buff *__hci_cmd_sync_sk(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout, + struct sock *sk) +{ + struct hci_request req; + struct sk_buff *skb; + int err = 0; + + bt_dev_dbg(hdev, ""); + + hci_req_init(&req, hdev); + + hci_cmd_sync_add(&req, opcode, plen, param, event, sk); + + hdev->req_status = HCI_REQ_PEND; + + err = hci_cmd_sync_run(&req); + if (err < 0) + return ERR_PTR(err); + + err = wait_event_interruptible_timeout(hdev->req_wait_q, + hdev->req_status != HCI_REQ_PEND, + timeout); + + if (err == -ERESTARTSYS) + return ERR_PTR(-EINTR); + + switch (hdev->req_status) { + case HCI_REQ_DONE: + err = -bt_to_errno(hdev->req_result); + break; + + case HCI_REQ_CANCELED: + err = -hdev->req_result; + break; + + default: + err = -ETIMEDOUT; + break; + } + + hdev->req_status = 0; + hdev->req_result = 0; + skb = hdev->req_skb; + hdev->req_skb = NULL; + + bt_dev_dbg(hdev, "end: err %d", err); + + if (err < 0) { + kfree_skb(skb); + return ERR_PTR(err); + } + + if (!skb) + return ERR_PTR(-ENODATA); + + return skb; +} +EXPORT_SYMBOL(__hci_cmd_sync_sk); + +/* This function requires the caller holds hdev->req_lock. */ +struct sk_buff *__hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout) +{ + return __hci_cmd_sync_sk(hdev, opcode, plen, param, 0, timeout, NULL); +} +EXPORT_SYMBOL(__hci_cmd_sync); + +/* Send HCI command and wait for command complete event */ +struct sk_buff *hci_cmd_sync(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout) +{ + struct sk_buff *skb; + + if (!test_bit(HCI_UP, &hdev->flags)) + return ERR_PTR(-ENETDOWN); + + bt_dev_dbg(hdev, "opcode 0x%4.4x plen %d", opcode, plen); + + hci_req_sync_lock(hdev); + skb = __hci_cmd_sync(hdev, opcode, plen, param, timeout); + hci_req_sync_unlock(hdev); + + return skb; +} +EXPORT_SYMBOL(hci_cmd_sync); + +/* This function requires the caller holds hdev->req_lock. */ +struct sk_buff *__hci_cmd_sync_ev(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout) +{ + return __hci_cmd_sync_sk(hdev, opcode, plen, param, event, timeout, + NULL); +} +EXPORT_SYMBOL(__hci_cmd_sync_ev); + +/* This function requires the caller holds hdev->req_lock. */ +int __hci_cmd_sync_status_sk(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u8 event, u32 timeout, + struct sock *sk) +{ + struct sk_buff *skb; + u8 status; + + skb = __hci_cmd_sync_sk(hdev, opcode, plen, param, event, timeout, sk); + if (IS_ERR_OR_NULL(skb)) { + bt_dev_err(hdev, "Opcode 0x%4x failed: %ld", opcode, + PTR_ERR(skb)); + return PTR_ERR(skb); + } + + status = skb->data[0]; + + kfree_skb(skb); + + return status; +} +EXPORT_SYMBOL(__hci_cmd_sync_status_sk); + +int __hci_cmd_sync_status(struct hci_dev *hdev, u16 opcode, u32 plen, + const void *param, u32 timeout) +{ + return __hci_cmd_sync_status_sk(hdev, opcode, plen, param, 0, timeout, + NULL); +} +EXPORT_SYMBOL(__hci_cmd_sync_status); + +static void hci_cmd_sync_work(struct work_struct *work) +{ + struct hci_dev *hdev = container_of(work, struct hci_dev, cmd_sync_work); + struct hci_cmd_sync_work_entry *entry; + hci_cmd_sync_work_func_t func; + hci_cmd_sync_work_destroy_t destroy; + void *data; + + bt_dev_dbg(hdev, ""); + + mutex_lock(&hdev->cmd_sync_work_lock); + entry = list_first_entry(&hdev->cmd_sync_work_list, + struct hci_cmd_sync_work_entry, list); + if (entry) { + list_del(&entry->list); + func = entry->func; + data = entry->data; + destroy = entry->destroy; + kfree(entry); + } else { + func = NULL; + data = NULL; + destroy = NULL; + } + mutex_unlock(&hdev->cmd_sync_work_lock); + + if (func) { + int err; + + hci_req_sync_lock(hdev); + + err = func(hdev, data); + + if (destroy) + destroy(hdev, data, err); + + hci_req_sync_unlock(hdev); + } +} + +void hci_cmd_sync_init(struct hci_dev *hdev) +{ + INIT_WORK(&hdev->cmd_sync_work, hci_cmd_sync_work); + INIT_LIST_HEAD(&hdev->cmd_sync_work_list); + mutex_init(&hdev->cmd_sync_work_lock); +} + +void hci_cmd_sync_clear(struct hci_dev *hdev) +{ + struct hci_cmd_sync_work_entry *entry, *tmp; + + cancel_work_sync(&hdev->cmd_sync_work); + + list_for_each_entry_safe(entry, tmp, &hdev->cmd_sync_work_list, list) { + if (entry->destroy) + entry->destroy(hdev, entry->data, -ECANCELED); + + list_del(&entry->list); + kfree(entry); + } +} + +int hci_cmd_sync_queue(struct hci_dev *hdev, hci_cmd_sync_work_func_t func, + void *data, hci_cmd_sync_work_destroy_t destroy) +{ + struct hci_cmd_sync_work_entry *entry; + + entry = kmalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + return -ENOMEM; + + entry->func = func; + entry->data = data; + entry->destroy = destroy; + + mutex_lock(&hdev->cmd_sync_work_lock); + list_add_tail(&entry->list, &hdev->cmd_sync_work_list); + mutex_unlock(&hdev->cmd_sync_work_lock); + + queue_work(hdev->req_workqueue, &hdev->cmd_sync_work); + + return 0; +} +EXPORT_SYMBOL(hci_cmd_sync_queue); From 161510ccf91c961638940b03abb1ee804be53a97 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:39 -0700 Subject: [PATCH 0032/2048] Bluetooth: hci_sync: Make use of hci_cmd_sync_queue set 1 This make use of hci_cmd_sync_queue for the following MGMT commands: Set Device Class Set Device ID Add UUID Remove UUID tools/mgmt-tester -s "Set Device Class" Test Summary ------------ Set Device Class - Success 1 Passed Set Device Class - Success 2 Passed Set Device Class - Invalid parameters 1 Passed Total: 3, Passed: 3 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.0599 seconds tools/mgmt-tester -s "Set Device ID" Test Summary ------------ Set Device ID - Success 1 Passed Set Device ID - Success 2 Passed Set Device ID - Disable Passed Set Device ID - Power off and Power on Passed Set Device ID - SSP off and Power on Passed Set Device ID - Invalid Parameter Passed Total: 6, Passed: 6 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.107 seconds tools/mgmt-tester -s "Add UUID" Test Summary ------------ Add UUID - UUID-16 1 Passed Add UUID - UUID-16 multiple 1 Passed Add UUID - UUID-16 partial 1 Passed Add UUID - UUID-32 1 Passed Add UUID - UUID-32 multiple 1 Passed Add UUID - UUID-32 partial 1 Passed Add UUID - UUID-128 1 Passed Add UUID - UUID-128 multiple 1 Passed Add UUID - UUID-128 partial 1 Passed Add UUID - UUID mix Passed Total: 10, Passed: 10 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.198 seconds tools/mgmt-tester -s "Remove UUID" Test Summary ------------ Remove UUID - Success 1 Passed Remove UUID - All UUID - Success 2 Passed Remove UUID - Power Off - Success 3 Passed Remove UUID - Power Off and On - Success 4 Passed Remove UUID - Not Exist - Invalid Params 1 Passed Total: 5, Passed: 5 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.0908 seconds Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 3 + net/bluetooth/hci_request.c | 4 +- net/bluetooth/hci_request.h | 6 + net/bluetooth/hci_sync.c | 72 ++++++++++++ net/bluetooth/mgmt.c | 192 +++++++++++++++---------------- net/bluetooth/mgmt_util.c | 15 ++- net/bluetooth/mgmt_util.h | 3 + 7 files changed, 193 insertions(+), 102 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index fcfdeb3cbd7c..fe77ff97bc8e 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -40,3 +40,6 @@ void hci_cmd_sync_clear(struct hci_dev *hdev); int hci_cmd_sync_queue(struct hci_dev *hdev, hci_cmd_sync_work_func_t func, void *data, hci_cmd_sync_work_destroy_t destroy); + +int hci_update_eir_sync(struct hci_dev *hdev); +int hci_update_class_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index c98340bba5c7..b1b33432c68d 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -97,8 +97,8 @@ int hci_req_run_skb(struct hci_request *req, hci_req_complete_skb_t complete) return req_run(req, NULL, complete); } -static void hci_req_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, - struct sk_buff *skb) +void hci_req_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, + struct sk_buff *skb) { bt_dev_dbg(hdev, "result 0x%2.2x", result); diff --git a/net/bluetooth/hci_request.h b/net/bluetooth/hci_request.h index 74c5de5970a7..ba75c2da70f4 100644 --- a/net/bluetooth/hci_request.h +++ b/net/bluetooth/hci_request.h @@ -22,6 +22,10 @@ #include +#define HCI_REQ_DONE 0 +#define HCI_REQ_PEND 1 +#define HCI_REQ_CANCELED 2 + #define hci_req_sync_lock(hdev) mutex_lock(&hdev->req_lock) #define hci_req_sync_unlock(hdev) mutex_unlock(&hdev->req_lock) @@ -44,6 +48,8 @@ void hci_req_purge(struct hci_request *req); bool hci_req_status_pend(struct hci_dev *hdev); int hci_req_run(struct hci_request *req, hci_req_complete_t complete); int hci_req_run_skb(struct hci_request *req, hci_req_complete_skb_t complete); +void hci_req_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, + struct sk_buff *skb); void hci_req_add(struct hci_request *req, u16 opcode, u32 plen, const void *param); void hci_req_add_ev(struct hci_request *req, u16 opcode, u32 plen, diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index b2048287fe90..8546cbcae10c 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -11,6 +11,7 @@ #include "hci_request.h" #include "smp.h" +#include "eir.h" static void hci_cmd_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, struct sk_buff *skb) @@ -328,3 +329,74 @@ int hci_cmd_sync_queue(struct hci_dev *hdev, hci_cmd_sync_work_func_t func, return 0; } EXPORT_SYMBOL(hci_cmd_sync_queue); + +int hci_update_eir_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_eir cp; + + bt_dev_dbg(hdev, ""); + + if (!hdev_is_powered(hdev)) + return 0; + + if (!lmp_ext_inq_capable(hdev)) + return 0; + + if (!hci_dev_test_flag(hdev, HCI_SSP_ENABLED)) + return 0; + + if (hci_dev_test_flag(hdev, HCI_SERVICE_CACHE)) + return 0; + + memset(&cp, 0, sizeof(cp)); + + eir_create(hdev, cp.data); + + if (memcmp(cp.data, hdev->eir, sizeof(cp.data)) == 0) + return 0; + + memcpy(hdev->eir, cp.data, sizeof(cp.data)); + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_EIR, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); +} + +static u8 get_service_classes(struct hci_dev *hdev) +{ + struct bt_uuid *uuid; + u8 val = 0; + + list_for_each_entry(uuid, &hdev->uuids, list) + val |= uuid->svc_hint; + + return val; +} + +int hci_update_class_sync(struct hci_dev *hdev) +{ + u8 cod[3]; + + bt_dev_dbg(hdev, ""); + + if (!hdev_is_powered(hdev)) + return 0; + + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + return 0; + + if (hci_dev_test_flag(hdev, HCI_SERVICE_CACHE)) + return 0; + + cod[0] = hdev->minor_class; + cod[1] = hdev->major_class; + cod[2] = get_service_classes(hdev); + + if (hci_dev_test_flag(hdev, HCI_LIMITED_DISCOVERABLE)) + cod[1] |= 0x20; + + if (memcmp(cod, hdev->dev_class, 3) == 0) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_CLASS_OF_DEV, + sizeof(cod), cod, HCI_CMD_TIMEOUT); +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 44683443300c..ec6aa7af482f 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -276,10 +276,39 @@ static const u8 mgmt_status_table[] = { MGMT_STATUS_CONNECT_FAILED, /* MAC Connection Failed */ }; -static u8 mgmt_status(u8 hci_status) +static u8 mgmt_errno_status(int err) { - if (hci_status < ARRAY_SIZE(mgmt_status_table)) - return mgmt_status_table[hci_status]; + switch (err) { + case 0: + return MGMT_STATUS_SUCCESS; + case -EPERM: + return MGMT_STATUS_REJECTED; + case -EINVAL: + return MGMT_STATUS_INVALID_PARAMS; + case -EOPNOTSUPP: + return MGMT_STATUS_NOT_SUPPORTED; + case -EBUSY: + return MGMT_STATUS_BUSY; + case -ETIMEDOUT: + return MGMT_STATUS_AUTH_FAILED; + case -ENOMEM: + return MGMT_STATUS_NO_RESOURCES; + case -EISCONN: + return MGMT_STATUS_ALREADY_CONNECTED; + case -ENOTCONN: + return MGMT_STATUS_DISCONNECTED; + } + + return MGMT_STATUS_FAILED; +} + +static u8 mgmt_status(int err) +{ + if (err < 0) + return mgmt_errno_status(err); + + if (err < ARRAY_SIZE(mgmt_status_table)) + return mgmt_status_table[err]; return MGMT_STATUS_FAILED; } @@ -951,25 +980,23 @@ bool mgmt_get_connectable(struct hci_dev *hdev) return hci_dev_test_flag(hdev, HCI_CONNECTABLE); } +static int service_cache_sync(struct hci_dev *hdev, void *data) +{ + hci_update_eir_sync(hdev); + hci_update_class_sync(hdev); + + return 0; +} + static void service_cache_off(struct work_struct *work) { struct hci_dev *hdev = container_of(work, struct hci_dev, service_cache.work); - struct hci_request req; if (!hci_dev_test_and_clear_flag(hdev, HCI_SERVICE_CACHE)) return; - hci_req_init(&req, hdev); - - hci_dev_lock(hdev); - - __hci_req_update_eir(&req); - __hci_req_update_class(&req); - - hci_dev_unlock(hdev); - - hci_req_run(&req, NULL); + hci_cmd_sync_queue(hdev, service_cache_sync, NULL, NULL); } static void rpa_expired(struct work_struct *work) @@ -2075,37 +2102,33 @@ static u8 get_uuid_size(const u8 *uuid) return 16; } -static void mgmt_class_complete(struct hci_dev *hdev, u16 mgmt_op, u8 status) +static void mgmt_class_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; - hci_dev_lock(hdev); - - cmd = pending_find(mgmt_op, hdev); - if (!cmd) - goto unlock; + bt_dev_dbg(hdev, "err %d", err); mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status), hdev->dev_class, 3); + mgmt_status(err), hdev->dev_class, 3); - mgmt_pending_remove(cmd); - -unlock: - hci_dev_unlock(hdev); + mgmt_pending_free(cmd); } -static void add_uuid_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static int add_uuid_sync(struct hci_dev *hdev, void *data) { - bt_dev_dbg(hdev, "status 0x%02x", status); + int err; - mgmt_class_complete(hdev, MGMT_OP_ADD_UUID, status); + err = hci_update_class_sync(hdev); + if (err) + return err; + + return hci_update_eir_sync(hdev); } static int add_uuid(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_cp_add_uuid *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; struct bt_uuid *uuid; int err; @@ -2131,28 +2154,17 @@ static int add_uuid(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) list_add_tail(&uuid->list, &hdev->uuids); - hci_req_init(&req, hdev); - - __hci_req_update_class(&req); - __hci_req_update_eir(&req); - - err = hci_req_run(&req, add_uuid_complete); - if (err < 0) { - if (err != -ENODATA) - goto failed; - - err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_ADD_UUID, 0, - hdev->dev_class, 3); - goto failed; - } - - cmd = mgmt_pending_add(sk, MGMT_OP_ADD_UUID, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_ADD_UUID, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; } - err = 0; + err = hci_cmd_sync_queue(hdev, add_uuid_sync, cmd, mgmt_class_complete); + if (err < 0) { + mgmt_pending_free(cmd); + goto failed; + } failed: hci_dev_unlock(hdev); @@ -2173,11 +2185,15 @@ static bool enable_service_cache(struct hci_dev *hdev) return false; } -static void remove_uuid_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static int remove_uuid_sync(struct hci_dev *hdev, void *data) { - bt_dev_dbg(hdev, "status 0x%02x", status); + int err; - mgmt_class_complete(hdev, MGMT_OP_REMOVE_UUID, status); + err = hci_update_class_sync(hdev); + if (err) + return err; + + return hci_update_eir_sync(hdev); } static int remove_uuid(struct sock *sk, struct hci_dev *hdev, void *data, @@ -2187,7 +2203,6 @@ static int remove_uuid(struct sock *sk, struct hci_dev *hdev, void *data, struct mgmt_pending_cmd *cmd; struct bt_uuid *match, *tmp; u8 bt_uuid_any[] = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }; - struct hci_request req; int err, found; bt_dev_dbg(hdev, "sock %p", sk); @@ -2231,39 +2246,35 @@ static int remove_uuid(struct sock *sk, struct hci_dev *hdev, void *data, } update_class: - hci_req_init(&req, hdev); - - __hci_req_update_class(&req); - __hci_req_update_eir(&req); - - err = hci_req_run(&req, remove_uuid_complete); - if (err < 0) { - if (err != -ENODATA) - goto unlock; - - err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_REMOVE_UUID, 0, - hdev->dev_class, 3); - goto unlock; - } - - cmd = mgmt_pending_add(sk, MGMT_OP_REMOVE_UUID, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_REMOVE_UUID, hdev, data, len); if (!cmd) { err = -ENOMEM; goto unlock; } - err = 0; + err = hci_cmd_sync_queue(hdev, remove_uuid_sync, cmd, + mgmt_class_complete); + if (err < 0) + mgmt_pending_free(cmd); unlock: hci_dev_unlock(hdev); return err; } -static void set_class_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static int set_class_sync(struct hci_dev *hdev, void *data) { - bt_dev_dbg(hdev, "status 0x%02x", status); + int err = 0; - mgmt_class_complete(hdev, MGMT_OP_SET_DEV_CLASS, status); + if (hci_dev_test_and_clear_flag(hdev, HCI_SERVICE_CACHE)) { + cancel_delayed_work_sync(&hdev->service_cache); + err = hci_update_eir_sync(hdev); + } + + if (err) + return err; + + return hci_update_class_sync(hdev); } static int set_dev_class(struct sock *sk, struct hci_dev *hdev, void *data, @@ -2271,7 +2282,6 @@ static int set_dev_class(struct sock *sk, struct hci_dev *hdev, void *data, { struct mgmt_cp_set_dev_class *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); @@ -2303,34 +2313,16 @@ static int set_dev_class(struct sock *sk, struct hci_dev *hdev, void *data, goto unlock; } - hci_req_init(&req, hdev); - - if (hci_dev_test_and_clear_flag(hdev, HCI_SERVICE_CACHE)) { - hci_dev_unlock(hdev); - cancel_delayed_work_sync(&hdev->service_cache); - hci_dev_lock(hdev); - __hci_req_update_eir(&req); - } - - __hci_req_update_class(&req); - - err = hci_req_run(&req, set_class_complete); - if (err < 0) { - if (err != -ENODATA) - goto unlock; - - err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_SET_DEV_CLASS, 0, - hdev->dev_class, 3); - goto unlock; - } - - cmd = mgmt_pending_add(sk, MGMT_OP_SET_DEV_CLASS, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_SET_DEV_CLASS, hdev, data, len); if (!cmd) { err = -ENOMEM; goto unlock; } - err = 0; + err = hci_cmd_sync_queue(hdev, set_class_sync, cmd, + mgmt_class_complete); + if (err < 0) + mgmt_pending_free(cmd); unlock: hci_dev_unlock(hdev); @@ -5494,11 +5486,15 @@ done: return err; } +static int set_device_id_sync(struct hci_dev *hdev, void *data) +{ + return hci_update_eir_sync(hdev); +} + static int set_device_id(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_cp_set_device_id *cp = data; - struct hci_request req; int err; __u16 source; @@ -5520,9 +5516,7 @@ static int set_device_id(struct sock *sk, struct hci_dev *hdev, void *data, err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_SET_DEVICE_ID, 0, NULL, 0); - hci_req_init(&req, hdev); - __hci_req_update_eir(&req); - hci_req_run(&req, NULL); + hci_cmd_sync_queue(hdev, set_device_id_sync, NULL, NULL); hci_dev_unlock(hdev); diff --git a/net/bluetooth/mgmt_util.c b/net/bluetooth/mgmt_util.c index 0d0a6d77b9e8..83875f2a0604 100644 --- a/net/bluetooth/mgmt_util.c +++ b/net/bluetooth/mgmt_util.c @@ -227,7 +227,7 @@ void mgmt_pending_foreach(u16 opcode, struct hci_dev *hdev, } } -struct mgmt_pending_cmd *mgmt_pending_add(struct sock *sk, u16 opcode, +struct mgmt_pending_cmd *mgmt_pending_new(struct sock *sk, u16 opcode, struct hci_dev *hdev, void *data, u16 len) { @@ -251,6 +251,19 @@ struct mgmt_pending_cmd *mgmt_pending_add(struct sock *sk, u16 opcode, cmd->sk = sk; sock_hold(sk); + return cmd; +} + +struct mgmt_pending_cmd *mgmt_pending_add(struct sock *sk, u16 opcode, + struct hci_dev *hdev, + void *data, u16 len) +{ + struct mgmt_pending_cmd *cmd; + + cmd = mgmt_pending_new(sk, opcode, hdev, data, len); + if (!cmd) + return NULL; + list_add(&cmd->list, &hdev->mgmt_pending); return cmd; diff --git a/net/bluetooth/mgmt_util.h b/net/bluetooth/mgmt_util.h index 6559f189213c..9dc24ba0d51a 100644 --- a/net/bluetooth/mgmt_util.h +++ b/net/bluetooth/mgmt_util.h @@ -49,5 +49,8 @@ void mgmt_pending_foreach(u16 opcode, struct hci_dev *hdev, struct mgmt_pending_cmd *mgmt_pending_add(struct sock *sk, u16 opcode, struct hci_dev *hdev, void *data, u16 len); +struct mgmt_pending_cmd *mgmt_pending_new(struct sock *sk, u16 opcode, + struct hci_dev *hdev, + void *data, u16 len); void mgmt_pending_free(struct mgmt_pending_cmd *cmd); void mgmt_pending_remove(struct mgmt_pending_cmd *cmd); From cba6b758711cab946c787f7c15be92cc749b8e1f Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:40 -0700 Subject: [PATCH 0033/2048] Bluetooth: hci_sync: Make use of hci_cmd_sync_queue set 2 This make use of hci_cmd_sync_queue for the following MGMT commands: Add Advertising Remove Advertising Add Extended Advertising Parameters Add Extended Advertising Data mgmt-tester -s "Add Advertising" Test Summary ------------ Add Advertising - Failure: LE off Passed Add Advertising - Invalid Params 1 (AD too long) Passed Add Advertising - Invalid Params 2 (Malformed len) Passed Add Advertising - Invalid Params 3 (Malformed len) Passed Add Advertising - Invalid Params 4 (Malformed len) Passed Add Advertising - Invalid Params 5 (AD too long) Passed Add Advertising - Invalid Params 6 (ScRsp too long) Passed Add Advertising - Invalid Params 7 (Malformed len) Passed Add Advertising - Invalid Params 8 (Malformed len) Passed Add Advertising - Invalid Params 9 (Malformed len) Passed Add Advertising - Invalid Params 10 (ScRsp too long) Passed Add Advertising - Rejected (Timeout, !Powered) Passed Add Advertising - Success 1 (Powered, Add Adv Inst) Passed Add Advertising - Success 2 (!Powered, Add Adv Inst) Passed Add Advertising - Success 3 (!Powered, Adv Enable) Passed Add Advertising - Success 4 (Set Adv on override) Passed Add Advertising - Success 5 (Set Adv off override) Passed Add Advertising - Success 6 (Scan Rsp Dta, Adv ok) Passed Add Advertising - Success 7 (Scan Rsp Dta, Scan ok) Passed Add Advertising - Success 8 (Connectable Flag) Passed Add Advertising - Success 9 (General Discov Flag) Passed Add Advertising - Success 10 (Limited Discov Flag) Passed Add Advertising - Success 11 (Managed Flags) Passed Add Advertising - Success 12 (TX Power Flag) Passed Add Advertising - Success 13 (ADV_SCAN_IND) Passed Add Advertising - Success 14 (ADV_NONCONN_IND) Passed Add Advertising - Success 15 (ADV_IND) Passed Add Advertising - Success 16 (Connectable -> on) Passed Add Advertising - Success 17 (Connectable -> off) Passed Add Advertising - Success 18 (Power -> off, Remove) Passed Add Advertising - Success 19 (Power -> off, Keep) Passed Add Advertising - Success 20 (Add Adv override) Passed Add Advertising - Success 21 (Timeout expires) Passed Add Advertising - Success 22 (LE -> off, Remove) Passed Add Advertising - Success (Empty ScRsp) Passed Add Advertising - Success (ScRsp only) Passed Add Advertising - Invalid Params (ScRsp too long) Passed Add Advertising - Success (ScRsp appear) Passed Add Advertising - Invalid Params (ScRsp appear long) Passed Add Advertising - Success (Appear is null) Passed Add Advertising - Success (Name is null) Passed Add Advertising - Success (Complete name) Passed Add Advertising - Success (Shortened name) Passed Add Advertising - Success (Short name) Passed Add Advertising - Success (Name + data) Passed Add Advertising - Invalid Params (Name + data) Passed Add Advertising - Success (Name+data+appear) Passed Total: 47, Passed: 47 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 2.17 seconds mgmt-tester -s "Remove Advertising" Test Summary ------------ Remove Advertising - Invalid Params 1 Passed Remove Advertising - Success 1 Passed Remove Advertising - Success 2 Passed Total: 3, Passed: 3 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.0585 seconds mgmt-tester -s "Ext Adv MGMT Params" Test Summary: ------------ Ext Adv MGMT Params - Unpowered Passed Ext Adv MGMT Params - Invalid parameters Passed Ext Adv MGMT Params - Success Passed Ext Adv MGMT Params - (5.0) Success Passed Total: 4, Passed: 4 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.0746 seconds mgmt-tester -s "Ext Adv MGMT -" Test Summary ------------ Ext Adv MGMT - Data set without Params Passed Ext Adv MGMT - AD Data (5.0) Invalid parameters Passed Ext Adv MGMT - AD Data (5.0) Success Passed Ext Adv MGMT - AD Scan Response (5.0) Success Passed Total: 4, Passed: 4 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.0805 seconds Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/bluetooth.h | 2 + include/net/bluetooth/hci_sync.h | 20 + net/bluetooth/hci_event.c | 57 ++ net/bluetooth/hci_sync.c | 890 ++++++++++++++++++++++++++++++ net/bluetooth/mgmt.c | 395 +++++++------ 5 files changed, 1160 insertions(+), 204 deletions(-) diff --git a/include/net/bluetooth/bluetooth.h b/include/net/bluetooth/bluetooth.h index 3271870fd85e..2f31e571f34c 100644 --- a/include/net/bluetooth/bluetooth.h +++ b/include/net/bluetooth/bluetooth.h @@ -380,6 +380,7 @@ typedef void (*hci_req_complete_skb_t)(struct hci_dev *hdev, u8 status, #define HCI_REQ_SKB BIT(1) struct hci_ctrl { + struct sock *sk; u16 opcode; u8 req_flags; u8 req_event; @@ -405,6 +406,7 @@ struct bt_skb_cb { #define hci_skb_pkt_type(skb) bt_cb((skb))->pkt_type #define hci_skb_expect(skb) bt_cb((skb))->expect #define hci_skb_opcode(skb) bt_cb((skb))->hci.opcode +#define hci_skb_sk(skb) bt_cb((skb))->hci.sk static inline struct sk_buff *bt_skb_alloc(unsigned int len, gfp_t how) { diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index fe77ff97bc8e..143829cd1b29 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -43,3 +43,23 @@ int hci_cmd_sync_queue(struct hci_dev *hdev, hci_cmd_sync_work_func_t func, int hci_update_eir_sync(struct hci_dev *hdev); int hci_update_class_sync(struct hci_dev *hdev); + +int hci_update_eir_sync(struct hci_dev *hdev); +int hci_update_class_sync(struct hci_dev *hdev); + +int hci_update_random_address_sync(struct hci_dev *hdev, bool require_privacy, + bool rpa, u8 *own_addr_type); + +int hci_update_scan_rsp_data_sync(struct hci_dev *hdev, u8 instance); +int hci_update_adv_data_sync(struct hci_dev *hdev, u8 instance); +int hci_schedule_adv_instance_sync(struct hci_dev *hdev, u8 instance, + bool force); + +int hci_setup_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance); +int hci_start_ext_adv_sync(struct hci_dev *hdev, u8 instance); +int hci_enable_ext_advertising_sync(struct hci_dev *hdev, u8 instance); +int hci_enable_advertising_sync(struct hci_dev *hdev); + +int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, + u8 instance, bool force); +int hci_disable_advertising_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 894670419a27..1ac3eea7ac8e 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -1239,6 +1239,55 @@ static void hci_cc_le_set_adv_set_random_addr(struct hci_dev *hdev, hci_dev_unlock(hdev); } +static void hci_cc_le_remove_adv_set(struct hci_dev *hdev, struct sk_buff *skb) +{ + __u8 status = *((__u8 *)skb->data); + u8 *instance; + int err; + + if (status) + return; + + instance = hci_sent_cmd_data(hdev, HCI_OP_LE_REMOVE_ADV_SET); + if (!instance) + return; + + hci_dev_lock(hdev); + + err = hci_remove_adv_instance(hdev, *instance); + if (!err) + mgmt_advertising_removed(hci_skb_sk(hdev->sent_cmd), hdev, + *instance); + + hci_dev_unlock(hdev); +} + +static void hci_cc_le_clear_adv_sets(struct hci_dev *hdev, struct sk_buff *skb) +{ + __u8 status = *((__u8 *)skb->data); + struct adv_info *adv, *n; + int err; + + if (status) + return; + + if (!hci_sent_cmd_data(hdev, HCI_OP_LE_CLEAR_ADV_SETS)) + return; + + hci_dev_lock(hdev); + + list_for_each_entry_safe(adv, n, &hdev->adv_instances, list) { + u8 instance = adv->instance; + + err = hci_remove_adv_instance(hdev, instance); + if (!err) + mgmt_advertising_removed(hci_skb_sk(hdev->sent_cmd), + hdev, instance); + } + + hci_dev_unlock(hdev); +} + static void hci_cc_le_read_transmit_power(struct hci_dev *hdev, struct sk_buff *skb) { @@ -3725,6 +3774,14 @@ static void hci_cmd_complete_evt(struct hci_dev *hdev, struct sk_buff *skb, hci_cc_le_set_adv_set_random_addr(hdev, skb); break; + case HCI_OP_LE_REMOVE_ADV_SET: + hci_cc_le_remove_adv_set(hdev, skb); + break; + + case HCI_OP_LE_CLEAR_ADV_SETS: + hci_cc_le_clear_adv_sets(hdev, skb); + break; + case HCI_OP_LE_READ_TRANSMIT_POWER: hci_cc_le_read_transmit_power(hdev, skb); break; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 8546cbcae10c..6ccd9de87cd6 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -24,6 +24,16 @@ static void hci_cmd_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, hdev->req_result = result; hdev->req_status = HCI_REQ_DONE; + if (skb) { + struct sock *sk = hci_skb_sk(skb); + + /* Drop sk reference if set */ + if (sk) + sock_put(sk); + + hdev->req_skb = skb_get(skb); + } + wake_up_interruptible(&hdev->req_wait_q); } @@ -51,6 +61,14 @@ static struct sk_buff *hci_cmd_sync_alloc(struct hci_dev *hdev, u16 opcode, hci_skb_pkt_type(skb) = HCI_COMMAND_PKT; hci_skb_opcode(skb) = opcode; + /* Grab a reference if command needs to be associated with a sock (e.g. + * likely mgmt socket that initiated the command). + */ + if (sk) { + hci_skb_sk(skb) = sk; + sock_hold(sk); + } + return skb; } @@ -400,3 +418,875 @@ int hci_update_class_sync(struct hci_dev *hdev) return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_CLASS_OF_DEV, sizeof(cod), cod, HCI_CMD_TIMEOUT); } + +static bool is_advertising_allowed(struct hci_dev *hdev, bool connectable) +{ + /* If there is no connection we are OK to advertise. */ + if (hci_conn_num(hdev, LE_LINK) == 0) + return true; + + /* Check le_states if there is any connection in peripheral role. */ + if (hdev->conn_hash.le_num_peripheral > 0) { + /* Peripheral connection state and non connectable mode + * bit 20. + */ + if (!connectable && !(hdev->le_states[2] & 0x10)) + return false; + + /* Peripheral connection state and connectable mode bit 38 + * and scannable bit 21. + */ + if (connectable && (!(hdev->le_states[4] & 0x40) || + !(hdev->le_states[2] & 0x20))) + return false; + } + + /* Check le_states if there is any connection in central role. */ + if (hci_conn_num(hdev, LE_LINK) != hdev->conn_hash.le_num_peripheral) { + /* Central connection state and non connectable mode bit 18. */ + if (!connectable && !(hdev->le_states[2] & 0x02)) + return false; + + /* Central connection state and connectable mode bit 35 and + * scannable 19. + */ + if (connectable && (!(hdev->le_states[4] & 0x08) || + !(hdev->le_states[2] & 0x08))) + return false; + } + + return true; +} + +static bool adv_use_rpa(struct hci_dev *hdev, uint32_t flags) +{ + /* If privacy is not enabled don't use RPA */ + if (!hci_dev_test_flag(hdev, HCI_PRIVACY)) + return false; + + /* If basic privacy mode is enabled use RPA */ + if (!hci_dev_test_flag(hdev, HCI_LIMITED_PRIVACY)) + return true; + + /* If limited privacy mode is enabled don't use RPA if we're + * both discoverable and bondable. + */ + if ((flags & MGMT_ADV_FLAG_DISCOV) && + hci_dev_test_flag(hdev, HCI_BONDABLE)) + return false; + + /* We're neither bondable nor discoverable in the limited + * privacy mode, therefore use RPA. + */ + return true; +} + +static int hci_set_random_addr_sync(struct hci_dev *hdev, bdaddr_t *rpa) +{ + /* If we're advertising or initiating an LE connection we can't + * go ahead and change the random address at this time. This is + * because the eventual initiator address used for the + * subsequently created connection will be undefined (some + * controllers use the new address and others the one we had + * when the operation started). + * + * In this kind of scenario skip the update and let the random + * address be updated at the next cycle. + */ + if (hci_dev_test_flag(hdev, HCI_LE_ADV) || + hci_lookup_le_connect(hdev)) { + bt_dev_dbg(hdev, "Deferring random address update"); + hci_dev_set_flag(hdev, HCI_RPA_EXPIRED); + return 0; + } + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_RANDOM_ADDR, + 6, rpa, HCI_CMD_TIMEOUT); +} + +int hci_update_random_address_sync(struct hci_dev *hdev, bool require_privacy, + bool rpa, u8 *own_addr_type) +{ + int err; + + /* If privacy is enabled use a resolvable private address. If + * current RPA has expired or there is something else than + * the current RPA in use, then generate a new one. + */ + if (rpa) { + /* If Controller supports LL Privacy use own address type is + * 0x03 + */ + if (use_ll_privacy(hdev) && + hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + *own_addr_type = ADDR_LE_DEV_RANDOM_RESOLVED; + else + *own_addr_type = ADDR_LE_DEV_RANDOM; + + /* Check if RPA is valid */ + if (rpa_valid(hdev)) + return 0; + + err = smp_generate_rpa(hdev, hdev->irk, &hdev->rpa); + if (err < 0) { + bt_dev_err(hdev, "failed to generate new RPA"); + return err; + } + + err = hci_set_random_addr_sync(hdev, &hdev->rpa); + if (err) + return err; + + return 0; + } + + /* In case of required privacy without resolvable private address, + * use an non-resolvable private address. This is useful for active + * scanning and non-connectable advertising. + */ + if (require_privacy) { + bdaddr_t nrpa; + + while (true) { + /* The non-resolvable private address is generated + * from random six bytes with the two most significant + * bits cleared. + */ + get_random_bytes(&nrpa, 6); + nrpa.b[5] &= 0x3f; + + /* The non-resolvable private address shall not be + * equal to the public address. + */ + if (bacmp(&hdev->bdaddr, &nrpa)) + break; + } + + *own_addr_type = ADDR_LE_DEV_RANDOM; + + return hci_set_random_addr_sync(hdev, &nrpa); + } + + /* If forcing static address is in use or there is no public + * address use the static address as random address (but skip + * the HCI command if the current random address is already the + * static one. + * + * In case BR/EDR has been disabled on a dual-mode controller + * and a static address has been configured, then use that + * address instead of the public BR/EDR address. + */ + if (hci_dev_test_flag(hdev, HCI_FORCE_STATIC_ADDR) || + !bacmp(&hdev->bdaddr, BDADDR_ANY) || + (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED) && + bacmp(&hdev->static_addr, BDADDR_ANY))) { + *own_addr_type = ADDR_LE_DEV_RANDOM; + if (bacmp(&hdev->static_addr, &hdev->random_addr)) + return hci_set_random_addr_sync(hdev, + &hdev->static_addr); + return 0; + } + + /* Neither privacy nor static address is being used so use a + * public address. + */ + *own_addr_type = ADDR_LE_DEV_PUBLIC; + + return 0; +} + +static int hci_disable_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance) +{ + struct hci_cp_le_set_ext_adv_enable *cp; + struct hci_cp_ext_adv_set *set; + u8 data[sizeof(*cp) + sizeof(*set) * 1]; + u8 size; + + /* If request specifies an instance that doesn't exist, fail */ + if (instance > 0) { + struct adv_info *adv; + + adv = hci_find_adv_instance(hdev, instance); + if (!adv) + return -EINVAL; + + /* If not enabled there is nothing to do */ + if (!adv->enabled) + return 0; + } + + memset(data, 0, sizeof(data)); + + cp = (void *)data; + set = (void *)cp->data; + + /* Instance 0x00 indicates all advertising instances will be disabled */ + cp->num_of_sets = !!instance; + cp->enable = 0x00; + + set->handle = instance; + + size = sizeof(*cp) + sizeof(*set) * cp->num_of_sets; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_ADV_ENABLE, + size, data, HCI_CMD_TIMEOUT); +} + +static int hci_set_adv_set_random_addr_sync(struct hci_dev *hdev, u8 instance, + bdaddr_t *random_addr) +{ + struct hci_cp_le_set_adv_set_rand_addr cp; + int err; + + if (!instance) { + /* Instance 0x00 doesn't have an adv_info, instead it uses + * hdev->random_addr to track its address so whenever it needs + * to be updated this also set the random address since + * hdev->random_addr is shared with scan state machine. + */ + err = hci_set_random_addr_sync(hdev, random_addr); + if (err) + return err; + } + + memset(&cp, 0, sizeof(cp)); + + cp.handle = instance; + bacpy(&cp.bdaddr, random_addr); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_SET_RAND_ADDR, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +int hci_setup_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance) +{ + struct hci_cp_le_set_ext_adv_params cp; + bool connectable; + u32 flags; + bdaddr_t random_addr; + u8 own_addr_type; + int err; + struct adv_info *adv; + bool secondary_adv; + + if (instance > 0) { + adv = hci_find_adv_instance(hdev, instance); + if (!adv) + return -EINVAL; + } else { + adv = NULL; + } + + /* Updating parameters of an active instance will return a + * Command Disallowed error, so we must first disable the + * instance if it is active. + */ + if (adv && !adv->pending) { + err = hci_disable_ext_adv_instance_sync(hdev, instance); + if (err) + return err; + } + + flags = hci_adv_instance_flags(hdev, instance); + + /* If the "connectable" instance flag was not set, then choose between + * ADV_IND and ADV_NONCONN_IND based on the global connectable setting. + */ + connectable = (flags & MGMT_ADV_FLAG_CONNECTABLE) || + mgmt_get_connectable(hdev); + + if (!is_advertising_allowed(hdev, connectable)) + return -EPERM; + + /* Set require_privacy to true only when non-connectable + * advertising is used. In that case it is fine to use a + * non-resolvable private address. + */ + err = hci_get_random_address(hdev, !connectable, + adv_use_rpa(hdev, flags), adv, + &own_addr_type, &random_addr); + if (err < 0) + return err; + + memset(&cp, 0, sizeof(cp)); + + if (adv) { + hci_cpu_to_le24(adv->min_interval, cp.min_interval); + hci_cpu_to_le24(adv->max_interval, cp.max_interval); + cp.tx_power = adv->tx_power; + } else { + hci_cpu_to_le24(hdev->le_adv_min_interval, cp.min_interval); + hci_cpu_to_le24(hdev->le_adv_max_interval, cp.max_interval); + cp.tx_power = HCI_ADV_TX_POWER_NO_PREFERENCE; + } + + secondary_adv = (flags & MGMT_ADV_FLAG_SEC_MASK); + + if (connectable) { + if (secondary_adv) + cp.evt_properties = cpu_to_le16(LE_EXT_ADV_CONN_IND); + else + cp.evt_properties = cpu_to_le16(LE_LEGACY_ADV_IND); + } else if (hci_adv_instance_is_scannable(hdev, instance) || + (flags & MGMT_ADV_PARAM_SCAN_RSP)) { + if (secondary_adv) + cp.evt_properties = cpu_to_le16(LE_EXT_ADV_SCAN_IND); + else + cp.evt_properties = cpu_to_le16(LE_LEGACY_ADV_SCAN_IND); + } else { + if (secondary_adv) + cp.evt_properties = cpu_to_le16(LE_EXT_ADV_NON_CONN_IND); + else + cp.evt_properties = cpu_to_le16(LE_LEGACY_NONCONN_IND); + } + + cp.own_addr_type = own_addr_type; + cp.channel_map = hdev->le_adv_channel_map; + cp.handle = instance; + + if (flags & MGMT_ADV_FLAG_SEC_2M) { + cp.primary_phy = HCI_ADV_PHY_1M; + cp.secondary_phy = HCI_ADV_PHY_2M; + } else if (flags & MGMT_ADV_FLAG_SEC_CODED) { + cp.primary_phy = HCI_ADV_PHY_CODED; + cp.secondary_phy = HCI_ADV_PHY_CODED; + } else { + /* In all other cases use 1M */ + cp.primary_phy = HCI_ADV_PHY_1M; + cp.secondary_phy = HCI_ADV_PHY_1M; + } + + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_ADV_PARAMS, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (err) + return err; + + if ((own_addr_type == ADDR_LE_DEV_RANDOM || + own_addr_type == ADDR_LE_DEV_RANDOM_RESOLVED) && + bacmp(&random_addr, BDADDR_ANY)) { + /* Check if random address need to be updated */ + if (adv) { + if (!bacmp(&random_addr, &adv->random_addr)) + return 0; + } else { + if (!bacmp(&random_addr, &hdev->random_addr)) + return 0; + } + + return hci_set_adv_set_random_addr_sync(hdev, instance, + &random_addr); + } + + return 0; +} + +static int hci_set_ext_scan_rsp_data_sync(struct hci_dev *hdev, u8 instance) +{ + struct { + struct hci_cp_le_set_ext_scan_rsp_data cp; + u8 data[HCI_MAX_EXT_AD_LENGTH]; + } pdu; + u8 len; + + memset(&pdu, 0, sizeof(pdu)); + + len = eir_create_scan_rsp(hdev, instance, pdu.data); + + if (hdev->scan_rsp_data_len == len && + !memcmp(pdu.data, hdev->scan_rsp_data, len)) + return 0; + + memcpy(hdev->scan_rsp_data, pdu.data, len); + hdev->scan_rsp_data_len = len; + + pdu.cp.handle = instance; + pdu.cp.length = len; + pdu.cp.operation = LE_SET_ADV_DATA_OP_COMPLETE; + pdu.cp.frag_pref = LE_SET_ADV_DATA_NO_FRAG; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_SCAN_RSP_DATA, + sizeof(pdu.cp) + len, &pdu.cp, + HCI_CMD_TIMEOUT); +} + +static int __hci_set_scan_rsp_data_sync(struct hci_dev *hdev, u8 instance) +{ + struct hci_cp_le_set_scan_rsp_data cp; + u8 len; + + memset(&cp, 0, sizeof(cp)); + + len = eir_create_scan_rsp(hdev, instance, cp.data); + + if (hdev->scan_rsp_data_len == len && + !memcmp(cp.data, hdev->scan_rsp_data, len)) + return 0; + + memcpy(hdev->scan_rsp_data, cp.data, sizeof(cp.data)); + hdev->scan_rsp_data_len = len; + + cp.length = len; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_SCAN_RSP_DATA, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +int hci_update_scan_rsp_data_sync(struct hci_dev *hdev, u8 instance) +{ + if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED)) + return 0; + + if (ext_adv_capable(hdev)) + return hci_set_ext_scan_rsp_data_sync(hdev, instance); + + return __hci_set_scan_rsp_data_sync(hdev, instance); +} + +int hci_enable_ext_advertising_sync(struct hci_dev *hdev, u8 instance) +{ + struct hci_cp_le_set_ext_adv_enable *cp; + struct hci_cp_ext_adv_set *set; + u8 data[sizeof(*cp) + sizeof(*set) * 1]; + struct adv_info *adv; + + if (instance > 0) { + adv = hci_find_adv_instance(hdev, instance); + if (!adv) + return -EINVAL; + /* If already enabled there is nothing to do */ + if (adv->enabled) + return 0; + } else { + adv = NULL; + } + + cp = (void *)data; + set = (void *)cp->data; + + memset(cp, 0, sizeof(*cp)); + + cp->enable = 0x01; + cp->num_of_sets = 0x01; + + memset(set, 0, sizeof(*set)); + + set->handle = instance; + + /* Set duration per instance since controller is responsible for + * scheduling it. + */ + if (adv && adv->duration) { + u16 duration = adv->timeout * MSEC_PER_SEC; + + /* Time = N * 10 ms */ + set->duration = cpu_to_le16(duration / 10); + } + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_ADV_ENABLE, + sizeof(*cp) + + sizeof(*set) * cp->num_of_sets, + data, HCI_CMD_TIMEOUT); +} + +int hci_start_ext_adv_sync(struct hci_dev *hdev, u8 instance) +{ + int err; + + err = hci_setup_ext_adv_instance_sync(hdev, instance); + if (err) + return err; + + err = hci_set_ext_scan_rsp_data_sync(hdev, instance); + if (err) + return err; + + return hci_enable_ext_advertising_sync(hdev, instance); +} + +static int hci_start_adv_sync(struct hci_dev *hdev, u8 instance) +{ + int err; + + if (ext_adv_capable(hdev)) + return hci_start_ext_adv_sync(hdev, instance); + + err = hci_update_adv_data_sync(hdev, instance); + if (err) + return err; + + err = hci_update_scan_rsp_data_sync(hdev, instance); + if (err) + return err; + + return hci_enable_advertising_sync(hdev); +} + +int hci_enable_advertising_sync(struct hci_dev *hdev) +{ + struct adv_info *adv_instance; + struct hci_cp_le_set_adv_param cp; + u8 own_addr_type, enable = 0x01; + bool connectable; + u16 adv_min_interval, adv_max_interval; + u32 flags; + u8 status; + + flags = hci_adv_instance_flags(hdev, hdev->cur_adv_instance); + adv_instance = hci_find_adv_instance(hdev, hdev->cur_adv_instance); + + /* If the "connectable" instance flag was not set, then choose between + * ADV_IND and ADV_NONCONN_IND based on the global connectable setting. + */ + connectable = (flags & MGMT_ADV_FLAG_CONNECTABLE) || + mgmt_get_connectable(hdev); + + if (!is_advertising_allowed(hdev, connectable)) + return -EINVAL; + + if (hci_dev_test_flag(hdev, HCI_LE_ADV)) { + status = hci_disable_advertising_sync(hdev); + if (status) + return status; + } + + /* Clear the HCI_LE_ADV bit temporarily so that the + * hci_update_random_address knows that it's safe to go ahead + * and write a new random address. The flag will be set back on + * as soon as the SET_ADV_ENABLE HCI command completes. + */ + hci_dev_clear_flag(hdev, HCI_LE_ADV); + + /* Set require_privacy to true only when non-connectable + * advertising is used. In that case it is fine to use a + * non-resolvable private address. + */ + status = hci_update_random_address_sync(hdev, !connectable, + adv_use_rpa(hdev, flags), + &own_addr_type); + if (status) + return status; + + memset(&cp, 0, sizeof(cp)); + + if (adv_instance) { + adv_min_interval = adv_instance->min_interval; + adv_max_interval = adv_instance->max_interval; + } else { + adv_min_interval = hdev->le_adv_min_interval; + adv_max_interval = hdev->le_adv_max_interval; + } + + if (connectable) { + cp.type = LE_ADV_IND; + } else { + if (hci_adv_instance_is_scannable(hdev, hdev->cur_adv_instance)) + cp.type = LE_ADV_SCAN_IND; + else + cp.type = LE_ADV_NONCONN_IND; + + if (!hci_dev_test_flag(hdev, HCI_DISCOVERABLE) || + hci_dev_test_flag(hdev, HCI_LIMITED_DISCOVERABLE)) { + adv_min_interval = DISCOV_LE_FAST_ADV_INT_MIN; + adv_max_interval = DISCOV_LE_FAST_ADV_INT_MAX; + } + } + + cp.min_interval = cpu_to_le16(adv_min_interval); + cp.max_interval = cpu_to_le16(adv_max_interval); + cp.own_address_type = own_addr_type; + cp.channel_map = hdev->le_adv_channel_map; + + status = __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_PARAM, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (status) + return status; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_ENABLE, + sizeof(enable), &enable, HCI_CMD_TIMEOUT); +} + +static int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, + struct sock *sk) +{ + int err; + + if (!ext_adv_capable(hdev)) + return 0; + + err = hci_disable_ext_adv_instance_sync(hdev, instance); + if (err) + return err; + + /* If request specifies an instance that doesn't exist, fail */ + if (instance > 0 && !hci_find_adv_instance(hdev, instance)) + return -EINVAL; + + return __hci_cmd_sync_status_sk(hdev, HCI_OP_LE_REMOVE_ADV_SET, + sizeof(instance), &instance, 0, + HCI_CMD_TIMEOUT, sk); +} + +static void cancel_adv_timeout(struct hci_dev *hdev) +{ + if (hdev->adv_instance_timeout) { + hdev->adv_instance_timeout = 0; + cancel_delayed_work(&hdev->adv_instance_expire); + } +} + +static int hci_set_ext_adv_data_sync(struct hci_dev *hdev, u8 instance) +{ + struct { + struct hci_cp_le_set_ext_adv_data cp; + u8 data[HCI_MAX_EXT_AD_LENGTH]; + } pdu; + u8 len; + + memset(&pdu, 0, sizeof(pdu)); + + len = eir_create_adv_data(hdev, instance, pdu.data); + + /* There's nothing to do if the data hasn't changed */ + if (hdev->adv_data_len == len && + memcmp(pdu.data, hdev->adv_data, len) == 0) + return 0; + + memcpy(hdev->adv_data, pdu.data, len); + hdev->adv_data_len = len; + + pdu.cp.length = len; + pdu.cp.handle = instance; + pdu.cp.operation = LE_SET_ADV_DATA_OP_COMPLETE; + pdu.cp.frag_pref = LE_SET_ADV_DATA_NO_FRAG; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_ADV_DATA, + sizeof(pdu.cp) + len, &pdu.cp, + HCI_CMD_TIMEOUT); +} + +static int hci_set_adv_data_sync(struct hci_dev *hdev, u8 instance) +{ + struct hci_cp_le_set_adv_data cp; + u8 len; + + memset(&cp, 0, sizeof(cp)); + + len = eir_create_adv_data(hdev, instance, cp.data); + + /* There's nothing to do if the data hasn't changed */ + if (hdev->adv_data_len == len && + memcmp(cp.data, hdev->adv_data, len) == 0) + return 0; + + memcpy(hdev->adv_data, cp.data, sizeof(cp.data)); + hdev->adv_data_len = len; + + cp.length = len; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_DATA, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +int hci_update_adv_data_sync(struct hci_dev *hdev, u8 instance) +{ + if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED)) + return 0; + + if (ext_adv_capable(hdev)) + return hci_set_ext_adv_data_sync(hdev, instance); + + return hci_set_adv_data_sync(hdev, instance); +} + +int hci_schedule_adv_instance_sync(struct hci_dev *hdev, u8 instance, + bool force) +{ + struct adv_info *adv = NULL; + u16 timeout; + + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || + list_empty(&hdev->adv_instances)) + return -EPERM; + + if (hdev->adv_instance_timeout) + return -EBUSY; + + adv = hci_find_adv_instance(hdev, instance); + if (!adv) + return -ENOENT; + + /* A zero timeout means unlimited advertising. As long as there is + * only one instance, duration should be ignored. We still set a timeout + * in case further instances are being added later on. + * + * If the remaining lifetime of the instance is more than the duration + * then the timeout corresponds to the duration, otherwise it will be + * reduced to the remaining instance lifetime. + */ + if (adv->timeout == 0 || adv->duration <= adv->remaining_time) + timeout = adv->duration; + else + timeout = adv->remaining_time; + + /* The remaining time is being reduced unless the instance is being + * advertised without time limit. + */ + if (adv->timeout) + adv->remaining_time = adv->remaining_time - timeout; + + /* Only use work for scheduling instances with legacy advertising */ + if (!ext_adv_capable(hdev)) { + hdev->adv_instance_timeout = timeout; + queue_delayed_work(hdev->req_workqueue, + &hdev->adv_instance_expire, + msecs_to_jiffies(timeout * 1000)); + } + + /* If we're just re-scheduling the same instance again then do not + * execute any HCI commands. This happens when a single instance is + * being advertised. + */ + if (!force && hdev->cur_adv_instance == instance && + hci_dev_test_flag(hdev, HCI_LE_ADV)) + return 0; + + hdev->cur_adv_instance = instance; + + return hci_start_adv_sync(hdev, instance); +} + +static int hci_clear_adv_sets_sync(struct hci_dev *hdev, struct sock *sk) +{ + int err; + + if (!ext_adv_capable(hdev)) + return 0; + + /* Disable instance 0x00 to disable all instances */ + err = hci_disable_ext_adv_instance_sync(hdev, 0x00); + if (err) + return err; + + return __hci_cmd_sync_status_sk(hdev, HCI_OP_LE_CLEAR_ADV_SETS, + 0, NULL, 0, HCI_CMD_TIMEOUT, sk); +} + +static int hci_clear_adv_sync(struct hci_dev *hdev, struct sock *sk, bool force) +{ + struct adv_info *adv, *n; + + if (ext_adv_capable(hdev)) + /* Remove all existing sets */ + return hci_clear_adv_sets_sync(hdev, sk); + + /* This is safe as long as there is no command send while the lock is + * held. + */ + hci_dev_lock(hdev); + + /* Cleanup non-ext instances */ + list_for_each_entry_safe(adv, n, &hdev->adv_instances, list) { + u8 instance = adv->instance; + int err; + + if (!(force || adv->timeout)) + continue; + + err = hci_remove_adv_instance(hdev, instance); + if (!err) + mgmt_advertising_removed(sk, hdev, instance); + } + + hci_dev_unlock(hdev); + + return 0; +} + +static int hci_remove_adv_sync(struct hci_dev *hdev, u8 instance, + struct sock *sk) +{ + int err; + + /* If we use extended advertising, instance has to be removed first. */ + if (ext_adv_capable(hdev)) + return hci_remove_ext_adv_instance_sync(hdev, instance, sk); + + /* This is safe as long as there is no command send while the lock is + * held. + */ + hci_dev_lock(hdev); + + err = hci_remove_adv_instance(hdev, instance); + if (!err) + mgmt_advertising_removed(sk, hdev, instance); + + hci_dev_unlock(hdev); + + return err; +} + +/* For a single instance: + * - force == true: The instance will be removed even when its remaining + * lifetime is not zero. + * - force == false: the instance will be deactivated but kept stored unless + * the remaining lifetime is zero. + * + * For instance == 0x00: + * - force == true: All instances will be removed regardless of their timeout + * setting. + * - force == false: Only instances that have a timeout will be removed. + */ +int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, + u8 instance, bool force) +{ + struct adv_info *next = NULL; + int err; + + /* Cancel any timeout concerning the removed instance(s). */ + if (!instance || hdev->cur_adv_instance == instance) + cancel_adv_timeout(hdev); + + /* Get the next instance to advertise BEFORE we remove + * the current one. This can be the same instance again + * if there is only one instance. + */ + if (hdev->cur_adv_instance == instance) + next = hci_get_next_instance(hdev, instance); + + if (!instance) { + err = hci_clear_adv_sync(hdev, sk, force); + if (err) + return err; + } else { + struct adv_info *adv = hci_find_adv_instance(hdev, instance); + + if (force || (adv && adv->timeout && !adv->remaining_time)) { + /* Don't advertise a removed instance. */ + if (next && next->instance == instance) + next = NULL; + + err = hci_remove_adv_sync(hdev, instance, sk); + if (err) + return err; + } + } + + if (!hdev_is_powered(hdev) || hci_dev_test_flag(hdev, HCI_ADVERTISING)) + return 0; + + if (next && !ext_adv_capable(hdev)) + hci_schedule_adv_instance_sync(hdev, next->instance, false); + + return 0; +} + +int hci_disable_advertising_sync(struct hci_dev *hdev) +{ + u8 enable = 0x00; + + if (ext_adv_capable(hdev)) + return hci_disable_ext_adv_instance_sync(hdev, 0x00); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_ENABLE, + sizeof(enable), &enable, HCI_CMD_TIMEOUT); +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ec6aa7af482f..6e481ca86303 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -999,11 +999,22 @@ static void service_cache_off(struct work_struct *work) hci_cmd_sync_queue(hdev, service_cache_sync, NULL, NULL); } +static int rpa_expired_sync(struct hci_dev *hdev, void *data) +{ + /* The generation of a new RPA and programming it into the + * controller happens in the hci_req_enable_advertising() + * function. + */ + if (ext_adv_capable(hdev)) + return hci_start_ext_adv_sync(hdev, hdev->cur_adv_instance); + else + return hci_enable_advertising_sync(hdev); +} + static void rpa_expired(struct work_struct *work) { struct hci_dev *hdev = container_of(work, struct hci_dev, rpa_expired.work); - struct hci_request req; bt_dev_dbg(hdev, ""); @@ -1012,16 +1023,7 @@ static void rpa_expired(struct work_struct *work) if (!hci_dev_test_flag(hdev, HCI_ADVERTISING)) return; - /* The generation of a new RPA and programming it into the - * controller happens in the hci_req_enable_advertising() - * function. - */ - hci_req_init(&req, hdev); - if (ext_adv_capable(hdev)) - __hci_req_start_ext_adv(&req, hdev->cur_adv_instance); - else - __hci_req_enable_advertising(&req); - hci_req_run(&req, NULL); + hci_cmd_sync_queue(hdev, rpa_expired_sync, NULL, NULL); } static void mgmt_init_hdev(struct sock *sk, struct hci_dev *hdev) @@ -7873,58 +7875,66 @@ static bool adv_busy(struct hci_dev *hdev) pending_find(MGMT_OP_ADD_EXT_ADV_DATA, hdev)); } -static void add_advertising_complete(struct hci_dev *hdev, u8 status, - u16 opcode) +static void add_adv_complete(struct hci_dev *hdev, struct sock *sk, u8 instance, + int err) { - struct mgmt_pending_cmd *cmd; - struct mgmt_cp_add_advertising *cp; - struct mgmt_rp_add_advertising rp; - struct adv_info *adv_instance, *n; - u8 instance; + struct adv_info *adv, *n; - bt_dev_dbg(hdev, "status %u", status); + bt_dev_dbg(hdev, "err %d", err); hci_dev_lock(hdev); - cmd = pending_find(MGMT_OP_ADD_ADVERTISING, hdev); - if (!cmd) - cmd = pending_find(MGMT_OP_ADD_EXT_ADV_DATA, hdev); + list_for_each_entry_safe(adv, n, &hdev->adv_instances, list) { + u8 instance; - list_for_each_entry_safe(adv_instance, n, &hdev->adv_instances, list) { - if (!adv_instance->pending) + if (!adv->pending) continue; - if (!status) { - adv_instance->pending = false; + if (!err) { + adv->pending = false; continue; } - instance = adv_instance->instance; + instance = adv->instance; if (hdev->cur_adv_instance == instance) cancel_adv_timeout(hdev); hci_remove_adv_instance(hdev, instance); - mgmt_advertising_removed(cmd ? cmd->sk : NULL, hdev, instance); + mgmt_advertising_removed(sk, hdev, instance); } - if (!cmd) - goto unlock; + hci_dev_unlock(hdev); +} + +static void add_advertising_complete(struct hci_dev *hdev, void *data, int err) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_advertising *cp = cmd->param; + struct mgmt_rp_add_advertising rp; + + memset(&rp, 0, sizeof(rp)); - cp = cmd->param; rp.instance = cp->instance; - if (status) + if (err) mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status)); + mgmt_status(err)); else mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status), &rp, sizeof(rp)); + mgmt_status(err), &rp, sizeof(rp)); - mgmt_pending_remove(cmd); + add_adv_complete(hdev, cmd->sk, cp->instance, err); -unlock: - hci_dev_unlock(hdev); + mgmt_pending_free(cmd); +} + +static int add_advertising_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_advertising *cp = cmd->param; + + return hci_schedule_adv_instance_sync(hdev, cp->instance, true); } static int add_advertising(struct sock *sk, struct hci_dev *hdev, @@ -7940,7 +7950,6 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, struct adv_info *next_instance; int err; struct mgmt_pending_cmd *cmd; - struct hci_request req; bt_dev_dbg(hdev, "sock %p", sk); @@ -8048,25 +8057,19 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, /* We're good to go, update advertising data, parameters, and start * advertising. */ - cmd = mgmt_pending_add(sk, MGMT_OP_ADD_ADVERTISING, hdev, data, + cmd = mgmt_pending_new(sk, MGMT_OP_ADD_ADVERTISING, hdev, data, data_len); if (!cmd) { err = -ENOMEM; goto unlock; } - hci_req_init(&req, hdev); + cp->instance = schedule_instance; - err = __hci_req_schedule_adv_instance(&req, schedule_instance, true); - - if (!err) - err = hci_req_run(&req, add_advertising_complete); - - if (err < 0) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, - MGMT_STATUS_FAILED); - mgmt_pending_remove(cmd); - } + err = hci_cmd_sync_queue(hdev, add_advertising_sync, cmd, + add_advertising_complete); + if (err < 0) + mgmt_pending_free(cmd); unlock: hci_dev_unlock(hdev); @@ -8074,30 +8077,25 @@ unlock: return err; } -static void add_ext_adv_params_complete(struct hci_dev *hdev, u8 status, - u16 opcode) +static void add_ext_adv_params_complete(struct hci_dev *hdev, void *data, + int err) { - struct mgmt_pending_cmd *cmd; - struct mgmt_cp_add_ext_adv_params *cp; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_ext_adv_params *cp = cmd->param; struct mgmt_rp_add_ext_adv_params rp; - struct adv_info *adv_instance; + struct adv_info *adv; u32 flags; BT_DBG("%s", hdev->name); hci_dev_lock(hdev); - cmd = pending_find(MGMT_OP_ADD_EXT_ADV_PARAMS, hdev); - if (!cmd) - goto unlock; - - cp = cmd->param; - adv_instance = hci_find_adv_instance(hdev, cp->instance); - if (!adv_instance) + adv = hci_find_adv_instance(hdev, cp->instance); + if (!adv) goto unlock; rp.instance = cp->instance; - rp.tx_power = adv_instance->tx_power; + rp.tx_power = adv->tx_power; /* While we're at it, inform userspace of the available space for this * advertisement, given the flags that will be used. @@ -8106,39 +8104,44 @@ static void add_ext_adv_params_complete(struct hci_dev *hdev, u8 status, rp.max_adv_data_len = tlv_data_max_len(hdev, flags, true); rp.max_scan_rsp_len = tlv_data_max_len(hdev, flags, false); - if (status) { + if (err) { /* If this advertisement was previously advertising and we * failed to update it, we signal that it has been removed and * delete its structure */ - if (!adv_instance->pending) + if (!adv->pending) mgmt_advertising_removed(cmd->sk, hdev, cp->instance); hci_remove_adv_instance(hdev, cp->instance); mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status)); - + mgmt_status(err)); } else { mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status), &rp, sizeof(rp)); + mgmt_status(err), &rp, sizeof(rp)); } unlock: if (cmd) - mgmt_pending_remove(cmd); + mgmt_pending_free(cmd); hci_dev_unlock(hdev); } +static int add_ext_adv_params_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_ext_adv_params *cp = cmd->param; + + return hci_setup_ext_adv_instance_sync(hdev, cp->instance); +} + static int add_ext_adv_params(struct sock *sk, struct hci_dev *hdev, void *data, u16 data_len) { struct mgmt_cp_add_ext_adv_params *cp = data; struct mgmt_rp_add_ext_adv_params rp; struct mgmt_pending_cmd *cmd = NULL; - struct adv_info *adv_instance; - struct hci_request req; u32 flags, min_interval, max_interval; u16 timeout, duration; u8 status; @@ -8220,29 +8223,18 @@ static int add_ext_adv_params(struct sock *sk, struct hci_dev *hdev, /* Submit request for advertising params if ext adv available */ if (ext_adv_capable(hdev)) { - hci_req_init(&req, hdev); - adv_instance = hci_find_adv_instance(hdev, cp->instance); - - /* Updating parameters of an active instance will return a - * Command Disallowed error, so we must first disable the - * instance if it is active. - */ - if (!adv_instance->pending) - __hci_req_disable_ext_adv_instance(&req, cp->instance); - - __hci_req_setup_ext_adv_instance(&req, cp->instance); - - err = hci_req_run(&req, add_ext_adv_params_complete); - - if (!err) - cmd = mgmt_pending_add(sk, MGMT_OP_ADD_EXT_ADV_PARAMS, - hdev, data, data_len); + cmd = mgmt_pending_new(sk, MGMT_OP_ADD_EXT_ADV_PARAMS, hdev, + data, data_len); if (!cmd) { err = -ENOMEM; hci_remove_adv_instance(hdev, cp->instance); goto unlock; } + err = hci_cmd_sync_queue(hdev, add_ext_adv_params_sync, cmd, + add_ext_adv_params_complete); + if (err < 0) + mgmt_pending_free(cmd); } else { rp.instance = cp->instance; rp.tx_power = HCI_ADV_TX_POWER_NO_PREFERENCE; @@ -8259,6 +8251,49 @@ unlock: return err; } +static void add_ext_adv_data_complete(struct hci_dev *hdev, void *data, int err) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_ext_adv_data *cp = cmd->param; + struct mgmt_rp_add_advertising rp; + + add_adv_complete(hdev, cmd->sk, cp->instance, err); + + memset(&rp, 0, sizeof(rp)); + + rp.instance = cp->instance; + + if (err) + mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, + mgmt_status(err)); + else + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, + mgmt_status(err), &rp, sizeof(rp)); + + mgmt_pending_free(cmd); +} + +static int add_ext_adv_data_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_add_ext_adv_data *cp = cmd->param; + int err; + + if (ext_adv_capable(hdev)) { + err = hci_update_adv_data_sync(hdev, cp->instance); + if (err) + return err; + + err = hci_update_scan_rsp_data_sync(hdev, cp->instance); + if (err) + return err; + + return hci_enable_ext_advertising_sync(hdev, cp->instance); + } + + return hci_schedule_adv_instance_sync(hdev, cp->instance, true); +} + static int add_ext_adv_data(struct sock *sk, struct hci_dev *hdev, void *data, u16 data_len) { @@ -8269,7 +8304,6 @@ static int add_ext_adv_data(struct sock *sk, struct hci_dev *hdev, void *data, struct adv_info *adv_instance; int err = 0; struct mgmt_pending_cmd *cmd; - struct hci_request req; BT_DBG("%s", hdev->name); @@ -8311,78 +8345,52 @@ static int add_ext_adv_data(struct sock *sk, struct hci_dev *hdev, void *data, cp->data, cp->scan_rsp_len, cp->data + cp->adv_data_len); - /* We're good to go, update advertising data, parameters, and start - * advertising. - */ - - hci_req_init(&req, hdev); - - hci_req_add(&req, HCI_OP_READ_LOCAL_NAME, 0, NULL); - - if (ext_adv_capable(hdev)) { - __hci_req_update_adv_data(&req, cp->instance); - __hci_req_update_scan_rsp_data(&req, cp->instance); - __hci_req_enable_ext_advertising(&req, cp->instance); - - } else { - /* If using software rotation, determine next instance to use */ - - if (hdev->cur_adv_instance == cp->instance) { - /* If the currently advertised instance is being changed - * then cancel the current advertising and schedule the - * next instance. If there is only one instance then the - * overridden advertising data will be visible right - * away - */ - cancel_adv_timeout(hdev); - - next_instance = hci_get_next_instance(hdev, - cp->instance); - if (next_instance) - schedule_instance = next_instance->instance; - } else if (!hdev->adv_instance_timeout) { - /* Immediately advertise the new instance if no other - * instance is currently being advertised. - */ - schedule_instance = cp->instance; - } - - /* If the HCI_ADVERTISING flag is set or there is no instance to - * be advertised then we have no HCI communication to make. - * Simply return. + /* If using software rotation, determine next instance to use */ + if (hdev->cur_adv_instance == cp->instance) { + /* If the currently advertised instance is being changed + * then cancel the current advertising and schedule the + * next instance. If there is only one instance then the + * overridden advertising data will be visible right + * away */ - if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || - !schedule_instance) { - if (adv_instance->pending) { - mgmt_advertising_added(sk, hdev, cp->instance); - adv_instance->pending = false; - } - rp.instance = cp->instance; - err = mgmt_cmd_complete(sk, hdev->id, - MGMT_OP_ADD_EXT_ADV_DATA, - MGMT_STATUS_SUCCESS, &rp, - sizeof(rp)); - goto unlock; - } + cancel_adv_timeout(hdev); - err = __hci_req_schedule_adv_instance(&req, schedule_instance, - true); + next_instance = hci_get_next_instance(hdev, cp->instance); + if (next_instance) + schedule_instance = next_instance->instance; + } else if (!hdev->adv_instance_timeout) { + /* Immediately advertise the new instance if no other + * instance is currently being advertised. + */ + schedule_instance = cp->instance; } - cmd = mgmt_pending_add(sk, MGMT_OP_ADD_EXT_ADV_DATA, hdev, data, + /* If the HCI_ADVERTISING flag is set or there is no instance to + * be advertised then we have no HCI communication to make. + * Simply return. + */ + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || !schedule_instance) { + if (adv_instance->pending) { + mgmt_advertising_added(sk, hdev, cp->instance); + adv_instance->pending = false; + } + rp.instance = cp->instance; + err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, + MGMT_STATUS_SUCCESS, &rp, sizeof(rp)); + goto unlock; + } + + cmd = mgmt_pending_new(sk, MGMT_OP_ADD_EXT_ADV_DATA, hdev, data, data_len); if (!cmd) { err = -ENOMEM; goto clear_new_instance; } - if (!err) - err = hci_req_run(&req, add_advertising_complete); - + err = hci_cmd_sync_queue(hdev, add_ext_adv_data_sync, cmd, + add_ext_adv_data_complete); if (err < 0) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_EXT_ADV_DATA, - MGMT_STATUS_FAILED); - mgmt_pending_remove(cmd); + mgmt_pending_free(cmd); goto clear_new_instance; } @@ -8405,43 +8413,49 @@ unlock: return err; } -static void remove_advertising_complete(struct hci_dev *hdev, u8 status, - u16 opcode) +static void remove_advertising_complete(struct hci_dev *hdev, void *data, + int err) { - struct mgmt_pending_cmd *cmd; - struct mgmt_cp_remove_advertising *cp; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_remove_advertising *cp = cmd->param; struct mgmt_rp_remove_advertising rp; - bt_dev_dbg(hdev, "status %u", status); + bt_dev_dbg(hdev, "err %d", err); - hci_dev_lock(hdev); - - /* A failure status here only means that we failed to disable - * advertising. Otherwise, the advertising instance has been removed, - * so report success. - */ - cmd = pending_find(MGMT_OP_REMOVE_ADVERTISING, hdev); - if (!cmd) - goto unlock; - - cp = cmd->param; + memset(&rp, 0, sizeof(rp)); rp.instance = cp->instance; - mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, MGMT_STATUS_SUCCESS, - &rp, sizeof(rp)); - mgmt_pending_remove(cmd); + if (err) + mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, + mgmt_status(err)); + else + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, + MGMT_STATUS_SUCCESS, &rp, sizeof(rp)); -unlock: - hci_dev_unlock(hdev); + mgmt_pending_free(cmd); +} + +static int remove_advertising_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_remove_advertising *cp = cmd->param; + int err; + + err = hci_remove_advertising_sync(hdev, cmd->sk, cp->instance, true); + if (err) + return err; + + if (list_empty(&hdev->adv_instances)) + err = hci_disable_advertising_sync(hdev); + + return err; } static int remove_advertising(struct sock *sk, struct hci_dev *hdev, void *data, u16 data_len) { struct mgmt_cp_remove_advertising *cp = data; - struct mgmt_rp_remove_advertising rp; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); @@ -8476,44 +8490,17 @@ static int remove_advertising(struct sock *sk, struct hci_dev *hdev, goto unlock; } - hci_req_init(&req, hdev); - - /* If we use extended advertising, instance is disabled and removed */ - if (ext_adv_capable(hdev)) { - __hci_req_disable_ext_adv_instance(&req, cp->instance); - __hci_req_remove_ext_adv_instance(&req, cp->instance); - } - - hci_req_clear_adv_instance(hdev, sk, &req, cp->instance, true); - - if (list_empty(&hdev->adv_instances)) - __hci_req_disable_advertising(&req); - - /* If no HCI commands have been collected so far or the HCI_ADVERTISING - * flag is set or the device isn't powered then we have no HCI - * communication to make. Simply return. - */ - if (skb_queue_empty(&req.cmd_q) || - !hdev_is_powered(hdev) || - hci_dev_test_flag(hdev, HCI_ADVERTISING)) { - hci_req_purge(&req); - rp.instance = cp->instance; - err = mgmt_cmd_complete(sk, hdev->id, - MGMT_OP_REMOVE_ADVERTISING, - MGMT_STATUS_SUCCESS, &rp, sizeof(rp)); - goto unlock; - } - - cmd = mgmt_pending_add(sk, MGMT_OP_REMOVE_ADVERTISING, hdev, data, + cmd = mgmt_pending_new(sk, MGMT_OP_REMOVE_ADVERTISING, hdev, data, data_len); if (!cmd) { err = -ENOMEM; goto unlock; } - err = hci_req_run(&req, remove_advertising_complete); + err = hci_cmd_sync_queue(hdev, remove_advertising_sync, cmd, + remove_advertising_complete); if (err < 0) - mgmt_pending_remove(cmd); + mgmt_pending_free(cmd); unlock: hci_dev_unlock(hdev); From e8907f76544ffe225ab95d70f7313267b1d0c76d Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:41 -0700 Subject: [PATCH 0034/2048] Bluetooth: hci_sync: Make use of hci_cmd_sync_queue set 3 This make use of hci_cmd_sync_queue for the following MGMT commands: Add Device Remove Device Tested with: mgmt-tester -s "Add Device" Test Summary ------------ Add Device - Invalid Params 1 Passed Add Device - Invalid Params 2 Passed Add Device - Invalid Params 3 Passed Add Device - Invalid Params 4 Passed Add Device - Success 1 Passed Add Device - Success 2 Passed Add Device - Success 3 Passed Add Device - Success 4 Passed Add Device - Success 5 Passed Add Device - Success 6 - Add to whitelist Passed Add Device - Success 7 - Add to resolv list Passed Add Device - Success 8 - Enable resolv list Passed Total: 12, Passed: 12 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 0.209 seconds mgmt-tester -s "Remove Device" Test Summary ------------ Remove Device - Invalid Params 1 Passed Remove Device - Invalid Params 2 Passed Remove Device - Invalid Params 3 Passed Remove Device - Success 1 Passed Remove Device - Success 2 Passed Remove Device - Success 3 Passed Remove Device - Success 4 Passed Remove Device - Success 5 Passed Remove Device - Success 6 - All Devices Passed Remove Device - Success 7 - Remove from whitelist Passed Remove Device - Success 8 - Remove from resolv list Passed Total: 11, Passed: 11 (100.0%), Failed: 0, Not Run: 0 Overall execution time: 4.26 seconds Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 2 + net/bluetooth/hci_sync.c | 617 +++++++++++++++++++++++++++++++ net/bluetooth/mgmt.c | 19 +- 3 files changed, 634 insertions(+), 4 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 143829cd1b29..b0708f900d89 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -63,3 +63,5 @@ int hci_enable_advertising_sync(struct hci_dev *hdev); int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, u8 instance, bool force); int hci_disable_advertising_sync(struct hci_dev *hdev); + +int hci_update_passive_scan_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 6ccd9de87cd6..94667efc4395 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1290,3 +1290,620 @@ int hci_disable_advertising_sync(struct hci_dev *hdev) return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADV_ENABLE, sizeof(enable), &enable, HCI_CMD_TIMEOUT); } + +static int hci_le_set_ext_scan_enable_sync(struct hci_dev *hdev, u8 val, + u8 filter_dup) +{ + struct hci_cp_le_set_ext_scan_enable cp; + + memset(&cp, 0, sizeof(cp)); + cp.enable = val; + cp.filter_dup = filter_dup; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_SCAN_ENABLE, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_set_scan_enable_sync(struct hci_dev *hdev, u8 val, + u8 filter_dup) +{ + struct hci_cp_le_set_scan_enable cp; + + if (use_ext_scan(hdev)) + return hci_le_set_ext_scan_enable_sync(hdev, val, filter_dup); + + memset(&cp, 0, sizeof(cp)); + cp.enable = val; + cp.filter_dup = filter_dup; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_SCAN_ENABLE, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_set_addr_resolution_enable_sync(struct hci_dev *hdev, u8 val) +{ + if (!use_ll_privacy(hdev) || + !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADDR_RESOLV_ENABLE, + sizeof(val), &val, HCI_CMD_TIMEOUT); +} + +int hci_scan_disable_sync(struct hci_dev *hdev, bool rpa_le_conn) +{ + int err; + + /* If controller is not scanning we are done. */ + if (!hci_dev_test_flag(hdev, HCI_LE_SCAN)) + return 0; + + if (hdev->scanning_paused) { + bt_dev_dbg(hdev, "Scanning is paused for suspend"); + return 0; + } + + if (hdev->suspended) + set_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); + + err = hci_le_set_scan_enable_sync(hdev, LE_SCAN_DISABLE, 0x00); + if (err) { + bt_dev_err(hdev, "Unable to disable scanning: %d", err); + return err; + } + + if (rpa_le_conn) { + err = hci_le_set_addr_resolution_enable_sync(hdev, 0x00); + if (err) + bt_dev_err(hdev, "Unable to disable LL privacy: %d", + err); + } + + return err; +} + +static bool scan_use_rpa(struct hci_dev *hdev) +{ + return hci_dev_test_flag(hdev, HCI_PRIVACY); +} + +static void hci_start_interleave_scan(struct hci_dev *hdev) +{ + hdev->interleave_scan_state = INTERLEAVE_SCAN_NO_FILTER; + queue_delayed_work(hdev->req_workqueue, + &hdev->interleave_scan, 0); +} + +static bool is_interleave_scanning(struct hci_dev *hdev) +{ + return hdev->interleave_scan_state != INTERLEAVE_SCAN_NONE; +} + +static void cancel_interleave_scan(struct hci_dev *hdev) +{ + bt_dev_dbg(hdev, "cancelling interleave scan"); + + cancel_delayed_work_sync(&hdev->interleave_scan); + + hdev->interleave_scan_state = INTERLEAVE_SCAN_NONE; +} + +/* Return true if interleave_scan wasn't started until exiting this function, + * otherwise, return false + */ +static bool hci_update_interleaved_scan_sync(struct hci_dev *hdev) +{ + /* Do interleaved scan only if all of the following are true: + * - There is at least one ADV monitor + * - At least one pending LE connection or one device to be scanned for + * - Monitor offloading is not supported + * If so, we should alternate between allowlist scan and one without + * any filters to save power. + */ + bool use_interleaving = hci_is_adv_monitoring(hdev) && + !(list_empty(&hdev->pend_le_conns) && + list_empty(&hdev->pend_le_reports)) && + hci_get_adv_monitor_offload_ext(hdev) == + HCI_ADV_MONITOR_EXT_NONE; + bool is_interleaving = is_interleave_scanning(hdev); + + if (use_interleaving && !is_interleaving) { + hci_start_interleave_scan(hdev); + bt_dev_dbg(hdev, "starting interleave scan"); + return true; + } + + if (!use_interleaving && is_interleaving) + cancel_interleave_scan(hdev); + + return false; +} + +/* Removes connection to resolve list if needed.*/ +static int hci_le_del_resolve_list_sync(struct hci_dev *hdev, + bdaddr_t *bdaddr, u8 bdaddr_type) +{ + struct hci_cp_le_del_from_resolv_list cp; + struct bdaddr_list_with_irk *entry; + + if (!use_ll_privacy(hdev) || + !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + return 0; + + /* Check if the IRK has been programmed */ + entry = hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, bdaddr, + bdaddr_type); + if (!entry) + return 0; + + cp.bdaddr_type = bdaddr_type; + bacpy(&cp.bdaddr, bdaddr); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_DEL_FROM_RESOLV_LIST, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_del_accept_list_sync(struct hci_dev *hdev, + bdaddr_t *bdaddr, u8 bdaddr_type) +{ + struct hci_cp_le_del_from_accept_list cp; + int err; + + /* Check if device is on accept list before removing it */ + if (!hci_bdaddr_list_lookup(&hdev->le_accept_list, bdaddr, bdaddr_type)) + return 0; + + cp.bdaddr_type = bdaddr_type; + bacpy(&cp.bdaddr, bdaddr); + + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_DEL_FROM_ACCEPT_LIST, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (err) { + bt_dev_err(hdev, "Unable to remove from allow list: %d", err); + return err; + } + + bt_dev_dbg(hdev, "Remove %pMR (0x%x) from allow list", &cp.bdaddr, + cp.bdaddr_type); + + return hci_le_del_resolve_list_sync(hdev, &cp.bdaddr, cp.bdaddr_type); +} + +/* Adds connection to resolve list if needed.*/ +static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, + struct hci_conn_params *params) +{ + struct hci_cp_le_add_to_resolv_list cp; + struct smp_irk *irk; + struct bdaddr_list_with_irk *entry; + + if (!use_ll_privacy(hdev) || + !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + return 0; + + irk = hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type); + if (!irk) + return 0; + + /* Check if the IK has _not_ been programmed yet. */ + entry = hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, + ¶ms->addr, + params->addr_type); + if (entry) + return 0; + + cp.bdaddr_type = params->addr_type; + bacpy(&cp.bdaddr, ¶ms->addr); + memcpy(cp.peer_irk, irk->val, 16); + + if (hci_dev_test_flag(hdev, HCI_PRIVACY)) + memcpy(cp.local_irk, hdev->irk, 16); + else + memset(cp.local_irk, 0, 16); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_ADD_TO_RESOLV_LIST, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +/* Adds connection to allow list if needed, if the device uses RPA (has IRK) + * this attempts to program the device in the resolving list as well. + */ +static int hci_le_add_accept_list_sync(struct hci_dev *hdev, + struct hci_conn_params *params, + u8 *num_entries, bool allow_rpa) +{ + struct hci_cp_le_add_to_accept_list cp; + int err; + + /* Already in accept list */ + if (hci_bdaddr_list_lookup(&hdev->le_accept_list, ¶ms->addr, + params->addr_type)) + return 0; + + /* Select filter policy to accept all advertising */ + if (*num_entries >= hdev->le_accept_list_size) + return -ENOSPC; + + /* Accept list can not be used with RPAs */ + if (!allow_rpa && + !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && + hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type)) { + return -EINVAL; + } + + /* During suspend, only wakeable devices can be in acceptlist */ + if (hdev->suspended && !hci_conn_test_flag(HCI_CONN_FLAG_REMOTE_WAKEUP, + params->current_flags)) + return 0; + + *num_entries += 1; + cp.bdaddr_type = params->addr_type; + bacpy(&cp.bdaddr, ¶ms->addr); + + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_ADD_TO_ACCEPT_LIST, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (err) { + bt_dev_err(hdev, "Unable to add to allow list: %d", err); + return err; + } + + bt_dev_dbg(hdev, "Add %pMR (0x%x) to allow list", &cp.bdaddr, + cp.bdaddr_type); + + return hci_le_add_resolve_list_sync(hdev, params); +} + +static u8 hci_update_accept_list_sync(struct hci_dev *hdev) +{ + struct hci_conn_params *params; + struct bdaddr_list *b, *t; + u8 num_entries = 0; + bool pend_conn, pend_report; + /* We allow acceptlisting even with RPAs in suspend. In the worst case, + * we won't be able to wake from devices that use the privacy1.2 + * features. Additionally, once we support privacy1.2 and IRK + * offloading, we can update this to also check for those conditions. + */ + bool allow_rpa = hdev->suspended; + + if (use_ll_privacy(hdev) && + hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + allow_rpa = true; + + /* Go through the current accept list programmed into the + * controller one by one and check if that address is still + * in the list of pending connections or list of devices to + * report. If not present in either list, then remove it from + * the controller. + */ + list_for_each_entry_safe(b, t, &hdev->le_accept_list, list) { + pend_conn = hci_pend_le_action_lookup(&hdev->pend_le_conns, + &b->bdaddr, + b->bdaddr_type); + pend_report = hci_pend_le_action_lookup(&hdev->pend_le_reports, + &b->bdaddr, + b->bdaddr_type); + + /* If the device is not likely to connect or report, + * remove it from the acceptlist. + */ + if (!pend_conn && !pend_report) { + hci_le_del_accept_list_sync(hdev, &b->bdaddr, + b->bdaddr_type); + continue; + } + + /* Accept list can not be used with RPAs */ + if (!allow_rpa && + !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && + hci_find_irk_by_addr(hdev, &b->bdaddr, b->bdaddr_type)) { + return 0x00; + } + + num_entries++; + } + + /* Since all no longer valid accept list entries have been + * removed, walk through the list of pending connections + * and ensure that any new device gets programmed into + * the controller. + * + * If the list of the devices is larger than the list of + * available accept list entries in the controller, then + * just abort and return filer policy value to not use the + * accept list. + */ + list_for_each_entry(params, &hdev->pend_le_conns, action) { + if (hci_le_add_accept_list_sync(hdev, params, &num_entries, + allow_rpa)) + return 0x00; + } + + /* After adding all new pending connections, walk through + * the list of pending reports and also add these to the + * accept list if there is still space. Abort if space runs out. + */ + list_for_each_entry(params, &hdev->pend_le_reports, action) { + if (hci_le_add_accept_list_sync(hdev, params, &num_entries, + allow_rpa)) + return 0x00; + } + + /* Use the allowlist unless the following conditions are all true: + * - We are not currently suspending + * - There are 1 or more ADV monitors registered and it's not offloaded + * - Interleaved scanning is not currently using the allowlist + */ + if (!idr_is_empty(&hdev->adv_monitors_idr) && !hdev->suspended && + hci_get_adv_monitor_offload_ext(hdev) == HCI_ADV_MONITOR_EXT_NONE && + hdev->interleave_scan_state != INTERLEAVE_SCAN_ALLOWLIST) + return 0x00; + + /* Select filter policy to use accept list */ + return 0x01; +} + +/* Returns true if an le connection is in the scanning state */ +static inline bool hci_is_le_conn_scanning(struct hci_dev *hdev) +{ + struct hci_conn_hash *h = &hdev->conn_hash; + struct hci_conn *c; + + rcu_read_lock(); + + list_for_each_entry_rcu(c, &h->list, list) { + if (c->type == LE_LINK && c->state == BT_CONNECT && + test_bit(HCI_CONN_SCANNING, &c->flags)) { + rcu_read_unlock(); + return true; + } + } + + rcu_read_unlock(); + + return false; +} + +static int hci_le_set_ext_scan_param_sync(struct hci_dev *hdev, u8 type, + u16 interval, u16 window, + u8 own_addr_type, u8 filter_policy) +{ + struct hci_cp_le_set_ext_scan_params *cp; + struct hci_cp_le_scan_phy_params *phy; + u8 data[sizeof(*cp) + sizeof(*phy) * 2]; + u8 num_phy = 0; + + cp = (void *)data; + phy = (void *)cp->data; + + memset(data, 0, sizeof(data)); + + cp->own_addr_type = own_addr_type; + cp->filter_policy = filter_policy; + + if (scan_1m(hdev) || scan_2m(hdev)) { + cp->scanning_phys |= LE_SCAN_PHY_1M; + + phy->type = type; + phy->interval = cpu_to_le16(interval); + phy->window = cpu_to_le16(window); + + num_phy++; + phy++; + } + + if (scan_coded(hdev)) { + cp->scanning_phys |= LE_SCAN_PHY_CODED; + + phy->type = type; + phy->interval = cpu_to_le16(interval); + phy->window = cpu_to_le16(window); + + num_phy++; + phy++; + } + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EXT_SCAN_PARAMS, + sizeof(*cp) + sizeof(*phy) * num_phy, + data, HCI_CMD_TIMEOUT); +} + +static int hci_le_set_scan_param_sync(struct hci_dev *hdev, u8 type, + u16 interval, u16 window, + u8 own_addr_type, u8 filter_policy) +{ + struct hci_cp_le_set_scan_param cp; + + if (use_ext_scan(hdev)) + return hci_le_set_ext_scan_param_sync(hdev, type, interval, + window, own_addr_type, + filter_policy); + + memset(&cp, 0, sizeof(cp)); + cp.type = type; + cp.interval = cpu_to_le16(interval); + cp.window = cpu_to_le16(window); + cp.own_address_type = own_addr_type; + cp.filter_policy = filter_policy; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_SCAN_PARAM, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, + u16 window, u8 own_addr_type, u8 filter_policy, + bool addr_resolv) +{ + int err; + + if (hdev->scanning_paused) { + bt_dev_dbg(hdev, "Scanning is paused for suspend"); + return 0; + } + + if (addr_resolv) { + err = hci_le_set_addr_resolution_enable_sync(hdev, 0x01); + if (err) + return err; + } + + err = hci_le_set_scan_param_sync(hdev, type, interval, window, + own_addr_type, filter_policy); + if (err) + return err; + + return hci_le_set_scan_enable_sync(hdev, LE_SCAN_ENABLE, + LE_SCAN_FILTER_DUP_ENABLE); +} + +/* Ensure to call hci_scan_disable_sync first to disable the controller based + * address resolution to be able to reconfigure resolving list. + */ +int hci_passive_scan_sync(struct hci_dev *hdev) +{ + u8 own_addr_type; + u8 filter_policy; + u16 window, interval; + /* Background scanning should run with address resolution */ + bool addr_resolv = true; + + if (hdev->scanning_paused) { + bt_dev_dbg(hdev, "Scanning is paused for suspend"); + return 0; + } + + /* Set require_privacy to false since no SCAN_REQ are send + * during passive scanning. Not using an non-resolvable address + * here is important so that peer devices using direct + * advertising with our address will be correctly reported + * by the controller. + */ + if (hci_update_random_address_sync(hdev, false, scan_use_rpa(hdev), + &own_addr_type)) + return 0; + + if (hdev->enable_advmon_interleave_scan && + hci_update_interleaved_scan_sync(hdev)) + return 0; + + bt_dev_dbg(hdev, "interleave state %d", hdev->interleave_scan_state); + /* Adding or removing entries from the accept list must + * happen before enabling scanning. The controller does + * not allow accept list modification while scanning. + */ + filter_policy = hci_update_accept_list_sync(hdev); + + /* When the controller is using random resolvable addresses and + * with that having LE privacy enabled, then controllers with + * Extended Scanner Filter Policies support can now enable support + * for handling directed advertising. + * + * So instead of using filter polices 0x00 (no acceptlist) + * and 0x01 (acceptlist enabled) use the new filter policies + * 0x02 (no acceptlist) and 0x03 (acceptlist enabled). + */ + if (hci_dev_test_flag(hdev, HCI_PRIVACY) && + (hdev->le_features[0] & HCI_LE_EXT_SCAN_POLICY)) + filter_policy |= 0x02; + + if (hdev->suspended) { + window = hdev->le_scan_window_suspend; + interval = hdev->le_scan_int_suspend; + + set_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks); + } else if (hci_is_le_conn_scanning(hdev)) { + window = hdev->le_scan_window_connect; + interval = hdev->le_scan_int_connect; + } else if (hci_is_adv_monitoring(hdev)) { + window = hdev->le_scan_window_adv_monitor; + interval = hdev->le_scan_int_adv_monitor; + } else { + window = hdev->le_scan_window; + interval = hdev->le_scan_interval; + } + + bt_dev_dbg(hdev, "LE passive scan with acceptlist = %d", filter_policy); + + return hci_start_scan_sync(hdev, LE_SCAN_PASSIVE, interval, window, + own_addr_type, filter_policy, addr_resolv); +} + +/* This function controls the passive scanning based on hdev->pend_le_conns + * list. If there are pending LE connection we start the background scanning, + * otherwise we stop it. + */ +int hci_update_passive_scan_sync(struct hci_dev *hdev) +{ + int err; + + if (!test_bit(HCI_UP, &hdev->flags) || + test_bit(HCI_INIT, &hdev->flags) || + hci_dev_test_flag(hdev, HCI_SETUP) || + hci_dev_test_flag(hdev, HCI_CONFIG) || + hci_dev_test_flag(hdev, HCI_AUTO_OFF) || + hci_dev_test_flag(hdev, HCI_UNREGISTER)) + return 0; + + /* No point in doing scanning if LE support hasn't been enabled */ + if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED)) + return 0; + + /* If discovery is active don't interfere with it */ + if (hdev->discovery.state != DISCOVERY_STOPPED) + return 0; + + /* Reset RSSI and UUID filters when starting background scanning + * since these filters are meant for service discovery only. + * + * The Start Discovery and Start Service Discovery operations + * ensure to set proper values for RSSI threshold and UUID + * filter list. So it is safe to just reset them here. + */ + hci_discovery_filter_clear(hdev); + + bt_dev_dbg(hdev, "ADV monitoring is %s", + hci_is_adv_monitoring(hdev) ? "on" : "off"); + + if (list_empty(&hdev->pend_le_conns) && + list_empty(&hdev->pend_le_reports) && + !hci_is_adv_monitoring(hdev)) { + /* If there is no pending LE connections or devices + * to be scanned for or no ADV monitors, we should stop the + * background scanning. + */ + + bt_dev_dbg(hdev, "stopping background scanning"); + + err = hci_scan_disable_sync(hdev, false); + if (err) + bt_dev_err(hdev, "stop background scanning failed: %d", + err); + } else { + /* If there is at least one pending LE connection, we should + * keep the background scan running. + */ + + /* If controller is connecting, we should not start scanning + * since some controllers are not able to scan and connect at + * the same time. + */ + if (hci_lookup_le_connect(hdev)) + return 0; + + err = hci_scan_disable_sync(hdev, false); + if (err) { + bt_dev_err(hdev, "stop background scanning failed: %d", + err); + return err; + } + + bt_dev_dbg(hdev, "start background scanning"); + + err = hci_passive_scan_sync(hdev); + if (err) + bt_dev_err(hdev, "start background scanning failed: %d", + err); + } + + return err; +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 6e481ca86303..ef61b152c627 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -6927,6 +6927,11 @@ static void device_added(struct sock *sk, struct hci_dev *hdev, mgmt_event(MGMT_EV_DEVICE_ADDED, hdev, &ev, sizeof(ev), sk); } +static int add_device_sync(struct hci_dev *hdev, void *data) +{ + return hci_update_passive_scan_sync(hdev); +} + static int add_device(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -7009,7 +7014,9 @@ static int add_device(struct sock *sk, struct hci_dev *hdev, current_flags = params->current_flags; } - hci_update_background_scan(hdev); + err = hci_cmd_sync_queue(hdev, add_device_sync, NULL, NULL); + if (err < 0) + goto unlock; added: device_added(sk, hdev, &cp->addr.bdaddr, cp->addr.type, cp->action); @@ -7036,6 +7043,11 @@ static void device_removed(struct sock *sk, struct hci_dev *hdev, mgmt_event(MGMT_EV_DEVICE_REMOVED, hdev, &ev, sizeof(ev), sk); } +static int remove_device_sync(struct hci_dev *hdev, void *data) +{ + return hci_update_passive_scan_sync(hdev); +} + static int remove_device(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -7115,7 +7127,6 @@ static int remove_device(struct sock *sk, struct hci_dev *hdev, list_del(¶ms->action); list_del(¶ms->list); kfree(params); - hci_update_background_scan(hdev); device_removed(sk, hdev, &cp->addr.bdaddr, cp->addr.type); } else { @@ -7152,10 +7163,10 @@ static int remove_device(struct sock *sk, struct hci_dev *hdev, } bt_dev_dbg(hdev, "All LE connection parameters were removed"); - - hci_update_background_scan(hdev); } + hci_cmd_sync_queue(hdev, remove_device_sync, NULL, NULL); + complete: err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_REMOVE_DEVICE, MGMT_STATUS_SUCCESS, &cp->addr, From ad383c2c65a5baf16e334cd40a013cc302176891 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:42 -0700 Subject: [PATCH 0035/2048] Bluetooth: hci_sync: Enable advertising when LL privacy is enabled This enables advertising when LL privacy is enabled and changes the command sequence when resolving list is updated to also account for when advertising is enabled using the following sequence: If there are devices to scan: Disable Scanning -> Update Accept List -> use_ll_privacy((Disable Advertising) -> Disable Resolving List -> Update Resolving List -> Enable Resolving List -> (Enable Advertising)) -> Enable Scanning Otherwise: Disable Scanning Errors during the Update Accept List stage are handled gracefully by restoring any previous state (e.g. advertising) and disabling the use of accept list as either accept list or resolving list could not be updated. Tested with: mgmt-tester -s "LL Privacy" Test Summary ------------ LL Privacy - Add Device 1 (Add to WL) Passed LL Privacy - Add Device 2 (Add to RL) Passed LL Privacy - Add Device 3 (Enable RL) Passed LL Privacy - Add Device 4 (2 Devices to WL) Passed LL Privacy - Add Device 5 (2 Devices to RL) Passed LL Privacy - Add Device 6 (RL is full) Passed LL Privacy - Add Device 7 (WL is full) Passed LL Privacy - Add Device 8 (Disable Adv) Passed LL Privacy - Add Device 9 (Multi Adv) Passed LL Privacy - Add Device 10 (Multi Dev and Multi Adv) Passed LL Privacy - Remove Device 1 (Remove from WL) Passed LL Privacy - Remove Device 2 (Remove from RL) Passed LL Privacy - Remove Device 3 (Disable RL) Passed LL Privacy - Remove Device 4 (Disable Adv) Passed LL Privacy - Remove Device 5 (Multi Adv) Passed LL Privacy - Start Discovery 1 (Disable RL) Passed LL Privacy - Start Discovery 2 (Disable RL) Passed Total: 18, Passed: 18 (100.0%), Failed: 0, Not Run: 0 Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 5 +- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_event.c | 4 +- net/bluetooth/hci_request.c | 26 +-- net/bluetooth/hci_sync.c | 270 ++++++++++++++++++++++--------- net/bluetooth/mgmt.c | 39 +---- 6 files changed, 214 insertions(+), 131 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 17ddea51d161..a6b075203cbe 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -1465,8 +1465,11 @@ void hci_conn_del_sysfs(struct hci_conn *conn); #define scan_coded(dev) (((dev)->le_tx_def_phys & HCI_LE_SET_PHY_CODED) || \ ((dev)->le_rx_def_phys & HCI_LE_SET_PHY_CODED)) +#define ll_privacy_capable(dev) ((dev)->le_features[0] & HCI_LE_LL_PRIVACY) + /* Use LL Privacy based address resolution if supported */ -#define use_ll_privacy(dev) ((dev)->le_features[0] & HCI_LE_LL_PRIVACY) +#define use_ll_privacy(dev) (ll_privacy_capable(dev) && \ + hci_dev_test_flag(dev, HCI_ENABLE_LL_PRIVACY)) /* Use enhanced synchronous connection if command is supported */ #define enhanced_sco_capable(dev) ((dev)->commands[29] & 0x08) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index b0708f900d89..ec727eb18e90 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -65,3 +65,4 @@ int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, int hci_disable_advertising_sync(struct hci_dev *hdev); int hci_update_passive_scan_sync(struct hci_dev *hdev); +int hci_update_passive_scan(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 1ac3eea7ac8e..e45e62b4a0d8 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -5501,9 +5501,7 @@ static void hci_le_enh_conn_complete_evt(struct hci_dev *hdev, le16_to_cpu(ev->latency), le16_to_cpu(ev->supervision_timeout)); - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && - hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) + if (hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) hci_req_disable_address_resolution(hdev); } diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index b1b33432c68d..cb36b2bd1109 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -511,9 +511,7 @@ void hci_req_add_le_scan_disable(struct hci_request *req, bool rpa_le_conn) } /* Disable address resolution */ - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && - hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION) && !rpa_le_conn) { + if (hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION) && !rpa_le_conn) { __u8 enable = 0x00; hci_req_add(req, HCI_OP_LE_SET_ADDR_RESOLV_ENABLE, 1, &enable); @@ -532,8 +530,7 @@ static void del_from_accept_list(struct hci_request *req, bdaddr_t *bdaddr, cp.bdaddr_type); hci_req_add(req, HCI_OP_LE_DEL_FROM_ACCEPT_LIST, sizeof(cp), &cp); - if (use_ll_privacy(req->hdev) && - hci_dev_test_flag(req->hdev, HCI_ENABLE_LL_PRIVACY)) { + if (use_ll_privacy(req->hdev)) { struct smp_irk *irk; irk = hci_find_irk_by_addr(req->hdev, bdaddr, bdaddr_type); @@ -586,8 +583,7 @@ static int add_to_accept_list(struct hci_request *req, cp.bdaddr_type); hci_req_add(req, HCI_OP_LE_ADD_TO_ACCEPT_LIST, sizeof(cp), &cp); - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) { + if (use_ll_privacy(hdev)) { struct smp_irk *irk; irk = hci_find_irk_by_addr(hdev, ¶ms->addr, @@ -626,8 +622,7 @@ static u8 update_accept_list(struct hci_request *req) */ bool allow_rpa = hdev->suspended; - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (use_ll_privacy(hdev)) allow_rpa = true; /* Go through the current accept list programmed into the @@ -716,9 +711,7 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, return; } - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && - addr_resolv) { + if (use_ll_privacy(hdev) && addr_resolv) { u8 enable = 0x01; hci_req_add(req, HCI_OP_LE_SET_ADDR_RESOLV_ENABLE, 1, &enable); @@ -1480,8 +1473,7 @@ void hci_req_disable_address_resolution(struct hci_dev *hdev) struct hci_request req; __u8 enable = 0x00; - if (!use_ll_privacy(hdev) && - !hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) + if (!hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) return; hci_req_init(&req, hdev); @@ -1624,8 +1616,7 @@ int hci_get_random_address(struct hci_dev *hdev, bool require_privacy, /* If Controller supports LL Privacy use own address type is * 0x03 */ - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (use_ll_privacy(hdev)) *own_addr_type = ADDR_LE_DEV_RANDOM_RESOLVED; else *own_addr_type = ADDR_LE_DEV_RANDOM; @@ -2092,8 +2083,7 @@ int hci_update_random_address(struct hci_request *req, bool require_privacy, /* If Controller supports LL Privacy use own address type is * 0x03 */ - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (use_ll_privacy(hdev)) *own_addr_type = ADDR_LE_DEV_RANDOM_RESOLVED; else *own_addr_type = ADDR_LE_DEV_RANDOM; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 94667efc4395..c9aba2fd0301 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -517,8 +517,7 @@ int hci_update_random_address_sync(struct hci_dev *hdev, bool require_privacy, /* If Controller supports LL Privacy use own address type is * 0x03 */ - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (use_ll_privacy(hdev)) *own_addr_type = ADDR_LE_DEV_RANDOM_RESOLVED; else *own_addr_type = ADDR_LE_DEV_RANDOM; @@ -931,6 +930,10 @@ int hci_enable_advertising_sync(struct hci_dev *hdev) u32 flags; u8 status; + if (ext_adv_capable(hdev)) + return hci_enable_ext_advertising_sync(hdev, + hdev->cur_adv_instance); + flags = hci_adv_instance_flags(hdev, hdev->cur_adv_instance); adv_instance = hci_find_adv_instance(hdev, hdev->cur_adv_instance); @@ -943,11 +946,9 @@ int hci_enable_advertising_sync(struct hci_dev *hdev) if (!is_advertising_allowed(hdev, connectable)) return -EINVAL; - if (hci_dev_test_flag(hdev, HCI_LE_ADV)) { - status = hci_disable_advertising_sync(hdev); - if (status) - return status; - } + status = hci_disable_advertising_sync(hdev); + if (status) + return status; /* Clear the HCI_LE_ADV bit temporarily so that the * hci_update_random_address knows that it's safe to go ahead @@ -1284,6 +1285,10 @@ int hci_disable_advertising_sync(struct hci_dev *hdev) { u8 enable = 0x00; + /* If controller is not advertising we are done. */ + if (!hci_dev_test_flag(hdev, HCI_LE_ADV)) + return 0; + if (ext_adv_capable(hdev)) return hci_disable_ext_adv_instance_sync(hdev, 0x00); @@ -1322,15 +1327,18 @@ static int hci_le_set_scan_enable_sync(struct hci_dev *hdev, u8 val, static int hci_le_set_addr_resolution_enable_sync(struct hci_dev *hdev, u8 val) { - if (!use_ll_privacy(hdev) || - !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (!use_ll_privacy(hdev)) + return 0; + + /* If controller is not/already resolving we are done. */ + if (val == hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) return 0; return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_ADDR_RESOLV_ENABLE, sizeof(val), &val, HCI_CMD_TIMEOUT); } -int hci_scan_disable_sync(struct hci_dev *hdev, bool rpa_le_conn) +int hci_scan_disable_sync(struct hci_dev *hdev) { int err; @@ -1352,13 +1360,6 @@ int hci_scan_disable_sync(struct hci_dev *hdev, bool rpa_le_conn) return err; } - if (rpa_le_conn) { - err = hci_le_set_addr_resolution_enable_sync(hdev, 0x00); - if (err) - bt_dev_err(hdev, "Unable to disable LL privacy: %d", - err); - } - return err; } @@ -1426,8 +1427,7 @@ static int hci_le_del_resolve_list_sync(struct hci_dev *hdev, struct hci_cp_le_del_from_resolv_list cp; struct bdaddr_list_with_irk *entry; - if (!use_ll_privacy(hdev) || - !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (!use_ll_privacy(hdev)) return 0; /* Check if the IRK has been programmed */ @@ -1456,6 +1456,11 @@ static int hci_le_del_accept_list_sync(struct hci_dev *hdev, cp.bdaddr_type = bdaddr_type; bacpy(&cp.bdaddr, bdaddr); + /* Ignore errors when removing from resolving list as that is likely + * that the device was never added. + */ + hci_le_del_resolve_list_sync(hdev, &cp.bdaddr, cp.bdaddr_type); + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_DEL_FROM_ACCEPT_LIST, sizeof(cp), &cp, HCI_CMD_TIMEOUT); if (err) { @@ -1466,7 +1471,7 @@ static int hci_le_del_accept_list_sync(struct hci_dev *hdev, bt_dev_dbg(hdev, "Remove %pMR (0x%x) from allow list", &cp.bdaddr, cp.bdaddr_type); - return hci_le_del_resolve_list_sync(hdev, &cp.bdaddr, cp.bdaddr_type); + return 0; } /* Adds connection to resolve list if needed.*/ @@ -1477,8 +1482,7 @@ static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, struct smp_irk *irk; struct bdaddr_list_with_irk *entry; - if (!use_ll_privacy(hdev) || - !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) + if (!use_ll_privacy(hdev)) return 0; irk = hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type); @@ -1510,7 +1514,7 @@ static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, */ static int hci_le_add_accept_list_sync(struct hci_dev *hdev, struct hci_conn_params *params, - u8 *num_entries, bool allow_rpa) + u8 *num_entries) { struct hci_cp_le_add_to_accept_list cp; int err; @@ -1525,8 +1529,7 @@ static int hci_le_add_accept_list_sync(struct hci_dev *hdev, return -ENOSPC; /* Accept list can not be used with RPAs */ - if (!allow_rpa && - !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && + if (!use_ll_privacy(hdev) && hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type)) { return -EINVAL; } @@ -1536,6 +1539,16 @@ static int hci_le_add_accept_list_sync(struct hci_dev *hdev, params->current_flags)) return 0; + /* Attempt to program the device in the resolving list first to avoid + * having to rollback in case it fails since the resolving list is + * dynamic it can probably be smaller than the accept list. + */ + err = hci_le_add_resolve_list_sync(hdev, params); + if (err) { + bt_dev_err(hdev, "Unable to add to resolve list: %d", err); + return err; + } + *num_entries += 1; cp.bdaddr_type = params->addr_type; bacpy(&cp.bdaddr, ¶ms->addr); @@ -1544,31 +1557,129 @@ static int hci_le_add_accept_list_sync(struct hci_dev *hdev, sizeof(cp), &cp, HCI_CMD_TIMEOUT); if (err) { bt_dev_err(hdev, "Unable to add to allow list: %d", err); + /* Rollback the device from the resolving list */ + hci_le_del_resolve_list_sync(hdev, &cp.bdaddr, cp.bdaddr_type); return err; } bt_dev_dbg(hdev, "Add %pMR (0x%x) to allow list", &cp.bdaddr, cp.bdaddr_type); - return hci_le_add_resolve_list_sync(hdev, params); + return 0; } +/* This function disables all advertising instances (including 0x00) */ +static int hci_pause_advertising_sync(struct hci_dev *hdev) +{ + int err; + + /* If there are no instances or advertising has already been paused + * there is nothing to do. + */ + if (!hdev->adv_instance_cnt || hdev->advertising_paused) + return 0; + + bt_dev_dbg(hdev, "Pausing advertising instances"); + + /* Call to disable any advertisements active on the controller. + * This will succeed even if no advertisements are configured. + */ + err = hci_disable_advertising_sync(hdev); + if (err) + return err; + + /* If we are using software rotation, pause the loop */ + if (!ext_adv_capable(hdev)) + cancel_adv_timeout(hdev); + + hdev->advertising_paused = true; + + return 0; +} + +/* This function disables all user advertising instances (excluding 0x00) */ +static int hci_resume_advertising_sync(struct hci_dev *hdev) +{ + struct adv_info *adv, *tmp; + int err; + + /* If advertising has not been paused there is nothing to do. */ + if (!hdev->advertising_paused) + return 0; + + bt_dev_dbg(hdev, "Resuming advertising instances"); + + if (ext_adv_capable(hdev)) { + /* Call for each tracked instance to be re-enabled */ + list_for_each_entry_safe(adv, tmp, &hdev->adv_instances, list) { + err = hci_enable_ext_advertising_sync(hdev, + adv->instance); + if (!err) + continue; + + /* If the instance cannot be resumed remove it */ + hci_remove_ext_adv_instance_sync(hdev, adv->instance, + NULL); + } + } else { + /* Schedule for most recent instance to be restarted and begin + * the software rotation loop + */ + err = hci_schedule_adv_instance_sync(hdev, + hdev->cur_adv_instance, + true); + } + + hdev->advertising_paused = false; + + return err; +} + +/* Device must not be scanning when updating the accept list. + * + * Update is done using the following sequence: + * + * use_ll_privacy((Disable Advertising) -> Disable Resolving List) -> + * Remove Devices From Accept List -> + * (has IRK && use_ll_privacy(Remove Devices From Resolving List))-> + * Add Devices to Accept List -> + * (has IRK && use_ll_privacy(Remove Devices From Resolving List)) -> + * use_ll_privacy(Enable Resolving List -> (Enable Advertising)) -> + * Enable Scanning + * + * In case of failure advertising shall be restored to its original state and + * return would disable accept list since either accept or resolving list could + * not be programmed. + * + */ static u8 hci_update_accept_list_sync(struct hci_dev *hdev) { struct hci_conn_params *params; struct bdaddr_list *b, *t; u8 num_entries = 0; bool pend_conn, pend_report; - /* We allow acceptlisting even with RPAs in suspend. In the worst case, - * we won't be able to wake from devices that use the privacy1.2 - * features. Additionally, once we support privacy1.2 and IRK - * offloading, we can update this to also check for those conditions. - */ - bool allow_rpa = hdev->suspended; + int err; - if (use_ll_privacy(hdev) && - hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - allow_rpa = true; + /* Pause advertising if resolving list can be used as controllers are + * cannot accept resolving list modifications while advertising. + */ + if (use_ll_privacy(hdev)) { + err = hci_pause_advertising_sync(hdev); + if (err) { + bt_dev_err(hdev, "pause advertising failed: %d", err); + return 0x00; + } + } + + /* Disable address resolution while reprogramming accept list since + * devices that do have an IRK will be programmed in the resolving list + * when LL Privacy is enabled. + */ + err = hci_le_set_addr_resolution_enable_sync(hdev, 0x00); + if (err) { + bt_dev_err(hdev, "Unable to disable LL privacy: %d", err); + goto done; + } /* Go through the current accept list programmed into the * controller one by one and check if that address is still @@ -1593,13 +1704,6 @@ static u8 hci_update_accept_list_sync(struct hci_dev *hdev) continue; } - /* Accept list can not be used with RPAs */ - if (!allow_rpa && - !hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY) && - hci_find_irk_by_addr(hdev, &b->bdaddr, b->bdaddr_type)) { - return 0x00; - } - num_entries++; } @@ -1614,9 +1718,9 @@ static u8 hci_update_accept_list_sync(struct hci_dev *hdev) * accept list. */ list_for_each_entry(params, &hdev->pend_le_conns, action) { - if (hci_le_add_accept_list_sync(hdev, params, &num_entries, - allow_rpa)) - return 0x00; + err = hci_le_add_accept_list_sync(hdev, params, &num_entries); + if (err) + goto done; } /* After adding all new pending connections, walk through @@ -1624,9 +1728,9 @@ static u8 hci_update_accept_list_sync(struct hci_dev *hdev) * accept list if there is still space. Abort if space runs out. */ list_for_each_entry(params, &hdev->pend_le_reports, action) { - if (hci_le_add_accept_list_sync(hdev, params, &num_entries, - allow_rpa)) - return 0x00; + err = hci_le_add_accept_list_sync(hdev, params, &num_entries); + if (err) + goto done; } /* Use the allowlist unless the following conditions are all true: @@ -1637,10 +1741,20 @@ static u8 hci_update_accept_list_sync(struct hci_dev *hdev) if (!idr_is_empty(&hdev->adv_monitors_idr) && !hdev->suspended && hci_get_adv_monitor_offload_ext(hdev) == HCI_ADV_MONITOR_EXT_NONE && hdev->interleave_scan_state != INTERLEAVE_SCAN_ALLOWLIST) - return 0x00; + err = -EINVAL; + +done: + /* Enable address resolution when LL Privacy is enabled. */ + err = hci_le_set_addr_resolution_enable_sync(hdev, 0x01); + if (err) + bt_dev_err(hdev, "Unable to enable LL privacy: %d", err); + + /* Resume advertising if it was paused */ + if (use_ll_privacy(hdev)) + hci_resume_advertising_sync(hdev); /* Select filter policy to use accept list */ - return 0x01; + return err ? 0x00 : 0x01; } /* Returns true if an le connection is in the scanning state */ @@ -1731,8 +1845,7 @@ static int hci_le_set_scan_param_sync(struct hci_dev *hdev, u8 type, } static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, - u16 window, u8 own_addr_type, u8 filter_policy, - bool addr_resolv) + u16 window, u8 own_addr_type, u8 filter_policy) { int err; @@ -1741,12 +1854,6 @@ static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, return 0; } - if (addr_resolv) { - err = hci_le_set_addr_resolution_enable_sync(hdev, 0x01); - if (err) - return err; - } - err = hci_le_set_scan_param_sync(hdev, type, interval, window, own_addr_type, filter_policy); if (err) @@ -1756,22 +1863,24 @@ static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, LE_SCAN_FILTER_DUP_ENABLE); } -/* Ensure to call hci_scan_disable_sync first to disable the controller based - * address resolution to be able to reconfigure resolving list. - */ int hci_passive_scan_sync(struct hci_dev *hdev) { u8 own_addr_type; u8 filter_policy; u16 window, interval; - /* Background scanning should run with address resolution */ - bool addr_resolv = true; + int err; if (hdev->scanning_paused) { bt_dev_dbg(hdev, "Scanning is paused for suspend"); return 0; } + err = hci_scan_disable_sync(hdev); + if (err) { + bt_dev_err(hdev, "disable scanning failed: %d", err); + return err; + } + /* Set require_privacy to false since no SCAN_REQ are send * during passive scanning. Not using an non-resolvable address * here is important so that peer devices using direct @@ -1787,6 +1896,7 @@ int hci_passive_scan_sync(struct hci_dev *hdev) return 0; bt_dev_dbg(hdev, "interleave state %d", hdev->interleave_scan_state); + /* Adding or removing entries from the accept list must * happen before enabling scanning. The controller does * not allow accept list modification while scanning. @@ -1825,12 +1935,23 @@ int hci_passive_scan_sync(struct hci_dev *hdev) bt_dev_dbg(hdev, "LE passive scan with acceptlist = %d", filter_policy); return hci_start_scan_sync(hdev, LE_SCAN_PASSIVE, interval, window, - own_addr_type, filter_policy, addr_resolv); + own_addr_type, filter_policy); } /* This function controls the passive scanning based on hdev->pend_le_conns * list. If there are pending LE connection we start the background scanning, - * otherwise we stop it. + * otherwise we stop it in the following sequence: + * + * If there are devices to scan: + * + * Disable Scanning -> Update Accept List -> + * use_ll_privacy((Disable Advertising) -> Disable Resolving List -> + * Update Resolving List -> Enable Resolving List -> (Enable Advertising)) -> + * Enable Scanning + * + * Otherwise: + * + * Disable Scanning */ int hci_update_passive_scan_sync(struct hci_dev *hdev) { @@ -1874,7 +1995,7 @@ int hci_update_passive_scan_sync(struct hci_dev *hdev) bt_dev_dbg(hdev, "stopping background scanning"); - err = hci_scan_disable_sync(hdev, false); + err = hci_scan_disable_sync(hdev); if (err) bt_dev_err(hdev, "stop background scanning failed: %d", err); @@ -1890,13 +2011,6 @@ int hci_update_passive_scan_sync(struct hci_dev *hdev) if (hci_lookup_le_connect(hdev)) return 0; - err = hci_scan_disable_sync(hdev, false); - if (err) { - bt_dev_err(hdev, "stop background scanning failed: %d", - err); - return err; - } - bt_dev_dbg(hdev, "start background scanning"); err = hci_passive_scan_sync(hdev); @@ -1907,3 +2021,13 @@ int hci_update_passive_scan_sync(struct hci_dev *hdev) return err; } + +static int update_passive_scan_sync(struct hci_dev *hdev, void *data) +{ + return hci_update_passive_scan_sync(hdev); +} + +int hci_update_passive_scan(struct hci_dev *hdev) +{ + return hci_cmd_sync_queue(hdev, update_passive_scan_sync, NULL, NULL); +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ef61b152c627..ca2273a8979d 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -839,12 +839,7 @@ static u32 get_supported_settings(struct hci_dev *hdev) settings |= MGMT_SETTING_SECURE_CONN; settings |= MGMT_SETTING_PRIVACY; settings |= MGMT_SETTING_STATIC_ADDRESS; - - /* When the experimental feature for LL Privacy support is - * enabled, then advertising is no longer supported. - */ - if (!hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - settings |= MGMT_SETTING_ADVERTISING; + settings |= MGMT_SETTING_ADVERTISING; } if (test_bit(HCI_QUIRK_EXTERNAL_CONFIG, &hdev->quirks) || @@ -3846,7 +3841,7 @@ static int read_exp_features_info(struct sock *sk, struct hci_dev *hdev, idx++; } - if (hdev && use_ll_privacy(hdev)) { + if (hdev && ll_privacy_capable(hdev)) { if (hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) flags = BIT(0) | BIT(1); else @@ -5621,13 +5616,6 @@ static int set_advertising(struct sock *sk, struct hci_dev *hdev, void *data, return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_ADVERTISING, status); - /* Enabling the experimental LL Privay support disables support for - * advertising. - */ - if (hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_ADVERTISING, - MGMT_STATUS_NOT_SUPPORTED); - if (cp->val != 0x00 && cp->val != 0x01 && cp->val != 0x02) return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_ADVERTISING, MGMT_STATUS_INVALID_PARAMS); @@ -7723,13 +7711,6 @@ static int read_adv_features(struct sock *sk, struct hci_dev *hdev, return mgmt_cmd_status(sk, hdev->id, MGMT_OP_READ_ADV_FEATURES, MGMT_STATUS_REJECTED); - /* Enabling the experimental LL Privay support disables support for - * advertising. - */ - if (hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_READ_ADV_FEATURES, - MGMT_STATUS_NOT_SUPPORTED); - hci_dev_lock(hdev); rp_len = sizeof(*rp) + hdev->adv_instance_cnt; @@ -7969,13 +7950,6 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, status); - /* Enabling the experimental LL Privay support disables support for - * advertising. - */ - if (hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, - MGMT_STATUS_NOT_SUPPORTED); - if (cp->instance < 1 || cp->instance > hdev->le_num_of_adv_sets) return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, MGMT_STATUS_INVALID_PARAMS); @@ -8471,13 +8445,6 @@ static int remove_advertising(struct sock *sk, struct hci_dev *hdev, bt_dev_dbg(hdev, "sock %p", sk); - /* Enabling the experimental LL Privay support disables support for - * advertising. - */ - if (hci_dev_test_flag(hdev, HCI_ENABLE_LL_PRIVACY)) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_REMOVE_ADVERTISING, - MGMT_STATUS_NOT_SUPPORTED); - hci_dev_lock(hdev); if (cp->instance && !hci_find_adv_instance(hdev, cp->instance)) { @@ -8788,7 +8755,7 @@ void mgmt_power_on(struct hci_dev *hdev, int err) if (!err) { restart_le_actions(hdev); - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); } mgmt_pending_foreach(MGMT_OP_SET_POWERED, hdev, settings_rsp, &match); From 5bee2fd6bcaaaa9f8f415afc48ed8c1083d8a303 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:43 -0700 Subject: [PATCH 0036/2048] Bluetooth: hci_sync: Rework background scan This replaces the use of hci_update_background_scan with hci_update_passive_scan which runs from cmd_work_sync and deal properly with resolving list when LL privacy is enabled. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_conn.c | 6 +++--- net/bluetooth/hci_core.c | 14 +++++++------- net/bluetooth/hci_event.c | 4 ++-- net/bluetooth/hci_sync.c | 9 +++++++++ net/bluetooth/mgmt.c | 8 ++++---- net/bluetooth/msft.c | 2 +- 6 files changed, 26 insertions(+), 17 deletions(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index bd669c95b9a7..1783ec5f6e3e 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -108,7 +108,7 @@ static void hci_connect_le_scan_cleanup(struct hci_conn *conn) break; } - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); } static void hci_conn_cleanup(struct hci_conn *conn) @@ -913,7 +913,7 @@ void hci_le_conn_failed(struct hci_conn *conn, u8 status) /* Since we may have temporarily stopped the background scanning in * favor of connection establishment, we should restart it. */ - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); /* Re-enable advertising in case this was a failed connection * attempt as a peripheral. @@ -1411,7 +1411,7 @@ struct hci_conn *hci_connect_le_scan(struct hci_dev *hdev, bdaddr_t *dst, conn->conn_timeout = conn_timeout; conn->conn_reason = conn_reason; - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); done: hci_conn_hold(conn); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index fdc1cce5eb5b..8c54b1d4d41b 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -975,7 +975,7 @@ void hci_discovery_set_state(struct hci_dev *hdev, int state) switch (state) { case DISCOVERY_STOPPED: - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); if (old_state != DISCOVERY_STARTING) mgmt_discovering(hdev, 0); @@ -1850,7 +1850,7 @@ done: return ret; } -static void hci_update_scan_state(struct hci_dev *hdev, u8 scan) +static void hci_update_passive_scan_state(struct hci_dev *hdev, u8 scan) { bool conn_changed, discov_changed; @@ -1951,7 +1951,7 @@ int hci_dev_cmd(unsigned int cmd, void __user *arg) * get correctly modified as this was a non-mgmt change. */ if (!err) - hci_update_scan_state(hdev, dr.dev_opt); + hci_update_passive_scan_state(hdev, dr.dev_opt); break; case HCISETLINKPOL: @@ -3096,7 +3096,7 @@ bool hci_add_adv_monitor(struct hci_dev *hdev, struct adv_monitor *monitor, switch (hci_get_adv_monitor_offload_ext(hdev)) { case HCI_ADV_MONITOR_EXT_NONE: - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); bt_dev_dbg(hdev, "%s add monitor status %d", hdev->name, *err); /* Message was not forwarded to controller - not an error */ return false; @@ -3160,7 +3160,7 @@ bool hci_remove_single_adv_monitor(struct hci_dev *hdev, u16 handle, int *err) pending = hci_remove_adv_monitor(hdev, monitor, handle, err); if (!*err && !pending) - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); bt_dev_dbg(hdev, "%s remove monitor handle %d, status %d, %spending", hdev->name, handle, *err, pending ? "" : "not "); @@ -3192,7 +3192,7 @@ bool hci_remove_all_adv_monitor(struct hci_dev *hdev, int *err) } if (update) - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); bt_dev_dbg(hdev, "%s remove all monitors status %d, %spending", hdev->name, *err, pending ? "" : "not "); @@ -3486,7 +3486,7 @@ void hci_conn_params_del(struct hci_dev *hdev, bdaddr_t *addr, u8 addr_type) hci_conn_params_free(params); - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); BT_DBG("addr %pMR (type %u)", addr, addr_type); } diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index e45e62b4a0d8..d5e4a82b7020 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -3028,7 +3028,7 @@ static void hci_disconn_complete_evt(struct hci_dev *hdev, struct sk_buff *skb) case HCI_AUTO_CONN_ALWAYS: list_del_init(¶ms->action); list_add(¶ms->action, &hdev->pend_le_conns); - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); break; default: @@ -5471,7 +5471,7 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, } unlock: - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); hci_dev_unlock(hdev); } diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index c9aba2fd0301..6d53a876de85 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2029,5 +2029,14 @@ static int update_passive_scan_sync(struct hci_dev *hdev, void *data) int hci_update_passive_scan(struct hci_dev *hdev) { + /* Only queue if it would have any effect */ + if (!test_bit(HCI_UP, &hdev->flags) || + test_bit(HCI_INIT, &hdev->flags) || + hci_dev_test_flag(hdev, HCI_SETUP) || + hci_dev_test_flag(hdev, HCI_CONFIG) || + hci_dev_test_flag(hdev, HCI_AUTO_OFF) || + hci_dev_test_flag(hdev, HCI_UNREGISTER)) + return 0; + return hci_cmd_sync_queue(hdev, update_passive_scan_sync, NULL, NULL); } diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ca2273a8979d..d09a757a5851 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1585,7 +1585,7 @@ static int set_connectable_update_settings(struct hci_dev *hdev, if (changed) { hci_req_update_scan(hdev); - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); return new_settings(hdev, sk); } @@ -1943,7 +1943,7 @@ static void le_enable_complete(struct hci_dev *hdev, u8 status, u16 opcode) __hci_req_update_scan_rsp_data(&req, 0x00); } hci_req_run(&req, NULL); - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); } unlock: @@ -4488,7 +4488,7 @@ int mgmt_add_adv_patterns_monitor_complete(struct hci_dev *hdev, u8 status) hdev->adv_monitors_cnt++; if (monitor->state == ADV_MONITOR_STATE_NOT_REGISTERED) monitor->state = ADV_MONITOR_STATE_REGISTERED; - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); } err = mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, @@ -4714,7 +4714,7 @@ int mgmt_remove_adv_monitor_complete(struct hci_dev *hdev, u8 status) rp.monitor_handle = cp->monitor_handle; if (!status) - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); err = mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, mgmt_status(status), &rp, sizeof(rp)); diff --git a/net/bluetooth/msft.c b/net/bluetooth/msft.c index 255cffa554ee..5205d9410ff5 100644 --- a/net/bluetooth/msft.c +++ b/net/bluetooth/msft.c @@ -168,7 +168,7 @@ static void reregister_monitor(struct hci_dev *hdev, int handle) if (!monitor) { /* All monitors have been reregistered */ msft->reregistering = false; - hci_update_background_scan(hdev); + hci_update_passive_scan(hdev); return; } From cf75ad8b41d2aa06f98f365d42a3ae8b059daddd Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:44 -0700 Subject: [PATCH 0037/2048] Bluetooth: hci_sync: Convert MGMT_SET_POWERED This make use of hci_cmd_sync_queue when MGMT_SET_POWERED is used so all commands are run within hdev->cmd_sync_work instead of hdev->power_on_work and hdev->power_off_work. In addition to that the power on sequence now takes into account if local IRK needs to be programmed in the resolving list. Tested with: tools/mgmt-tester -s "Set powered" Test Summary ------------ Set powered on - Success Passed Set powered on - Invalid parameters 1 Passed Set powered on - Invalid parameters 2 Passed Set powered on - Invalid parameters 3 Passed Set powered on - Invalid index Passed Set powered on - Privacy and Advertising Passed Set powered off - Success Passed Set powered off - Class of Device Passed Set powered off - Invalid parameters 1 Passed Set powered off - Invalid parameters 2 Passed Set powered off - Invalid parameters 3 Passed Total: 11, Passed: 11 (100.0%), Failed: 0, Not Run: 0 Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 6 + net/bluetooth/hci_core.c | 45 ++- net/bluetooth/hci_request.c | 3 +- net/bluetooth/hci_sync.c | 584 ++++++++++++++++++++++++++++++- net/bluetooth/mgmt.c | 151 ++++---- 5 files changed, 688 insertions(+), 101 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index ec727eb18e90..00203077e656 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -66,3 +66,9 @@ int hci_disable_advertising_sync(struct hci_dev *hdev); int hci_update_passive_scan_sync(struct hci_dev *hdev); int hci_update_passive_scan(struct hci_dev *hdev); + +int hci_dev_open_sync(struct hci_dev *hdev); +int hci_dev_close_sync(struct hci_dev *hdev); + +int hci_powered_update_sync(struct hci_dev *hdev); +int hci_set_powered_sync(struct hci_dev *hdev, u8 val); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 8c54b1d4d41b..420ed6a02337 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -1315,14 +1315,13 @@ static void hci_dev_get_bd_addr_from_property(struct hci_dev *hdev) bacpy(&hdev->public_addr, &ba); } -static int hci_dev_do_open(struct hci_dev *hdev) +/* TODO: Move this function into hci_sync.c */ +int hci_dev_open_sync(struct hci_dev *hdev) { int ret = 0; BT_DBG("%s %p", hdev->name, hdev); - hci_req_sync_lock(hdev); - if (hci_dev_test_flag(hdev, HCI_UNREGISTER)) { ret = -ENODEV; goto done; @@ -1489,8 +1488,7 @@ setup_failed: !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && hci_dev_test_flag(hdev, HCI_MGMT) && hdev->dev_type == HCI_PRIMARY) { - ret = __hci_req_hci_power_on(hdev); - mgmt_power_on(hdev, ret); + ret = hci_powered_update_sync(hdev); } } else { /* Init failed, cleanup */ @@ -1522,6 +1520,19 @@ setup_failed: } done: + return ret; +} + +static int hci_dev_do_open(struct hci_dev *hdev) +{ + int ret = 0; + + BT_DBG("%s %p", hdev->name, hdev); + + hci_req_sync_lock(hdev); + + ret = hci_dev_open_sync(hdev); + hci_req_sync_unlock(hdev); return ret; } @@ -1600,7 +1611,8 @@ static void hci_pend_le_actions_clear(struct hci_dev *hdev) BT_DBG("All LE pending actions cleared"); } -int hci_dev_do_close(struct hci_dev *hdev) +/* TODO: Move this function into hci_sync.c */ +int hci_dev_close_sync(struct hci_dev *hdev) { bool auto_off; int err = 0; @@ -1611,7 +1623,6 @@ int hci_dev_do_close(struct hci_dev *hdev) cancel_delayed_work(&hdev->ncmd_timer); hci_request_cancel_all(hdev); - hci_req_sync_lock(hdev); if (!hci_dev_test_flag(hdev, HCI_UNREGISTER) && !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && @@ -1623,7 +1634,6 @@ int hci_dev_do_close(struct hci_dev *hdev) if (!test_and_clear_bit(HCI_UP, &hdev->flags)) { cancel_delayed_work_sync(&hdev->cmd_timer); - hci_req_sync_unlock(hdev); return err; } @@ -1729,9 +1739,22 @@ int hci_dev_do_close(struct hci_dev *hdev) bacpy(&hdev->random_addr, BDADDR_ANY); hci_codec_list_clear(&hdev->local_codecs); + hci_dev_put(hdev); + return err; +} + +int hci_dev_do_close(struct hci_dev *hdev) +{ + int err; + + BT_DBG("%s %p", hdev->name, hdev); + + hci_req_sync_lock(hdev); + + err = hci_dev_close_sync(hdev); + hci_req_sync_unlock(hdev); - hci_dev_put(hdev); return err; } @@ -2133,9 +2156,7 @@ static void hci_power_on(struct work_struct *work) hci_dev_test_flag(hdev, HCI_MGMT) && hci_dev_test_and_clear_flag(hdev, HCI_AUTO_OFF)) { cancel_delayed_work(&hdev->power_off); - hci_req_sync_lock(hdev); - err = __hci_req_hci_power_on(hdev); - hci_req_sync_unlock(hdev); + err = hci_powered_update_sync(hdev); mgmt_power_on(hdev, err); return; } diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index cb36b2bd1109..46fa9c3bdb3e 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -1794,7 +1794,8 @@ int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance) hci_req_add(req, HCI_OP_LE_SET_EXT_ADV_PARAMS, sizeof(cp), &cp); - if (own_addr_type == ADDR_LE_DEV_RANDOM && + if ((own_addr_type == ADDR_LE_DEV_RANDOM || + own_addr_type == ADDR_LE_DEV_RANDOM_RESOLVED) && bacmp(&random_addr, BDADDR_ANY)) { struct hci_cp_le_set_adv_set_rand_addr cp; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 6d53a876de85..ee5de1f8d133 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -739,6 +739,17 @@ int hci_setup_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance) cp.evt_properties = cpu_to_le16(LE_LEGACY_NONCONN_IND); } + /* If Own_Address_Type equals 0x02 or 0x03, the Peer_Address parameter + * contains the peer’s Identity Address and the Peer_Address_Type + * parameter contains the peer’s Identity Type (i.e., 0x00 or 0x01). + * These parameters are used to locate the corresponding local IRK in + * the resolving list; this IRK is used to generate their own address + * used in the advertisement. + */ + if (own_addr_type == ADDR_LE_DEV_RANDOM_RESOLVED) + hci_copy_identity_address(hdev, &cp.peer_addr, + &cp.peer_addr_type); + cp.own_addr_type = own_addr_type; cp.channel_map = hdev->le_adv_channel_map; cp.handle = instance; @@ -1105,8 +1116,7 @@ int hci_schedule_adv_instance_sync(struct hci_dev *hdev, u8 instance, struct adv_info *adv = NULL; u16 timeout; - if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || - list_empty(&hdev->adv_instances)) + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) && !ext_adv_capable(hdev)) return -EPERM; if (hdev->adv_instance_timeout) @@ -1474,7 +1484,9 @@ static int hci_le_del_accept_list_sync(struct hci_dev *hdev, return 0; } -/* Adds connection to resolve list if needed.*/ +/* Adds connection to resolve list if needed. + * Setting params to NULL programs local hdev->irk + */ static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, struct hci_conn_params *params) { @@ -1485,6 +1497,18 @@ static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, if (!use_ll_privacy(hdev)) return 0; + /* Attempt to program local identity address, type and irk if params is + * NULL. + */ + if (!params) { + if (!hci_dev_test_flag(hdev, HCI_PRIVACY)) + return 0; + + hci_copy_identity_address(hdev, &cp.bdaddr, &cp.bdaddr_type); + memcpy(cp.peer_irk, hdev->irk, 16); + goto done; + } + irk = hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type); if (!irk) return 0; @@ -1500,6 +1524,7 @@ static int hci_le_add_resolve_list_sync(struct hci_dev *hdev, bacpy(&cp.bdaddr, ¶ms->addr); memcpy(cp.peer_irk, irk->val, 16); +done: if (hci_dev_test_flag(hdev, HCI_PRIVACY)) memcpy(cp.local_irk, hdev->irk, 16); else @@ -2040,3 +2065,556 @@ int hci_update_passive_scan(struct hci_dev *hdev) return hci_cmd_sync_queue(hdev, update_passive_scan_sync, NULL, NULL); } + +static int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val) +{ + if (!bredr_sc_enabled(hdev) || lmp_host_sc_capable(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SC_SUPPORT, + sizeof(val), &val, HCI_CMD_TIMEOUT); +} + +static int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) +{ + int err; + + if (!hci_dev_test_flag(hdev, HCI_SSP_ENABLED) || + lmp_host_ssp_capable(hdev)) + return 0; + + err = __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SSP_MODE, + sizeof(mode), &mode, HCI_CMD_TIMEOUT); + if (err) + return err; + + return hci_write_sc_support_sync(hdev, 0x01); +} + +static int hci_write_le_host_supported_sync(struct hci_dev *hdev, u8 le, + u8 simul) +{ + struct hci_cp_write_le_host_supported cp; + + if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED) || + !lmp_bredr_capable(hdev)) + return 0; + + /* Check first if we already have the right host state + * (host features set) + */ + if (le == lmp_host_le_capable(hdev) && + simul == lmp_host_le_br_capable(hdev)) + return 0; + + memset(&cp, 0, sizeof(cp)); + + cp.le = le; + cp.simul = simul; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_LE_HOST_SUPPORTED, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_powered_update_adv_sync(struct hci_dev *hdev) +{ + struct adv_info *adv, *tmp; + int err; + + if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED)) + return 0; + + /* If RPA Resolution has not been enable yet it means the + * resolving list is empty and we should attempt to program the + * local IRK in order to support using own_addr_type + * ADDR_LE_DEV_RANDOM_RESOLVED (0x03). + */ + if (!hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) { + hci_le_add_resolve_list_sync(hdev, NULL); + hci_le_set_addr_resolution_enable_sync(hdev, 0x01); + } + + /* Make sure the controller has a good default for + * advertising data. This also applies to the case + * where BR/EDR was toggled during the AUTO_OFF phase. + */ + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || + list_empty(&hdev->adv_instances)) { + if (ext_adv_capable(hdev)) { + err = hci_setup_ext_adv_instance_sync(hdev, 0x00); + if (!err) + hci_update_scan_rsp_data_sync(hdev, 0x00); + } else { + err = hci_update_adv_data_sync(hdev, 0x00); + if (!err) + hci_update_scan_rsp_data_sync(hdev, 0x00); + } + + if (hci_dev_test_flag(hdev, HCI_ADVERTISING)) + hci_enable_advertising_sync(hdev); + } + + /* Call for each tracked instance to be scheduled */ + list_for_each_entry_safe(adv, tmp, &hdev->adv_instances, list) + hci_schedule_adv_instance_sync(hdev, adv->instance, true); + + return 0; +} + +static int hci_write_auth_enable_sync(struct hci_dev *hdev) +{ + u8 link_sec; + + link_sec = hci_dev_test_flag(hdev, HCI_LINK_SECURITY); + if (link_sec == test_bit(HCI_AUTH, &hdev->flags)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_AUTH_ENABLE, + sizeof(link_sec), &link_sec, + HCI_CMD_TIMEOUT); +} + +static int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable) +{ + struct hci_cp_write_page_scan_activity cp; + u8 type; + int err = 0; + + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + return 0; + + if (hdev->hci_ver < BLUETOOTH_VER_1_2) + return 0; + + memset(&cp, 0, sizeof(cp)); + + if (enable) { + type = PAGE_SCAN_TYPE_INTERLACED; + + /* 160 msec page scan interval */ + cp.interval = cpu_to_le16(0x0100); + } else { + type = hdev->def_page_scan_type; + cp.interval = cpu_to_le16(hdev->def_page_scan_int); + } + + cp.window = cpu_to_le16(hdev->def_page_scan_window); + + if (__cpu_to_le16(hdev->page_scan_interval) != cp.interval || + __cpu_to_le16(hdev->page_scan_window) != cp.window) { + err = __hci_cmd_sync_status(hdev, + HCI_OP_WRITE_PAGE_SCAN_ACTIVITY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (err) + return err; + } + + if (hdev->page_scan_type != type) + err = __hci_cmd_sync_status(hdev, + HCI_OP_WRITE_PAGE_SCAN_TYPE, + sizeof(type), &type, + HCI_CMD_TIMEOUT); + + return err; +} + +static bool disconnected_accept_list_entries(struct hci_dev *hdev) +{ + struct bdaddr_list *b; + + list_for_each_entry(b, &hdev->accept_list, list) { + struct hci_conn *conn; + + conn = hci_conn_hash_lookup_ba(hdev, ACL_LINK, &b->bdaddr); + if (!conn) + return true; + + if (conn->state != BT_CONNECTED && conn->state != BT_CONFIG) + return true; + } + + return false; +} + +static int hci_write_scan_enable_sync(struct hci_dev *hdev, u8 val) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SCAN_ENABLE, + sizeof(val), &val, + HCI_CMD_TIMEOUT); +} + +static int hci_update_scan_sync(struct hci_dev *hdev) +{ + u8 scan; + + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + return 0; + + if (!hdev_is_powered(hdev)) + return 0; + + if (mgmt_powering_down(hdev)) + return 0; + + if (hdev->scanning_paused) + return 0; + + if (hci_dev_test_flag(hdev, HCI_CONNECTABLE) || + disconnected_accept_list_entries(hdev)) + scan = SCAN_PAGE; + else + scan = SCAN_DISABLED; + + if (hci_dev_test_flag(hdev, HCI_DISCOVERABLE)) + scan |= SCAN_INQUIRY; + + if (test_bit(HCI_PSCAN, &hdev->flags) == !!(scan & SCAN_PAGE) && + test_bit(HCI_ISCAN, &hdev->flags) == !!(scan & SCAN_INQUIRY)) + return 0; + + return hci_write_scan_enable_sync(hdev, scan); +} + +static int hci_update_name_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_local_name cp; + + memset(&cp, 0, sizeof(cp)); + + memcpy(cp.name, hdev->dev_name, sizeof(cp.name)); + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_LOCAL_NAME, + sizeof(cp), &cp, + HCI_CMD_TIMEOUT); +} + +/* This function perform powered update HCI command sequence after the HCI init + * sequence which end up resetting all states, the sequence is as follows: + * + * HCI_SSP_ENABLED(Enable SSP) + * HCI_LE_ENABLED(Enable LE) + * HCI_LE_ENABLED(use_ll_privacy(Add local IRK to Resolving List) -> + * Update adv data) + * Enable Authentication + * lmp_bredr_capable(Set Fast Connectable -> Set Scan Type -> Set Class -> + * Set Name -> Set EIR) + */ +int hci_powered_update_sync(struct hci_dev *hdev) +{ + int err; + + /* Register the available SMP channels (BR/EDR and LE) only when + * successfully powering on the controller. This late + * registration is required so that LE SMP can clearly decide if + * the public address or static address is used. + */ + smp_register(hdev); + + err = hci_write_ssp_mode_sync(hdev, 0x01); + if (err) + return err; + + err = hci_write_le_host_supported_sync(hdev, 0x01, 0x00); + if (err) + return err; + + err = hci_powered_update_adv_sync(hdev); + if (err) + return err; + + err = hci_write_auth_enable_sync(hdev); + if (err) + return err; + + if (lmp_bredr_capable(hdev)) { + if (hci_dev_test_flag(hdev, HCI_FAST_CONNECTABLE)) + hci_write_fast_connectable_sync(hdev, true); + else + hci_write_fast_connectable_sync(hdev, false); + hci_update_scan_sync(hdev); + hci_update_class_sync(hdev); + hci_update_name_sync(hdev); + hci_update_eir_sync(hdev); + } + + return 0; +} + +/* This function perform power on HCI command sequence as follows: + * + * If controller is already up (HCI_UP) performs hci_powered_update_sync + * sequence otherwise run hci_dev_open_sync which will follow with + * hci_powered_update_sync after the init sequence is completed. + */ +static int hci_power_on_sync(struct hci_dev *hdev) +{ + int err; + + if (test_bit(HCI_UP, &hdev->flags) && + hci_dev_test_flag(hdev, HCI_MGMT) && + hci_dev_test_and_clear_flag(hdev, HCI_AUTO_OFF)) { + cancel_delayed_work(&hdev->power_off); + return hci_powered_update_sync(hdev); + } + + err = hci_dev_open_sync(hdev); + if (err < 0) + return err; + + /* During the HCI setup phase, a few error conditions are + * ignored and they need to be checked now. If they are still + * valid, it is important to return the device back off. + */ + if (hci_dev_test_flag(hdev, HCI_RFKILLED) || + hci_dev_test_flag(hdev, HCI_UNCONFIGURED) || + (hdev->dev_type == HCI_PRIMARY && + !bacmp(&hdev->bdaddr, BDADDR_ANY) && + !bacmp(&hdev->static_addr, BDADDR_ANY))) { + hci_dev_clear_flag(hdev, HCI_AUTO_OFF); + hci_dev_close_sync(hdev); + } else if (hci_dev_test_flag(hdev, HCI_AUTO_OFF)) { + queue_delayed_work(hdev->req_workqueue, &hdev->power_off, + HCI_AUTO_OFF_TIMEOUT); + } + + if (hci_dev_test_and_clear_flag(hdev, HCI_SETUP)) { + /* For unconfigured devices, set the HCI_RAW flag + * so that userspace can easily identify them. + */ + if (hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) + set_bit(HCI_RAW, &hdev->flags); + + /* For fully configured devices, this will send + * the Index Added event. For unconfigured devices, + * it will send Unconfigued Index Added event. + * + * Devices with HCI_QUIRK_RAW_DEVICE are ignored + * and no event will be send. + */ + mgmt_index_added(hdev); + } else if (hci_dev_test_and_clear_flag(hdev, HCI_CONFIG)) { + /* When the controller is now configured, then it + * is important to clear the HCI_RAW flag. + */ + if (!hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) + clear_bit(HCI_RAW, &hdev->flags); + + /* Powering on the controller with HCI_CONFIG set only + * happens with the transition from unconfigured to + * configured. This will send the Index Added event. + */ + mgmt_index_added(hdev); + } + + return 0; +} + +static int hci_remote_name_cancel_sync(struct hci_dev *hdev, bdaddr_t *addr) +{ + struct hci_cp_remote_name_req_cancel cp; + + memset(&cp, 0, sizeof(cp)); + bacpy(&cp.bdaddr, addr); + + return __hci_cmd_sync_status(hdev, HCI_OP_REMOTE_NAME_REQ_CANCEL, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_stop_discovery_sync(struct hci_dev *hdev) +{ + struct discovery_state *d = &hdev->discovery; + struct inquiry_entry *e; + int err; + + bt_dev_dbg(hdev, "state %u", hdev->discovery.state); + + if (d->state == DISCOVERY_FINDING || d->state == DISCOVERY_STOPPING) { + if (test_bit(HCI_INQUIRY, &hdev->flags)) { + err = __hci_cmd_sync_status(hdev, HCI_OP_INQUIRY_CANCEL, + 0, NULL, HCI_CMD_TIMEOUT); + if (err) + return err; + } + + if (hci_dev_test_flag(hdev, HCI_LE_SCAN)) { + cancel_delayed_work(&hdev->le_scan_disable); + cancel_delayed_work(&hdev->le_scan_restart); + + err = hci_scan_disable_sync(hdev); + if (err) + return err; + } + + } else { + err = hci_scan_disable_sync(hdev); + if (err) + return err; + } + + /* No further actions needed for LE-only discovery */ + if (d->type == DISCOV_TYPE_LE) + return 0; + + if (d->state == DISCOVERY_RESOLVING || d->state == DISCOVERY_STOPPING) { + e = hci_inquiry_cache_lookup_resolve(hdev, BDADDR_ANY, + NAME_PENDING); + if (!e) + return 0; + + return hci_remote_name_cancel_sync(hdev, &e->data.bdaddr); + } + + return 0; +} + +static int hci_disconnect_phy_link_sync(struct hci_dev *hdev, u16 handle, + u8 reason) +{ + struct hci_cp_disconn_phy_link cp; + + memset(&cp, 0, sizeof(cp)); + cp.phy_handle = HCI_PHY_HANDLE(handle); + cp.reason = reason; + + return __hci_cmd_sync_status(hdev, HCI_OP_DISCONN_PHY_LINK, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_disconnect_sync(struct hci_dev *hdev, struct hci_conn *conn, + u8 reason) +{ + struct hci_cp_disconnect cp; + + if (conn->type == AMP_LINK) + return hci_disconnect_phy_link_sync(hdev, conn->handle, reason); + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(conn->handle); + cp.reason = reason; + + return __hci_cmd_sync_status(hdev, HCI_OP_DISCONNECT, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_connect_cancel_sync(struct hci_dev *hdev, + struct hci_conn *conn) +{ + if (test_bit(HCI_CONN_SCANNING, &conn->flags)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_CREATE_CONN_CANCEL, + 6, &conn->dst, HCI_CMD_TIMEOUT); +} + +static int hci_connect_cancel_sync(struct hci_dev *hdev, struct hci_conn *conn) +{ + if (conn->type == LE_LINK) + return hci_le_connect_cancel_sync(hdev, conn); + + if (hdev->hci_ver < BLUETOOTH_VER_1_2) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_CREATE_CONN_CANCEL, + 6, &conn->dst, HCI_CMD_TIMEOUT); +} + +static int hci_reject_sco_sync(struct hci_dev *hdev, struct hci_conn *conn, + u8 reason) +{ + struct hci_cp_reject_sync_conn_req cp; + + memset(&cp, 0, sizeof(cp)); + bacpy(&cp.bdaddr, &conn->dst); + cp.reason = reason; + + /* SCO rejection has its own limited set of + * allowed error values (0x0D-0x0F). + */ + if (reason < 0x0d || reason > 0x0f) + cp.reason = HCI_ERROR_REJ_LIMITED_RESOURCES; + + return __hci_cmd_sync_status(hdev, HCI_OP_REJECT_SYNC_CONN_REQ, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_reject_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, + u8 reason) +{ + struct hci_cp_reject_conn_req cp; + + if (conn->type == SCO_LINK || conn->type == ESCO_LINK) + return hci_reject_sco_sync(hdev, conn, reason); + + memset(&cp, 0, sizeof(cp)); + bacpy(&cp.bdaddr, &conn->dst); + cp.reason = reason; + + return __hci_cmd_sync_status(hdev, HCI_OP_REJECT_CONN_REQ, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, + u8 reason) +{ + switch (conn->state) { + case BT_CONNECTED: + case BT_CONFIG: + return hci_disconnect_sync(hdev, conn, reason); + case BT_CONNECT: + return hci_connect_cancel_sync(hdev, conn); + case BT_CONNECT2: + return hci_reject_conn_sync(hdev, conn, reason); + default: + conn->state = BT_CLOSED; + break; + } + + return 0; +} + +/* This function perform power off HCI command sequence as follows: + * + * Clear Advertising + * Stop Discovery + * Disconnect all connections + * hci_dev_close_sync + */ +static int hci_power_off_sync(struct hci_dev *hdev) +{ + struct hci_conn *conn; + int err; + + /* If controller is already down there is nothing to do */ + if (!test_bit(HCI_UP, &hdev->flags)) + return 0; + + if (test_bit(HCI_ISCAN, &hdev->flags) || + test_bit(HCI_PSCAN, &hdev->flags)) { + err = hci_write_scan_enable_sync(hdev, 0x00); + if (err) + return err; + } + + err = hci_clear_adv_sync(hdev, NULL, false); + if (err) + return err; + + err = hci_stop_discovery_sync(hdev); + if (err) + return err; + + list_for_each_entry(conn, &hdev->conn_hash.list, list) { + /* 0x15 == Terminated due to Power Off */ + hci_abort_conn_sync(hdev, conn, 0x15); + } + + return hci_dev_close_sync(hdev); +} + +int hci_set_powered_sync(struct hci_dev *hdev, u8 val) +{ + if (val) + return hci_power_on_sync(hdev); + + return hci_power_off_sync(hdev); +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index d09a757a5851..9490663ee3fb 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1155,16 +1155,6 @@ static int send_settings_rsp(struct sock *sk, u16 opcode, struct hci_dev *hdev) sizeof(settings)); } -static void clean_up_hci_complete(struct hci_dev *hdev, u8 status, u16 opcode) -{ - bt_dev_dbg(hdev, "status 0x%02x", status); - - if (hci_conn_count(hdev) == 0) { - cancel_delayed_work(&hdev->power_off); - queue_work(hdev->req_workqueue, &hdev->power_off.work); - } -} - void mgmt_advertising_added(struct sock *sk, struct hci_dev *hdev, u8 instance) { struct mgmt_ev_advertising_added ev; @@ -1192,38 +1182,77 @@ static void cancel_adv_timeout(struct hci_dev *hdev) } } -static int clean_up_hci_state(struct hci_dev *hdev) +/* This function requires the caller holds hdev->lock */ +static void restart_le_actions(struct hci_dev *hdev) { - struct hci_request req; - struct hci_conn *conn; - bool discov_stopped; - int err; + struct hci_conn_params *p; - hci_req_init(&req, hdev); + list_for_each_entry(p, &hdev->le_conn_params, list) { + /* Needed for AUTO_OFF case where might not "really" + * have been powered off. + */ + list_del_init(&p->action); - if (test_bit(HCI_ISCAN, &hdev->flags) || - test_bit(HCI_PSCAN, &hdev->flags)) { - u8 scan = 0x00; - hci_req_add(&req, HCI_OP_WRITE_SCAN_ENABLE, 1, &scan); + switch (p->auto_connect) { + case HCI_AUTO_CONN_DIRECT: + case HCI_AUTO_CONN_ALWAYS: + list_add(&p->action, &hdev->pend_le_conns); + break; + case HCI_AUTO_CONN_REPORT: + list_add(&p->action, &hdev->pend_le_reports); + break; + default: + break; + } + } +} + +static int new_settings(struct hci_dev *hdev, struct sock *skip) +{ + __le32 ev = cpu_to_le32(get_current_settings(hdev)); + + return mgmt_limited_event(MGMT_EV_NEW_SETTINGS, hdev, &ev, + sizeof(ev), HCI_MGMT_SETTING_EVENTS, skip); +} + +static void mgmt_set_powered_complete(struct hci_dev *hdev, void *data, int err) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + + bt_dev_dbg(hdev, "err %d", err); + + if (!err) { + if (cp->val) { + hci_dev_lock(hdev); + restart_le_actions(hdev); + hci_update_passive_scan(hdev); + hci_dev_unlock(hdev); + } + + send_settings_rsp(cmd->sk, cmd->opcode, hdev); + + /* Only call new_setting for power on as power off is deferred + * to hdev->power_off work which does call hci_dev_do_close. + */ + if (cp->val) + new_settings(hdev, cmd->sk); + } else { + mgmt_cmd_status(cmd->sk, hdev->id, MGMT_OP_SET_POWERED, + mgmt_status(err)); } - hci_req_clear_adv_instance(hdev, NULL, NULL, 0x00, false); + mgmt_pending_free(cmd); +} - if (hci_dev_test_flag(hdev, HCI_LE_ADV)) - __hci_req_disable_advertising(&req); +static int set_powered_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; - discov_stopped = hci_req_stop_discovery(&req); + BT_DBG("%s", hdev->name); - list_for_each_entry(conn, &hdev->conn_hash.list, list) { - /* 0x15 == Terminated due to Power Off */ - __hci_abort_conn(&req, conn, 0x15); - } - - err = hci_req_run(&req, clean_up_hci_complete); - if (!err && discov_stopped) - hci_discovery_set_state(hdev, DISCOVERY_STOPPING); - - return err; + return hci_set_powered_sync(hdev, cp->val); } static int set_powered(struct sock *sk, struct hci_dev *hdev, void *data, @@ -1252,43 +1281,20 @@ static int set_powered(struct sock *sk, struct hci_dev *hdev, void *data, goto failed; } - cmd = mgmt_pending_add(sk, MGMT_OP_SET_POWERED, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_SET_POWERED, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; } - if (cp->val) { - queue_work(hdev->req_workqueue, &hdev->power_on); - err = 0; - } else { - /* Disconnect connections, stop scans, etc */ - err = clean_up_hci_state(hdev); - if (!err) - queue_delayed_work(hdev->req_workqueue, &hdev->power_off, - HCI_POWER_OFF_TIMEOUT); - - /* ENODATA means there were no HCI commands queued */ - if (err == -ENODATA) { - cancel_delayed_work(&hdev->power_off); - queue_work(hdev->req_workqueue, &hdev->power_off.work); - err = 0; - } - } + err = hci_cmd_sync_queue(hdev, set_powered_sync, cmd, + mgmt_set_powered_complete); failed: hci_dev_unlock(hdev); return err; } -static int new_settings(struct hci_dev *hdev, struct sock *skip) -{ - __le32 ev = cpu_to_le32(get_current_settings(hdev)); - - return mgmt_limited_event(MGMT_EV_NEW_SETTINGS, hdev, &ev, - sizeof(ev), HCI_MGMT_SETTING_EVENTS, skip); -} - int mgmt_new_settings(struct hci_dev *hdev) { return new_settings(hdev, NULL); @@ -8720,31 +8726,6 @@ void mgmt_index_removed(struct hci_dev *hdev) HCI_MGMT_EXT_INDEX_EVENTS); } -/* This function requires the caller holds hdev->lock */ -static void restart_le_actions(struct hci_dev *hdev) -{ - struct hci_conn_params *p; - - list_for_each_entry(p, &hdev->le_conn_params, list) { - /* Needed for AUTO_OFF case where might not "really" - * have been powered off. - */ - list_del_init(&p->action); - - switch (p->auto_connect) { - case HCI_AUTO_CONN_DIRECT: - case HCI_AUTO_CONN_ALWAYS: - list_add(&p->action, &hdev->pend_le_conns); - break; - case HCI_AUTO_CONN_REPORT: - list_add(&p->action, &hdev->pend_le_reports); - break; - default: - break; - } - } -} - void mgmt_power_on(struct hci_dev *hdev, int err) { struct cmd_lookup match = { NULL, hdev }; From abfeea476c68afea54c9c050a2d3b19d5d2ee873 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:45 -0700 Subject: [PATCH 0038/2048] Bluetooth: hci_sync: Convert MGMT_OP_START_DISCOVERY This make use of hci_cmd_sync_queue for MGMT_OP_START_DISCOVERY, MGMT_OP_START_SERVICE_DISCOVERY and MGMT_OP_STOP_DISCOVERY to use hci_cmd_sync_queue so they no longer depend on hdev->discov_update work to send any commands. Tested with: tools/mgmt-tester -s "Start Discovery" Test Summary ------------ Start Discovery - Not powered 1 Passed Start Discovery - Invalid parameters 1 Passed Start Discovery - Not supported 1 Passed Start Discovery - Success 1 Passed Start Discovery - Success 2 Passed Start Discovery - Power Off 1 Passed Start Discovery BREDR LE - (Ext Scan Enable) Passed Start Discovery LE - (Ext Scan Enable) Passed Start Discovery LE - (Ext Scan Param) Passed Start Discovery - (2m, Scan Param) Passed Start Discovery - (coded, Scan Param) Passed Start Discovery - (1m, 2m, coded, Scan Param) Passed LL Privacy - Start Discovery 1 (Disable RL) Passed LL Privacy - Start Discovery 2 (Disable RL) Passed Total: 14, Passed: 14 (100.0%), Failed: 0, Not Run: 0 tools/mgmt-tester -s "Start Service" Test Summary ------------ Start Service Discovery - Not powered 1 Passed Start Service Discovery - Invalid parameters 1 Passed Start Service Discovery - Not supported 1 Passed Start Service Discovery - Success 1 Passed Start Service Discovery - Success 2 Passed Total: 5, Passed: 5 (100.0%), Failed: 0, Not Run: 0 tools/mgmt-tester -s "Stop Discovery" Test Summary ------------ Stop Discovery - Success 1 Passed Stop Discovery - BR/EDR (Inquiry) Success 1 Passed Stop Discovery - Rejected 1 Passed Stop Discovery - Invalid parameters 1 Passed Stop Discovery - (Ext Scan Disable) Passed Total: 5, Passed: 5 (100.0%), Failed: 0, Not Run: 0 Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 4 + net/bluetooth/hci_conn.c | 4 +- net/bluetooth/hci_event.c | 15 +- net/bluetooth/hci_sync.c | 238 +++++++++++++++++++++++++++++-- net/bluetooth/mgmt.c | 105 ++++++++++---- 5 files changed, 310 insertions(+), 56 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 00203077e656..c4fa77321b31 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -59,6 +59,7 @@ int hci_setup_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance); int hci_start_ext_adv_sync(struct hci_dev *hdev, u8 instance); int hci_enable_ext_advertising_sync(struct hci_dev *hdev, u8 instance); int hci_enable_advertising_sync(struct hci_dev *hdev); +int hci_enable_advertising(struct hci_dev *hdev); int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, u8 instance, bool force); @@ -72,3 +73,6 @@ int hci_dev_close_sync(struct hci_dev *hdev); int hci_powered_update_sync(struct hci_dev *hdev); int hci_set_powered_sync(struct hci_dev *hdev, u8 val); + +int hci_start_discovery_sync(struct hci_dev *hdev); +int hci_stop_discovery_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 1783ec5f6e3e..dbd737b62f98 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -915,10 +915,10 @@ void hci_le_conn_failed(struct hci_conn *conn, u8 status) */ hci_update_passive_scan(hdev); - /* Re-enable advertising in case this was a failed connection + /* Enable advertising in case this was a failed connection * attempt as a peripheral. */ - hci_req_reenable_advertising(hdev); + hci_enable_advertising(hdev); } static void create_le_conn_complete(struct hci_dev *hdev, u8 status, u16 opcode) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index d5e4a82b7020..64008697aa23 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -1514,16 +1514,10 @@ static void le_set_scan_enable_complete(struct hci_dev *hdev, u8 enable) /* The HCI_LE_SCAN_INTERRUPTED flag indicates that we * interrupted scanning due to a connect request. Mark - * therefore discovery as stopped. If this was not - * because of a connect request advertising might have - * been disabled because of active scanning, so - * re-enable it again if necessary. + * therefore discovery as stopped. */ if (hci_dev_test_and_clear_flag(hdev, HCI_LE_SCAN_INTERRUPTED)) hci_discovery_set_state(hdev, DISCOVERY_STOPPED); - else if (!hci_dev_test_flag(hdev, HCI_LE_ADV) && - hdev->discovery.state == DISCOVERY_FINDING) - hci_req_reenable_advertising(hdev); break; @@ -2440,7 +2434,7 @@ static void hci_cs_disconnect(struct hci_dev *hdev, u8 status) if (conn->type == LE_LINK && conn->role == HCI_ROLE_SLAVE) { hdev->cur_adv_instance = conn->adv_instance; - hci_req_reenable_advertising(hdev); + hci_enable_advertising(hdev); } /* If the disconnection failed for any reason, the upper layer @@ -3050,7 +3044,7 @@ static void hci_disconn_complete_evt(struct hci_dev *hdev, struct sk_buff *skb) */ if (conn->type == LE_LINK && conn->role == HCI_ROLE_SLAVE) { hdev->cur_adv_instance = conn->adv_instance; - hci_req_reenable_advertising(hdev); + hci_enable_advertising(hdev); } hci_conn_del(conn); @@ -5500,9 +5494,6 @@ static void hci_le_enh_conn_complete_evt(struct hci_dev *hdev, le16_to_cpu(ev->interval), le16_to_cpu(ev->latency), le16_to_cpu(ev->supervision_timeout)); - - if (hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION)) - hci_req_disable_address_resolution(hdev); } static void hci_le_ext_adv_term_evt(struct hci_dev *hdev, struct sk_buff *skb) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index ee5de1f8d133..2f988d7f0008 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -189,9 +189,6 @@ struct sk_buff *__hci_cmd_sync_sk(struct hci_dev *hdev, u16 opcode, u32 plen, return ERR_PTR(err); } - if (!skb) - return ERR_PTR(-ENODATA); - return skb; } EXPORT_SYMBOL(__hci_cmd_sync_sk); @@ -241,12 +238,19 @@ int __hci_cmd_sync_status_sk(struct hci_dev *hdev, u16 opcode, u32 plen, u8 status; skb = __hci_cmd_sync_sk(hdev, opcode, plen, param, event, timeout, sk); - if (IS_ERR_OR_NULL(skb)) { + if (IS_ERR(skb)) { bt_dev_err(hdev, "Opcode 0x%4x failed: %ld", opcode, PTR_ERR(skb)); return PTR_ERR(skb); } + /* If command return a status event skb will be set to NULL as there are + * no parameters, in case of failure IS_ERR(skb) would have be set to + * the actual error would be found with PTR_ERR(skb). + */ + if (!skb) + return 0; + status = skb->data[0]; kfree_skb(skb); @@ -1017,8 +1021,22 @@ int hci_enable_advertising_sync(struct hci_dev *hdev) sizeof(enable), &enable, HCI_CMD_TIMEOUT); } -static int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, - struct sock *sk) +static int enable_advertising_sync(struct hci_dev *hdev, void *data) +{ + return hci_enable_advertising_sync(hdev); +} + +int hci_enable_advertising(struct hci_dev *hdev) +{ + if (!hci_dev_test_flag(hdev, HCI_ADVERTISING) && + list_empty(&hdev->adv_instances)) + return 0; + + return hci_cmd_sync_queue(hdev, enable_advertising_sync, NULL, NULL); +} + +int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, + struct sock *sk) { int err; @@ -1622,7 +1640,7 @@ static int hci_pause_advertising_sync(struct hci_dev *hdev) return 0; } -/* This function disables all user advertising instances (excluding 0x00) */ +/* This function enables all user advertising instances (excluding 0x00) */ static int hci_resume_advertising_sync(struct hci_dev *hdev) { struct adv_info *adv, *tmp; @@ -1870,7 +1888,8 @@ static int hci_le_set_scan_param_sync(struct hci_dev *hdev, u8 type, } static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, - u16 window, u8 own_addr_type, u8 filter_policy) + u16 window, u8 own_addr_type, u8 filter_policy, + u8 filter_dup) { int err; @@ -1884,8 +1903,7 @@ static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, if (err) return err; - return hci_le_set_scan_enable_sync(hdev, LE_SCAN_ENABLE, - LE_SCAN_FILTER_DUP_ENABLE); + return hci_le_set_scan_enable_sync(hdev, LE_SCAN_ENABLE, filter_dup); } int hci_passive_scan_sync(struct hci_dev *hdev) @@ -1960,7 +1978,8 @@ int hci_passive_scan_sync(struct hci_dev *hdev) bt_dev_dbg(hdev, "LE passive scan with acceptlist = %d", filter_policy); return hci_start_scan_sync(hdev, LE_SCAN_PASSIVE, interval, window, - own_addr_type, filter_policy); + own_addr_type, filter_policy, + LE_SCAN_FILTER_DUP_ENABLE); } /* This function controls the passive scanning based on hdev->pend_le_conns @@ -2420,7 +2439,7 @@ static int hci_remote_name_cancel_sync(struct hci_dev *hdev, bdaddr_t *addr) sizeof(cp), &cp, HCI_CMD_TIMEOUT); } -static int hci_stop_discovery_sync(struct hci_dev *hdev) +int hci_stop_discovery_sync(struct hci_dev *hdev) { struct discovery_state *d = &hdev->discovery; struct inquiry_entry *e; @@ -2451,6 +2470,10 @@ static int hci_stop_discovery_sync(struct hci_dev *hdev) return err; } + /* Resume advertising if it was paused */ + if (use_ll_privacy(hdev)) + hci_resume_advertising_sync(hdev); + /* No further actions needed for LE-only discovery */ if (d->type == DISCOV_TYPE_LE) return 0; @@ -2618,3 +2641,194 @@ int hci_set_powered_sync(struct hci_dev *hdev, u8 val) return hci_power_off_sync(hdev); } + +static int hci_inquiry_sync(struct hci_dev *hdev, u8 length) +{ + const u8 giac[3] = { 0x33, 0x8b, 0x9e }; + const u8 liac[3] = { 0x00, 0x8b, 0x9e }; + struct hci_cp_inquiry cp; + + bt_dev_dbg(hdev, ""); + + if (hci_dev_test_flag(hdev, HCI_INQUIRY)) + return 0; + + hci_dev_lock(hdev); + hci_inquiry_cache_flush(hdev); + hci_dev_unlock(hdev); + + memset(&cp, 0, sizeof(cp)); + + if (hdev->discovery.limited) + memcpy(&cp.lap, liac, sizeof(cp.lap)); + else + memcpy(&cp.lap, giac, sizeof(cp.lap)); + + cp.length = length; + + return __hci_cmd_sync_status(hdev, HCI_OP_INQUIRY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_active_scan_sync(struct hci_dev *hdev, uint16_t interval) +{ + u8 own_addr_type; + /* Accept list is not used for discovery */ + u8 filter_policy = 0x00; + /* Default is to enable duplicates filter */ + u8 filter_dup = LE_SCAN_FILTER_DUP_ENABLE; + int err; + + bt_dev_dbg(hdev, ""); + + /* If controller is scanning, it means the passive scanning is + * running. Thus, we should temporarily stop it in order to set the + * discovery scanning parameters. + */ + err = hci_scan_disable_sync(hdev); + if (err) { + bt_dev_err(hdev, "Unable to disable scanning: %d", err); + return err; + } + + cancel_interleave_scan(hdev); + + /* Pause advertising since active scanning disables address resolution + * which advertising depend on in order to generate its RPAs. + */ + if (use_ll_privacy(hdev)) { + err = hci_pause_advertising_sync(hdev); + if (err) { + bt_dev_err(hdev, "pause advertising failed: %d", err); + goto failed; + } + } + + /* Disable address resolution while doing active scanning since the + * accept list shall not be used and all reports shall reach the host + * anyway. + */ + err = hci_le_set_addr_resolution_enable_sync(hdev, 0x00); + if (err) { + bt_dev_err(hdev, "Unable to disable Address Resolution: %d", + err); + goto failed; + } + + /* All active scans will be done with either a resolvable private + * address (when privacy feature has been enabled) or non-resolvable + * private address. + */ + err = hci_update_random_address_sync(hdev, true, scan_use_rpa(hdev), + &own_addr_type); + if (err < 0) + own_addr_type = ADDR_LE_DEV_PUBLIC; + + if (hci_is_adv_monitoring(hdev)) { + /* Duplicate filter should be disabled when some advertisement + * monitor is activated, otherwise AdvMon can only receive one + * advertisement for one peer(*) during active scanning, and + * might report loss to these peers. + * + * Note that different controllers have different meanings of + * |duplicate|. Some of them consider packets with the same + * address as duplicate, and others consider packets with the + * same address and the same RSSI as duplicate. Although in the + * latter case we don't need to disable duplicate filter, but + * it is common to have active scanning for a short period of + * time, the power impact should be neglectable. + */ + filter_dup = LE_SCAN_FILTER_DUP_DISABLE; + } + + err = hci_start_scan_sync(hdev, LE_SCAN_ACTIVE, interval, + hdev->le_scan_window_discovery, + own_addr_type, filter_policy, filter_dup); + if (!err) + return err; + +failed: + /* Resume advertising if it was paused */ + if (use_ll_privacy(hdev)) + hci_resume_advertising_sync(hdev); + + /* Resume passive scanning */ + hci_update_passive_scan_sync(hdev); + return err; +} + +static int hci_start_interleaved_discovery_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + err = hci_active_scan_sync(hdev, hdev->le_scan_int_discovery * 2); + if (err) + return err; + + return hci_inquiry_sync(hdev, DISCOV_BREDR_INQUIRY_LEN); +} + +int hci_start_discovery_sync(struct hci_dev *hdev) +{ + unsigned long timeout; + int err; + + bt_dev_dbg(hdev, "type %u", hdev->discovery.type); + + switch (hdev->discovery.type) { + case DISCOV_TYPE_BREDR: + return hci_inquiry_sync(hdev, DISCOV_BREDR_INQUIRY_LEN); + case DISCOV_TYPE_INTERLEAVED: + /* When running simultaneous discovery, the LE scanning time + * should occupy the whole discovery time sine BR/EDR inquiry + * and LE scanning are scheduled by the controller. + * + * For interleaving discovery in comparison, BR/EDR inquiry + * and LE scanning are done sequentially with separate + * timeouts. + */ + if (test_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, + &hdev->quirks)) { + timeout = msecs_to_jiffies(DISCOV_LE_TIMEOUT); + /* During simultaneous discovery, we double LE scan + * interval. We must leave some time for the controller + * to do BR/EDR inquiry. + */ + err = hci_start_interleaved_discovery_sync(hdev); + break; + } + + timeout = msecs_to_jiffies(hdev->discov_interleaved_timeout); + err = hci_active_scan_sync(hdev, hdev->le_scan_int_discovery); + break; + case DISCOV_TYPE_LE: + timeout = msecs_to_jiffies(DISCOV_LE_TIMEOUT); + err = hci_active_scan_sync(hdev, hdev->le_scan_int_discovery); + break; + default: + return -EINVAL; + } + + if (err) + return err; + + bt_dev_dbg(hdev, "timeout %u ms", jiffies_to_msecs(timeout)); + + /* When service discovery is used and the controller has a + * strict duplicate filter, it is important to remember the + * start and duration of the scan. This is required for + * restarting scanning during the discovery phase. + */ + if (test_bit(HCI_QUIRK_STRICT_DUPLICATE_FILTER, &hdev->quirks) && + hdev->discovery.result_filtering) { + hdev->discovery.scan_start = jiffies; + hdev->discovery.scan_duration = timeout; + } + + queue_delayed_work(hdev->req_workqueue, &hdev->le_scan_disable, + timeout); + + return 0; +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 9490663ee3fb..db2e5a2f4e03 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5111,6 +5111,32 @@ static bool discovery_type_is_valid(struct hci_dev *hdev, uint8_t type, return true; } +static void start_discovery_complete(struct hci_dev *hdev, void *data, int err) +{ + struct mgmt_pending_cmd *cmd = data; + + bt_dev_dbg(hdev, "err %d", err); + + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, mgmt_status(err), + cmd->param, 1); + mgmt_pending_free(cmd); + + /* Handle suspend notifier */ + if (test_and_clear_bit(SUSPEND_UNPAUSE_DISCOVERY, + hdev->suspend_tasks)) { + bt_dev_dbg(hdev, "Unpaused discovery"); + wake_up(&hdev->suspend_wait_q); + } + + hci_discovery_set_state(hdev, err ? DISCOVERY_STOPPED : + DISCOVERY_FINDING); +} + +static int start_discovery_sync(struct hci_dev *hdev, void *data) +{ + return hci_start_discovery_sync(hdev); +} + static int start_discovery_internal(struct sock *sk, struct hci_dev *hdev, u16 op, void *data, u16 len) { @@ -5162,17 +5188,20 @@ static int start_discovery_internal(struct sock *sk, struct hci_dev *hdev, else hdev->discovery.limited = false; - cmd = mgmt_pending_add(sk, op, hdev, data, len); + cmd = mgmt_pending_new(sk, op, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; } - cmd->cmd_complete = generic_cmd_complete; + err = hci_cmd_sync_queue(hdev, start_discovery_sync, cmd, + start_discovery_complete); + if (err < 0) { + mgmt_pending_free(cmd); + goto failed; + } hci_discovery_set_state(hdev, DISCOVERY_STARTING); - queue_work(hdev->req_workqueue, &hdev->discov_update); - err = 0; failed: hci_dev_unlock(hdev); @@ -5194,13 +5223,6 @@ static int start_limited_discovery(struct sock *sk, struct hci_dev *hdev, data, len); } -static int service_discovery_cmd_complete(struct mgmt_pending_cmd *cmd, - u8 status) -{ - return mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, status, - cmd->param, 1); -} - static int start_service_discovery(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -5269,15 +5291,13 @@ static int start_service_discovery(struct sock *sk, struct hci_dev *hdev, goto failed; } - cmd = mgmt_pending_add(sk, MGMT_OP_START_SERVICE_DISCOVERY, + cmd = mgmt_pending_new(sk, MGMT_OP_START_SERVICE_DISCOVERY, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; } - cmd->cmd_complete = service_discovery_cmd_complete; - /* Clear the discovery filter first to free any previously * allocated memory for the UUID list. */ @@ -5301,9 +5321,14 @@ static int start_service_discovery(struct sock *sk, struct hci_dev *hdev, } } + err = hci_cmd_sync_queue(hdev, start_discovery_sync, cmd, + start_discovery_complete); + if (err < 0) { + mgmt_pending_free(cmd); + goto failed; + } + hci_discovery_set_state(hdev, DISCOVERY_STARTING); - queue_work(hdev->req_workqueue, &hdev->discov_update); - err = 0; failed: hci_dev_unlock(hdev); @@ -5333,6 +5358,31 @@ void mgmt_stop_discovery_complete(struct hci_dev *hdev, u8 status) } } +static void stop_discovery_complete(struct hci_dev *hdev, void *data, int err) +{ + struct mgmt_pending_cmd *cmd = data; + + bt_dev_dbg(hdev, "err %d", err); + + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, mgmt_status(err), + cmd->param, 1); + mgmt_pending_free(cmd); + + /* Handle suspend notifier */ + if (test_and_clear_bit(SUSPEND_PAUSE_DISCOVERY, hdev->suspend_tasks)) { + bt_dev_dbg(hdev, "Paused discovery"); + wake_up(&hdev->suspend_wait_q); + } + + if (!err) + hci_discovery_set_state(hdev, DISCOVERY_STOPPED); +} + +static int stop_discovery_sync(struct hci_dev *hdev, void *data) +{ + return hci_stop_discovery_sync(hdev); +} + static int stop_discovery(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -5358,17 +5408,20 @@ static int stop_discovery(struct sock *sk, struct hci_dev *hdev, void *data, goto unlock; } - cmd = mgmt_pending_add(sk, MGMT_OP_STOP_DISCOVERY, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_STOP_DISCOVERY, hdev, data, len); if (!cmd) { err = -ENOMEM; goto unlock; } - cmd->cmd_complete = generic_cmd_complete; + err = hci_cmd_sync_queue(hdev, stop_discovery_sync, cmd, + stop_discovery_complete); + if (err < 0) { + mgmt_pending_free(cmd); + goto unlock; + } hci_discovery_set_state(hdev, DISCOVERY_STOPPING); - queue_work(hdev->req_workqueue, &hdev->discov_update); - err = 0; unlock: hci_dev_unlock(hdev); @@ -5803,16 +5856,8 @@ static int set_scan_params(struct sock *sk, struct hci_dev *hdev, * loaded. */ if (hci_dev_test_flag(hdev, HCI_LE_SCAN) && - hdev->discovery.state == DISCOVERY_STOPPED) { - struct hci_request req; - - hci_req_init(&req, hdev); - - hci_req_add_le_scan_disable(&req, false); - hci_req_add_le_passive_scan(&req); - - hci_req_run(&req, NULL); - } + hdev->discovery.state == DISCOVERY_STOPPED) + hci_update_passive_scan(hdev); hci_dev_unlock(hdev); From 353a0249c3f60365c55d53e2d068de4f43669a22 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:46 -0700 Subject: [PATCH 0039/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_FAST_CONNECTABLE This creates a synchronized Write Fast Connectable call and attaches it to the MGMT_OP_SET_FAST_CONNECTABLE management opcode. Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 2 + net/bluetooth/hci_sync.c | 2 +- net/bluetooth/mgmt.c | 72 ++++++++++++++------------------ 3 files changed, 34 insertions(+), 42 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index c4fa77321b31..1fb66b6f8a34 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -68,6 +68,8 @@ int hci_disable_advertising_sync(struct hci_dev *hdev); int hci_update_passive_scan_sync(struct hci_dev *hdev); int hci_update_passive_scan(struct hci_dev *hdev); +int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); + int hci_dev_open_sync(struct hci_dev *hdev); int hci_dev_close_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 2f988d7f0008..f7c30b4e9aca 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2193,7 +2193,7 @@ static int hci_write_auth_enable_sync(struct hci_dev *hdev) HCI_CMD_TIMEOUT); } -static int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable) +int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable) { struct hci_cp_write_page_scan_activity cp; u8 type; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index db2e5a2f4e03..098ce324f20f 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5864,22 +5864,15 @@ static int set_scan_params(struct sock *sk, struct hci_dev *hdev, return err; } -static void fast_connectable_complete(struct hci_dev *hdev, u8 status, - u16 opcode) +static void fast_connectable_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; - bt_dev_dbg(hdev, "status 0x%02x", status); + bt_dev_dbg(hdev, "err %d", err); - hci_dev_lock(hdev); - - cmd = pending_find(MGMT_OP_SET_FAST_CONNECTABLE, hdev); - if (!cmd) - goto unlock; - - if (status) { + if (err) { mgmt_cmd_status(cmd->sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, - mgmt_status(status)); + mgmt_status(err)); } else { struct mgmt_mode *cp = cmd->param; @@ -5892,10 +5885,15 @@ static void fast_connectable_complete(struct hci_dev *hdev, u8 status, new_settings(hdev, cmd->sk); } - mgmt_pending_remove(cmd); + mgmt_pending_free(cmd); +} -unlock: - hci_dev_unlock(hdev); +static int write_fast_connectable_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + + return hci_write_fast_connectable_sync(hdev, cp->val); } static int set_fast_connectable(struct sock *sk, struct hci_dev *hdev, @@ -5903,58 +5901,50 @@ static int set_fast_connectable(struct sock *sk, struct hci_dev *hdev, { struct mgmt_mode *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED) || hdev->hci_ver < BLUETOOTH_VER_1_2) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, + return mgmt_cmd_status(sk, hdev->id, + MGMT_OP_SET_FAST_CONNECTABLE, MGMT_STATUS_NOT_SUPPORTED); if (cp->val != 0x00 && cp->val != 0x01) - return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, + return mgmt_cmd_status(sk, hdev->id, + MGMT_OP_SET_FAST_CONNECTABLE, MGMT_STATUS_INVALID_PARAMS); hci_dev_lock(hdev); - if (pending_find(MGMT_OP_SET_FAST_CONNECTABLE, hdev)) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, - MGMT_STATUS_BUSY); - goto unlock; - } - if (!!cp->val == hci_dev_test_flag(hdev, HCI_FAST_CONNECTABLE)) { - err = send_settings_rsp(sk, MGMT_OP_SET_FAST_CONNECTABLE, - hdev); + err = send_settings_rsp(sk, MGMT_OP_SET_FAST_CONNECTABLE, hdev); goto unlock; } if (!hdev_is_powered(hdev)) { hci_dev_change_flag(hdev, HCI_FAST_CONNECTABLE); - err = send_settings_rsp(sk, MGMT_OP_SET_FAST_CONNECTABLE, - hdev); + err = send_settings_rsp(sk, MGMT_OP_SET_FAST_CONNECTABLE, hdev); new_settings(hdev, sk); goto unlock; } - cmd = mgmt_pending_add(sk, MGMT_OP_SET_FAST_CONNECTABLE, hdev, - data, len); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_SET_FAST_CONNECTABLE, hdev, data, + len); + if (!cmd) err = -ENOMEM; - goto unlock; - } + else + err = hci_cmd_sync_queue(hdev, write_fast_connectable_sync, cmd, + fast_connectable_complete); - hci_req_init(&req, hdev); - - __hci_req_write_fast_connectable(&req, cp->val); - - err = hci_req_run(&req, fast_connectable_complete); if (err < 0) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, - MGMT_STATUS_FAILED); - mgmt_pending_remove(cmd); + mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, + MGMT_STATUS_FAILED); + + + if (cmd) + mgmt_pending_free(cmd); } unlock: From 451d95a98c5a350da2f2c5447cc17115a5b94c8e Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:47 -0700 Subject: [PATCH 0040/2048] Bluetooth: hci_sync: Enable synch'd set_bredr Uses previously written: hci_write_fast_connectable_sync hci_update_scan_sync hci_update_adv_data_sync Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_sync.c | 2 +- net/bluetooth/mgmt.c | 72 ++++++++++++++++---------------- 3 files changed, 37 insertions(+), 38 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 1fb66b6f8a34..03ffe95415fc 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -69,6 +69,7 @@ int hci_update_passive_scan_sync(struct hci_dev *hdev); int hci_update_passive_scan(struct hci_dev *hdev); int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); +int hci_update_scan_sync(struct hci_dev *hdev); int hci_dev_open_sync(struct hci_dev *hdev); int hci_dev_close_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index f7c30b4e9aca..39fdb51c9698 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2262,7 +2262,7 @@ static int hci_write_scan_enable_sync(struct hci_dev *hdev, u8 val) HCI_CMD_TIMEOUT); } -static int hci_update_scan_sync(struct hci_dev *hdev) +int hci_update_scan_sync(struct hci_dev *hdev) { u8 scan; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 098ce324f20f..1e244783c5e1 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5953,20 +5953,14 @@ unlock: return err; } -static void set_bredr_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static void set_bredr_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; - bt_dev_dbg(hdev, "status 0x%02x", status); + bt_dev_dbg(hdev, "err %d", err); - hci_dev_lock(hdev); - - cmd = pending_find(MGMT_OP_SET_BREDR, hdev); - if (!cmd) - goto unlock; - - if (status) { - u8 mgmt_err = mgmt_status(status); + if (err) { + u8 mgmt_err = mgmt_status(err); /* We need to restore the flag if related HCI commands * failed. @@ -5979,17 +5973,31 @@ static void set_bredr_complete(struct hci_dev *hdev, u8 status, u16 opcode) new_settings(hdev, cmd->sk); } - mgmt_pending_remove(cmd); + mgmt_pending_free(cmd); +} -unlock: - hci_dev_unlock(hdev); +static int set_bredr_sync(struct hci_dev *hdev, void *data) +{ + int status; + + status = hci_write_fast_connectable_sync(hdev, false); + + if (!status) + status = hci_update_scan_sync(hdev); + + /* Since only the advertising data flags will change, there + * is no need to update the scan response data. + */ + if (!status) + status = hci_update_adv_data_sync(hdev, hdev->cur_adv_instance); + + return status; } static int set_bredr(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_mode *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); @@ -6061,15 +6069,19 @@ static int set_bredr(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) } } - if (pending_find(MGMT_OP_SET_BREDR, hdev)) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_BREDR, - MGMT_STATUS_BUSY); - goto unlock; - } - - cmd = mgmt_pending_add(sk, MGMT_OP_SET_BREDR, hdev, data, len); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_SET_BREDR, hdev, data, len); + if (!cmd) err = -ENOMEM; + else + err = hci_cmd_sync_queue(hdev, set_bredr_sync, cmd, + set_bredr_complete); + + if (err < 0) { + mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_BREDR, + MGMT_STATUS_FAILED); + if (cmd) + mgmt_pending_free(cmd); + goto unlock; } @@ -6078,20 +6090,6 @@ static int set_bredr(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) */ hci_dev_set_flag(hdev, HCI_BREDR_ENABLED); - hci_req_init(&req, hdev); - - __hci_req_write_fast_connectable(&req, false); - __hci_req_update_scan(&req); - - /* Since only the advertising data flags will change, there - * is no need to update the scan response data. - */ - __hci_req_update_adv_data(&req, hdev->cur_adv_instance); - - err = hci_req_run(&req, set_bredr_complete); - if (err < 0) - mgmt_pending_remove(cmd); - unlock: hci_dev_unlock(hdev); return err; From 47db6b42991e6d5645d0938e43085aaf88cdfba4 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:48 -0700 Subject: [PATCH 0041/2048] Bluetooth: hci_sync: Convert MGMT_OP_GET_CONN_INFO Synchronous version of MGMT_OP_GET_CONN_INFO Implements: hci_read_rssi_sync hci_read_tx_power_sync Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 2 + net/bluetooth/hci_sync.c | 19 ++++ net/bluetooth/mgmt.c | 156 ++++++++++++------------------- 3 files changed, 83 insertions(+), 94 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 03ffe95415fc..156cb5faf64f 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -67,6 +67,8 @@ int hci_disable_advertising_sync(struct hci_dev *hdev); int hci_update_passive_scan_sync(struct hci_dev *hdev); int hci_update_passive_scan(struct hci_dev *hdev); +int hci_read_rssi_sync(struct hci_dev *hdev, __le16 handle); +int hci_read_tx_power_sync(struct hci_dev *hdev, __le16 handle, u8 type); int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); int hci_update_scan_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 39fdb51c9698..ad819ff4e30d 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1309,6 +1309,25 @@ int hci_remove_advertising_sync(struct hci_dev *hdev, struct sock *sk, return 0; } +int hci_read_rssi_sync(struct hci_dev *hdev, __le16 handle) +{ + struct hci_cp_read_rssi cp; + + cp.handle = handle; + return __hci_cmd_sync_status(hdev, HCI_OP_READ_RSSI, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +int hci_read_tx_power_sync(struct hci_dev *hdev, __le16 handle, u8 type) +{ + struct hci_cp_read_tx_power cp; + + cp.handle = handle; + cp.type = type; + return __hci_cmd_sync_status(hdev, HCI_OP_READ_TX_POWER, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + int hci_disable_advertising_sync(struct hci_dev *hdev) { u8 enable = 0x00; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 1e244783c5e1..d809ab5e81e3 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5942,7 +5942,6 @@ static int set_fast_connectable(struct sock *sk, struct hci_dev *hdev, mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_FAST_CONNECTABLE, MGMT_STATUS_FAILED); - if (cmd) mgmt_pending_free(cmd); } @@ -6534,14 +6533,19 @@ static int load_long_term_keys(struct sock *sk, struct hci_dev *hdev, return err; } -static int conn_info_cmd_complete(struct mgmt_pending_cmd *cmd, u8 status) +static void get_conn_info_complete(struct hci_dev *hdev, void *data, int err) { + struct mgmt_pending_cmd *cmd = data; struct hci_conn *conn = cmd->user_data; + struct mgmt_cp_get_conn_info *cp = cmd->param; struct mgmt_rp_get_conn_info rp; - int err; + u8 status; - memcpy(&rp.addr, cmd->param, sizeof(rp.addr)); + bt_dev_dbg(hdev, "err %d", err); + memcpy(&rp.addr, &cp->addr.bdaddr, sizeof(rp.addr)); + + status = mgmt_status(err); if (status == MGMT_STATUS_SUCCESS) { rp.rssi = conn->rssi; rp.tx_power = conn->tx_power; @@ -6552,67 +6556,58 @@ static int conn_info_cmd_complete(struct mgmt_pending_cmd *cmd, u8 status) rp.max_tx_power = HCI_TX_POWER_INVALID; } - err = mgmt_cmd_complete(cmd->sk, cmd->index, MGMT_OP_GET_CONN_INFO, - status, &rp, sizeof(rp)); + mgmt_cmd_complete(cmd->sk, cmd->index, MGMT_OP_GET_CONN_INFO, status, + &rp, sizeof(rp)); - hci_conn_drop(conn); - hci_conn_put(conn); + if (conn) { + hci_conn_drop(conn); + hci_conn_put(conn); + } - return err; + mgmt_pending_free(cmd); } -static void conn_info_refresh_complete(struct hci_dev *hdev, u8 hci_status, - u16 opcode) +static int get_conn_info_sync(struct hci_dev *hdev, void *data) { - struct hci_cp_read_rssi *cp; - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_get_conn_info *cp = cmd->param; struct hci_conn *conn; - u16 handle; - u8 status; + int err; + __le16 handle; - bt_dev_dbg(hdev, "status 0x%02x", hci_status); + /* Make sure we are still connected */ + if (cp->addr.type == BDADDR_BREDR) + conn = hci_conn_hash_lookup_ba(hdev, ACL_LINK, + &cp->addr.bdaddr); + else + conn = hci_conn_hash_lookup_ba(hdev, LE_LINK, &cp->addr.bdaddr); - hci_dev_lock(hdev); + if (!conn || conn != cmd->user_data || conn->state != BT_CONNECTED) { + if (cmd->user_data) { + hci_conn_drop(cmd->user_data); + hci_conn_put(cmd->user_data); + cmd->user_data = NULL; + } + return MGMT_STATUS_NOT_CONNECTED; + } - /* Commands sent in request are either Read RSSI or Read Transmit Power - * Level so we check which one was last sent to retrieve connection - * handle. Both commands have handle as first parameter so it's safe to - * cast data on the same command struct. - * - * First command sent is always Read RSSI and we fail only if it fails. - * In other case we simply override error to indicate success as we - * already remembered if TX power value is actually valid. + handle = cpu_to_le16(conn->handle); + + /* Refresh RSSI each time */ + err = hci_read_rssi_sync(hdev, handle); + + /* For LE links TX power does not change thus we don't need to + * query for it once value is known. */ - cp = hci_sent_cmd_data(hdev, HCI_OP_READ_RSSI); - if (!cp) { - cp = hci_sent_cmd_data(hdev, HCI_OP_READ_TX_POWER); - status = MGMT_STATUS_SUCCESS; - } else { - status = mgmt_status(hci_status); - } + if (!err && (!bdaddr_type_is_le(cp->addr.type) || + conn->tx_power == HCI_TX_POWER_INVALID)) + err = hci_read_tx_power_sync(hdev, handle, 0x00); - if (!cp) { - bt_dev_err(hdev, "invalid sent_cmd in conn_info response"); - goto unlock; - } + /* Max TX power needs to be read only once per connection */ + if (!err && conn->max_tx_power == HCI_TX_POWER_INVALID) + err = hci_read_tx_power_sync(hdev, handle, 0x01); - handle = __le16_to_cpu(cp->handle); - conn = hci_conn_hash_lookup_handle(hdev, handle); - if (!conn) { - bt_dev_err(hdev, "unknown handle (%u) in conn_info response", - handle); - goto unlock; - } - - cmd = pending_find_data(MGMT_OP_GET_CONN_INFO, hdev, conn); - if (!cmd) - goto unlock; - - cmd->cmd_complete(cmd, status); - mgmt_pending_remove(cmd); - -unlock: - hci_dev_unlock(hdev); + return err; } static int get_conn_info(struct sock *sk, struct hci_dev *hdev, void *data, @@ -6657,12 +6652,6 @@ static int get_conn_info(struct sock *sk, struct hci_dev *hdev, void *data, goto unlock; } - if (pending_find_data(MGMT_OP_GET_CONN_INFO, hdev, conn)) { - err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_GET_CONN_INFO, - MGMT_STATUS_BUSY, &rp, sizeof(rp)); - goto unlock; - } - /* To avoid client trying to guess when to poll again for information we * calculate conn info age as random value between min/max set in hdev. */ @@ -6676,49 +6665,28 @@ static int get_conn_info(struct sock *sk, struct hci_dev *hdev, void *data, if (time_after(jiffies, conn->conn_info_timestamp + msecs_to_jiffies(conn_info_age)) || !conn->conn_info_timestamp) { - struct hci_request req; - struct hci_cp_read_tx_power req_txp_cp; - struct hci_cp_read_rssi req_rssi_cp; struct mgmt_pending_cmd *cmd; - hci_req_init(&req, hdev); - req_rssi_cp.handle = cpu_to_le16(conn->handle); - hci_req_add(&req, HCI_OP_READ_RSSI, sizeof(req_rssi_cp), - &req_rssi_cp); - - /* For LE links TX power does not change thus we don't need to - * query for it once value is known. - */ - if (!bdaddr_type_is_le(cp->addr.type) || - conn->tx_power == HCI_TX_POWER_INVALID) { - req_txp_cp.handle = cpu_to_le16(conn->handle); - req_txp_cp.type = 0x00; - hci_req_add(&req, HCI_OP_READ_TX_POWER, - sizeof(req_txp_cp), &req_txp_cp); - } - - /* Max TX power needs to be read only once per connection */ - if (conn->max_tx_power == HCI_TX_POWER_INVALID) { - req_txp_cp.handle = cpu_to_le16(conn->handle); - req_txp_cp.type = 0x01; - hci_req_add(&req, HCI_OP_READ_TX_POWER, - sizeof(req_txp_cp), &req_txp_cp); - } - - err = hci_req_run(&req, conn_info_refresh_complete); - if (err < 0) - goto unlock; - - cmd = mgmt_pending_add(sk, MGMT_OP_GET_CONN_INFO, hdev, - data, len); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_GET_CONN_INFO, hdev, data, + len); + if (!cmd) err = -ENOMEM; + else + err = hci_cmd_sync_queue(hdev, get_conn_info_sync, + cmd, get_conn_info_complete); + + if (err < 0) { + mgmt_cmd_complete(sk, hdev->id, MGMT_OP_GET_CONN_INFO, + MGMT_STATUS_FAILED, &rp, sizeof(rp)); + + if (cmd) + mgmt_pending_free(cmd); + goto unlock; } hci_conn_hold(conn); cmd->user_data = hci_conn_get(conn); - cmd->cmd_complete = conn_info_cmd_complete; conn->conn_info_timestamp = jiffies; } else { From 2f2eb0c9de2eb69969aaf04feffb69310d3804b2 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:49 -0700 Subject: [PATCH 0042/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_SECURE_CONN Synchronous version of MGMT_OP_SET_SECURE_CONN. Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_sync.c | 18 ++++++++- net/bluetooth/mgmt.c | 66 ++++++++++++++++---------------- 3 files changed, 49 insertions(+), 36 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 156cb5faf64f..4b27a89cc57e 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -69,6 +69,7 @@ int hci_update_passive_scan_sync(struct hci_dev *hdev); int hci_update_passive_scan(struct hci_dev *hdev); int hci_read_rssi_sync(struct hci_dev *hdev, __le16 handle); int hci_read_tx_power_sync(struct hci_dev *hdev, __le16 handle, u8 type); +int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val); int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); int hci_update_scan_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index ad819ff4e30d..dc27d6652188 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2104,13 +2104,27 @@ int hci_update_passive_scan(struct hci_dev *hdev) return hci_cmd_sync_queue(hdev, update_passive_scan_sync, NULL, NULL); } -static int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val) +int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val) { + int err; + if (!bredr_sc_enabled(hdev) || lmp_host_sc_capable(hdev)) return 0; - return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SC_SUPPORT, + err = __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SC_SUPPORT, sizeof(val), &val, HCI_CMD_TIMEOUT); + + if (!err) { + if (val) { + hdev->features[1][0] |= LMP_HOST_SC; + hci_dev_set_flag(hdev, HCI_SC_ENABLED); + } else { + hdev->features[1][0] &= ~LMP_HOST_SC; + hci_dev_clear_flag(hdev, HCI_SC_ENABLED); + } + } + + return err; } static int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index d809ab5e81e3..656e34617029 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -6094,23 +6094,18 @@ unlock: return err; } -static void sc_enable_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static void set_secure_conn_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; struct mgmt_mode *cp; - bt_dev_dbg(hdev, "status %u", status); + bt_dev_dbg(hdev, "err %d", err); - hci_dev_lock(hdev); + if (err) { + u8 mgmt_err = mgmt_status(err); - cmd = pending_find(MGMT_OP_SET_SECURE_CONN, hdev); - if (!cmd) - goto unlock; - - if (status) { - mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, - mgmt_status(status)); - goto remove; + mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, mgmt_err); + goto done; } cp = cmd->param; @@ -6130,13 +6125,23 @@ static void sc_enable_complete(struct hci_dev *hdev, u8 status, u16 opcode) break; } - send_settings_rsp(cmd->sk, MGMT_OP_SET_SECURE_CONN, hdev); + send_settings_rsp(cmd->sk, cmd->opcode, hdev); new_settings(hdev, cmd->sk); -remove: - mgmt_pending_remove(cmd); -unlock: - hci_dev_unlock(hdev); +done: + mgmt_pending_free(cmd); +} + +static int set_secure_conn_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + u8 val = !!cp->val; + + /* Force write of val */ + hci_dev_set_flag(hdev, HCI_SC_ENABLED); + + return hci_write_sc_support_sync(hdev, val); } static int set_secure_conn(struct sock *sk, struct hci_dev *hdev, @@ -6144,7 +6149,6 @@ static int set_secure_conn(struct sock *sk, struct hci_dev *hdev, { struct mgmt_mode *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; u8 val; int err; @@ -6163,7 +6167,7 @@ static int set_secure_conn(struct sock *sk, struct hci_dev *hdev, if (cp->val != 0x00 && cp->val != 0x01 && cp->val != 0x02) return mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_SECURE_CONN, - MGMT_STATUS_INVALID_PARAMS); + MGMT_STATUS_INVALID_PARAMS); hci_dev_lock(hdev); @@ -6194,12 +6198,6 @@ static int set_secure_conn(struct sock *sk, struct hci_dev *hdev, goto failed; } - if (pending_find(MGMT_OP_SET_SECURE_CONN, hdev)) { - err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_SECURE_CONN, - MGMT_STATUS_BUSY); - goto failed; - } - val = !!cp->val; if (val == hci_dev_test_flag(hdev, HCI_SC_ENABLED) && @@ -6208,18 +6206,18 @@ static int set_secure_conn(struct sock *sk, struct hci_dev *hdev, goto failed; } - cmd = mgmt_pending_add(sk, MGMT_OP_SET_SECURE_CONN, hdev, data, len); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_SET_SECURE_CONN, hdev, data, len); + if (!cmd) err = -ENOMEM; - goto failed; - } + else + err = hci_cmd_sync_queue(hdev, set_secure_conn_sync, cmd, + set_secure_conn_complete); - hci_req_init(&req, hdev); - hci_req_add(&req, HCI_OP_WRITE_SC_SUPPORT, 1, &val); - err = hci_req_run(&req, sc_enable_complete); if (err < 0) { - mgmt_pending_remove(cmd); - goto failed; + mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_SECURE_CONN, + MGMT_STATUS_FAILED); + if (cmd) + mgmt_pending_free(cmd); } failed: From 5a750137466400276560458db040c143cbc254ed Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:50 -0700 Subject: [PATCH 0043/2048] Bluetooth: hci_sync: Convert MGMT_OP_GET_CLOCK_INFO Synchronous version of MGMT_OP_GET_CLOCK_INFO. Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_sync.c | 6 ++ net/bluetooth/mgmt.c | 129 +++++++++++++------------------ 3 files changed, 62 insertions(+), 74 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 4b27a89cc57e..cf54f8f14edb 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -70,6 +70,7 @@ int hci_update_passive_scan(struct hci_dev *hdev); int hci_read_rssi_sync(struct hci_dev *hdev, __le16 handle); int hci_read_tx_power_sync(struct hci_dev *hdev, __le16 handle, u8 type); int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val); +int hci_read_clock_sync(struct hci_dev *hdev, struct hci_cp_read_clock *cp); int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); int hci_update_scan_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index dc27d6652188..37595b393ac1 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1318,6 +1318,12 @@ int hci_read_rssi_sync(struct hci_dev *hdev, __le16 handle) sizeof(cp), &cp, HCI_CMD_TIMEOUT); } +int hci_read_clock_sync(struct hci_dev *hdev, struct hci_cp_read_clock *cp) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_CLOCK, + sizeof(*cp), cp, HCI_CMD_TIMEOUT); +} + int hci_read_tx_power_sync(struct hci_dev *hdev, __le16 handle, u8 type) { struct hci_cp_read_tx_power cp; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 656e34617029..7ee0a4beab08 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -927,13 +927,6 @@ static struct mgmt_pending_cmd *pending_find(u16 opcode, struct hci_dev *hdev) return mgmt_pending_find(HCI_CHANNEL_CONTROL, opcode, hdev); } -static struct mgmt_pending_cmd *pending_find_data(u16 opcode, - struct hci_dev *hdev, - const void *data) -{ - return mgmt_pending_find_data(HCI_CHANNEL_CONTROL, opcode, hdev, data); -} - u8 mgmt_get_adv_discov_flags(struct hci_dev *hdev) { struct mgmt_pending_cmd *cmd; @@ -6702,82 +6695,76 @@ unlock: return err; } -static int clock_info_cmd_complete(struct mgmt_pending_cmd *cmd, u8 status) +static void get_clock_info_complete(struct hci_dev *hdev, void *data, int err) { - struct hci_conn *conn = cmd->user_data; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_get_clock_info *cp = cmd->param; struct mgmt_rp_get_clock_info rp; - struct hci_dev *hdev; - int err; + struct hci_conn *conn = cmd->user_data; + u8 status = mgmt_status(err); + + bt_dev_dbg(hdev, "err %d", err); memset(&rp, 0, sizeof(rp)); - memcpy(&rp.addr, cmd->param, sizeof(rp.addr)); + bacpy(&rp.addr.bdaddr, &cp->addr.bdaddr); + rp.addr.type = cp->addr.type; - if (status) + if (err) goto complete; - hdev = hci_dev_get(cmd->index); - if (hdev) { - rp.local_clock = cpu_to_le32(hdev->clock); - hci_dev_put(hdev); - } + rp.local_clock = cpu_to_le32(hdev->clock); if (conn) { rp.piconet_clock = cpu_to_le32(conn->clock); rp.accuracy = cpu_to_le16(conn->clock_accuracy); + hci_conn_drop(conn); + hci_conn_put(conn); } complete: - err = mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, status, &rp, - sizeof(rp)); + mgmt_cmd_complete(cmd->sk, cmd->index, cmd->opcode, status, &rp, + sizeof(rp)); + + mgmt_pending_free(cmd); +} + +static int get_clock_info_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_get_clock_info *cp = cmd->param; + struct hci_cp_read_clock hci_cp; + struct hci_conn *conn = cmd->user_data; + int err; + + memset(&hci_cp, 0, sizeof(hci_cp)); + err = hci_read_clock_sync(hdev, &hci_cp); if (conn) { - hci_conn_drop(conn); - hci_conn_put(conn); + /* Make sure connection still exists */ + conn = hci_conn_hash_lookup_ba(hdev, ACL_LINK, + &cp->addr.bdaddr); + + if (conn && conn == cmd->user_data && + conn->state == BT_CONNECTED) { + hci_cp.handle = cpu_to_le16(conn->handle); + hci_cp.which = 0x01; /* Piconet clock */ + err = hci_read_clock_sync(hdev, &hci_cp); + } else if (cmd->user_data) { + hci_conn_drop(cmd->user_data); + hci_conn_put(cmd->user_data); + cmd->user_data = NULL; + } } return err; } -static void get_clock_info_complete(struct hci_dev *hdev, u8 status, u16 opcode) -{ - struct hci_cp_read_clock *hci_cp; - struct mgmt_pending_cmd *cmd; - struct hci_conn *conn; - - bt_dev_dbg(hdev, "status %u", status); - - hci_dev_lock(hdev); - - hci_cp = hci_sent_cmd_data(hdev, HCI_OP_READ_CLOCK); - if (!hci_cp) - goto unlock; - - if (hci_cp->which) { - u16 handle = __le16_to_cpu(hci_cp->handle); - conn = hci_conn_hash_lookup_handle(hdev, handle); - } else { - conn = NULL; - } - - cmd = pending_find_data(MGMT_OP_GET_CLOCK_INFO, hdev, conn); - if (!cmd) - goto unlock; - - cmd->cmd_complete(cmd, mgmt_status(status)); - mgmt_pending_remove(cmd); - -unlock: - hci_dev_unlock(hdev); -} - static int get_clock_info(struct sock *sk, struct hci_dev *hdev, void *data, - u16 len) + u16 len) { struct mgmt_cp_get_clock_info *cp = data; struct mgmt_rp_get_clock_info rp; - struct hci_cp_read_clock hci_cp; struct mgmt_pending_cmd *cmd; - struct hci_request req; struct hci_conn *conn; int err; @@ -6815,31 +6802,25 @@ static int get_clock_info(struct sock *sk, struct hci_dev *hdev, void *data, conn = NULL; } - cmd = mgmt_pending_add(sk, MGMT_OP_GET_CLOCK_INFO, hdev, data, len); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_GET_CLOCK_INFO, hdev, data, len); + if (!cmd) err = -ENOMEM; - goto unlock; - } + else + err = hci_cmd_sync_queue(hdev, get_clock_info_sync, cmd, + get_clock_info_complete); - cmd->cmd_complete = clock_info_cmd_complete; + if (err < 0) { + err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_GET_CLOCK_INFO, + MGMT_STATUS_FAILED, &rp, sizeof(rp)); - hci_req_init(&req, hdev); + if (cmd) + mgmt_pending_free(cmd); - memset(&hci_cp, 0, sizeof(hci_cp)); - hci_req_add(&req, HCI_OP_READ_CLOCK, sizeof(hci_cp), &hci_cp); - - if (conn) { + } else if (conn) { hci_conn_hold(conn); cmd->user_data = hci_conn_get(conn); - - hci_cp.handle = cpu_to_le16(conn->handle); - hci_cp.which = 0x01; /* Piconet clock */ - hci_req_add(&req, HCI_OP_READ_CLOCK, sizeof(hci_cp), &hci_cp); } - err = hci_req_run(&req, get_clock_info_complete); - if (err < 0) - mgmt_pending_remove(cmd); unlock: hci_dev_unlock(hdev); From d81a494c43df66f053f7d1ec612e057eb99f34d4 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:51 -0700 Subject: [PATCH 0044/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_LE Uses: hci_disable_advertising_sync hci_remove_ext_adv_instance_sync hci_write_le_host_supported_sync hci_setup_ext_adv_instance_sync hci_update_scan_rsp_data_sync Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 4 ++ net/bluetooth/hci_sync.c | 3 +- net/bluetooth/mgmt.c | 89 ++++++++++++++++---------------- 3 files changed, 50 insertions(+), 46 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index cf54f8f14edb..d969693c33b5 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -75,6 +75,10 @@ int hci_read_clock_sync(struct hci_dev *hdev, struct hci_cp_read_clock *cp); int hci_write_fast_connectable_sync(struct hci_dev *hdev, bool enable); int hci_update_scan_sync(struct hci_dev *hdev); +int hci_write_le_host_supported_sync(struct hci_dev *hdev, u8 le, u8 simul); +int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, + struct sock *sk); + int hci_dev_open_sync(struct hci_dev *hdev); int hci_dev_close_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 37595b393ac1..877f5b97a0f2 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2149,8 +2149,7 @@ static int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) return hci_write_sc_support_sync(hdev, 0x01); } -static int hci_write_le_host_supported_sync(struct hci_dev *hdev, u8 le, - u8 simul) +int hci_write_le_host_supported_sync(struct hci_dev *hdev, u8 le, u8 simul) { struct hci_cp_write_le_host_supported cp; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 7ee0a4beab08..82369db92bb8 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1902,18 +1902,17 @@ unlock: return err; } -static void le_enable_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static void set_le_complete(struct hci_dev *hdev, void *data, int err) { struct cmd_lookup match = { NULL, hdev }; + u8 status = mgmt_status(err); - hci_dev_lock(hdev); + bt_dev_dbg(hdev, "err %d", err); if (status) { - u8 mgmt_err = mgmt_status(status); - mgmt_pending_foreach(MGMT_OP_SET_LE, hdev, cmd_status_rsp, - &mgmt_err); - goto unlock; + &status); + return; } mgmt_pending_foreach(MGMT_OP_SET_LE, hdev, settings_rsp, &match); @@ -1922,39 +1921,54 @@ static void le_enable_complete(struct hci_dev *hdev, u8 status, u16 opcode) if (match.sk) sock_put(match.sk); +} + +static int set_le_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + u8 val = !!cp->val; + int err; + + if (!val) { + if (hci_dev_test_flag(hdev, HCI_LE_ADV)) + hci_disable_advertising_sync(hdev); + + if (ext_adv_capable(hdev)) + hci_remove_ext_adv_instance_sync(hdev, 0, cmd->sk); + } else { + hci_dev_set_flag(hdev, HCI_LE_ENABLED); + } + + err = hci_write_le_host_supported_sync(hdev, val, 0); /* Make sure the controller has a good default for * advertising data. Restrict the update to when LE * has actually been enabled. During power on, the * update in powered_update_hci will take care of it. */ - if (hci_dev_test_flag(hdev, HCI_LE_ENABLED)) { - struct hci_request req; - hci_req_init(&req, hdev); + if (!err && hci_dev_test_flag(hdev, HCI_LE_ENABLED)) { if (ext_adv_capable(hdev)) { - int err; + int status; - err = __hci_req_setup_ext_adv_instance(&req, 0x00); - if (!err) - __hci_req_update_scan_rsp_data(&req, 0x00); + status = hci_setup_ext_adv_instance_sync(hdev, 0x00); + if (!status) + hci_update_scan_rsp_data_sync(hdev, 0x00); } else { - __hci_req_update_adv_data(&req, 0x00); - __hci_req_update_scan_rsp_data(&req, 0x00); + hci_update_adv_data_sync(hdev, 0x00); + hci_update_scan_rsp_data_sync(hdev, 0x00); } - hci_req_run(&req, NULL); + hci_update_passive_scan(hdev); } -unlock: - hci_dev_unlock(hdev); + return err; } static int set_le(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_mode *cp = data; - struct hci_cp_write_le_host_supported hci_cp; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; u8 val, enabled; @@ -2024,33 +2038,20 @@ static int set_le(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) } cmd = mgmt_pending_add(sk, MGMT_OP_SET_LE, hdev, data, len); - if (!cmd) { + if (!cmd) err = -ENOMEM; - goto unlock; + else + err = hci_cmd_sync_queue(hdev, set_le_sync, cmd, + set_le_complete); + + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_LE, + MGMT_STATUS_FAILED); + + if (cmd) + mgmt_pending_remove(cmd); } - hci_req_init(&req, hdev); - - memset(&hci_cp, 0, sizeof(hci_cp)); - - if (val) { - hci_cp.le = val; - hci_cp.simul = 0x00; - } else { - if (hci_dev_test_flag(hdev, HCI_LE_ADV)) - __hci_req_disable_advertising(&req); - - if (ext_adv_capable(hdev)) - __hci_req_clear_ext_adv_sets(&req); - } - - hci_req_add(&req, HCI_OP_WRITE_LE_HOST_SUPPORTED, sizeof(hci_cp), - &hci_cp); - - err = hci_req_run(&req, le_enable_complete); - if (err < 0) - mgmt_pending_remove(cmd); - unlock: hci_dev_unlock(hdev); return err; From f892244b05bf6a99e48db14f8bbd96db16bcfa69 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:52 -0700 Subject: [PATCH 0045/2048] Bluetooth: hci_sync: Convert MGMT_OP_READ_LOCAL_OOB_DATA New functions: hci_read_local_oob_data_sync This function requires all of the data from the cmd cmplt event to be passed up to the caller via the skb. mgmt-tester paths: Read Local OOB Data - Not powered Read Local OOB Data - Legacy pairing Read Local OOB Data - Success SSP Read Local OOB Data - Success SC Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 2 + net/bluetooth/hci_sync.c | 9 ++++ net/bluetooth/mgmt.c | 72 ++++++++++++++++++++------------ net/bluetooth/mgmt_util.h | 1 + 4 files changed, 58 insertions(+), 26 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index d969693c33b5..a381621a56a1 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -78,6 +78,8 @@ int hci_update_scan_sync(struct hci_dev *hdev); int hci_write_le_host_supported_sync(struct hci_dev *hdev, u8 le, u8 simul); int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, struct sock *sk); +struct sk_buff *hci_read_local_oob_data_sync(struct hci_dev *hdev, bool ext, + struct sock *sk); int hci_dev_open_sync(struct hci_dev *hdev); int hci_dev_close_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 877f5b97a0f2..a0a5c27d111c 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1703,6 +1703,15 @@ static int hci_resume_advertising_sync(struct hci_dev *hdev) return err; } +struct sk_buff *hci_read_local_oob_data_sync(struct hci_dev *hdev, + bool extended, struct sock *sk) +{ + u16 opcode = extended ? HCI_OP_READ_LOCAL_OOB_EXT_DATA : + HCI_OP_READ_LOCAL_OOB_DATA; + + return __hci_cmd_sync_sk(hdev, opcode, 0, NULL, 0, HCI_CMD_TIMEOUT, sk); +} + /* Device must not be scanning when updating the accept list. * * Update is done using the following sequence: diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 82369db92bb8..6bd75ec7f45e 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -4793,28 +4793,33 @@ unlock: status); } -static void read_local_oob_data_complete(struct hci_dev *hdev, u8 status, - u16 opcode, struct sk_buff *skb) +static void read_local_oob_data_complete(struct hci_dev *hdev, void *data, int err) { struct mgmt_rp_read_local_oob_data mgmt_rp; size_t rp_size = sizeof(mgmt_rp); - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; + struct sk_buff *skb = cmd->skb; + u8 status = mgmt_status(err); - bt_dev_dbg(hdev, "status %u", status); + if (!status) { + if (!skb) + status = MGMT_STATUS_FAILED; + else if (IS_ERR(skb)) + status = mgmt_status(PTR_ERR(skb)); + else + status = mgmt_status(skb->data[0]); + } - cmd = pending_find(MGMT_OP_READ_LOCAL_OOB_DATA, hdev); - if (!cmd) - return; + bt_dev_dbg(hdev, "status %d", status); - if (status || !skb) { - mgmt_cmd_status(cmd->sk, hdev->id, MGMT_OP_READ_LOCAL_OOB_DATA, - status ? mgmt_status(status) : MGMT_STATUS_FAILED); + if (status) { + mgmt_cmd_status(cmd->sk, hdev->id, MGMT_OP_READ_LOCAL_OOB_DATA, status); goto remove; } memset(&mgmt_rp, 0, sizeof(mgmt_rp)); - if (opcode == HCI_OP_READ_LOCAL_OOB_DATA) { + if (!bredr_sc_enabled(hdev)) { struct hci_rp_read_local_oob_data *rp = (void *) skb->data; if (skb->len < sizeof(*rp)) { @@ -4849,14 +4854,31 @@ static void read_local_oob_data_complete(struct hci_dev *hdev, u8 status, MGMT_STATUS_SUCCESS, &mgmt_rp, rp_size); remove: - mgmt_pending_remove(cmd); + if (skb && !IS_ERR(skb)) + kfree_skb(skb); + + mgmt_pending_free(cmd); +} + +static int read_local_oob_data_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + + if (bredr_sc_enabled(hdev)) + cmd->skb = hci_read_local_oob_data_sync(hdev, true, cmd->sk); + else + cmd->skb = hci_read_local_oob_data_sync(hdev, false, cmd->sk); + + if (IS_ERR(cmd->skb)) + return PTR_ERR(cmd->skb); + else + return 0; } static int read_local_oob_data(struct sock *sk, struct hci_dev *hdev, void *data, u16 data_len) { struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); @@ -4881,22 +4903,20 @@ static int read_local_oob_data(struct sock *sk, struct hci_dev *hdev, goto unlock; } - cmd = mgmt_pending_add(sk, MGMT_OP_READ_LOCAL_OOB_DATA, hdev, NULL, 0); - if (!cmd) { + cmd = mgmt_pending_new(sk, MGMT_OP_READ_LOCAL_OOB_DATA, hdev, NULL, 0); + if (!cmd) err = -ENOMEM; - goto unlock; - } - - hci_req_init(&req, hdev); - - if (bredr_sc_enabled(hdev)) - hci_req_add(&req, HCI_OP_READ_LOCAL_OOB_EXT_DATA, 0, NULL); else - hci_req_add(&req, HCI_OP_READ_LOCAL_OOB_DATA, 0, NULL); + err = hci_cmd_sync_queue(hdev, read_local_oob_data_sync, cmd, + read_local_oob_data_complete); - err = hci_req_run_skb(&req, read_local_oob_data_complete); - if (err < 0) - mgmt_pending_remove(cmd); + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_READ_LOCAL_OOB_DATA, + MGMT_STATUS_FAILED); + + if (cmd) + mgmt_pending_free(cmd); + } unlock: hci_dev_unlock(hdev); diff --git a/net/bluetooth/mgmt_util.h b/net/bluetooth/mgmt_util.h index 9dc24ba0d51a..63b965eaaaac 100644 --- a/net/bluetooth/mgmt_util.h +++ b/net/bluetooth/mgmt_util.h @@ -27,6 +27,7 @@ struct mgmt_pending_cmd { void *param; size_t param_len; struct sock *sk; + struct sk_buff *skb; void *user_data; int (*cmd_complete)(struct mgmt_pending_cmd *cmd, u8 status); }; From 177e77a30e469af96e301658facf6a12a6e9e200 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:53 -0700 Subject: [PATCH 0046/2048] Bluetooth: hci_sync: Convert MGMT_OP_READ_LOCAL_OOB_EXT_DATA Uses existing *_sync functions mgmt-test paths: Read Local OOB Ext Data - Invalid index Read Local OOB Ext Data - Legacy pairing Read Local OOB Ext Data - Success SSP Read Local OOB Ext Data - Success SC Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- net/bluetooth/mgmt.c | 37 ++++++++++++++++++++----------------- 1 file changed, 20 insertions(+), 17 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 6bd75ec7f45e..4ac63e402ef6 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -7364,22 +7364,28 @@ unlock: return err; } -static void read_local_oob_ext_data_complete(struct hci_dev *hdev, u8 status, - u16 opcode, struct sk_buff *skb) +static void read_local_oob_ext_data_complete(struct hci_dev *hdev, void *data, + int err) { const struct mgmt_cp_read_local_oob_ext_data *mgmt_cp; struct mgmt_rp_read_local_oob_ext_data *mgmt_rp; u8 *h192, *r192, *h256, *r256; - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; + struct sk_buff *skb = cmd->skb; + u8 status = mgmt_status(err); u16 eir_len; - int err; + + if (!status) { + if (!skb) + status = MGMT_STATUS_FAILED; + else if (IS_ERR(skb)) + status = mgmt_status(PTR_ERR(skb)); + else + status = mgmt_status(skb->data[0]); + } bt_dev_dbg(hdev, "status %u", status); - cmd = pending_find(MGMT_OP_READ_LOCAL_OOB_EXT_DATA, hdev); - if (!cmd) - return; - mgmt_cp = cmd->param; if (status) { @@ -7390,7 +7396,7 @@ static void read_local_oob_ext_data_complete(struct hci_dev *hdev, u8 status, r192 = NULL; h256 = NULL; r256 = NULL; - } else if (opcode == HCI_OP_READ_LOCAL_OOB_DATA) { + } else if (!bredr_sc_enabled(hdev)) { struct hci_rp_read_local_oob_data *rp; if (skb->len != sizeof(*rp)) { @@ -7471,6 +7477,9 @@ send_rsp: mgmt_rp, sizeof(*mgmt_rp) + eir_len, HCI_MGMT_OOB_DATA_EVENTS, cmd->sk); done: + if (skb && !IS_ERR(skb)) + kfree_skb(skb); + kfree(mgmt_rp); mgmt_pending_remove(cmd); } @@ -7479,7 +7488,6 @@ static int read_local_ssp_oob_req(struct hci_dev *hdev, struct sock *sk, struct mgmt_cp_read_local_oob_ext_data *cp) { struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; cmd = mgmt_pending_add(sk, MGMT_OP_READ_LOCAL_OOB_EXT_DATA, hdev, @@ -7487,14 +7495,9 @@ static int read_local_ssp_oob_req(struct hci_dev *hdev, struct sock *sk, if (!cmd) return -ENOMEM; - hci_req_init(&req, hdev); + err = hci_cmd_sync_queue(hdev, read_local_oob_data_sync, cmd, + read_local_oob_ext_data_complete); - if (bredr_sc_enabled(hdev)) - hci_req_add(&req, HCI_OP_READ_LOCAL_OOB_EXT_DATA, 0, NULL); - else - hci_req_add(&req, HCI_OP_READ_LOCAL_OOB_DATA, 0, NULL); - - err = hci_req_run_skb(&req, read_local_oob_ext_data_complete); if (err < 0) { mgmt_pending_remove(cmd); return err; From 6f6ff38a1e140d7ca3a733556b0027eec1fe4ef1 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:54 -0700 Subject: [PATCH 0047/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_LOCAL_NAME Uses existing *_sync functions, but made hci_update_name_sync non-static. mgmt-test paths: Set Advertising on - Local name 1 Set Advertising on - Name + Appear 1 Set Local Name - Success 1 Set Local Name - Success 2 Set Local Name - Success 3 Add Advertising - Success (Empty ScRsp) Add Advertising - Success (Complete name) Add Advertising - Success (Shortened name) Add Advertising - Success (Short name) Add Advertising - Success (Name + data) Add Advertising - Invalid Params (Name + data) Add Advertising - Success (Name+data+appear) Read Ext Controller Info 3 Read Ext Controller Info 4 Read Ext Controller Info 5 Add Ext Advertising - Success (Empty ScRsp) Add Ext Advertising - Success (Complete name) Add Ext Advertising - Success (Shortened name) Add Ext Advertising - Success (Short name) Add Ext Advertising - Success (Name + data) Add Ext Advertising - Invalid Params (Name + data) Add Ext Advertising - Success (Name+data+appear) Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_sync.c | 2 +- net/bluetooth/mgmt.c | 67 ++++++++++++++++---------------- 3 files changed, 35 insertions(+), 35 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index a381621a56a1..d9f2e3182ed8 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -46,6 +46,7 @@ int hci_update_class_sync(struct hci_dev *hdev); int hci_update_eir_sync(struct hci_dev *hdev); int hci_update_class_sync(struct hci_dev *hdev); +int hci_update_name_sync(struct hci_dev *hdev); int hci_update_random_address_sync(struct hci_dev *hdev, bool require_privacy, bool rpa, u8 *own_addr_type); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index a0a5c27d111c..47beabe48257 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2341,7 +2341,7 @@ int hci_update_scan_sync(struct hci_dev *hdev) return hci_write_scan_enable_sync(hdev, scan); } -static int hci_update_name_sync(struct hci_dev *hdev) +int hci_update_name_sync(struct hci_dev *hdev) { struct hci_cp_write_local_name cp; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 4ac63e402ef6..2585eab1fe21 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -3246,24 +3246,17 @@ static void adv_expire(struct hci_dev *hdev, u32 flags) hci_req_run(&req, NULL); } -static void set_name_complete(struct hci_dev *hdev, u8 status, u16 opcode) +static void set_name_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_cp_set_local_name *cp; - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_set_local_name *cp = cmd->param; + u8 status = mgmt_status(err); - bt_dev_dbg(hdev, "status 0x%02x", status); - - hci_dev_lock(hdev); - - cmd = pending_find(MGMT_OP_SET_LOCAL_NAME, hdev); - if (!cmd) - goto unlock; - - cp = cmd->param; + bt_dev_dbg(hdev, "err %d", err); if (status) { mgmt_cmd_status(cmd->sk, hdev->id, MGMT_OP_SET_LOCAL_NAME, - mgmt_status(status)); + status); } else { mgmt_cmd_complete(cmd->sk, hdev->id, MGMT_OP_SET_LOCAL_NAME, 0, cp, sizeof(*cp)); @@ -3273,9 +3266,22 @@ static void set_name_complete(struct hci_dev *hdev, u8 status, u16 opcode) } mgmt_pending_remove(cmd); +} -unlock: - hci_dev_unlock(hdev); +static int set_name_sync(struct hci_dev *hdev, void *data) +{ + if (lmp_bredr_capable(hdev)) { + hci_update_name_sync(hdev); + hci_update_eir_sync(hdev); + } + + /* The name is stored in the scan response data and so + * no need to update the advertising data here. + */ + if (lmp_le_capable(hdev) && hci_dev_test_flag(hdev, HCI_ADVERTISING)) + hci_update_scan_rsp_data_sync(hdev, hdev->cur_adv_instance); + + return 0; } static int set_local_name(struct sock *sk, struct hci_dev *hdev, void *data, @@ -3283,7 +3289,6 @@ static int set_local_name(struct sock *sk, struct hci_dev *hdev, void *data, { struct mgmt_cp_set_local_name *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; int err; bt_dev_dbg(hdev, "sock %p", sk); @@ -3319,30 +3324,24 @@ static int set_local_name(struct sock *sk, struct hci_dev *hdev, void *data, } cmd = mgmt_pending_add(sk, MGMT_OP_SET_LOCAL_NAME, hdev, data, len); - if (!cmd) { + if (!cmd) err = -ENOMEM; + else + err = hci_cmd_sync_queue(hdev, set_name_sync, cmd, + set_name_complete); + + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_LOCAL_NAME, + MGMT_STATUS_FAILED); + + if (cmd) + mgmt_pending_remove(cmd); + goto failed; } memcpy(hdev->dev_name, cp->name, sizeof(hdev->dev_name)); - hci_req_init(&req, hdev); - - if (lmp_bredr_capable(hdev)) { - __hci_req_update_name(&req); - __hci_req_update_eir(&req); - } - - /* The name is stored in the scan response data and so - * no need to update the advertising data here. - */ - if (lmp_le_capable(hdev) && hci_dev_test_flag(hdev, HCI_ADVERTISING)) - __hci_req_update_scan_rsp_data(&req, hdev->cur_adv_instance); - - err = hci_req_run(&req, set_name_complete); - if (err < 0) - mgmt_pending_remove(cmd); - failed: hci_dev_unlock(hdev); return err; From 71efbb08b538d925733e0e07d4f48d996292ac5e Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:55 -0700 Subject: [PATCH 0048/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_PHY_CONFIGURATION mgmt-test paths: Set PHY 2m Success Set PHY coded Succcess Set PHY 1m 2m coded Succcess Set PHY 2m tx success Set PHY 2m rx success Set PHY Invalid Param Start Discovery - (2m, Scan Param) Start Discovery - (coded, Scan Param) Start Discovery - (1m, 2m, coded, Scan Param) Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- net/bluetooth/mgmt.c | 120 ++++++++++++++++++++++++------------------- 1 file changed, 68 insertions(+), 52 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 2585eab1fe21..b8b3176e05b6 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -3414,23 +3414,26 @@ int mgmt_phy_configuration_changed(struct hci_dev *hdev, struct sock *skip) sizeof(ev), skip); } -static void set_default_phy_complete(struct hci_dev *hdev, u8 status, - u16 opcode, struct sk_buff *skb) +static void set_default_phy_complete(struct hci_dev *hdev, void *data, int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; + struct sk_buff *skb = cmd->skb; + u8 status = mgmt_status(err); - bt_dev_dbg(hdev, "status 0x%02x", status); + if (!status) { + if (!skb) + status = MGMT_STATUS_FAILED; + else if (IS_ERR(skb)) + status = mgmt_status(PTR_ERR(skb)); + else + status = mgmt_status(skb->data[0]); + } - hci_dev_lock(hdev); - - cmd = pending_find(MGMT_OP_SET_PHY_CONFIGURATION, hdev); - if (!cmd) - goto unlock; + bt_dev_dbg(hdev, "status %d", status); if (status) { mgmt_cmd_status(cmd->sk, hdev->id, - MGMT_OP_SET_PHY_CONFIGURATION, - mgmt_status(status)); + MGMT_OP_SET_PHY_CONFIGURATION, status); } else { mgmt_cmd_complete(cmd->sk, hdev->id, MGMT_OP_SET_PHY_CONFIGURATION, 0, @@ -3439,19 +3442,56 @@ static void set_default_phy_complete(struct hci_dev *hdev, u8 status, mgmt_phy_configuration_changed(hdev, cmd->sk); } - mgmt_pending_remove(cmd); + if (skb && !IS_ERR(skb)) + kfree_skb(skb); -unlock: - hci_dev_unlock(hdev); + mgmt_pending_remove(cmd); +} + +static int set_default_phy_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_cp_set_phy_configuration *cp = cmd->param; + struct hci_cp_le_set_default_phy cp_phy; + u32 selected_phys = __le32_to_cpu(cp->selected_phys); + + memset(&cp_phy, 0, sizeof(cp_phy)); + + if (!(selected_phys & MGMT_PHY_LE_TX_MASK)) + cp_phy.all_phys |= 0x01; + + if (!(selected_phys & MGMT_PHY_LE_RX_MASK)) + cp_phy.all_phys |= 0x02; + + if (selected_phys & MGMT_PHY_LE_1M_TX) + cp_phy.tx_phys |= HCI_LE_SET_PHY_1M; + + if (selected_phys & MGMT_PHY_LE_2M_TX) + cp_phy.tx_phys |= HCI_LE_SET_PHY_2M; + + if (selected_phys & MGMT_PHY_LE_CODED_TX) + cp_phy.tx_phys |= HCI_LE_SET_PHY_CODED; + + if (selected_phys & MGMT_PHY_LE_1M_RX) + cp_phy.rx_phys |= HCI_LE_SET_PHY_1M; + + if (selected_phys & MGMT_PHY_LE_2M_RX) + cp_phy.rx_phys |= HCI_LE_SET_PHY_2M; + + if (selected_phys & MGMT_PHY_LE_CODED_RX) + cp_phy.rx_phys |= HCI_LE_SET_PHY_CODED; + + cmd->skb = __hci_cmd_sync(hdev, HCI_OP_LE_SET_DEFAULT_PHY, + sizeof(cp_phy), &cp_phy, HCI_CMD_TIMEOUT); + + return 0; } static int set_phy_configuration(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_cp_set_phy_configuration *cp = data; - struct hci_cp_le_set_default_phy cp_phy; struct mgmt_pending_cmd *cmd; - struct hci_request req; u32 selected_phys, configurable_phys, supported_phys, unconfigure_phys; u16 pkt_type = (HCI_DH1 | HCI_DM1); bool changed = false; @@ -3555,45 +3595,21 @@ static int set_phy_configuration(struct sock *sk, struct hci_dev *hdev, cmd = mgmt_pending_add(sk, MGMT_OP_SET_PHY_CONFIGURATION, hdev, data, len); - if (!cmd) { + if (!cmd) err = -ENOMEM; - goto unlock; + else + err = hci_cmd_sync_queue(hdev, set_default_phy_sync, cmd, + set_default_phy_complete); + + if (err < 0) { + err = mgmt_cmd_status(sk, hdev->id, + MGMT_OP_SET_PHY_CONFIGURATION, + MGMT_STATUS_FAILED); + + if (cmd) + mgmt_pending_remove(cmd); } - hci_req_init(&req, hdev); - - memset(&cp_phy, 0, sizeof(cp_phy)); - - if (!(selected_phys & MGMT_PHY_LE_TX_MASK)) - cp_phy.all_phys |= 0x01; - - if (!(selected_phys & MGMT_PHY_LE_RX_MASK)) - cp_phy.all_phys |= 0x02; - - if (selected_phys & MGMT_PHY_LE_1M_TX) - cp_phy.tx_phys |= HCI_LE_SET_PHY_1M; - - if (selected_phys & MGMT_PHY_LE_2M_TX) - cp_phy.tx_phys |= HCI_LE_SET_PHY_2M; - - if (selected_phys & MGMT_PHY_LE_CODED_TX) - cp_phy.tx_phys |= HCI_LE_SET_PHY_CODED; - - if (selected_phys & MGMT_PHY_LE_1M_RX) - cp_phy.rx_phys |= HCI_LE_SET_PHY_1M; - - if (selected_phys & MGMT_PHY_LE_2M_RX) - cp_phy.rx_phys |= HCI_LE_SET_PHY_2M; - - if (selected_phys & MGMT_PHY_LE_CODED_RX) - cp_phy.rx_phys |= HCI_LE_SET_PHY_CODED; - - hci_req_add(&req, HCI_OP_LE_SET_DEFAULT_PHY, sizeof(cp_phy), &cp_phy); - - err = hci_req_run_skb(&req, set_default_phy_complete); - if (err < 0) - mgmt_pending_remove(cmd); - unlock: hci_dev_unlock(hdev); From 26ac4c56f03f64a6634c5f2323c22e550b7e76b4 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:56 -0700 Subject: [PATCH 0049/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_ADVERTISING mgmt-test paths: Set powered on - Privacy and Advertising Set Advertising on - Success 2 Set Advertising on - Appearance 1 Set Advertising on - Local name 1 Set Advertising on - Name + Appear 1 Add Advertising - Success 4 Add Advertising - Success 5 Add Ext Advertising - Success 4 Add Ext Advertising - Success 5 Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- net/bluetooth/mgmt.c | 103 ++++++++++++++++++++----------------------- 1 file changed, 48 insertions(+), 55 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index b8b3176e05b6..48a901c93141 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5608,29 +5608,25 @@ static int set_device_id(struct sock *sk, struct hci_dev *hdev, void *data, return err; } -static void enable_advertising_instance(struct hci_dev *hdev, u8 status, - u16 opcode) +static void enable_advertising_instance(struct hci_dev *hdev, int err) { - bt_dev_dbg(hdev, "status %u", status); + if (err) + bt_dev_err(hdev, "failed to re-configure advertising %d", err); + else + bt_dev_dbg(hdev, "status %d", err); } -static void set_advertising_complete(struct hci_dev *hdev, u8 status, - u16 opcode) +static void set_advertising_complete(struct hci_dev *hdev, void *data, int err) { struct cmd_lookup match = { NULL, hdev }; - struct hci_request req; u8 instance; struct adv_info *adv_instance; - int err; - - hci_dev_lock(hdev); + u8 status = mgmt_status(err); if (status) { - u8 mgmt_err = mgmt_status(status); - mgmt_pending_foreach(MGMT_OP_SET_ADVERTISING, hdev, - cmd_status_rsp, &mgmt_err); - goto unlock; + cmd_status_rsp, &status); + return; } if (hci_dev_test_flag(hdev, HCI_LE_ADV)) @@ -5662,30 +5658,55 @@ static void set_advertising_complete(struct hci_dev *hdev, u8 status, */ if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || list_empty(&hdev->adv_instances)) - goto unlock; + return; instance = hdev->cur_adv_instance; if (!instance) { adv_instance = list_first_entry_or_null(&hdev->adv_instances, struct adv_info, list); if (!adv_instance) - goto unlock; + return; instance = adv_instance->instance; } - hci_req_init(&req, hdev); + err = hci_schedule_adv_instance_sync(hdev, instance, true); - err = __hci_req_schedule_adv_instance(&req, instance, true); + enable_advertising_instance(hdev, err); +} - if (!err) - err = hci_req_run(&req, enable_advertising_instance); +static int set_adv_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + u8 val = !!cp->val; - if (err) - bt_dev_err(hdev, "failed to re-configure advertising"); + if (cp->val == 0x02) + hci_dev_set_flag(hdev, HCI_ADVERTISING_CONNECTABLE); + else + hci_dev_clear_flag(hdev, HCI_ADVERTISING_CONNECTABLE); -unlock: - hci_dev_unlock(hdev); + cancel_adv_timeout(hdev); + + if (val) { + /* Switch to instance "0" for the Set Advertising setting. + * We cannot use update_[adv|scan_rsp]_data() here as the + * HCI_ADVERTISING flag is not yet set. + */ + hdev->cur_adv_instance = 0x00; + + if (ext_adv_capable(hdev)) { + hci_start_ext_adv_sync(hdev, 0x00); + } else { + hci_update_adv_data_sync(hdev, 0x00); + hci_update_scan_rsp_data_sync(hdev, 0x00); + hci_enable_advertising_sync(hdev); + } + } else { + hci_disable_advertising_sync(hdev); + } + + return 0; } static int set_advertising(struct sock *sk, struct hci_dev *hdev, void *data, @@ -5693,7 +5714,6 @@ static int set_advertising(struct sock *sk, struct hci_dev *hdev, void *data, { struct mgmt_mode *cp = data; struct mgmt_pending_cmd *cmd; - struct hci_request req; u8 val, status; int err; @@ -5759,40 +5779,13 @@ static int set_advertising(struct sock *sk, struct hci_dev *hdev, void *data, } cmd = mgmt_pending_add(sk, MGMT_OP_SET_ADVERTISING, hdev, data, len); - if (!cmd) { + if (!cmd) err = -ENOMEM; - goto unlock; - } - - hci_req_init(&req, hdev); - - if (cp->val == 0x02) - hci_dev_set_flag(hdev, HCI_ADVERTISING_CONNECTABLE); else - hci_dev_clear_flag(hdev, HCI_ADVERTISING_CONNECTABLE); + err = hci_cmd_sync_queue(hdev, set_adv_sync, cmd, + set_advertising_complete); - cancel_adv_timeout(hdev); - - if (val) { - /* Switch to instance "0" for the Set Advertising setting. - * We cannot use update_[adv|scan_rsp]_data() here as the - * HCI_ADVERTISING flag is not yet set. - */ - hdev->cur_adv_instance = 0x00; - - if (ext_adv_capable(hdev)) { - __hci_req_start_ext_adv(&req, 0x00); - } else { - __hci_req_update_adv_data(&req, 0x00); - __hci_req_update_scan_rsp_data(&req, 0x00); - __hci_req_enable_advertising(&req); - } - } else { - __hci_req_disable_advertising(&req); - } - - err = hci_req_run(&req, set_advertising_complete); - if (err < 0) + if (err < 0 && cmd) mgmt_pending_remove(cmd); unlock: From 5e233ed59cc4e1c5f2fa44a7eb77151ee1ded8e3 Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:57 -0700 Subject: [PATCH 0050/2048] Bluetooth: hci_sync: Convert adv_expire mgmt-test paths: Set Advertising on - Appearance 1 Set Advertising on - Local name 1 Set Advertising on - Name + Appear 1 Set Local Name - Success 2 Set Local Name - Success 3 Add Advertising - Success (Empty ScRsp) Add Advertising - Success (ScRsp appear) Add Advertising - Invalid Params (ScRsp appear long) Add Advertising - Success (Complete name) Add Advertising - Success (Shortened name) Add Advertising - Success (Short name) Add Advertising - Success (Name + data) Add Advertising - Invalid Params (Name + data) Add Advertising - Success (Name+data+appear) Set appearance - BR/EDR LE Set appearance - LE only Add Ext Advertising - Success (Empty ScRsp) Add Ext Advertising - Success (ScRsp appear) Add Ext Advertising - Invalid Params (ScRsp appear long) Add Ext Advertising - Success (Complete name) Add Ext Advertising - Success (Shortened name) Add Ext Advertising - Success (Short name) Add Ext Advertising - Success (Name + data) Add Ext Advertising - Invalid Params (Name + data) Add Ext Advertising - Success (Name+data+appear) Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- net/bluetooth/mgmt.c | 33 +++++++++++++++++++-------------- 1 file changed, 19 insertions(+), 14 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 48a901c93141..ee144fadcab6 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -3217,33 +3217,32 @@ static int user_passkey_neg_reply(struct sock *sk, struct hci_dev *hdev, HCI_OP_USER_PASSKEY_NEG_REPLY, 0); } -static void adv_expire(struct hci_dev *hdev, u32 flags) +static int adv_expire_sync(struct hci_dev *hdev, u32 flags) { struct adv_info *adv_instance; - struct hci_request req; - int err; adv_instance = hci_find_adv_instance(hdev, hdev->cur_adv_instance); if (!adv_instance) - return; + return 0; /* stop if current instance doesn't need to be changed */ if (!(adv_instance->flags & flags)) - return; + return 0; cancel_adv_timeout(hdev); adv_instance = hci_get_next_instance(hdev, adv_instance->instance); if (!adv_instance) - return; + return 0; - hci_req_init(&req, hdev); - err = __hci_req_schedule_adv_instance(&req, adv_instance->instance, - true); - if (err) - return; + hci_schedule_adv_instance_sync(hdev, adv_instance->instance, true); - hci_req_run(&req, NULL); + return 0; +} + +static int name_changed_sync(struct hci_dev *hdev, void *data) +{ + return adv_expire_sync(hdev, MGMT_ADV_FLAG_LOCAL_NAME); } static void set_name_complete(struct hci_dev *hdev, void *data, int err) @@ -3262,7 +3261,7 @@ static void set_name_complete(struct hci_dev *hdev, void *data, int err) cp, sizeof(*cp)); if (hci_dev_test_flag(hdev, HCI_LE_ADV)) - adv_expire(hdev, MGMT_ADV_FLAG_LOCAL_NAME); + hci_cmd_sync_queue(hdev, name_changed_sync, NULL, NULL); } mgmt_pending_remove(cmd); @@ -3347,6 +3346,11 @@ failed: return err; } +static int appearance_changed_sync(struct hci_dev *hdev, void *data) +{ + return adv_expire_sync(hdev, MGMT_ADV_FLAG_APPEARANCE); +} + static int set_appearance(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -3368,7 +3372,8 @@ static int set_appearance(struct sock *sk, struct hci_dev *hdev, void *data, hdev->appearance = appearance; if (hci_dev_test_flag(hdev, HCI_LE_ADV)) - adv_expire(hdev, MGMT_ADV_FLAG_APPEARANCE); + hci_cmd_sync_queue(hdev, appearance_changed_sync, NULL, + NULL); ext_info_changed(hdev, sk); } From 3244845c6307fa6f4fa2eabe5259d2c93c837dce Mon Sep 17 00:00:00 2001 From: Brian Gix Date: Wed, 27 Oct 2021 16:58:58 -0700 Subject: [PATCH 0051/2048] Bluetooth: hci_sync: Convert MGMT_OP_SSP mgmt-tester paths: Set SSP on - Success 2 Set Device ID - SSP off and Power on Signed-off-by: Brian Gix Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 1 - include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_event.c | 4 +- net/bluetooth/hci_sync.c | 7 +- net/bluetooth/mgmt.c | 150 +++++++++++++++---------------- 5 files changed, 80 insertions(+), 83 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index a6b075203cbe..3e53c845ab0e 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -1806,7 +1806,6 @@ int mgmt_user_passkey_notify(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 entered); void mgmt_auth_failed(struct hci_conn *conn, u8 status); void mgmt_auth_enable_complete(struct hci_dev *hdev, u8 status); -void mgmt_ssp_enable_complete(struct hci_dev *hdev, u8 enable, u8 status); void mgmt_set_class_of_dev_complete(struct hci_dev *hdev, u8 *dev_class, u8 status); void mgmt_set_local_name_complete(struct hci_dev *hdev, u8 *name, u8 status); diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index d9f2e3182ed8..db96546d40c8 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -47,6 +47,7 @@ int hci_update_class_sync(struct hci_dev *hdev); int hci_update_eir_sync(struct hci_dev *hdev); int hci_update_class_sync(struct hci_dev *hdev); int hci_update_name_sync(struct hci_dev *hdev); +int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode); int hci_update_random_address_sync(struct hci_dev *hdev, bool require_privacy, bool rpa, u8 *own_addr_type); diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 64008697aa23..0f5761b94f39 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -545,9 +545,7 @@ static void hci_cc_write_ssp_mode(struct hci_dev *hdev, struct sk_buff *skb) hdev->features[1][0] &= ~LMP_HOST_SSP; } - if (hci_dev_test_flag(hdev, HCI_MGMT)) - mgmt_ssp_enable_complete(hdev, sent->mode, status); - else if (!status) { + if (!status) { if (sent->mode) hci_dev_set_flag(hdev, HCI_SSP_ENABLED); else diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 47beabe48257..8b63b4e5c96a 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -2142,7 +2142,7 @@ int hci_write_sc_support_sync(struct hci_dev *hdev, u8 val) return err; } -static int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) +int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) { int err; @@ -2150,6 +2150,11 @@ static int hci_write_ssp_mode_sync(struct hci_dev *hdev, u8 mode) lmp_host_ssp_capable(hdev)) return 0; + if (!mode && hci_dev_test_flag(hdev, HCI_USE_DEBUG_KEYS)) { + __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SSP_DEBUG_MODE, + sizeof(mode), &mode, HCI_CMD_TIMEOUT); + } + err = __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SSP_MODE, sizeof(mode), &mode, HCI_CMD_TIMEOUT); if (err) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ee144fadcab6..bfa08eb4c2b6 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1760,6 +1760,69 @@ failed: return err; } +static void set_ssp_complete(struct hci_dev *hdev, void *data, int err) +{ + struct cmd_lookup match = { NULL, hdev }; + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + u8 enable = cp->val; + bool changed; + + if (err) { + u8 mgmt_err = mgmt_status(err); + + if (enable && hci_dev_test_and_clear_flag(hdev, + HCI_SSP_ENABLED)) { + hci_dev_clear_flag(hdev, HCI_HS_ENABLED); + new_settings(hdev, NULL); + } + + mgmt_pending_foreach(MGMT_OP_SET_SSP, hdev, cmd_status_rsp, + &mgmt_err); + return; + } + + if (enable) { + changed = !hci_dev_test_and_set_flag(hdev, HCI_SSP_ENABLED); + } else { + changed = hci_dev_test_and_clear_flag(hdev, HCI_SSP_ENABLED); + + if (!changed) + changed = hci_dev_test_and_clear_flag(hdev, + HCI_HS_ENABLED); + else + hci_dev_clear_flag(hdev, HCI_HS_ENABLED); + } + + mgmt_pending_foreach(MGMT_OP_SET_SSP, hdev, settings_rsp, &match); + + if (changed) + new_settings(hdev, match.sk); + + if (match.sk) + sock_put(match.sk); + + hci_update_eir_sync(hdev); +} + +static int set_ssp_sync(struct hci_dev *hdev, void *data) +{ + struct mgmt_pending_cmd *cmd = data; + struct mgmt_mode *cp = cmd->param; + bool changed = false; + int err; + + if (cp->val) + changed = !hci_dev_test_and_set_flag(hdev, HCI_SSP_ENABLED); + + err = hci_write_ssp_mode_sync(hdev, cp->val); + + if (!err && changed) + hci_dev_clear_flag(hdev, HCI_SSP_ENABLED); + + return err; +} + static int set_ssp(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { struct mgmt_mode *cp = data; @@ -1821,19 +1884,18 @@ static int set_ssp(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) } cmd = mgmt_pending_add(sk, MGMT_OP_SET_SSP, hdev, data, len); - if (!cmd) { + if (!cmd) err = -ENOMEM; - goto failed; - } + else + err = hci_cmd_sync_queue(hdev, set_ssp_sync, cmd, + set_ssp_complete); - if (!cp->val && hci_dev_test_flag(hdev, HCI_USE_DEBUG_KEYS)) - hci_send_cmd(hdev, HCI_OP_WRITE_SSP_DEBUG_MODE, - sizeof(cp->val), &cp->val); - - err = hci_send_cmd(hdev, HCI_OP_WRITE_SSP_MODE, 1, &cp->val); if (err < 0) { - mgmt_pending_remove(cmd); - goto failed; + err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_SSP, + MGMT_STATUS_FAILED); + + if (cmd) + mgmt_pending_remove(cmd); } failed: @@ -9309,74 +9371,6 @@ void mgmt_auth_enable_complete(struct hci_dev *hdev, u8 status) sock_put(match.sk); } -static void clear_eir(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - struct hci_cp_write_eir cp; - - if (!lmp_ext_inq_capable(hdev)) - return; - - memset(hdev->eir, 0, sizeof(hdev->eir)); - - memset(&cp, 0, sizeof(cp)); - - hci_req_add(req, HCI_OP_WRITE_EIR, sizeof(cp), &cp); -} - -void mgmt_ssp_enable_complete(struct hci_dev *hdev, u8 enable, u8 status) -{ - struct cmd_lookup match = { NULL, hdev }; - struct hci_request req; - bool changed = false; - - if (status) { - u8 mgmt_err = mgmt_status(status); - - if (enable && hci_dev_test_and_clear_flag(hdev, - HCI_SSP_ENABLED)) { - hci_dev_clear_flag(hdev, HCI_HS_ENABLED); - new_settings(hdev, NULL); - } - - mgmt_pending_foreach(MGMT_OP_SET_SSP, hdev, cmd_status_rsp, - &mgmt_err); - return; - } - - if (enable) { - changed = !hci_dev_test_and_set_flag(hdev, HCI_SSP_ENABLED); - } else { - changed = hci_dev_test_and_clear_flag(hdev, HCI_SSP_ENABLED); - if (!changed) - changed = hci_dev_test_and_clear_flag(hdev, - HCI_HS_ENABLED); - else - hci_dev_clear_flag(hdev, HCI_HS_ENABLED); - } - - mgmt_pending_foreach(MGMT_OP_SET_SSP, hdev, settings_rsp, &match); - - if (changed) - new_settings(hdev, match.sk); - - if (match.sk) - sock_put(match.sk); - - hci_req_init(&req, hdev); - - if (hci_dev_test_flag(hdev, HCI_SSP_ENABLED)) { - if (hci_dev_test_flag(hdev, HCI_USE_DEBUG_KEYS)) - hci_req_add(&req, HCI_OP_WRITE_SSP_DEBUG_MODE, - sizeof(enable), &enable); - __hci_req_update_eir(&req); - } else { - clear_eir(&req); - } - - hci_req_run(&req, NULL); -} - static void sk_lookup(struct mgmt_pending_cmd *cmd, void *data) { struct cmd_lookup *match = data; From d0b137062b2de75b264b84143d21c98abc5f5ad2 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:58:59 -0700 Subject: [PATCH 0052/2048] Bluetooth: hci_sync: Rework init stages This moves the init stages to use the hci_sync infra and in addition to that have the stages as function tables so it is easier to change the command sequence. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_core.c | 1203 +--------------------- net/bluetooth/hci_sync.c | 1659 +++++++++++++++++++++++++++++- 3 files changed, 1657 insertions(+), 1206 deletions(-) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index db96546d40c8..487e4981cce1 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -83,6 +83,7 @@ int hci_remove_ext_adv_instance_sync(struct hci_dev *hdev, u8 instance, struct sk_buff *hci_read_local_oob_data_sync(struct hci_dev *hdev, bool ext, struct sock *sk); +int hci_reset_sync(struct hci_dev *hdev); int hci_dev_open_sync(struct hci_dev *hdev); int hci_dev_close_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 420ed6a02337..e0c0aa779dbc 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -62,824 +62,6 @@ DEFINE_MUTEX(hci_cb_list_lock); /* HCI ID Numbering */ static DEFINE_IDA(hci_index_ida); -static int hci_reset_req(struct hci_request *req, unsigned long opt) -{ - BT_DBG("%s %ld", req->hdev->name, opt); - - /* Reset device */ - set_bit(HCI_RESET, &req->hdev->flags); - hci_req_add(req, HCI_OP_RESET, 0, NULL); - return 0; -} - -static void bredr_init(struct hci_request *req) -{ - req->hdev->flow_ctl_mode = HCI_FLOW_CTL_MODE_PACKET_BASED; - - /* Read Local Supported Features */ - hci_req_add(req, HCI_OP_READ_LOCAL_FEATURES, 0, NULL); - - /* Read Local Version */ - hci_req_add(req, HCI_OP_READ_LOCAL_VERSION, 0, NULL); - - /* Read BD Address */ - hci_req_add(req, HCI_OP_READ_BD_ADDR, 0, NULL); -} - -static void amp_init1(struct hci_request *req) -{ - req->hdev->flow_ctl_mode = HCI_FLOW_CTL_MODE_BLOCK_BASED; - - /* Read Local Version */ - hci_req_add(req, HCI_OP_READ_LOCAL_VERSION, 0, NULL); - - /* Read Local Supported Commands */ - hci_req_add(req, HCI_OP_READ_LOCAL_COMMANDS, 0, NULL); - - /* Read Local AMP Info */ - hci_req_add(req, HCI_OP_READ_LOCAL_AMP_INFO, 0, NULL); - - /* Read Data Blk size */ - hci_req_add(req, HCI_OP_READ_DATA_BLOCK_SIZE, 0, NULL); - - /* Read Flow Control Mode */ - hci_req_add(req, HCI_OP_READ_FLOW_CONTROL_MODE, 0, NULL); - - /* Read Location Data */ - hci_req_add(req, HCI_OP_READ_LOCATION_DATA, 0, NULL); -} - -static int amp_init2(struct hci_request *req) -{ - /* Read Local Supported Features. Not all AMP controllers - * support this so it's placed conditionally in the second - * stage init. - */ - if (req->hdev->commands[14] & 0x20) - hci_req_add(req, HCI_OP_READ_LOCAL_FEATURES, 0, NULL); - - return 0; -} - -static int hci_init1_req(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - - BT_DBG("%s %ld", hdev->name, opt); - - /* Reset */ - if (!test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks)) - hci_reset_req(req, 0); - - switch (hdev->dev_type) { - case HCI_PRIMARY: - bredr_init(req); - break; - case HCI_AMP: - amp_init1(req); - break; - default: - bt_dev_err(hdev, "Unknown device type %d", hdev->dev_type); - break; - } - - return 0; -} - -static void bredr_setup(struct hci_request *req) -{ - __le16 param; - __u8 flt_type; - - /* Read Buffer Size (ACL mtu, max pkt, etc.) */ - hci_req_add(req, HCI_OP_READ_BUFFER_SIZE, 0, NULL); - - /* Read Class of Device */ - hci_req_add(req, HCI_OP_READ_CLASS_OF_DEV, 0, NULL); - - /* Read Local Name */ - hci_req_add(req, HCI_OP_READ_LOCAL_NAME, 0, NULL); - - /* Read Voice Setting */ - hci_req_add(req, HCI_OP_READ_VOICE_SETTING, 0, NULL); - - /* Read Number of Supported IAC */ - hci_req_add(req, HCI_OP_READ_NUM_SUPPORTED_IAC, 0, NULL); - - /* Read Current IAC LAP */ - hci_req_add(req, HCI_OP_READ_CURRENT_IAC_LAP, 0, NULL); - - /* Clear Event Filters */ - flt_type = HCI_FLT_CLEAR_ALL; - hci_req_add(req, HCI_OP_SET_EVENT_FLT, 1, &flt_type); - - /* Connection accept timeout ~20 secs */ - param = cpu_to_le16(0x7d00); - hci_req_add(req, HCI_OP_WRITE_CA_TIMEOUT, 2, ¶m); -} - -static void le_setup(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - - /* Read LE Buffer Size */ - hci_req_add(req, HCI_OP_LE_READ_BUFFER_SIZE, 0, NULL); - - /* Read LE Local Supported Features */ - hci_req_add(req, HCI_OP_LE_READ_LOCAL_FEATURES, 0, NULL); - - /* Read LE Supported States */ - hci_req_add(req, HCI_OP_LE_READ_SUPPORTED_STATES, 0, NULL); - - /* LE-only controllers have LE implicitly enabled */ - if (!lmp_bredr_capable(hdev)) - hci_dev_set_flag(hdev, HCI_LE_ENABLED); -} - -static void hci_setup_event_mask(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - - /* The second byte is 0xff instead of 0x9f (two reserved bits - * disabled) since a Broadcom 1.2 dongle doesn't respond to the - * command otherwise. - */ - u8 events[8] = { 0xff, 0xff, 0xfb, 0xff, 0x00, 0x00, 0x00, 0x00 }; - - /* CSR 1.1 dongles does not accept any bitfield so don't try to set - * any event mask for pre 1.2 devices. - */ - if (hdev->hci_ver < BLUETOOTH_VER_1_2) - return; - - if (lmp_bredr_capable(hdev)) { - events[4] |= 0x01; /* Flow Specification Complete */ - } else { - /* Use a different default for LE-only devices */ - memset(events, 0, sizeof(events)); - events[1] |= 0x20; /* Command Complete */ - events[1] |= 0x40; /* Command Status */ - events[1] |= 0x80; /* Hardware Error */ - - /* If the controller supports the Disconnect command, enable - * the corresponding event. In addition enable packet flow - * control related events. - */ - if (hdev->commands[0] & 0x20) { - events[0] |= 0x10; /* Disconnection Complete */ - events[2] |= 0x04; /* Number of Completed Packets */ - events[3] |= 0x02; /* Data Buffer Overflow */ - } - - /* If the controller supports the Read Remote Version - * Information command, enable the corresponding event. - */ - if (hdev->commands[2] & 0x80) - events[1] |= 0x08; /* Read Remote Version Information - * Complete - */ - - if (hdev->le_features[0] & HCI_LE_ENCRYPTION) { - events[0] |= 0x80; /* Encryption Change */ - events[5] |= 0x80; /* Encryption Key Refresh Complete */ - } - } - - if (lmp_inq_rssi_capable(hdev) || - test_bit(HCI_QUIRK_FIXUP_INQUIRY_MODE, &hdev->quirks)) - events[4] |= 0x02; /* Inquiry Result with RSSI */ - - if (lmp_ext_feat_capable(hdev)) - events[4] |= 0x04; /* Read Remote Extended Features Complete */ - - if (lmp_esco_capable(hdev)) { - events[5] |= 0x08; /* Synchronous Connection Complete */ - events[5] |= 0x10; /* Synchronous Connection Changed */ - } - - if (lmp_sniffsubr_capable(hdev)) - events[5] |= 0x20; /* Sniff Subrating */ - - if (lmp_pause_enc_capable(hdev)) - events[5] |= 0x80; /* Encryption Key Refresh Complete */ - - if (lmp_ext_inq_capable(hdev)) - events[5] |= 0x40; /* Extended Inquiry Result */ - - if (lmp_no_flush_capable(hdev)) - events[7] |= 0x01; /* Enhanced Flush Complete */ - - if (lmp_lsto_capable(hdev)) - events[6] |= 0x80; /* Link Supervision Timeout Changed */ - - if (lmp_ssp_capable(hdev)) { - events[6] |= 0x01; /* IO Capability Request */ - events[6] |= 0x02; /* IO Capability Response */ - events[6] |= 0x04; /* User Confirmation Request */ - events[6] |= 0x08; /* User Passkey Request */ - events[6] |= 0x10; /* Remote OOB Data Request */ - events[6] |= 0x20; /* Simple Pairing Complete */ - events[7] |= 0x04; /* User Passkey Notification */ - events[7] |= 0x08; /* Keypress Notification */ - events[7] |= 0x10; /* Remote Host Supported - * Features Notification - */ - } - - if (lmp_le_capable(hdev)) - events[7] |= 0x20; /* LE Meta-Event */ - - hci_req_add(req, HCI_OP_SET_EVENT_MASK, sizeof(events), events); -} - -static int hci_init2_req(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - - if (hdev->dev_type == HCI_AMP) - return amp_init2(req); - - if (lmp_bredr_capable(hdev)) - bredr_setup(req); - else - hci_dev_clear_flag(hdev, HCI_BREDR_ENABLED); - - if (lmp_le_capable(hdev)) - le_setup(req); - - /* All Bluetooth 1.2 and later controllers should support the - * HCI command for reading the local supported commands. - * - * Unfortunately some controllers indicate Bluetooth 1.2 support, - * but do not have support for this command. If that is the case, - * the driver can quirk the behavior and skip reading the local - * supported commands. - */ - if (hdev->hci_ver > BLUETOOTH_VER_1_1 && - !test_bit(HCI_QUIRK_BROKEN_LOCAL_COMMANDS, &hdev->quirks)) - hci_req_add(req, HCI_OP_READ_LOCAL_COMMANDS, 0, NULL); - - if (lmp_ssp_capable(hdev)) { - /* When SSP is available, then the host features page - * should also be available as well. However some - * controllers list the max_page as 0 as long as SSP - * has not been enabled. To achieve proper debugging - * output, force the minimum max_page to 1 at least. - */ - hdev->max_page = 0x01; - - if (hci_dev_test_flag(hdev, HCI_SSP_ENABLED)) { - u8 mode = 0x01; - - hci_req_add(req, HCI_OP_WRITE_SSP_MODE, - sizeof(mode), &mode); - } else { - struct hci_cp_write_eir cp; - - memset(hdev->eir, 0, sizeof(hdev->eir)); - memset(&cp, 0, sizeof(cp)); - - hci_req_add(req, HCI_OP_WRITE_EIR, sizeof(cp), &cp); - } - } - - if (lmp_inq_rssi_capable(hdev) || - test_bit(HCI_QUIRK_FIXUP_INQUIRY_MODE, &hdev->quirks)) { - u8 mode; - - /* If Extended Inquiry Result events are supported, then - * they are clearly preferred over Inquiry Result with RSSI - * events. - */ - mode = lmp_ext_inq_capable(hdev) ? 0x02 : 0x01; - - hci_req_add(req, HCI_OP_WRITE_INQUIRY_MODE, 1, &mode); - } - - if (lmp_inq_tx_pwr_capable(hdev)) - hci_req_add(req, HCI_OP_READ_INQ_RSP_TX_POWER, 0, NULL); - - if (lmp_ext_feat_capable(hdev)) { - struct hci_cp_read_local_ext_features cp; - - cp.page = 0x01; - hci_req_add(req, HCI_OP_READ_LOCAL_EXT_FEATURES, - sizeof(cp), &cp); - } - - if (hci_dev_test_flag(hdev, HCI_LINK_SECURITY)) { - u8 enable = 1; - hci_req_add(req, HCI_OP_WRITE_AUTH_ENABLE, sizeof(enable), - &enable); - } - - return 0; -} - -static void hci_setup_link_policy(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - struct hci_cp_write_def_link_policy cp; - u16 link_policy = 0; - - if (lmp_rswitch_capable(hdev)) - link_policy |= HCI_LP_RSWITCH; - if (lmp_hold_capable(hdev)) - link_policy |= HCI_LP_HOLD; - if (lmp_sniff_capable(hdev)) - link_policy |= HCI_LP_SNIFF; - if (lmp_park_capable(hdev)) - link_policy |= HCI_LP_PARK; - - cp.policy = cpu_to_le16(link_policy); - hci_req_add(req, HCI_OP_WRITE_DEF_LINK_POLICY, sizeof(cp), &cp); -} - -static void hci_set_le_support(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - struct hci_cp_write_le_host_supported cp; - - /* LE-only devices do not support explicit enablement */ - if (!lmp_bredr_capable(hdev)) - return; - - memset(&cp, 0, sizeof(cp)); - - if (hci_dev_test_flag(hdev, HCI_LE_ENABLED)) { - cp.le = 0x01; - cp.simul = 0x00; - } - - if (cp.le != lmp_host_le_capable(hdev)) - hci_req_add(req, HCI_OP_WRITE_LE_HOST_SUPPORTED, sizeof(cp), - &cp); -} - -static void hci_set_event_mask_page_2(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - u8 events[8] = { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; - bool changed = false; - - /* If Connectionless Peripheral Broadcast central role is supported - * enable all necessary events for it. - */ - if (lmp_cpb_central_capable(hdev)) { - events[1] |= 0x40; /* Triggered Clock Capture */ - events[1] |= 0x80; /* Synchronization Train Complete */ - events[2] |= 0x10; /* Peripheral Page Response Timeout */ - events[2] |= 0x20; /* CPB Channel Map Change */ - changed = true; - } - - /* If Connectionless Peripheral Broadcast peripheral role is supported - * enable all necessary events for it. - */ - if (lmp_cpb_peripheral_capable(hdev)) { - events[2] |= 0x01; /* Synchronization Train Received */ - events[2] |= 0x02; /* CPB Receive */ - events[2] |= 0x04; /* CPB Timeout */ - events[2] |= 0x08; /* Truncated Page Complete */ - changed = true; - } - - /* Enable Authenticated Payload Timeout Expired event if supported */ - if (lmp_ping_capable(hdev) || hdev->le_features[0] & HCI_LE_PING) { - events[2] |= 0x80; - changed = true; - } - - /* Some Broadcom based controllers indicate support for Set Event - * Mask Page 2 command, but then actually do not support it. Since - * the default value is all bits set to zero, the command is only - * required if the event mask has to be changed. In case no change - * to the event mask is needed, skip this command. - */ - if (changed) - hci_req_add(req, HCI_OP_SET_EVENT_MASK_PAGE_2, - sizeof(events), events); -} - -static int hci_init3_req(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - u8 p; - - hci_setup_event_mask(req); - - if (hdev->commands[6] & 0x20 && - !test_bit(HCI_QUIRK_BROKEN_STORED_LINK_KEY, &hdev->quirks)) { - struct hci_cp_read_stored_link_key cp; - - bacpy(&cp.bdaddr, BDADDR_ANY); - cp.read_all = 0x01; - hci_req_add(req, HCI_OP_READ_STORED_LINK_KEY, sizeof(cp), &cp); - } - - if (hdev->commands[5] & 0x10) - hci_setup_link_policy(req); - - if (hdev->commands[8] & 0x01) - hci_req_add(req, HCI_OP_READ_PAGE_SCAN_ACTIVITY, 0, NULL); - - if (hdev->commands[18] & 0x04 && - !test_bit(HCI_QUIRK_BROKEN_ERR_DATA_REPORTING, &hdev->quirks)) - hci_req_add(req, HCI_OP_READ_DEF_ERR_DATA_REPORTING, 0, NULL); - - /* Some older Broadcom based Bluetooth 1.2 controllers do not - * support the Read Page Scan Type command. Check support for - * this command in the bit mask of supported commands. - */ - if (hdev->commands[13] & 0x01) - hci_req_add(req, HCI_OP_READ_PAGE_SCAN_TYPE, 0, NULL); - - if (lmp_le_capable(hdev)) { - u8 events[8]; - - memset(events, 0, sizeof(events)); - - if (hdev->le_features[0] & HCI_LE_ENCRYPTION) - events[0] |= 0x10; /* LE Long Term Key Request */ - - /* If controller supports the Connection Parameters Request - * Link Layer Procedure, enable the corresponding event. - */ - if (hdev->le_features[0] & HCI_LE_CONN_PARAM_REQ_PROC) - events[0] |= 0x20; /* LE Remote Connection - * Parameter Request - */ - - /* If the controller supports the Data Length Extension - * feature, enable the corresponding event. - */ - if (hdev->le_features[0] & HCI_LE_DATA_LEN_EXT) - events[0] |= 0x40; /* LE Data Length Change */ - - /* If the controller supports LL Privacy feature, enable - * the corresponding event. - */ - if (hdev->le_features[0] & HCI_LE_LL_PRIVACY) - events[1] |= 0x02; /* LE Enhanced Connection - * Complete - */ - - /* If the controller supports Extended Scanner Filter - * Policies, enable the corresponding event. - */ - if (hdev->le_features[0] & HCI_LE_EXT_SCAN_POLICY) - events[1] |= 0x04; /* LE Direct Advertising - * Report - */ - - /* If the controller supports Channel Selection Algorithm #2 - * feature, enable the corresponding event. - */ - if (hdev->le_features[1] & HCI_LE_CHAN_SEL_ALG2) - events[2] |= 0x08; /* LE Channel Selection - * Algorithm - */ - - /* If the controller supports the LE Set Scan Enable command, - * enable the corresponding advertising report event. - */ - if (hdev->commands[26] & 0x08) - events[0] |= 0x02; /* LE Advertising Report */ - - /* If the controller supports the LE Create Connection - * command, enable the corresponding event. - */ - if (hdev->commands[26] & 0x10) - events[0] |= 0x01; /* LE Connection Complete */ - - /* If the controller supports the LE Connection Update - * command, enable the corresponding event. - */ - if (hdev->commands[27] & 0x04) - events[0] |= 0x04; /* LE Connection Update - * Complete - */ - - /* If the controller supports the LE Read Remote Used Features - * command, enable the corresponding event. - */ - if (hdev->commands[27] & 0x20) - events[0] |= 0x08; /* LE Read Remote Used - * Features Complete - */ - - /* If the controller supports the LE Read Local P-256 - * Public Key command, enable the corresponding event. - */ - if (hdev->commands[34] & 0x02) - events[0] |= 0x80; /* LE Read Local P-256 - * Public Key Complete - */ - - /* If the controller supports the LE Generate DHKey - * command, enable the corresponding event. - */ - if (hdev->commands[34] & 0x04) - events[1] |= 0x01; /* LE Generate DHKey Complete */ - - /* If the controller supports the LE Set Default PHY or - * LE Set PHY commands, enable the corresponding event. - */ - if (hdev->commands[35] & (0x20 | 0x40)) - events[1] |= 0x08; /* LE PHY Update Complete */ - - /* If the controller supports LE Set Extended Scan Parameters - * and LE Set Extended Scan Enable commands, enable the - * corresponding event. - */ - if (use_ext_scan(hdev)) - events[1] |= 0x10; /* LE Extended Advertising - * Report - */ - - /* If the controller supports the LE Extended Advertising - * command, enable the corresponding event. - */ - if (ext_adv_capable(hdev)) - events[2] |= 0x02; /* LE Advertising Set - * Terminated - */ - - hci_req_add(req, HCI_OP_LE_SET_EVENT_MASK, sizeof(events), - events); - - /* Read LE Advertising Channel TX Power */ - if ((hdev->commands[25] & 0x40) && !ext_adv_capable(hdev)) { - /* HCI TS spec forbids mixing of legacy and extended - * advertising commands wherein READ_ADV_TX_POWER is - * also included. So do not call it if extended adv - * is supported otherwise controller will return - * COMMAND_DISALLOWED for extended commands. - */ - hci_req_add(req, HCI_OP_LE_READ_ADV_TX_POWER, 0, NULL); - } - - if (hdev->commands[38] & 0x80) { - /* Read LE Min/Max Tx Power*/ - hci_req_add(req, HCI_OP_LE_READ_TRANSMIT_POWER, - 0, NULL); - } - - if (hdev->commands[26] & 0x40) { - /* Read LE Accept List Size */ - hci_req_add(req, HCI_OP_LE_READ_ACCEPT_LIST_SIZE, - 0, NULL); - } - - if (hdev->commands[26] & 0x80) { - /* Clear LE Accept List */ - hci_req_add(req, HCI_OP_LE_CLEAR_ACCEPT_LIST, 0, NULL); - } - - if (hdev->commands[34] & 0x40) { - /* Read LE Resolving List Size */ - hci_req_add(req, HCI_OP_LE_READ_RESOLV_LIST_SIZE, - 0, NULL); - } - - if (hdev->commands[34] & 0x20) { - /* Clear LE Resolving List */ - hci_req_add(req, HCI_OP_LE_CLEAR_RESOLV_LIST, 0, NULL); - } - - if (hdev->commands[35] & 0x04) { - __le16 rpa_timeout = cpu_to_le16(hdev->rpa_timeout); - - /* Set RPA timeout */ - hci_req_add(req, HCI_OP_LE_SET_RPA_TIMEOUT, 2, - &rpa_timeout); - } - - if (hdev->le_features[0] & HCI_LE_DATA_LEN_EXT) { - /* Read LE Maximum Data Length */ - hci_req_add(req, HCI_OP_LE_READ_MAX_DATA_LEN, 0, NULL); - - /* Read LE Suggested Default Data Length */ - hci_req_add(req, HCI_OP_LE_READ_DEF_DATA_LEN, 0, NULL); - } - - if (ext_adv_capable(hdev)) { - /* Read LE Number of Supported Advertising Sets */ - hci_req_add(req, HCI_OP_LE_READ_NUM_SUPPORTED_ADV_SETS, - 0, NULL); - } - - hci_set_le_support(req); - } - - /* Read features beyond page 1 if available */ - for (p = 2; p < HCI_MAX_PAGES && p <= hdev->max_page; p++) { - struct hci_cp_read_local_ext_features cp; - - cp.page = p; - hci_req_add(req, HCI_OP_READ_LOCAL_EXT_FEATURES, - sizeof(cp), &cp); - } - - return 0; -} - -static int hci_init4_req(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - - /* Some Broadcom based Bluetooth controllers do not support the - * Delete Stored Link Key command. They are clearly indicating its - * absence in the bit mask of supported commands. - * - * Check the supported commands and only if the command is marked - * as supported send it. If not supported assume that the controller - * does not have actual support for stored link keys which makes this - * command redundant anyway. - * - * Some controllers indicate that they support handling deleting - * stored link keys, but they don't. The quirk lets a driver - * just disable this command. - */ - if (hdev->commands[6] & 0x80 && - !test_bit(HCI_QUIRK_BROKEN_STORED_LINK_KEY, &hdev->quirks)) { - struct hci_cp_delete_stored_link_key cp; - - bacpy(&cp.bdaddr, BDADDR_ANY); - cp.delete_all = 0x01; - hci_req_add(req, HCI_OP_DELETE_STORED_LINK_KEY, - sizeof(cp), &cp); - } - - /* Set event mask page 2 if the HCI command for it is supported */ - if (hdev->commands[22] & 0x04) - hci_set_event_mask_page_2(req); - - /* Read local pairing options if the HCI command is supported */ - if (hdev->commands[41] & 0x08) - hci_req_add(req, HCI_OP_READ_LOCAL_PAIRING_OPTS, 0, NULL); - - /* Get MWS transport configuration if the HCI command is supported */ - if (hdev->commands[30] & 0x08) - hci_req_add(req, HCI_OP_GET_MWS_TRANSPORT_CONFIG, 0, NULL); - - /* Check for Synchronization Train support */ - if (lmp_sync_train_capable(hdev)) - hci_req_add(req, HCI_OP_READ_SYNC_TRAIN_PARAMS, 0, NULL); - - /* Enable Secure Connections if supported and configured */ - if (hci_dev_test_flag(hdev, HCI_SSP_ENABLED) && - bredr_sc_enabled(hdev)) { - u8 support = 0x01; - - hci_req_add(req, HCI_OP_WRITE_SC_SUPPORT, - sizeof(support), &support); - } - - /* Set erroneous data reporting if supported to the wideband speech - * setting value - */ - if (hdev->commands[18] & 0x08 && - !test_bit(HCI_QUIRK_BROKEN_ERR_DATA_REPORTING, &hdev->quirks)) { - bool enabled = hci_dev_test_flag(hdev, - HCI_WIDEBAND_SPEECH_ENABLED); - - if (enabled != - (hdev->err_data_reporting == ERR_DATA_REPORTING_ENABLED)) { - struct hci_cp_write_def_err_data_reporting cp; - - cp.err_data_reporting = enabled ? - ERR_DATA_REPORTING_ENABLED : - ERR_DATA_REPORTING_DISABLED; - - hci_req_add(req, HCI_OP_WRITE_DEF_ERR_DATA_REPORTING, - sizeof(cp), &cp); - } - } - - /* Set Suggested Default Data Length to maximum if supported */ - if (hdev->le_features[0] & HCI_LE_DATA_LEN_EXT) { - struct hci_cp_le_write_def_data_len cp; - - cp.tx_len = cpu_to_le16(hdev->le_max_tx_len); - cp.tx_time = cpu_to_le16(hdev->le_max_tx_time); - hci_req_add(req, HCI_OP_LE_WRITE_DEF_DATA_LEN, sizeof(cp), &cp); - } - - /* Set Default PHY parameters if command is supported */ - if (hdev->commands[35] & 0x20) { - struct hci_cp_le_set_default_phy cp; - - cp.all_phys = 0x00; - cp.tx_phys = hdev->le_tx_def_phys; - cp.rx_phys = hdev->le_rx_def_phys; - - hci_req_add(req, HCI_OP_LE_SET_DEFAULT_PHY, sizeof(cp), &cp); - } - - return 0; -} - -static int __hci_init(struct hci_dev *hdev) -{ - int err; - - err = __hci_req_sync(hdev, hci_init1_req, 0, HCI_INIT_TIMEOUT, NULL); - if (err < 0) - return err; - - if (hci_dev_test_flag(hdev, HCI_SETUP)) - hci_debugfs_create_basic(hdev); - - err = __hci_req_sync(hdev, hci_init2_req, 0, HCI_INIT_TIMEOUT, NULL); - if (err < 0) - return err; - - /* HCI_PRIMARY covers both single-mode LE, BR/EDR and dual-mode - * BR/EDR/LE type controllers. AMP controllers only need the - * first two stages of init. - */ - if (hdev->dev_type != HCI_PRIMARY) - return 0; - - err = __hci_req_sync(hdev, hci_init3_req, 0, HCI_INIT_TIMEOUT, NULL); - if (err < 0) - return err; - - err = __hci_req_sync(hdev, hci_init4_req, 0, HCI_INIT_TIMEOUT, NULL); - if (err < 0) - return err; - - /* Read local codec list if the HCI command is supported */ - if (hdev->commands[45] & 0x04) - hci_read_supported_codecs_v2(hdev); - else if (hdev->commands[29] & 0x20) - hci_read_supported_codecs(hdev); - - /* This function is only called when the controller is actually in - * configured state. When the controller is marked as unconfigured, - * this initialization procedure is not run. - * - * It means that it is possible that a controller runs through its - * setup phase and then discovers missing settings. If that is the - * case, then this function will not be called. It then will only - * be called during the config phase. - * - * So only when in setup phase or config phase, create the debugfs - * entries and register the SMP channels. - */ - if (!hci_dev_test_flag(hdev, HCI_SETUP) && - !hci_dev_test_flag(hdev, HCI_CONFIG)) - return 0; - - hci_debugfs_create_common(hdev); - - if (lmp_bredr_capable(hdev)) - hci_debugfs_create_bredr(hdev); - - if (lmp_le_capable(hdev)) - hci_debugfs_create_le(hdev); - - return 0; -} - -static int hci_init0_req(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - - BT_DBG("%s %ld", hdev->name, opt); - - /* Reset */ - if (!test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks)) - hci_reset_req(req, 0); - - /* Read Local Version */ - hci_req_add(req, HCI_OP_READ_LOCAL_VERSION, 0, NULL); - - /* Read BD Address */ - if (hdev->set_bdaddr) - hci_req_add(req, HCI_OP_READ_BD_ADDR, 0, NULL); - - return 0; -} - -static int __hci_unconf_init(struct hci_dev *hdev) -{ - int err; - - if (test_bit(HCI_QUIRK_RAW_DEVICE, &hdev->quirks)) - return 0; - - err = __hci_req_sync(hdev, hci_init0_req, 0, HCI_INIT_TIMEOUT, NULL); - if (err < 0) - return err; - - if (hci_dev_test_flag(hdev, HCI_SETUP)) - hci_debugfs_create_basic(hdev); - - return 0; -} - static int hci_scan_req(struct hci_request *req, unsigned long opt) { __u8 scan = opt; @@ -1289,240 +471,6 @@ done: return err; } -/** - * hci_dev_get_bd_addr_from_property - Get the Bluetooth Device Address - * (BD_ADDR) for a HCI device from - * a firmware node property. - * @hdev: The HCI device - * - * Search the firmware node for 'local-bd-address'. - * - * All-zero BD addresses are rejected, because those could be properties - * that exist in the firmware tables, but were not updated by the firmware. For - * example, the DTS could define 'local-bd-address', with zero BD addresses. - */ -static void hci_dev_get_bd_addr_from_property(struct hci_dev *hdev) -{ - struct fwnode_handle *fwnode = dev_fwnode(hdev->dev.parent); - bdaddr_t ba; - int ret; - - ret = fwnode_property_read_u8_array(fwnode, "local-bd-address", - (u8 *)&ba, sizeof(ba)); - if (ret < 0 || !bacmp(&ba, BDADDR_ANY)) - return; - - bacpy(&hdev->public_addr, &ba); -} - -/* TODO: Move this function into hci_sync.c */ -int hci_dev_open_sync(struct hci_dev *hdev) -{ - int ret = 0; - - BT_DBG("%s %p", hdev->name, hdev); - - if (hci_dev_test_flag(hdev, HCI_UNREGISTER)) { - ret = -ENODEV; - goto done; - } - - if (!hci_dev_test_flag(hdev, HCI_SETUP) && - !hci_dev_test_flag(hdev, HCI_CONFIG)) { - /* Check for rfkill but allow the HCI setup stage to - * proceed (which in itself doesn't cause any RF activity). - */ - if (hci_dev_test_flag(hdev, HCI_RFKILLED)) { - ret = -ERFKILL; - goto done; - } - - /* Check for valid public address or a configured static - * random address, but let the HCI setup proceed to - * be able to determine if there is a public address - * or not. - * - * In case of user channel usage, it is not important - * if a public address or static random address is - * available. - * - * This check is only valid for BR/EDR controllers - * since AMP controllers do not have an address. - */ - if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && - hdev->dev_type == HCI_PRIMARY && - !bacmp(&hdev->bdaddr, BDADDR_ANY) && - !bacmp(&hdev->static_addr, BDADDR_ANY)) { - ret = -EADDRNOTAVAIL; - goto done; - } - } - - if (test_bit(HCI_UP, &hdev->flags)) { - ret = -EALREADY; - goto done; - } - - if (hdev->open(hdev)) { - ret = -EIO; - goto done; - } - - set_bit(HCI_RUNNING, &hdev->flags); - hci_sock_dev_event(hdev, HCI_DEV_OPEN); - - atomic_set(&hdev->cmd_cnt, 1); - set_bit(HCI_INIT, &hdev->flags); - - if (hci_dev_test_flag(hdev, HCI_SETUP) || - test_bit(HCI_QUIRK_NON_PERSISTENT_SETUP, &hdev->quirks)) { - bool invalid_bdaddr; - - hci_sock_dev_event(hdev, HCI_DEV_SETUP); - - if (hdev->setup) - ret = hdev->setup(hdev); - - /* The transport driver can set the quirk to mark the - * BD_ADDR invalid before creating the HCI device or in - * its setup callback. - */ - invalid_bdaddr = test_bit(HCI_QUIRK_INVALID_BDADDR, - &hdev->quirks); - - if (ret) - goto setup_failed; - - if (test_bit(HCI_QUIRK_USE_BDADDR_PROPERTY, &hdev->quirks)) { - if (!bacmp(&hdev->public_addr, BDADDR_ANY)) - hci_dev_get_bd_addr_from_property(hdev); - - if (bacmp(&hdev->public_addr, BDADDR_ANY) && - hdev->set_bdaddr) { - ret = hdev->set_bdaddr(hdev, - &hdev->public_addr); - - /* If setting of the BD_ADDR from the device - * property succeeds, then treat the address - * as valid even if the invalid BD_ADDR - * quirk indicates otherwise. - */ - if (!ret) - invalid_bdaddr = false; - } - } - -setup_failed: - /* The transport driver can set these quirks before - * creating the HCI device or in its setup callback. - * - * For the invalid BD_ADDR quirk it is possible that - * it becomes a valid address if the bootloader does - * provide it (see above). - * - * In case any of them is set, the controller has to - * start up as unconfigured. - */ - if (test_bit(HCI_QUIRK_EXTERNAL_CONFIG, &hdev->quirks) || - invalid_bdaddr) - hci_dev_set_flag(hdev, HCI_UNCONFIGURED); - - /* For an unconfigured controller it is required to - * read at least the version information provided by - * the Read Local Version Information command. - * - * If the set_bdaddr driver callback is provided, then - * also the original Bluetooth public device address - * will be read using the Read BD Address command. - */ - if (hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) - ret = __hci_unconf_init(hdev); - } - - if (hci_dev_test_flag(hdev, HCI_CONFIG)) { - /* If public address change is configured, ensure that - * the address gets programmed. If the driver does not - * support changing the public address, fail the power - * on procedure. - */ - if (bacmp(&hdev->public_addr, BDADDR_ANY) && - hdev->set_bdaddr) - ret = hdev->set_bdaddr(hdev, &hdev->public_addr); - else - ret = -EADDRNOTAVAIL; - } - - if (!ret) { - if (!hci_dev_test_flag(hdev, HCI_UNCONFIGURED) && - !hci_dev_test_flag(hdev, HCI_USER_CHANNEL)) { - ret = __hci_init(hdev); - if (!ret && hdev->post_init) - ret = hdev->post_init(hdev); - } - } - - /* If the HCI Reset command is clearing all diagnostic settings, - * then they need to be reprogrammed after the init procedure - * completed. - */ - if (test_bit(HCI_QUIRK_NON_PERSISTENT_DIAG, &hdev->quirks) && - !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && - hci_dev_test_flag(hdev, HCI_VENDOR_DIAG) && hdev->set_diag) - ret = hdev->set_diag(hdev, true); - - msft_do_open(hdev); - aosp_do_open(hdev); - - clear_bit(HCI_INIT, &hdev->flags); - - if (!ret) { - hci_dev_hold(hdev); - hci_dev_set_flag(hdev, HCI_RPA_EXPIRED); - hci_adv_instances_set_rpa_expired(hdev, true); - set_bit(HCI_UP, &hdev->flags); - hci_sock_dev_event(hdev, HCI_DEV_UP); - hci_leds_update_powered(hdev, true); - if (!hci_dev_test_flag(hdev, HCI_SETUP) && - !hci_dev_test_flag(hdev, HCI_CONFIG) && - !hci_dev_test_flag(hdev, HCI_UNCONFIGURED) && - !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && - hci_dev_test_flag(hdev, HCI_MGMT) && - hdev->dev_type == HCI_PRIMARY) { - ret = hci_powered_update_sync(hdev); - } - } else { - /* Init failed, cleanup */ - flush_work(&hdev->tx_work); - - /* Since hci_rx_work() is possible to awake new cmd_work - * it should be flushed first to avoid unexpected call of - * hci_cmd_work() - */ - flush_work(&hdev->rx_work); - flush_work(&hdev->cmd_work); - - skb_queue_purge(&hdev->cmd_q); - skb_queue_purge(&hdev->rx_q); - - if (hdev->flush) - hdev->flush(hdev); - - if (hdev->sent_cmd) { - kfree_skb(hdev->sent_cmd); - hdev->sent_cmd = NULL; - } - - clear_bit(HCI_RUNNING, &hdev->flags); - hci_sock_dev_event(hdev, HCI_DEV_CLOSE); - - hdev->close(hdev); - hdev->flags &= BIT(HCI_RAW); - } - -done: - return ret; -} - static int hci_dev_do_open(struct hci_dev *hdev) { int ret = 0; @@ -1594,155 +542,6 @@ done: return err; } -/* This function requires the caller holds hdev->lock */ -static void hci_pend_le_actions_clear(struct hci_dev *hdev) -{ - struct hci_conn_params *p; - - list_for_each_entry(p, &hdev->le_conn_params, list) { - if (p->conn) { - hci_conn_drop(p->conn); - hci_conn_put(p->conn); - p->conn = NULL; - } - list_del_init(&p->action); - } - - BT_DBG("All LE pending actions cleared"); -} - -/* TODO: Move this function into hci_sync.c */ -int hci_dev_close_sync(struct hci_dev *hdev) -{ - bool auto_off; - int err = 0; - - BT_DBG("%s %p", hdev->name, hdev); - - cancel_delayed_work(&hdev->power_off); - cancel_delayed_work(&hdev->ncmd_timer); - - hci_request_cancel_all(hdev); - - if (!hci_dev_test_flag(hdev, HCI_UNREGISTER) && - !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && - test_bit(HCI_UP, &hdev->flags)) { - /* Execute vendor specific shutdown routine */ - if (hdev->shutdown) - err = hdev->shutdown(hdev); - } - - if (!test_and_clear_bit(HCI_UP, &hdev->flags)) { - cancel_delayed_work_sync(&hdev->cmd_timer); - return err; - } - - hci_leds_update_powered(hdev, false); - - /* Flush RX and TX works */ - flush_work(&hdev->tx_work); - flush_work(&hdev->rx_work); - - if (hdev->discov_timeout > 0) { - hdev->discov_timeout = 0; - hci_dev_clear_flag(hdev, HCI_DISCOVERABLE); - hci_dev_clear_flag(hdev, HCI_LIMITED_DISCOVERABLE); - } - - if (hci_dev_test_and_clear_flag(hdev, HCI_SERVICE_CACHE)) - cancel_delayed_work(&hdev->service_cache); - - if (hci_dev_test_flag(hdev, HCI_MGMT)) { - struct adv_info *adv_instance; - - cancel_delayed_work_sync(&hdev->rpa_expired); - - list_for_each_entry(adv_instance, &hdev->adv_instances, list) - cancel_delayed_work_sync(&adv_instance->rpa_expired_cb); - } - - /* Avoid potential lockdep warnings from the *_flush() calls by - * ensuring the workqueue is empty up front. - */ - drain_workqueue(hdev->workqueue); - - hci_dev_lock(hdev); - - hci_discovery_set_state(hdev, DISCOVERY_STOPPED); - - auto_off = hci_dev_test_and_clear_flag(hdev, HCI_AUTO_OFF); - - if (!auto_off && hdev->dev_type == HCI_PRIMARY && - !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && - hci_dev_test_flag(hdev, HCI_MGMT)) - __mgmt_power_off(hdev); - - hci_inquiry_cache_flush(hdev); - hci_pend_le_actions_clear(hdev); - hci_conn_hash_flush(hdev); - hci_dev_unlock(hdev); - - smp_unregister(hdev); - - hci_sock_dev_event(hdev, HCI_DEV_DOWN); - - aosp_do_close(hdev); - msft_do_close(hdev); - - if (hdev->flush) - hdev->flush(hdev); - - /* Reset device */ - skb_queue_purge(&hdev->cmd_q); - atomic_set(&hdev->cmd_cnt, 1); - if (test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks) && - !auto_off && !hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) { - set_bit(HCI_INIT, &hdev->flags); - __hci_req_sync(hdev, hci_reset_req, 0, HCI_CMD_TIMEOUT, NULL); - clear_bit(HCI_INIT, &hdev->flags); - } - - /* flush cmd work */ - flush_work(&hdev->cmd_work); - - /* Drop queues */ - skb_queue_purge(&hdev->rx_q); - skb_queue_purge(&hdev->cmd_q); - skb_queue_purge(&hdev->raw_q); - - /* Drop last sent command */ - if (hdev->sent_cmd) { - cancel_delayed_work_sync(&hdev->cmd_timer); - kfree_skb(hdev->sent_cmd); - hdev->sent_cmd = NULL; - } - - clear_bit(HCI_RUNNING, &hdev->flags); - hci_sock_dev_event(hdev, HCI_DEV_CLOSE); - - if (test_and_clear_bit(SUSPEND_POWERING_DOWN, hdev->suspend_tasks)) - wake_up(&hdev->suspend_wait_q); - - /* After this point our queues are empty - * and no tasks are scheduled. */ - hdev->close(hdev); - - /* Clear flags */ - hdev->flags &= BIT(HCI_RAW); - hci_dev_clear_volatile_flags(hdev); - - /* Controller radio is available but is currently powered down */ - hdev->amp_status = AMP_STATUS_POWERED_DOWN; - - memset(hdev->eir, 0, sizeof(hdev->eir)); - memset(hdev->dev_class, 0, sizeof(hdev->dev_class)); - bacpy(&hdev->random_addr, BDADDR_ANY); - hci_codec_list_clear(&hdev->local_codecs); - - hci_dev_put(hdev); - return err; -} - int hci_dev_do_close(struct hci_dev *hdev) { int err; @@ -1810,7 +609,7 @@ static int hci_dev_do_reset(struct hci_dev *hdev) atomic_set(&hdev->cmd_cnt, 1); hdev->acl_cnt = 0; hdev->sco_cnt = 0; hdev->le_cnt = 0; - ret = __hci_req_sync(hdev, hci_reset_req, 0, HCI_INIT_TIMEOUT, NULL); + ret = hci_reset_sync(hdev); hci_req_sync_unlock(hdev); return ret; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 8b63b4e5c96a..4e0a7719b58d 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -5,13 +5,19 @@ * Copyright (C) 2021 Intel Corporation */ +#include + #include #include #include #include "hci_request.h" +#include "hci_debugfs.h" #include "smp.h" #include "eir.h" +#include "msft.h" +#include "aosp.h" +#include "leds.h" static void hci_cmd_sync_complete(struct hci_dev *hdev, u8 result, u16 opcode, struct sk_buff *skb) @@ -144,7 +150,7 @@ struct sk_buff *__hci_cmd_sync_sk(struct hci_dev *hdev, u16 opcode, u32 plen, struct sk_buff *skb; int err = 0; - bt_dev_dbg(hdev, ""); + bt_dev_dbg(hdev, "Opcode 0x%4x", opcode); hci_req_init(&req, hdev); @@ -2411,6 +2417,1643 @@ int hci_powered_update_sync(struct hci_dev *hdev) return 0; } +/** + * hci_dev_get_bd_addr_from_property - Get the Bluetooth Device Address + * (BD_ADDR) for a HCI device from + * a firmware node property. + * @hdev: The HCI device + * + * Search the firmware node for 'local-bd-address'. + * + * All-zero BD addresses are rejected, because those could be properties + * that exist in the firmware tables, but were not updated by the firmware. For + * example, the DTS could define 'local-bd-address', with zero BD addresses. + */ +static void hci_dev_get_bd_addr_from_property(struct hci_dev *hdev) +{ + struct fwnode_handle *fwnode = dev_fwnode(hdev->dev.parent); + bdaddr_t ba; + int ret; + + ret = fwnode_property_read_u8_array(fwnode, "local-bd-address", + (u8 *)&ba, sizeof(ba)); + if (ret < 0 || !bacmp(&ba, BDADDR_ANY)) + return; + + bacpy(&hdev->public_addr, &ba); +} + +struct hci_init_stage { + int (*func)(struct hci_dev *hdev); +}; + +/* Run init stage NULL terminated function table */ +static int hci_init_stage_sync(struct hci_dev *hdev, + const struct hci_init_stage *stage) +{ + size_t i; + + for (i = 0; stage[i].func; i++) { + int err; + + err = stage[i].func(hdev); + if (err) + return err; + } + + return 0; +} + +/* Read Local Version */ +static int hci_read_local_version_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_VERSION, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read BD Address */ +static int hci_read_bd_addr_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_BD_ADDR, + 0, NULL, HCI_CMD_TIMEOUT); +} + +#define HCI_INIT(_func) \ +{ \ + .func = _func, \ +} + +static const struct hci_init_stage hci_init0[] = { + /* HCI_OP_READ_LOCAL_VERSION */ + HCI_INIT(hci_read_local_version_sync), + /* HCI_OP_READ_BD_ADDR */ + HCI_INIT(hci_read_bd_addr_sync), + {} +}; + +int hci_reset_sync(struct hci_dev *hdev) +{ + int err; + + set_bit(HCI_RESET, &hdev->flags); + + err = __hci_cmd_sync_status(hdev, HCI_OP_RESET, 0, NULL, + HCI_CMD_TIMEOUT); + if (err) + return err; + + return 0; +} + +static int hci_init0_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + /* Reset */ + if (!test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks)) { + err = hci_reset_sync(hdev); + if (err) + return err; + } + + return hci_init_stage_sync(hdev, hci_init0); +} + +static int hci_unconf_init_sync(struct hci_dev *hdev) +{ + int err; + + if (test_bit(HCI_QUIRK_RAW_DEVICE, &hdev->quirks)) + return 0; + + err = hci_init0_sync(hdev); + if (err < 0) + return err; + + if (hci_dev_test_flag(hdev, HCI_SETUP)) + hci_debugfs_create_basic(hdev); + + return 0; +} + +/* Read Local Supported Features. */ +static int hci_read_local_features_sync(struct hci_dev *hdev) +{ + /* Not all AMP controllers support this command */ + if (hdev->dev_type == HCI_AMP && !(hdev->commands[14] & 0x20)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_FEATURES, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* BR Controller init stage 1 command sequence */ +static const struct hci_init_stage br_init1[] = { + /* HCI_OP_READ_LOCAL_FEATURES */ + HCI_INIT(hci_read_local_features_sync), + /* HCI_OP_READ_LOCAL_VERSION */ + HCI_INIT(hci_read_local_version_sync), + /* HCI_OP_READ_BD_ADDR */ + HCI_INIT(hci_read_bd_addr_sync), + {} +}; + +/* Read Local Commands */ +static int hci_read_local_cmds_sync(struct hci_dev *hdev) +{ + /* All Bluetooth 1.2 and later controllers should support the + * HCI command for reading the local supported commands. + * + * Unfortunately some controllers indicate Bluetooth 1.2 support, + * but do not have support for this command. If that is the case, + * the driver can quirk the behavior and skip reading the local + * supported commands. + */ + if (hdev->hci_ver > BLUETOOTH_VER_1_1 && + !test_bit(HCI_QUIRK_BROKEN_LOCAL_COMMANDS, &hdev->quirks)) + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_COMMANDS, + 0, NULL, HCI_CMD_TIMEOUT); + + return 0; +} + +/* Read Local AMP Info */ +static int hci_read_local_amp_info_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_AMP_INFO, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Data Blk size */ +static int hci_read_data_block_size_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_DATA_BLOCK_SIZE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Flow Control Mode */ +static int hci_read_flow_control_mode_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_FLOW_CONTROL_MODE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Location Data */ +static int hci_read_location_data_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCATION_DATA, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* AMP Controller init stage 1 command sequence */ +static const struct hci_init_stage amp_init1[] = { + /* HCI_OP_READ_LOCAL_VERSION */ + HCI_INIT(hci_read_local_version_sync), + /* HCI_OP_READ_LOCAL_COMMANDS */ + HCI_INIT(hci_read_local_cmds_sync), + /* HCI_OP_READ_LOCAL_AMP_INFO */ + HCI_INIT(hci_read_local_amp_info_sync), + /* HCI_OP_READ_DATA_BLOCK_SIZE */ + HCI_INIT(hci_read_data_block_size_sync), + /* HCI_OP_READ_FLOW_CONTROL_MODE */ + HCI_INIT(hci_read_flow_control_mode_sync), + /* HCI_OP_READ_LOCATION_DATA */ + HCI_INIT(hci_read_location_data_sync), +}; + +static int hci_init1_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + /* Reset */ + if (!test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks)) { + err = hci_reset_sync(hdev); + if (err) + return err; + } + + switch (hdev->dev_type) { + case HCI_PRIMARY: + hdev->flow_ctl_mode = HCI_FLOW_CTL_MODE_PACKET_BASED; + return hci_init_stage_sync(hdev, br_init1); + case HCI_AMP: + hdev->flow_ctl_mode = HCI_FLOW_CTL_MODE_BLOCK_BASED; + return hci_init_stage_sync(hdev, amp_init1); + default: + bt_dev_err(hdev, "Unknown device type %d", hdev->dev_type); + break; + } + + return 0; +} + +/* AMP Controller init stage 2 command sequence */ +static const struct hci_init_stage amp_init2[] = { + /* HCI_OP_READ_LOCAL_FEATURES */ + HCI_INIT(hci_read_local_features_sync), +}; + +/* Read Buffer Size (ACL mtu, max pkt, etc.) */ +static int hci_read_buffer_size_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_BUFFER_SIZE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Class of Device */ +static int hci_read_dev_class_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_CLASS_OF_DEV, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Local Name */ +static int hci_read_local_name_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_NAME, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Voice Setting */ +static int hci_read_voice_setting_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_VOICE_SETTING, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Number of Supported IAC */ +static int hci_read_num_supported_iac_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_NUM_SUPPORTED_IAC, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read Current IAC LAP */ +static int hci_read_current_iac_lap_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_READ_CURRENT_IAC_LAP, + 0, NULL, HCI_CMD_TIMEOUT); +} + +static int hci_set_event_filter_sync(struct hci_dev *hdev, u8 flt_type, + u8 cond_type, bdaddr_t *bdaddr, + u8 auto_accept) +{ + struct hci_cp_set_event_filter cp; + + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + return 0; + + memset(&cp, 0, sizeof(cp)); + cp.flt_type = flt_type; + + if (flt_type != HCI_FLT_CLEAR_ALL) { + cp.cond_type = cond_type; + bacpy(&cp.addr_conn_flt.bdaddr, bdaddr); + cp.addr_conn_flt.auto_accept = auto_accept; + } + + return __hci_cmd_sync_status(hdev, HCI_OP_SET_EVENT_FLT, + flt_type == HCI_FLT_CLEAR_ALL ? + sizeof(cp.flt_type) : sizeof(cp), &cp, + HCI_CMD_TIMEOUT); +} + +static int hci_clear_event_filter_sync(struct hci_dev *hdev) +{ + if (!hci_dev_test_flag(hdev, HCI_EVENT_FILTER_CONFIGURED)) + return 0; + + return hci_set_event_filter_sync(hdev, HCI_FLT_CLEAR_ALL, 0x00, + BDADDR_ANY, 0x00); +} + +/* Connection accept timeout ~20 secs */ +static int hci_write_ca_timeout_sync(struct hci_dev *hdev) +{ + __le16 param = cpu_to_le16(0x7d00); + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_CA_TIMEOUT, + sizeof(param), ¶m, HCI_CMD_TIMEOUT); +} + +/* BR Controller init stage 2 command sequence */ +static const struct hci_init_stage br_init2[] = { + /* HCI_OP_READ_BUFFER_SIZE */ + HCI_INIT(hci_read_buffer_size_sync), + /* HCI_OP_READ_CLASS_OF_DEV */ + HCI_INIT(hci_read_dev_class_sync), + /* HCI_OP_READ_LOCAL_NAME */ + HCI_INIT(hci_read_local_name_sync), + /* HCI_OP_READ_VOICE_SETTING */ + HCI_INIT(hci_read_voice_setting_sync), + /* HCI_OP_READ_NUM_SUPPORTED_IAC */ + HCI_INIT(hci_read_num_supported_iac_sync), + /* HCI_OP_READ_CURRENT_IAC_LAP */ + HCI_INIT(hci_read_current_iac_lap_sync), + /* HCI_OP_SET_EVENT_FLT */ + HCI_INIT(hci_clear_event_filter_sync), + /* HCI_OP_WRITE_CA_TIMEOUT */ + HCI_INIT(hci_write_ca_timeout_sync), + {} +}; + +static int hci_write_ssp_mode_1_sync(struct hci_dev *hdev) +{ + u8 mode = 0x01; + + if (!lmp_ssp_capable(hdev) || !hci_dev_test_flag(hdev, HCI_SSP_ENABLED)) + return 0; + + /* When SSP is available, then the host features page + * should also be available as well. However some + * controllers list the max_page as 0 as long as SSP + * has not been enabled. To achieve proper debugging + * output, force the minimum max_page to 1 at least. + */ + hdev->max_page = 0x01; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SSP_MODE, + sizeof(mode), &mode, HCI_CMD_TIMEOUT); +} + +static int hci_write_eir_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_eir cp; + + if (!lmp_ssp_capable(hdev) || hci_dev_test_flag(hdev, HCI_SSP_ENABLED)) + return 0; + + memset(hdev->eir, 0, sizeof(hdev->eir)); + memset(&cp, 0, sizeof(cp)); + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_EIR, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); +} + +static int hci_write_inquiry_mode_sync(struct hci_dev *hdev) +{ + u8 mode; + + if (!lmp_inq_rssi_capable(hdev) && + !test_bit(HCI_QUIRK_FIXUP_INQUIRY_MODE, &hdev->quirks)) + return 0; + + /* If Extended Inquiry Result events are supported, then + * they are clearly preferred over Inquiry Result with RSSI + * events. + */ + mode = lmp_ext_inq_capable(hdev) ? 0x02 : 0x01; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_INQUIRY_MODE, + sizeof(mode), &mode, HCI_CMD_TIMEOUT); +} + +static int hci_read_inq_rsp_tx_power_sync(struct hci_dev *hdev) +{ + if (!lmp_inq_tx_pwr_capable(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_INQ_RSP_TX_POWER, + 0, NULL, HCI_CMD_TIMEOUT); +} + +static int hci_read_local_ext_features_sync(struct hci_dev *hdev, u8 page) +{ + struct hci_cp_read_local_ext_features cp; + + if (!lmp_ext_feat_capable(hdev)) + return 0; + + memset(&cp, 0, sizeof(cp)); + cp.page = page; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_EXT_FEATURES, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_read_local_ext_features_1_sync(struct hci_dev *hdev) +{ + return hci_read_local_ext_features_sync(hdev, 0x01); +} + +/* HCI Controller init stage 2 command sequence */ +static const struct hci_init_stage hci_init2[] = { + /* HCI_OP_READ_LOCAL_COMMANDS */ + HCI_INIT(hci_read_local_cmds_sync), + /* HCI_OP_WRITE_SSP_MODE */ + HCI_INIT(hci_write_ssp_mode_1_sync), + /* HCI_OP_WRITE_EIR */ + HCI_INIT(hci_write_eir_sync), + /* HCI_OP_WRITE_INQUIRY_MODE */ + HCI_INIT(hci_write_inquiry_mode_sync), + /* HCI_OP_READ_INQ_RSP_TX_POWER */ + HCI_INIT(hci_read_inq_rsp_tx_power_sync), + /* HCI_OP_READ_LOCAL_EXT_FEATURES */ + HCI_INIT(hci_read_local_ext_features_1_sync), + /* HCI_OP_WRITE_AUTH_ENABLE */ + HCI_INIT(hci_write_auth_enable_sync), + {} +}; + +/* Read LE Buffer Size */ +static int hci_le_read_buffer_size_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_BUFFER_SIZE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read LE Local Supported Features */ +static int hci_le_read_local_features_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_LOCAL_FEATURES, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read LE Supported States */ +static int hci_le_read_supported_states_sync(struct hci_dev *hdev) +{ + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_SUPPORTED_STATES, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* LE Controller init stage 2 command sequence */ +static const struct hci_init_stage le_init2[] = { + /* HCI_OP_LE_READ_BUFFER_SIZE */ + HCI_INIT(hci_le_read_buffer_size_sync), + /* HCI_OP_LE_READ_LOCAL_FEATURES */ + HCI_INIT(hci_le_read_local_features_sync), + /* HCI_OP_LE_READ_SUPPORTED_STATES */ + HCI_INIT(hci_le_read_supported_states_sync), + {} +}; + +static int hci_init2_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + if (hdev->dev_type == HCI_AMP) + return hci_init_stage_sync(hdev, amp_init2); + + if (lmp_bredr_capable(hdev)) { + err = hci_init_stage_sync(hdev, br_init2); + if (err) + return err; + } else { + hci_dev_clear_flag(hdev, HCI_BREDR_ENABLED); + } + + if (lmp_le_capable(hdev)) { + err = hci_init_stage_sync(hdev, le_init2); + if (err) + return err; + /* LE-only controllers have LE implicitly enabled */ + if (!lmp_bredr_capable(hdev)) + hci_dev_set_flag(hdev, HCI_LE_ENABLED); + } + + return hci_init_stage_sync(hdev, hci_init2); +} + +static int hci_set_event_mask_sync(struct hci_dev *hdev) +{ + /* The second byte is 0xff instead of 0x9f (two reserved bits + * disabled) since a Broadcom 1.2 dongle doesn't respond to the + * command otherwise. + */ + u8 events[8] = { 0xff, 0xff, 0xfb, 0xff, 0x00, 0x00, 0x00, 0x00 }; + + /* CSR 1.1 dongles does not accept any bitfield so don't try to set + * any event mask for pre 1.2 devices. + */ + if (hdev->hci_ver < BLUETOOTH_VER_1_2) + return 0; + + if (lmp_bredr_capable(hdev)) { + events[4] |= 0x01; /* Flow Specification Complete */ + } else { + /* Use a different default for LE-only devices */ + memset(events, 0, sizeof(events)); + events[1] |= 0x20; /* Command Complete */ + events[1] |= 0x40; /* Command Status */ + events[1] |= 0x80; /* Hardware Error */ + + /* If the controller supports the Disconnect command, enable + * the corresponding event. In addition enable packet flow + * control related events. + */ + if (hdev->commands[0] & 0x20) { + events[0] |= 0x10; /* Disconnection Complete */ + events[2] |= 0x04; /* Number of Completed Packets */ + events[3] |= 0x02; /* Data Buffer Overflow */ + } + + /* If the controller supports the Read Remote Version + * Information command, enable the corresponding event. + */ + if (hdev->commands[2] & 0x80) + events[1] |= 0x08; /* Read Remote Version Information + * Complete + */ + + if (hdev->le_features[0] & HCI_LE_ENCRYPTION) { + events[0] |= 0x80; /* Encryption Change */ + events[5] |= 0x80; /* Encryption Key Refresh Complete */ + } + } + + if (lmp_inq_rssi_capable(hdev) || + test_bit(HCI_QUIRK_FIXUP_INQUIRY_MODE, &hdev->quirks)) + events[4] |= 0x02; /* Inquiry Result with RSSI */ + + if (lmp_ext_feat_capable(hdev)) + events[4] |= 0x04; /* Read Remote Extended Features Complete */ + + if (lmp_esco_capable(hdev)) { + events[5] |= 0x08; /* Synchronous Connection Complete */ + events[5] |= 0x10; /* Synchronous Connection Changed */ + } + + if (lmp_sniffsubr_capable(hdev)) + events[5] |= 0x20; /* Sniff Subrating */ + + if (lmp_pause_enc_capable(hdev)) + events[5] |= 0x80; /* Encryption Key Refresh Complete */ + + if (lmp_ext_inq_capable(hdev)) + events[5] |= 0x40; /* Extended Inquiry Result */ + + if (lmp_no_flush_capable(hdev)) + events[7] |= 0x01; /* Enhanced Flush Complete */ + + if (lmp_lsto_capable(hdev)) + events[6] |= 0x80; /* Link Supervision Timeout Changed */ + + if (lmp_ssp_capable(hdev)) { + events[6] |= 0x01; /* IO Capability Request */ + events[6] |= 0x02; /* IO Capability Response */ + events[6] |= 0x04; /* User Confirmation Request */ + events[6] |= 0x08; /* User Passkey Request */ + events[6] |= 0x10; /* Remote OOB Data Request */ + events[6] |= 0x20; /* Simple Pairing Complete */ + events[7] |= 0x04; /* User Passkey Notification */ + events[7] |= 0x08; /* Keypress Notification */ + events[7] |= 0x10; /* Remote Host Supported + * Features Notification + */ + } + + if (lmp_le_capable(hdev)) + events[7] |= 0x20; /* LE Meta-Event */ + + return __hci_cmd_sync_status(hdev, HCI_OP_SET_EVENT_MASK, + sizeof(events), events, HCI_CMD_TIMEOUT); +} + +static int hci_read_stored_link_key_sync(struct hci_dev *hdev) +{ + struct hci_cp_read_stored_link_key cp; + + if (!(hdev->commands[6] & 0x20) || + test_bit(HCI_QUIRK_BROKEN_STORED_LINK_KEY, &hdev->quirks)) + return 0; + + memset(&cp, 0, sizeof(cp)); + bacpy(&cp.bdaddr, BDADDR_ANY); + cp.read_all = 0x01; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_STORED_LINK_KEY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_setup_link_policy_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_def_link_policy cp; + u16 link_policy = 0; + + if (!(hdev->commands[5] & 0x10)) + return 0; + + memset(&cp, 0, sizeof(cp)); + + if (lmp_rswitch_capable(hdev)) + link_policy |= HCI_LP_RSWITCH; + if (lmp_hold_capable(hdev)) + link_policy |= HCI_LP_HOLD; + if (lmp_sniff_capable(hdev)) + link_policy |= HCI_LP_SNIFF; + if (lmp_park_capable(hdev)) + link_policy |= HCI_LP_PARK; + + cp.policy = cpu_to_le16(link_policy); + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_DEF_LINK_POLICY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_read_page_scan_activity_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[8] & 0x01)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_PAGE_SCAN_ACTIVITY, + 0, NULL, HCI_CMD_TIMEOUT); +} + +static int hci_read_def_err_data_reporting_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[18] & 0x04) || + test_bit(HCI_QUIRK_BROKEN_ERR_DATA_REPORTING, &hdev->quirks)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_DEF_ERR_DATA_REPORTING, + 0, NULL, HCI_CMD_TIMEOUT); +} + +static int hci_read_page_scan_type_sync(struct hci_dev *hdev) +{ + /* Some older Broadcom based Bluetooth 1.2 controllers do not + * support the Read Page Scan Type command. Check support for + * this command in the bit mask of supported commands. + */ + if (!(hdev->commands[13] & 0x01)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_PAGE_SCAN_TYPE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read features beyond page 1 if available */ +static int hci_read_local_ext_features_all_sync(struct hci_dev *hdev) +{ + u8 page; + int err; + + if (!lmp_ext_feat_capable(hdev)) + return 0; + + for (page = 2; page < HCI_MAX_PAGES && page <= hdev->max_page; + page++) { + err = hci_read_local_ext_features_sync(hdev, page); + if (err) + return err; + } + + return 0; +} + +/* HCI Controller init stage 3 command sequence */ +static const struct hci_init_stage hci_init3[] = { + /* HCI_OP_SET_EVENT_MASK */ + HCI_INIT(hci_set_event_mask_sync), + /* HCI_OP_READ_STORED_LINK_KEY */ + HCI_INIT(hci_read_stored_link_key_sync), + /* HCI_OP_WRITE_DEF_LINK_POLICY */ + HCI_INIT(hci_setup_link_policy_sync), + /* HCI_OP_READ_PAGE_SCAN_ACTIVITY */ + HCI_INIT(hci_read_page_scan_activity_sync), + /* HCI_OP_READ_DEF_ERR_DATA_REPORTING */ + HCI_INIT(hci_read_def_err_data_reporting_sync), + /* HCI_OP_READ_PAGE_SCAN_TYPE */ + HCI_INIT(hci_read_page_scan_type_sync), + /* HCI_OP_READ_LOCAL_EXT_FEATURES */ + HCI_INIT(hci_read_local_ext_features_all_sync), + {} +}; + +static int hci_le_set_event_mask_sync(struct hci_dev *hdev) +{ + u8 events[8]; + + if (!lmp_le_capable(hdev)) + return 0; + + memset(events, 0, sizeof(events)); + + if (hdev->le_features[0] & HCI_LE_ENCRYPTION) + events[0] |= 0x10; /* LE Long Term Key Request */ + + /* If controller supports the Connection Parameters Request + * Link Layer Procedure, enable the corresponding event. + */ + if (hdev->le_features[0] & HCI_LE_CONN_PARAM_REQ_PROC) + /* LE Remote Connection Parameter Request */ + events[0] |= 0x20; + + /* If the controller supports the Data Length Extension + * feature, enable the corresponding event. + */ + if (hdev->le_features[0] & HCI_LE_DATA_LEN_EXT) + events[0] |= 0x40; /* LE Data Length Change */ + + /* If the controller supports LL Privacy feature, enable + * the corresponding event. + */ + if (hdev->le_features[0] & HCI_LE_LL_PRIVACY) + events[1] |= 0x02; /* LE Enhanced Connection Complete */ + + /* If the controller supports Extended Scanner Filter + * Policies, enable the corresponding event. + */ + if (hdev->le_features[0] & HCI_LE_EXT_SCAN_POLICY) + events[1] |= 0x04; /* LE Direct Advertising Report */ + + /* If the controller supports Channel Selection Algorithm #2 + * feature, enable the corresponding event. + */ + if (hdev->le_features[1] & HCI_LE_CHAN_SEL_ALG2) + events[2] |= 0x08; /* LE Channel Selection Algorithm */ + + /* If the controller supports the LE Set Scan Enable command, + * enable the corresponding advertising report event. + */ + if (hdev->commands[26] & 0x08) + events[0] |= 0x02; /* LE Advertising Report */ + + /* If the controller supports the LE Create Connection + * command, enable the corresponding event. + */ + if (hdev->commands[26] & 0x10) + events[0] |= 0x01; /* LE Connection Complete */ + + /* If the controller supports the LE Connection Update + * command, enable the corresponding event. + */ + if (hdev->commands[27] & 0x04) + events[0] |= 0x04; /* LE Connection Update Complete */ + + /* If the controller supports the LE Read Remote Used Features + * command, enable the corresponding event. + */ + if (hdev->commands[27] & 0x20) + /* LE Read Remote Used Features Complete */ + events[0] |= 0x08; + + /* If the controller supports the LE Read Local P-256 + * Public Key command, enable the corresponding event. + */ + if (hdev->commands[34] & 0x02) + /* LE Read Local P-256 Public Key Complete */ + events[0] |= 0x80; + + /* If the controller supports the LE Generate DHKey + * command, enable the corresponding event. + */ + if (hdev->commands[34] & 0x04) + events[1] |= 0x01; /* LE Generate DHKey Complete */ + + /* If the controller supports the LE Set Default PHY or + * LE Set PHY commands, enable the corresponding event. + */ + if (hdev->commands[35] & (0x20 | 0x40)) + events[1] |= 0x08; /* LE PHY Update Complete */ + + /* If the controller supports LE Set Extended Scan Parameters + * and LE Set Extended Scan Enable commands, enable the + * corresponding event. + */ + if (use_ext_scan(hdev)) + events[1] |= 0x10; /* LE Extended Advertising Report */ + + /* If the controller supports the LE Extended Advertising + * command, enable the corresponding event. + */ + if (ext_adv_capable(hdev)) + events[2] |= 0x02; /* LE Advertising Set Terminated */ + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_EVENT_MASK, + sizeof(events), events, HCI_CMD_TIMEOUT); +} + +/* Read LE Advertising Channel TX Power */ +static int hci_le_read_adv_tx_power_sync(struct hci_dev *hdev) +{ + if ((hdev->commands[25] & 0x40) && !ext_adv_capable(hdev)) { + /* HCI TS spec forbids mixing of legacy and extended + * advertising commands wherein READ_ADV_TX_POWER is + * also included. So do not call it if extended adv + * is supported otherwise controller will return + * COMMAND_DISALLOWED for extended commands. + */ + return __hci_cmd_sync_status(hdev, + HCI_OP_LE_READ_ADV_TX_POWER, + 0, NULL, HCI_CMD_TIMEOUT); + } + + return 0; +} + +/* Read LE Min/Max Tx Power*/ +static int hci_le_read_tx_power_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[38] & 0x80)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_TRANSMIT_POWER, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Read LE Accept List Size */ +static int hci_le_read_accept_list_size_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[26] & 0x40)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_ACCEPT_LIST_SIZE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Clear LE Accept List */ +static int hci_le_clear_accept_list_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[26] & 0x80)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_CLEAR_ACCEPT_LIST, 0, NULL, + HCI_CMD_TIMEOUT); +} + +/* Read LE Resolving List Size */ +static int hci_le_read_resolv_list_size_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[34] & 0x40)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_RESOLV_LIST_SIZE, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Clear LE Resolving List */ +static int hci_le_clear_resolv_list_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[34] & 0x20)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_CLEAR_RESOLV_LIST, 0, NULL, + HCI_CMD_TIMEOUT); +} + +/* Set RPA timeout */ +static int hci_le_set_rpa_timeout_sync(struct hci_dev *hdev) +{ + __le16 timeout = cpu_to_le16(hdev->rpa_timeout); + + if (!(hdev->commands[35] & 0x04)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_RPA_TIMEOUT, + sizeof(timeout), &timeout, + HCI_CMD_TIMEOUT); +} + +/* Read LE Maximum Data Length */ +static int hci_le_read_max_data_len_sync(struct hci_dev *hdev) +{ + if (!(hdev->le_features[0] & HCI_LE_DATA_LEN_EXT)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_MAX_DATA_LEN, 0, NULL, + HCI_CMD_TIMEOUT); +} + +/* Read LE Suggested Default Data Length */ +static int hci_le_read_def_data_len_sync(struct hci_dev *hdev) +{ + if (!(hdev->le_features[0] & HCI_LE_DATA_LEN_EXT)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_DEF_DATA_LEN, 0, NULL, + HCI_CMD_TIMEOUT); +} + +/* Read LE Number of Supported Advertising Sets */ +static int hci_le_read_num_support_adv_sets_sync(struct hci_dev *hdev) +{ + if (!ext_adv_capable(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, + HCI_OP_LE_READ_NUM_SUPPORTED_ADV_SETS, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Write LE Host Supported */ +static int hci_set_le_support_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_le_host_supported cp; + + /* LE-only devices do not support explicit enablement */ + if (!lmp_bredr_capable(hdev)) + return 0; + + memset(&cp, 0, sizeof(cp)); + + if (hci_dev_test_flag(hdev, HCI_LE_ENABLED)) { + cp.le = 0x01; + cp.simul = 0x00; + } + + if (cp.le == lmp_host_le_capable(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_LE_HOST_SUPPORTED, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +/* LE Controller init stage 3 command sequence */ +static const struct hci_init_stage le_init3[] = { + /* HCI_OP_LE_SET_EVENT_MASK */ + HCI_INIT(hci_le_set_event_mask_sync), + /* HCI_OP_LE_READ_ADV_TX_POWER */ + HCI_INIT(hci_le_read_adv_tx_power_sync), + /* HCI_OP_LE_READ_TRANSMIT_POWER */ + HCI_INIT(hci_le_read_tx_power_sync), + /* HCI_OP_LE_READ_ACCEPT_LIST_SIZE */ + HCI_INIT(hci_le_read_accept_list_size_sync), + /* HCI_OP_LE_CLEAR_ACCEPT_LIST */ + HCI_INIT(hci_le_clear_accept_list_sync), + /* HCI_OP_LE_READ_RESOLV_LIST_SIZE */ + HCI_INIT(hci_le_read_resolv_list_size_sync), + /* HCI_OP_LE_CLEAR_RESOLV_LIST */ + HCI_INIT(hci_le_clear_resolv_list_sync), + /* HCI_OP_LE_SET_RPA_TIMEOUT */ + HCI_INIT(hci_le_set_rpa_timeout_sync), + /* HCI_OP_LE_READ_MAX_DATA_LEN */ + HCI_INIT(hci_le_read_max_data_len_sync), + /* HCI_OP_LE_READ_DEF_DATA_LEN */ + HCI_INIT(hci_le_read_def_data_len_sync), + /* HCI_OP_LE_READ_NUM_SUPPORTED_ADV_SETS */ + HCI_INIT(hci_le_read_num_support_adv_sets_sync), + /* HCI_OP_WRITE_LE_HOST_SUPPORTED */ + HCI_INIT(hci_set_le_support_sync), + {} +}; + +static int hci_init3_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + err = hci_init_stage_sync(hdev, hci_init3); + if (err) + return err; + + if (lmp_le_capable(hdev)) + return hci_init_stage_sync(hdev, le_init3); + + return 0; +} + +static int hci_delete_stored_link_key_sync(struct hci_dev *hdev) +{ + struct hci_cp_delete_stored_link_key cp; + + /* Some Broadcom based Bluetooth controllers do not support the + * Delete Stored Link Key command. They are clearly indicating its + * absence in the bit mask of supported commands. + * + * Check the supported commands and only if the command is marked + * as supported send it. If not supported assume that the controller + * does not have actual support for stored link keys which makes this + * command redundant anyway. + * + * Some controllers indicate that they support handling deleting + * stored link keys, but they don't. The quirk lets a driver + * just disable this command. + */ + if (!(hdev->commands[6] & 0x80) || + test_bit(HCI_QUIRK_BROKEN_STORED_LINK_KEY, &hdev->quirks)) + return 0; + + memset(&cp, 0, sizeof(cp)); + bacpy(&cp.bdaddr, BDADDR_ANY); + cp.delete_all = 0x01; + + return __hci_cmd_sync_status(hdev, HCI_OP_DELETE_STORED_LINK_KEY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_set_event_mask_page_2_sync(struct hci_dev *hdev) +{ + u8 events[8] = { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; + bool changed = false; + + /* Set event mask page 2 if the HCI command for it is supported */ + if (!(hdev->commands[22] & 0x04)) + return 0; + + /* If Connectionless Peripheral Broadcast central role is supported + * enable all necessary events for it. + */ + if (lmp_cpb_central_capable(hdev)) { + events[1] |= 0x40; /* Triggered Clock Capture */ + events[1] |= 0x80; /* Synchronization Train Complete */ + events[2] |= 0x10; /* Peripheral Page Response Timeout */ + events[2] |= 0x20; /* CPB Channel Map Change */ + changed = true; + } + + /* If Connectionless Peripheral Broadcast peripheral role is supported + * enable all necessary events for it. + */ + if (lmp_cpb_peripheral_capable(hdev)) { + events[2] |= 0x01; /* Synchronization Train Received */ + events[2] |= 0x02; /* CPB Receive */ + events[2] |= 0x04; /* CPB Timeout */ + events[2] |= 0x08; /* Truncated Page Complete */ + changed = true; + } + + /* Enable Authenticated Payload Timeout Expired event if supported */ + if (lmp_ping_capable(hdev) || hdev->le_features[0] & HCI_LE_PING) { + events[2] |= 0x80; + changed = true; + } + + /* Some Broadcom based controllers indicate support for Set Event + * Mask Page 2 command, but then actually do not support it. Since + * the default value is all bits set to zero, the command is only + * required if the event mask has to be changed. In case no change + * to the event mask is needed, skip this command. + */ + if (!changed) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_SET_EVENT_MASK_PAGE_2, + sizeof(events), events, HCI_CMD_TIMEOUT); +} + +/* Read local codec list if the HCI command is supported */ +static int hci_read_local_codecs_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[29] & 0x20)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_CODECS, 0, NULL, + HCI_CMD_TIMEOUT); +} + +/* Read local pairing options if the HCI command is supported */ +static int hci_read_local_pairing_opts_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[41] & 0x08)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_LOCAL_PAIRING_OPTS, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Get MWS transport configuration if the HCI command is supported */ +static int hci_get_mws_transport_config_sync(struct hci_dev *hdev) +{ + if (!(hdev->commands[30] & 0x08)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_GET_MWS_TRANSPORT_CONFIG, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Check for Synchronization Train support */ +static int hci_read_sync_train_params_sync(struct hci_dev *hdev) +{ + if (!lmp_sync_train_capable(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_READ_SYNC_TRAIN_PARAMS, + 0, NULL, HCI_CMD_TIMEOUT); +} + +/* Enable Secure Connections if supported and configured */ +static int hci_write_sc_support_1_sync(struct hci_dev *hdev) +{ + u8 support = 0x01; + + if (!hci_dev_test_flag(hdev, HCI_SSP_ENABLED) || + !bredr_sc_enabled(hdev)) + return 0; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_SC_SUPPORT, + sizeof(support), &support, + HCI_CMD_TIMEOUT); +} + +/* Set erroneous data reporting if supported to the wideband speech + * setting value + */ +static int hci_set_err_data_report_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_def_err_data_reporting cp; + bool enabled = hci_dev_test_flag(hdev, HCI_WIDEBAND_SPEECH_ENABLED); + + if (!(hdev->commands[18] & 0x08) || + test_bit(HCI_QUIRK_BROKEN_ERR_DATA_REPORTING, &hdev->quirks)) + return 0; + + if (enabled == hdev->err_data_reporting) + return 0; + + memset(&cp, 0, sizeof(cp)); + cp.err_data_reporting = enabled ? ERR_DATA_REPORTING_ENABLED : + ERR_DATA_REPORTING_DISABLED; + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_DEF_ERR_DATA_REPORTING, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static const struct hci_init_stage hci_init4[] = { + /* HCI_OP_DELETE_STORED_LINK_KEY */ + HCI_INIT(hci_delete_stored_link_key_sync), + /* HCI_OP_SET_EVENT_MASK_PAGE_2 */ + HCI_INIT(hci_set_event_mask_page_2_sync), + /* HCI_OP_READ_LOCAL_CODECS */ + HCI_INIT(hci_read_local_codecs_sync), + /* HCI_OP_READ_LOCAL_PAIRING_OPTS */ + HCI_INIT(hci_read_local_pairing_opts_sync), + /* HCI_OP_GET_MWS_TRANSPORT_CONFIG */ + HCI_INIT(hci_get_mws_transport_config_sync), + /* HCI_OP_READ_SYNC_TRAIN_PARAMS */ + HCI_INIT(hci_read_sync_train_params_sync), + /* HCI_OP_WRITE_SC_SUPPORT */ + HCI_INIT(hci_write_sc_support_1_sync), + /* HCI_OP_WRITE_DEF_ERR_DATA_REPORTING */ + HCI_INIT(hci_set_err_data_report_sync), + {} +}; + +/* Set Suggested Default Data Length to maximum if supported */ +static int hci_le_set_write_def_data_len_sync(struct hci_dev *hdev) +{ + struct hci_cp_le_write_def_data_len cp; + + if (!(hdev->le_features[0] & HCI_LE_DATA_LEN_EXT)) + return 0; + + memset(&cp, 0, sizeof(cp)); + cp.tx_len = cpu_to_le16(hdev->le_max_tx_len); + cp.tx_time = cpu_to_le16(hdev->le_max_tx_time); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_WRITE_DEF_DATA_LEN, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +/* Set Default PHY parameters if command is supported */ +static int hci_le_set_default_phy_sync(struct hci_dev *hdev) +{ + struct hci_cp_le_set_default_phy cp; + + if (!(hdev->commands[35] & 0x20)) + return 0; + + memset(&cp, 0, sizeof(cp)); + cp.all_phys = 0x00; + cp.tx_phys = hdev->le_tx_def_phys; + cp.rx_phys = hdev->le_rx_def_phys; + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_SET_DEFAULT_PHY, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static const struct hci_init_stage le_init4[] = { + /* HCI_OP_LE_WRITE_DEF_DATA_LEN */ + HCI_INIT(hci_le_set_write_def_data_len_sync), + /* HCI_OP_LE_SET_DEFAULT_PHY */ + HCI_INIT(hci_le_set_default_phy_sync), + {} +}; + +static int hci_init4_sync(struct hci_dev *hdev) +{ + int err; + + bt_dev_dbg(hdev, ""); + + err = hci_init_stage_sync(hdev, hci_init4); + if (err) + return err; + + if (lmp_le_capable(hdev)) + return hci_init_stage_sync(hdev, le_init4); + + return 0; +} + +static int hci_init_sync(struct hci_dev *hdev) +{ + int err; + + err = hci_init1_sync(hdev); + if (err < 0) + return err; + + if (hci_dev_test_flag(hdev, HCI_SETUP)) + hci_debugfs_create_basic(hdev); + + err = hci_init2_sync(hdev); + if (err < 0) + return err; + + /* HCI_PRIMARY covers both single-mode LE, BR/EDR and dual-mode + * BR/EDR/LE type controllers. AMP controllers only need the + * first two stages of init. + */ + if (hdev->dev_type != HCI_PRIMARY) + return 0; + + err = hci_init3_sync(hdev); + if (err < 0) + return err; + + err = hci_init4_sync(hdev); + if (err < 0) + return err; + + /* This function is only called when the controller is actually in + * configured state. When the controller is marked as unconfigured, + * this initialization procedure is not run. + * + * It means that it is possible that a controller runs through its + * setup phase and then discovers missing settings. If that is the + * case, then this function will not be called. It then will only + * be called during the config phase. + * + * So only when in setup phase or config phase, create the debugfs + * entries and register the SMP channels. + */ + if (!hci_dev_test_flag(hdev, HCI_SETUP) && + !hci_dev_test_flag(hdev, HCI_CONFIG)) + return 0; + + hci_debugfs_create_common(hdev); + + if (lmp_bredr_capable(hdev)) + hci_debugfs_create_bredr(hdev); + + if (lmp_le_capable(hdev)) + hci_debugfs_create_le(hdev); + + return 0; +} + +int hci_dev_open_sync(struct hci_dev *hdev) +{ + int ret = 0; + + bt_dev_dbg(hdev, ""); + + if (hci_dev_test_flag(hdev, HCI_UNREGISTER)) { + ret = -ENODEV; + goto done; + } + + if (!hci_dev_test_flag(hdev, HCI_SETUP) && + !hci_dev_test_flag(hdev, HCI_CONFIG)) { + /* Check for rfkill but allow the HCI setup stage to + * proceed (which in itself doesn't cause any RF activity). + */ + if (hci_dev_test_flag(hdev, HCI_RFKILLED)) { + ret = -ERFKILL; + goto done; + } + + /* Check for valid public address or a configured static + * random address, but let the HCI setup proceed to + * be able to determine if there is a public address + * or not. + * + * In case of user channel usage, it is not important + * if a public address or static random address is + * available. + * + * This check is only valid for BR/EDR controllers + * since AMP controllers do not have an address. + */ + if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && + hdev->dev_type == HCI_PRIMARY && + !bacmp(&hdev->bdaddr, BDADDR_ANY) && + !bacmp(&hdev->static_addr, BDADDR_ANY)) { + ret = -EADDRNOTAVAIL; + goto done; + } + } + + if (test_bit(HCI_UP, &hdev->flags)) { + ret = -EALREADY; + goto done; + } + + if (hdev->open(hdev)) { + ret = -EIO; + goto done; + } + + set_bit(HCI_RUNNING, &hdev->flags); + hci_sock_dev_event(hdev, HCI_DEV_OPEN); + + atomic_set(&hdev->cmd_cnt, 1); + set_bit(HCI_INIT, &hdev->flags); + + if (hci_dev_test_flag(hdev, HCI_SETUP) || + test_bit(HCI_QUIRK_NON_PERSISTENT_SETUP, &hdev->quirks)) { + bool invalid_bdaddr; + + hci_sock_dev_event(hdev, HCI_DEV_SETUP); + + if (hdev->setup) + ret = hdev->setup(hdev); + + /* The transport driver can set the quirk to mark the + * BD_ADDR invalid before creating the HCI device or in + * its setup callback. + */ + invalid_bdaddr = test_bit(HCI_QUIRK_INVALID_BDADDR, + &hdev->quirks); + + if (ret) + goto setup_failed; + + if (test_bit(HCI_QUIRK_USE_BDADDR_PROPERTY, &hdev->quirks)) { + if (!bacmp(&hdev->public_addr, BDADDR_ANY)) + hci_dev_get_bd_addr_from_property(hdev); + + if (bacmp(&hdev->public_addr, BDADDR_ANY) && + hdev->set_bdaddr) { + ret = hdev->set_bdaddr(hdev, + &hdev->public_addr); + + /* If setting of the BD_ADDR from the device + * property succeeds, then treat the address + * as valid even if the invalid BD_ADDR + * quirk indicates otherwise. + */ + if (!ret) + invalid_bdaddr = false; + } + } + +setup_failed: + /* The transport driver can set these quirks before + * creating the HCI device or in its setup callback. + * + * For the invalid BD_ADDR quirk it is possible that + * it becomes a valid address if the bootloader does + * provide it (see above). + * + * In case any of them is set, the controller has to + * start up as unconfigured. + */ + if (test_bit(HCI_QUIRK_EXTERNAL_CONFIG, &hdev->quirks) || + invalid_bdaddr) + hci_dev_set_flag(hdev, HCI_UNCONFIGURED); + + /* For an unconfigured controller it is required to + * read at least the version information provided by + * the Read Local Version Information command. + * + * If the set_bdaddr driver callback is provided, then + * also the original Bluetooth public device address + * will be read using the Read BD Address command. + */ + if (hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) + ret = hci_unconf_init_sync(hdev); + } + + if (hci_dev_test_flag(hdev, HCI_CONFIG)) { + /* If public address change is configured, ensure that + * the address gets programmed. If the driver does not + * support changing the public address, fail the power + * on procedure. + */ + if (bacmp(&hdev->public_addr, BDADDR_ANY) && + hdev->set_bdaddr) + ret = hdev->set_bdaddr(hdev, &hdev->public_addr); + else + ret = -EADDRNOTAVAIL; + } + + if (!ret) { + if (!hci_dev_test_flag(hdev, HCI_UNCONFIGURED) && + !hci_dev_test_flag(hdev, HCI_USER_CHANNEL)) { + ret = hci_init_sync(hdev); + if (!ret && hdev->post_init) + ret = hdev->post_init(hdev); + } + } + + /* If the HCI Reset command is clearing all diagnostic settings, + * then they need to be reprogrammed after the init procedure + * completed. + */ + if (test_bit(HCI_QUIRK_NON_PERSISTENT_DIAG, &hdev->quirks) && + !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && + hci_dev_test_flag(hdev, HCI_VENDOR_DIAG) && hdev->set_diag) + ret = hdev->set_diag(hdev, true); + + msft_do_open(hdev); + aosp_do_open(hdev); + + clear_bit(HCI_INIT, &hdev->flags); + + if (!ret) { + hci_dev_hold(hdev); + hci_dev_set_flag(hdev, HCI_RPA_EXPIRED); + hci_adv_instances_set_rpa_expired(hdev, true); + set_bit(HCI_UP, &hdev->flags); + hci_sock_dev_event(hdev, HCI_DEV_UP); + hci_leds_update_powered(hdev, true); + if (!hci_dev_test_flag(hdev, HCI_SETUP) && + !hci_dev_test_flag(hdev, HCI_CONFIG) && + !hci_dev_test_flag(hdev, HCI_UNCONFIGURED) && + !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && + hci_dev_test_flag(hdev, HCI_MGMT) && + hdev->dev_type == HCI_PRIMARY) { + ret = hci_powered_update_sync(hdev); + } + } else { + /* Init failed, cleanup */ + flush_work(&hdev->tx_work); + + /* Since hci_rx_work() is possible to awake new cmd_work + * it should be flushed first to avoid unexpected call of + * hci_cmd_work() + */ + flush_work(&hdev->rx_work); + flush_work(&hdev->cmd_work); + + skb_queue_purge(&hdev->cmd_q); + skb_queue_purge(&hdev->rx_q); + + if (hdev->flush) + hdev->flush(hdev); + + if (hdev->sent_cmd) { + kfree_skb(hdev->sent_cmd); + hdev->sent_cmd = NULL; + } + + clear_bit(HCI_RUNNING, &hdev->flags); + hci_sock_dev_event(hdev, HCI_DEV_CLOSE); + + hdev->close(hdev); + hdev->flags &= BIT(HCI_RAW); + } + +done: + return ret; +} + +/* This function requires the caller holds hdev->lock */ +static void hci_pend_le_actions_clear(struct hci_dev *hdev) +{ + struct hci_conn_params *p; + + list_for_each_entry(p, &hdev->le_conn_params, list) { + if (p->conn) { + hci_conn_drop(p->conn); + hci_conn_put(p->conn); + p->conn = NULL; + } + list_del_init(&p->action); + } + + BT_DBG("All LE pending actions cleared"); +} + +int hci_dev_close_sync(struct hci_dev *hdev) +{ + bool auto_off; + int err = 0; + + bt_dev_dbg(hdev, ""); + + cancel_delayed_work(&hdev->power_off); + cancel_delayed_work(&hdev->ncmd_timer); + + hci_request_cancel_all(hdev); + + if (!hci_dev_test_flag(hdev, HCI_UNREGISTER) && + !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && + test_bit(HCI_UP, &hdev->flags)) { + /* Execute vendor specific shutdown routine */ + if (hdev->shutdown) + err = hdev->shutdown(hdev); + } + + if (!test_and_clear_bit(HCI_UP, &hdev->flags)) { + cancel_delayed_work_sync(&hdev->cmd_timer); + return err; + } + + hci_leds_update_powered(hdev, false); + + /* Flush RX and TX works */ + flush_work(&hdev->tx_work); + flush_work(&hdev->rx_work); + + if (hdev->discov_timeout > 0) { + hdev->discov_timeout = 0; + hci_dev_clear_flag(hdev, HCI_DISCOVERABLE); + hci_dev_clear_flag(hdev, HCI_LIMITED_DISCOVERABLE); + } + + if (hci_dev_test_and_clear_flag(hdev, HCI_SERVICE_CACHE)) + cancel_delayed_work(&hdev->service_cache); + + if (hci_dev_test_flag(hdev, HCI_MGMT)) { + struct adv_info *adv_instance; + + cancel_delayed_work_sync(&hdev->rpa_expired); + + list_for_each_entry(adv_instance, &hdev->adv_instances, list) + cancel_delayed_work_sync(&adv_instance->rpa_expired_cb); + } + + /* Avoid potential lockdep warnings from the *_flush() calls by + * ensuring the workqueue is empty up front. + */ + drain_workqueue(hdev->workqueue); + + hci_dev_lock(hdev); + + hci_discovery_set_state(hdev, DISCOVERY_STOPPED); + + auto_off = hci_dev_test_and_clear_flag(hdev, HCI_AUTO_OFF); + + if (!auto_off && hdev->dev_type == HCI_PRIMARY && + !hci_dev_test_flag(hdev, HCI_USER_CHANNEL) && + hci_dev_test_flag(hdev, HCI_MGMT)) + __mgmt_power_off(hdev); + + hci_inquiry_cache_flush(hdev); + hci_pend_le_actions_clear(hdev); + hci_conn_hash_flush(hdev); + hci_dev_unlock(hdev); + + smp_unregister(hdev); + + hci_sock_dev_event(hdev, HCI_DEV_DOWN); + + aosp_do_close(hdev); + msft_do_close(hdev); + + if (hdev->flush) + hdev->flush(hdev); + + /* Reset device */ + skb_queue_purge(&hdev->cmd_q); + atomic_set(&hdev->cmd_cnt, 1); + if (test_bit(HCI_QUIRK_RESET_ON_CLOSE, &hdev->quirks) && + !auto_off && !hci_dev_test_flag(hdev, HCI_UNCONFIGURED)) { + set_bit(HCI_INIT, &hdev->flags); + hci_reset_sync(hdev); + clear_bit(HCI_INIT, &hdev->flags); + } + + /* flush cmd work */ + flush_work(&hdev->cmd_work); + + /* Drop queues */ + skb_queue_purge(&hdev->rx_q); + skb_queue_purge(&hdev->cmd_q); + skb_queue_purge(&hdev->raw_q); + + /* Drop last sent command */ + if (hdev->sent_cmd) { + cancel_delayed_work_sync(&hdev->cmd_timer); + kfree_skb(hdev->sent_cmd); + hdev->sent_cmd = NULL; + } + + clear_bit(HCI_RUNNING, &hdev->flags); + hci_sock_dev_event(hdev, HCI_DEV_CLOSE); + + if (test_and_clear_bit(SUSPEND_POWERING_DOWN, hdev->suspend_tasks)) + wake_up(&hdev->suspend_wait_q); + + /* After this point our queues are empty and no tasks are scheduled. */ + hdev->close(hdev); + + /* Clear flags */ + hdev->flags &= BIT(HCI_RAW); + hci_dev_clear_volatile_flags(hdev); + + /* Controller radio is available but is currently powered down */ + hdev->amp_status = AMP_STATUS_POWERED_DOWN; + + memset(hdev->eir, 0, sizeof(hdev->eir)); + memset(hdev->dev_class, 0, sizeof(hdev->dev_class)); + bacpy(&hdev->random_addr, BDADDR_ANY); + + hci_dev_put(hdev); + return err; +} + /* This function perform power on HCI command sequence as follows: * * If controller is already up (HCI_UP) performs hci_powered_update_sync @@ -2567,8 +4210,17 @@ static int hci_disconnect_sync(struct hci_dev *hdev, struct hci_conn *conn, cp.handle = cpu_to_le16(conn->handle); cp.reason = reason; - return __hci_cmd_sync_status(hdev, HCI_OP_DISCONNECT, - sizeof(cp), &cp, HCI_CMD_TIMEOUT); + /* Wait for HCI_EV_DISCONN_COMPLETE not HCI_EV_CMD_STATUS when not + * suspending. + */ + if (!hdev->suspended) + return __hci_cmd_sync_status_sk(hdev, HCI_OP_DISCONNECT, + sizeof(cp), &cp, + HCI_EV_DISCONN_COMPLETE, + HCI_CMD_TIMEOUT, NULL); + + return __hci_cmd_sync_status(hdev, HCI_OP_DISCONNECT, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); } static int hci_le_connect_cancel_sync(struct hci_dev *hdev, @@ -2881,6 +4533,5 @@ int hci_start_discovery_sync(struct hci_dev *hdev) queue_delayed_work(hdev->req_workqueue, &hdev->le_scan_disable, timeout); - return 0; } From 182ee45da083db4e3e621541ccf255bfa9652214 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Wed, 27 Oct 2021 16:59:00 -0700 Subject: [PATCH 0053/2048] Bluetooth: hci_sync: Rework hci_suspend_notifier This makes hci_suspend_notifier use the hci_*_sync which can be executed synchronously which is allowed in the suspend_notifier and simplifies a lot of the handling since the status of each command can be checked inline so no other work need to be scheduled thus can be performed without using of a state machine. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 4 - include/net/bluetooth/hci_sync.h | 3 + net/bluetooth/hci_conn.c | 10 - net/bluetooth/hci_core.c | 106 +------ net/bluetooth/hci_event.c | 71 +++-- net/bluetooth/hci_request.c | 237 -------------- net/bluetooth/hci_sync.c | 294 +++++++++++++++++- net/bluetooth/mgmt.c | 39 +-- net/bluetooth/msft.c | 511 ++++++++++++++++++------------- net/bluetooth/msft.h | 15 +- 10 files changed, 651 insertions(+), 639 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 3e53c845ab0e..53a8c7d3a4bf 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -523,7 +523,6 @@ struct hci_dev { bool advertising_paused; struct notifier_block suspend_notifier; - struct work_struct suspend_prepare; enum suspended_state suspend_state_next; enum suspended_state suspend_state; bool scanning_paused; @@ -532,9 +531,6 @@ struct hci_dev { bdaddr_t wake_addr; u8 wake_addr_type; - wait_queue_head_t suspend_wait_q; - DECLARE_BITMAP(suspend_tasks, __SUSPEND_NUM_TASKS); - struct hci_conn_hash conn_hash; struct list_head mgmt_pending; diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 487e4981cce1..00b13e8ca800 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -92,3 +92,6 @@ int hci_set_powered_sync(struct hci_dev *hdev, u8 val); int hci_start_discovery_sync(struct hci_dev *hdev); int hci_stop_discovery_sync(struct hci_dev *hdev); + +int hci_suspend_sync(struct hci_dev *hdev); +int hci_resume_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index dbd737b62f98..cd6e1cf7e396 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -900,16 +900,6 @@ void hci_le_conn_failed(struct hci_conn *conn, u8 status) hci_conn_del(conn); - /* The suspend notifier is waiting for all devices to disconnect and an - * LE connect cancel will result in an hci_le_conn_failed. Once the last - * connection is deleted, we should also wake the suspend queue to - * complete suspend operations. - */ - if (list_empty(&hdev->conn_hash.list) && - test_and_clear_bit(SUSPEND_DISCONNECTING, hdev->suspend_tasks)) { - wake_up(&hdev->suspend_wait_q); - } - /* Since we may have temporarily stopped the background scanning in * favor of connection establishment, we should restart it. */ diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index e0c0aa779dbc..fdc0dcf8ee36 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -2374,61 +2374,6 @@ void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr, } } -static void hci_suspend_clear_tasks(struct hci_dev *hdev) -{ - int i; - - for (i = 0; i < __SUSPEND_NUM_TASKS; i++) - clear_bit(i, hdev->suspend_tasks); - - wake_up(&hdev->suspend_wait_q); -} - -static int hci_suspend_wait_event(struct hci_dev *hdev) -{ -#define WAKE_COND \ - (find_first_bit(hdev->suspend_tasks, __SUSPEND_NUM_TASKS) == \ - __SUSPEND_NUM_TASKS) - - int i; - int ret = wait_event_timeout(hdev->suspend_wait_q, - WAKE_COND, SUSPEND_NOTIFIER_TIMEOUT); - - if (ret == 0) { - bt_dev_err(hdev, "Timed out waiting for suspend events"); - for (i = 0; i < __SUSPEND_NUM_TASKS; ++i) { - if (test_bit(i, hdev->suspend_tasks)) - bt_dev_err(hdev, "Suspend timeout bit: %d", i); - clear_bit(i, hdev->suspend_tasks); - } - - ret = -ETIMEDOUT; - } else { - ret = 0; - } - - return ret; -} - -static void hci_prepare_suspend(struct work_struct *work) -{ - struct hci_dev *hdev = - container_of(work, struct hci_dev, suspend_prepare); - - hci_dev_lock(hdev); - hci_req_prepare_suspend(hdev, hdev->suspend_state_next); - hci_dev_unlock(hdev); -} - -static int hci_change_suspend_state(struct hci_dev *hdev, - enum suspended_state next) -{ - hdev->suspend_state_next = next; - set_bit(SUSPEND_PREPARE_NOTIFIER, hdev->suspend_tasks); - queue_work(hdev->req_workqueue, &hdev->suspend_prepare); - return hci_suspend_wait_event(hdev); -} - static void hci_clear_wake_reason(struct hci_dev *hdev) { hci_dev_lock(hdev); @@ -2565,7 +2510,6 @@ struct hci_dev *hci_alloc_dev_priv(int sizeof_priv) INIT_WORK(&hdev->tx_work, hci_tx_work); INIT_WORK(&hdev->power_on, hci_power_on); INIT_WORK(&hdev->error_reset, hci_error_reset); - INIT_WORK(&hdev->suspend_prepare, hci_prepare_suspend); hci_cmd_sync_init(hdev); @@ -2576,7 +2520,6 @@ struct hci_dev *hci_alloc_dev_priv(int sizeof_priv) skb_queue_head_init(&hdev->raw_q); init_waitqueue_head(&hdev->req_wait_q); - init_waitqueue_head(&hdev->suspend_wait_q); INIT_DELAYED_WORK(&hdev->cmd_timer, hci_cmd_timeout); INIT_DELAYED_WORK(&hdev->ncmd_timer, hci_ncmd_timeout); @@ -2729,11 +2672,8 @@ void hci_unregister_dev(struct hci_dev *hdev) hci_cmd_sync_clear(hdev); - if (!test_bit(HCI_QUIRK_NO_SUSPEND_NOTIFIER, &hdev->quirks)) { - hci_suspend_clear_tasks(hdev); + if (!test_bit(HCI_QUIRK_NO_SUSPEND_NOTIFIER, &hdev->quirks)) unregister_pm_notifier(&hdev->suspend_notifier); - cancel_work_sync(&hdev->suspend_prepare); - } msft_unregister(hdev); @@ -2800,7 +2740,6 @@ EXPORT_SYMBOL(hci_release_dev); int hci_suspend_dev(struct hci_dev *hdev) { int ret; - u8 state = BT_RUNNING; bt_dev_dbg(hdev, ""); @@ -2809,40 +2748,17 @@ int hci_suspend_dev(struct hci_dev *hdev) hci_dev_test_flag(hdev, HCI_UNREGISTER)) return 0; - /* If powering down, wait for completion. */ - if (mgmt_powering_down(hdev)) { - set_bit(SUSPEND_POWERING_DOWN, hdev->suspend_tasks); - ret = hci_suspend_wait_event(hdev); - if (ret) - goto done; - } + /* If powering down don't attempt to suspend */ + if (mgmt_powering_down(hdev)) + return 0; - /* Suspend consists of two actions: - * - First, disconnect everything and make the controller not - * connectable (disabling scanning) - * - Second, program event filter/accept list and enable scan - */ - ret = hci_change_suspend_state(hdev, BT_SUSPEND_DISCONNECT); - if (ret) - goto clear; + hci_req_sync_lock(hdev); + ret = hci_suspend_sync(hdev); + hci_req_sync_unlock(hdev); - state = BT_SUSPEND_DISCONNECT; - - /* Only configure accept list if device may wakeup. */ - if (hdev->wakeup && hdev->wakeup(hdev)) { - ret = hci_change_suspend_state(hdev, BT_SUSPEND_CONFIGURE_WAKE); - if (!ret) - state = BT_SUSPEND_CONFIGURE_WAKE; - } - -clear: hci_clear_wake_reason(hdev); - mgmt_suspending(hdev, state); + mgmt_suspending(hdev, hdev->suspend_state); -done: - /* We always allow suspend even if suspend preparation failed and - * attempt to recover in resume. - */ hci_sock_dev_event(hdev, HCI_DEV_SUSPEND); return ret; } @@ -2864,10 +2780,12 @@ int hci_resume_dev(struct hci_dev *hdev) if (mgmt_powering_down(hdev)) return 0; - ret = hci_change_suspend_state(hdev, BT_RUNNING); + hci_req_sync_lock(hdev); + ret = hci_resume_sync(hdev); + hci_req_sync_unlock(hdev); mgmt_resuming(hdev, hdev->wake_reason, &hdev->wake_addr, - hdev->wake_addr_type); + hdev->wake_addr_type); hci_sock_dev_event(hdev, HCI_DEV_RESUME); return ret; diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 0f5761b94f39..d4b75a6cfeee 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -2414,9 +2414,14 @@ static void hci_cs_exit_sniff_mode(struct hci_dev *hdev, __u8 status) static void hci_cs_disconnect(struct hci_dev *hdev, u8 status) { struct hci_cp_disconnect *cp; + struct hci_conn_params *params; struct hci_conn *conn; + bool mgmt_conn; - if (!status) + /* Wait for HCI_EV_DISCONN_COMPLETE if status 0x00 and not suspended + * otherwise cleanup the connection immediately. + */ + if (!status && !hdev->suspended) return; cp = hci_sent_cmd_data(hdev, HCI_OP_DISCONNECT); @@ -2426,7 +2431,10 @@ static void hci_cs_disconnect(struct hci_dev *hdev, u8 status) hci_dev_lock(hdev); conn = hci_conn_hash_lookup_handle(hdev, __le16_to_cpu(cp->handle)); - if (conn) { + if (!conn) + goto unlock; + + if (status) { mgmt_disconnect_failed(hdev, &conn->dst, conn->type, conn->dst_type, status); @@ -2435,14 +2443,48 @@ static void hci_cs_disconnect(struct hci_dev *hdev, u8 status) hci_enable_advertising(hdev); } - /* If the disconnection failed for any reason, the upper layer - * does not retry to disconnect in current implementation. - * Hence, we need to do some basic cleanup here and re-enable - * advertising if necessary. - */ - hci_conn_del(conn); + goto done; } + mgmt_conn = test_and_clear_bit(HCI_CONN_MGMT_CONNECTED, &conn->flags); + + if (conn->type == ACL_LINK) { + if (test_bit(HCI_CONN_FLUSH_KEY, &conn->flags)) + hci_remove_link_key(hdev, &conn->dst); + } + + params = hci_conn_params_lookup(hdev, &conn->dst, conn->dst_type); + if (params) { + switch (params->auto_connect) { + case HCI_AUTO_CONN_LINK_LOSS: + if (cp->reason != HCI_ERROR_CONNECTION_TIMEOUT) + break; + fallthrough; + + case HCI_AUTO_CONN_DIRECT: + case HCI_AUTO_CONN_ALWAYS: + list_del_init(¶ms->action); + list_add(¶ms->action, &hdev->pend_le_conns); + break; + + default: + break; + } + } + + mgmt_device_disconnected(hdev, &conn->dst, conn->type, conn->dst_type, + cp->reason, mgmt_conn); + + hci_disconn_cfm(conn, cp->reason); + +done: + /* If the disconnection failed for any reason, the upper layer + * does not retry to disconnect in current implementation. + * Hence, we need to do some basic cleanup here and re-enable + * advertising if necessary. + */ + hci_conn_del(conn); +unlock: hci_dev_unlock(hdev); } @@ -3047,14 +3089,6 @@ static void hci_disconn_complete_evt(struct hci_dev *hdev, struct sk_buff *skb) hci_conn_del(conn); - /* The suspend notifier is waiting for all devices to disconnect so - * clear the bit from pending tasks and inform the wait queue. - */ - if (list_empty(&hdev->conn_hash.list) && - test_and_clear_bit(SUSPEND_DISCONNECTING, hdev->suspend_tasks)) { - wake_up(&hdev->suspend_wait_q); - } - unlock: hci_dev_unlock(hdev); } @@ -5575,8 +5609,9 @@ static struct hci_conn *check_pending_le_conn(struct hci_dev *hdev, if (adv_type != LE_ADV_IND && adv_type != LE_ADV_DIRECT_IND) return NULL; - /* Ignore if the device is blocked */ - if (hci_bdaddr_list_lookup(&hdev->reject_list, addr, addr_type)) + /* Ignore if the device is blocked or hdev is suspended */ + if (hci_bdaddr_list_lookup(&hdev->reject_list, addr, addr_type) || + hdev->suspended) return NULL; /* Most controller will fail if we try to create new connections diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 46fa9c3bdb3e..8aa6e1840c9a 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -492,9 +492,6 @@ void hci_req_add_le_scan_disable(struct hci_request *req, bool rpa_le_conn) return; } - if (hdev->suspended) - set_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); - if (use_ext_scan(hdev)) { struct hci_cp_le_set_ext_scan_enable cp; @@ -868,8 +865,6 @@ void hci_req_add_le_passive_scan(struct hci_request *req) if (hdev->suspended) { window = hdev->le_scan_window_suspend; interval = hdev->le_scan_int_suspend; - - set_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks); } else if (hci_is_le_conn_scanning(hdev)) { window = hdev->le_scan_window_connect; interval = hdev->le_scan_int_connect; @@ -902,59 +897,6 @@ void hci_req_add_le_passive_scan(struct hci_request *req) addr_resolv); } -static void hci_req_clear_event_filter(struct hci_request *req) -{ - struct hci_cp_set_event_filter f; - - if (!hci_dev_test_flag(req->hdev, HCI_BREDR_ENABLED)) - return; - - if (hci_dev_test_flag(req->hdev, HCI_EVENT_FILTER_CONFIGURED)) { - memset(&f, 0, sizeof(f)); - f.flt_type = HCI_FLT_CLEAR_ALL; - hci_req_add(req, HCI_OP_SET_EVENT_FLT, 1, &f); - } -} - -static void hci_req_set_event_filter(struct hci_request *req) -{ - struct bdaddr_list_with_flags *b; - struct hci_cp_set_event_filter f; - struct hci_dev *hdev = req->hdev; - u8 scan = SCAN_DISABLED; - bool scanning = test_bit(HCI_PSCAN, &hdev->flags); - - if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) - return; - - /* Always clear event filter when starting */ - hci_req_clear_event_filter(req); - - list_for_each_entry(b, &hdev->accept_list, list) { - if (!hci_conn_test_flag(HCI_CONN_FLAG_REMOTE_WAKEUP, - b->current_flags)) - continue; - - memset(&f, 0, sizeof(f)); - bacpy(&f.addr_conn_flt.bdaddr, &b->bdaddr); - f.flt_type = HCI_FLT_CONN_SETUP; - f.cond_type = HCI_CONN_SETUP_ALLOW_BDADDR; - f.addr_conn_flt.auto_accept = HCI_CONN_SETUP_AUTO_ON; - - bt_dev_dbg(hdev, "Adding event filters for %pMR", &b->bdaddr); - hci_req_add(req, HCI_OP_SET_EVENT_FLT, sizeof(f), &f); - scan = SCAN_PAGE; - } - - if (scan && !scanning) { - set_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks); - hci_req_add(req, HCI_OP_WRITE_SCAN_ENABLE, 1, &scan); - } else if (!scan && scanning) { - set_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); - hci_req_add(req, HCI_OP_WRITE_SCAN_ENABLE, 1, &scan); - } -} - static void cancel_adv_timeout(struct hci_dev *hdev) { if (hdev->adv_instance_timeout) { @@ -1013,185 +955,6 @@ int hci_req_resume_adv_instances(struct hci_dev *hdev) return hci_req_run(&req, NULL); } -static void suspend_req_complete(struct hci_dev *hdev, u8 status, u16 opcode) -{ - bt_dev_dbg(hdev, "Request complete opcode=0x%x, status=0x%x", opcode, - status); - if (test_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks) || - test_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks)) { - clear_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks); - clear_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); - wake_up(&hdev->suspend_wait_q); - } - - if (test_bit(SUSPEND_SET_ADV_FILTER, hdev->suspend_tasks)) { - clear_bit(SUSPEND_SET_ADV_FILTER, hdev->suspend_tasks); - wake_up(&hdev->suspend_wait_q); - } -} - -static void hci_req_prepare_adv_monitor_suspend(struct hci_request *req, - bool suspending) -{ - struct hci_dev *hdev = req->hdev; - - switch (hci_get_adv_monitor_offload_ext(hdev)) { - case HCI_ADV_MONITOR_EXT_MSFT: - if (suspending) - msft_suspend(hdev); - else - msft_resume(hdev); - break; - default: - return; - } - - /* No need to block when enabling since it's on resume path */ - if (hdev->suspended && suspending) - set_bit(SUSPEND_SET_ADV_FILTER, hdev->suspend_tasks); -} - -/* Call with hci_dev_lock */ -void hci_req_prepare_suspend(struct hci_dev *hdev, enum suspended_state next) -{ - int old_state; - struct hci_conn *conn; - struct hci_request req; - u8 page_scan; - int disconnect_counter; - - if (next == hdev->suspend_state) { - bt_dev_dbg(hdev, "Same state before and after: %d", next); - goto done; - } - - hdev->suspend_state = next; - hci_req_init(&req, hdev); - - if (next == BT_SUSPEND_DISCONNECT) { - /* Mark device as suspended */ - hdev->suspended = true; - - /* Pause discovery if not already stopped */ - old_state = hdev->discovery.state; - if (old_state != DISCOVERY_STOPPED) { - set_bit(SUSPEND_PAUSE_DISCOVERY, hdev->suspend_tasks); - hci_discovery_set_state(hdev, DISCOVERY_STOPPING); - queue_work(hdev->req_workqueue, &hdev->discov_update); - } - - hdev->discovery_paused = true; - hdev->discovery_old_state = old_state; - - /* Stop directed advertising */ - old_state = hci_dev_test_flag(hdev, HCI_ADVERTISING); - if (old_state) { - set_bit(SUSPEND_PAUSE_ADVERTISING, hdev->suspend_tasks); - cancel_delayed_work(&hdev->discov_off); - queue_delayed_work(hdev->req_workqueue, - &hdev->discov_off, 0); - } - - /* Pause other advertisements */ - if (hdev->adv_instance_cnt) - __hci_req_pause_adv_instances(&req); - - hdev->advertising_paused = true; - hdev->advertising_old_state = old_state; - - /* Disable page scan if enabled */ - if (test_bit(HCI_PSCAN, &hdev->flags)) { - page_scan = SCAN_DISABLED; - hci_req_add(&req, HCI_OP_WRITE_SCAN_ENABLE, 1, - &page_scan); - set_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); - } - - /* Disable LE passive scan if enabled */ - if (hci_dev_test_flag(hdev, HCI_LE_SCAN)) { - cancel_interleave_scan(hdev); - hci_req_add_le_scan_disable(&req, false); - } - - /* Disable advertisement filters */ - hci_req_prepare_adv_monitor_suspend(&req, true); - - /* Prevent disconnects from causing scanning to be re-enabled */ - hdev->scanning_paused = true; - - /* Run commands before disconnecting */ - hci_req_run(&req, suspend_req_complete); - - disconnect_counter = 0; - /* Soft disconnect everything (power off) */ - list_for_each_entry(conn, &hdev->conn_hash.list, list) { - hci_disconnect(conn, HCI_ERROR_REMOTE_POWER_OFF); - disconnect_counter++; - } - - if (disconnect_counter > 0) { - bt_dev_dbg(hdev, - "Had %d disconnects. Will wait on them", - disconnect_counter); - set_bit(SUSPEND_DISCONNECTING, hdev->suspend_tasks); - } - } else if (next == BT_SUSPEND_CONFIGURE_WAKE) { - /* Unpause to take care of updating scanning params */ - hdev->scanning_paused = false; - /* Enable event filter for paired devices */ - hci_req_set_event_filter(&req); - /* Enable passive scan at lower duty cycle */ - __hci_update_background_scan(&req); - /* Pause scan changes again. */ - hdev->scanning_paused = true; - hci_req_run(&req, suspend_req_complete); - } else { - hdev->suspended = false; - hdev->scanning_paused = false; - - /* Clear any event filters and restore scan state */ - hci_req_clear_event_filter(&req); - __hci_req_update_scan(&req); - - /* Reset passive/background scanning to normal */ - __hci_update_background_scan(&req); - /* Enable all of the advertisement filters */ - hci_req_prepare_adv_monitor_suspend(&req, false); - - /* Unpause directed advertising */ - hdev->advertising_paused = false; - if (hdev->advertising_old_state) { - set_bit(SUSPEND_UNPAUSE_ADVERTISING, - hdev->suspend_tasks); - hci_dev_set_flag(hdev, HCI_ADVERTISING); - queue_work(hdev->req_workqueue, - &hdev->discoverable_update); - hdev->advertising_old_state = 0; - } - - /* Resume other advertisements */ - if (hdev->adv_instance_cnt) - __hci_req_resume_adv_instances(&req); - - /* Unpause discovery */ - hdev->discovery_paused = false; - if (hdev->discovery_old_state != DISCOVERY_STOPPED && - hdev->discovery_old_state != DISCOVERY_STOPPING) { - set_bit(SUSPEND_UNPAUSE_DISCOVERY, hdev->suspend_tasks); - hci_discovery_set_state(hdev, DISCOVERY_STARTING); - queue_work(hdev->req_workqueue, &hdev->discov_update); - } - - hci_req_run(&req, suspend_req_complete); - } - - hdev->suspend_state = next; - -done: - clear_bit(SUSPEND_PREPARE_NOTIFIER, hdev->suspend_tasks); - wake_up(&hdev->suspend_wait_q); -} - static bool adv_cur_instance_is_scannable(struct hci_dev *hdev) { return hci_adv_instance_is_scannable(hdev, hdev->cur_adv_instance); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 4e0a7719b58d..e3f44e621b7f 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1410,9 +1410,6 @@ int hci_scan_disable_sync(struct hci_dev *hdev) return 0; } - if (hdev->suspended) - set_bit(SUSPEND_SCAN_DISABLE, hdev->suspend_tasks); - err = hci_le_set_scan_enable_sync(hdev, LE_SCAN_DISABLE, 0x00); if (err) { bt_dev_err(hdev, "Unable to disable scanning: %d", err); @@ -1642,10 +1639,11 @@ static int hci_le_add_accept_list_sync(struct hci_dev *hdev, return 0; } -/* This function disables all advertising instances (including 0x00) */ +/* This function disables/pause all advertising instances */ static int hci_pause_advertising_sync(struct hci_dev *hdev) { int err; + int old_state; /* If there are no instances or advertising has already been paused * there is nothing to do. @@ -1653,6 +1651,21 @@ static int hci_pause_advertising_sync(struct hci_dev *hdev) if (!hdev->adv_instance_cnt || hdev->advertising_paused) return 0; + bt_dev_dbg(hdev, "Pausing directed advertising"); + + /* Stop directed advertising */ + old_state = hci_dev_test_flag(hdev, HCI_ADVERTISING); + if (old_state) { + /* When discoverable timeout triggers, then just make sure + * the limited discoverable flag is cleared. Even in the case + * of a timeout triggered from general discoverable, it is + * safe to unconditionally clear the flag. + */ + hci_dev_clear_flag(hdev, HCI_LIMITED_DISCOVERABLE); + hci_dev_clear_flag(hdev, HCI_DISCOVERABLE); + hdev->discov_timeout = 0; + } + bt_dev_dbg(hdev, "Pausing advertising instances"); /* Call to disable any advertisements active on the controller. @@ -1667,11 +1680,12 @@ static int hci_pause_advertising_sync(struct hci_dev *hdev) cancel_adv_timeout(hdev); hdev->advertising_paused = true; + hdev->advertising_old_state = old_state; return 0; } -/* This function enables all user advertising instances (excluding 0x00) */ +/* This function enables all user advertising instances */ static int hci_resume_advertising_sync(struct hci_dev *hdev) { struct adv_info *adv, *tmp; @@ -1681,6 +1695,14 @@ static int hci_resume_advertising_sync(struct hci_dev *hdev) if (!hdev->advertising_paused) return 0; + /* Resume directed advertising */ + hdev->advertising_paused = false; + if (hdev->advertising_old_state) { + hci_dev_set_flag(hdev, HCI_ADVERTISING); + queue_work(hdev->req_workqueue, &hdev->discoverable_update); + hdev->advertising_old_state = 0; + } + bt_dev_dbg(hdev, "Resuming advertising instances"); if (ext_adv_capable(hdev)) { @@ -2002,8 +2024,6 @@ int hci_passive_scan_sync(struct hci_dev *hdev) if (hdev->suspended) { window = hdev->le_scan_window_suspend; interval = hdev->le_scan_int_suspend; - - set_bit(SUSPEND_SCAN_ENABLE, hdev->suspend_tasks); } else if (hci_is_le_conn_scanning(hdev)) { window = hdev->le_scan_window_connect; interval = hdev->le_scan_int_connect; @@ -2937,6 +2957,13 @@ static int hci_set_event_mask_sync(struct hci_dev *hdev) if (lmp_bredr_capable(hdev)) { events[4] |= 0x01; /* Flow Specification Complete */ + + /* Don't set Disconnect Complete when suspended as that + * would wakeup the host when disconnecting due to + * suspend. + */ + if (hdev->suspended) + events[0] &= 0xef; } else { /* Use a different default for LE-only devices */ memset(events, 0, sizeof(events)); @@ -2949,7 +2976,12 @@ static int hci_set_event_mask_sync(struct hci_dev *hdev) * control related events. */ if (hdev->commands[0] & 0x20) { - events[0] |= 0x10; /* Disconnection Complete */ + /* Don't set Disconnect Complete when suspended as that + * would wakeup the host when disconnecting due to + * suspend. + */ + if (!hdev->suspended) + events[0] |= 0x10; /* Disconnection Complete */ events[2] |= 0x04; /* Number of Completed Packets */ events[3] |= 0x02; /* Data Buffer Overflow */ } @@ -4033,9 +4065,6 @@ int hci_dev_close_sync(struct hci_dev *hdev) clear_bit(HCI_RUNNING, &hdev->flags); hci_sock_dev_event(hdev, HCI_DEV_CLOSE); - if (test_and_clear_bit(SUSPEND_POWERING_DOWN, hdev->suspend_tasks)) - wake_up(&hdev->suspend_wait_q); - /* After this point our queues are empty and no tasks are scheduled. */ hdev->close(hdev); @@ -4299,6 +4328,20 @@ static int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, return 0; } +static int hci_disconnect_all_sync(struct hci_dev *hdev, u8 reason) +{ + struct hci_conn *conn, *tmp; + int err; + + list_for_each_entry_safe(conn, tmp, &hdev->conn_hash.list, list) { + err = hci_abort_conn_sync(hdev, conn, reason); + if (err) + return err; + } + + return err; +} + /* This function perform power off HCI command sequence as follows: * * Clear Advertising @@ -4308,7 +4351,6 @@ static int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, */ static int hci_power_off_sync(struct hci_dev *hdev) { - struct hci_conn *conn; int err; /* If controller is already down there is nothing to do */ @@ -4330,10 +4372,10 @@ static int hci_power_off_sync(struct hci_dev *hdev) if (err) return err; - list_for_each_entry(conn, &hdev->conn_hash.list, list) { - /* 0x15 == Terminated due to Power Off */ - hci_abort_conn_sync(hdev, conn, 0x15); - } + /* Terminated due to Power Off */ + err = hci_disconnect_all_sync(hdev, HCI_ERROR_REMOTE_POWER_OFF); + if (err) + return err; return hci_dev_close_sync(hdev); } @@ -4535,3 +4577,223 @@ int hci_start_discovery_sync(struct hci_dev *hdev) timeout); return 0; } + +static void hci_suspend_monitor_sync(struct hci_dev *hdev) +{ + switch (hci_get_adv_monitor_offload_ext(hdev)) { + case HCI_ADV_MONITOR_EXT_MSFT: + msft_suspend_sync(hdev); + break; + default: + return; + } +} + +/* This function disables discovery and mark it as paused */ +static int hci_pause_discovery_sync(struct hci_dev *hdev) +{ + int old_state = hdev->discovery.state; + int err; + + /* If discovery already stopped/stopping/paused there nothing to do */ + if (old_state == DISCOVERY_STOPPED || old_state == DISCOVERY_STOPPING || + hdev->discovery_paused) + return 0; + + hci_discovery_set_state(hdev, DISCOVERY_STOPPING); + err = hci_stop_discovery_sync(hdev); + if (err) + return err; + + hdev->discovery_paused = true; + hdev->discovery_old_state = old_state; + hci_discovery_set_state(hdev, DISCOVERY_STOPPED); + + return 0; +} + +static int hci_update_event_filter_sync(struct hci_dev *hdev) +{ + struct bdaddr_list_with_flags *b; + u8 scan = SCAN_DISABLED; + bool scanning = test_bit(HCI_PSCAN, &hdev->flags); + int err; + + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + return 0; + + /* Always clear event filter when starting */ + hci_clear_event_filter_sync(hdev); + + list_for_each_entry(b, &hdev->accept_list, list) { + if (!hci_conn_test_flag(HCI_CONN_FLAG_REMOTE_WAKEUP, + b->current_flags)) + continue; + + bt_dev_dbg(hdev, "Adding event filters for %pMR", &b->bdaddr); + + err = hci_set_event_filter_sync(hdev, HCI_FLT_CONN_SETUP, + HCI_CONN_SETUP_ALLOW_BDADDR, + &b->bdaddr, + HCI_CONN_SETUP_AUTO_ON); + if (err) + bt_dev_dbg(hdev, "Failed to set event filter for %pMR", + &b->bdaddr); + else + scan = SCAN_PAGE; + } + + if (scan && !scanning) + hci_write_scan_enable_sync(hdev, scan); + else if (!scan && scanning) + hci_write_scan_enable_sync(hdev, scan); + + return 0; +} + +/* This function performs the HCI suspend procedures in the follow order: + * + * Pause discovery (active scanning/inquiry) + * Pause Directed Advertising/Advertising + * Disconnect all connections + * Set suspend_status to BT_SUSPEND_DISCONNECT if hdev cannot wakeup + * otherwise: + * Update event mask (only set events that are allowed to wake up the host) + * Update event filter (with devices marked with HCI_CONN_FLAG_REMOTE_WAKEUP) + * Update passive scanning (lower duty cycle) + * Set suspend_status to BT_SUSPEND_CONFIGURE_WAKE + */ +int hci_suspend_sync(struct hci_dev *hdev) +{ + int err; + + /* If marked as suspended there nothing to do */ + if (hdev->suspended) + return 0; + + /* Mark device as suspended */ + hdev->suspended = true; + + /* Pause discovery if not already stopped */ + hci_pause_discovery_sync(hdev); + + /* Pause other advertisements */ + hci_pause_advertising_sync(hdev); + + /* Disable page scan if enabled */ + if (test_bit(HCI_PSCAN, &hdev->flags)) + hci_write_scan_enable_sync(hdev, SCAN_DISABLED); + + /* Suspend monitor filters */ + hci_suspend_monitor_sync(hdev); + + /* Prevent disconnects from causing scanning to be re-enabled */ + hdev->scanning_paused = true; + + /* Soft disconnect everything (power off) */ + err = hci_disconnect_all_sync(hdev, HCI_ERROR_REMOTE_POWER_OFF); + if (err) { + /* Set state to BT_RUNNING so resume doesn't notify */ + hdev->suspend_state = BT_RUNNING; + hci_resume_sync(hdev); + return err; + } + + /* Only configure accept list if disconnect succeeded and wake + * isn't being prevented. + */ + if (!hdev->wakeup || !hdev->wakeup(hdev)) { + hdev->suspend_state = BT_SUSPEND_DISCONNECT; + return 0; + } + + /* Unpause to take care of updating scanning params */ + hdev->scanning_paused = false; + + /* Update event mask so only the allowed event can wakeup the host */ + hci_set_event_mask_sync(hdev); + + /* Enable event filter for paired devices */ + hci_update_event_filter_sync(hdev); + + /* Update LE passive scan if enabled */ + hci_update_passive_scan_sync(hdev); + + /* Pause scan changes again. */ + hdev->scanning_paused = true; + + hdev->suspend_state = BT_SUSPEND_CONFIGURE_WAKE; + + return 0; +} + +/* This function resumes discovery */ +static int hci_resume_discovery_sync(struct hci_dev *hdev) +{ + int err; + + /* If discovery not paused there nothing to do */ + if (!hdev->discovery_paused) + return 0; + + hdev->discovery_paused = false; + + hci_discovery_set_state(hdev, DISCOVERY_STARTING); + + err = hci_start_discovery_sync(hdev); + + hci_discovery_set_state(hdev, err ? DISCOVERY_STOPPED : + DISCOVERY_FINDING); + + return err; +} + +static void hci_resume_monitor_sync(struct hci_dev *hdev) +{ + switch (hci_get_adv_monitor_offload_ext(hdev)) { + case HCI_ADV_MONITOR_EXT_MSFT: + msft_resume_sync(hdev); + break; + default: + return; + } +} + +/* This function performs the HCI suspend procedures in the follow order: + * + * Restore event mask + * Clear event filter + * Update passive scanning (normal duty cycle) + * Resume Directed Advertising/Advertising + * Resume discovery (active scanning/inquiry) + */ +int hci_resume_sync(struct hci_dev *hdev) +{ + /* If not marked as suspended there nothing to do */ + if (!hdev->suspended) + return 0; + + hdev->suspended = false; + hdev->scanning_paused = false; + + /* Restore event mask */ + hci_set_event_mask_sync(hdev); + + /* Clear any event filters and restore scan state */ + hci_clear_event_filter_sync(hdev); + hci_update_scan_sync(hdev); + + /* Reset passive scanning to normal */ + hci_update_passive_scan_sync(hdev); + + /* Resume monitor filters */ + hci_resume_monitor_sync(hdev); + + /* Resume other advertisements */ + hci_resume_advertising_sync(hdev); + + /* Resume discovery */ + hci_resume_discovery_sync(hdev); + + return 0; +} diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index bfa08eb4c2b6..a7d35c138713 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5171,13 +5171,6 @@ void mgmt_start_discovery_complete(struct hci_dev *hdev, u8 status) } hci_dev_unlock(hdev); - - /* Handle suspend notifier */ - if (test_and_clear_bit(SUSPEND_UNPAUSE_DISCOVERY, - hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Unpaused discovery"); - wake_up(&hdev->suspend_wait_q); - } } static bool discovery_type_is_valid(struct hci_dev *hdev, uint8_t type, @@ -5217,14 +5210,7 @@ static void start_discovery_complete(struct hci_dev *hdev, void *data, int err) cmd->param, 1); mgmt_pending_free(cmd); - /* Handle suspend notifier */ - if (test_and_clear_bit(SUSPEND_UNPAUSE_DISCOVERY, - hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Unpaused discovery"); - wake_up(&hdev->suspend_wait_q); - } - - hci_discovery_set_state(hdev, err ? DISCOVERY_STOPPED : + hci_discovery_set_state(hdev, err ? DISCOVERY_STOPPED: DISCOVERY_FINDING); } @@ -5446,12 +5432,6 @@ void mgmt_stop_discovery_complete(struct hci_dev *hdev, u8 status) } hci_dev_unlock(hdev); - - /* Handle suspend notifier */ - if (test_and_clear_bit(SUSPEND_PAUSE_DISCOVERY, hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Paused discovery"); - wake_up(&hdev->suspend_wait_q); - } } static void stop_discovery_complete(struct hci_dev *hdev, void *data, int err) @@ -5464,12 +5444,6 @@ static void stop_discovery_complete(struct hci_dev *hdev, void *data, int err) cmd->param, 1); mgmt_pending_free(cmd); - /* Handle suspend notifier */ - if (test_and_clear_bit(SUSPEND_PAUSE_DISCOVERY, hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Paused discovery"); - wake_up(&hdev->suspend_wait_q); - } - if (!err) hci_discovery_set_state(hdev, DISCOVERY_STOPPED); } @@ -5709,17 +5683,6 @@ static void set_advertising_complete(struct hci_dev *hdev, void *data, int err) if (match.sk) sock_put(match.sk); - /* Handle suspend notifier */ - if (test_and_clear_bit(SUSPEND_PAUSE_ADVERTISING, - hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Paused advertising"); - wake_up(&hdev->suspend_wait_q); - } else if (test_and_clear_bit(SUSPEND_UNPAUSE_ADVERTISING, - hdev->suspend_tasks)) { - bt_dev_dbg(hdev, "Unpaused advertising"); - wake_up(&hdev->suspend_wait_q); - } - /* If "Set Advertising" was just disabled and instance advertising was * set up earlier, then re-enable multi-instance advertising. */ diff --git a/net/bluetooth/msft.c b/net/bluetooth/msft.c index 5205d9410ff5..1122097e1e49 100644 --- a/net/bluetooth/msft.c +++ b/net/bluetooth/msft.c @@ -93,7 +93,7 @@ struct msft_data { struct list_head handle_map; __u16 pending_add_handle; __u16 pending_remove_handle; - __u8 reregistering; + __u8 resuming; __u8 suspending; __u8 filter_enabled; }; @@ -156,7 +156,6 @@ failed: return false; } -/* This function requires the caller holds hdev->lock */ static void reregister_monitor(struct hci_dev *hdev, int handle) { struct adv_monitor *monitor; @@ -166,8 +165,8 @@ static void reregister_monitor(struct hci_dev *hdev, int handle) while (1) { monitor = idr_get_next(&hdev->adv_monitors_idr, &handle); if (!monitor) { - /* All monitors have been reregistered */ - msft->reregistering = false; + /* All monitors have been resumed */ + msft->resuming = false; hci_update_passive_scan(hdev); return; } @@ -185,67 +184,317 @@ static void reregister_monitor(struct hci_dev *hdev, int handle) } } -/* This function requires the caller holds hdev->lock */ -static void remove_monitor_on_suspend(struct hci_dev *hdev, int handle) +/* is_mgmt = true matches the handle exposed to userspace via mgmt. + * is_mgmt = false matches the handle used by the msft controller. + * This function requires the caller holds hdev->lock + */ +static struct msft_monitor_advertisement_handle_data *msft_find_handle_data + (struct hci_dev *hdev, u16 handle, bool is_mgmt) { + struct msft_monitor_advertisement_handle_data *entry; + struct msft_data *msft = hdev->msft_data; + + list_for_each_entry(entry, &msft->handle_map, list) { + if (is_mgmt && entry->mgmt_handle == handle) + return entry; + if (!is_mgmt && entry->msft_handle == handle) + return entry; + } + + return NULL; +} + +static void msft_le_monitor_advertisement_cb(struct hci_dev *hdev, + u8 status, u16 opcode, + struct sk_buff *skb) +{ + struct msft_rp_le_monitor_advertisement *rp; struct adv_monitor *monitor; + struct msft_monitor_advertisement_handle_data *handle_data; + struct msft_data *msft = hdev->msft_data; + + hci_dev_lock(hdev); + + monitor = idr_find(&hdev->adv_monitors_idr, msft->pending_add_handle); + if (!monitor) { + bt_dev_err(hdev, "msft add advmon: monitor %u is not found!", + msft->pending_add_handle); + status = HCI_ERROR_UNSPECIFIED; + goto unlock; + } + + if (status) + goto unlock; + + rp = (struct msft_rp_le_monitor_advertisement *)skb->data; + if (skb->len < sizeof(*rp)) { + status = HCI_ERROR_UNSPECIFIED; + goto unlock; + } + + handle_data = kmalloc(sizeof(*handle_data), GFP_KERNEL); + if (!handle_data) { + status = HCI_ERROR_UNSPECIFIED; + goto unlock; + } + + handle_data->mgmt_handle = monitor->handle; + handle_data->msft_handle = rp->handle; + INIT_LIST_HEAD(&handle_data->list); + list_add(&handle_data->list, &msft->handle_map); + + monitor->state = ADV_MONITOR_STATE_OFFLOADED; + +unlock: + if (status && monitor) + hci_free_adv_monitor(hdev, monitor); + + hci_dev_unlock(hdev); + + if (!msft->resuming) + hci_add_adv_patterns_monitor_complete(hdev, status); +} + +static void msft_le_cancel_monitor_advertisement_cb(struct hci_dev *hdev, + u8 status, u16 opcode, + struct sk_buff *skb) +{ + struct msft_cp_le_cancel_monitor_advertisement *cp; + struct msft_rp_le_cancel_monitor_advertisement *rp; + struct adv_monitor *monitor; + struct msft_monitor_advertisement_handle_data *handle_data; struct msft_data *msft = hdev->msft_data; int err; + bool pending; - while (1) { - monitor = idr_get_next(&hdev->adv_monitors_idr, &handle); - if (!monitor) { - /* All monitors have been removed */ - msft->suspending = false; - hci_update_background_scan(hdev); + if (status) + goto done; + + rp = (struct msft_rp_le_cancel_monitor_advertisement *)skb->data; + if (skb->len < sizeof(*rp)) { + status = HCI_ERROR_UNSPECIFIED; + goto done; + } + + hci_dev_lock(hdev); + + cp = hci_sent_cmd_data(hdev, hdev->msft_opcode); + handle_data = msft_find_handle_data(hdev, cp->handle, false); + + if (handle_data) { + monitor = idr_find(&hdev->adv_monitors_idr, + handle_data->mgmt_handle); + + if (monitor && monitor->state == ADV_MONITOR_STATE_OFFLOADED) + monitor->state = ADV_MONITOR_STATE_REGISTERED; + + /* Do not free the monitor if it is being removed due to + * suspend. It will be re-monitored on resume. + */ + if (monitor && !msft->suspending) + hci_free_adv_monitor(hdev, monitor); + + list_del(&handle_data->list); + kfree(handle_data); + } + + /* If remove all monitors is required, we need to continue the process + * here because the earlier it was paused when waiting for the + * response from controller. + */ + if (msft->pending_remove_handle == 0) { + pending = hci_remove_all_adv_monitor(hdev, &err); + if (pending) { + hci_dev_unlock(hdev); return; } - msft->pending_remove_handle = (u16)handle; - err = __msft_remove_monitor(hdev, monitor, handle); + if (err) + status = HCI_ERROR_UNSPECIFIED; + } - /* If success, return and wait for monitor removed callback */ - if (!err) - return; + hci_dev_unlock(hdev); + +done: + if (!msft->suspending) + hci_remove_adv_monitor_complete(hdev, status); +} + +static int msft_remove_monitor_sync(struct hci_dev *hdev, + struct adv_monitor *monitor) +{ + struct msft_cp_le_cancel_monitor_advertisement cp; + struct msft_monitor_advertisement_handle_data *handle_data; + struct sk_buff *skb; + u8 status; + + handle_data = msft_find_handle_data(hdev, monitor->handle, true); + + /* If no matched handle, just remove without telling controller */ + if (!handle_data) + return -ENOENT; + + cp.sub_opcode = MSFT_OP_LE_CANCEL_MONITOR_ADVERTISEMENT; + cp.handle = handle_data->msft_handle; + + skb = __hci_cmd_sync(hdev, hdev->msft_opcode, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); + if (IS_ERR(skb)) + return PTR_ERR(skb); + + status = skb->data[0]; + skb_pull(skb, 1); + + msft_le_cancel_monitor_advertisement_cb(hdev, status, hdev->msft_opcode, + skb); + + return status; +} + +/* This function requires the caller holds hci_req_sync_lock */ +int msft_suspend_sync(struct hci_dev *hdev) +{ + struct msft_data *msft = hdev->msft_data; + struct adv_monitor *monitor; + int handle = 0; + + if (!msft || !msft_monitor_supported(hdev)) + return 0; + + msft->suspending = true; + + while (1) { + monitor = idr_get_next(&hdev->adv_monitors_idr, &handle); + if (!monitor) + break; + + msft_remove_monitor_sync(hdev, monitor); - /* Otherwise free the monitor and keep removing */ - hci_free_adv_monitor(hdev, monitor); handle++; } + + /* All monitors have been removed */ + msft->suspending = false; + + return 0; } -/* This function requires the caller holds hdev->lock */ -void msft_suspend(struct hci_dev *hdev) +static bool msft_monitor_rssi_valid(struct adv_monitor *monitor) { - struct msft_data *msft = hdev->msft_data; + struct adv_rssi_thresholds *r = &monitor->rssi; - if (!msft) - return; + if (r->high_threshold < MSFT_RSSI_THRESHOLD_VALUE_MIN || + r->high_threshold > MSFT_RSSI_THRESHOLD_VALUE_MAX || + r->low_threshold < MSFT_RSSI_THRESHOLD_VALUE_MIN || + r->low_threshold > MSFT_RSSI_THRESHOLD_VALUE_MAX) + return false; - if (msft_monitor_supported(hdev)) { - msft->suspending = true; - /* Quitely remove all monitors on suspend to avoid waking up - * the system. - */ - remove_monitor_on_suspend(hdev, 0); - } + /* High_threshold_timeout is not supported, + * once high_threshold is reached, events are immediately reported. + */ + if (r->high_threshold_timeout != 0) + return false; + + if (r->low_threshold_timeout > MSFT_RSSI_LOW_TIMEOUT_MAX) + return false; + + /* Sampling period from 0x00 to 0xFF are all allowed */ + return true; } -/* This function requires the caller holds hdev->lock */ -void msft_resume(struct hci_dev *hdev) +static bool msft_monitor_pattern_valid(struct adv_monitor *monitor) +{ + return msft_monitor_rssi_valid(monitor); + /* No additional check needed for pattern-based monitor */ +} + +static int msft_add_monitor_sync(struct hci_dev *hdev, + struct adv_monitor *monitor) +{ + struct msft_cp_le_monitor_advertisement *cp; + struct msft_le_monitor_advertisement_pattern_data *pattern_data; + struct msft_le_monitor_advertisement_pattern *pattern; + struct adv_pattern *entry; + size_t total_size = sizeof(*cp) + sizeof(*pattern_data); + ptrdiff_t offset = 0; + u8 pattern_count = 0; + struct sk_buff *skb; + u8 status; + + if (!msft_monitor_pattern_valid(monitor)) + return -EINVAL; + + list_for_each_entry(entry, &monitor->patterns, list) { + pattern_count++; + total_size += sizeof(*pattern) + entry->length; + } + + cp = kmalloc(total_size, GFP_KERNEL); + if (!cp) + return -ENOMEM; + + cp->sub_opcode = MSFT_OP_LE_MONITOR_ADVERTISEMENT; + cp->rssi_high = monitor->rssi.high_threshold; + cp->rssi_low = monitor->rssi.low_threshold; + cp->rssi_low_interval = (u8)monitor->rssi.low_threshold_timeout; + cp->rssi_sampling_period = monitor->rssi.sampling_period; + + cp->cond_type = MSFT_MONITOR_ADVERTISEMENT_TYPE_PATTERN; + + pattern_data = (void *)cp->data; + pattern_data->count = pattern_count; + + list_for_each_entry(entry, &monitor->patterns, list) { + pattern = (void *)(pattern_data->data + offset); + /* the length also includes data_type and offset */ + pattern->length = entry->length + 2; + pattern->data_type = entry->ad_type; + pattern->start_byte = entry->offset; + memcpy(pattern->pattern, entry->value, entry->length); + offset += sizeof(*pattern) + entry->length; + } + + skb = __hci_cmd_sync(hdev, hdev->msft_opcode, total_size, cp, + HCI_CMD_TIMEOUT); + kfree(cp); + + if (IS_ERR(skb)) + return PTR_ERR(skb); + + status = skb->data[0]; + skb_pull(skb, 1); + + msft_le_monitor_advertisement_cb(hdev, status, hdev->msft_opcode, skb); + + return status; +} + +/* This function requires the caller holds hci_req_sync_lock */ +int msft_resume_sync(struct hci_dev *hdev) { struct msft_data *msft = hdev->msft_data; + struct adv_monitor *monitor; + int handle = 0; - if (!msft) - return; + if (!msft || !msft_monitor_supported(hdev)) + return 0; - if (msft_monitor_supported(hdev)) { - msft->reregistering = true; - /* Monitors are removed on suspend, so we need to add all - * monitors on resume. - */ - reregister_monitor(hdev, 0); + msft->resuming = true; + + while (1) { + monitor = idr_get_next(&hdev->adv_monitors_idr, &handle); + if (!monitor) + break; + + msft_add_monitor_sync(hdev, monitor); + + handle++; } + + /* All monitors have been resumed */ + msft->resuming = false; + + return 0; } void msft_do_open(struct hci_dev *hdev) @@ -275,7 +524,7 @@ void msft_do_open(struct hci_dev *hdev) } if (msft_monitor_supported(hdev)) { - msft->reregistering = true; + msft->resuming = true; msft_set_filter_enable(hdev, true); /* Monitors get removed on power off, so we need to explicitly * tell the controller to re-monitor. @@ -381,151 +630,6 @@ __u64 msft_get_features(struct hci_dev *hdev) return msft ? msft->features : 0; } -/* is_mgmt = true matches the handle exposed to userspace via mgmt. - * is_mgmt = false matches the handle used by the msft controller. - * This function requires the caller holds hdev->lock - */ -static struct msft_monitor_advertisement_handle_data *msft_find_handle_data - (struct hci_dev *hdev, u16 handle, bool is_mgmt) -{ - struct msft_monitor_advertisement_handle_data *entry; - struct msft_data *msft = hdev->msft_data; - - list_for_each_entry(entry, &msft->handle_map, list) { - if (is_mgmt && entry->mgmt_handle == handle) - return entry; - if (!is_mgmt && entry->msft_handle == handle) - return entry; - } - - return NULL; -} - -static void msft_le_monitor_advertisement_cb(struct hci_dev *hdev, - u8 status, u16 opcode, - struct sk_buff *skb) -{ - struct msft_rp_le_monitor_advertisement *rp; - struct adv_monitor *monitor; - struct msft_monitor_advertisement_handle_data *handle_data; - struct msft_data *msft = hdev->msft_data; - - hci_dev_lock(hdev); - - monitor = idr_find(&hdev->adv_monitors_idr, msft->pending_add_handle); - if (!monitor) { - bt_dev_err(hdev, "msft add advmon: monitor %u is not found!", - msft->pending_add_handle); - status = HCI_ERROR_UNSPECIFIED; - goto unlock; - } - - if (status) - goto unlock; - - rp = (struct msft_rp_le_monitor_advertisement *)skb->data; - if (skb->len < sizeof(*rp)) { - status = HCI_ERROR_UNSPECIFIED; - goto unlock; - } - - handle_data = kmalloc(sizeof(*handle_data), GFP_KERNEL); - if (!handle_data) { - status = HCI_ERROR_UNSPECIFIED; - goto unlock; - } - - handle_data->mgmt_handle = monitor->handle; - handle_data->msft_handle = rp->handle; - INIT_LIST_HEAD(&handle_data->list); - list_add(&handle_data->list, &msft->handle_map); - - monitor->state = ADV_MONITOR_STATE_OFFLOADED; - -unlock: - if (status && monitor) - hci_free_adv_monitor(hdev, monitor); - - /* If in restart/reregister sequence, keep registering. */ - if (msft->reregistering) - reregister_monitor(hdev, msft->pending_add_handle + 1); - - hci_dev_unlock(hdev); - - if (!msft->reregistering) - hci_add_adv_patterns_monitor_complete(hdev, status); -} - -static void msft_le_cancel_monitor_advertisement_cb(struct hci_dev *hdev, - u8 status, u16 opcode, - struct sk_buff *skb) -{ - struct msft_cp_le_cancel_monitor_advertisement *cp; - struct msft_rp_le_cancel_monitor_advertisement *rp; - struct adv_monitor *monitor; - struct msft_monitor_advertisement_handle_data *handle_data; - struct msft_data *msft = hdev->msft_data; - int err; - bool pending; - - if (status) - goto done; - - rp = (struct msft_rp_le_cancel_monitor_advertisement *)skb->data; - if (skb->len < sizeof(*rp)) { - status = HCI_ERROR_UNSPECIFIED; - goto done; - } - - hci_dev_lock(hdev); - - cp = hci_sent_cmd_data(hdev, hdev->msft_opcode); - handle_data = msft_find_handle_data(hdev, cp->handle, false); - - if (handle_data) { - monitor = idr_find(&hdev->adv_monitors_idr, - handle_data->mgmt_handle); - - if (monitor && monitor->state == ADV_MONITOR_STATE_OFFLOADED) - monitor->state = ADV_MONITOR_STATE_REGISTERED; - - /* Do not free the monitor if it is being removed due to - * suspend. It will be re-monitored on resume. - */ - if (monitor && !msft->suspending) - hci_free_adv_monitor(hdev, monitor); - - list_del(&handle_data->list); - kfree(handle_data); - } - - /* If in suspend/remove sequence, keep removing. */ - if (msft->suspending) - remove_monitor_on_suspend(hdev, - msft->pending_remove_handle + 1); - - /* If remove all monitors is required, we need to continue the process - * here because the earlier it was paused when waiting for the - * response from controller. - */ - if (msft->pending_remove_handle == 0) { - pending = hci_remove_all_adv_monitor(hdev, &err); - if (pending) { - hci_dev_unlock(hdev); - return; - } - - if (err) - status = HCI_ERROR_UNSPECIFIED; - } - - hci_dev_unlock(hdev); - -done: - if (!msft->suspending) - hci_remove_adv_monitor_complete(hdev, status); -} - static void msft_le_set_advertisement_filter_enable_cb(struct hci_dev *hdev, u8 status, u16 opcode, struct sk_buff *skb) @@ -560,35 +664,6 @@ static void msft_le_set_advertisement_filter_enable_cb(struct hci_dev *hdev, hci_dev_unlock(hdev); } -static bool msft_monitor_rssi_valid(struct adv_monitor *monitor) -{ - struct adv_rssi_thresholds *r = &monitor->rssi; - - if (r->high_threshold < MSFT_RSSI_THRESHOLD_VALUE_MIN || - r->high_threshold > MSFT_RSSI_THRESHOLD_VALUE_MAX || - r->low_threshold < MSFT_RSSI_THRESHOLD_VALUE_MIN || - r->low_threshold > MSFT_RSSI_THRESHOLD_VALUE_MAX) - return false; - - /* High_threshold_timeout is not supported, - * once high_threshold is reached, events are immediately reported. - */ - if (r->high_threshold_timeout != 0) - return false; - - if (r->low_threshold_timeout > MSFT_RSSI_LOW_TIMEOUT_MAX) - return false; - - /* Sampling period from 0x00 to 0xFF are all allowed */ - return true; -} - -static bool msft_monitor_pattern_valid(struct adv_monitor *monitor) -{ - return msft_monitor_rssi_valid(monitor); - /* No additional check needed for pattern-based monitor */ -} - /* This function requires the caller holds hdev->lock */ static int __msft_add_monitor_pattern(struct hci_dev *hdev, struct adv_monitor *monitor) @@ -656,7 +731,7 @@ int msft_add_monitor_pattern(struct hci_dev *hdev, struct adv_monitor *monitor) if (!msft) return -EOPNOTSUPP; - if (msft->reregistering || msft->suspending) + if (msft->resuming || msft->suspending) return -EBUSY; return __msft_add_monitor_pattern(hdev, monitor); @@ -700,7 +775,7 @@ int msft_remove_monitor(struct hci_dev *hdev, struct adv_monitor *monitor, if (!msft) return -EOPNOTSUPP; - if (msft->reregistering || msft->suspending) + if (msft->resuming || msft->suspending) return -EBUSY; return __msft_remove_monitor(hdev, monitor, handle); diff --git a/net/bluetooth/msft.h b/net/bluetooth/msft.h index 59c6e081c789..b59b63dc0ea8 100644 --- a/net/bluetooth/msft.h +++ b/net/bluetooth/msft.h @@ -24,8 +24,8 @@ int msft_remove_monitor(struct hci_dev *hdev, struct adv_monitor *monitor, u16 handle); void msft_req_add_set_filter_enable(struct hci_request *req, bool enable); int msft_set_filter_enable(struct hci_dev *hdev, bool enable); -void msft_suspend(struct hci_dev *hdev); -void msft_resume(struct hci_dev *hdev); +int msft_suspend_sync(struct hci_dev *hdev); +int msft_resume_sync(struct hci_dev *hdev); bool msft_curve_validity(struct hci_dev *hdev); #else @@ -61,8 +61,15 @@ static inline int msft_set_filter_enable(struct hci_dev *hdev, bool enable) return -EOPNOTSUPP; } -static inline void msft_suspend(struct hci_dev *hdev) {} -static inline void msft_resume(struct hci_dev *hdev) {} +static inline int msft_suspend_sync(struct hci_dev *hdev) +{ + return -EOPNOTSUPP; +} + +static inline int msft_resume_sync(struct hci_dev *hdev) +{ + return -EOPNOTSUPP; +} static inline bool msft_curve_validity(struct hci_dev *hdev) { From d6cba4e6d0e202276b81117b96106859a4e2d56e Mon Sep 17 00:00:00 2001 From: Tim Jiang Date: Fri, 29 Oct 2021 11:21:21 +0800 Subject: [PATCH 0054/2048] Bluetooth: btusb: Add support using different nvm for variant WCN6855 controller the RF performance of wcn6855 soc chip from different foundries will be difference, so we should use different nvm to configure them. Signed-off-by: Tim Jiang Reviewed-by: Matthias Kaehlcke Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btusb.c | 53 ++++++++++++++++++++++++++++----------- 1 file changed, 39 insertions(+), 14 deletions(-) diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index 88bfdf1347a7..46d892bbde62 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2876,6 +2876,9 @@ static int btusb_set_bdaddr_wcn6855(struct hci_dev *hdev, #define QCA_DFU_TIMEOUT 3000 #define QCA_FLAG_MULTI_NVM 0x80 +#define WCN6855_2_0_RAM_VERSION_GF 0x400c1200 +#define WCN6855_2_1_RAM_VERSION_GF 0x400c1211 + struct qca_version { __le32 rom_version; __le32 patch_version; @@ -2907,6 +2910,7 @@ static const struct qca_device_info qca_devices_table[] = { { 0x00000302, 28, 4, 16 }, /* Rome 3.2 */ { 0x00130100, 40, 4, 16 }, /* WCN6855 1.0 */ { 0x00130200, 40, 4, 16 }, /* WCN6855 2.0 */ + { 0x00130201, 40, 4, 16 }, /* WCN6855 2.1 */ }; static int btusb_qca_send_vendor_req(struct usb_device *udev, u8 request, @@ -3061,6 +3065,40 @@ done: return err; } +static void btusb_generate_qca_nvm_name(char *fwname, size_t max_size, + const struct qca_version *ver) +{ + u32 rom_version = le32_to_cpu(ver->rom_version); + u16 flag = le16_to_cpu(ver->flag); + + if (((flag >> 8) & 0xff) == QCA_FLAG_MULTI_NVM) { + u16 board_id = le16_to_cpu(ver->board_id); + const char *variant; + + switch (le32_to_cpu(ver->ram_version)) { + case WCN6855_2_0_RAM_VERSION_GF: + case WCN6855_2_1_RAM_VERSION_GF: + variant = "_gf"; + break; + default: + variant = ""; + break; + } + + if (board_id == 0) { + snprintf(fwname, max_size, "qca/nvm_usb_%08x%s.bin", + rom_version, variant); + } else { + snprintf(fwname, max_size, "qca/nvm_usb_%08x%s_%04x.bin", + rom_version, variant, board_id); + } + } else { + snprintf(fwname, max_size, "qca/nvm_usb_%08x.bin", + rom_version); + } + +} + static int btusb_setup_qca_load_nvm(struct hci_dev *hdev, struct qca_version *ver, const struct qca_device_info *info) @@ -3069,20 +3107,7 @@ static int btusb_setup_qca_load_nvm(struct hci_dev *hdev, char fwname[64]; int err; - if (((ver->flag >> 8) & 0xff) == QCA_FLAG_MULTI_NVM) { - /* if boardid equal 0, use default nvm without surfix */ - if (le16_to_cpu(ver->board_id) == 0x0) { - snprintf(fwname, sizeof(fwname), "qca/nvm_usb_%08x.bin", - le32_to_cpu(ver->rom_version)); - } else { - snprintf(fwname, sizeof(fwname), "qca/nvm_usb_%08x_%04x.bin", - le32_to_cpu(ver->rom_version), - le16_to_cpu(ver->board_id)); - } - } else { - snprintf(fwname, sizeof(fwname), "qca/nvm_usb_%08x.bin", - le32_to_cpu(ver->rom_version)); - } + btusb_generate_qca_nvm_name(fwname, sizeof(fwname), ver); err = request_firmware(&fw, fwname, &hdev->dev); if (err) { From 82c434c103408842a87404e873992b7698b6df2b Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Thu, 28 Oct 2021 10:46:28 +0300 Subject: [PATCH 0055/2048] ath11k: set correct NL80211_FEATURE_DYNAMIC_SMPS for WCN6855 Commit 6f4d70308e5e ("ath11k: support SMPS configuration for 6 GHz") changed "if (ht_cap & WMI_HT_CAP_DYNAMIC_SMPS)" to "if (ht_cap & WMI_HT_CAP_DYNAMIC_SMPS || ar->supports_6ghz)" which means NL80211_FEATURE_DYNAMIC_SMPS is enabled for all chips which support 6 GHz. However, WCN6855 supports 6 GHz but it does not support feature NL80211_FEATURE_DYNAMIC_SMPS, and this can lead to MU-MIMO test failures for WCN6855. Disable NL80211_FEATURE_DYNAMIC_SMPS for WCN6855 since its ht_cap does not support WMI_HT_CAP_DYNAMIC_SMPS. Enable the feature only on QCN9074 as that's the only other device supporting 6 GHz band. Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1 Signed-off-by: Wen Gong Signed-off-by: Jouni Malinen Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20210914163726.38604-3-jouni@codeaurora.org --- drivers/net/wireless/ath/ath11k/core.c | 5 +++++ drivers/net/wireless/ath/ath11k/hw.h | 1 + drivers/net/wireless/ath/ath11k/mac.c | 3 ++- 3 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index b5a2af3ffc3e..7ee2ccc49c74 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -82,6 +82,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .fix_l1ss = true, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, + .supports_dynamic_smps_6ghz = false, }, { .hw_rev = ATH11K_HW_IPQ6018_HW10, @@ -131,6 +132,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .fix_l1ss = true, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, + .supports_dynamic_smps_6ghz = false, }, { .name = "qca6390 hw2.0", @@ -179,6 +181,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .fix_l1ss = true, .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, + .supports_dynamic_smps_6ghz = false, }, { .name = "qcn9074 hw1.0", @@ -227,6 +230,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .fix_l1ss = true, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, + .supports_dynamic_smps_6ghz = true, }, { .name = "wcn6855 hw2.0", @@ -275,6 +279,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .fix_l1ss = false, .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, + .supports_dynamic_smps_6ghz = false, }, }; diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h index 19223d36846e..6dcac596e3fe 100644 --- a/drivers/net/wireless/ath/ath11k/hw.h +++ b/drivers/net/wireless/ath/ath11k/hw.h @@ -176,6 +176,7 @@ struct ath11k_hw_params { bool fix_l1ss; u8 max_tx_ring; const struct ath11k_hw_hal_params *hal_params; + bool supports_dynamic_smps_6ghz; }; struct ath11k_hw_ops { diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 1cc55602787b..50a6f88115cd 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -7672,7 +7672,8 @@ static int __ath11k_mac_register(struct ath11k *ar) * for each band for a dual band capable radio. It will be tricky to * handle it when the ht capability different for each band. */ - if (ht_cap & WMI_HT_CAP_DYNAMIC_SMPS || ar->supports_6ghz) + if (ht_cap & WMI_HT_CAP_DYNAMIC_SMPS || + (ar->supports_6ghz && ab->hw_params.supports_dynamic_smps_6ghz)) ar->hw->wiphy->features |= NL80211_FEATURE_DYNAMIC_SMPS; ar->hw->wiphy->max_scan_ssids = WLAN_SCAN_PARAMS_MAX_SSID; From 27deb0f1570b0dbf465443857ce10ac6443d141d Mon Sep 17 00:00:00 2001 From: Christian Lamparter Date: Thu, 28 Oct 2021 10:46:28 +0300 Subject: [PATCH 0056/2048] ath10k: fetch (pre-)calibration data via nvmem subsystem MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ATH10K chips are used it wide range of routers, accesspoints, range extenders, network appliances. On these embedded devices, calibration data is often stored on the main system's flash and was out of reach for the driver. To bridge this gap, ath10k is getting extended to pull the (pre-)calibration data through nvmem subsystem. To do this, a nvmem-cell containing the information can either be specified in the platform data or via device-tree. Tested with: Netgear EX6150v2 (IPQ4018 - pre-calibration method) TP-Link Archer C7 v2 (QCA9880v2 - old calibration method) Cc: Robert Marko Cc: Thibaut VARÈNE Signed-off-by: Christian Lamparter Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211016234609.1568317-1-chunkeey@gmail.com --- drivers/net/wireless/ath/ath10k/core.c | 64 +++++++++++++++++++++++++- drivers/net/wireless/ath/ath10k/core.h | 6 +++ 2 files changed, 68 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath10k/core.c b/drivers/net/wireless/ath/ath10k/core.c index 5935e0973d14..72a366aa9f60 100644 --- a/drivers/net/wireless/ath/ath10k/core.c +++ b/drivers/net/wireless/ath/ath10k/core.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include "core.h" @@ -935,7 +936,8 @@ static int ath10k_core_get_board_id_from_otp(struct ath10k *ar) } if (ar->cal_mode == ATH10K_PRE_CAL_MODE_DT || - ar->cal_mode == ATH10K_PRE_CAL_MODE_FILE) + ar->cal_mode == ATH10K_PRE_CAL_MODE_FILE || + ar->cal_mode == ATH10K_PRE_CAL_MODE_NVMEM) bmi_board_id_param = BMI_PARAM_GET_FLASH_BOARD_ID; else bmi_board_id_param = BMI_PARAM_GET_EEPROM_BOARD_ID; @@ -1726,7 +1728,8 @@ static int ath10k_download_and_run_otp(struct ath10k *ar) /* As of now pre-cal is valid for 10_4 variants */ if (ar->cal_mode == ATH10K_PRE_CAL_MODE_DT || - ar->cal_mode == ATH10K_PRE_CAL_MODE_FILE) + ar->cal_mode == ATH10K_PRE_CAL_MODE_FILE || + ar->cal_mode == ATH10K_PRE_CAL_MODE_NVMEM) bmi_otp_exe_param = BMI_PARAM_FLASH_SECTION_ALL; ret = ath10k_bmi_execute(ar, address, bmi_otp_exe_param, &result); @@ -1853,6 +1856,39 @@ out_free: return ret; } +static int ath10k_download_cal_nvmem(struct ath10k *ar, const char *cell_name) +{ + struct nvmem_cell *cell; + void *buf; + size_t len; + int ret; + + cell = devm_nvmem_cell_get(ar->dev, cell_name); + if (IS_ERR(cell)) { + ret = PTR_ERR(cell); + return ret; + } + + buf = nvmem_cell_read(cell, &len); + if (IS_ERR(buf)) + return PTR_ERR(buf); + + if (ar->hw_params.cal_data_len != len) { + kfree(buf); + ath10k_warn(ar, "invalid calibration data length in nvmem-cell '%s': %zu != %u\n", + cell_name, len, ar->hw_params.cal_data_len); + return -EMSGSIZE; + } + + ret = ath10k_download_board_data(ar, buf, len); + kfree(buf); + if (ret) + ath10k_warn(ar, "failed to download calibration data from nvmem-cell '%s': %d\n", + cell_name, ret); + + return ret; +} + int ath10k_core_fetch_firmware_api_n(struct ath10k *ar, const char *name, struct ath10k_fw_file *fw_file) { @@ -2087,6 +2123,18 @@ static int ath10k_core_pre_cal_download(struct ath10k *ar) { int ret; + ret = ath10k_download_cal_nvmem(ar, "pre-calibration"); + if (ret == 0) { + ar->cal_mode = ATH10K_PRE_CAL_MODE_NVMEM; + goto success; + } else if (ret == -EPROBE_DEFER) { + return ret; + } + + ath10k_dbg(ar, ATH10K_DBG_BOOT, + "boot did not find a pre-calibration nvmem-cell, try file next: %d\n", + ret); + ret = ath10k_download_cal_file(ar, ar->pre_cal_file); if (ret == 0) { ar->cal_mode = ATH10K_PRE_CAL_MODE_FILE; @@ -2153,6 +2201,18 @@ static int ath10k_download_cal_data(struct ath10k *ar) "pre cal download procedure failed, try cal file: %d\n", ret); + ret = ath10k_download_cal_nvmem(ar, "calibration"); + if (ret == 0) { + ar->cal_mode = ATH10K_CAL_MODE_NVMEM; + goto done; + } else if (ret == -EPROBE_DEFER) { + return ret; + } + + ath10k_dbg(ar, ATH10K_DBG_BOOT, + "boot did not find a calibration nvmem-cell, try file next: %d\n", + ret); + ret = ath10k_download_cal_file(ar, ar->cal_file); if (ret == 0) { ar->cal_mode = ATH10K_CAL_MODE_FILE; diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h index 5aeff2d9f6cf..9f6680b3be0a 100644 --- a/drivers/net/wireless/ath/ath10k/core.h +++ b/drivers/net/wireless/ath/ath10k/core.h @@ -877,8 +877,10 @@ enum ath10k_cal_mode { ATH10K_CAL_MODE_FILE, ATH10K_CAL_MODE_OTP, ATH10K_CAL_MODE_DT, + ATH10K_CAL_MODE_NVMEM, ATH10K_PRE_CAL_MODE_FILE, ATH10K_PRE_CAL_MODE_DT, + ATH10K_PRE_CAL_MODE_NVMEM, ATH10K_CAL_MODE_EEPROM, }; @@ -898,10 +900,14 @@ static inline const char *ath10k_cal_mode_str(enum ath10k_cal_mode mode) return "otp"; case ATH10K_CAL_MODE_DT: return "dt"; + case ATH10K_CAL_MODE_NVMEM: + return "nvmem"; case ATH10K_PRE_CAL_MODE_FILE: return "pre-cal-file"; case ATH10K_PRE_CAL_MODE_DT: return "pre-cal-dt"; + case ATH10K_PRE_CAL_MODE_NVMEM: + return "pre-cal-nvmem"; case ATH10K_CAL_MODE_EEPROM: return "eeprom"; } From df008741dd62bd3bc8733fe568415fb01b3e65c5 Mon Sep 17 00:00:00 2001 From: Benjamin Li Date: Wed, 27 Oct 2021 10:03:03 -0700 Subject: [PATCH 0057/2048] wcn36xx: add debug prints for sw_scan start/complete Add some MAC debug prints for more easily demarcating a software scan when parsing logs. Signed-off-by: Benjamin Li Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211027170306.555535-2-benl@squareup.com --- drivers/net/wireless/ath/wcn36xx/main.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c index b04533bbc3a4..06558f5cdeb9 100644 --- a/drivers/net/wireless/ath/wcn36xx/main.c +++ b/drivers/net/wireless/ath/wcn36xx/main.c @@ -707,6 +707,8 @@ static void wcn36xx_sw_scan_start(struct ieee80211_hw *hw, struct wcn36xx *wcn = hw->priv; struct wcn36xx_vif *vif_priv = wcn36xx_vif_to_priv(vif); + wcn36xx_dbg(WCN36XX_DBG_MAC, "sw_scan_start"); + wcn->sw_scan = true; wcn->sw_scan_vif = vif; wcn->sw_scan_channel = 0; @@ -721,6 +723,8 @@ static void wcn36xx_sw_scan_complete(struct ieee80211_hw *hw, { struct wcn36xx *wcn = hw->priv; + wcn36xx_dbg(WCN36XX_DBG_MAC, "sw_scan_complete"); + /* ensure that any scan session is finished */ wcn36xx_smd_finish_scan(wcn, HAL_SYS_MODE_SCAN, wcn->sw_scan_vif); wcn->sw_scan = false; From f02e1cc2a84693a649d94a7584291d88f31fd5fa Mon Sep 17 00:00:00 2001 From: Benjamin Li Date: Wed, 27 Oct 2021 10:03:04 -0700 Subject: [PATCH 0058/2048] wcn36xx: implement flush op to speed up connected scan Without ieee80211_ops->flush implemented to empty HW queues, mac80211 will do a 100ms dead wait after stopping SW queues, before leaving the operating channel to resume a software connected scan[1]. (see ieee80211_scan_state_resume) This wait is correctly included in the calculation for whether or not we've exceeded max off-channel time, as it occurs after sending the null frame with PS bit set. Thus, with 125 ms max off-channel time we only have 25 ms of scan time, which technically isn't even enough to scan one channel (although mac80211 always scans at least one channel per off- channel window). Moreover, for passive probes we end up spending at least 100 ms + 111 ms (IEEE80211_PASSIVE_CHANNEL_TIME) "off-channel"[2], which exceeds the listen interval of 200 ms that we provide in our association request frame. That's technically out-of-spec. [1]: Until recently, wcn36xx performed software (rather than FW-offloaded) scanning when 5GHz channels are requested. This apparent limitation is now resolved -- see commit 1395f8a6a4d5 ("wcn36xx: Enable hardware scan offload for 5Ghz band"). [2]: in quotes because about 100 ms of it is still on-channel but with PS set Signed-off-by: Benjamin Li Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211027170306.555535-3-benl@squareup.com --- drivers/net/wireless/ath/wcn36xx/dxe.c | 47 +++++++++++++++++++++++++ drivers/net/wireless/ath/wcn36xx/dxe.h | 1 + drivers/net/wireless/ath/wcn36xx/main.c | 11 ++++++ 3 files changed, 59 insertions(+) diff --git a/drivers/net/wireless/ath/wcn36xx/dxe.c b/drivers/net/wireless/ath/wcn36xx/dxe.c index aff04ef66266..fd627c9f3d40 100644 --- a/drivers/net/wireless/ath/wcn36xx/dxe.c +++ b/drivers/net/wireless/ath/wcn36xx/dxe.c @@ -834,6 +834,53 @@ unlock: return ret; } +static bool _wcn36xx_dxe_tx_channel_is_empty(struct wcn36xx_dxe_ch *ch) +{ + unsigned long flags; + struct wcn36xx_dxe_ctl *ctl_bd_start, *ctl_skb_start; + struct wcn36xx_dxe_ctl *ctl_bd, *ctl_skb; + bool ret = true; + + spin_lock_irqsave(&ch->lock, flags); + + /* Loop through ring buffer looking for nonempty entries. */ + ctl_bd_start = ch->head_blk_ctl; + ctl_bd = ctl_bd_start; + ctl_skb_start = ctl_bd_start->next; + ctl_skb = ctl_skb_start; + do { + if (ctl_skb->skb) { + ret = false; + goto unlock; + } + ctl_bd = ctl_skb->next; + ctl_skb = ctl_bd->next; + } while (ctl_skb != ctl_skb_start); + +unlock: + spin_unlock_irqrestore(&ch->lock, flags); + return ret; +} + +int wcn36xx_dxe_tx_flush(struct wcn36xx *wcn) +{ + int i = 0; + + /* Called with mac80211 queues stopped. Wait for empty HW queues. */ + do { + if (_wcn36xx_dxe_tx_channel_is_empty(&wcn->dxe_tx_l_ch) && + _wcn36xx_dxe_tx_channel_is_empty(&wcn->dxe_tx_h_ch)) { + return 0; + } + /* This ieee80211_ops callback is specifically allowed to + * sleep. + */ + usleep_range(1000, 1100); + } while (++i < 100); + + return -EBUSY; +} + int wcn36xx_dxe_init(struct wcn36xx *wcn) { int reg_data = 0, ret; diff --git a/drivers/net/wireless/ath/wcn36xx/dxe.h b/drivers/net/wireless/ath/wcn36xx/dxe.h index 31b81b7547a3..26a31edf52e9 100644 --- a/drivers/net/wireless/ath/wcn36xx/dxe.h +++ b/drivers/net/wireless/ath/wcn36xx/dxe.h @@ -466,5 +466,6 @@ int wcn36xx_dxe_tx_frame(struct wcn36xx *wcn, struct wcn36xx_tx_bd *bd, struct sk_buff *skb, bool is_low); +int wcn36xx_dxe_tx_flush(struct wcn36xx *wcn); void wcn36xx_dxe_tx_ack_ind(struct wcn36xx *wcn, u32 status); #endif /* _DXE_H_ */ diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c index 06558f5cdeb9..a8b841511484 100644 --- a/drivers/net/wireless/ath/wcn36xx/main.c +++ b/drivers/net/wireless/ath/wcn36xx/main.c @@ -1281,6 +1281,16 @@ static void wcn36xx_ipv6_addr_change(struct ieee80211_hw *hw, } #endif +static void wcn36xx_flush(struct ieee80211_hw *hw, struct ieee80211_vif *vif, + u32 queues, bool drop) +{ + struct wcn36xx *wcn = hw->priv; + + if (wcn36xx_dxe_tx_flush(wcn)) { + wcn36xx_err("Failed to flush hardware tx queues\n"); + } +} + static const struct ieee80211_ops wcn36xx_ops = { .start = wcn36xx_start, .stop = wcn36xx_stop, @@ -1308,6 +1318,7 @@ static const struct ieee80211_ops wcn36xx_ops = { #if IS_ENABLED(CONFIG_IPV6) .ipv6_addr_change = wcn36xx_ipv6_addr_change, #endif + .flush = wcn36xx_flush, CFG80211_TESTMODE_CMD(wcn36xx_tm_cmd) }; From 8f1ba8b0ee2679f0b3d22d2a5c1bc70c436fd872 Mon Sep 17 00:00:00 2001 From: Benjamin Li Date: Wed, 27 Oct 2021 10:03:05 -0700 Subject: [PATCH 0059/2048] wcn36xx: ensure pairing of init_scan/finish_scan and start_scan/end_scan An SMD capture from the downstream prima driver on WCN3680B shows the following command sequence for connected scans: - init_scan_req - start_scan_req, channel 1 - end_scan_req, channel 1 - start_scan_req, channel 2 - ... - end_scan_req, channel 3 - finish_scan_req - init_scan_req - start_scan_req, channel 4 - ... - end_scan_req, channel 6 - finish_scan_req - ... - end_scan_req, channel 165 - finish_scan_req Upstream currently never calls wcn36xx_smd_end_scan, and in some cases[1] still sends finish_scan_req twice in a row or before init_scan_req. A typical connected scan looks like this: - init_scan_req - start_scan_req, channel 1 - finish_scan_req - init_scan_req - start_scan_req, channel 2 - ... - start_scan_req, channel 165 - finish_scan_req - finish_scan_req This patch cleans up scanning so that init/finish and start/end are always paired together and correctly nested. - init_scan_req - start_scan_req, channel 1 - end_scan_req, channel 1 - finish_scan_req - init_scan_req - start_scan_req, channel 2 - end_scan_req, channel 2 - ... - start_scan_req, channel 165 - end_scan_req, channel 165 - finish_scan_req Note that upstream will not do batching of 3 active-probe scans before returning to the operating channel, and this patch does not change that. To match downstream in this aspect, adjust IEEE80211_PROBE_DELAY and/or the 125ms max off-channel time in ieee80211_scan_state_decision. [1]: commit d195d7aac09b ("wcn36xx: Ensure finish scan is not requested before start scan") addressed one case of finish_scan_req being sent without a preceding init_scan_req (the case of the operating channel coinciding with the first scan channel); two other cases are: 1) if SW scan is started and aborted immediately, without scanning any channels, we send a finish_scan_req without ever sending init_scan_req, and 2) as SW scan logic always returns us to the operating channel before calling wcn36xx_sw_scan_complete, finish_scan_req is always sent twice at the end of a SW scan Fixes: 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware") Signed-off-by: Benjamin Li Tested-by: Bryan O'Donoghue Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211027170306.555535-4-benl@squareup.com --- drivers/net/wireless/ath/wcn36xx/main.c | 34 +++++++++++++++++----- drivers/net/wireless/ath/wcn36xx/smd.c | 4 +++ drivers/net/wireless/ath/wcn36xx/wcn36xx.h | 1 + 3 files changed, 32 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c index a8b841511484..0110f5062296 100644 --- a/drivers/net/wireless/ath/wcn36xx/main.c +++ b/drivers/net/wireless/ath/wcn36xx/main.c @@ -402,6 +402,7 @@ static void wcn36xx_change_opchannel(struct wcn36xx *wcn, int ch) static int wcn36xx_config(struct ieee80211_hw *hw, u32 changed) { struct wcn36xx *wcn = hw->priv; + int ret; wcn36xx_dbg(WCN36XX_DBG_MAC, "mac config changed 0x%08x\n", changed); @@ -417,17 +418,31 @@ static int wcn36xx_config(struct ieee80211_hw *hw, u32 changed) * want to receive/transmit regular data packets, then * simply stop the scan session and exit PS mode. */ - wcn36xx_smd_finish_scan(wcn, HAL_SYS_MODE_SCAN, - wcn->sw_scan_vif); - wcn->sw_scan_channel = 0; + if (wcn->sw_scan_channel) + wcn36xx_smd_end_scan(wcn, wcn->sw_scan_channel); + if (wcn->sw_scan_init) { + wcn36xx_smd_finish_scan(wcn, HAL_SYS_MODE_SCAN, + wcn->sw_scan_vif); + } } else if (wcn->sw_scan) { /* A scan is ongoing, do not change the operating * channel, but start a scan session on the channel. */ - wcn36xx_smd_init_scan(wcn, HAL_SYS_MODE_SCAN, - wcn->sw_scan_vif); + if (wcn->sw_scan_channel) + wcn36xx_smd_end_scan(wcn, wcn->sw_scan_channel); + if (!wcn->sw_scan_init) { + /* This can fail if we are unable to notify the + * operating channel. + */ + ret = wcn36xx_smd_init_scan(wcn, + HAL_SYS_MODE_SCAN, + wcn->sw_scan_vif); + if (ret) { + mutex_unlock(&wcn->conf_mutex); + return -EIO; + } + } wcn36xx_smd_start_scan(wcn, ch); - wcn->sw_scan_channel = ch; } else { wcn36xx_change_opchannel(wcn, ch); } @@ -726,7 +741,12 @@ static void wcn36xx_sw_scan_complete(struct ieee80211_hw *hw, wcn36xx_dbg(WCN36XX_DBG_MAC, "sw_scan_complete"); /* ensure that any scan session is finished */ - wcn36xx_smd_finish_scan(wcn, HAL_SYS_MODE_SCAN, wcn->sw_scan_vif); + if (wcn->sw_scan_channel) + wcn36xx_smd_end_scan(wcn, wcn->sw_scan_channel); + if (wcn->sw_scan_init) { + wcn36xx_smd_finish_scan(wcn, HAL_SYS_MODE_SCAN, + wcn->sw_scan_vif); + } wcn->sw_scan = false; wcn->sw_scan_opchannel = 0; } diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c b/drivers/net/wireless/ath/wcn36xx/smd.c index ed45e2cf039b..e44506e4f1d5 100644 --- a/drivers/net/wireless/ath/wcn36xx/smd.c +++ b/drivers/net/wireless/ath/wcn36xx/smd.c @@ -722,6 +722,7 @@ int wcn36xx_smd_init_scan(struct wcn36xx *wcn, enum wcn36xx_hal_sys_mode mode, wcn36xx_err("hal_init_scan response failed err=%d\n", ret); goto out; } + wcn->sw_scan_init = true; out: mutex_unlock(&wcn->hal_mutex); return ret; @@ -752,6 +753,7 @@ int wcn36xx_smd_start_scan(struct wcn36xx *wcn, u8 scan_channel) wcn36xx_err("hal_start_scan response failed err=%d\n", ret); goto out; } + wcn->sw_scan_channel = scan_channel; out: mutex_unlock(&wcn->hal_mutex); return ret; @@ -782,6 +784,7 @@ int wcn36xx_smd_end_scan(struct wcn36xx *wcn, u8 scan_channel) wcn36xx_err("hal_end_scan response failed err=%d\n", ret); goto out; } + wcn->sw_scan_channel = 0; out: mutex_unlock(&wcn->hal_mutex); return ret; @@ -823,6 +826,7 @@ int wcn36xx_smd_finish_scan(struct wcn36xx *wcn, wcn36xx_err("hal_finish_scan response failed err=%d\n", ret); goto out; } + wcn->sw_scan_init = false; out: mutex_unlock(&wcn->hal_mutex); return ret; diff --git a/drivers/net/wireless/ath/wcn36xx/wcn36xx.h b/drivers/net/wireless/ath/wcn36xx/wcn36xx.h index 1c8d918137da..fbd0558c2c19 100644 --- a/drivers/net/wireless/ath/wcn36xx/wcn36xx.h +++ b/drivers/net/wireless/ath/wcn36xx/wcn36xx.h @@ -248,6 +248,7 @@ struct wcn36xx { struct cfg80211_scan_request *scan_req; bool sw_scan; u8 sw_scan_opchannel; + bool sw_scan_init; u8 sw_scan_channel; struct ieee80211_vif *sw_scan_vif; struct mutex scan_lock; From 588b45c88ae130fe373a8c50edaf54735c3f4fe3 Mon Sep 17 00:00:00 2001 From: Bryan O'Donoghue Date: Thu, 28 Oct 2021 00:25:29 +0100 Subject: [PATCH 0060/2048] wcn36xx: Indicate beacon not connection loss on MISSED_BEACON_IND Firmware can trigger a missed beacon indication, this is not the same as a lost signal. Flag to Linux the missed beacon and let the WiFi stack decide for itself if the link is up or down by sending its own probe to determine this. We should only be signalling the link is lost when the firmware indicates Fixes: 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware") Signed-off-by: Bryan O'Donoghue Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211027232529.657764-1-bryan.odonoghue@linaro.org --- drivers/net/wireless/ath/wcn36xx/smd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c b/drivers/net/wireless/ath/wcn36xx/smd.c index e44506e4f1d5..d3285a504429 100644 --- a/drivers/net/wireless/ath/wcn36xx/smd.c +++ b/drivers/net/wireless/ath/wcn36xx/smd.c @@ -2736,7 +2736,7 @@ static int wcn36xx_smd_missed_beacon_ind(struct wcn36xx *wcn, wcn36xx_dbg(WCN36XX_DBG_HAL, "beacon missed bss_index %d\n", tmp->bss_index); vif = wcn36xx_priv_to_vif(tmp); - ieee80211_connection_loss(vif); + ieee80211_beacon_loss(vif); } return 0; } @@ -2751,7 +2751,7 @@ static int wcn36xx_smd_missed_beacon_ind(struct wcn36xx *wcn, wcn36xx_dbg(WCN36XX_DBG_HAL, "beacon missed bss_index %d\n", rsp->bss_index); vif = wcn36xx_priv_to_vif(tmp); - ieee80211_connection_loss(vif); + ieee80211_beacon_loss(vif); return 0; } } From 22f9ba7fee107903ec9a4f8f150ced196f6a1615 Mon Sep 17 00:00:00 2001 From: Yang Guang Date: Thu, 28 Oct 2021 01:04:51 +0000 Subject: [PATCH 0061/2048] ath9k: use swap() to make code cleaner Using swap() make it more readable. Reported-by: Zeal Robot Signed-off-by: Yang Guang Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211028010451.7754-1-yang.guang5@zte.com.cn --- drivers/net/wireless/ath/ath9k/ar9003_calib.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/ar9003_calib.c b/drivers/net/wireless/ath/ath9k/ar9003_calib.c index 7e27a06e5df1..dc24da1ff00b 100644 --- a/drivers/net/wireless/ath/ath9k/ar9003_calib.c +++ b/drivers/net/wireless/ath/ath9k/ar9003_calib.c @@ -1005,24 +1005,20 @@ static void __ar955x_tx_iq_cal_sort(struct ath_hw *ah, int i, int nmeasurement) { struct ath_common *common = ath9k_hw_common(ah); - int im, ix, iy, temp; + int im, ix, iy; for (im = 0; im < nmeasurement; im++) { for (ix = 0; ix < MAXIQCAL - 1; ix++) { for (iy = ix + 1; iy <= MAXIQCAL - 1; iy++) { if (coeff->mag_coeff[i][im][iy] < coeff->mag_coeff[i][im][ix]) { - temp = coeff->mag_coeff[i][im][ix]; - coeff->mag_coeff[i][im][ix] = - coeff->mag_coeff[i][im][iy]; - coeff->mag_coeff[i][im][iy] = temp; + swap(coeff->mag_coeff[i][im][ix], + coeff->mag_coeff[i][im][iy]); } if (coeff->phs_coeff[i][im][iy] < coeff->phs_coeff[i][im][ix]) { - temp = coeff->phs_coeff[i][im][ix]; - coeff->phs_coeff[i][im][ix] = - coeff->phs_coeff[i][im][iy]; - coeff->phs_coeff[i][im][iy] = temp; + swap(coeff->phs_coeff[i][im][ix], + coeff->phs_coeff[i][im][iy]); } } } From 27592ca1fadf2d2a44bb24ecc5da604fd840d633 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Mon, 1 Nov 2021 09:52:40 -0700 Subject: [PATCH 0062/2048] Bluetooth: hci_sync: Fix missing static warnings This fixes the following warnings: >> net/bluetooth/hci_sync.c:1333:5: warning: no previous prototype for 'hci_scan_disable_sync' [-Wmissing-prototypes] 1333 | int hci_scan_disable_sync(struct hci_dev *hdev, bool rpa_le_conn) | ^~~~~~~~~~~~~~~~~~~~~ >> net/bluetooth/hci_sync.c:1762:5: warning: no previous prototype for 'hci_passive_scan_sync' [-Wmissing-prototypes] 1762 | int hci_passive_scan_sync(struct hci_dev *hdev) | ^~~~~~~~~~~~~~~~~~~~~ Fixes: e8907f76544f ("Bluetooth: hci_sync: Make use of hci_cmd_sync_queue set 3") Reported-by: kernel test robot Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_sync.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index e3f44e621b7f..fb7ecf24b89a 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1397,7 +1397,7 @@ static int hci_le_set_addr_resolution_enable_sync(struct hci_dev *hdev, u8 val) sizeof(val), &val, HCI_CMD_TIMEOUT); } -int hci_scan_disable_sync(struct hci_dev *hdev) +static int hci_scan_disable_sync(struct hci_dev *hdev) { int err; @@ -1968,7 +1968,7 @@ static int hci_start_scan_sync(struct hci_dev *hdev, u8 type, u16 interval, return hci_le_set_scan_enable_sync(hdev, LE_SCAN_ENABLE, filter_dup); } -int hci_passive_scan_sync(struct hci_dev *hdev) +static int hci_passive_scan_sync(struct hci_dev *hdev) { u8 own_addr_type; u8 filter_policy; From f16a491c65d9eb19398b25aefc10c2d3313d17b3 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Mon, 1 Nov 2021 15:14:41 -0700 Subject: [PATCH 0063/2048] Bluetooth: hci_sync: Fix not setting adv set duration 10bbffa3e88e attempted to fix the use of rotation duration as advertising duration but it didn't change the if condition which still uses the duration instead of the timeout. Fixes: 10bbffa3e88e ("Bluetooth: Fix using advertising instance duration as timeout") Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_sync.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index fb7ecf24b89a..b794605dc882 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -895,7 +895,7 @@ int hci_enable_ext_advertising_sync(struct hci_dev *hdev, u8 instance) /* Set duration per instance since controller is responsible for * scheduling it. */ - if (adv && adv->duration) { + if (adv && adv->timeout) { u16 duration = adv->timeout * MSEC_PER_SEC; /* Time = N * 10 ms */ From 749a6c594203b0e9ac59e3d8da492a8ac6a80510 Mon Sep 17 00:00:00 2001 From: Joseph Hwang Date: Tue, 2 Nov 2021 15:19:28 +0800 Subject: [PATCH 0064/2048] Bluetooth: Add struct of reading AOSP vendor capabilities This patch adds the struct of reading AOSP vendor capabilities. New capabilities are added incrementally. Note that the version_supported octets will be used to determine whether a capability has been defined for the version. Signed-off-by: Joseph Hwang Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 1 + net/bluetooth/aosp.c | 81 +++++++++++++++++++++++++++++++- 2 files changed, 81 insertions(+), 1 deletion(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 53a8c7d3a4bf..b5f061882c10 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -603,6 +603,7 @@ struct hci_dev { #if IS_ENABLED(CONFIG_BT_AOSPEXT) bool aosp_capable; + bool aosp_quality_report; #endif int (*open)(struct hci_dev *hdev); diff --git a/net/bluetooth/aosp.c b/net/bluetooth/aosp.c index a1b7762335a5..871e889b7e17 100644 --- a/net/bluetooth/aosp.c +++ b/net/bluetooth/aosp.c @@ -8,9 +8,43 @@ #include "aosp.h" +/* Command complete parameters of LE_Get_Vendor_Capabilities_Command + * The parameters grow over time. The base version that declares the + * version_supported field is v0.95. Refer to + * https://cs.android.com/android/platform/superproject/+/master:system/ + * bt/gd/hci/controller.cc;l=452?q=le_get_vendor_capabilities_handler + */ +struct aosp_rp_le_get_vendor_capa { + /* v0.95: 15 octets */ + __u8 status; + __u8 max_advt_instances; + __u8 offloaded_resolution_of_private_address; + __le16 total_scan_results_storage; + __u8 max_irk_list_sz; + __u8 filtering_support; + __u8 max_filter; + __u8 activity_energy_info_support; + __le16 version_supported; + __le16 total_num_of_advt_tracked; + __u8 extended_scan_support; + __u8 debug_logging_supported; + /* v0.96: 16 octets */ + __u8 le_address_generation_offloading_support; + /* v0.98: 21 octets */ + __le32 a2dp_source_offload_capability_mask; + __u8 bluetooth_quality_report_support; + /* v1.00: 25 octets */ + __le32 dynamic_audio_buffer_support; +} __packed; + +#define VENDOR_CAPA_BASE_SIZE 15 +#define VENDOR_CAPA_0_98_SIZE 21 + void aosp_do_open(struct hci_dev *hdev) { struct sk_buff *skb; + struct aosp_rp_le_get_vendor_capa *rp; + u16 version_supported; if (!hdev->aosp_capable) return; @@ -20,9 +54,54 @@ void aosp_do_open(struct hci_dev *hdev) /* LE Get Vendor Capabilities Command */ skb = __hci_cmd_sync(hdev, hci_opcode_pack(0x3f, 0x153), 0, NULL, HCI_CMD_TIMEOUT); - if (IS_ERR(skb)) + if (IS_ERR(skb)) { + bt_dev_err(hdev, "AOSP get vendor capabilities (%ld)", + PTR_ERR(skb)); return; + } + /* A basic length check */ + if (skb->len < VENDOR_CAPA_BASE_SIZE) + goto length_error; + + rp = (struct aosp_rp_le_get_vendor_capa *)skb->data; + + version_supported = le16_to_cpu(rp->version_supported); + /* AOSP displays the verion number like v0.98, v1.00, etc. */ + bt_dev_info(hdev, "AOSP extensions version v%u.%02u", + version_supported >> 8, version_supported & 0xff); + + /* Do not support very old versions. */ + if (version_supported < 95) { + bt_dev_warn(hdev, "AOSP capabilities version %u too old", + version_supported); + goto done; + } + + if (version_supported < 98) { + bt_dev_warn(hdev, "AOSP quality report is not supported"); + goto done; + } + + if (skb->len < VENDOR_CAPA_0_98_SIZE) + goto length_error; + + /* The bluetooth_quality_report_support is defined at version + * v0.98. Refer to + * https://cs.android.com/android/platform/superproject/+/ + * master:system/bt/gd/hci/controller.cc;l=477 + */ + if (rp->bluetooth_quality_report_support) { + hdev->aosp_quality_report = true; + bt_dev_info(hdev, "AOSP quality report is supported"); + } + + goto done; + +length_error: + bt_dev_err(hdev, "AOSP capabilities length %d too short", skb->len); + +done: kfree_skb(skb); } From 258f56d11bbbf39df5bc5faf0119d28be528f27d Mon Sep 17 00:00:00 2001 From: Joseph Hwang Date: Tue, 2 Nov 2021 15:19:29 +0800 Subject: [PATCH 0065/2048] Bluetooth: aosp: Support AOSP Bluetooth Quality Report This patch adds the support of the AOSP Bluetooth Quality Report (BQR) events. Multiple vendors have supported the AOSP Bluetooth Quality Report. When a Bluetooth controller supports the capability, it can enable the aosp capability through hci_set_aosp_capable. Then hci_core will set up the hdev->aosp_set_quality_report callback through aosp_do_open if the controller responds to support the quality report capability. Note that Intel also supports a distinct telemetry quality report specification. Intel sets up the hdev->set_quality_report callback in the btusb driver module. Reviewed-by: Miao-chen Chou Signed-off-by: Joseph Hwang Signed-off-by: Marcel Holtmann --- net/bluetooth/aosp.c | 87 ++++++++++++++++++++++++++++++++++++++++++++ net/bluetooth/aosp.h | 13 +++++++ net/bluetooth/mgmt.c | 17 ++++++--- 3 files changed, 112 insertions(+), 5 deletions(-) diff --git a/net/bluetooth/aosp.c b/net/bluetooth/aosp.c index 871e889b7e17..432ae3aac9e3 100644 --- a/net/bluetooth/aosp.c +++ b/net/bluetooth/aosp.c @@ -112,3 +112,90 @@ void aosp_do_close(struct hci_dev *hdev) bt_dev_dbg(hdev, "Cleanup of AOSP extension"); } + +/* BQR command */ +#define BQR_OPCODE hci_opcode_pack(0x3f, 0x015e) + +/* BQR report action */ +#define REPORT_ACTION_ADD 0x00 +#define REPORT_ACTION_DELETE 0x01 +#define REPORT_ACTION_CLEAR 0x02 + +/* BQR event masks */ +#define QUALITY_MONITORING BIT(0) +#define APPRAOCHING_LSTO BIT(1) +#define A2DP_AUDIO_CHOPPY BIT(2) +#define SCO_VOICE_CHOPPY BIT(3) + +#define DEFAULT_BQR_EVENT_MASK (QUALITY_MONITORING | APPRAOCHING_LSTO | \ + A2DP_AUDIO_CHOPPY | SCO_VOICE_CHOPPY) + +/* Reporting at milliseconds so as not to stress the controller too much. + * Range: 0 ~ 65535 ms + */ +#define DEFALUT_REPORT_INTERVAL_MS 5000 + +struct aosp_bqr_cp { + __u8 report_action; + __u32 event_mask; + __u16 min_report_interval; +} __packed; + +static int enable_quality_report(struct hci_dev *hdev) +{ + struct sk_buff *skb; + struct aosp_bqr_cp cp; + + cp.report_action = REPORT_ACTION_ADD; + cp.event_mask = DEFAULT_BQR_EVENT_MASK; + cp.min_report_interval = DEFALUT_REPORT_INTERVAL_MS; + + skb = __hci_cmd_sync(hdev, BQR_OPCODE, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); + if (IS_ERR(skb)) { + bt_dev_err(hdev, "Enabling Android BQR failed (%ld)", + PTR_ERR(skb)); + return PTR_ERR(skb); + } + + kfree_skb(skb); + return 0; +} + +static int disable_quality_report(struct hci_dev *hdev) +{ + struct sk_buff *skb; + struct aosp_bqr_cp cp = { 0 }; + + cp.report_action = REPORT_ACTION_CLEAR; + + skb = __hci_cmd_sync(hdev, BQR_OPCODE, sizeof(cp), &cp, + HCI_CMD_TIMEOUT); + if (IS_ERR(skb)) { + bt_dev_err(hdev, "Disabling Android BQR failed (%ld)", + PTR_ERR(skb)); + return PTR_ERR(skb); + } + + kfree_skb(skb); + return 0; +} + +bool aosp_has_quality_report(struct hci_dev *hdev) +{ + return hdev->aosp_quality_report; +} + +int aosp_set_quality_report(struct hci_dev *hdev, bool enable) +{ + if (!aosp_has_quality_report(hdev)) + return -EOPNOTSUPP; + + bt_dev_dbg(hdev, "quality report enable %d", enable); + + /* Enable or disable the quality report feature. */ + if (enable) + return enable_quality_report(hdev); + else + return disable_quality_report(hdev); +} diff --git a/net/bluetooth/aosp.h b/net/bluetooth/aosp.h index 328fc6d39f70..2fd8886d51b2 100644 --- a/net/bluetooth/aosp.h +++ b/net/bluetooth/aosp.h @@ -8,9 +8,22 @@ void aosp_do_open(struct hci_dev *hdev); void aosp_do_close(struct hci_dev *hdev); +bool aosp_has_quality_report(struct hci_dev *hdev); +int aosp_set_quality_report(struct hci_dev *hdev, bool enable); + #else static inline void aosp_do_open(struct hci_dev *hdev) {} static inline void aosp_do_close(struct hci_dev *hdev) {} +static inline bool aosp_has_quality_report(struct hci_dev *hdev) +{ + return false; +} + +static inline int aosp_set_quality_report(struct hci_dev *hdev, bool enable) +{ + return -EOPNOTSUPP; +} + #endif diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index a7d35c138713..06384d761928 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -39,6 +39,7 @@ #include "mgmt_config.h" #include "msft.h" #include "eir.h" +#include "aosp.h" #define MGMT_VERSION 1 #define MGMT_REVISION 21 @@ -3934,7 +3935,8 @@ static int read_exp_features_info(struct sock *sk, struct hci_dev *hdev, idx++; } - if (hdev && hdev->set_quality_report) { + if (hdev && (aosp_has_quality_report(hdev) || + hdev->set_quality_report)) { if (hci_dev_test_flag(hdev, HCI_QUALITY_REPORT)) flags = BIT(0); else @@ -4198,7 +4200,7 @@ static int set_quality_report_func(struct sock *sk, struct hci_dev *hdev, val = !!cp->param[0]; changed = (val != hci_dev_test_flag(hdev, HCI_QUALITY_REPORT)); - if (!hdev->set_quality_report) { + if (!aosp_has_quality_report(hdev) && !hdev->set_quality_report) { err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_EXP_FEATURE, MGMT_STATUS_NOT_SUPPORTED); @@ -4206,13 +4208,18 @@ static int set_quality_report_func(struct sock *sk, struct hci_dev *hdev, } if (changed) { - err = hdev->set_quality_report(hdev, val); + if (hdev->set_quality_report) + err = hdev->set_quality_report(hdev, val); + else + err = aosp_set_quality_report(hdev, val); + if (err) { err = mgmt_cmd_status(sk, hdev->id, MGMT_OP_SET_EXP_FEATURE, MGMT_STATUS_FAILED); goto unlock_quality_report; } + if (val) hci_dev_set_flag(hdev, HCI_QUALITY_REPORT); else @@ -4224,8 +4231,8 @@ static int set_quality_report_func(struct sock *sk, struct hci_dev *hdev, memcpy(rp.uuid, quality_report_uuid, 16); rp.flags = cpu_to_le32(val ? BIT(0) : 0); hci_sock_set_flag(sk, HCI_MGMT_EXP_FEATURE_EVENTS); - err = mgmt_cmd_complete(sk, hdev->id, - MGMT_OP_SET_EXP_FEATURE, 0, + + err = mgmt_cmd_complete(sk, hdev->id, MGMT_OP_SET_EXP_FEATURE, 0, &rp, sizeof(rp)); if (changed) From 60f270753960291895cdd07d360c4e09c56c4596 Mon Sep 17 00:00:00 2001 From: Dave Marchevsky Date: Mon, 1 Nov 2021 15:43:54 -0700 Subject: [PATCH 0066/2048] bpftool: Migrate -1 err checks of libbpf fn calls Per [0], callers of libbpf functions with LIBBPF_STRICT_DIRECT_ERRS set should handle negative error codes of various values (e.g. -EINVAL). Migrate two callsites which were explicitly checking for -1 only to handle the new scheme. [0]: https://github.com/libbpf/libbpf/wiki/Libbpf-1.0-migration-guide#direct-error-code-returning-libbpf_strict_direct_errs Signed-off-by: Dave Marchevsky Signed-off-by: Andrii Nakryiko Acked-by: Quentin Monnet Link: https://lore.kernel.org/bpf/20211101224357.2651181-2-davemarchevsky@fb.com --- tools/bpf/bpftool/btf_dumper.c | 2 +- tools/bpf/bpftool/struct_ops.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/bpf/bpftool/btf_dumper.c b/tools/bpf/bpftool/btf_dumper.c index 9c25286a5c73..6934e8634b94 100644 --- a/tools/bpf/bpftool/btf_dumper.c +++ b/tools/bpf/bpftool/btf_dumper.c @@ -52,7 +52,7 @@ static int dump_prog_id_as_func_ptr(const struct btf_dumper *d, /* Get the bpf_prog's name. Obtain from func_info. */ prog_fd = bpf_prog_get_fd_by_id(prog_id); - if (prog_fd == -1) + if (prog_fd < 0) goto print; prog_info = bpf_program__get_prog_info_linear(prog_fd, diff --git a/tools/bpf/bpftool/struct_ops.c b/tools/bpf/bpftool/struct_ops.c index ab2d2290569a..20f803dce2e4 100644 --- a/tools/bpf/bpftool/struct_ops.c +++ b/tools/bpf/bpftool/struct_ops.c @@ -252,7 +252,7 @@ static struct res do_one_id(const char *id_str, work_func func, void *data, } fd = bpf_map_get_fd_by_id(id); - if (fd == -1) { + if (fd < 0) { p_err("can't get map by id (%lu): %s", id, strerror(errno)); res.nr_errs++; return res; From c59765cfd193382b00454b1a4424cb78d4c065e2 Mon Sep 17 00:00:00 2001 From: Dave Marchevsky Date: Mon, 1 Nov 2021 15:43:55 -0700 Subject: [PATCH 0067/2048] bpftool: Use bpf_obj_get_info_by_fd directly To prepare for impending deprecation of libbpf's bpf_program__get_prog_info_linear, migrate uses of this function to use bpf_obj_get_info_by_fd. Since the profile_target_name and dump_prog_id_as_func_ptr helpers were only looking at the first func_info, avoid grabbing the rest to save a malloc. For do_dump, add a more full-featured helper, but avoid free/realloc of buffer when possible for multi-prog dumps. Signed-off-by: Dave Marchevsky Signed-off-by: Andrii Nakryiko Acked-by: Quentin Monnet Link: https://lore.kernel.org/bpf/20211101224357.2651181-3-davemarchevsky@fb.com --- tools/bpf/bpftool/btf_dumper.c | 40 +++++---- tools/bpf/bpftool/prog.c | 159 ++++++++++++++++++++++++++------- 2 files changed, 149 insertions(+), 50 deletions(-) diff --git a/tools/bpf/bpftool/btf_dumper.c b/tools/bpf/bpftool/btf_dumper.c index 6934e8634b94..f5dddf8ef404 100644 --- a/tools/bpf/bpftool/btf_dumper.c +++ b/tools/bpf/bpftool/btf_dumper.c @@ -32,14 +32,16 @@ static int dump_prog_id_as_func_ptr(const struct btf_dumper *d, const struct btf_type *func_proto, __u32 prog_id) { - struct bpf_prog_info_linear *prog_info = NULL; const struct btf_type *func_type; + int prog_fd = -1, func_sig_len; + struct bpf_prog_info info = {}; + __u32 info_len = sizeof(info); const char *prog_name = NULL; - struct bpf_func_info *finfo; struct btf *prog_btf = NULL; - struct bpf_prog_info *info; - int prog_fd, func_sig_len; + struct bpf_func_info finfo; + __u32 finfo_rec_size; char prog_str[1024]; + int err; /* Get the ptr's func_proto */ func_sig_len = btf_dump_func(d->btf, prog_str, func_proto, NULL, 0, @@ -55,22 +57,27 @@ static int dump_prog_id_as_func_ptr(const struct btf_dumper *d, if (prog_fd < 0) goto print; - prog_info = bpf_program__get_prog_info_linear(prog_fd, - 1UL << BPF_PROG_INFO_FUNC_INFO); - close(prog_fd); - if (IS_ERR(prog_info)) { - prog_info = NULL; + err = bpf_obj_get_info_by_fd(prog_fd, &info, &info_len); + if (err) goto print; - } - info = &prog_info->info; - if (!info->btf_id || !info->nr_func_info) + if (!info.btf_id || !info.nr_func_info) goto print; - prog_btf = btf__load_from_kernel_by_id(info->btf_id); + + finfo_rec_size = info.func_info_rec_size; + memset(&info, 0, sizeof(info)); + info.nr_func_info = 1; + info.func_info_rec_size = finfo_rec_size; + info.func_info = ptr_to_u64(&finfo); + + err = bpf_obj_get_info_by_fd(prog_fd, &info, &info_len); + if (err) + goto print; + + prog_btf = btf__load_from_kernel_by_id(info.btf_id); if (libbpf_get_error(prog_btf)) goto print; - finfo = u64_to_ptr(info->func_info); - func_type = btf__type_by_id(prog_btf, finfo->type_id); + func_type = btf__type_by_id(prog_btf, finfo.type_id); if (!func_type || !btf_is_func(func_type)) goto print; @@ -92,7 +99,8 @@ print: prog_str[sizeof(prog_str) - 1] = '\0'; jsonw_string(d->jw, prog_str); btf__free(prog_btf); - free(prog_info); + if (prog_fd >= 0) + close(prog_fd); return 0; } diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index 515d22952602..dea7a49ec26e 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -100,6 +100,76 @@ static enum bpf_attach_type parse_attach_type(const char *str) return __MAX_BPF_ATTACH_TYPE; } +static int prep_prog_info(struct bpf_prog_info *const info, enum dump_mode mode, + void **info_data, size_t *const info_data_sz) +{ + struct bpf_prog_info holder = {}; + size_t needed = 0; + void *ptr; + + if (mode == DUMP_JITED) { + holder.jited_prog_len = info->jited_prog_len; + needed += info->jited_prog_len; + } else { + holder.xlated_prog_len = info->xlated_prog_len; + needed += info->xlated_prog_len; + } + + holder.nr_jited_ksyms = info->nr_jited_ksyms; + needed += info->nr_jited_ksyms * sizeof(__u64); + + holder.nr_jited_func_lens = info->nr_jited_func_lens; + needed += info->nr_jited_func_lens * sizeof(__u32); + + holder.nr_func_info = info->nr_func_info; + holder.func_info_rec_size = info->func_info_rec_size; + needed += info->nr_func_info * info->func_info_rec_size; + + holder.nr_line_info = info->nr_line_info; + holder.line_info_rec_size = info->line_info_rec_size; + needed += info->nr_line_info * info->line_info_rec_size; + + holder.nr_jited_line_info = info->nr_jited_line_info; + holder.jited_line_info_rec_size = info->jited_line_info_rec_size; + needed += info->nr_jited_line_info * info->jited_line_info_rec_size; + + if (needed > *info_data_sz) { + ptr = realloc(*info_data, needed); + if (!ptr) + return -1; + + *info_data = ptr; + *info_data_sz = needed; + } + ptr = *info_data; + + if (mode == DUMP_JITED) { + holder.jited_prog_insns = ptr_to_u64(ptr); + ptr += holder.jited_prog_len; + } else { + holder.xlated_prog_insns = ptr_to_u64(ptr); + ptr += holder.xlated_prog_len; + } + + holder.jited_ksyms = ptr_to_u64(ptr); + ptr += holder.nr_jited_ksyms * sizeof(__u64); + + holder.jited_func_lens = ptr_to_u64(ptr); + ptr += holder.nr_jited_func_lens * sizeof(__u32); + + holder.func_info = ptr_to_u64(ptr); + ptr += holder.nr_func_info * holder.func_info_rec_size; + + holder.line_info = ptr_to_u64(ptr); + ptr += holder.nr_line_info * holder.line_info_rec_size; + + holder.jited_line_info = ptr_to_u64(ptr); + ptr += holder.nr_jited_line_info * holder.jited_line_info_rec_size; + + *info = holder; + return 0; +} + static void print_boot_time(__u64 nsecs, char *buf, unsigned int size) { struct timespec real_time_ts, boot_time_ts; @@ -803,16 +873,18 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, static int do_dump(int argc, char **argv) { - struct bpf_prog_info_linear *info_linear; + struct bpf_prog_info info; + __u32 info_len = sizeof(info); + size_t info_data_sz = 0; + void *info_data = NULL; char *filepath = NULL; bool opcodes = false; bool visual = false; enum dump_mode mode; bool linum = false; - int *fds = NULL; int nb_fds, i = 0; + int *fds = NULL; int err = -1; - __u64 arrays; if (is_prefix(*argv, "jited")) { if (disasm_init()) @@ -872,43 +944,44 @@ static int do_dump(int argc, char **argv) goto exit_close; } - if (mode == DUMP_JITED) - arrays = 1UL << BPF_PROG_INFO_JITED_INSNS; - else - arrays = 1UL << BPF_PROG_INFO_XLATED_INSNS; - - arrays |= 1UL << BPF_PROG_INFO_JITED_KSYMS; - arrays |= 1UL << BPF_PROG_INFO_JITED_FUNC_LENS; - arrays |= 1UL << BPF_PROG_INFO_FUNC_INFO; - arrays |= 1UL << BPF_PROG_INFO_LINE_INFO; - arrays |= 1UL << BPF_PROG_INFO_JITED_LINE_INFO; - if (json_output && nb_fds > 1) jsonw_start_array(json_wtr); /* root array */ for (i = 0; i < nb_fds; i++) { - info_linear = bpf_program__get_prog_info_linear(fds[i], arrays); - if (IS_ERR_OR_NULL(info_linear)) { + memset(&info, 0, sizeof(info)); + + err = bpf_obj_get_info_by_fd(fds[i], &info, &info_len); + if (err) { + p_err("can't get prog info: %s", strerror(errno)); + break; + } + + err = prep_prog_info(&info, mode, &info_data, &info_data_sz); + if (err) { + p_err("can't grow prog info_data"); + break; + } + + err = bpf_obj_get_info_by_fd(fds[i], &info, &info_len); + if (err) { p_err("can't get prog info: %s", strerror(errno)); break; } if (json_output && nb_fds > 1) { jsonw_start_object(json_wtr); /* prog object */ - print_prog_header_json(&info_linear->info); + print_prog_header_json(&info); jsonw_name(json_wtr, "insns"); } else if (nb_fds > 1) { - print_prog_header_plain(&info_linear->info); + print_prog_header_plain(&info); } - err = prog_dump(&info_linear->info, mode, filepath, opcodes, - visual, linum); + err = prog_dump(&info, mode, filepath, opcodes, visual, linum); if (json_output && nb_fds > 1) jsonw_end_object(json_wtr); /* prog object */ else if (i != nb_fds - 1 && nb_fds > 1) printf("\n"); - free(info_linear); if (err) break; close(fds[i]); @@ -920,6 +993,7 @@ exit_close: for (; i < nb_fds; i++) close(fds[i]); exit_free: + free(info_data); free(fds); return err; } @@ -2016,41 +2090,58 @@ static void profile_print_readings(void) static char *profile_target_name(int tgt_fd) { - struct bpf_prog_info_linear *info_linear; - struct bpf_func_info *func_info; + struct bpf_func_info func_info; + struct bpf_prog_info info = {}; + __u32 info_len = sizeof(info); const struct btf_type *t; + __u32 func_info_rec_size; struct btf *btf = NULL; char *name = NULL; + int err; - info_linear = bpf_program__get_prog_info_linear( - tgt_fd, 1UL << BPF_PROG_INFO_FUNC_INFO); - if (IS_ERR_OR_NULL(info_linear)) { - p_err("failed to get info_linear for prog FD %d", tgt_fd); - return NULL; + err = bpf_obj_get_info_by_fd(tgt_fd, &info, &info_len); + if (err) { + p_err("failed to bpf_obj_get_info_by_fd for prog FD %d", tgt_fd); + goto out; } - if (info_linear->info.btf_id == 0) { + if (info.btf_id == 0) { p_err("prog FD %d doesn't have valid btf", tgt_fd); goto out; } - btf = btf__load_from_kernel_by_id(info_linear->info.btf_id); + func_info_rec_size = info.func_info_rec_size; + if (info.nr_func_info == 0) { + p_err("bpf_obj_get_info_by_fd for prog FD %d found 0 func_info", tgt_fd); + goto out; + } + + memset(&info, 0, sizeof(info)); + info.nr_func_info = 1; + info.func_info_rec_size = func_info_rec_size; + info.func_info = ptr_to_u64(&func_info); + + err = bpf_obj_get_info_by_fd(tgt_fd, &info, &info_len); + if (err) { + p_err("failed to get func_info for prog FD %d", tgt_fd); + goto out; + } + + btf = btf__load_from_kernel_by_id(info.btf_id); if (libbpf_get_error(btf)) { p_err("failed to load btf for prog FD %d", tgt_fd); goto out; } - func_info = u64_to_ptr(info_linear->info.func_info); - t = btf__type_by_id(btf, func_info[0].type_id); + t = btf__type_by_id(btf, func_info.type_id); if (!t) { p_err("btf %d doesn't have type %d", - info_linear->info.btf_id, func_info[0].type_id); + info.btf_id, func_info.type_id); goto out; } name = strdup(btf__name_by_offset(btf, t->name_off)); out: btf__free(btf); - free(info_linear); return name; } From 199e06fe832ddca80c2167661acab0e9dec657c4 Mon Sep 17 00:00:00 2001 From: Dave Marchevsky Date: Mon, 1 Nov 2021 15:43:56 -0700 Subject: [PATCH 0068/2048] perf: Pull in bpf_program__get_prog_info_linear To prepare for impending deprecation of libbpf's bpf_program__get_prog_info_linear, pull in the function and associated helpers into the perf codebase and migrate existing uses to the perf copy. Since libbpf's deprecated definitions will still be visible to perf, it is necessary to rename perf's definitions. Signed-off-by: Dave Marchevsky Signed-off-by: Andrii Nakryiko Acked-by: Song Liu Acked-by: Quentin Monnet Link: https://lore.kernel.org/bpf/20211101224357.2651181-4-davemarchevsky@fb.com --- .../Documentation/perf.data-file-format.txt | 2 +- tools/perf/util/Build | 1 + tools/perf/util/annotate.c | 3 +- tools/perf/util/bpf-event.c | 41 ++- tools/perf/util/bpf-event.h | 2 +- tools/perf/util/bpf-utils.c | 261 ++++++++++++++++++ tools/perf/util/bpf-utils.h | 76 +++++ tools/perf/util/bpf_counter.c | 6 +- tools/perf/util/dso.c | 1 + tools/perf/util/env.c | 1 + tools/perf/util/header.c | 13 +- 11 files changed, 374 insertions(+), 33 deletions(-) create mode 100644 tools/perf/util/bpf-utils.c create mode 100644 tools/perf/util/bpf-utils.h diff --git a/tools/perf/Documentation/perf.data-file-format.txt b/tools/perf/Documentation/perf.data-file-format.txt index e6ff8c898ada..f56d0e0fbff6 100644 --- a/tools/perf/Documentation/perf.data-file-format.txt +++ b/tools/perf/Documentation/perf.data-file-format.txt @@ -346,7 +346,7 @@ to special needs. HEADER_BPF_PROG_INFO = 25, -struct bpf_prog_info_linear, which contains detailed information about +struct perf_bpil, which contains detailed information about a BPF program, including type, id, tag, jited/xlated instructions, etc. HEADER_BPF_BTF = 26, diff --git a/tools/perf/util/Build b/tools/perf/util/Build index f2914d5bed6e..ee42da1d3639 100644 --- a/tools/perf/util/Build +++ b/tools/perf/util/Build @@ -201,6 +201,7 @@ endif perf-y += perf-hooks.o perf-$(CONFIG_LIBBPF) += bpf-event.o +perf-$(CONFIG_LIBBPF) += bpf-utils.o perf-$(CONFIG_CXX) += c++/ diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 0bae061b2d6d..f0e5a236b7e3 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -28,6 +28,7 @@ #include "evsel.h" #include "evlist.h" #include "bpf-event.h" +#include "bpf-utils.h" #include "block-range.h" #include "string2.h" #include "util/event.h" @@ -1700,12 +1701,12 @@ static int symbol__disassemble_bpf(struct symbol *sym, { struct annotation *notes = symbol__annotation(sym); struct annotation_options *opts = args->options; - struct bpf_prog_info_linear *info_linear; struct bpf_prog_linfo *prog_linfo = NULL; struct bpf_prog_info_node *info_node; int len = sym->end - sym->start; disassembler_ftype disassemble; struct map *map = args->ms.map; + struct perf_bpil *info_linear; struct disassemble_info info; struct dso *dso = map->dso; int pc = 0, count, sub_id; diff --git a/tools/perf/util/bpf-event.c b/tools/perf/util/bpf-event.c index 388847bab6d9..a27badb0a53a 100644 --- a/tools/perf/util/bpf-event.c +++ b/tools/perf/util/bpf-event.c @@ -10,6 +10,7 @@ #include #include #include "bpf-event.h" +#include "bpf-utils.h" #include "debug.h" #include "dso.h" #include "symbol.h" @@ -32,8 +33,6 @@ struct btf * __weak btf__load_from_kernel_by_id(__u32 id) return err ? ERR_PTR(err) : btf; } -#define ptr_to_u64(ptr) ((__u64)(unsigned long)(ptr)) - static int snprintf_hex(char *buf, size_t size, unsigned char *data, size_t len) { int ret = 0; @@ -48,9 +47,9 @@ static int machine__process_bpf_event_load(struct machine *machine, union perf_event *event, struct perf_sample *sample __maybe_unused) { - struct bpf_prog_info_linear *info_linear; struct bpf_prog_info_node *info_node; struct perf_env *env = machine->env; + struct perf_bpil *info_linear; int id = event->bpf.id; unsigned int i; @@ -175,9 +174,9 @@ static int perf_event__synthesize_one_bpf_prog(struct perf_session *session, { struct perf_record_ksymbol *ksymbol_event = &event->ksymbol; struct perf_record_bpf_event *bpf_event = &event->bpf; - struct bpf_prog_info_linear *info_linear; struct perf_tool *tool = session->tool; struct bpf_prog_info_node *info_node; + struct perf_bpil *info_linear; struct bpf_prog_info *info; struct btf *btf = NULL; struct perf_env *env; @@ -191,15 +190,15 @@ static int perf_event__synthesize_one_bpf_prog(struct perf_session *session, */ env = session->data ? &session->header.env : &perf_env; - arrays = 1UL << BPF_PROG_INFO_JITED_KSYMS; - arrays |= 1UL << BPF_PROG_INFO_JITED_FUNC_LENS; - arrays |= 1UL << BPF_PROG_INFO_FUNC_INFO; - arrays |= 1UL << BPF_PROG_INFO_PROG_TAGS; - arrays |= 1UL << BPF_PROG_INFO_JITED_INSNS; - arrays |= 1UL << BPF_PROG_INFO_LINE_INFO; - arrays |= 1UL << BPF_PROG_INFO_JITED_LINE_INFO; + arrays = 1UL << PERF_BPIL_JITED_KSYMS; + arrays |= 1UL << PERF_BPIL_JITED_FUNC_LENS; + arrays |= 1UL << PERF_BPIL_FUNC_INFO; + arrays |= 1UL << PERF_BPIL_PROG_TAGS; + arrays |= 1UL << PERF_BPIL_JITED_INSNS; + arrays |= 1UL << PERF_BPIL_LINE_INFO; + arrays |= 1UL << PERF_BPIL_JITED_LINE_INFO; - info_linear = bpf_program__get_prog_info_linear(fd, arrays); + info_linear = get_bpf_prog_info_linear(fd, arrays); if (IS_ERR_OR_NULL(info_linear)) { info_linear = NULL; pr_debug("%s: failed to get BPF program info. aborting\n", __func__); @@ -452,8 +451,8 @@ int perf_event__synthesize_bpf_events(struct perf_session *session, static void perf_env__add_bpf_info(struct perf_env *env, u32 id) { - struct bpf_prog_info_linear *info_linear; struct bpf_prog_info_node *info_node; + struct perf_bpil *info_linear; struct btf *btf = NULL; u64 arrays; u32 btf_id; @@ -463,15 +462,15 @@ static void perf_env__add_bpf_info(struct perf_env *env, u32 id) if (fd < 0) return; - arrays = 1UL << BPF_PROG_INFO_JITED_KSYMS; - arrays |= 1UL << BPF_PROG_INFO_JITED_FUNC_LENS; - arrays |= 1UL << BPF_PROG_INFO_FUNC_INFO; - arrays |= 1UL << BPF_PROG_INFO_PROG_TAGS; - arrays |= 1UL << BPF_PROG_INFO_JITED_INSNS; - arrays |= 1UL << BPF_PROG_INFO_LINE_INFO; - arrays |= 1UL << BPF_PROG_INFO_JITED_LINE_INFO; + arrays = 1UL << PERF_BPIL_JITED_KSYMS; + arrays |= 1UL << PERF_BPIL_JITED_FUNC_LENS; + arrays |= 1UL << PERF_BPIL_FUNC_INFO; + arrays |= 1UL << PERF_BPIL_PROG_TAGS; + arrays |= 1UL << PERF_BPIL_JITED_INSNS; + arrays |= 1UL << PERF_BPIL_LINE_INFO; + arrays |= 1UL << PERF_BPIL_JITED_LINE_INFO; - info_linear = bpf_program__get_prog_info_linear(fd, arrays); + info_linear = get_bpf_prog_info_linear(fd, arrays); if (IS_ERR_OR_NULL(info_linear)) { pr_debug("%s: failed to get BPF program info. aborting\n", __func__); goto out; diff --git a/tools/perf/util/bpf-event.h b/tools/perf/util/bpf-event.h index 68f315c3df5b..144a8a24cc69 100644 --- a/tools/perf/util/bpf-event.h +++ b/tools/perf/util/bpf-event.h @@ -19,7 +19,7 @@ struct evlist; struct target; struct bpf_prog_info_node { - struct bpf_prog_info_linear *info_linear; + struct perf_bpil *info_linear; struct rb_node rb_node; }; diff --git a/tools/perf/util/bpf-utils.c b/tools/perf/util/bpf-utils.c new file mode 100644 index 000000000000..e271e05e51bc --- /dev/null +++ b/tools/perf/util/bpf-utils.c @@ -0,0 +1,261 @@ +// SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +#ifndef _GNU_SOURCE +#define _GNU_SOURCE +#endif + +#include +#include +#include +#include +#include +#include "bpf-utils.h" +#include "debug.h" + +struct bpil_array_desc { + int array_offset; /* e.g. offset of jited_prog_insns */ + int count_offset; /* e.g. offset of jited_prog_len */ + int size_offset; /* > 0: offset of rec size, + * < 0: fix size of -size_offset + */ +}; + +static struct bpil_array_desc bpil_array_desc[] = { + [PERF_BPIL_JITED_INSNS] = { + offsetof(struct bpf_prog_info, jited_prog_insns), + offsetof(struct bpf_prog_info, jited_prog_len), + -1, + }, + [PERF_BPIL_XLATED_INSNS] = { + offsetof(struct bpf_prog_info, xlated_prog_insns), + offsetof(struct bpf_prog_info, xlated_prog_len), + -1, + }, + [PERF_BPIL_MAP_IDS] = { + offsetof(struct bpf_prog_info, map_ids), + offsetof(struct bpf_prog_info, nr_map_ids), + -(int)sizeof(__u32), + }, + [PERF_BPIL_JITED_KSYMS] = { + offsetof(struct bpf_prog_info, jited_ksyms), + offsetof(struct bpf_prog_info, nr_jited_ksyms), + -(int)sizeof(__u64), + }, + [PERF_BPIL_JITED_FUNC_LENS] = { + offsetof(struct bpf_prog_info, jited_func_lens), + offsetof(struct bpf_prog_info, nr_jited_func_lens), + -(int)sizeof(__u32), + }, + [PERF_BPIL_FUNC_INFO] = { + offsetof(struct bpf_prog_info, func_info), + offsetof(struct bpf_prog_info, nr_func_info), + offsetof(struct bpf_prog_info, func_info_rec_size), + }, + [PERF_BPIL_LINE_INFO] = { + offsetof(struct bpf_prog_info, line_info), + offsetof(struct bpf_prog_info, nr_line_info), + offsetof(struct bpf_prog_info, line_info_rec_size), + }, + [PERF_BPIL_JITED_LINE_INFO] = { + offsetof(struct bpf_prog_info, jited_line_info), + offsetof(struct bpf_prog_info, nr_jited_line_info), + offsetof(struct bpf_prog_info, jited_line_info_rec_size), + }, + [PERF_BPIL_PROG_TAGS] = { + offsetof(struct bpf_prog_info, prog_tags), + offsetof(struct bpf_prog_info, nr_prog_tags), + -(int)sizeof(__u8) * BPF_TAG_SIZE, + }, + +}; + +static __u32 bpf_prog_info_read_offset_u32(struct bpf_prog_info *info, + int offset) +{ + __u32 *array = (__u32 *)info; + + if (offset >= 0) + return array[offset / sizeof(__u32)]; + return -(int)offset; +} + +static __u64 bpf_prog_info_read_offset_u64(struct bpf_prog_info *info, + int offset) +{ + __u64 *array = (__u64 *)info; + + if (offset >= 0) + return array[offset / sizeof(__u64)]; + return -(int)offset; +} + +static void bpf_prog_info_set_offset_u32(struct bpf_prog_info *info, int offset, + __u32 val) +{ + __u32 *array = (__u32 *)info; + + if (offset >= 0) + array[offset / sizeof(__u32)] = val; +} + +static void bpf_prog_info_set_offset_u64(struct bpf_prog_info *info, int offset, + __u64 val) +{ + __u64 *array = (__u64 *)info; + + if (offset >= 0) + array[offset / sizeof(__u64)] = val; +} + +struct perf_bpil * +get_bpf_prog_info_linear(int fd, __u64 arrays) +{ + struct bpf_prog_info info = {}; + struct perf_bpil *info_linear; + __u32 info_len = sizeof(info); + __u32 data_len = 0; + int i, err; + void *ptr; + + if (arrays >> PERF_BPIL_LAST_ARRAY) + return ERR_PTR(-EINVAL); + + /* step 1: get array dimensions */ + err = bpf_obj_get_info_by_fd(fd, &info, &info_len); + if (err) { + pr_debug("can't get prog info: %s", strerror(errno)); + return ERR_PTR(-EFAULT); + } + + /* step 2: calculate total size of all arrays */ + for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) { + bool include_array = (arrays & (1UL << i)) > 0; + struct bpil_array_desc *desc; + __u32 count, size; + + desc = bpil_array_desc + i; + + /* kernel is too old to support this field */ + if (info_len < desc->array_offset + sizeof(__u32) || + info_len < desc->count_offset + sizeof(__u32) || + (desc->size_offset > 0 && info_len < (__u32)desc->size_offset)) + include_array = false; + + if (!include_array) { + arrays &= ~(1UL << i); /* clear the bit */ + continue; + } + + count = bpf_prog_info_read_offset_u32(&info, desc->count_offset); + size = bpf_prog_info_read_offset_u32(&info, desc->size_offset); + + data_len += count * size; + } + + /* step 3: allocate continuous memory */ + data_len = roundup(data_len, sizeof(__u64)); + info_linear = malloc(sizeof(struct perf_bpil) + data_len); + if (!info_linear) + return ERR_PTR(-ENOMEM); + + /* step 4: fill data to info_linear->info */ + info_linear->arrays = arrays; + memset(&info_linear->info, 0, sizeof(info)); + ptr = info_linear->data; + + for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) { + struct bpil_array_desc *desc; + __u32 count, size; + + if ((arrays & (1UL << i)) == 0) + continue; + + desc = bpil_array_desc + i; + count = bpf_prog_info_read_offset_u32(&info, desc->count_offset); + size = bpf_prog_info_read_offset_u32(&info, desc->size_offset); + bpf_prog_info_set_offset_u32(&info_linear->info, + desc->count_offset, count); + bpf_prog_info_set_offset_u32(&info_linear->info, + desc->size_offset, size); + bpf_prog_info_set_offset_u64(&info_linear->info, + desc->array_offset, + ptr_to_u64(ptr)); + ptr += count * size; + } + + /* step 5: call syscall again to get required arrays */ + err = bpf_obj_get_info_by_fd(fd, &info_linear->info, &info_len); + if (err) { + pr_debug("can't get prog info: %s", strerror(errno)); + free(info_linear); + return ERR_PTR(-EFAULT); + } + + /* step 6: verify the data */ + for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) { + struct bpil_array_desc *desc; + __u32 v1, v2; + + if ((arrays & (1UL << i)) == 0) + continue; + + desc = bpil_array_desc + i; + v1 = bpf_prog_info_read_offset_u32(&info, desc->count_offset); + v2 = bpf_prog_info_read_offset_u32(&info_linear->info, + desc->count_offset); + if (v1 != v2) + pr_warning("%s: mismatch in element count\n", __func__); + + v1 = bpf_prog_info_read_offset_u32(&info, desc->size_offset); + v2 = bpf_prog_info_read_offset_u32(&info_linear->info, + desc->size_offset); + if (v1 != v2) + pr_warning("%s: mismatch in rec size\n", __func__); + } + + /* step 7: update info_len and data_len */ + info_linear->info_len = sizeof(struct bpf_prog_info); + info_linear->data_len = data_len; + + return info_linear; +} + +void bpil_addr_to_offs(struct perf_bpil *info_linear) +{ + int i; + + for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) { + struct bpil_array_desc *desc; + __u64 addr, offs; + + if ((info_linear->arrays & (1UL << i)) == 0) + continue; + + desc = bpil_array_desc + i; + addr = bpf_prog_info_read_offset_u64(&info_linear->info, + desc->array_offset); + offs = addr - ptr_to_u64(info_linear->data); + bpf_prog_info_set_offset_u64(&info_linear->info, + desc->array_offset, offs); + } +} + +void bpil_offs_to_addr(struct perf_bpil *info_linear) +{ + int i; + + for (i = PERF_BPIL_FIRST_ARRAY; i < PERF_BPIL_LAST_ARRAY; ++i) { + struct bpil_array_desc *desc; + __u64 addr, offs; + + if ((info_linear->arrays & (1UL << i)) == 0) + continue; + + desc = bpil_array_desc + i; + offs = bpf_prog_info_read_offset_u64(&info_linear->info, + desc->array_offset); + addr = offs + ptr_to_u64(info_linear->data); + bpf_prog_info_set_offset_u64(&info_linear->info, + desc->array_offset, addr); + } +} diff --git a/tools/perf/util/bpf-utils.h b/tools/perf/util/bpf-utils.h new file mode 100644 index 000000000000..86a5055cdfad --- /dev/null +++ b/tools/perf/util/bpf-utils.h @@ -0,0 +1,76 @@ +/* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */ + +#ifndef __PERF_BPF_UTILS_H +#define __PERF_BPF_UTILS_H + +#define ptr_to_u64(ptr) ((__u64)(unsigned long)(ptr)) + +#ifdef HAVE_LIBBPF_SUPPORT + +#include + +/* + * Get bpf_prog_info in continuous memory + * + * struct bpf_prog_info has multiple arrays. The user has option to choose + * arrays to fetch from kernel. The following APIs provide an uniform way to + * fetch these data. All arrays in bpf_prog_info are stored in a single + * continuous memory region. This makes it easy to store the info in a + * file. + * + * Before writing perf_bpil to files, it is necessary to + * translate pointers in bpf_prog_info to offsets. Helper functions + * bpil_addr_to_offs() and bpil_offs_to_addr() + * are introduced to switch between pointers and offsets. + * + * Examples: + * # To fetch map_ids and prog_tags: + * __u64 arrays = (1UL << PERF_BPIL_MAP_IDS) | + * (1UL << PERF_BPIL_PROG_TAGS); + * struct perf_bpil *info_linear = + * get_bpf_prog_info_linear(fd, arrays); + * + * # To save data in file + * bpil_addr_to_offs(info_linear); + * write(f, info_linear, sizeof(*info_linear) + info_linear->data_len); + * + * # To read data from file + * read(f, info_linear, ); + * bpil_offs_to_addr(info_linear); + */ +enum perf_bpil_array_types { + PERF_BPIL_FIRST_ARRAY = 0, + PERF_BPIL_JITED_INSNS = 0, + PERF_BPIL_XLATED_INSNS, + PERF_BPIL_MAP_IDS, + PERF_BPIL_JITED_KSYMS, + PERF_BPIL_JITED_FUNC_LENS, + PERF_BPIL_FUNC_INFO, + PERF_BPIL_LINE_INFO, + PERF_BPIL_JITED_LINE_INFO, + PERF_BPIL_PROG_TAGS, + PERF_BPIL_LAST_ARRAY, +}; + +struct perf_bpil { + /* size of struct bpf_prog_info, when the tool is compiled */ + __u32 info_len; + /* total bytes allocated for data, round up to 8 bytes */ + __u32 data_len; + /* which arrays are included in data */ + __u64 arrays; + struct bpf_prog_info info; + __u8 data[]; +}; + +struct perf_bpil * +get_bpf_prog_info_linear(int fd, __u64 arrays); + +void +bpil_addr_to_offs(struct perf_bpil *info_linear); + +void +bpil_offs_to_addr(struct perf_bpil *info_linear); + +#endif /* HAVE_LIBBPF_SUPPORT */ +#endif /* __PERF_BPF_UTILS_H */ diff --git a/tools/perf/util/bpf_counter.c b/tools/perf/util/bpf_counter.c index ba0f20853651..2b04df8c5f87 100644 --- a/tools/perf/util/bpf_counter.c +++ b/tools/perf/util/bpf_counter.c @@ -13,6 +13,7 @@ #include #include "bpf_counter.h" +#include "bpf-utils.h" #include "counts.h" #include "debug.h" #include "evsel.h" @@ -61,14 +62,13 @@ static int bpf_program_profiler__destroy(struct evsel *evsel) static char *bpf_target_prog_name(int tgt_fd) { - struct bpf_prog_info_linear *info_linear; struct bpf_func_info *func_info; + struct perf_bpil *info_linear; const struct btf_type *t; struct btf *btf = NULL; char *name = NULL; - info_linear = bpf_program__get_prog_info_linear( - tgt_fd, 1UL << BPF_PROG_INFO_FUNC_INFO); + info_linear = get_bpf_prog_info_linear(tgt_fd, 1UL << PERF_BPIL_FUNC_INFO); if (IS_ERR_OR_NULL(info_linear)) { pr_debug("failed to get info_linear for prog FD %d\n", tgt_fd); return NULL; diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c index 9ed9a5676d35..9cc8a1772b4b 100644 --- a/tools/perf/util/dso.c +++ b/tools/perf/util/dso.c @@ -14,6 +14,7 @@ #ifdef HAVE_LIBBPF_SUPPORT #include #include "bpf-event.h" +#include "bpf-utils.h" #endif #include "compress.h" #include "env.h" diff --git a/tools/perf/util/env.c b/tools/perf/util/env.c index cf773f0dec38..17f1dd0680b4 100644 --- a/tools/perf/util/env.c +++ b/tools/perf/util/env.c @@ -16,6 +16,7 @@ struct perf_env perf_env; #ifdef HAVE_LIBBPF_SUPPORT #include "bpf-event.h" +#include "bpf-utils.h" #include void perf_env__insert_bpf_prog_info(struct perf_env *env, diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 1c7414f66655..56511db8fa03 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -48,6 +48,7 @@ #include "util/util.h" // perf_exe() #include "cputopo.h" #include "bpf-event.h" +#include "bpf-utils.h" #include "clockid.h" #include "pmu-hybrid.h" @@ -1006,17 +1007,17 @@ static int write_bpf_prog_info(struct feat_fd *ff, node = rb_entry(next, struct bpf_prog_info_node, rb_node); next = rb_next(&node->rb_node); - len = sizeof(struct bpf_prog_info_linear) + + len = sizeof(struct perf_bpil) + node->info_linear->data_len; /* before writing to file, translate address to offset */ - bpf_program__bpil_addr_to_offs(node->info_linear); + bpil_addr_to_offs(node->info_linear); ret = do_write(ff, node->info_linear, len); /* * translate back to address even when do_write() fails, * so that this function never changes the data. */ - bpf_program__bpil_offs_to_addr(node->info_linear); + bpil_offs_to_addr(node->info_linear); if (ret < 0) goto out; } @@ -3018,9 +3019,9 @@ static int process_dir_format(struct feat_fd *ff, #ifdef HAVE_LIBBPF_SUPPORT static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused) { - struct bpf_prog_info_linear *info_linear; struct bpf_prog_info_node *info_node; struct perf_env *env = &ff->ph->env; + struct perf_bpil *info_linear; u32 count, i; int err = -1; @@ -3049,7 +3050,7 @@ static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused) goto out; } - info_linear = malloc(sizeof(struct bpf_prog_info_linear) + + info_linear = malloc(sizeof(struct perf_bpil) + data_len); if (!info_linear) goto out; @@ -3071,7 +3072,7 @@ static int process_bpf_prog_info(struct feat_fd *ff, void *data __maybe_unused) goto out; /* after reading from file, translate offset to address */ - bpf_program__bpil_offs_to_addr(info_linear); + bpil_offs_to_addr(info_linear); info_node->info_linear = info_linear; perf_env__insert_bpf_prog_info(env, info_node); } From f5aafbc2af51931668799a9c5080c8e35cbb571f Mon Sep 17 00:00:00 2001 From: Dave Marchevsky Date: Mon, 1 Nov 2021 15:43:57 -0700 Subject: [PATCH 0069/2048] libbpf: Deprecate bpf_program__get_prog_info_linear As part of the road to libbpf 1.0, and discussed in libbpf issue tracker [0], bpf_program__get_prog_info_linear and its associated structs and helper functions should be deprecated. The functionality is too specific to the needs of 'perf', and there's little/no out-of-tree usage to preclude introduction of a more general helper in the future. [0] Closes: https://github.com/libbpf/libbpf/issues/313 Signed-off-by: Dave Marchevsky Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211101224357.2651181-5-davemarchevsky@fb.com --- tools/lib/bpf/libbpf.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 9de0f299706b..797f5f8a0e20 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -918,12 +918,15 @@ struct bpf_prog_info_linear { __u8 data[]; }; +LIBBPF_DEPRECATED_SINCE(0, 6, "use a custom linear prog_info wrapper") LIBBPF_API struct bpf_prog_info_linear * bpf_program__get_prog_info_linear(int fd, __u64 arrays); +LIBBPF_DEPRECATED_SINCE(0, 6, "use a custom linear prog_info wrapper") LIBBPF_API void bpf_program__bpil_addr_to_offs(struct bpf_prog_info_linear *info_linear); +LIBBPF_DEPRECATED_SINCE(0, 6, "use a custom linear prog_info wrapper") LIBBPF_API void bpf_program__bpil_offs_to_addr(struct bpf_prog_info_linear *info_linear); From 833907876be55205d0ec153dcd819c014404ee16 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 10:32:09 -0700 Subject: [PATCH 0070/2048] libbpf: Detect corrupted ELF symbols section Prevent divide-by-zero if ELF is corrupted and has zero sh_entsize. Reported by oss-fuzz project. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211103173213.1376990-2-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index a1bea1953df6..71f5a009010a 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -3555,7 +3555,7 @@ static int bpf_object__collect_externs(struct bpf_object *obj) scn = elf_sec_by_idx(obj, obj->efile.symbols_shndx); sh = elf_sec_hdr(obj, scn); - if (!sh) + if (!sh || sh->sh_entsize != sizeof(Elf64_Sym)) return -LIBBPF_ERRNO__FORMAT; dummy_var_btf_id = add_dummy_ksym_var(obj->btf); From 88918dc12dc357a06d8d722a684617b1c87a4654 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 10:32:10 -0700 Subject: [PATCH 0071/2048] libbpf: Improve sanity checking during BTF fix up If BTF is corrupted DATASEC's variable type ID might be incorrect. Prevent this easy to detect situation with extra NULL check. Reported by oss-fuzz project. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211103173213.1376990-3-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 71f5a009010a..f836a1936597 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -2752,13 +2752,12 @@ static int btf_fixup_datasec(struct bpf_object *obj, struct btf *btf, for (i = 0, vsi = btf_var_secinfos(t); i < vars; i++, vsi++) { t_var = btf__type_by_id(btf, vsi->type); - var = btf_var(t_var); - - if (!btf_is_var(t_var)) { + if (!t_var || !btf_is_var(t_var)) { pr_debug("Non-VAR type seen in section %s\n", name); return -EINVAL; } + var = btf_var(t_var); if (var->linkage == BTF_VAR_STATIC) continue; From 62554d52e71797eefa3fc15b54008038837bb2d4 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 10:32:11 -0700 Subject: [PATCH 0072/2048] libbpf: Validate that .BTF and .BTF.ext sections contain data .BTF and .BTF.ext ELF sections should have SHT_PROGBITS type and contain data. If they are not, ELF is invalid or corrupted, so bail out. Otherwise this can lead to data->d_buf being NULL and SIGSEGV later on. Reported by oss-fuzz project. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211103173213.1376990-4-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index f836a1936597..0dc6465271ce 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -3270,8 +3270,12 @@ static int bpf_object__elf_collect(struct bpf_object *obj) } else if (strcmp(name, MAPS_ELF_SEC) == 0) { obj->efile.btf_maps_shndx = idx; } else if (strcmp(name, BTF_ELF_SEC) == 0) { + if (sh->sh_type != SHT_PROGBITS) + return -LIBBPF_ERRNO__FORMAT; btf_data = data; } else if (strcmp(name, BTF_EXT_ELF_SEC) == 0) { + if (sh->sh_type != SHT_PROGBITS) + return -LIBBPF_ERRNO__FORMAT; btf_ext_data = data; } else if (sh->sh_type == SHT_SYMTAB) { /* already processed during the first pass above */ From 0d6988e16a12ebd41d3e268992211b0ceba44ed7 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 10:32:12 -0700 Subject: [PATCH 0073/2048] libbpf: Fix section counting logic e_shnum does include section #0 and as such is exactly the number of ELF sections that we need to allocate memory for to use section indices as array indices. Fix the off-by-one error. This is purely accounting fix, previously we were overallocating one too many array items. But no correctness errors otherwise. Fixes: 25bbbd7a444b ("libbpf: Remove assumptions about uniqueness of .rodata/.data/.bss maps") Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211103173213.1376990-5-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 0dc6465271ce..ecfea6c20042 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -3190,11 +3190,11 @@ static int bpf_object__elf_collect(struct bpf_object *obj) Elf_Scn *scn; Elf64_Shdr *sh; - /* ELF section indices are 1-based, so allocate +1 element to keep - * indexing simple. Also include 0th invalid section into sec_cnt for - * simpler and more traditional iteration logic. + /* ELF section indices are 0-based, but sec #0 is special "invalid" + * section. e_shnum does include sec #0, so e_shnum is the necessary + * size of an array to keep all the sections. */ - obj->efile.sec_cnt = 1 + obj->efile.ehdr->e_shnum; + obj->efile.sec_cnt = obj->efile.ehdr->e_shnum; obj->efile.secs = calloc(obj->efile.sec_cnt, sizeof(*obj->efile.secs)); if (!obj->efile.secs) return -ENOMEM; From b7332d2820d394dd2ac127df1567b4da597355a1 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 10:32:13 -0700 Subject: [PATCH 0074/2048] libbpf: Improve ELF relo sanitization Add few sanity checks for relocations to prevent div-by-zero and out-of-bounds array accesses in libbpf. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211103173213.1376990-6-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index ecfea6c20042..86a44735230e 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -3306,6 +3306,10 @@ static int bpf_object__elf_collect(struct bpf_object *obj) } else if (sh->sh_type == SHT_REL) { int targ_sec_idx = sh->sh_info; /* points to other section */ + if (sh->sh_entsize != sizeof(Elf64_Rel) || + targ_sec_idx >= obj->efile.sec_cnt) + return -LIBBPF_ERRNO__FORMAT; + /* Only do relo for section with exec instructions */ if (!section_have_execinstr(obj, targ_sec_idx) && strcmp(name, ".rel" STRUCT_OPS_SEC) && @@ -4025,7 +4029,7 @@ static int bpf_object__collect_prog_relos(struct bpf_object *obj, Elf64_Shdr *shdr, Elf_Data *data) { const char *relo_sec_name, *sec_name; - size_t sec_idx = shdr->sh_info; + size_t sec_idx = shdr->sh_info, sym_idx; struct bpf_program *prog; struct reloc_desc *relos; int err, i, nrels; @@ -4036,6 +4040,9 @@ bpf_object__collect_prog_relos(struct bpf_object *obj, Elf64_Shdr *shdr, Elf_Dat Elf64_Sym *sym; Elf64_Rel *rel; + if (sec_idx >= obj->efile.sec_cnt) + return -EINVAL; + scn = elf_sec_by_idx(obj, sec_idx); scn_data = elf_sec_data(obj, scn); @@ -4055,16 +4062,23 @@ bpf_object__collect_prog_relos(struct bpf_object *obj, Elf64_Shdr *shdr, Elf_Dat return -LIBBPF_ERRNO__FORMAT; } - sym = elf_sym_by_idx(obj, ELF64_R_SYM(rel->r_info)); + sym_idx = ELF64_R_SYM(rel->r_info); + sym = elf_sym_by_idx(obj, sym_idx); if (!sym) { - pr_warn("sec '%s': symbol 0x%zx not found for relo #%d\n", - relo_sec_name, (size_t)ELF64_R_SYM(rel->r_info), i); + pr_warn("sec '%s': symbol #%zu not found for relo #%d\n", + relo_sec_name, sym_idx, i); + return -LIBBPF_ERRNO__FORMAT; + } + + if (sym->st_shndx >= obj->efile.sec_cnt) { + pr_warn("sec '%s': corrupted symbol #%zu pointing to invalid section #%zu for relo #%d\n", + relo_sec_name, sym_idx, (size_t)sym->st_shndx, i); return -LIBBPF_ERRNO__FORMAT; } if (rel->r_offset % BPF_INSN_SZ || rel->r_offset >= scn_data->d_size) { pr_warn("sec '%s': invalid offset 0x%zx for relo #%d\n", - relo_sec_name, (size_t)ELF64_R_SYM(rel->r_info), i); + relo_sec_name, (size_t)rel->r_offset, i); return -LIBBPF_ERRNO__FORMAT; } From be2f2d1680dfb36793ea8d3110edd4a1db496352 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Tue, 2 Nov 2021 22:14:49 -0700 Subject: [PATCH 0075/2048] libbpf: Deprecate bpf_program__load() API Mark bpf_program__load() as deprecated ([0]) since v0.6. Also rename few internal program loading bpf_object helper functions to have more consistent naming. [0] Closes: https://github.com/libbpf/libbpf/issues/301 Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103051449.1884903-1-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 36 ++++++++++++++++++++++-------------- tools/lib/bpf/libbpf.h | 4 ++-- 2 files changed, 24 insertions(+), 16 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 86a44735230e..7fcea11ecaa9 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -6428,12 +6428,12 @@ static int libbpf_preload_prog(struct bpf_program *prog, return 0; } -static int -load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, - char *license, __u32 kern_version, int *pfd) +static int bpf_object_load_prog_instance(struct bpf_object *obj, struct bpf_program *prog, + struct bpf_insn *insns, int insns_cnt, + const char *license, __u32 kern_version, + int *prog_fd) { struct bpf_prog_load_params load_attr = {}; - struct bpf_object *obj = prog->obj; char *cp, errmsg[STRERR_BUFSIZE]; size_t log_buf_size = 0; char *log_buf = NULL; @@ -6494,7 +6494,7 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt, if (obj->gen_loader) { bpf_gen__prog_load(obj->gen_loader, &load_attr, prog - obj->programs); - *pfd = -1; + *prog_fd = -1; return 0; } retry_load: @@ -6532,7 +6532,7 @@ retry_load: } } - *pfd = ret; + *prog_fd = ret; ret = 0; goto out; } @@ -6608,11 +6608,12 @@ static int bpf_program__record_externs(struct bpf_program *prog) return 0; } -int bpf_program__load(struct bpf_program *prog, char *license, __u32 kern_ver) +static int bpf_object_load_prog(struct bpf_object *obj, struct bpf_program *prog, + const char *license, __u32 kern_ver) { int err = 0, fd, i; - if (prog->obj->loaded) { + if (obj->loaded) { pr_warn("prog '%s': can't load after object was loaded\n", prog->name); return libbpf_err(-EINVAL); } @@ -6638,10 +6639,11 @@ int bpf_program__load(struct bpf_program *prog, char *license, __u32 kern_ver) pr_warn("prog '%s': inconsistent nr(%d) != 1\n", prog->name, prog->instances.nr); } - if (prog->obj->gen_loader) + if (obj->gen_loader) bpf_program__record_externs(prog); - err = load_program(prog, prog->insns, prog->insns_cnt, - license, kern_ver, &fd); + err = bpf_object_load_prog_instance(obj, prog, + prog->insns, prog->insns_cnt, + license, kern_ver, &fd); if (!err) prog->instances.fds[0] = fd; goto out; @@ -6669,8 +6671,9 @@ int bpf_program__load(struct bpf_program *prog, char *license, __u32 kern_ver) continue; } - err = load_program(prog, result.new_insn_ptr, - result.new_insn_cnt, license, kern_ver, &fd); + err = bpf_object_load_prog_instance(obj, prog, + result.new_insn_ptr, result.new_insn_cnt, + license, kern_ver, &fd); if (err) { pr_warn("Loading the %dth instance of program '%s' failed\n", i, prog->name); @@ -6687,6 +6690,11 @@ out: return libbpf_err(err); } +int bpf_program__load(struct bpf_program *prog, const char *license, __u32 kern_ver) +{ + return bpf_object_load_prog(prog->obj, prog, license, kern_ver); +} + static int bpf_object__load_progs(struct bpf_object *obj, int log_level) { @@ -6710,7 +6718,7 @@ bpf_object__load_progs(struct bpf_object *obj, int log_level) continue; } prog->log_level |= log_level; - err = bpf_program__load(prog, obj->license, obj->kern_version); + err = bpf_object_load_prog(obj, prog, obj->license, obj->kern_version); if (err) return err; } diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 797f5f8a0e20..a364c379b998 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -262,8 +262,8 @@ LIBBPF_API const struct bpf_insn *bpf_program__insns(const struct bpf_program *p */ LIBBPF_API size_t bpf_program__insn_cnt(const struct bpf_program *prog); -LIBBPF_API int bpf_program__load(struct bpf_program *prog, char *license, - __u32 kern_version); +LIBBPF_DEPRECATED_SINCE(0, 6, "use bpf_object__load() instead") +LIBBPF_API int bpf_program__load(struct bpf_program *prog, const char *license, __u32 kern_version); LIBBPF_API int bpf_program__fd(const struct bpf_program *prog); LIBBPF_DEPRECATED_SINCE(0, 7, "multi-instance bpf_program support is deprecated") LIBBPF_API int bpf_program__pin_instance(struct bpf_program *prog, From b8b5cb55f5d3f03cc1479a3768d68173a10359ad Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Fri, 5 Nov 2021 12:10:55 -0700 Subject: [PATCH 0076/2048] libbpf: Fix non-C89 loop variable declaration in gen_loader.c Fix the `int i` declaration inside the for statement. This is non-C89 compliant. See [0] for user report breaking BCC build. [0] https://github.com/libbpf/libbpf/issues/403 Fixes: 18f4fccbf314 ("libbpf: Update gen_loader to emit BTF_KIND_FUNC relocations") Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/bpf/20211105191055.3324874-1-andrii@kernel.org --- tools/lib/bpf/gen_loader.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/lib/bpf/gen_loader.c b/tools/lib/bpf/gen_loader.c index 502dea53a742..2e10776b6d85 100644 --- a/tools/lib/bpf/gen_loader.c +++ b/tools/lib/bpf/gen_loader.c @@ -584,8 +584,9 @@ void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, bool is_weak, static struct ksym_desc *get_ksym_desc(struct bpf_gen *gen, struct ksym_relo_desc *relo) { struct ksym_desc *kdesc; + int i; - for (int i = 0; i < gen->nr_ksyms; i++) { + for (i = 0; i < gen->nr_ksyms; i++) { if (!strcmp(gen->ksyms[i].name, relo->name)) { gen->ksyms[i].ref++; return &gen->ksyms[i]; From be80e9cdbca8ac66d09e0e24e0bd41d992362a0b Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:34 -0700 Subject: [PATCH 0077/2048] libbpf: Rename DECLARE_LIBBPF_OPTS into LIBBPF_OPTS It's confusing that libbpf-provided helper macro doesn't start with LIBBPF. Also "declare" vs "define" is confusing terminology, I can never remember and always have to look up previous examples. Bypass both issues by renaming DECLARE_LIBBPF_OPTS into a short and clean LIBBPF_OPTS. To avoid breaking existing code, provide: #define DECLARE_LIBBPF_OPTS LIBBPF_OPTS in libbpf_legacy.h. We can decide later if we ever want to remove it or we'll keep it forever because it doesn't add any maintainability burden. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-2-andrii@kernel.org --- tools/lib/bpf/bpf.h | 1 + tools/lib/bpf/libbpf_common.h | 2 +- tools/lib/bpf/libbpf_legacy.h | 1 + 3 files changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h index 6fffb3cdf39b..f35146c1d9a9 100644 --- a/tools/lib/bpf/bpf.h +++ b/tools/lib/bpf/bpf.h @@ -29,6 +29,7 @@ #include #include "libbpf_common.h" +#include "libbpf_legacy.h" #ifdef __cplusplus extern "C" { diff --git a/tools/lib/bpf/libbpf_common.h b/tools/lib/bpf/libbpf_common.h index aaa1efbf6f51..0967112b933a 100644 --- a/tools/lib/bpf/libbpf_common.h +++ b/tools/lib/bpf/libbpf_common.h @@ -54,7 +54,7 @@ * including any extra padding, it with memset() and then assigns initial * values provided by users in struct initializer-syntax as varargs. */ -#define DECLARE_LIBBPF_OPTS(TYPE, NAME, ...) \ +#define LIBBPF_OPTS(TYPE, NAME, ...) \ struct TYPE NAME = ({ \ memset(&NAME, 0, sizeof(struct TYPE)); \ (struct TYPE) { \ diff --git a/tools/lib/bpf/libbpf_legacy.h b/tools/lib/bpf/libbpf_legacy.h index 5ba5c9beccfa..bb03c568af7b 100644 --- a/tools/lib/bpf/libbpf_legacy.h +++ b/tools/lib/bpf/libbpf_legacy.h @@ -69,6 +69,7 @@ enum libbpf_strict_mode { LIBBPF_API int libbpf_set_strict_mode(enum libbpf_strict_mode mode); +#define DECLARE_LIBBPF_OPTS LIBBPF_OPTS #ifdef __cplusplus } /* extern "C" */ From 45493cbaf59e3c9482e0e6a2646b362fff45db8b Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:35 -0700 Subject: [PATCH 0078/2048] libbpf: Pass number of prog load attempts explicitly Allow to control number of BPF_PROG_LOAD attempts from outside the sys_bpf_prog_load() helper. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-3-andrii@kernel.org --- tools/lib/bpf/bpf.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/tools/lib/bpf/bpf.c b/tools/lib/bpf/bpf.c index c09cbb868c9f..8e6a23c42560 100644 --- a/tools/lib/bpf/bpf.c +++ b/tools/lib/bpf/bpf.c @@ -74,14 +74,15 @@ static inline int sys_bpf_fd(enum bpf_cmd cmd, union bpf_attr *attr, return ensure_good_fd(fd); } -static inline int sys_bpf_prog_load(union bpf_attr *attr, unsigned int size) +#define PROG_LOAD_ATTEMPTS 5 + +static inline int sys_bpf_prog_load(union bpf_attr *attr, unsigned int size, int attempts) { - int retries = 5; int fd; do { fd = sys_bpf_fd(BPF_PROG_LOAD, attr, size); - } while (fd < 0 && errno == EAGAIN && retries-- > 0); + } while (fd < 0 && errno == EAGAIN && --attempts > 0); return fd; } @@ -304,7 +305,7 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) memcpy(attr.prog_name, load_attr->name, min(strlen(load_attr->name), (size_t)BPF_OBJ_NAME_LEN - 1)); - fd = sys_bpf_prog_load(&attr, sizeof(attr)); + fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); if (fd >= 0) return fd; @@ -345,7 +346,7 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) break; } - fd = sys_bpf_prog_load(&attr, sizeof(attr)); + fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); if (fd >= 0) goto done; } @@ -359,7 +360,7 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) attr.log_level = 1; load_attr->log_buf[0] = 0; - fd = sys_bpf_prog_load(&attr, sizeof(attr)); + fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); done: /* free() doesn't affect errno, so we don't need to restore it */ free(finfo); @@ -449,7 +450,7 @@ int bpf_verify_program(enum bpf_prog_type type, const struct bpf_insn *insns, attr.kern_version = kern_version; attr.prog_flags = prog_flags; - fd = sys_bpf_prog_load(&attr, sizeof(attr)); + fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); return libbpf_err_errno(fd); } From d10ef2b825cffd0807dd733fdfd6a5bea32270d7 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:36 -0700 Subject: [PATCH 0079/2048] libbpf: Unify low-level BPF_PROG_LOAD APIs into bpf_prog_load() Add a new unified OPTS-based low-level API for program loading, bpf_prog_load() ([0]). bpf_prog_load() accepts few "mandatory" parameters as input arguments (program type, name, license, instructions) and all the other optional (as in not required to specify for all types of BPF programs) fields into struct bpf_prog_load_opts. This makes all the other non-extensible APIs variant for BPF_PROG_LOAD obsolete and they are slated for deprecation in libbpf v0.7: - bpf_load_program(); - bpf_load_program_xattr(); - bpf_verify_program(). Implementation-wise, internal helper libbpf__bpf_prog_load is refactored to become a public bpf_prog_load() API. struct bpf_prog_load_params used internally is replaced by public struct bpf_prog_load_opts. Unfortunately, while conceptually all this is pretty straightforward, the biggest complication comes from the already existing bpf_prog_load() *high-level* API, which has nothing to do with BPF_PROG_LOAD command. We try really hard to have a new API named bpf_prog_load(), though, because it maps naturally to BPF_PROG_LOAD command. For that, we rename old bpf_prog_load() into bpf_prog_load_deprecated() and mark it as COMPAT_VERSION() for shared library users compiled against old version of libbpf. Statically linked users and shared lib users compiled against new version of libbpf headers will get "rerouted" to bpf_prog_deprecated() through a macro helper that decides whether to use new or old bpf_prog_load() based on number of input arguments (see ___libbpf_overload in libbpf_common.h). To test that existing bpf_prog_load()-using code compiles and works as expected, I've compiled and ran selftests as is. I had to remove (locally) selftest/bpf/Makefile -Dbpf_prog_load=bpf_prog_test_load hack because it was conflicting with the macro-based overload approach. I don't expect anyone else to do something like this in practice, though. This is testing-specific way to replace bpf_prog_load() calls with special testing variant of it, which adds extra prog_flags value. After testing I kept this selftests hack, but ensured that we use a new bpf_prog_load_deprecated name for this. This patch also marks bpf_prog_load() and bpf_prog_load_xattr() as deprecated. bpf_object interface has to be used for working with struct bpf_program. Libbpf doesn't support loading just a bpf_program. The silver lining is that when we get to libbpf 1.0 all these complication will be gone and we'll have one clean bpf_prog_load() low-level API with no backwards compatibility hackery surrounding it. [0] Closes: https://github.com/libbpf/libbpf/issues/284 Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103220845.2676888-4-andrii@kernel.org --- tools/lib/bpf/bpf.c | 149 ++++++++++++++++----------- tools/lib/bpf/bpf.h | 73 ++++++++++++- tools/lib/bpf/bpf_gen_internal.h | 8 +- tools/lib/bpf/gen_loader.c | 30 +++--- tools/lib/bpf/libbpf.c | 51 +++++---- tools/lib/bpf/libbpf.h | 5 +- tools/lib/bpf/libbpf.map | 2 + tools/lib/bpf/libbpf_common.h | 12 +++ tools/lib/bpf/libbpf_internal.h | 31 ------ tools/testing/selftests/bpf/Makefile | 2 +- 10 files changed, 223 insertions(+), 140 deletions(-) diff --git a/tools/lib/bpf/bpf.c b/tools/lib/bpf/bpf.c index 8e6a23c42560..8f2a701cb079 100644 --- a/tools/lib/bpf/bpf.c +++ b/tools/lib/bpf/bpf.c @@ -28,6 +28,7 @@ #include #include #include +#include #include "bpf.h" #include "libbpf.h" #include "libbpf_internal.h" @@ -254,58 +255,91 @@ alloc_zero_tailing_info(const void *orecord, __u32 cnt, return info; } -int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) +DEFAULT_VERSION(bpf_prog_load_v0_6_0, bpf_prog_load, LIBBPF_0.6.0) +int bpf_prog_load_v0_6_0(enum bpf_prog_type prog_type, + const char *prog_name, const char *license, + const struct bpf_insn *insns, size_t insn_cnt, + const struct bpf_prog_load_opts *opts) { void *finfo = NULL, *linfo = NULL; + const char *func_info, *line_info; + __u32 log_size, log_level, attach_prog_fd, attach_btf_obj_fd; + __u32 func_info_rec_size, line_info_rec_size; + int fd, attempts; union bpf_attr attr; - int fd; + char *log_buf; - if (!load_attr->log_buf != !load_attr->log_buf_sz) + if (!OPTS_VALID(opts, bpf_prog_load_opts)) return libbpf_err(-EINVAL); - if (load_attr->log_level > (4 | 2 | 1) || (load_attr->log_level && !load_attr->log_buf)) + attempts = OPTS_GET(opts, attempts, 0); + if (attempts < 0) return libbpf_err(-EINVAL); + if (attempts == 0) + attempts = PROG_LOAD_ATTEMPTS; memset(&attr, 0, sizeof(attr)); - attr.prog_type = load_attr->prog_type; - attr.expected_attach_type = load_attr->expected_attach_type; - if (load_attr->attach_prog_fd) - attr.attach_prog_fd = load_attr->attach_prog_fd; + attr.prog_type = prog_type; + attr.expected_attach_type = OPTS_GET(opts, expected_attach_type, 0); + + attr.prog_btf_fd = OPTS_GET(opts, prog_btf_fd, 0); + attr.prog_flags = OPTS_GET(opts, prog_flags, 0); + attr.prog_ifindex = OPTS_GET(opts, prog_ifindex, 0); + attr.kern_version = OPTS_GET(opts, kern_version, 0); + + if (prog_name) + strncat(attr.prog_name, prog_name, sizeof(attr.prog_name) - 1); + attr.license = ptr_to_u64(license); + + if (insn_cnt > UINT_MAX) + return libbpf_err(-E2BIG); + + attr.insns = ptr_to_u64(insns); + attr.insn_cnt = (__u32)insn_cnt; + + attach_prog_fd = OPTS_GET(opts, attach_prog_fd, 0); + attach_btf_obj_fd = OPTS_GET(opts, attach_btf_obj_fd, 0); + + if (attach_prog_fd && attach_btf_obj_fd) + return libbpf_err(-EINVAL); + + attr.attach_btf_id = OPTS_GET(opts, attach_btf_id, 0); + if (attach_prog_fd) + attr.attach_prog_fd = attach_prog_fd; else - attr.attach_btf_obj_fd = load_attr->attach_btf_obj_fd; - attr.attach_btf_id = load_attr->attach_btf_id; + attr.attach_btf_obj_fd = attach_btf_obj_fd; - attr.prog_ifindex = load_attr->prog_ifindex; - attr.kern_version = load_attr->kern_version; + log_buf = OPTS_GET(opts, log_buf, NULL); + log_size = OPTS_GET(opts, log_size, 0); + log_level = OPTS_GET(opts, log_level, 0); - attr.insn_cnt = (__u32)load_attr->insn_cnt; - attr.insns = ptr_to_u64(load_attr->insns); - attr.license = ptr_to_u64(load_attr->license); + if (!!log_buf != !!log_size) + return libbpf_err(-EINVAL); + if (log_level > (4 | 2 | 1)) + return libbpf_err(-EINVAL); + if (log_level && !log_buf) + return libbpf_err(-EINVAL); - attr.log_level = load_attr->log_level; - if (attr.log_level) { - attr.log_buf = ptr_to_u64(load_attr->log_buf); - attr.log_size = load_attr->log_buf_sz; - } + attr.log_level = log_level; + attr.log_buf = ptr_to_u64(log_buf); + attr.log_size = log_size; - attr.prog_btf_fd = load_attr->prog_btf_fd; - attr.prog_flags = load_attr->prog_flags; + func_info_rec_size = OPTS_GET(opts, func_info_rec_size, 0); + func_info = OPTS_GET(opts, func_info, NULL); + attr.func_info_rec_size = func_info_rec_size; + attr.func_info = ptr_to_u64(func_info); + attr.func_info_cnt = OPTS_GET(opts, func_info_cnt, 0); - attr.func_info_rec_size = load_attr->func_info_rec_size; - attr.func_info_cnt = load_attr->func_info_cnt; - attr.func_info = ptr_to_u64(load_attr->func_info); + line_info_rec_size = OPTS_GET(opts, line_info_rec_size, 0); + line_info = OPTS_GET(opts, line_info, NULL); + attr.line_info_rec_size = line_info_rec_size; + attr.line_info = ptr_to_u64(line_info); + attr.line_info_cnt = OPTS_GET(opts, line_info_cnt, 0); - attr.line_info_rec_size = load_attr->line_info_rec_size; - attr.line_info_cnt = load_attr->line_info_cnt; - attr.line_info = ptr_to_u64(load_attr->line_info); - attr.fd_array = ptr_to_u64(load_attr->fd_array); + attr.fd_array = ptr_to_u64(OPTS_GET(opts, fd_array, NULL)); - if (load_attr->name) - memcpy(attr.prog_name, load_attr->name, - min(strlen(load_attr->name), (size_t)BPF_OBJ_NAME_LEN - 1)); - - fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); + fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts); if (fd >= 0) return fd; @@ -315,11 +349,11 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) */ while (errno == E2BIG && (!finfo || !linfo)) { if (!finfo && attr.func_info_cnt && - attr.func_info_rec_size < load_attr->func_info_rec_size) { + attr.func_info_rec_size < func_info_rec_size) { /* try with corrected func info records */ - finfo = alloc_zero_tailing_info(load_attr->func_info, - load_attr->func_info_cnt, - load_attr->func_info_rec_size, + finfo = alloc_zero_tailing_info(func_info, + attr.func_info_cnt, + func_info_rec_size, attr.func_info_rec_size); if (!finfo) { errno = E2BIG; @@ -327,13 +361,12 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) } attr.func_info = ptr_to_u64(finfo); - attr.func_info_rec_size = load_attr->func_info_rec_size; + attr.func_info_rec_size = func_info_rec_size; } else if (!linfo && attr.line_info_cnt && - attr.line_info_rec_size < - load_attr->line_info_rec_size) { - linfo = alloc_zero_tailing_info(load_attr->line_info, - load_attr->line_info_cnt, - load_attr->line_info_rec_size, + attr.line_info_rec_size < line_info_rec_size) { + linfo = alloc_zero_tailing_info(line_info, + attr.line_info_cnt, + line_info_rec_size, attr.line_info_rec_size); if (!linfo) { errno = E2BIG; @@ -341,26 +374,26 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr) } attr.line_info = ptr_to_u64(linfo); - attr.line_info_rec_size = load_attr->line_info_rec_size; + attr.line_info_rec_size = line_info_rec_size; } else { break; } - fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); + fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts); if (fd >= 0) goto done; } - if (load_attr->log_level || !load_attr->log_buf) + if (log_level || !log_buf) goto done; /* Try again with log */ - attr.log_buf = ptr_to_u64(load_attr->log_buf); - attr.log_size = load_attr->log_buf_sz; + log_buf[0] = 0; + attr.log_buf = ptr_to_u64(log_buf); + attr.log_size = log_size; attr.log_level = 1; - load_attr->log_buf[0] = 0; - fd = sys_bpf_prog_load(&attr, sizeof(attr), PROG_LOAD_ATTEMPTS); + fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts); done: /* free() doesn't affect errno, so we don't need to restore it */ free(finfo); @@ -371,14 +404,13 @@ done: int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, char *log_buf, size_t log_buf_sz) { - struct bpf_prog_load_params p = {}; + LIBBPF_OPTS(bpf_prog_load_opts, p); if (!load_attr || !log_buf != !log_buf_sz) return libbpf_err(-EINVAL); - p.prog_type = load_attr->prog_type; p.expected_attach_type = load_attr->expected_attach_type; - switch (p.prog_type) { + switch (load_attr->prog_type) { case BPF_PROG_TYPE_STRUCT_OPS: case BPF_PROG_TYPE_LSM: p.attach_btf_id = load_attr->attach_btf_id; @@ -392,12 +424,9 @@ int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, p.prog_ifindex = load_attr->prog_ifindex; p.kern_version = load_attr->kern_version; } - p.insn_cnt = load_attr->insns_cnt; - p.insns = load_attr->insns; - p.license = load_attr->license; p.log_level = load_attr->log_level; p.log_buf = log_buf; - p.log_buf_sz = log_buf_sz; + p.log_size = log_buf_sz; p.prog_btf_fd = load_attr->prog_btf_fd; p.func_info_rec_size = load_attr->func_info_rec_size; p.func_info_cnt = load_attr->func_info_cnt; @@ -405,10 +434,10 @@ int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, p.line_info_rec_size = load_attr->line_info_rec_size; p.line_info_cnt = load_attr->line_info_cnt; p.line_info = load_attr->line_info; - p.name = load_attr->name; p.prog_flags = load_attr->prog_flags; - return libbpf__bpf_prog_load(&p); + return bpf_prog_load(load_attr->prog_type, load_attr->name, load_attr->license, + load_attr->insns, load_attr->insns_cnt, &p); } int bpf_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h index f35146c1d9a9..079cc81ac51e 100644 --- a/tools/lib/bpf/bpf.h +++ b/tools/lib/bpf/bpf.h @@ -72,6 +72,71 @@ LIBBPF_API int bpf_create_map_in_map(enum bpf_map_type map_type, int inner_map_fd, int max_entries, __u32 map_flags); +struct bpf_prog_load_opts { + size_t sz; /* size of this struct for forward/backward compatibility */ + + /* libbpf can retry BPF_PROG_LOAD command if bpf() syscall returns + * -EAGAIN. This field determines how many attempts libbpf has to + * make. If not specified, libbpf will use default value of 5. + */ + int attempts; + + enum bpf_attach_type expected_attach_type; + __u32 prog_btf_fd; + __u32 prog_flags; + __u32 prog_ifindex; + __u32 kern_version; + + __u32 attach_btf_id; + __u32 attach_prog_fd; + __u32 attach_btf_obj_fd; + + const int *fd_array; + + /* .BTF.ext func info data */ + const void *func_info; + __u32 func_info_cnt; + __u32 func_info_rec_size; + + /* .BTF.ext line info data */ + const void *line_info; + __u32 line_info_cnt; + __u32 line_info_rec_size; + + /* verifier log options */ + __u32 log_level; + __u32 log_size; + char *log_buf; +}; +#define bpf_prog_load_opts__last_field log_buf + +LIBBPF_API int bpf_prog_load(enum bpf_prog_type prog_type, + const char *prog_name, const char *license, + const struct bpf_insn *insns, size_t insn_cnt, + const struct bpf_prog_load_opts *opts); +/* this "specialization" should go away in libbpf 1.0 */ +LIBBPF_API int bpf_prog_load_v0_6_0(enum bpf_prog_type prog_type, + const char *prog_name, const char *license, + const struct bpf_insn *insns, size_t insn_cnt, + const struct bpf_prog_load_opts *opts); + +/* This is an elaborate way to not conflict with deprecated bpf_prog_load() + * API, defined in libbpf.h. Once we hit libbpf 1.0, all this will be gone. + * With this approach, if someone is calling bpf_prog_load() with + * 4 arguments, they will use the deprecated API, which keeps backwards + * compatibility (both source code and binary). If bpf_prog_load() is called + * with 6 arguments, though, it gets redirected to __bpf_prog_load. + * So looking forward to libbpf 1.0 when this hack will be gone and + * __bpf_prog_load() will be called just bpf_prog_load(). + */ +#ifndef bpf_prog_load +#define bpf_prog_load(...) ___libbpf_overload(___bpf_prog_load, __VA_ARGS__) +#define ___bpf_prog_load4(file, type, pobj, prog_fd) \ + bpf_prog_load_deprecated(file, type, pobj, prog_fd) +#define ___bpf_prog_load6(prog_type, prog_name, license, insns, insn_cnt, opts) \ + bpf_prog_load(prog_type, prog_name, license, insns, insn_cnt, opts) +#endif /* bpf_prog_load */ + struct bpf_load_program_attr { enum bpf_prog_type prog_type; enum bpf_attach_type expected_attach_type; @@ -103,13 +168,15 @@ struct bpf_load_program_attr { /* Recommend log buffer size */ #define BPF_LOG_BUF_SIZE (UINT32_MAX >> 8) /* verifier maximum in kernels <= 5.1 */ -LIBBPF_API int -bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, - char *log_buf, size_t log_buf_sz); +LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_prog_load() instead") +LIBBPF_API int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, + char *log_buf, size_t log_buf_sz); +LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_prog_load() instead") LIBBPF_API int bpf_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, size_t insns_cnt, const char *license, __u32 kern_version, char *log_buf, size_t log_buf_sz); +LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_prog_load() instead") LIBBPF_API int bpf_verify_program(enum bpf_prog_type type, const struct bpf_insn *insns, size_t insns_cnt, __u32 prog_flags, diff --git a/tools/lib/bpf/bpf_gen_internal.h b/tools/lib/bpf/bpf_gen_internal.h index d26e5472fe50..75ca9fb857b2 100644 --- a/tools/lib/bpf/bpf_gen_internal.h +++ b/tools/lib/bpf/bpf_gen_internal.h @@ -3,6 +3,8 @@ #ifndef __BPF_GEN_INTERNAL_H #define __BPF_GEN_INTERNAL_H +#include "bpf.h" + struct ksym_relo_desc { const char *name; int kind; @@ -50,8 +52,10 @@ int bpf_gen__finish(struct bpf_gen *gen); void bpf_gen__free(struct bpf_gen *gen); void bpf_gen__load_btf(struct bpf_gen *gen, const void *raw_data, __u32 raw_size); void bpf_gen__map_create(struct bpf_gen *gen, struct bpf_create_map_params *map_attr, int map_idx); -struct bpf_prog_load_params; -void bpf_gen__prog_load(struct bpf_gen *gen, struct bpf_prog_load_params *load_attr, int prog_idx); +void bpf_gen__prog_load(struct bpf_gen *gen, + enum bpf_prog_type prog_type, const char *prog_name, + const char *license, struct bpf_insn *insns, size_t insn_cnt, + struct bpf_prog_load_opts *load_attr, int prog_idx); void bpf_gen__map_update_elem(struct bpf_gen *gen, int map_idx, void *value, __u32 value_size); void bpf_gen__map_freeze(struct bpf_gen *gen, int map_idx); void bpf_gen__record_attach_target(struct bpf_gen *gen, const char *name, enum bpf_attach_type type); diff --git a/tools/lib/bpf/gen_loader.c b/tools/lib/bpf/gen_loader.c index 2e10776b6d85..7b73f97b1fa1 100644 --- a/tools/lib/bpf/gen_loader.c +++ b/tools/lib/bpf/gen_loader.c @@ -901,27 +901,27 @@ static void cleanup_relos(struct bpf_gen *gen, int insns) } void bpf_gen__prog_load(struct bpf_gen *gen, - struct bpf_prog_load_params *load_attr, int prog_idx) + enum bpf_prog_type prog_type, const char *prog_name, + const char *license, struct bpf_insn *insns, size_t insn_cnt, + struct bpf_prog_load_opts *load_attr, int prog_idx) { int attr_size = offsetofend(union bpf_attr, fd_array); - int prog_load_attr, license, insns, func_info, line_info; + int prog_load_attr, license_off, insns_off, func_info, line_info; union bpf_attr attr; memset(&attr, 0, attr_size); - pr_debug("gen: prog_load: type %d insns_cnt %zd\n", - load_attr->prog_type, load_attr->insn_cnt); + pr_debug("gen: prog_load: type %d insns_cnt %zd\n", prog_type, insn_cnt); /* add license string to blob of bytes */ - license = add_data(gen, load_attr->license, strlen(load_attr->license) + 1); + license_off = add_data(gen, license, strlen(license) + 1); /* add insns to blob of bytes */ - insns = add_data(gen, load_attr->insns, - load_attr->insn_cnt * sizeof(struct bpf_insn)); + insns_off = add_data(gen, insns, insn_cnt * sizeof(struct bpf_insn)); - attr.prog_type = load_attr->prog_type; + attr.prog_type = prog_type; attr.expected_attach_type = load_attr->expected_attach_type; attr.attach_btf_id = load_attr->attach_btf_id; attr.prog_ifindex = load_attr->prog_ifindex; attr.kern_version = 0; - attr.insn_cnt = (__u32)load_attr->insn_cnt; + attr.insn_cnt = (__u32)insn_cnt; attr.prog_flags = load_attr->prog_flags; attr.func_info_rec_size = load_attr->func_info_rec_size; @@ -934,15 +934,15 @@ void bpf_gen__prog_load(struct bpf_gen *gen, line_info = add_data(gen, load_attr->line_info, attr.line_info_cnt * attr.line_info_rec_size); - memcpy(attr.prog_name, load_attr->name, - min((unsigned)strlen(load_attr->name), BPF_OBJ_NAME_LEN - 1)); + memcpy(attr.prog_name, prog_name, + min((unsigned)strlen(prog_name), BPF_OBJ_NAME_LEN - 1)); prog_load_attr = add_data(gen, &attr, attr_size); /* populate union bpf_attr with a pointer to license */ - emit_rel_store(gen, attr_field(prog_load_attr, license), license); + emit_rel_store(gen, attr_field(prog_load_attr, license), license_off); /* populate union bpf_attr with a pointer to instructions */ - emit_rel_store(gen, attr_field(prog_load_attr, insns), insns); + emit_rel_store(gen, attr_field(prog_load_attr, insns), insns_off); /* populate union bpf_attr with a pointer to func_info */ emit_rel_store(gen, attr_field(prog_load_attr, func_info), func_info); @@ -974,12 +974,12 @@ void bpf_gen__prog_load(struct bpf_gen *gen, emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_0, BPF_REG_7, offsetof(union bpf_attr, attach_btf_obj_fd))); } - emit_relos(gen, insns); + emit_relos(gen, insns_off); /* emit PROG_LOAD command */ emit_sys_bpf(gen, BPF_PROG_LOAD, prog_load_attr, attr_size); debug_ret(gen, "prog_load %s insn_cnt %d", attr.prog_name, attr.insn_cnt); /* successful or not, close btf module FDs used in extern ksyms and attach_btf_obj_fd */ - cleanup_relos(gen, insns); + cleanup_relos(gen, insns_off); if (gen->attach_kind) emit_sys_close_blob(gen, attr_field(prog_load_attr, attach_btf_obj_fd)); diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 7fcea11ecaa9..7a82b81b8859 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -221,7 +221,7 @@ struct reloc_desc { struct bpf_sec_def; typedef int (*init_fn_t)(struct bpf_program *prog, long cookie); -typedef int (*preload_fn_t)(struct bpf_program *prog, struct bpf_prog_load_params *attr, long cookie); +typedef int (*preload_fn_t)(struct bpf_program *prog, struct bpf_prog_load_opts *opts, long cookie); typedef struct bpf_link *(*attach_fn_t)(const struct bpf_program *prog, long cookie); /* stored as sec_def->cookie for all libbpf-supported SEC()s */ @@ -6391,16 +6391,16 @@ static int libbpf_find_attach_btf_id(struct bpf_program *prog, const char *attac /* this is called as prog->sec_def->preload_fn for libbpf-supported sec_defs */ static int libbpf_preload_prog(struct bpf_program *prog, - struct bpf_prog_load_params *attr, long cookie) + struct bpf_prog_load_opts *opts, long cookie) { enum sec_def_flags def = cookie; /* old kernels might not support specifying expected_attach_type */ if ((def & SEC_EXP_ATTACH_OPT) && !kernel_supports(prog->obj, FEAT_EXP_ATTACH_TYPE)) - attr->expected_attach_type = 0; + opts->expected_attach_type = 0; if (def & SEC_SLEEPABLE) - attr->prog_flags |= BPF_F_SLEEPABLE; + opts->prog_flags |= BPF_F_SLEEPABLE; if ((prog->type == BPF_PROG_TYPE_TRACING || prog->type == BPF_PROG_TYPE_LSM || @@ -6419,11 +6419,11 @@ static int libbpf_preload_prog(struct bpf_program *prog, /* but by now libbpf common logic is not utilizing * prog->atach_btf_obj_fd/prog->attach_btf_id anymore because - * this callback is called after attrs were populated by - * libbpf, so this callback has to update attr explicitly here + * this callback is called after opts were populated by + * libbpf, so this callback has to update opts explicitly here */ - attr->attach_btf_obj_fd = btf_obj_fd; - attr->attach_btf_id = btf_type_id; + opts->attach_btf_obj_fd = btf_obj_fd; + opts->attach_btf_id = btf_type_id; } return 0; } @@ -6433,7 +6433,8 @@ static int bpf_object_load_prog_instance(struct bpf_object *obj, struct bpf_prog const char *license, __u32 kern_version, int *prog_fd) { - struct bpf_prog_load_params load_attr = {}; + LIBBPF_OPTS(bpf_prog_load_opts, load_attr); + const char *prog_name = NULL; char *cp, errmsg[STRERR_BUFSIZE]; size_t log_buf_size = 0; char *log_buf = NULL; @@ -6452,13 +6453,9 @@ static int bpf_object_load_prog_instance(struct bpf_object *obj, struct bpf_prog if (!insns || !insns_cnt) return -EINVAL; - load_attr.prog_type = prog->type; load_attr.expected_attach_type = prog->expected_attach_type; if (kernel_supports(obj, FEAT_PROG_NAME)) - load_attr.name = prog->name; - load_attr.insns = insns; - load_attr.insn_cnt = insns_cnt; - load_attr.license = license; + prog_name = prog->name; load_attr.attach_btf_id = prog->attach_btf_id; load_attr.attach_prog_fd = prog->attach_prog_fd; load_attr.attach_btf_obj_fd = prog->attach_btf_obj_fd; @@ -6492,7 +6489,8 @@ static int bpf_object_load_prog_instance(struct bpf_object *obj, struct bpf_prog } if (obj->gen_loader) { - bpf_gen__prog_load(obj->gen_loader, &load_attr, + bpf_gen__prog_load(obj->gen_loader, prog->type, prog->name, + license, insns, insns_cnt, &load_attr, prog - obj->programs); *prog_fd = -1; return 0; @@ -6507,8 +6505,8 @@ retry_load: } load_attr.log_buf = log_buf; - load_attr.log_buf_sz = log_buf_size; - ret = libbpf__bpf_prog_load(&load_attr); + load_attr.log_size = log_buf_size; + ret = bpf_prog_load(prog->type, prog_name, license, insns, insns_cnt, &load_attr); if (ret >= 0) { if (log_buf && load_attr.log_level) @@ -6554,19 +6552,19 @@ retry_load: pr_warn("-- BEGIN DUMP LOG ---\n"); pr_warn("\n%s\n", log_buf); pr_warn("-- END LOG --\n"); - } else if (load_attr.insn_cnt >= BPF_MAXINSNS) { - pr_warn("Program too large (%zu insns), at most %d insns\n", - load_attr.insn_cnt, BPF_MAXINSNS); + } else if (insns_cnt >= BPF_MAXINSNS) { + pr_warn("Program too large (%d insns), at most %d insns\n", + insns_cnt, BPF_MAXINSNS); ret = -LIBBPF_ERRNO__PROG2BIG; - } else if (load_attr.prog_type != BPF_PROG_TYPE_KPROBE) { + } else if (prog->type != BPF_PROG_TYPE_KPROBE) { /* Wrong program type? */ int fd; - load_attr.prog_type = BPF_PROG_TYPE_KPROBE; load_attr.expected_attach_type = 0; load_attr.log_buf = NULL; - load_attr.log_buf_sz = 0; - fd = libbpf__bpf_prog_load(&load_attr); + load_attr.log_size = 0; + fd = bpf_prog_load(BPF_PROG_TYPE_KPROBE, prog_name, license, + insns, insns_cnt, &load_attr); if (fd >= 0) { close(fd); ret = -LIBBPF_ERRNO__PROGTYPE; @@ -9170,8 +9168,9 @@ long libbpf_get_error(const void *ptr) return -errno; } -int bpf_prog_load(const char *file, enum bpf_prog_type type, - struct bpf_object **pobj, int *prog_fd) +COMPAT_VERSION(bpf_prog_load_deprecated, bpf_prog_load, LIBBPF_0.0.1) +int bpf_prog_load_deprecated(const char *file, enum bpf_prog_type type, + struct bpf_object **pobj, int *prog_fd) { struct bpf_prog_load_attr attr; diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index a364c379b998..bbc828667b22 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -676,8 +676,9 @@ struct bpf_prog_load_attr { LIBBPF_API int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr, struct bpf_object **pobj, int *prog_fd); -LIBBPF_API int bpf_prog_load(const char *file, enum bpf_prog_type type, - struct bpf_object **pobj, int *prog_fd); +LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__open() and bpf_object__load() instead") +LIBBPF_API int bpf_prog_load_deprecated(const char *file, enum bpf_prog_type type, + struct bpf_object **pobj, int *prog_fd); /* XDP related API */ struct xdp_link_info { diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index 43580eb47740..b895861a13c0 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -395,6 +395,8 @@ LIBBPF_0.6.0 { bpf_object__next_program; bpf_object__prev_map; bpf_object__prev_program; + bpf_prog_load_deprecated; + bpf_prog_load; bpf_program__insn_cnt; bpf_program__insns; btf__add_btf; diff --git a/tools/lib/bpf/libbpf_common.h b/tools/lib/bpf/libbpf_common.h index 0967112b933a..b21cefc9c3b6 100644 --- a/tools/lib/bpf/libbpf_common.h +++ b/tools/lib/bpf/libbpf_common.h @@ -41,6 +41,18 @@ #define __LIBBPF_MARK_DEPRECATED_0_7(X) #endif +/* This set of internal macros allows to do "function overloading" based on + * number of arguments provided by used in backwards-compatible way during the + * transition to libbpf 1.0 + * It's ugly but necessary evil that will be cleaned up when we get to 1.0. + * See bpf_prog_load() overload for example. + */ +#define ___libbpf_cat(A, B) A ## B +#define ___libbpf_select(NAME, NUM) ___libbpf_cat(NAME, NUM) +#define ___libbpf_nth(_1, _2, _3, _4, _5, _6, N, ...) N +#define ___libbpf_cnt(...) ___libbpf_nth(__VA_ARGS__, 6, 5, 4, 3, 2, 1) +#define ___libbpf_overload(NAME, ...) ___libbpf_select(NAME, ___libbpf_cnt(__VA_ARGS__))(__VA_ARGS__) + /* Helper macro to declare and initialize libbpf options struct * * This dance with uninitialized declaration, followed by memset to zero, diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index aeb79e3a8ff9..c1e34794b829 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -276,37 +276,6 @@ int parse_cpu_mask_file(const char *fcpu, bool **mask, int *mask_sz); int libbpf__load_raw_btf(const char *raw_types, size_t types_len, const char *str_sec, size_t str_len); -struct bpf_prog_load_params { - enum bpf_prog_type prog_type; - enum bpf_attach_type expected_attach_type; - const char *name; - const struct bpf_insn *insns; - size_t insn_cnt; - const char *license; - __u32 kern_version; - __u32 attach_prog_fd; - __u32 attach_btf_obj_fd; - __u32 attach_btf_id; - __u32 prog_ifindex; - __u32 prog_btf_fd; - __u32 prog_flags; - - __u32 func_info_rec_size; - const void *func_info; - __u32 func_info_cnt; - - __u32 line_info_rec_size; - const void *line_info; - __u32 line_info_cnt; - - __u32 log_level; - char *log_buf; - size_t log_buf_sz; - int *fd_array; -}; - -int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr); - struct bpf_create_map_params { const char *name; enum bpf_map_type map_type; diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 54b0a41a3775..c4497a4af3fe 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -24,7 +24,7 @@ SAN_CFLAGS ?= CFLAGS += -g -O0 -rdynamic -Wall $(GENFLAGS) $(SAN_CFLAGS) \ -I$(CURDIR) -I$(INCLUDE_DIR) -I$(GENDIR) -I$(LIBDIR) \ -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT) \ - -Dbpf_prog_load=bpf_prog_test_load \ + -Dbpf_prog_load_deprecated=bpf_prog_test_load \ -Dbpf_load_program=bpf_test_load_program LDLIBS += -lcap -lelf -lz -lrt -lpthread From e32660ac6fd6bd3c9d249644330d968c6ef61b07 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:37 -0700 Subject: [PATCH 0080/2048] libbpf: Remove internal use of deprecated bpf_prog_load() variants Remove all the internal uses of bpf_load_program_xattr(), which is slated for deprecation in v0.7. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103220845.2676888-5-andrii@kernel.org --- tools/lib/bpf/bpf.c | 8 ++- tools/lib/bpf/libbpf.c | 119 +++++++++++----------------------- tools/lib/bpf/libbpf_probes.c | 20 +++--- tools/lib/bpf/xsk.c | 34 ++++------ 4 files changed, 64 insertions(+), 117 deletions(-) diff --git a/tools/lib/bpf/bpf.c b/tools/lib/bpf/bpf.c index 8f2a701cb079..4ff45749f8cf 100644 --- a/tools/lib/bpf/bpf.c +++ b/tools/lib/bpf/bpf.c @@ -401,8 +401,12 @@ done: return libbpf_err_errno(fd); } +__attribute__((alias("bpf_load_program_xattr2"))) int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, - char *log_buf, size_t log_buf_sz) + char *log_buf, size_t log_buf_sz); + +static int bpf_load_program_xattr2(const struct bpf_load_program_attr *load_attr, + char *log_buf, size_t log_buf_sz) { LIBBPF_OPTS(bpf_prog_load_opts, p); @@ -456,7 +460,7 @@ int bpf_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, load_attr.license = license; load_attr.kern_version = kern_version; - return bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz); + return bpf_load_program_xattr2(&load_attr, log_buf, log_buf_sz); } int bpf_verify_program(enum bpf_prog_type type, const struct bpf_insn *insns, diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 7a82b81b8859..5751cade0f66 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -4282,30 +4282,20 @@ int bpf_map__resize(struct bpf_map *map, __u32 max_entries) static int bpf_object__probe_loading(struct bpf_object *obj) { - struct bpf_load_program_attr attr; char *cp, errmsg[STRERR_BUFSIZE]; struct bpf_insn insns[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; - int ret; + int ret, insn_cnt = ARRAY_SIZE(insns); if (obj->gen_loader) return 0; /* make sure basic loading works */ - - memset(&attr, 0, sizeof(attr)); - attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER; - attr.insns = insns; - attr.insns_cnt = ARRAY_SIZE(insns); - attr.license = "GPL"; - - ret = bpf_load_program_xattr(&attr, NULL, 0); - if (ret < 0) { - attr.prog_type = BPF_PROG_TYPE_TRACEPOINT; - ret = bpf_load_program_xattr(&attr, NULL, 0); - } + ret = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL", insns, insn_cnt, NULL); + if (ret < 0) + ret = bpf_prog_load(BPF_PROG_TYPE_TRACEPOINT, NULL, "GPL", insns, insn_cnt, NULL); if (ret < 0) { ret = errno; cp = libbpf_strerror_r(ret, errmsg, sizeof(errmsg)); @@ -4329,28 +4319,19 @@ static int probe_fd(int fd) static int probe_kern_prog_name(void) { - struct bpf_load_program_attr attr; struct bpf_insn insns[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; - int ret; + int ret, insn_cnt = ARRAY_SIZE(insns); /* make sure loading with name works */ - - memset(&attr, 0, sizeof(attr)); - attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER; - attr.insns = insns; - attr.insns_cnt = ARRAY_SIZE(insns); - attr.license = "GPL"; - attr.name = "test"; - ret = bpf_load_program_xattr(&attr, NULL, 0); + ret = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, "test", "GPL", insns, insn_cnt, NULL); return probe_fd(ret); } static int probe_kern_global_data(void) { - struct bpf_load_program_attr prg_attr; struct bpf_create_map_attr map_attr; char *cp, errmsg[STRERR_BUFSIZE]; struct bpf_insn insns[] = { @@ -4359,7 +4340,7 @@ static int probe_kern_global_data(void) BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; - int ret, map; + int ret, map, insn_cnt = ARRAY_SIZE(insns); memset(&map_attr, 0, sizeof(map_attr)); map_attr.map_type = BPF_MAP_TYPE_ARRAY; @@ -4378,13 +4359,7 @@ static int probe_kern_global_data(void) insns[0].imm = map; - memset(&prg_attr, 0, sizeof(prg_attr)); - prg_attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER; - prg_attr.insns = insns; - prg_attr.insns_cnt = ARRAY_SIZE(insns); - prg_attr.license = "GPL"; - - ret = bpf_load_program_xattr(&prg_attr, NULL, 0); + ret = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL", insns, insn_cnt, NULL); close(map); return probe_fd(ret); } @@ -4500,30 +4475,24 @@ static int probe_kern_array_mmap(void) static int probe_kern_exp_attach_type(void) { - struct bpf_load_program_attr attr; + LIBBPF_OPTS(bpf_prog_load_opts, opts, .expected_attach_type = BPF_CGROUP_INET_SOCK_CREATE); struct bpf_insn insns[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; + int fd, insn_cnt = ARRAY_SIZE(insns); - memset(&attr, 0, sizeof(attr)); /* use any valid combination of program type and (optional) * non-zero expected attach type (i.e., not a BPF_CGROUP_INET_INGRESS) * to see if kernel supports expected_attach_type field for * BPF_PROG_LOAD command */ - attr.prog_type = BPF_PROG_TYPE_CGROUP_SOCK; - attr.expected_attach_type = BPF_CGROUP_INET_SOCK_CREATE; - attr.insns = insns; - attr.insns_cnt = ARRAY_SIZE(insns); - attr.license = "GPL"; - - return probe_fd(bpf_load_program_xattr(&attr, NULL, 0)); + fd = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SOCK, NULL, "GPL", insns, insn_cnt, &opts); + return probe_fd(fd); } static int probe_kern_probe_read_kernel(void) { - struct bpf_load_program_attr attr; struct bpf_insn insns[] = { BPF_MOV64_REG(BPF_REG_1, BPF_REG_10), /* r1 = r10 (fp) */ BPF_ALU64_IMM(BPF_ADD, BPF_REG_1, -8), /* r1 += -8 */ @@ -4532,26 +4501,21 @@ static int probe_kern_probe_read_kernel(void) BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_probe_read_kernel), BPF_EXIT_INSN(), }; + int fd, insn_cnt = ARRAY_SIZE(insns); - memset(&attr, 0, sizeof(attr)); - attr.prog_type = BPF_PROG_TYPE_KPROBE; - attr.insns = insns; - attr.insns_cnt = ARRAY_SIZE(insns); - attr.license = "GPL"; - - return probe_fd(bpf_load_program_xattr(&attr, NULL, 0)); + fd = bpf_prog_load(BPF_PROG_TYPE_KPROBE, NULL, "GPL", insns, insn_cnt, NULL); + return probe_fd(fd); } static int probe_prog_bind_map(void) { - struct bpf_load_program_attr prg_attr; struct bpf_create_map_attr map_attr; char *cp, errmsg[STRERR_BUFSIZE]; struct bpf_insn insns[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; - int ret, map, prog; + int ret, map, prog, insn_cnt = ARRAY_SIZE(insns); memset(&map_attr, 0, sizeof(map_attr)); map_attr.map_type = BPF_MAP_TYPE_ARRAY; @@ -4568,13 +4532,7 @@ static int probe_prog_bind_map(void) return ret; } - memset(&prg_attr, 0, sizeof(prg_attr)); - prg_attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER; - prg_attr.insns = insns; - prg_attr.insns_cnt = ARRAY_SIZE(insns); - prg_attr.license = "GPL"; - - prog = bpf_load_program_xattr(&prg_attr, NULL, 0); + prog = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL", insns, insn_cnt, NULL); if (prog < 0) { close(map); return 0; @@ -4619,19 +4577,14 @@ static int probe_module_btf(void) static int probe_perf_link(void) { - struct bpf_load_program_attr attr; struct bpf_insn insns[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; int prog_fd, link_fd, err; - memset(&attr, 0, sizeof(attr)); - attr.prog_type = BPF_PROG_TYPE_TRACEPOINT; - attr.insns = insns; - attr.insns_cnt = ARRAY_SIZE(insns); - attr.license = "GPL"; - prog_fd = bpf_load_program_xattr(&attr, NULL, 0); + prog_fd = bpf_prog_load(BPF_PROG_TYPE_TRACEPOINT, NULL, "GPL", + insns, ARRAY_SIZE(insns), NULL); if (prog_fd < 0) return -errno; @@ -9168,22 +9121,12 @@ long libbpf_get_error(const void *ptr) return -errno; } -COMPAT_VERSION(bpf_prog_load_deprecated, bpf_prog_load, LIBBPF_0.0.1) -int bpf_prog_load_deprecated(const char *file, enum bpf_prog_type type, - struct bpf_object **pobj, int *prog_fd) -{ - struct bpf_prog_load_attr attr; - - memset(&attr, 0, sizeof(struct bpf_prog_load_attr)); - attr.file = file; - attr.prog_type = type; - attr.expected_attach_type = 0; - - return bpf_prog_load_xattr(&attr, pobj, prog_fd); -} - +__attribute__((alias("bpf_prog_load_xattr2"))) int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr, - struct bpf_object **pobj, int *prog_fd) + struct bpf_object **pobj, int *prog_fd); + +static int bpf_prog_load_xattr2(const struct bpf_prog_load_attr *attr, + struct bpf_object **pobj, int *prog_fd) { struct bpf_object_open_attr open_attr = {}; struct bpf_program *prog, *first_prog = NULL; @@ -9254,6 +9197,20 @@ int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr, return 0; } +COMPAT_VERSION(bpf_prog_load_deprecated, bpf_prog_load, LIBBPF_0.0.1) +int bpf_prog_load_deprecated(const char *file, enum bpf_prog_type type, + struct bpf_object **pobj, int *prog_fd) +{ + struct bpf_prog_load_attr attr; + + memset(&attr, 0, sizeof(struct bpf_prog_load_attr)); + attr.file = file; + attr.prog_type = type; + attr.expected_attach_type = 0; + + return bpf_prog_load_xattr2(&attr, pobj, prog_fd); +} + struct bpf_link { int (*detach)(struct bpf_link *link); void (*dealloc)(struct bpf_link *link); diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c index 68f2dbf364aa..02c401e314c7 100644 --- a/tools/lib/bpf/libbpf_probes.c +++ b/tools/lib/bpf/libbpf_probes.c @@ -68,21 +68,21 @@ static void probe_load(enum bpf_prog_type prog_type, const struct bpf_insn *insns, size_t insns_cnt, char *buf, size_t buf_len, __u32 ifindex) { - struct bpf_load_program_attr xattr = {}; + LIBBPF_OPTS(bpf_prog_load_opts, opts); int fd; switch (prog_type) { case BPF_PROG_TYPE_CGROUP_SOCK_ADDR: - xattr.expected_attach_type = BPF_CGROUP_INET4_CONNECT; + opts.expected_attach_type = BPF_CGROUP_INET4_CONNECT; break; case BPF_PROG_TYPE_CGROUP_SOCKOPT: - xattr.expected_attach_type = BPF_CGROUP_GETSOCKOPT; + opts.expected_attach_type = BPF_CGROUP_GETSOCKOPT; break; case BPF_PROG_TYPE_SK_LOOKUP: - xattr.expected_attach_type = BPF_SK_LOOKUP; + opts.expected_attach_type = BPF_SK_LOOKUP; break; case BPF_PROG_TYPE_KPROBE: - xattr.kern_version = get_kernel_version(); + opts.kern_version = get_kernel_version(); break; case BPF_PROG_TYPE_UNSPEC: case BPF_PROG_TYPE_SOCKET_FILTER: @@ -115,13 +115,11 @@ probe_load(enum bpf_prog_type prog_type, const struct bpf_insn *insns, break; } - xattr.prog_type = prog_type; - xattr.insns = insns; - xattr.insns_cnt = insns_cnt; - xattr.license = "GPL"; - xattr.prog_ifindex = ifindex; + opts.prog_ifindex = ifindex; + opts.log_buf = buf; + opts.log_size = buf_len; - fd = bpf_load_program_xattr(&xattr, buf, buf_len); + fd = bpf_prog_load(prog_type, NULL, "GPL", insns, insns_cnt, NULL); if (fd >= 0) close(fd); } diff --git a/tools/lib/bpf/xsk.c b/tools/lib/bpf/xsk.c index 81f8fbc85e70..fdb22f5405c9 100644 --- a/tools/lib/bpf/xsk.c +++ b/tools/lib/bpf/xsk.c @@ -364,7 +364,6 @@ int xsk_umem__create_v0_0_2(struct xsk_umem **umem_ptr, void *umem_area, static enum xsk_prog get_xsk_prog(void) { enum xsk_prog detected = XSK_PROG_FALLBACK; - struct bpf_load_program_attr prog_attr; struct bpf_create_map_attr map_attr; __u32 size_out, retval, duration; char data_in = 0, data_out; @@ -375,7 +374,7 @@ static enum xsk_prog get_xsk_prog(void) BPF_EMIT_CALL(BPF_FUNC_redirect_map), BPF_EXIT_INSN(), }; - int prog_fd, map_fd, ret; + int prog_fd, map_fd, ret, insn_cnt = ARRAY_SIZE(insns); memset(&map_attr, 0, sizeof(map_attr)); map_attr.map_type = BPF_MAP_TYPE_XSKMAP; @@ -389,13 +388,7 @@ static enum xsk_prog get_xsk_prog(void) insns[0].imm = map_fd; - memset(&prog_attr, 0, sizeof(prog_attr)); - prog_attr.prog_type = BPF_PROG_TYPE_XDP; - prog_attr.insns = insns; - prog_attr.insns_cnt = ARRAY_SIZE(insns); - prog_attr.license = "GPL"; - - prog_fd = bpf_load_program_xattr(&prog_attr, NULL, 0); + prog_fd = bpf_prog_load(BPF_PROG_TYPE_XDP, NULL, "GPL", insns, insn_cnt, NULL); if (prog_fd < 0) { close(map_fd); return detected; @@ -495,10 +488,13 @@ static int xsk_load_xdp_prog(struct xsk_socket *xsk) }; struct bpf_insn *progs[] = {prog, prog_redirect_flags}; enum xsk_prog option = get_xsk_prog(); + LIBBPF_OPTS(bpf_prog_load_opts, opts, + .log_buf = log_buf, + .log_size = log_buf_size, + ); - prog_fd = bpf_load_program(BPF_PROG_TYPE_XDP, progs[option], insns_cnt[option], - "LGPL-2.1 or BSD-2-Clause", 0, log_buf, - log_buf_size); + prog_fd = bpf_prog_load(BPF_PROG_TYPE_XDP, NULL, "LGPL-2.1 or BSD-2-Clause", + progs[option], insns_cnt[option], &opts); if (prog_fd < 0) { pr_warn("BPF log buffer:\n%s", log_buf); return prog_fd; @@ -725,14 +721,12 @@ static int xsk_link_lookup(int ifindex, __u32 *prog_id, int *link_fd) static bool xsk_probe_bpf_link(void) { - DECLARE_LIBBPF_OPTS(bpf_link_create_opts, opts, - .flags = XDP_FLAGS_SKB_MODE); - struct bpf_load_program_attr prog_attr; + LIBBPF_OPTS(bpf_link_create_opts, opts, .flags = XDP_FLAGS_SKB_MODE); struct bpf_insn insns[2] = { BPF_MOV64_IMM(BPF_REG_0, XDP_PASS), BPF_EXIT_INSN() }; - int prog_fd, link_fd = -1; + int prog_fd, link_fd = -1, insn_cnt = ARRAY_SIZE(insns); int ifindex_lo = 1; bool ret = false; int err; @@ -744,13 +738,7 @@ static bool xsk_probe_bpf_link(void) if (link_fd >= 0) return true; - memset(&prog_attr, 0, sizeof(prog_attr)); - prog_attr.prog_type = BPF_PROG_TYPE_XDP; - prog_attr.insns = insns; - prog_attr.insns_cnt = ARRAY_SIZE(insns); - prog_attr.license = "GPL"; - - prog_fd = bpf_load_program_xattr(&prog_attr, NULL, 0); + prog_fd = bpf_prog_load(BPF_PROG_TYPE_XDP, NULL, "GPL", insns, insn_cnt, NULL); if (prog_fd < 0) return ret; From bcc40fc0021d4b7c016f8bcf62bd4e21251fdee8 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:38 -0700 Subject: [PATCH 0081/2048] libbpf: Stop using to-be-deprecated APIs Remove all the internal uses of libbpf APIs that are slated to be deprecated in v0.7. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103220845.2676888-6-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 5751cade0f66..dfd15cc60ea7 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -7709,7 +7709,7 @@ int bpf_object__pin_maps(struct bpf_object *obj, const char *path) return 0; err_unpin_maps: - while ((map = bpf_map__prev(map, obj))) { + while ((map = bpf_object__prev_map(obj, map))) { if (!map->pin_path) continue; @@ -7789,7 +7789,7 @@ int bpf_object__pin_programs(struct bpf_object *obj, const char *path) return 0; err_unpin_programs: - while ((prog = bpf_program__prev(prog, obj))) { + while ((prog = bpf_object__prev_program(obj, prog))) { char buf[PATH_MAX]; int len; @@ -8130,9 +8130,11 @@ int bpf_program__set_autoload(struct bpf_program *prog, bool autoload) return 0; } +static int bpf_program_nth_fd(const struct bpf_program *prog, int n); + int bpf_program__fd(const struct bpf_program *prog) { - return bpf_program__nth_fd(prog, 0); + return bpf_program_nth_fd(prog, 0); } size_t bpf_program__size(const struct bpf_program *prog) @@ -8178,7 +8180,10 @@ int bpf_program__set_prep(struct bpf_program *prog, int nr_instances, return 0; } -int bpf_program__nth_fd(const struct bpf_program *prog, int n) +__attribute__((alias("bpf_program_nth_fd"))) +int bpf_program__nth_fd(const struct bpf_program *prog, int n); + +static int bpf_program_nth_fd(const struct bpf_program *prog, int n) { int fd; From a3c7c7e8050fc299b42fa3d89bac253a8dfa5c0c Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:39 -0700 Subject: [PATCH 0082/2048] bpftool: Stop using deprecated bpf_load_program() Switch to bpf_prog_load() instead. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103220845.2676888-7-andrii@kernel.org --- tools/bpf/bpftool/feature.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/bpf/bpftool/feature.c b/tools/bpf/bpftool/feature.c index ade44577688e..5397077d0d9e 100644 --- a/tools/bpf/bpftool/feature.c +++ b/tools/bpf/bpftool/feature.c @@ -467,7 +467,7 @@ static bool probe_bpf_syscall(const char *define_prefix) { bool res; - bpf_load_program(BPF_PROG_TYPE_UNSPEC, NULL, 0, NULL, 0, NULL, 0); + bpf_prog_load(BPF_PROG_TYPE_UNSPEC, NULL, NULL, NULL, 0, NULL); res = (errno != ENOSYS); print_bool_feature("have_bpf_syscall", From 5c5edcdebfcf3a95257b0d8ef27a60af0e0ea03a Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:40 -0700 Subject: [PATCH 0083/2048] libbpf: Remove deprecation attribute from struct bpf_prog_prep_result This deprecation annotation has no effect because for struct deprecation attribute has to be declared after struct definition. But instead of moving it to the end of struct definition, remove it. When deprecation will go in effect at libbpf v0.7, this deprecation attribute will cause libbpf's own source code compilation to trigger deprecation warnings, which is unavoidable because libbpf still has to support that API. So keep deprecation of APIs, but don't mark structs used in API as deprecated. Fixes: e21d585cb3db ("libbpf: Deprecate multi-instance bpf_program APIs") Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-8-andrii@kernel.org --- tools/lib/bpf/libbpf.h | 1 - 1 file changed, 1 deletion(-) diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index bbc828667b22..039058763173 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -431,7 +431,6 @@ bpf_program__attach_iter(const struct bpf_program *prog, * one instance. In this case bpf_program__fd(prog) is equal to * bpf_program__nth_fd(prog, 0). */ -LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_program__insns() for getting bpf_program instructions") struct bpf_prog_prep_result { /* * If not NULL, load new instruction array. From 3d1d62397f4a12dedee09727b26cd5a4b254ebb7 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:41 -0700 Subject: [PATCH 0084/2048] selftests/bpf: Fix non-strict SEC() program sections Fix few more SEC() definitions that were previously missed. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-9-andrii@kernel.org --- tools/testing/selftests/bpf/progs/test_l4lb.c | 2 +- tools/testing/selftests/bpf/progs/test_l4lb_noinline.c | 2 +- tools/testing/selftests/bpf/progs/test_map_lock.c | 2 +- tools/testing/selftests/bpf/progs/test_queue_stack_map.h | 2 +- tools/testing/selftests/bpf/progs/test_skb_ctx.c | 2 +- tools/testing/selftests/bpf/progs/test_spin_lock.c | 2 +- tools/testing/selftests/bpf/progs/test_tcp_estats.c | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/test_l4lb.c b/tools/testing/selftests/bpf/progs/test_l4lb.c index 04fee08863cb..c26057ec46dc 100644 --- a/tools/testing/selftests/bpf/progs/test_l4lb.c +++ b/tools/testing/selftests/bpf/progs/test_l4lb.c @@ -448,7 +448,7 @@ static __always_inline int process_packet(void *data, __u64 off, void *data_end, return bpf_redirect(ifindex, 0); } -SEC("l4lb-demo") +SEC("tc") int balancer_ingress(struct __sk_buff *ctx) { void *data_end = (void *)(long)ctx->data_end; diff --git a/tools/testing/selftests/bpf/progs/test_l4lb_noinline.c b/tools/testing/selftests/bpf/progs/test_l4lb_noinline.c index b9e2753f4f91..19e4d2071c60 100644 --- a/tools/testing/selftests/bpf/progs/test_l4lb_noinline.c +++ b/tools/testing/selftests/bpf/progs/test_l4lb_noinline.c @@ -447,7 +447,7 @@ static __noinline int process_packet(void *data, __u64 off, void *data_end, return bpf_redirect(ifindex, 0); } -SEC("l4lb-demo") +SEC("tc") int balancer_ingress(struct __sk_buff *ctx) { void *data_end = (void *)(long)ctx->data_end; diff --git a/tools/testing/selftests/bpf/progs/test_map_lock.c b/tools/testing/selftests/bpf/progs/test_map_lock.c index b5c07ae7b68f..acf073db9e8b 100644 --- a/tools/testing/selftests/bpf/progs/test_map_lock.c +++ b/tools/testing/selftests/bpf/progs/test_map_lock.c @@ -30,7 +30,7 @@ struct { __type(value, struct array_elem); } array_map SEC(".maps"); -SEC("map_lock_demo") +SEC("cgroup/skb") int bpf_map_lock_test(struct __sk_buff *skb) { struct hmap_elem zero = {}, *val; diff --git a/tools/testing/selftests/bpf/progs/test_queue_stack_map.h b/tools/testing/selftests/bpf/progs/test_queue_stack_map.h index 0fcd3ff0e38a..648e8cab7a23 100644 --- a/tools/testing/selftests/bpf/progs/test_queue_stack_map.h +++ b/tools/testing/selftests/bpf/progs/test_queue_stack_map.h @@ -24,7 +24,7 @@ struct { __uint(value_size, sizeof(__u32)); } map_out SEC(".maps"); -SEC("test") +SEC("tc") int _test(struct __sk_buff *skb) { void *data_end = (void *)(long)skb->data_end; diff --git a/tools/testing/selftests/bpf/progs/test_skb_ctx.c b/tools/testing/selftests/bpf/progs/test_skb_ctx.c index 1d61b36e6067..c482110cfc95 100644 --- a/tools/testing/selftests/bpf/progs/test_skb_ctx.c +++ b/tools/testing/selftests/bpf/progs/test_skb_ctx.c @@ -5,7 +5,7 @@ char _license[] SEC("license") = "GPL"; -SEC("skb_ctx") +SEC("tc") int process(struct __sk_buff *skb) { #pragma clang loop unroll(full) diff --git a/tools/testing/selftests/bpf/progs/test_spin_lock.c b/tools/testing/selftests/bpf/progs/test_spin_lock.c index 0d31a3b3505f..7e88309d3229 100644 --- a/tools/testing/selftests/bpf/progs/test_spin_lock.c +++ b/tools/testing/selftests/bpf/progs/test_spin_lock.c @@ -45,7 +45,7 @@ struct { #define CREDIT_PER_NS(delta, rate) (((delta) * rate) >> 20) -SEC("spin_lock_demo") +SEC("tc") int bpf_sping_lock_test(struct __sk_buff *skb) { volatile int credit = 0, max_credit = 100, pkt_len = 64; diff --git a/tools/testing/selftests/bpf/progs/test_tcp_estats.c b/tools/testing/selftests/bpf/progs/test_tcp_estats.c index 2c5c602c6011..e2ae049c2f85 100644 --- a/tools/testing/selftests/bpf/progs/test_tcp_estats.c +++ b/tools/testing/selftests/bpf/progs/test_tcp_estats.c @@ -244,7 +244,7 @@ static __always_inline void send_basic_event(struct sock *sk, bpf_map_update_elem(&ev_record_map, &key, &ev, BPF_ANY); } -SEC("dummy_tracepoint") +SEC("tp/dummy/tracepoint") int _dummy_tracepoint(struct dummy_tracepoint_args *arg) { if (!arg->sock) From d8e86407e5fc6c3da1e336f89bd3e9bbc1c0cf60 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:42 -0700 Subject: [PATCH 0085/2048] selftests/bpf: Convert legacy prog load APIs to bpf_prog_load() Convert all the uses of legacy low-level BPF program loading APIs (mostly bpf_load_program_xattr(), but also some bpf_verify_program()) to bpf_prog_load() uses. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211103220845.2676888-10-andrii@kernel.org --- .../testing/selftests/bpf/prog_tests/align.c | 11 ++++-- .../selftests/bpf/prog_tests/fexit_stress.c | 33 ++++++++-------- .../raw_tp_writable_reject_nbd_invalid.c | 14 +++---- .../bpf/prog_tests/raw_tp_writable_test_run.c | 29 +++++++------- .../selftests/bpf/prog_tests/sockopt.c | 19 +++++----- tools/testing/selftests/bpf/test_lru_map.c | 9 +---- tools/testing/selftests/bpf/test_sock.c | 19 +++++----- tools/testing/selftests/bpf/test_sock_addr.c | 13 +++---- tools/testing/selftests/bpf/test_sysctl.c | 22 ++++------- tools/testing/selftests/bpf/test_verifier.c | 38 +++++++++---------- 10 files changed, 97 insertions(+), 110 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/align.c b/tools/testing/selftests/bpf/prog_tests/align.c index 5861446d0777..837f67c6bfda 100644 --- a/tools/testing/selftests/bpf/prog_tests/align.c +++ b/tools/testing/selftests/bpf/prog_tests/align.c @@ -594,6 +594,12 @@ static int do_test_single(struct bpf_align_test *test) struct bpf_insn *prog = test->insns; int prog_type = test->prog_type; char bpf_vlog_copy[32768]; + LIBBPF_OPTS(bpf_prog_load_opts, opts, + .prog_flags = BPF_F_STRICT_ALIGNMENT, + .log_buf = bpf_vlog, + .log_size = sizeof(bpf_vlog), + .log_level = 2, + ); const char *line_ptr; int cur_line = -1; int prog_len, i; @@ -601,9 +607,8 @@ static int do_test_single(struct bpf_align_test *test) int ret; prog_len = probe_filter_length(prog); - fd_prog = bpf_verify_program(prog_type ? : BPF_PROG_TYPE_SOCKET_FILTER, - prog, prog_len, BPF_F_STRICT_ALIGNMENT, - "GPL", 0, bpf_vlog, sizeof(bpf_vlog), 2); + fd_prog = bpf_prog_load(prog_type ? : BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL", + prog, prog_len, &opts); if (fd_prog < 0 && test->result != REJECT) { printf("Failed to load program.\n"); printf("%s", bpf_vlog); diff --git a/tools/testing/selftests/bpf/prog_tests/fexit_stress.c b/tools/testing/selftests/bpf/prog_tests/fexit_stress.c index 7c9b62e971f1..e4cede6b4b2d 100644 --- a/tools/testing/selftests/bpf/prog_tests/fexit_stress.c +++ b/tools/testing/selftests/bpf/prog_tests/fexit_stress.c @@ -20,34 +20,33 @@ void test_fexit_stress(void) BPF_EXIT_INSN(), }; - struct bpf_load_program_attr load_attr = { - .prog_type = BPF_PROG_TYPE_TRACING, - .license = "GPL", - .insns = trace_program, - .insns_cnt = sizeof(trace_program) / sizeof(struct bpf_insn), + LIBBPF_OPTS(bpf_prog_load_opts, trace_opts, .expected_attach_type = BPF_TRACE_FEXIT, - }; + .log_buf = error, + .log_size = sizeof(error), + ); const struct bpf_insn skb_program[] = { BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), }; - struct bpf_load_program_attr skb_load_attr = { - .prog_type = BPF_PROG_TYPE_SOCKET_FILTER, - .license = "GPL", - .insns = skb_program, - .insns_cnt = sizeof(skb_program) / sizeof(struct bpf_insn), - }; + LIBBPF_OPTS(bpf_prog_load_opts, skb_opts, + .log_buf = error, + .log_size = sizeof(error), + ); err = libbpf_find_vmlinux_btf_id("bpf_fentry_test1", - load_attr.expected_attach_type); + trace_opts.expected_attach_type); if (CHECK(err <= 0, "find_vmlinux_btf_id", "failed: %d\n", err)) goto out; - load_attr.attach_btf_id = err; + trace_opts.attach_btf_id = err; for (i = 0; i < CNT; i++) { - fexit_fd[i] = bpf_load_program_xattr(&load_attr, error, sizeof(error)); + fexit_fd[i] = bpf_prog_load(BPF_PROG_TYPE_TRACING, NULL, "GPL", + trace_program, + sizeof(trace_program) / sizeof(struct bpf_insn), + &trace_opts); if (CHECK(fexit_fd[i] < 0, "fexit loaded", "failed: %d errno %d\n", fexit_fd[i], errno)) goto out; @@ -57,7 +56,9 @@ void test_fexit_stress(void) goto out; } - filter_fd = bpf_load_program_xattr(&skb_load_attr, error, sizeof(error)); + filter_fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL", + skb_program, sizeof(skb_program) / sizeof(struct bpf_insn), + &skb_opts); if (CHECK(filter_fd < 0, "test_program_loaded", "failed: %d errno %d\n", filter_fd, errno)) goto out; diff --git a/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c b/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c index 9807336a3016..e2f1445b0e10 100644 --- a/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c +++ b/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c @@ -18,15 +18,15 @@ void test_raw_tp_writable_reject_nbd_invalid(void) BPF_EXIT_INSN(), }; - struct bpf_load_program_attr load_attr = { - .prog_type = BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, - .license = "GPL v2", - .insns = program, - .insns_cnt = sizeof(program) / sizeof(struct bpf_insn), + LIBBPF_OPTS(bpf_prog_load_opts, opts, .log_level = 2, - }; + .log_buf = error, + .log_size = sizeof(error), + ); - bpf_fd = bpf_load_program_xattr(&load_attr, error, sizeof(error)); + bpf_fd = bpf_prog_load(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, NULL, "GPL v2", + program, sizeof(program) / sizeof(struct bpf_insn), + &opts); if (CHECK(bpf_fd < 0, "bpf_raw_tracepoint_writable load", "failed: %d errno %d\n", bpf_fd, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c b/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c index ddefa1192e5d..239baccabccb 100644 --- a/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c +++ b/tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c @@ -17,15 +17,15 @@ void serial_test_raw_tp_writable_test_run(void) BPF_EXIT_INSN(), }; - struct bpf_load_program_attr load_attr = { - .prog_type = BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, - .license = "GPL v2", - .insns = trace_program, - .insns_cnt = sizeof(trace_program) / sizeof(struct bpf_insn), + LIBBPF_OPTS(bpf_prog_load_opts, trace_opts, .log_level = 2, - }; + .log_buf = error, + .log_size = sizeof(error), + ); - int bpf_fd = bpf_load_program_xattr(&load_attr, error, sizeof(error)); + int bpf_fd = bpf_prog_load(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, NULL, "GPL v2", + trace_program, sizeof(trace_program) / sizeof(struct bpf_insn), + &trace_opts); if (CHECK(bpf_fd < 0, "bpf_raw_tracepoint_writable loaded", "failed: %d errno %d\n", bpf_fd, errno)) return; @@ -35,15 +35,14 @@ void serial_test_raw_tp_writable_test_run(void) BPF_EXIT_INSN(), }; - struct bpf_load_program_attr skb_load_attr = { - .prog_type = BPF_PROG_TYPE_SOCKET_FILTER, - .license = "GPL v2", - .insns = skb_program, - .insns_cnt = sizeof(skb_program) / sizeof(struct bpf_insn), - }; + LIBBPF_OPTS(bpf_prog_load_opts, skb_opts, + .log_buf = error, + .log_size = sizeof(error), + ); - int filter_fd = - bpf_load_program_xattr(&skb_load_attr, error, sizeof(error)); + int filter_fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL v2", + skb_program, sizeof(skb_program) / sizeof(struct bpf_insn), + &skb_opts); if (CHECK(filter_fd < 0, "test_program_loaded", "failed: %d errno %d\n", filter_fd, errno)) goto out_bpffd; diff --git a/tools/testing/selftests/bpf/prog_tests/sockopt.c b/tools/testing/selftests/bpf/prog_tests/sockopt.c index 3e8517a8395a..cd09f4c7dd92 100644 --- a/tools/testing/selftests/bpf/prog_tests/sockopt.c +++ b/tools/testing/selftests/bpf/prog_tests/sockopt.c @@ -852,22 +852,21 @@ static struct sockopt_test { static int load_prog(const struct bpf_insn *insns, enum bpf_attach_type expected_attach_type) { - struct bpf_load_program_attr attr = { - .prog_type = BPF_PROG_TYPE_CGROUP_SOCKOPT, + LIBBPF_OPTS(bpf_prog_load_opts, opts, .expected_attach_type = expected_attach_type, - .insns = insns, - .license = "GPL", .log_level = 2, - }; - int fd; + .log_buf = bpf_log_buf, + .log_size = sizeof(bpf_log_buf), + ); + int fd, insns_cnt = 0; for (; - insns[attr.insns_cnt].code != (BPF_JMP | BPF_EXIT); - attr.insns_cnt++) { + insns[insns_cnt].code != (BPF_JMP | BPF_EXIT); + insns_cnt++) { } - attr.insns_cnt++; + insns_cnt++; - fd = bpf_load_program_xattr(&attr, bpf_log_buf, sizeof(bpf_log_buf)); + fd = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SOCKOPT, NULL, "GPL", insns, insns_cnt, &opts); if (verbose && fd < 0) fprintf(stderr, "%s\n", bpf_log_buf); diff --git a/tools/testing/selftests/bpf/test_lru_map.c b/tools/testing/selftests/bpf/test_lru_map.c index 7e9049fa3edf..7f3d1d8460b4 100644 --- a/tools/testing/selftests/bpf/test_lru_map.c +++ b/tools/testing/selftests/bpf/test_lru_map.c @@ -42,7 +42,6 @@ static int create_map(int map_type, int map_flags, unsigned int size) static int bpf_map_lookup_elem_with_ref_bit(int fd, unsigned long long key, void *value) { - struct bpf_load_program_attr prog; struct bpf_create_map_attr map; struct bpf_insn insns[] = { BPF_LD_MAP_VALUE(BPF_REG_9, 0, 0), @@ -76,13 +75,7 @@ static int bpf_map_lookup_elem_with_ref_bit(int fd, unsigned long long key, insns[0].imm = mfd; - memset(&prog, 0, sizeof(prog)); - prog.prog_type = BPF_PROG_TYPE_SCHED_CLS; - prog.insns = insns; - prog.insns_cnt = ARRAY_SIZE(insns); - prog.license = "GPL"; - - pfd = bpf_load_program_xattr(&prog, NULL, 0); + pfd = bpf_prog_load(BPF_PROG_TYPE_SCHED_CLS, NULL, "GPL", insns, ARRAY_SIZE(insns), NULL); if (pfd < 0) { close(mfd); return -1; diff --git a/tools/testing/selftests/bpf/test_sock.c b/tools/testing/selftests/bpf/test_sock.c index 9613f7538840..e8edd3dd3ec2 100644 --- a/tools/testing/selftests/bpf/test_sock.c +++ b/tools/testing/selftests/bpf/test_sock.c @@ -328,18 +328,17 @@ static size_t probe_prog_length(const struct bpf_insn *fp) static int load_sock_prog(const struct bpf_insn *prog, enum bpf_attach_type attach_type) { - struct bpf_load_program_attr attr; - int ret; + LIBBPF_OPTS(bpf_prog_load_opts, opts); + int ret, insn_cnt; - memset(&attr, 0, sizeof(struct bpf_load_program_attr)); - attr.prog_type = BPF_PROG_TYPE_CGROUP_SOCK; - attr.expected_attach_type = attach_type; - attr.insns = prog; - attr.insns_cnt = probe_prog_length(attr.insns); - attr.license = "GPL"; - attr.log_level = 2; + insn_cnt = probe_prog_length(prog); - ret = bpf_load_program_xattr(&attr, bpf_log_buf, BPF_LOG_BUF_SIZE); + opts.expected_attach_type = attach_type; + opts.log_buf = bpf_log_buf; + opts.log_size = BPF_LOG_BUF_SIZE; + opts.log_level = 2; + + ret = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SOCK, NULL, "GPL", prog, insn_cnt, &opts); if (verbose && ret < 0) fprintf(stderr, "%s\n", bpf_log_buf); diff --git a/tools/testing/selftests/bpf/test_sock_addr.c b/tools/testing/selftests/bpf/test_sock_addr.c index aa3f185fcb89..05c9e4944c01 100644 --- a/tools/testing/selftests/bpf/test_sock_addr.c +++ b/tools/testing/selftests/bpf/test_sock_addr.c @@ -645,17 +645,14 @@ static int mk_sockaddr(int domain, const char *ip, unsigned short port, static int load_insns(const struct sock_addr_test *test, const struct bpf_insn *insns, size_t insns_cnt) { - struct bpf_load_program_attr load_attr; + LIBBPF_OPTS(bpf_prog_load_opts, opts); int ret; - memset(&load_attr, 0, sizeof(struct bpf_load_program_attr)); - load_attr.prog_type = BPF_PROG_TYPE_CGROUP_SOCK_ADDR; - load_attr.expected_attach_type = test->expected_attach_type; - load_attr.insns = insns; - load_attr.insns_cnt = insns_cnt; - load_attr.license = "GPL"; + opts.expected_attach_type = test->expected_attach_type; + opts.log_buf = bpf_log_buf; + opts.log_size = BPF_LOG_BUF_SIZE; - ret = bpf_load_program_xattr(&load_attr, bpf_log_buf, BPF_LOG_BUF_SIZE); + ret = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SOCK_ADDR, NULL, "GPL", insns, insns_cnt, &opts); if (ret < 0 && test->expected_result != LOAD_REJECT) { log_err(">>> Loading program error.\n" ">>> Verifier output:\n%s\n-------\n", bpf_log_buf); diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index a3bb6d399daa..4a395d7a8ea9 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -1435,14 +1435,10 @@ static int load_sysctl_prog_insns(struct sysctl_test *test, const char *sysctl_path) { struct bpf_insn *prog = test->insns; - struct bpf_load_program_attr attr; - int ret; + LIBBPF_OPTS(bpf_prog_load_opts, opts); + int ret, insn_cnt; - memset(&attr, 0, sizeof(struct bpf_load_program_attr)); - attr.prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL; - attr.insns = prog; - attr.insns_cnt = probe_prog_length(attr.insns); - attr.license = "GPL"; + insn_cnt = probe_prog_length(prog); if (test->fixup_value_insn) { char buf[128]; @@ -1465,7 +1461,10 @@ static int load_sysctl_prog_insns(struct sysctl_test *test, return -1; } - ret = bpf_load_program_xattr(&attr, bpf_log_buf, BPF_LOG_BUF_SIZE); + opts.log_buf = bpf_log_buf; + opts.log_size = BPF_LOG_BUF_SIZE; + + ret = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SYSCTL, NULL, "GPL", prog, insn_cnt, &opts); if (ret < 0 && test->result != LOAD_REJECT) { log_err(">>> Loading program error.\n" ">>> Verifier output:\n%s\n-------\n", bpf_log_buf); @@ -1476,15 +1475,10 @@ static int load_sysctl_prog_insns(struct sysctl_test *test, static int load_sysctl_prog_file(struct sysctl_test *test) { - struct bpf_prog_load_attr attr; struct bpf_object *obj; int prog_fd; - memset(&attr, 0, sizeof(struct bpf_prog_load_attr)); - attr.file = test->prog_file; - attr.prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL; - - if (bpf_prog_load_xattr(&attr, &obj, &prog_fd)) { + if (bpf_prog_test_load(test->prog_file, BPF_PROG_TYPE_CGROUP_SYSCTL, &obj, &prog_fd)) { if (test->result != LOAD_REJECT) log_err(">>> Loading program (%s) error.\n", test->prog_file); diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index 25afe423b3f0..e512b715a785 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -498,8 +498,7 @@ static int create_prog_dummy_simple(enum bpf_prog_type prog_type, int ret) BPF_EXIT_INSN(), }; - return bpf_load_program(prog_type, prog, - ARRAY_SIZE(prog), "GPL", 0, NULL, 0); + return bpf_prog_load(prog_type, NULL, "GPL", prog, ARRAY_SIZE(prog), NULL); } static int create_prog_dummy_loop(enum bpf_prog_type prog_type, int mfd, @@ -514,8 +513,7 @@ static int create_prog_dummy_loop(enum bpf_prog_type prog_type, int mfd, BPF_EXIT_INSN(), }; - return bpf_load_program(prog_type, prog, - ARRAY_SIZE(prog), "GPL", 0, NULL, 0); + return bpf_prog_load(prog_type, NULL, "GPL", prog, ARRAY_SIZE(prog), NULL); } static int create_prog_array(enum bpf_prog_type prog_type, uint32_t max_elem, @@ -1045,7 +1043,7 @@ static void do_test_single(struct bpf_test *test, bool unpriv, int fd_prog, expected_ret, alignment_prevented_execution; int prog_len, prog_type = test->prog_type; struct bpf_insn *prog = test->insns; - struct bpf_load_program_attr attr; + LIBBPF_OPTS(bpf_prog_load_opts, opts); int run_errs, run_successes; int map_fds[MAX_NR_MAPS]; const char *expected_err; @@ -1085,32 +1083,34 @@ static void do_test_single(struct bpf_test *test, bool unpriv, test->result_unpriv : test->result; expected_err = unpriv && test->errstr_unpriv ? test->errstr_unpriv : test->errstr; - memset(&attr, 0, sizeof(attr)); - attr.prog_type = prog_type; - attr.expected_attach_type = test->expected_attach_type; - attr.insns = prog; - attr.insns_cnt = prog_len; - attr.license = "GPL"; + + opts.expected_attach_type = test->expected_attach_type; if (verbose) - attr.log_level = 1; + opts.log_level = 1; else if (expected_ret == VERBOSE_ACCEPT) - attr.log_level = 2; + opts.log_level = 2; else - attr.log_level = 4; - attr.prog_flags = pflags; + opts.log_level = 4; + opts.prog_flags = pflags; if (prog_type == BPF_PROG_TYPE_TRACING && test->kfunc) { - attr.attach_btf_id = libbpf_find_vmlinux_btf_id(test->kfunc, - attr.expected_attach_type); - if (attr.attach_btf_id < 0) { + int attach_btf_id; + + attach_btf_id = libbpf_find_vmlinux_btf_id(test->kfunc, + opts.expected_attach_type); + if (attach_btf_id < 0) { printf("FAIL\nFailed to find BTF ID for '%s'!\n", test->kfunc); (*errors)++; return; } + + opts.attach_btf_id = attach_btf_id; } - fd_prog = bpf_load_program_xattr(&attr, bpf_vlog, sizeof(bpf_vlog)); + opts.log_buf = bpf_vlog; + opts.log_size = sizeof(bpf_vlog); + fd_prog = bpf_prog_load(prog_type, NULL, "GPL", prog, prog_len, &opts); saved_errno = errno; /* BPF_PROG_TYPE_TRACING requires more setup and From f87c1930ac2951d7fb3bacb523c24046c81015ed Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:43 -0700 Subject: [PATCH 0086/2048] selftests/bpf: Merge test_stub.c into testing_helpers.c Move testing prog and object load wrappers (bpf_prog_test_load and bpf_test_load_program) into testing_helpers.{c,h} and get rid of otherwise useless test_stub.c. Make testing_helpers.c available to non-test_progs binaries as well. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-11-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 31 ++++++----- tools/testing/selftests/bpf/test_stub.c | 44 --------------- tools/testing/selftests/bpf/testing_helpers.c | 55 +++++++++++++++++++ tools/testing/selftests/bpf/testing_helpers.h | 6 ++ 4 files changed, 77 insertions(+), 59 deletions(-) delete mode 100644 tools/testing/selftests/bpf/test_stub.c diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index c4497a4af3fe..5588c622d266 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -178,10 +178,6 @@ $(OUTPUT)/bpf_testmod.ko: $(VMLINUX_BTF) $(wildcard bpf_testmod/Makefile bpf_tes $(Q)$(MAKE) $(submake_extras) -C bpf_testmod $(Q)cp bpf_testmod/bpf_testmod.ko $@ -$(OUTPUT)/test_stub.o: test_stub.c $(BPFOBJ) - $(call msg,CC,,$@) - $(Q)$(CC) -c $(CFLAGS) -o $@ $< - DEFAULT_BPFTOOL := $(HOST_SCRATCH_DIR)/sbin/bpftool $(OUTPUT)/runqslower: $(BPFOBJ) | $(DEFAULT_BPFTOOL) $(RUNQSLOWER_OUTPUT) @@ -194,18 +190,23 @@ $(OUTPUT)/runqslower: $(BPFOBJ) | $(DEFAULT_BPFTOOL) $(RUNQSLOWER_OUTPUT) TEST_GEN_PROGS_EXTENDED += $(DEFAULT_BPFTOOL) -$(TEST_GEN_PROGS) $(TEST_GEN_PROGS_EXTENDED): $(OUTPUT)/test_stub.o $(BPFOBJ) +$(TEST_GEN_PROGS) $(TEST_GEN_PROGS_EXTENDED): $(BPFOBJ) -$(OUTPUT)/test_dev_cgroup: cgroup_helpers.c -$(OUTPUT)/test_skb_cgroup_id_user: cgroup_helpers.c -$(OUTPUT)/test_sock: cgroup_helpers.c -$(OUTPUT)/test_sock_addr: cgroup_helpers.c -$(OUTPUT)/test_sockmap: cgroup_helpers.c -$(OUTPUT)/test_tcpnotify_user: cgroup_helpers.c trace_helpers.c -$(OUTPUT)/get_cgroup_id_user: cgroup_helpers.c -$(OUTPUT)/test_cgroup_storage: cgroup_helpers.c -$(OUTPUT)/test_sock_fields: cgroup_helpers.c -$(OUTPUT)/test_sysctl: cgroup_helpers.c +$(OUTPUT)/test_dev_cgroup: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_skb_cgroup_id_user: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_sock: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_sock_addr: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_sockmap: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_tcpnotify_user: cgroup_helpers.c trace_helpers.c testing_helpers.o +$(OUTPUT)/get_cgroup_id_user: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_cgroup_storage: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_sock_fields: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_sysctl: cgroup_helpers.c testing_helpers.o +$(OUTPUT)/test_tag: testing_helpers.o +$(OUTPUT)/test_lirc_mode2_user: testing_helpers.o +$(OUTPUT)/xdping: testing_helpers.o +$(OUTPUT)/flow_dissector_load: testing_helpers.o +$(OUTPUT)/test_maps: testing_helpers.o BPFTOOL ?= $(DEFAULT_BPFTOOL) $(DEFAULT_BPFTOOL): $(wildcard $(BPFTOOLDIR)/*.[ch] $(BPFTOOLDIR)/Makefile) \ diff --git a/tools/testing/selftests/bpf/test_stub.c b/tools/testing/selftests/bpf/test_stub.c deleted file mode 100644 index 47e132726203..000000000000 --- a/tools/testing/selftests/bpf/test_stub.c +++ /dev/null @@ -1,44 +0,0 @@ -// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) -/* Copyright (C) 2019 Netronome Systems, Inc. */ - -#include -#include -#include - -int extra_prog_load_log_flags = 0; - -int bpf_prog_test_load(const char *file, enum bpf_prog_type type, - struct bpf_object **pobj, int *prog_fd) -{ - struct bpf_prog_load_attr attr; - - memset(&attr, 0, sizeof(struct bpf_prog_load_attr)); - attr.file = file; - attr.prog_type = type; - attr.expected_attach_type = 0; - attr.prog_flags = BPF_F_TEST_RND_HI32; - attr.log_level = extra_prog_load_log_flags; - - return bpf_prog_load_xattr(&attr, pobj, prog_fd); -} - -int bpf_test_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, - size_t insns_cnt, const char *license, - __u32 kern_version, char *log_buf, - size_t log_buf_sz) -{ - struct bpf_load_program_attr load_attr; - - memset(&load_attr, 0, sizeof(struct bpf_load_program_attr)); - load_attr.prog_type = type; - load_attr.expected_attach_type = 0; - load_attr.name = NULL; - load_attr.insns = insns; - load_attr.insns_cnt = insns_cnt; - load_attr.license = license; - load_attr.kern_version = kern_version; - load_attr.prog_flags = BPF_F_TEST_RND_HI32; - load_attr.log_level = extra_prog_load_log_flags; - - return bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz); -} diff --git a/tools/testing/selftests/bpf/testing_helpers.c b/tools/testing/selftests/bpf/testing_helpers.c index 800d503e5cb4..ef61d43adfe4 100644 --- a/tools/testing/selftests/bpf/testing_helpers.c +++ b/tools/testing/selftests/bpf/testing_helpers.c @@ -1,7 +1,11 @@ // SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +/* Copyright (C) 2019 Netronome Systems, Inc. */ /* Copyright (C) 2020 Facebook, Inc. */ #include +#include #include +#include +#include #include "testing_helpers.h" int parse_num_list(const char *s, bool **num_set, int *num_set_len) @@ -78,3 +82,54 @@ __u32 link_info_prog_id(const struct bpf_link *link, struct bpf_link_info *info) } return info->prog_id; } + +int extra_prog_load_log_flags = 0; + +int bpf_prog_test_load(const char *file, enum bpf_prog_type type, + struct bpf_object **pobj, int *prog_fd) +{ + struct bpf_object *obj; + struct bpf_program *prog; + int err; + + obj = bpf_object__open(file); + if (!obj) + return -errno; + + prog = bpf_object__next_program(obj, NULL); + if (!prog) { + err = -ENOENT; + goto err_out; + } + + if (type != BPF_PROG_TYPE_UNSPEC) + bpf_program__set_type(prog, type); + + err = bpf_object__load(obj); + if (err) + goto err_out; + + *pobj = obj; + *prog_fd = bpf_program__fd(prog); + + return 0; +err_out: + bpf_object__close(obj); + return err; +} + +int bpf_test_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, + size_t insns_cnt, const char *license, + __u32 kern_version, char *log_buf, + size_t log_buf_sz) +{ + LIBBPF_OPTS(bpf_prog_load_opts, opts, + .kern_version = kern_version, + .prog_flags = BPF_F_TEST_RND_HI32, + .log_level = extra_prog_load_log_flags, + .log_buf = log_buf, + .log_size = log_buf_sz, + ); + + return bpf_prog_load(type, NULL, license, insns, insns_cnt, &opts); +} diff --git a/tools/testing/selftests/bpf/testing_helpers.h b/tools/testing/selftests/bpf/testing_helpers.h index d4f8e749611b..f46ebc476ee8 100644 --- a/tools/testing/selftests/bpf/testing_helpers.h +++ b/tools/testing/selftests/bpf/testing_helpers.h @@ -6,3 +6,9 @@ int parse_num_list(const char *s, bool **set, int *set_len); __u32 link_info_prog_id(const struct bpf_link *link, struct bpf_link_info *info); +int bpf_prog_test_load(const char *file, enum bpf_prog_type type, + struct bpf_object **pobj, int *prog_fd); +int bpf_test_load_program(enum bpf_prog_type type, const struct bpf_insn *insns, + size_t insns_cnt, const char *license, + __u32 kern_version, char *log_buf, + size_t log_buf_sz); From cbdb1461dcf45765a036e9f6975ffe19e69bdc33 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:44 -0700 Subject: [PATCH 0087/2048] selftests/bpf: Use explicit bpf_prog_test_load() calls everywhere -Dbpf_prog_load_deprecated=bpf_prog_test_load trick is both ugly and breaks when deprecation goes into effect due to macro magic. Convert all the uses to explicit bpf_prog_test_load() calls which avoid deprecation errors and makes everything less magical. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-12-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 2 +- .../selftests/bpf/flow_dissector_load.h | 3 ++- .../testing/selftests/bpf/get_cgroup_id_user.c | 5 +++-- .../selftests/bpf/prog_tests/bpf_obj_id.c | 2 +- .../selftests/bpf/prog_tests/fexit_bpf2bpf.c | 8 ++++---- .../bpf/prog_tests/get_stack_raw_tp.c | 4 ++-- .../selftests/bpf/prog_tests/global_data.c | 2 +- .../bpf/prog_tests/global_func_args.c | 2 +- .../selftests/bpf/prog_tests/kfree_skb.c | 2 +- .../selftests/bpf/prog_tests/l4lb_all.c | 2 +- .../bpf/prog_tests/load_bytes_relative.c | 2 +- .../selftests/bpf/prog_tests/map_lock.c | 4 ++-- .../selftests/bpf/prog_tests/pkt_access.c | 2 +- .../selftests/bpf/prog_tests/pkt_md_access.c | 2 +- .../selftests/bpf/prog_tests/queue_stack_map.c | 2 +- .../testing/selftests/bpf/prog_tests/skb_ctx.c | 2 +- .../selftests/bpf/prog_tests/skb_helpers.c | 2 +- .../selftests/bpf/prog_tests/spinlock.c | 4 ++-- .../selftests/bpf/prog_tests/stacktrace_map.c | 2 +- .../bpf/prog_tests/stacktrace_map_raw_tp.c | 2 +- .../selftests/bpf/prog_tests/tailcalls.c | 18 +++++++++--------- .../bpf/prog_tests/task_fd_query_rawtp.c | 2 +- .../bpf/prog_tests/task_fd_query_tp.c | 4 ++-- .../selftests/bpf/prog_tests/tcp_estats.c | 2 +- .../selftests/bpf/prog_tests/tp_attach_query.c | 2 +- tools/testing/selftests/bpf/prog_tests/xdp.c | 2 +- .../selftests/bpf/prog_tests/xdp_adjust_tail.c | 6 +++--- .../selftests/bpf/prog_tests/xdp_attach.c | 6 +++--- .../selftests/bpf/prog_tests/xdp_info.c | 2 +- .../selftests/bpf/prog_tests/xdp_perf.c | 2 +- .../selftests/bpf/progs/fexit_bpf2bpf.c | 2 +- tools/testing/selftests/bpf/test_dev_cgroup.c | 3 ++- .../selftests/bpf/test_lirc_mode2_user.c | 6 ++++-- tools/testing/selftests/bpf/test_maps.c | 7 ++++--- tools/testing/selftests/bpf/test_sysctl.c | 1 + .../selftests/bpf/test_tcpnotify_user.c | 3 ++- tools/testing/selftests/bpf/xdping.c | 3 ++- 37 files changed, 68 insertions(+), 59 deletions(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 5588c622d266..2016c583ed20 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -24,7 +24,6 @@ SAN_CFLAGS ?= CFLAGS += -g -O0 -rdynamic -Wall $(GENFLAGS) $(SAN_CFLAGS) \ -I$(CURDIR) -I$(INCLUDE_DIR) -I$(GENDIR) -I$(LIBDIR) \ -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT) \ - -Dbpf_prog_load_deprecated=bpf_prog_test_load \ -Dbpf_load_program=bpf_test_load_program LDLIBS += -lcap -lelf -lz -lrt -lpthread @@ -207,6 +206,7 @@ $(OUTPUT)/test_lirc_mode2_user: testing_helpers.o $(OUTPUT)/xdping: testing_helpers.o $(OUTPUT)/flow_dissector_load: testing_helpers.o $(OUTPUT)/test_maps: testing_helpers.o +$(OUTPUT)/test_verifier: testing_helpers.o BPFTOOL ?= $(DEFAULT_BPFTOOL) $(DEFAULT_BPFTOOL): $(wildcard $(BPFTOOLDIR)/*.[ch] $(BPFTOOLDIR)/Makefile) \ diff --git a/tools/testing/selftests/bpf/flow_dissector_load.h b/tools/testing/selftests/bpf/flow_dissector_load.h index 9d0acc2fc6cc..f40b585f4e7e 100644 --- a/tools/testing/selftests/bpf/flow_dissector_load.h +++ b/tools/testing/selftests/bpf/flow_dissector_load.h @@ -4,6 +4,7 @@ #include #include +#include "testing_helpers.h" static inline int bpf_flow_load(struct bpf_object **obj, const char *path, @@ -18,7 +19,7 @@ static inline int bpf_flow_load(struct bpf_object **obj, int prog_array_fd; int ret, fd, i; - ret = bpf_prog_load(path, BPF_PROG_TYPE_FLOW_DISSECTOR, obj, + ret = bpf_prog_test_load(path, BPF_PROG_TYPE_FLOW_DISSECTOR, obj, prog_fd); if (ret) return ret; diff --git a/tools/testing/selftests/bpf/get_cgroup_id_user.c b/tools/testing/selftests/bpf/get_cgroup_id_user.c index 99628e1a1e58..3a7b82bd9e94 100644 --- a/tools/testing/selftests/bpf/get_cgroup_id_user.c +++ b/tools/testing/selftests/bpf/get_cgroup_id_user.c @@ -19,6 +19,7 @@ #include #include "cgroup_helpers.h" +#include "testing_helpers.h" #include "bpf_rlimit.h" #define CHECK(condition, tag, format...) ({ \ @@ -66,8 +67,8 @@ int main(int argc, char **argv) if (CHECK(cgroup_fd < 0, "cgroup_setup_and_join", "err %d errno %d\n", cgroup_fd, errno)) return 1; - err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); - if (CHECK(err, "bpf_prog_load", "err %d errno %d\n", err, errno)) + err = bpf_prog_test_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); + if (CHECK(err, "bpf_prog_test_load", "err %d errno %d\n", err, errno)) goto cleanup_cgroup_env; cgidmap_fd = bpf_find_map(__func__, obj, "cg_ids"); diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_obj_id.c b/tools/testing/selftests/bpf/prog_tests/bpf_obj_id.c index eb8eeebe6935..0a6c5f00abd4 100644 --- a/tools/testing/selftests/bpf/prog_tests/bpf_obj_id.c +++ b/tools/testing/selftests/bpf/prog_tests/bpf_obj_id.c @@ -48,7 +48,7 @@ void serial_test_bpf_obj_id(void) bzero(zeros, sizeof(zeros)); for (i = 0; i < nr_iters; i++) { now = time(NULL); - err = bpf_prog_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, + err = bpf_prog_test_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &objs[i], &prog_fds[i]); /* test_obj_id.o is a dumb prog. It should never fail * to load. diff --git a/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c b/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c index 9cff14a23bb7..fdd603ebda28 100644 --- a/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c +++ b/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c @@ -65,7 +65,7 @@ static void test_fexit_bpf2bpf_common(const char *obj_file, int err, tgt_fd, i; struct btf *btf; - err = bpf_prog_load(target_obj_file, BPF_PROG_TYPE_UNSPEC, + err = bpf_prog_test_load(target_obj_file, BPF_PROG_TYPE_UNSPEC, &tgt_obj, &tgt_fd); if (!ASSERT_OK(err, "tgt_prog_load")) return; @@ -224,7 +224,7 @@ static int test_second_attach(struct bpf_object *obj) if (CHECK(!prog, "find_prog", "prog %s not found\n", prog_name)) return -ENOENT; - err = bpf_prog_load(tgt_obj_file, BPF_PROG_TYPE_UNSPEC, + err = bpf_prog_test_load(tgt_obj_file, BPF_PROG_TYPE_UNSPEC, &tgt_obj, &tgt_fd); if (CHECK(err, "second_prog_load", "file %s err %d errno %d\n", tgt_obj_file, err, errno)) @@ -274,7 +274,7 @@ static void test_fmod_ret_freplace(void) __u32 duration = 0; int err, pkt_fd, attach_prog_fd; - err = bpf_prog_load(tgt_name, BPF_PROG_TYPE_UNSPEC, + err = bpf_prog_test_load(tgt_name, BPF_PROG_TYPE_UNSPEC, &pkt_obj, &pkt_fd); /* the target prog should load fine */ if (CHECK(err, "tgt_prog_load", "file %s err %d errno %d\n", @@ -341,7 +341,7 @@ static void test_obj_load_failure_common(const char *obj_file, int err, pkt_fd; __u32 duration = 0; - err = bpf_prog_load(target_obj_file, BPF_PROG_TYPE_UNSPEC, + err = bpf_prog_test_load(target_obj_file, BPF_PROG_TYPE_UNSPEC, &pkt_obj, &pkt_fd); /* the target prog should load fine */ if (CHECK(err, "tgt_prog_load", "file %s err %d errno %d\n", diff --git a/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c b/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c index 522237aa4470..569fcc6ed660 100644 --- a/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c +++ b/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c @@ -94,11 +94,11 @@ void test_get_stack_raw_tp(void) struct bpf_map *map; cpu_set_t cpu_set; - err = bpf_prog_load(file_err, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file_err, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); if (CHECK(err >= 0, "prog_load raw tp", "err %d errno %d\n", err, errno)) return; - err = bpf_prog_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); if (CHECK(err, "prog_load raw tp", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/global_data.c b/tools/testing/selftests/bpf/prog_tests/global_data.c index afd8639f9a94..9da131b32e13 100644 --- a/tools/testing/selftests/bpf/prog_tests/global_data.c +++ b/tools/testing/selftests/bpf/prog_tests/global_data.c @@ -136,7 +136,7 @@ void test_global_data(void) struct bpf_object *obj; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK(err, "load program", "error %d loading %s\n", err, file)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/global_func_args.c b/tools/testing/selftests/bpf/prog_tests/global_func_args.c index 8bcc2869102f..93a2439237b0 100644 --- a/tools/testing/selftests/bpf/prog_tests/global_func_args.c +++ b/tools/testing/selftests/bpf/prog_tests/global_func_args.c @@ -44,7 +44,7 @@ void test_global_func_args(void) struct bpf_object *obj; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); if (CHECK(err, "load program", "error %d loading %s\n", err, file)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/kfree_skb.c b/tools/testing/selftests/bpf/prog_tests/kfree_skb.c index 01e51d16c8b8..885413ed5c96 100644 --- a/tools/testing/selftests/bpf/prog_tests/kfree_skb.c +++ b/tools/testing/selftests/bpf/prog_tests/kfree_skb.c @@ -74,7 +74,7 @@ void serial_test_kfree_skb(void) const int zero = 0; bool test_ok[2]; - err = bpf_prog_load("./test_pkt_access.o", BPF_PROG_TYPE_SCHED_CLS, + err = bpf_prog_test_load("./test_pkt_access.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &tattr.prog_fd); if (CHECK(err, "prog_load sched cls", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/l4lb_all.c b/tools/testing/selftests/bpf/prog_tests/l4lb_all.c index 8073105548ff..540ef28fabff 100644 --- a/tools/testing/selftests/bpf/prog_tests/l4lb_all.c +++ b/tools/testing/selftests/bpf/prog_tests/l4lb_all.c @@ -30,7 +30,7 @@ static void test_l4lb(const char *file) char buf[128]; u32 *magic = (u32 *)buf; - err = bpf_prog_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/load_bytes_relative.c b/tools/testing/selftests/bpf/prog_tests/load_bytes_relative.c index 5a2a689dbb68..4e0b2ec057aa 100644 --- a/tools/testing/selftests/bpf/prog_tests/load_bytes_relative.c +++ b/tools/testing/selftests/bpf/prog_tests/load_bytes_relative.c @@ -27,7 +27,7 @@ void test_load_bytes_relative(void) if (CHECK_FAIL(server_fd < 0)) goto close_cgroup_fd; - err = bpf_prog_load("./load_bytes_relative.o", BPF_PROG_TYPE_CGROUP_SKB, + err = bpf_prog_test_load("./load_bytes_relative.o", BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); if (CHECK_FAIL(err)) goto close_server_fd; diff --git a/tools/testing/selftests/bpf/prog_tests/map_lock.c b/tools/testing/selftests/bpf/prog_tests/map_lock.c index ce17b1ed8709..23d19e9cf26a 100644 --- a/tools/testing/selftests/bpf/prog_tests/map_lock.c +++ b/tools/testing/selftests/bpf/prog_tests/map_lock.c @@ -53,9 +53,9 @@ void test_map_lock(void) int err = 0, key = 0, i; void *ret; - err = bpf_prog_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); if (CHECK_FAIL(err)) { - printf("test_map_lock:bpf_prog_load errno %d\n", errno); + printf("test_map_lock:bpf_prog_test_load errno %d\n", errno); goto close_prog; } map_fd[0] = bpf_find_map(__func__, obj, "hash_map"); diff --git a/tools/testing/selftests/bpf/prog_tests/pkt_access.c b/tools/testing/selftests/bpf/prog_tests/pkt_access.c index 44b514fabccd..6628710ec3c6 100644 --- a/tools/testing/selftests/bpf/prog_tests/pkt_access.c +++ b/tools/testing/selftests/bpf/prog_tests/pkt_access.c @@ -9,7 +9,7 @@ void test_pkt_access(void) __u32 duration, retval; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/pkt_md_access.c b/tools/testing/selftests/bpf/prog_tests/pkt_md_access.c index 939015cd6dba..c9d2d6a1bfcc 100644 --- a/tools/testing/selftests/bpf/prog_tests/pkt_md_access.c +++ b/tools/testing/selftests/bpf/prog_tests/pkt_md_access.c @@ -9,7 +9,7 @@ void test_pkt_md_access(void) __u32 duration, retval; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/queue_stack_map.c b/tools/testing/selftests/bpf/prog_tests/queue_stack_map.c index f47e7b1cb32c..8ccba3ab70ee 100644 --- a/tools/testing/selftests/bpf/prog_tests/queue_stack_map.c +++ b/tools/testing/selftests/bpf/prog_tests/queue_stack_map.c @@ -27,7 +27,7 @@ static void test_queue_stack_map_by_type(int type) else return; - err = bpf_prog_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/skb_ctx.c b/tools/testing/selftests/bpf/prog_tests/skb_ctx.c index c437e6ba8fe2..d3106078838c 100644 --- a/tools/testing/selftests/bpf/prog_tests/skb_ctx.c +++ b/tools/testing/selftests/bpf/prog_tests/skb_ctx.c @@ -32,7 +32,7 @@ void test_skb_ctx(void) int err; int i; - err = bpf_prog_load("./test_skb_ctx.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("./test_skb_ctx.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &tattr.prog_fd); if (CHECK_ATTR(err, "load", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/skb_helpers.c b/tools/testing/selftests/bpf/prog_tests/skb_helpers.c index f302ad84a298..6f802a1c0800 100644 --- a/tools/testing/selftests/bpf/prog_tests/skb_helpers.c +++ b/tools/testing/selftests/bpf/prog_tests/skb_helpers.c @@ -20,7 +20,7 @@ void test_skb_helpers(void) struct bpf_object *obj; int err; - err = bpf_prog_load("./test_skb_helpers.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("./test_skb_helpers.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &tattr.prog_fd); if (CHECK_ATTR(err, "load", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/spinlock.c b/tools/testing/selftests/bpf/prog_tests/spinlock.c index 7577a77a4c4c..6307f5d2b417 100644 --- a/tools/testing/selftests/bpf/prog_tests/spinlock.c +++ b/tools/testing/selftests/bpf/prog_tests/spinlock.c @@ -24,9 +24,9 @@ void test_spinlock(void) int err = 0, i; void *ret; - err = bpf_prog_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_CGROUP_SKB, &obj, &prog_fd); if (CHECK_FAIL(err)) { - printf("test_spin_lock:bpf_prog_load errno %d\n", errno); + printf("test_spin_lock:bpf_prog_test_load errno %d\n", errno); goto close_prog; } for (i = 0; i < 4; i++) diff --git a/tools/testing/selftests/bpf/prog_tests/stacktrace_map.c b/tools/testing/selftests/bpf/prog_tests/stacktrace_map.c index 04b476bd62b9..337493d74ec5 100644 --- a/tools/testing/selftests/bpf/prog_tests/stacktrace_map.c +++ b/tools/testing/selftests/bpf/prog_tests/stacktrace_map.c @@ -12,7 +12,7 @@ void test_stacktrace_map(void) struct bpf_object *obj; struct bpf_link *link; - err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); if (CHECK(err, "prog_load", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/stacktrace_map_raw_tp.c b/tools/testing/selftests/bpf/prog_tests/stacktrace_map_raw_tp.c index 4fd30bb651ad..063a14a2060d 100644 --- a/tools/testing/selftests/bpf/prog_tests/stacktrace_map_raw_tp.c +++ b/tools/testing/selftests/bpf/prog_tests/stacktrace_map_raw_tp.c @@ -12,7 +12,7 @@ void test_stacktrace_map_raw_tp(void) struct bpf_object *obj; struct bpf_link *link = NULL; - err = bpf_prog_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); if (CHECK(err, "prog_load raw tp", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/tailcalls.c b/tools/testing/selftests/bpf/prog_tests/tailcalls.c index 9825f1f7bfcc..5dc0f425bd11 100644 --- a/tools/testing/selftests/bpf/prog_tests/tailcalls.c +++ b/tools/testing/selftests/bpf/prog_tests/tailcalls.c @@ -16,7 +16,7 @@ static void test_tailcall_1(void) char prog_name[32]; char buff[128] = {}; - err = bpf_prog_load("tailcall1.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("tailcall1.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -154,7 +154,7 @@ static void test_tailcall_2(void) char prog_name[32]; char buff[128] = {}; - err = bpf_prog_load("tailcall2.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("tailcall2.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -228,7 +228,7 @@ static void test_tailcall_count(const char *which) __u32 retval, duration; char buff[128] = {}; - err = bpf_prog_load(which, BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load(which, BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -324,7 +324,7 @@ static void test_tailcall_4(void) char buff[128] = {}; char prog_name[32]; - err = bpf_prog_load("tailcall4.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("tailcall4.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -412,7 +412,7 @@ static void test_tailcall_5(void) char buff[128] = {}; char prog_name[32]; - err = bpf_prog_load("tailcall5.o", BPF_PROG_TYPE_SCHED_CLS, &obj, + err = bpf_prog_test_load("tailcall5.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -498,7 +498,7 @@ static void test_tailcall_bpf2bpf_1(void) __u32 retval, duration; char prog_name[32]; - err = bpf_prog_load("tailcall_bpf2bpf1.o", BPF_PROG_TYPE_SCHED_CLS, + err = bpf_prog_test_load("tailcall_bpf2bpf1.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -582,7 +582,7 @@ static void test_tailcall_bpf2bpf_2(void) __u32 retval, duration; char buff[128] = {}; - err = bpf_prog_load("tailcall_bpf2bpf2.o", BPF_PROG_TYPE_SCHED_CLS, + err = bpf_prog_test_load("tailcall_bpf2bpf2.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -660,7 +660,7 @@ static void test_tailcall_bpf2bpf_3(void) __u32 retval, duration; char prog_name[32]; - err = bpf_prog_load("tailcall_bpf2bpf3.o", BPF_PROG_TYPE_SCHED_CLS, + err = bpf_prog_test_load("tailcall_bpf2bpf3.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -757,7 +757,7 @@ static void test_tailcall_bpf2bpf_4(bool noise) __u32 retval, duration; char prog_name[32]; - err = bpf_prog_load("tailcall_bpf2bpf4.o", BPF_PROG_TYPE_SCHED_CLS, + err = bpf_prog_test_load("tailcall_bpf2bpf4.o", BPF_PROG_TYPE_SCHED_CLS, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/task_fd_query_rawtp.c b/tools/testing/selftests/bpf/prog_tests/task_fd_query_rawtp.c index 1bdc1d86a50c..17947c9e1d66 100644 --- a/tools/testing/selftests/bpf/prog_tests/task_fd_query_rawtp.c +++ b/tools/testing/selftests/bpf/prog_tests/task_fd_query_rawtp.c @@ -11,7 +11,7 @@ void test_task_fd_query_rawtp(void) __u32 duration = 0; char buf[256]; - err = bpf_prog_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_RAW_TRACEPOINT, &obj, &prog_fd); if (CHECK(err, "prog_load raw tp", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/task_fd_query_tp.c b/tools/testing/selftests/bpf/prog_tests/task_fd_query_tp.c index 3f131b8fe328..c2a98a7a8dfc 100644 --- a/tools/testing/selftests/bpf/prog_tests/task_fd_query_tp.c +++ b/tools/testing/selftests/bpf/prog_tests/task_fd_query_tp.c @@ -13,8 +13,8 @@ static void test_task_fd_query_tp_core(const char *probe_name, __u32 duration = 0; char buf[256]; - err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); - if (CHECK(err, "bpf_prog_load", "err %d errno %d\n", err, errno)) + err = bpf_prog_test_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); + if (CHECK(err, "bpf_prog_test_load", "err %d errno %d\n", err, errno)) goto close_prog; snprintf(buf, sizeof(buf), diff --git a/tools/testing/selftests/bpf/prog_tests/tcp_estats.c b/tools/testing/selftests/bpf/prog_tests/tcp_estats.c index 594307dffd13..11bf755be4c9 100644 --- a/tools/testing/selftests/bpf/prog_tests/tcp_estats.c +++ b/tools/testing/selftests/bpf/prog_tests/tcp_estats.c @@ -8,7 +8,7 @@ void test_tcp_estats(void) struct bpf_object *obj; __u32 duration = 0; - err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd); CHECK(err, "", "err %d errno %d\n", err, errno); if (err) return; diff --git a/tools/testing/selftests/bpf/prog_tests/tp_attach_query.c b/tools/testing/selftests/bpf/prog_tests/tp_attach_query.c index 8652d0a46c87..39e79291c82b 100644 --- a/tools/testing/selftests/bpf/prog_tests/tp_attach_query.c +++ b/tools/testing/selftests/bpf/prog_tests/tp_attach_query.c @@ -35,7 +35,7 @@ void serial_test_tp_attach_query(void) query = malloc(sizeof(*query) + sizeof(__u32) * num_progs); for (i = 0; i < num_progs; i++) { - err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj[i], + err = bpf_prog_test_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj[i], &prog_fd[i]); if (CHECK(err, "prog_load", "err %d errno %d\n", err, errno)) goto cleanup1; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp.c b/tools/testing/selftests/bpf/prog_tests/xdp.c index 48921ff74850..7a7ef9d4e151 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp.c @@ -16,7 +16,7 @@ void test_xdp(void) __u32 duration, retval, size; int err, prog_fd, map_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_adjust_tail.c b/tools/testing/selftests/bpf/prog_tests/xdp_adjust_tail.c index f529e3c923ae..3f5a17c38be5 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp_adjust_tail.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp_adjust_tail.c @@ -10,7 +10,7 @@ static void test_xdp_adjust_tail_shrink(void) int err, prog_fd; char buf[128]; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -38,7 +38,7 @@ static void test_xdp_adjust_tail_grow(void) __u32 duration, retval, size, expect_sz; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); if (CHECK_FAIL(err)) return; @@ -75,7 +75,7 @@ static void test_xdp_adjust_tail_grow2(void) .data_size_out = 0, /* Per test */ }; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &tattr.prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &tattr.prog_fd); if (CHECK_ATTR(err, "load", "err %d errno %d\n", err, errno)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_attach.c b/tools/testing/selftests/bpf/prog_tests/xdp_attach.c index 4c4057262cd8..c6fa390e3aa1 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp_attach.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp_attach.c @@ -16,7 +16,7 @@ void serial_test_xdp_attach(void) len = sizeof(info); - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj1, &fd1); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj1, &fd1); if (CHECK_FAIL(err)) return; err = bpf_obj_get_info_by_fd(fd1, &info, &len); @@ -24,7 +24,7 @@ void serial_test_xdp_attach(void) goto out_1; id1 = info.id; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj2, &fd2); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj2, &fd2); if (CHECK_FAIL(err)) goto out_1; @@ -34,7 +34,7 @@ void serial_test_xdp_attach(void) goto out_2; id2 = info.id; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj3, &fd3); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj3, &fd3); if (CHECK_FAIL(err)) goto out_2; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_info.c b/tools/testing/selftests/bpf/prog_tests/xdp_info.c index 4e2a4fd56f67..abe48e82e1dc 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp_info.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp_info.c @@ -29,7 +29,7 @@ void serial_test_xdp_info(void) /* Setup prog */ - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_perf.c b/tools/testing/selftests/bpf/prog_tests/xdp_perf.c index 7185bee16fe4..15a3900e4370 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp_perf.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp_perf.c @@ -9,7 +9,7 @@ void test_xdp_perf(void) char in[128], out[128]; int err, prog_fd; - err = bpf_prog_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); + err = bpf_prog_test_load(file, BPF_PROG_TYPE_XDP, &obj, &prog_fd); if (CHECK_FAIL(err)) return; diff --git a/tools/testing/selftests/bpf/progs/fexit_bpf2bpf.c b/tools/testing/selftests/bpf/progs/fexit_bpf2bpf.c index 49a84a3a2306..48cd14b43741 100644 --- a/tools/testing/selftests/bpf/progs/fexit_bpf2bpf.c +++ b/tools/testing/selftests/bpf/progs/fexit_bpf2bpf.c @@ -73,7 +73,7 @@ int test_subprog2(struct args_subprog2 *ctx) __builtin_preserve_access_index(&skb->len)); ret = ctx->ret; - /* bpf_prog_load() loads "test_pkt_access.o" with BPF_F_TEST_RND_HI32 + /* bpf_prog_test_load() loads "test_pkt_access.o" with BPF_F_TEST_RND_HI32 * which randomizes upper 32 bits after BPF_ALU32 insns. * Hence after 'w0 <<= 1' upper bits of $rax are random. * That is expected and correct. Trim them. diff --git a/tools/testing/selftests/bpf/test_dev_cgroup.c b/tools/testing/selftests/bpf/test_dev_cgroup.c index 804dddd97d4c..c299d3452695 100644 --- a/tools/testing/selftests/bpf/test_dev_cgroup.c +++ b/tools/testing/selftests/bpf/test_dev_cgroup.c @@ -14,6 +14,7 @@ #include #include "cgroup_helpers.h" +#include "testing_helpers.h" #include "bpf_rlimit.h" #define DEV_CGROUP_PROG "./dev_cgroup.o" @@ -27,7 +28,7 @@ int main(int argc, char **argv) int prog_fd, cgroup_fd; __u32 prog_cnt; - if (bpf_prog_load(DEV_CGROUP_PROG, BPF_PROG_TYPE_CGROUP_DEVICE, + if (bpf_prog_test_load(DEV_CGROUP_PROG, BPF_PROG_TYPE_CGROUP_DEVICE, &obj, &prog_fd)) { printf("Failed to load DEV_CGROUP program\n"); goto out; diff --git a/tools/testing/selftests/bpf/test_lirc_mode2_user.c b/tools/testing/selftests/bpf/test_lirc_mode2_user.c index fb5fd6841ef3..ebf68dce5504 100644 --- a/tools/testing/selftests/bpf/test_lirc_mode2_user.c +++ b/tools/testing/selftests/bpf/test_lirc_mode2_user.c @@ -45,6 +45,8 @@ #include #include +#include "testing_helpers.h" + int main(int argc, char **argv) { struct bpf_object *obj; @@ -58,8 +60,8 @@ int main(int argc, char **argv) return 2; } - ret = bpf_prog_load("test_lirc_mode2_kern.o", - BPF_PROG_TYPE_LIRC_MODE2, &obj, &progfd); + ret = bpf_prog_test_load("test_lirc_mode2_kern.o", + BPF_PROG_TYPE_LIRC_MODE2, &obj, &progfd); if (ret) { printf("Failed to load bpf program\n"); return 1; diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c index c7a36a9378f8..8b31bc1a801d 100644 --- a/tools/testing/selftests/bpf/test_maps.c +++ b/tools/testing/selftests/bpf/test_maps.c @@ -25,6 +25,7 @@ #include "bpf_util.h" #include "bpf_rlimit.h" #include "test_maps.h" +#include "testing_helpers.h" #ifndef ENOTSUPP #define ENOTSUPP 524 @@ -830,21 +831,21 @@ static void test_sockmap(unsigned int tasks, void *data) } /* Load SK_SKB program and Attach */ - err = bpf_prog_load(SOCKMAP_PARSE_PROG, + err = bpf_prog_test_load(SOCKMAP_PARSE_PROG, BPF_PROG_TYPE_SK_SKB, &obj, &parse_prog); if (err) { printf("Failed to load SK_SKB parse prog\n"); goto out_sockmap; } - err = bpf_prog_load(SOCKMAP_TCP_MSG_PROG, + err = bpf_prog_test_load(SOCKMAP_TCP_MSG_PROG, BPF_PROG_TYPE_SK_MSG, &obj, &msg_prog); if (err) { printf("Failed to load SK_SKB msg prog\n"); goto out_sockmap; } - err = bpf_prog_load(SOCKMAP_VERDICT_PROG, + err = bpf_prog_test_load(SOCKMAP_VERDICT_PROG, BPF_PROG_TYPE_SK_SKB, &obj, &verdict_prog); if (err) { printf("Failed to load SK_SKB verdict prog\n"); diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 4a395d7a8ea9..4f6cf833b522 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -17,6 +17,7 @@ #include "bpf_rlimit.h" #include "bpf_util.h" #include "cgroup_helpers.h" +#include "testing_helpers.h" #define CG_PATH "/foo" #define MAX_INSNS 512 diff --git a/tools/testing/selftests/bpf/test_tcpnotify_user.c b/tools/testing/selftests/bpf/test_tcpnotify_user.c index 4a39304cc5a6..63111cb082fe 100644 --- a/tools/testing/selftests/bpf/test_tcpnotify_user.c +++ b/tools/testing/selftests/bpf/test_tcpnotify_user.c @@ -25,6 +25,7 @@ #include "test_tcpnotify.h" #include "trace_helpers.h" +#include "testing_helpers.h" #define SOCKET_BUFFER_SIZE (getpagesize() < 8192L ? getpagesize() : 8192L) @@ -92,7 +93,7 @@ int main(int argc, char **argv) if (cg_fd < 0) goto err; - if (bpf_prog_load(file, BPF_PROG_TYPE_SOCK_OPS, &obj, &prog_fd)) { + if (bpf_prog_test_load(file, BPF_PROG_TYPE_SOCK_OPS, &obj, &prog_fd)) { printf("FAILED: load_bpf_file failed for: %s\n", file); goto err; } diff --git a/tools/testing/selftests/bpf/xdping.c b/tools/testing/selftests/bpf/xdping.c index 30f12637f4e4..baa870a759a2 100644 --- a/tools/testing/selftests/bpf/xdping.c +++ b/tools/testing/selftests/bpf/xdping.c @@ -22,6 +22,7 @@ #include "bpf/libbpf.h" #include "xdping.h" +#include "testing_helpers.h" static int ifindex; static __u32 xdp_flags = XDP_FLAGS_UPDATE_IF_NOEXIST; @@ -173,7 +174,7 @@ int main(int argc, char **argv) snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]); - if (bpf_prog_load(filename, BPF_PROG_TYPE_XDP, &obj, &prog_fd)) { + if (bpf_prog_test_load(filename, BPF_PROG_TYPE_XDP, &obj, &prog_fd)) { fprintf(stderr, "load of %s failed\n", filename); return 1; } From f19ddfe0360a1aa64db0b4a41f59e1ade3f6d288 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 3 Nov 2021 15:08:45 -0700 Subject: [PATCH 0088/2048] selftests/bpf: Use explicit bpf_test_load_program() helper calls Remove the second part of prog loading testing helper re-definition: -Dbpf_load_program=bpf_test_load_program This completes the clean up of deprecated libbpf program loading APIs. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Acked-by: Dave Marchevsky Link: https://lore.kernel.org/bpf/20211103220845.2676888-13-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 3 +-- .../selftests/bpf/prog_tests/cgroup_attach_autodetach.c | 2 +- tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c | 2 +- .../testing/selftests/bpf/prog_tests/cgroup_attach_override.c | 2 +- .../selftests/bpf/prog_tests/flow_dissector_load_bytes.c | 2 +- .../selftests/bpf/prog_tests/flow_dissector_reattach.c | 4 ++-- tools/testing/selftests/bpf/prog_tests/signal_pending.c | 2 +- tools/testing/selftests/bpf/test_cgroup_storage.c | 3 ++- tools/testing/selftests/bpf/test_tag.c | 3 ++- 9 files changed, 12 insertions(+), 11 deletions(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 2016c583ed20..e19cc6473936 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -23,8 +23,7 @@ BPF_GCC ?= $(shell command -v bpf-gcc;) SAN_CFLAGS ?= CFLAGS += -g -O0 -rdynamic -Wall $(GENFLAGS) $(SAN_CFLAGS) \ -I$(CURDIR) -I$(INCLUDE_DIR) -I$(GENDIR) -I$(LIBDIR) \ - -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT) \ - -Dbpf_load_program=bpf_test_load_program + -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT) LDLIBS += -lcap -lelf -lz -lrt -lpthread # Silence some warnings when compiled with clang diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_autodetach.c b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_autodetach.c index 5de485c7370f..858916d11e2e 100644 --- a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_autodetach.c +++ b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_autodetach.c @@ -16,7 +16,7 @@ static int prog_load(void) }; size_t insns_cnt = sizeof(prog) / sizeof(struct bpf_insn); - return bpf_load_program(BPF_PROG_TYPE_CGROUP_SKB, + return bpf_test_load_program(BPF_PROG_TYPE_CGROUP_SKB, prog, insns_cnt, "GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE); } diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c index 731bea84d8ed..de9c3e12b0ea 100644 --- a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c +++ b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c @@ -66,7 +66,7 @@ static int prog_load_cnt(int verdict, int val) size_t insns_cnt = sizeof(prog) / sizeof(struct bpf_insn); int ret; - ret = bpf_load_program(BPF_PROG_TYPE_CGROUP_SKB, + ret = bpf_test_load_program(BPF_PROG_TYPE_CGROUP_SKB, prog, insns_cnt, "GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE); diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_override.c b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_override.c index 10d3c33821a7..356547e849e2 100644 --- a/tools/testing/selftests/bpf/prog_tests/cgroup_attach_override.c +++ b/tools/testing/selftests/bpf/prog_tests/cgroup_attach_override.c @@ -18,7 +18,7 @@ static int prog_load(int verdict) }; size_t insns_cnt = sizeof(prog) / sizeof(struct bpf_insn); - return bpf_load_program(BPF_PROG_TYPE_CGROUP_SKB, + return bpf_test_load_program(BPF_PROG_TYPE_CGROUP_SKB, prog, insns_cnt, "GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE); } diff --git a/tools/testing/selftests/bpf/prog_tests/flow_dissector_load_bytes.c b/tools/testing/selftests/bpf/prog_tests/flow_dissector_load_bytes.c index 6093728497c7..93ac3f28226c 100644 --- a/tools/testing/selftests/bpf/prog_tests/flow_dissector_load_bytes.c +++ b/tools/testing/selftests/bpf/prog_tests/flow_dissector_load_bytes.c @@ -30,7 +30,7 @@ void serial_test_flow_dissector_load_bytes(void) /* make sure bpf_skb_load_bytes is not allowed from skb-less context */ - fd = bpf_load_program(BPF_PROG_TYPE_FLOW_DISSECTOR, prog, + fd = bpf_test_load_program(BPF_PROG_TYPE_FLOW_DISSECTOR, prog, ARRAY_SIZE(prog), "GPL", 0, NULL, 0); CHECK(fd < 0, "flow_dissector-bpf_skb_load_bytes-load", diff --git a/tools/testing/selftests/bpf/prog_tests/flow_dissector_reattach.c b/tools/testing/selftests/bpf/prog_tests/flow_dissector_reattach.c index f0c6c226aba8..7c79462d2702 100644 --- a/tools/testing/selftests/bpf/prog_tests/flow_dissector_reattach.c +++ b/tools/testing/selftests/bpf/prog_tests/flow_dissector_reattach.c @@ -47,9 +47,9 @@ static int load_prog(enum bpf_prog_type type) }; int fd; - fd = bpf_load_program(type, prog, ARRAY_SIZE(prog), "GPL", 0, NULL, 0); + fd = bpf_test_load_program(type, prog, ARRAY_SIZE(prog), "GPL", 0, NULL, 0); if (CHECK_FAIL(fd < 0)) - perror("bpf_load_program"); + perror("bpf_test_load_program"); return fd; } diff --git a/tools/testing/selftests/bpf/prog_tests/signal_pending.c b/tools/testing/selftests/bpf/prog_tests/signal_pending.c index fdfdcff6cbef..aecfe662c070 100644 --- a/tools/testing/selftests/bpf/prog_tests/signal_pending.c +++ b/tools/testing/selftests/bpf/prog_tests/signal_pending.c @@ -22,7 +22,7 @@ static void test_signal_pending_by_type(enum bpf_prog_type prog_type) prog[i] = BPF_ALU64_IMM(BPF_MOV, BPF_REG_0, 0); prog[ARRAY_SIZE(prog) - 1] = BPF_EXIT_INSN(); - prog_fd = bpf_load_program(prog_type, prog, ARRAY_SIZE(prog), + prog_fd = bpf_test_load_program(prog_type, prog, ARRAY_SIZE(prog), "GPL", 0, NULL, 0); CHECK(prog_fd < 0, "test-run", "errno %d\n", errno); diff --git a/tools/testing/selftests/bpf/test_cgroup_storage.c b/tools/testing/selftests/bpf/test_cgroup_storage.c index 0cda61da5d39..a63787e7bb1a 100644 --- a/tools/testing/selftests/bpf/test_cgroup_storage.c +++ b/tools/testing/selftests/bpf/test_cgroup_storage.c @@ -8,6 +8,7 @@ #include "bpf_rlimit.h" #include "cgroup_helpers.h" +#include "testing_helpers.h" char bpf_log_buf[BPF_LOG_BUF_SIZE]; @@ -66,7 +67,7 @@ int main(int argc, char **argv) prog[0].imm = percpu_map_fd; prog[7].imm = map_fd; - prog_fd = bpf_load_program(BPF_PROG_TYPE_CGROUP_SKB, + prog_fd = bpf_test_load_program(BPF_PROG_TYPE_CGROUP_SKB, prog, insns_cnt, "GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE); if (prog_fd < 0) { diff --git a/tools/testing/selftests/bpf/test_tag.c b/tools/testing/selftests/bpf/test_tag.c index 6272c784ca2a..5c7bea525626 100644 --- a/tools/testing/selftests/bpf/test_tag.c +++ b/tools/testing/selftests/bpf/test_tag.c @@ -21,6 +21,7 @@ #include "../../../include/linux/filter.h" #include "bpf_rlimit.h" +#include "testing_helpers.h" static struct bpf_insn prog[BPF_MAXINSNS]; @@ -57,7 +58,7 @@ static int bpf_try_load_prog(int insns, int fd_map, int fd_prog; bpf_filler(insns, fd_map); - fd_prog = bpf_load_program(BPF_PROG_TYPE_SCHED_CLS, prog, insns, "", 0, + fd_prog = bpf_test_load_program(BPF_PROG_TYPE_SCHED_CLS, prog, insns, "", 0, NULL, 0); assert(fd_prog > 0); if (fd_map > 0) From 2a2cb45b727b7a1041f3d3d93414b774e66454bb Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:13 -0800 Subject: [PATCH 0089/2048] selftests/bpf: Pass sanitizer flags to linker through LDFLAGS When adding -fsanitize=address to SAN_CFLAGS, it has to be passed both to compiler through CFLAGS as well as linker through LDFLAGS. Add SAN_CFLAGS into LDFLAGS to allow building selftests with ASAN. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-2-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index e19cc6473936..0468ea57650d 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -24,6 +24,7 @@ SAN_CFLAGS ?= CFLAGS += -g -O0 -rdynamic -Wall $(GENFLAGS) $(SAN_CFLAGS) \ -I$(CURDIR) -I$(INCLUDE_DIR) -I$(GENDIR) -I$(LIBDIR) \ -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT) +LDFLAGS += $(SAN_CFLAGS) LDLIBS += -lcap -lelf -lz -lrt -lpthread # Silence some warnings when compiled with clang From 8f7b239ea8cfdc8e64c875ee417fed41431a1f37 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:14 -0800 Subject: [PATCH 0090/2048] libbpf: Free up resources used by inner map definition It's not enough to just free(map->inner_map), as inner_map itself can have extra memory allocated, like map name. Fixes: 646f02ffdd49 ("libbpf: Add BTF-defined map-in-map support") Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-3-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index dfd15cc60ea7..d869ebee1e27 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -9009,7 +9009,10 @@ int bpf_map__set_inner_map_fd(struct bpf_map *map, int fd) pr_warn("error: inner_map_fd already specified\n"); return libbpf_err(-EINVAL); } - zfree(&map->inner_map); + if (map->inner_map) { + bpf_map__destroy(map->inner_map); + zfree(&map->inner_map); + } map->inner_map_fd = fd; return 0; } From 8ba285874913da21ca39a46376e9cc5ce0f45f94 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:15 -0800 Subject: [PATCH 0091/2048] selftests/bpf: Fix memory leaks in btf_type_c_dump() helper Free up memory and resources used by temporary allocated memstream and btf_dump instance. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-4-andrii@kernel.org --- tools/testing/selftests/bpf/btf_helpers.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/btf_helpers.c b/tools/testing/selftests/bpf/btf_helpers.c index b5b6b013a245..3d1a748d09d8 100644 --- a/tools/testing/selftests/bpf/btf_helpers.c +++ b/tools/testing/selftests/bpf/btf_helpers.c @@ -251,18 +251,23 @@ const char *btf_type_c_dump(const struct btf *btf) d = btf_dump__new(btf, NULL, &opts, btf_dump_printf); if (libbpf_get_error(d)) { fprintf(stderr, "Failed to create btf_dump instance: %ld\n", libbpf_get_error(d)); - return NULL; + goto err_out; } for (i = 1; i < btf__type_cnt(btf); i++) { err = btf_dump__dump_type(d, i); if (err) { fprintf(stderr, "Failed to dump type [%d]: %d\n", i, err); - return NULL; + goto err_out; } } + btf_dump__free(d); fflush(buf_file); fclose(buf_file); return buf; +err_out: + btf_dump__free(d); + fclose(buf_file); + return NULL; } From b8b26e585f3a0fbcee1032c622f046787da57390 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:16 -0800 Subject: [PATCH 0092/2048] selftests/bpf: Free per-cpu values array in bpf_iter selftest Array holding per-cpu values wasn't freed. Fix that. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211107165521.9240-5-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/bpf_iter.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c index 9454331aaf85..3e10abce3e5a 100644 --- a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c +++ b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c @@ -699,14 +699,13 @@ static void test_bpf_percpu_hash_map(void) char buf[64]; void *val; - val = malloc(8 * bpf_num_possible_cpus()); - skel = bpf_iter_bpf_percpu_hash_map__open(); if (CHECK(!skel, "bpf_iter_bpf_percpu_hash_map__open", "skeleton open failed\n")) return; skel->rodata->num_cpus = bpf_num_possible_cpus(); + val = malloc(8 * bpf_num_possible_cpus()); err = bpf_iter_bpf_percpu_hash_map__load(skel); if (CHECK(!skel, "bpf_iter_bpf_percpu_hash_map__load", @@ -770,6 +769,7 @@ free_link: bpf_link__destroy(link); out: bpf_iter_bpf_percpu_hash_map__destroy(skel); + free(val); } static void test_bpf_array_map(void) @@ -870,14 +870,13 @@ static void test_bpf_percpu_array_map(void) void *val; int len; - val = malloc(8 * bpf_num_possible_cpus()); - skel = bpf_iter_bpf_percpu_array_map__open(); if (CHECK(!skel, "bpf_iter_bpf_percpu_array_map__open", "skeleton open failed\n")) return; skel->rodata->num_cpus = bpf_num_possible_cpus(); + val = malloc(8 * bpf_num_possible_cpus()); err = bpf_iter_bpf_percpu_array_map__load(skel); if (CHECK(!skel, "bpf_iter_bpf_percpu_array_map__load", @@ -933,6 +932,7 @@ free_link: bpf_link__destroy(link); out: bpf_iter_bpf_percpu_array_map__destroy(skel); + free(val); } /* An iterator program deletes all local storage in a map. */ From 5309b516bcc6f76dda0e44a7a1824324277093d6 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:17 -0800 Subject: [PATCH 0093/2048] selftests/bpf: Free inner strings index in btf selftest Inner array of allocated strings wasn't freed on success. Now it's always freed. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-6-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/btf.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/btf.c b/tools/testing/selftests/bpf/prog_tests/btf.c index ac596cb06e40..ebd1aa4d09d6 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf.c +++ b/tools/testing/selftests/bpf/prog_tests/btf.c @@ -4046,11 +4046,9 @@ static void *btf_raw_create(const struct btf_header *hdr, next_str_idx < strs_cnt ? strs_idx[next_str_idx] : NULL; done: + free(strs_idx); if (err) { - if (raw_btf) - free(raw_btf); - if (strs_idx) - free(strs_idx); + free(raw_btf); return NULL; } return raw_btf; From f79587520a6007a3734b23a3c2eb4c62aa457533 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:18 -0800 Subject: [PATCH 0094/2048] selftests/bpf: Clean up btf and btf_dump in dump_datasec test Free up used resources at the end and on error. Also make it more obvious that there is btf__parse() call that creates struct btf instance. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-7-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/btf_dump.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/btf_dump.c b/tools/testing/selftests/bpf/prog_tests/btf_dump.c index aa76360d8f49..a04961942dfa 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_dump.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_dump.c @@ -814,21 +814,25 @@ static void test_btf_datasec(struct btf *btf, struct btf_dump *d, char *str, static void test_btf_dump_datasec_data(char *str) { - struct btf *btf = btf__parse("xdping_kern.o", NULL); + struct btf *btf; struct btf_dump_opts opts = { .ctx = str }; char license[4] = "GPL"; struct btf_dump *d; + btf = btf__parse("xdping_kern.o", NULL); if (!ASSERT_OK_PTR(btf, "xdping_kern.o BTF not found")) return; d = btf_dump__new(btf, NULL, &opts, btf_dump_snprintf); if (!ASSERT_OK_PTR(d, "could not create BTF dump")) - return; + goto out; test_btf_datasec(btf, d, str, "license", "SEC(\"license\") char[4] _license = (char[4])['G','P','L',];", license, sizeof(license)); +out: + btf_dump__free(d); + btf__free(btf); } void test_btf_dump() { From f92321d706a810b89a905e04658e38931c4bb0e0 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:19 -0800 Subject: [PATCH 0095/2048] selftests/bpf: Avoid duplicate btf__parse() call btf__parse() is repeated after successful setup, leaving the first instance leaked. Remove redundant and premature call. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-8-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/core_reloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/prog_tests/core_reloc.c b/tools/testing/selftests/bpf/prog_tests/core_reloc.c index 55ec85ba7375..1041d0c593f6 100644 --- a/tools/testing/selftests/bpf/prog_tests/core_reloc.c +++ b/tools/testing/selftests/bpf/prog_tests/core_reloc.c @@ -433,7 +433,7 @@ static int setup_type_id_case_local(struct core_reloc_test_case *test) static int setup_type_id_case_success(struct core_reloc_test_case *test) { struct core_reloc_type_id_output *exp = (void *)test->output; - struct btf *targ_btf = btf__parse(test->btf_src_file, NULL); + struct btf *targ_btf; int err; err = setup_type_id_case_local(test); From f91231eeeed752119f49eb6620cae44ec745a007 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:20 -0800 Subject: [PATCH 0096/2048] selftests/bpf: Destroy XDP link correctly bpf_link__detach() was confused with bpf_link__destroy() and leaves leaked FD in the process. Fix the problem. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-9-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/migrate_reuseport.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/migrate_reuseport.c b/tools/testing/selftests/bpf/prog_tests/migrate_reuseport.c index 7589c03fd26b..eb2feaac81fe 100644 --- a/tools/testing/selftests/bpf/prog_tests/migrate_reuseport.c +++ b/tools/testing/selftests/bpf/prog_tests/migrate_reuseport.c @@ -204,8 +204,8 @@ static int pass_ack(struct migrate_reuseport_test_case *test_case) { int err; - err = bpf_link__detach(test_case->link); - if (!ASSERT_OK(err, "bpf_link__detach")) + err = bpf_link__destroy(test_case->link); + if (!ASSERT_OK(err, "bpf_link__destroy")) return -1; test_case->link = NULL; From 8c7a95520184b6677ca6075e12df9c208d57d088 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Sun, 7 Nov 2021 08:55:21 -0800 Subject: [PATCH 0097/2048] selftests/bpf: Fix bpf_object leak in skb_ctx selftest skb_ctx selftest didn't close bpf_object implicitly allocated by bpf_prog_test_load() helper. Fix the problem by explicitly calling bpf_object__close() at the end of the test. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Reviewed-by: Hengqi Chen Link: https://lore.kernel.org/bpf/20211107165521.9240-10-andrii@kernel.org --- tools/testing/selftests/bpf/prog_tests/skb_ctx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/skb_ctx.c b/tools/testing/selftests/bpf/prog_tests/skb_ctx.c index d3106078838c..b5319ba2ee27 100644 --- a/tools/testing/selftests/bpf/prog_tests/skb_ctx.c +++ b/tools/testing/selftests/bpf/prog_tests/skb_ctx.c @@ -111,4 +111,6 @@ void test_skb_ctx(void) "ctx_out_mark", "skb->mark == %u, expected %d\n", skb.mark, 10); + + bpf_object__close(obj); } From 7c7e3d31e7856a8260a254f8c71db416f7f9f5a1 Mon Sep 17 00:00:00 2001 From: Song Liu Date: Fri, 5 Nov 2021 16:23:29 -0700 Subject: [PATCH 0098/2048] bpf: Introduce helper bpf_find_vma In some profiler use cases, it is necessary to map an address to the backing file, e.g., a shared library. bpf_find_vma helper provides a flexible way to achieve this. bpf_find_vma maps an address of a task to the vma (vm_area_struct) for this address, and feed the vma to an callback BPF function. The callback function is necessary here, as we need to ensure mmap_sem is unlocked. It is necessary to lock mmap_sem for find_vma. To lock and unlock mmap_sem safely when irqs are disable, we use the same mechanism as stackmap with build_id. Specifically, when irqs are disabled, the unlocked is postponed in an irq_work. Refactor stackmap.c so that the irq_work is shared among bpf_find_vma and stackmap helpers. Signed-off-by: Song Liu Signed-off-by: Alexei Starovoitov Tested-by: Hengqi Chen Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211105232330.1936330-2-songliubraving@fb.com --- include/linux/bpf.h | 1 + include/uapi/linux/bpf.h | 20 +++++++++ kernel/bpf/btf.c | 5 ++- kernel/bpf/mmap_unlock_work.h | 65 +++++++++++++++++++++++++++ kernel/bpf/stackmap.c | 80 +++------------------------------- kernel/bpf/task_iter.c | 76 +++++++++++++++++++++++++++++--- kernel/bpf/verifier.c | 34 +++++++++++++++ kernel/trace/bpf_trace.c | 2 + tools/include/uapi/linux/bpf.h | 20 +++++++++ 9 files changed, 222 insertions(+), 81 deletions(-) create mode 100644 kernel/bpf/mmap_unlock_work.h diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 2be6dfd68df9..df3410bff4b0 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -2157,6 +2157,7 @@ extern const struct bpf_func_proto bpf_btf_find_by_name_kind_proto; extern const struct bpf_func_proto bpf_sk_setsockopt_proto; extern const struct bpf_func_proto bpf_sk_getsockopt_proto; extern const struct bpf_func_proto bpf_kallsyms_lookup_name_proto; +extern const struct bpf_func_proto bpf_find_vma_proto; const struct bpf_func_proto *tracing_prog_func_proto( enum bpf_func_id func_id, const struct bpf_prog *prog); diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index ba5af15e25f5..509eee5f0393 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -4938,6 +4938,25 @@ union bpf_attr { * **-ENOENT** if symbol is not found. * * **-EPERM** if caller does not have permission to obtain kernel address. + * + * long bpf_find_vma(struct task_struct *task, u64 addr, void *callback_fn, void *callback_ctx, u64 flags) + * Description + * Find vma of *task* that contains *addr*, call *callback_fn* + * function with *task*, *vma*, and *callback_ctx*. + * The *callback_fn* should be a static function and + * the *callback_ctx* should be a pointer to the stack. + * The *flags* is used to control certain aspects of the helper. + * Currently, the *flags* must be 0. + * + * The expected callback signature is + * + * long (\*callback_fn)(struct task_struct \*task, struct vm_area_struct \*vma, void \*callback_ctx); + * + * Return + * 0 on success. + * **-ENOENT** if *task->mm* is NULL, or no vma contains *addr*. + * **-EBUSY** if failed to try lock mmap_lock. + * **-EINVAL** for invalid **flags**. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5120,6 +5139,7 @@ union bpf_attr { FN(trace_vprintk), \ FN(skc_to_unix_sock), \ FN(kallsyms_lookup_name), \ + FN(find_vma), \ /* */ /* integer value in 'imm' field of BPF_CALL instruction selects which helper diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index dbc3ad07e21b..cdb0fba65600 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -6342,7 +6342,10 @@ const struct bpf_func_proto bpf_btf_find_by_name_kind_proto = { .arg4_type = ARG_ANYTHING, }; -BTF_ID_LIST_GLOBAL_SINGLE(btf_task_struct_ids, struct, task_struct) +BTF_ID_LIST_GLOBAL(btf_task_struct_ids) +BTF_ID(struct, task_struct) +BTF_ID(struct, file) +BTF_ID(struct, vm_area_struct) /* BTF ID set registration API for modules */ diff --git a/kernel/bpf/mmap_unlock_work.h b/kernel/bpf/mmap_unlock_work.h new file mode 100644 index 000000000000..5d18d7d85bef --- /dev/null +++ b/kernel/bpf/mmap_unlock_work.h @@ -0,0 +1,65 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* Copyright (c) 2021 Facebook + */ + +#ifndef __MMAP_UNLOCK_WORK_H__ +#define __MMAP_UNLOCK_WORK_H__ +#include + +/* irq_work to run mmap_read_unlock() in irq_work */ +struct mmap_unlock_irq_work { + struct irq_work irq_work; + struct mm_struct *mm; +}; + +DECLARE_PER_CPU(struct mmap_unlock_irq_work, mmap_unlock_work); + +/* + * We cannot do mmap_read_unlock() when the irq is disabled, because of + * risk to deadlock with rq_lock. To look up vma when the irqs are + * disabled, we need to run mmap_read_unlock() in irq_work. We use a + * percpu variable to do the irq_work. If the irq_work is already used + * by another lookup, we fall over. + */ +static inline bool bpf_mmap_unlock_get_irq_work(struct mmap_unlock_irq_work **work_ptr) +{ + struct mmap_unlock_irq_work *work = NULL; + bool irq_work_busy = false; + + if (irqs_disabled()) { + if (!IS_ENABLED(CONFIG_PREEMPT_RT)) { + work = this_cpu_ptr(&mmap_unlock_work); + if (irq_work_is_busy(&work->irq_work)) { + /* cannot queue more up_read, fallback */ + irq_work_busy = true; + } + } else { + /* + * PREEMPT_RT does not allow to trylock mmap sem in + * interrupt disabled context. Force the fallback code. + */ + irq_work_busy = true; + } + } + + *work_ptr = work; + return irq_work_busy; +} + +static inline void bpf_mmap_unlock_mm(struct mmap_unlock_irq_work *work, struct mm_struct *mm) +{ + if (!work) { + mmap_read_unlock(mm); + } else { + work->mm = mm; + + /* The lock will be released once we're out of interrupt + * context. Tell lockdep that we've released it now so + * it doesn't complain that we forgot to release it. + */ + rwsem_release(&mm->mmap_lock.dep_map, _RET_IP_); + irq_work_queue(&work->irq_work); + } +} + +#endif /* __MMAP_UNLOCK_WORK_H__ */ diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index 6e75bbee39f0..1de0a1b03636 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -7,10 +7,10 @@ #include #include #include -#include #include #include #include "percpu_freelist.h" +#include "mmap_unlock_work.h" #define STACK_CREATE_FLAG_MASK \ (BPF_F_NUMA_NODE | BPF_F_RDONLY | BPF_F_WRONLY | \ @@ -31,25 +31,6 @@ struct bpf_stack_map { struct stack_map_bucket *buckets[]; }; -/* irq_work to run up_read() for build_id lookup in nmi context */ -struct stack_map_irq_work { - struct irq_work irq_work; - struct mm_struct *mm; -}; - -static void do_up_read(struct irq_work *entry) -{ - struct stack_map_irq_work *work; - - if (WARN_ON_ONCE(IS_ENABLED(CONFIG_PREEMPT_RT))) - return; - - work = container_of(entry, struct stack_map_irq_work, irq_work); - mmap_read_unlock_non_owner(work->mm); -} - -static DEFINE_PER_CPU(struct stack_map_irq_work, up_read_work); - static inline bool stack_map_use_build_id(struct bpf_map *map) { return (map->map_flags & BPF_F_STACK_BUILD_ID); @@ -149,35 +130,13 @@ static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs, u64 *ips, u32 trace_nr, bool user) { int i; + struct mmap_unlock_irq_work *work = NULL; + bool irq_work_busy = bpf_mmap_unlock_get_irq_work(&work); struct vm_area_struct *vma; - bool irq_work_busy = false; - struct stack_map_irq_work *work = NULL; - if (irqs_disabled()) { - if (!IS_ENABLED(CONFIG_PREEMPT_RT)) { - work = this_cpu_ptr(&up_read_work); - if (irq_work_is_busy(&work->irq_work)) { - /* cannot queue more up_read, fallback */ - irq_work_busy = true; - } - } else { - /* - * PREEMPT_RT does not allow to trylock mmap sem in - * interrupt disabled context. Force the fallback code. - */ - irq_work_busy = true; - } - } - - /* - * We cannot do up_read() when the irq is disabled, because of - * risk to deadlock with rq_lock. To do build_id lookup when the - * irqs are disabled, we need to run up_read() in irq_work. We use - * a percpu variable to do the irq_work. If the irq_work is - * already used by another lookup, we fall back to report ips. - * - * Same fallback is used for kernel stack (!user) on a stackmap - * with build_id. + /* If the irq_work is in use, fall back to report ips. Same + * fallback is used for kernel stack (!user) on a stackmap with + * build_id. */ if (!user || !current || !current->mm || irq_work_busy || !mmap_read_trylock(current->mm)) { @@ -203,19 +162,7 @@ static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs, - vma->vm_start; id_offs[i].status = BPF_STACK_BUILD_ID_VALID; } - - if (!work) { - mmap_read_unlock(current->mm); - } else { - work->mm = current->mm; - - /* The lock will be released once we're out of interrupt - * context. Tell lockdep that we've released it now so - * it doesn't complain that we forgot to release it. - */ - rwsem_release(¤t->mm->mmap_lock.dep_map, _RET_IP_); - irq_work_queue(&work->irq_work); - } + bpf_mmap_unlock_mm(work, current->mm); } static struct perf_callchain_entry * @@ -719,16 +666,3 @@ const struct bpf_map_ops stack_trace_map_ops = { .map_btf_name = "bpf_stack_map", .map_btf_id = &stack_trace_map_btf_id, }; - -static int __init stack_map_init(void) -{ - int cpu; - struct stack_map_irq_work *work; - - for_each_possible_cpu(cpu) { - work = per_cpu_ptr(&up_read_work, cpu); - init_irq_work(&work->irq_work, do_up_read); - } - return 0; -} -subsys_initcall(stack_map_init); diff --git a/kernel/bpf/task_iter.c b/kernel/bpf/task_iter.c index b48750bfba5a..f171479f7dd6 100644 --- a/kernel/bpf/task_iter.c +++ b/kernel/bpf/task_iter.c @@ -8,6 +8,7 @@ #include #include #include +#include "mmap_unlock_work.h" struct bpf_iter_seq_task_common { struct pid_namespace *ns; @@ -524,10 +525,6 @@ static const struct seq_operations task_vma_seq_ops = { .show = task_vma_seq_show, }; -BTF_ID_LIST(btf_task_file_ids) -BTF_ID(struct, file) -BTF_ID(struct, vm_area_struct) - static const struct bpf_iter_seq_info task_seq_info = { .seq_ops = &task_seq_ops, .init_seq_private = init_seq_pidns, @@ -586,9 +583,74 @@ static struct bpf_iter_reg task_vma_reg_info = { .seq_info = &task_vma_seq_info, }; +BPF_CALL_5(bpf_find_vma, struct task_struct *, task, u64, start, + bpf_callback_t, callback_fn, void *, callback_ctx, u64, flags) +{ + struct mmap_unlock_irq_work *work = NULL; + struct vm_area_struct *vma; + bool irq_work_busy = false; + struct mm_struct *mm; + int ret = -ENOENT; + + if (flags) + return -EINVAL; + + if (!task) + return -ENOENT; + + mm = task->mm; + if (!mm) + return -ENOENT; + + irq_work_busy = bpf_mmap_unlock_get_irq_work(&work); + + if (irq_work_busy || !mmap_read_trylock(mm)) + return -EBUSY; + + vma = find_vma(mm, start); + + if (vma && vma->vm_start <= start && vma->vm_end > start) { + callback_fn((u64)(long)task, (u64)(long)vma, + (u64)(long)callback_ctx, 0, 0); + ret = 0; + } + bpf_mmap_unlock_mm(work, mm); + return ret; +} + +const struct bpf_func_proto bpf_find_vma_proto = { + .func = bpf_find_vma, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_BTF_ID, + .arg1_btf_id = &btf_task_struct_ids[0], + .arg2_type = ARG_ANYTHING, + .arg3_type = ARG_PTR_TO_FUNC, + .arg4_type = ARG_PTR_TO_STACK_OR_NULL, + .arg5_type = ARG_ANYTHING, +}; + +DEFINE_PER_CPU(struct mmap_unlock_irq_work, mmap_unlock_work); + +static void do_mmap_read_unlock(struct irq_work *entry) +{ + struct mmap_unlock_irq_work *work; + + if (WARN_ON_ONCE(IS_ENABLED(CONFIG_PREEMPT_RT))) + return; + + work = container_of(entry, struct mmap_unlock_irq_work, irq_work); + mmap_read_unlock_non_owner(work->mm); +} + static int __init task_iter_init(void) { - int ret; + struct mmap_unlock_irq_work *work; + int ret, cpu; + + for_each_possible_cpu(cpu) { + work = per_cpu_ptr(&mmap_unlock_work, cpu); + init_irq_work(&work->irq_work, do_mmap_read_unlock); + } task_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; ret = bpf_iter_reg_target(&task_reg_info); @@ -596,13 +658,13 @@ static int __init task_iter_init(void) return ret; task_file_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; - task_file_reg_info.ctx_arg_info[1].btf_id = btf_task_file_ids[0]; + task_file_reg_info.ctx_arg_info[1].btf_id = btf_task_struct_ids[1]; ret = bpf_iter_reg_target(&task_file_reg_info); if (ret) return ret; task_vma_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; - task_vma_reg_info.ctx_arg_info[1].btf_id = btf_task_file_ids[1]; + task_vma_reg_info.ctx_arg_info[1].btf_id = btf_task_struct_ids[2]; return bpf_iter_reg_target(&task_vma_reg_info); } late_initcall(task_iter_init); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index f0dca726ebfd..1aafb43f61d1 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6132,6 +6132,33 @@ static int set_timer_callback_state(struct bpf_verifier_env *env, return 0; } +static int set_find_vma_callback_state(struct bpf_verifier_env *env, + struct bpf_func_state *caller, + struct bpf_func_state *callee, + int insn_idx) +{ + /* bpf_find_vma(struct task_struct *task, u64 addr, + * void *callback_fn, void *callback_ctx, u64 flags) + * (callback_fn)(struct task_struct *task, + * struct vm_area_struct *vma, void *callback_ctx); + */ + callee->regs[BPF_REG_1] = caller->regs[BPF_REG_1]; + + callee->regs[BPF_REG_2].type = PTR_TO_BTF_ID; + __mark_reg_known_zero(&callee->regs[BPF_REG_2]); + callee->regs[BPF_REG_2].btf = btf_vmlinux; + callee->regs[BPF_REG_2].btf_id = btf_task_struct_ids[2]; + + /* pointer to stack or null */ + callee->regs[BPF_REG_3] = caller->regs[BPF_REG_4]; + + /* unused */ + __mark_reg_not_init(env, &callee->regs[BPF_REG_4]); + __mark_reg_not_init(env, &callee->regs[BPF_REG_5]); + callee->in_callback_fn = true; + return 0; +} + static int prepare_func_exit(struct bpf_verifier_env *env, int *insn_idx) { struct bpf_verifier_state *state = env->cur_state; @@ -6489,6 +6516,13 @@ static int check_helper_call(struct bpf_verifier_env *env, struct bpf_insn *insn return -EINVAL; } + if (func_id == BPF_FUNC_find_vma) { + err = __check_func_call(env, insn, insn_idx_p, meta.subprogno, + set_find_vma_callback_state); + if (err < 0) + return -EINVAL; + } + if (func_id == BPF_FUNC_snprintf) { err = check_bpf_snprintf_call(env, regs); if (err < 0) diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 7396488793ff..390176a3031a 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -1208,6 +1208,8 @@ bpf_tracing_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_get_func_ip_proto_tracing; case BPF_FUNC_get_branch_snapshot: return &bpf_get_branch_snapshot_proto; + case BPF_FUNC_find_vma: + return &bpf_find_vma_proto; case BPF_FUNC_trace_vprintk: return bpf_get_trace_vprintk_proto(); default: diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index ba5af15e25f5..509eee5f0393 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -4938,6 +4938,25 @@ union bpf_attr { * **-ENOENT** if symbol is not found. * * **-EPERM** if caller does not have permission to obtain kernel address. + * + * long bpf_find_vma(struct task_struct *task, u64 addr, void *callback_fn, void *callback_ctx, u64 flags) + * Description + * Find vma of *task* that contains *addr*, call *callback_fn* + * function with *task*, *vma*, and *callback_ctx*. + * The *callback_fn* should be a static function and + * the *callback_ctx* should be a pointer to the stack. + * The *flags* is used to control certain aspects of the helper. + * Currently, the *flags* must be 0. + * + * The expected callback signature is + * + * long (\*callback_fn)(struct task_struct \*task, struct vm_area_struct \*vma, void \*callback_ctx); + * + * Return + * 0 on success. + * **-ENOENT** if *task->mm* is NULL, or no vma contains *addr*. + * **-EBUSY** if failed to try lock mmap_lock. + * **-EINVAL** for invalid **flags**. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5120,6 +5139,7 @@ union bpf_attr { FN(trace_vprintk), \ FN(skc_to_unix_sock), \ FN(kallsyms_lookup_name), \ + FN(find_vma), \ /* */ /* integer value in 'imm' field of BPF_CALL instruction selects which helper From f108662b27c96cdadfadd39f0c0d650704cd593d Mon Sep 17 00:00:00 2001 From: Song Liu Date: Fri, 5 Nov 2021 16:23:30 -0700 Subject: [PATCH 0099/2048] selftests/bpf: Add tests for bpf_find_vma Add tests for bpf_find_vma in perf_event program and kprobe program. The perf_event program is triggered from NMI context, so the second call of bpf_find_vma() will return -EBUSY (irq_work busy). The kprobe program, on the other hand, does not have this constraint. Also add tests for illegal writes to task or vma from the callback function. The verifier should reject both cases. Signed-off-by: Song Liu Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211105232330.1936330-3-songliubraving@fb.com --- .../selftests/bpf/prog_tests/find_vma.c | 117 ++++++++++++++++++ tools/testing/selftests/bpf/progs/find_vma.c | 69 +++++++++++ .../selftests/bpf/progs/find_vma_fail1.c | 29 +++++ .../selftests/bpf/progs/find_vma_fail2.c | 29 +++++ 4 files changed, 244 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/find_vma.c create mode 100644 tools/testing/selftests/bpf/progs/find_vma.c create mode 100644 tools/testing/selftests/bpf/progs/find_vma_fail1.c create mode 100644 tools/testing/selftests/bpf/progs/find_vma_fail2.c diff --git a/tools/testing/selftests/bpf/prog_tests/find_vma.c b/tools/testing/selftests/bpf/prog_tests/find_vma.c new file mode 100644 index 000000000000..b74b3c0c555a --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/find_vma.c @@ -0,0 +1,117 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ +#include +#include +#include +#include "find_vma.skel.h" +#include "find_vma_fail1.skel.h" +#include "find_vma_fail2.skel.h" + +static void test_and_reset_skel(struct find_vma *skel, int expected_find_zero_ret) +{ + ASSERT_EQ(skel->bss->found_vm_exec, 1, "found_vm_exec"); + ASSERT_EQ(skel->data->find_addr_ret, 0, "find_addr_ret"); + ASSERT_EQ(skel->data->find_zero_ret, expected_find_zero_ret, "find_zero_ret"); + ASSERT_OK_PTR(strstr(skel->bss->d_iname, "test_progs"), "find_test_progs"); + + skel->bss->found_vm_exec = 0; + skel->data->find_addr_ret = -1; + skel->data->find_zero_ret = -1; + skel->bss->d_iname[0] = 0; +} + +static int open_pe(void) +{ + struct perf_event_attr attr = {0}; + int pfd; + + /* create perf event */ + attr.size = sizeof(attr); + attr.type = PERF_TYPE_HARDWARE; + attr.config = PERF_COUNT_HW_CPU_CYCLES; + attr.freq = 1; + attr.sample_freq = 4000; + pfd = syscall(__NR_perf_event_open, &attr, 0, -1, -1, PERF_FLAG_FD_CLOEXEC); + + return pfd >= 0 ? pfd : -errno; +} + +static void test_find_vma_pe(struct find_vma *skel) +{ + struct bpf_link *link = NULL; + volatile int j = 0; + int pfd, i; + + pfd = open_pe(); + if (pfd < 0) { + if (pfd == -ENOENT || pfd == -EOPNOTSUPP) { + printf("%s:SKIP:no PERF_COUNT_HW_CPU_CYCLES\n", __func__); + test__skip(); + goto cleanup; + } + if (!ASSERT_GE(pfd, 0, "perf_event_open")) + goto cleanup; + } + + link = bpf_program__attach_perf_event(skel->progs.handle_pe, pfd); + if (!ASSERT_OK_PTR(link, "attach_perf_event")) + goto cleanup; + + for (i = 0; i < 1000000; ++i) + ++j; + + test_and_reset_skel(skel, -EBUSY /* in nmi, irq_work is busy */); +cleanup: + bpf_link__destroy(link); + close(pfd); +} + +static void test_find_vma_kprobe(struct find_vma *skel) +{ + int err; + + err = find_vma__attach(skel); + if (!ASSERT_OK(err, "get_branch_snapshot__attach")) + return; + + getpgid(skel->bss->target_pid); + test_and_reset_skel(skel, -ENOENT /* could not find vma for ptr 0 */); +} + +static void test_illegal_write_vma(void) +{ + struct find_vma_fail1 *skel; + + skel = find_vma_fail1__open_and_load(); + if (!ASSERT_ERR_PTR(skel, "find_vma_fail1__open_and_load")) + find_vma_fail1__destroy(skel); +} + +static void test_illegal_write_task(void) +{ + struct find_vma_fail2 *skel; + + skel = find_vma_fail2__open_and_load(); + if (!ASSERT_ERR_PTR(skel, "find_vma_fail2__open_and_load")) + find_vma_fail2__destroy(skel); +} + +void serial_test_find_vma(void) +{ + struct find_vma *skel; + + skel = find_vma__open_and_load(); + if (!ASSERT_OK_PTR(skel, "find_vma__open_and_load")) + return; + + skel->bss->target_pid = getpid(); + skel->bss->addr = (__u64)(uintptr_t)test_find_vma_pe; + + test_find_vma_pe(skel); + usleep(100000); /* allow the irq_work to finish */ + test_find_vma_kprobe(skel); + + find_vma__destroy(skel); + test_illegal_write_vma(); + test_illegal_write_task(); +} diff --git a/tools/testing/selftests/bpf/progs/find_vma.c b/tools/testing/selftests/bpf/progs/find_vma.c new file mode 100644 index 000000000000..38034fb82530 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/find_vma.c @@ -0,0 +1,69 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ +#include "vmlinux.h" +#include +#include + +char _license[] SEC("license") = "GPL"; + +struct callback_ctx { + int dummy; +}; + +#define VM_EXEC 0x00000004 +#define DNAME_INLINE_LEN 32 + +pid_t target_pid = 0; +char d_iname[DNAME_INLINE_LEN] = {0}; +__u32 found_vm_exec = 0; +__u64 addr = 0; +int find_zero_ret = -1; +int find_addr_ret = -1; + +static long check_vma(struct task_struct *task, struct vm_area_struct *vma, + struct callback_ctx *data) +{ + if (vma->vm_file) + bpf_probe_read_kernel_str(d_iname, DNAME_INLINE_LEN - 1, + vma->vm_file->f_path.dentry->d_iname); + + /* check for VM_EXEC */ + if (vma->vm_flags & VM_EXEC) + found_vm_exec = 1; + + return 0; +} + +SEC("raw_tp/sys_enter") +int handle_getpid(void) +{ + struct task_struct *task = bpf_get_current_task_btf(); + struct callback_ctx data = {}; + + if (task->pid != target_pid) + return 0; + + find_addr_ret = bpf_find_vma(task, addr, check_vma, &data, 0); + + /* this should return -ENOENT */ + find_zero_ret = bpf_find_vma(task, 0, check_vma, &data, 0); + return 0; +} + +SEC("perf_event") +int handle_pe(void) +{ + struct task_struct *task = bpf_get_current_task_btf(); + struct callback_ctx data = {}; + + if (task->pid != target_pid) + return 0; + + find_addr_ret = bpf_find_vma(task, addr, check_vma, &data, 0); + + /* In NMI, this should return -EBUSY, as the previous call is using + * the irq_work. + */ + find_zero_ret = bpf_find_vma(task, 0, check_vma, &data, 0); + return 0; +} diff --git a/tools/testing/selftests/bpf/progs/find_vma_fail1.c b/tools/testing/selftests/bpf/progs/find_vma_fail1.c new file mode 100644 index 000000000000..b3b326b8e2d1 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/find_vma_fail1.c @@ -0,0 +1,29 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ +#include "vmlinux.h" +#include + +char _license[] SEC("license") = "GPL"; + +struct callback_ctx { + int dummy; +}; + +static long write_vma(struct task_struct *task, struct vm_area_struct *vma, + struct callback_ctx *data) +{ + /* writing to vma, which is illegal */ + vma->vm_flags |= 0x55; + + return 0; +} + +SEC("raw_tp/sys_enter") +int handle_getpid(void) +{ + struct task_struct *task = bpf_get_current_task_btf(); + struct callback_ctx data = {}; + + bpf_find_vma(task, 0, write_vma, &data, 0); + return 0; +} diff --git a/tools/testing/selftests/bpf/progs/find_vma_fail2.c b/tools/testing/selftests/bpf/progs/find_vma_fail2.c new file mode 100644 index 000000000000..9bcf3203e26b --- /dev/null +++ b/tools/testing/selftests/bpf/progs/find_vma_fail2.c @@ -0,0 +1,29 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ +#include "vmlinux.h" +#include + +char _license[] SEC("license") = "GPL"; + +struct callback_ctx { + int dummy; +}; + +static long write_task(struct task_struct *task, struct vm_area_struct *vma, + struct callback_ctx *data) +{ + /* writing to task, which is illegal */ + task->mm = NULL; + + return 0; +} + +SEC("raw_tp/sys_enter") +int handle_getpid(void) +{ + struct task_struct *task = bpf_get_current_task_btf(); + struct callback_ctx data = {}; + + bpf_find_vma(task, 0, write_task, &data, 0); + return 0; +} From 89dcb1da611d9b3ff0728502d58372fdaae9ebff Mon Sep 17 00:00:00 2001 From: Bryan O'Donoghue Date: Fri, 5 Nov 2021 12:21:50 +0000 Subject: [PATCH 0100/2048] wcn36xx: Fix DMA channel enable/disable cycle Right now we have a broken sequence where we enable DMA channel interrupts which can be left enabled and never disabled if we hit an error path. Worse still when we unload the driver, the DMA channel interrupt bits are left intact. About the only saving grace here is that we do remember to disable the wcnss interrupt when unload the driver. Fixes: 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware") Signed-off-by: Bryan O'Donoghue Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211105122152.1580542-2-bryan.odonoghue@linaro.org --- drivers/net/wireless/ath/wcn36xx/dxe.c | 38 ++++++++++++++++++-------- 1 file changed, 27 insertions(+), 11 deletions(-) diff --git a/drivers/net/wireless/ath/wcn36xx/dxe.c b/drivers/net/wireless/ath/wcn36xx/dxe.c index fd627c9f3d40..d6c621518c7b 100644 --- a/drivers/net/wireless/ath/wcn36xx/dxe.c +++ b/drivers/net/wireless/ath/wcn36xx/dxe.c @@ -272,6 +272,21 @@ static int wcn36xx_dxe_enable_ch_int(struct wcn36xx *wcn, u16 wcn_ch) return 0; } +static void wcn36xx_dxe_disable_ch_int(struct wcn36xx *wcn, u16 wcn_ch) +{ + int reg_data = 0; + + wcn36xx_dxe_read_register(wcn, + WCN36XX_DXE_INT_MASK_REG, + ®_data); + + reg_data &= ~wcn_ch; + + wcn36xx_dxe_write_register(wcn, + WCN36XX_DXE_INT_MASK_REG, + (int)reg_data); +} + static int wcn36xx_dxe_fill_skb(struct device *dev, struct wcn36xx_dxe_ctl *ctl, gfp_t gfp) @@ -916,7 +931,6 @@ int wcn36xx_dxe_init(struct wcn36xx *wcn) WCN36XX_DXE_WQ_TX_L); wcn36xx_dxe_read_register(wcn, WCN36XX_DXE_REG_CH_EN, ®_data); - wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_L); /***************************************/ /* Init descriptors for TX HIGH channel */ @@ -940,9 +954,6 @@ int wcn36xx_dxe_init(struct wcn36xx *wcn) wcn36xx_dxe_read_register(wcn, WCN36XX_DXE_REG_CH_EN, ®_data); - /* Enable channel interrupts */ - wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_H); - /***************************************/ /* Init descriptors for RX LOW channel */ /***************************************/ @@ -952,7 +963,6 @@ int wcn36xx_dxe_init(struct wcn36xx *wcn) goto out_err_rxl_ch; } - /* For RX we need to preallocated buffers */ wcn36xx_dxe_ch_alloc_skb(wcn, &wcn->dxe_rx_l_ch); @@ -975,9 +985,6 @@ int wcn36xx_dxe_init(struct wcn36xx *wcn) WCN36XX_DXE_REG_CTL_RX_L, WCN36XX_DXE_CH_DEFAULT_CTL_RX_L); - /* Enable channel interrupts */ - wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_L); - /***************************************/ /* Init descriptors for RX HIGH channel */ /***************************************/ @@ -1009,15 +1016,18 @@ int wcn36xx_dxe_init(struct wcn36xx *wcn) WCN36XX_DXE_REG_CTL_RX_H, WCN36XX_DXE_CH_DEFAULT_CTL_RX_H); - /* Enable channel interrupts */ - wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_H); - ret = wcn36xx_dxe_request_irqs(wcn); if (ret < 0) goto out_err_irq; timer_setup(&wcn->tx_ack_timer, wcn36xx_dxe_tx_timer, 0); + /* Enable channel interrupts */ + wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_L); + wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_H); + wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_L); + wcn36xx_dxe_enable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_H); + return 0; out_err_irq: @@ -1034,6 +1044,12 @@ out_err_txh_ch: void wcn36xx_dxe_deinit(struct wcn36xx *wcn) { + /* Disable channel interrupts */ + wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_H); + wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_L); + wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_H); + wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_TX_L); + free_irq(wcn->tx_irq, wcn); free_irq(wcn->rx_irq, wcn); del_timer(&wcn->tx_ack_timer); From 3652096e5263ad67604b0323f71d133485f410e5 Mon Sep 17 00:00:00 2001 From: Bryan O'Donoghue Date: Fri, 5 Nov 2021 12:21:51 +0000 Subject: [PATCH 0101/2048] wcn36xx: Release DMA channel descriptor allocations When unloading the driver we are not releasing the DMA descriptors which we previously allocated. Fixes: 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware") Signed-off-by: Bryan O'Donoghue Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211105122152.1580542-3-bryan.odonoghue@linaro.org --- drivers/net/wireless/ath/wcn36xx/dxe.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/wireless/ath/wcn36xx/dxe.c b/drivers/net/wireless/ath/wcn36xx/dxe.c index d6c621518c7b..d6c951f7dec3 100644 --- a/drivers/net/wireless/ath/wcn36xx/dxe.c +++ b/drivers/net/wireless/ath/wcn36xx/dxe.c @@ -1061,4 +1061,9 @@ void wcn36xx_dxe_deinit(struct wcn36xx *wcn) wcn36xx_dxe_ch_free_skbs(wcn, &wcn->dxe_rx_l_ch); wcn36xx_dxe_ch_free_skbs(wcn, &wcn->dxe_rx_h_ch); + + wcn36xx_dxe_deinit_descs(wcn->dev, &wcn->dxe_tx_l_ch); + wcn36xx_dxe_deinit_descs(wcn->dev, &wcn->dxe_tx_h_ch); + wcn36xx_dxe_deinit_descs(wcn->dev, &wcn->dxe_rx_l_ch); + wcn36xx_dxe_deinit_descs(wcn->dev, &wcn->dxe_rx_h_ch); } From ed04ea76e69e7194f7489cebe23a32a68f39218d Mon Sep 17 00:00:00 2001 From: Bryan O'Donoghue Date: Fri, 5 Nov 2021 12:21:52 +0000 Subject: [PATCH 0102/2048] wcn36xx: Put DXE block into reset before freeing memory When deiniting the DXE hardware we should reset the block to ensure there is no spurious DMA write transaction from the downstream WCNSS to upstream MSM at a skbuff address we will have released. Fixes: 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware") Signed-off-by: Bryan O'Donoghue Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211105122152.1580542-4-bryan.odonoghue@linaro.org --- drivers/net/wireless/ath/wcn36xx/dxe.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/wireless/ath/wcn36xx/dxe.c b/drivers/net/wireless/ath/wcn36xx/dxe.c index d6c951f7dec3..4e9e13941c8f 100644 --- a/drivers/net/wireless/ath/wcn36xx/dxe.c +++ b/drivers/net/wireless/ath/wcn36xx/dxe.c @@ -1044,6 +1044,8 @@ out_err_txh_ch: void wcn36xx_dxe_deinit(struct wcn36xx *wcn) { + int reg_data = 0; + /* Disable channel interrupts */ wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_H); wcn36xx_dxe_disable_ch_int(wcn, WCN36XX_INT_MASK_CHAN_RX_L); @@ -1059,6 +1061,10 @@ void wcn36xx_dxe_deinit(struct wcn36xx *wcn) wcn->tx_ack_skb = NULL; } + /* Put the DXE block into reset before freeing memory */ + reg_data = WCN36XX_DXE_REG_RESET; + wcn36xx_dxe_write_register(wcn, WCN36XX_DXE_REG_CSR_RESET, reg_data); + wcn36xx_dxe_ch_free_skbs(wcn, &wcn->dxe_rx_l_ch); wcn36xx_dxe_ch_free_skbs(wcn, &wcn->dxe_rx_h_ch); From c9c5608fafe4dae975c9644c7d14c51ad3b0ed73 Mon Sep 17 00:00:00 2001 From: Benjamin Li Date: Wed, 3 Nov 2021 18:05:47 -0700 Subject: [PATCH 0103/2048] wcn36xx: populate band before determining rate on RX status.band is used in determination of status.rate -- for 5GHz on legacy rates there is a linear shift between the BD descriptor's rate field and the wcn36xx driver's rate table (wcn_5ghz_rates). We have a special clause to populate status.band for hardware scan offload frames. However, this block occurs after status.rate is already populated. Correctly handle this dependency by moving the band block before the rate block. This patch addresses kernel warnings & missing scan results for 5GHz APs that send their beacons/probe responses at the higher four legacy rates (24-54 Mbps), when using hardware scan offload: ------------[ cut here ]------------ WARNING: CPU: 0 PID: 0 at net/mac80211/rx.c:4532 ieee80211_rx_napi+0x744/0x8d8 Modules linked in: wcn36xx [...] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 4.19.107-g73909fa #1 Hardware name: Square, Inc. T2 (all variants) (DT) Call trace: dump_backtrace+0x0/0x148 show_stack+0x14/0x1c dump_stack+0xb8/0xf0 __warn+0x2ac/0x2d8 warn_slowpath_null+0x44/0x54 ieee80211_rx_napi+0x744/0x8d8 ieee80211_tasklet_handler+0xa4/0xe0 tasklet_action_common+0xe0/0x118 tasklet_action+0x20/0x28 __do_softirq+0x108/0x1ec irq_exit+0xd4/0xd8 __handle_domain_irq+0x84/0xbc gic_handle_irq+0x4c/0xb8 el1_irq+0xe8/0x190 lpm_cpuidle_enter+0x220/0x260 cpuidle_enter_state+0x114/0x1c0 cpuidle_enter+0x34/0x48 do_idle+0x150/0x268 cpu_startup_entry+0x20/0x24 rest_init+0xd4/0xe0 start_kernel+0x398/0x430 ---[ end trace ae28cb759352b403 ]--- Fixes: 8a27ca394782 ("wcn36xx: Correct band/freq reporting on RX") Signed-off-by: Benjamin Li Tested-by: Loic Poulain Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211104010548.1107405-2-benl@squareup.com --- drivers/net/wireless/ath/wcn36xx/txrx.c | 37 +++++++++++++------------ 1 file changed, 19 insertions(+), 18 deletions(-) diff --git a/drivers/net/wireless/ath/wcn36xx/txrx.c b/drivers/net/wireless/ath/wcn36xx/txrx.c index 75951ccbc840..f0a9f069a92a 100644 --- a/drivers/net/wireless/ath/wcn36xx/txrx.c +++ b/drivers/net/wireless/ath/wcn36xx/txrx.c @@ -314,8 +314,6 @@ int wcn36xx_rx_skb(struct wcn36xx *wcn, struct sk_buff *skb) fc = __le16_to_cpu(hdr->frame_control); sn = IEEE80211_SEQ_TO_SN(__le16_to_cpu(hdr->seq_ctrl)); - status.freq = WCN36XX_CENTER_FREQ(wcn); - status.band = WCN36XX_BAND(wcn); status.mactime = 10; status.signal = -get_rssi0(bd); status.antenna = 1; @@ -327,6 +325,25 @@ int wcn36xx_rx_skb(struct wcn36xx *wcn, struct sk_buff *skb) wcn36xx_dbg(WCN36XX_DBG_RX, "status.flags=%x\n", status.flag); + if (bd->scan_learn) { + /* If packet originate from hardware scanning, extract the + * band/channel from bd descriptor. + */ + u8 hwch = (bd->reserved0 << 4) + bd->rx_ch; + + if (bd->rf_band != 1 && hwch <= sizeof(ab_rx_ch_map) && hwch >= 1) { + status.band = NL80211_BAND_5GHZ; + status.freq = ieee80211_channel_to_frequency(ab_rx_ch_map[hwch - 1], + status.band); + } else { + status.band = NL80211_BAND_2GHZ; + status.freq = ieee80211_channel_to_frequency(hwch, status.band); + } + } else { + status.band = WCN36XX_BAND(wcn); + status.freq = WCN36XX_CENTER_FREQ(wcn); + } + if (bd->rate_id < ARRAY_SIZE(wcn36xx_rate_table)) { rate = &wcn36xx_rate_table[bd->rate_id]; status.encoding = rate->encoding; @@ -353,22 +370,6 @@ int wcn36xx_rx_skb(struct wcn36xx *wcn, struct sk_buff *skb) ieee80211_is_probe_resp(hdr->frame_control)) status.boottime_ns = ktime_get_boottime_ns(); - if (bd->scan_learn) { - /* If packet originates from hardware scanning, extract the - * band/channel from bd descriptor. - */ - u8 hwch = (bd->reserved0 << 4) + bd->rx_ch; - - if (bd->rf_band != 1 && hwch <= sizeof(ab_rx_ch_map) && hwch >= 1) { - status.band = NL80211_BAND_5GHZ; - status.freq = ieee80211_channel_to_frequency(ab_rx_ch_map[hwch - 1], - status.band); - } else { - status.band = NL80211_BAND_2GHZ; - status.freq = ieee80211_channel_to_frequency(hwch, status.band); - } - } - memcpy(IEEE80211_SKB_RXCB(skb), &status, sizeof(status)); if (ieee80211_is_beacon(hdr->frame_control)) { From cfdf6b19e750f7de8ae71a26932f63b52e3bf74c Mon Sep 17 00:00:00 2001 From: Benjamin Li Date: Wed, 3 Nov 2021 18:05:48 -0700 Subject: [PATCH 0104/2048] wcn36xx: fix RX BD rate mapping for 5GHz legacy rates The linear mapping between the BD rate field and the driver's 5GHz legacy rates table (wcn_5ghz_rates) does not only apply for the latter four rates -- it applies to all eight rates. Fixes: 6ea131acea98 ("wcn36xx: Fix warning due to bad rate_idx") Signed-off-by: Benjamin Li Tested-by: Loic Poulain Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211104010548.1107405-3-benl@squareup.com --- drivers/net/wireless/ath/wcn36xx/txrx.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/net/wireless/ath/wcn36xx/txrx.c b/drivers/net/wireless/ath/wcn36xx/txrx.c index f0a9f069a92a..dd58dde8c836 100644 --- a/drivers/net/wireless/ath/wcn36xx/txrx.c +++ b/drivers/net/wireless/ath/wcn36xx/txrx.c @@ -272,7 +272,6 @@ int wcn36xx_rx_skb(struct wcn36xx *wcn, struct sk_buff *skb) const struct wcn36xx_rate *rate; struct ieee80211_hdr *hdr; struct wcn36xx_rx_bd *bd; - struct ieee80211_supported_band *sband; u16 fc, sn; /* @@ -350,12 +349,11 @@ int wcn36xx_rx_skb(struct wcn36xx *wcn, struct sk_buff *skb) status.enc_flags = rate->encoding_flags; status.bw = rate->bw; status.rate_idx = rate->mcs_or_legacy_index; - sband = wcn->hw->wiphy->bands[status.band]; status.nss = 1; if (status.band == NL80211_BAND_5GHZ && status.encoding == RX_ENC_LEGACY && - status.rate_idx >= sband->n_bitrates) { + status.rate_idx >= 4) { /* no dsss rates in 5Ghz rates table */ status.rate_idx -= 4; } From b89ddf4cca43f1269093942cf5c4e457fd45c335 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 5 Nov 2021 16:50:45 +0000 Subject: [PATCH 0105/2048] arm64/bpf: Remove 128MB limit for BPF JIT programs Commit 91fc957c9b1d ("arm64/bpf: don't allocate BPF JIT programs in module memory") restricts BPF JIT program allocation to a 128MB region to ensure BPF programs are still in branching range of each other. However this restriction should not apply to the aarch64 JIT, since BPF_JMP | BPF_CALL are implemented as a 64-bit move into a register and then a BLR instruction - which has the effect of being able to call anything without proximity limitation. The practical reason to relax this restriction on JIT memory is that 128MB of JIT memory can be quickly exhausted, especially where PAGE_SIZE is 64KB - one page is needed per program. In cases where seccomp filters are applied to multiple VMs on VM launch - such filters are classic BPF but converted to BPF - this can severely limit the number of VMs that can be launched. In a world where we support BPF JIT always on, turning off the JIT isn't always an option either. Fixes: 91fc957c9b1d ("arm64/bpf: don't allocate BPF JIT programs in module memory") Suggested-by: Ard Biesheuvel Signed-off-by: Russell King Signed-off-by: Daniel Borkmann Tested-by: Alan Maguire Link: https://lore.kernel.org/bpf/1636131046-5982-2-git-send-email-alan.maguire@oracle.com --- arch/arm64/include/asm/extable.h | 9 --------- arch/arm64/include/asm/memory.h | 5 +---- arch/arm64/kernel/traps.c | 2 +- arch/arm64/mm/ptdump.c | 2 -- arch/arm64/net/bpf_jit_comp.c | 7 ++----- 5 files changed, 4 insertions(+), 21 deletions(-) diff --git a/arch/arm64/include/asm/extable.h b/arch/arm64/include/asm/extable.h index 8b300dd28def..72b0e71cc3de 100644 --- a/arch/arm64/include/asm/extable.h +++ b/arch/arm64/include/asm/extable.h @@ -33,15 +33,6 @@ do { \ (b)->data = (tmp).data; \ } while (0) -static inline bool in_bpf_jit(struct pt_regs *regs) -{ - if (!IS_ENABLED(CONFIG_BPF_JIT)) - return false; - - return regs->pc >= BPF_JIT_REGION_START && - regs->pc < BPF_JIT_REGION_END; -} - #ifdef CONFIG_BPF_JIT bool ex_handler_bpf(const struct exception_table_entry *ex, struct pt_regs *regs); diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h index 1b9a1e242612..0af70d9abede 100644 --- a/arch/arm64/include/asm/memory.h +++ b/arch/arm64/include/asm/memory.h @@ -44,11 +44,8 @@ #define _PAGE_OFFSET(va) (-(UL(1) << (va))) #define PAGE_OFFSET (_PAGE_OFFSET(VA_BITS)) #define KIMAGE_VADDR (MODULES_END) -#define BPF_JIT_REGION_START (_PAGE_END(VA_BITS_MIN)) -#define BPF_JIT_REGION_SIZE (SZ_128M) -#define BPF_JIT_REGION_END (BPF_JIT_REGION_START + BPF_JIT_REGION_SIZE) #define MODULES_END (MODULES_VADDR + MODULES_VSIZE) -#define MODULES_VADDR (BPF_JIT_REGION_END) +#define MODULES_VADDR (_PAGE_END(VA_BITS_MIN)) #define MODULES_VSIZE (SZ_128M) #define VMEMMAP_START (-(UL(1) << (VA_BITS - VMEMMAP_SHIFT))) #define VMEMMAP_END (VMEMMAP_START + VMEMMAP_SIZE) diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c index 7b21213a570f..e8986e6067a9 100644 --- a/arch/arm64/kernel/traps.c +++ b/arch/arm64/kernel/traps.c @@ -994,7 +994,7 @@ static struct break_hook bug_break_hook = { static int reserved_fault_handler(struct pt_regs *regs, unsigned int esr) { pr_err("%s generated an invalid instruction at %pS!\n", - in_bpf_jit(regs) ? "BPF JIT" : "Kernel text patching", + "Kernel text patching", (void *)instruction_pointer(regs)); /* We cannot handle this */ diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c index 1c403536c9bb..9bc4066c5bf3 100644 --- a/arch/arm64/mm/ptdump.c +++ b/arch/arm64/mm/ptdump.c @@ -41,8 +41,6 @@ static struct addr_marker address_markers[] = { { 0 /* KASAN_SHADOW_START */, "Kasan shadow start" }, { KASAN_SHADOW_END, "Kasan shadow end" }, #endif - { BPF_JIT_REGION_START, "BPF start" }, - { BPF_JIT_REGION_END, "BPF end" }, { MODULES_VADDR, "Modules start" }, { MODULES_END, "Modules end" }, { VMALLOC_START, "vmalloc() area" }, diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index 3a8a7140a9bf..86c9dc0681cc 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -1141,15 +1141,12 @@ out: u64 bpf_jit_alloc_exec_limit(void) { - return BPF_JIT_REGION_SIZE; + return VMALLOC_END - VMALLOC_START; } void *bpf_jit_alloc_exec(unsigned long size) { - return __vmalloc_node_range(size, PAGE_SIZE, BPF_JIT_REGION_START, - BPF_JIT_REGION_END, GFP_KERNEL, - PAGE_KERNEL, 0, NUMA_NO_NODE, - __builtin_return_address(0)); + return vmalloc(size); } void bpf_jit_free_exec(void *addr) From c23551c9c36ae394f9c53a5adf1944a943c65e0b Mon Sep 17 00:00:00 2001 From: Alan Maguire Date: Fri, 5 Nov 2021 16:50:46 +0000 Subject: [PATCH 0106/2048] selftests/bpf: Add exception handling selftests for tp_bpf program Exception handling is triggered in BPF tracing programs when a NULL pointer is dereferenced; the exception handler zeroes the target register and execution of the BPF program progresses. To test exception handling then, we need to trigger a NULL pointer dereference for a field which should never be zero; if it is, the only explanation is the exception handler ran. task->task_works is the NULL pointer chosen (for a new task from fork() no work is associated), and the task_works->func field should not be zero if task_works is non-NULL. The test verifies that task_works and task_works->func are 0. Signed-off-by: Alan Maguire Signed-off-by: Daniel Borkmann Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/1636131046-5982-3-git-send-email-alan.maguire@oracle.com --- .../selftests/bpf/prog_tests/exhandler.c | 43 +++++++++++++++++++ .../selftests/bpf/progs/exhandler_kern.c | 43 +++++++++++++++++++ 2 files changed, 86 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/exhandler.c create mode 100644 tools/testing/selftests/bpf/progs/exhandler_kern.c diff --git a/tools/testing/selftests/bpf/prog_tests/exhandler.c b/tools/testing/selftests/bpf/prog_tests/exhandler.c new file mode 100644 index 000000000000..118bb182ee20 --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/exhandler.c @@ -0,0 +1,43 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021, Oracle and/or its affiliates. */ + +#include + +/* Test that verifies exception handling is working. fork() + * triggers task_newtask tracepoint; that new task will have a + * NULL pointer task_works, and the associated task->task_works->func + * should not be NULL if task_works itself is non-NULL. + * + * So to verify exception handling we want to see a NULL task_works + * and task_works->func; if we see this we can conclude that the + * exception handler ran when we attempted to dereference task->task_works + * and zeroed the destination register. + */ +#include "exhandler_kern.skel.h" + +void test_exhandler(void) +{ + int err = 0, duration = 0, status; + struct exhandler_kern *skel; + pid_t cpid; + + skel = exhandler_kern__open_and_load(); + if (CHECK(!skel, "skel_load", "skeleton failed: %d\n", err)) + goto cleanup; + + skel->bss->test_pid = getpid(); + + err = exhandler_kern__attach(skel); + if (!ASSERT_OK(err, "attach")) + goto cleanup; + cpid = fork(); + if (!ASSERT_GT(cpid, -1, "fork failed")) + goto cleanup; + if (cpid == 0) + _exit(0); + waitpid(cpid, &status, 0); + + ASSERT_NEQ(skel->bss->exception_triggered, 0, "verify exceptions occurred"); +cleanup: + exhandler_kern__destroy(skel); +} diff --git a/tools/testing/selftests/bpf/progs/exhandler_kern.c b/tools/testing/selftests/bpf/progs/exhandler_kern.c new file mode 100644 index 000000000000..f5ca142abf8f --- /dev/null +++ b/tools/testing/selftests/bpf/progs/exhandler_kern.c @@ -0,0 +1,43 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021, Oracle and/or its affiliates. */ + +#include "vmlinux.h" + +#include +#include +#include + +char _license[] SEC("license") = "GPL"; + +unsigned int exception_triggered; +int test_pid; + +/* TRACE_EVENT(task_newtask, + * TP_PROTO(struct task_struct *p, u64 clone_flags) + */ +SEC("tp_btf/task_newtask") +int BPF_PROG(trace_task_newtask, struct task_struct *task, u64 clone_flags) +{ + int pid = bpf_get_current_pid_tgid() >> 32; + struct callback_head *work; + void *func; + + if (test_pid != pid) + return 0; + + /* To verify we hit an exception we dereference task->task_works->func. + * If task work has been added, + * - task->task_works is non-NULL; and + * - task->task_works->func is non-NULL also (the callback function + * must be specified for the task work. + * + * However, for a newly-created task, task->task_works is NULLed, + * so we know the exception handler triggered if task_works is + * NULL and func is NULL. + */ + work = task->task_works; + func = work->func; + if (!work && !func) + exception_triggered++; + return 0; +} From 3a74ac2d1159716f35c944639f71b33fa16084c8 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Sat, 6 Nov 2021 05:12:40 +0530 Subject: [PATCH 0107/2048] libbpf: Compile using -std=gnu89 The minimum supported C standard version is C89, with use of GNU extensions, hence make sure to catch any instances that would break the build for this mode by passing -std=gnu89. Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211105234243.390179-4-memxor@gmail.com --- tools/lib/bpf/Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/lib/bpf/Makefile b/tools/lib/bpf/Makefile index b393b5e82380..5f7086fae31c 100644 --- a/tools/lib/bpf/Makefile +++ b/tools/lib/bpf/Makefile @@ -84,6 +84,7 @@ else endif # Append required CFLAGS +override CFLAGS += -std=gnu89 override CFLAGS += $(EXTRA_WARNINGS) -Wno-switch-enum override CFLAGS += -Werror -Wall override CFLAGS += $(INCLUDES) From af3d89649bb69bd5be273cf6c001cd19c2604ca2 Mon Sep 17 00:00:00 2001 From: Kalle Valo Date: Mon, 8 Nov 2021 14:38:25 +0200 Subject: [PATCH 0108/2048] ath11k: convert ath11k_wmi_pdev_set_ps_mode() to use enum wmi_sta_ps_mode It's more descriptive to use the actual enum used by the firmware instead of a boolean so change ath11k_wmi_pdev_set_ps_mode() to use a boolean. Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1 Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211108123826.8463-1-kvalo@codeaurora.org --- drivers/net/wireless/ath/ath11k/mac.c | 3 ++- drivers/net/wireless/ath/ath11k/wmi.c | 7 ++++--- drivers/net/wireless/ath/ath11k/wmi.h | 3 ++- 3 files changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 50a6f88115cd..a1d85a09ec35 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -5638,7 +5638,8 @@ static int ath11k_mac_op_add_interface(struct ieee80211_hw *hw, goto err_peer_del; } - ret = ath11k_wmi_pdev_set_ps_mode(ar, arvif->vdev_id, false); + ret = ath11k_wmi_pdev_set_ps_mode(ar, arvif->vdev_id, + WMI_STA_PS_MODE_DISABLED); if (ret) { ath11k_warn(ar->ab, "failed to disable vdev %d ps mode: %d\n", arvif->vdev_id, ret); diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 5ae2ef4680d6..b9bfe0407cd7 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -1244,7 +1244,8 @@ int ath11k_wmi_pdev_set_param(struct ath11k *ar, u32 param_id, return ret; } -int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, u32 enable) +int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, + enum wmi_sta_ps_mode psmode) { struct ath11k_pdev_wmi *wmi = ar->wmi; struct wmi_pdev_set_ps_mode_cmd *cmd; @@ -1259,7 +1260,7 @@ int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, u32 enable) cmd->tlv_header = FIELD_PREP(WMI_TLV_TAG, WMI_TAG_STA_POWERSAVE_MODE_CMD) | FIELD_PREP(WMI_TLV_LEN, sizeof(*cmd) - TLV_HDR_SIZE); cmd->vdev_id = vdev_id; - cmd->sta_ps_mode = enable; + cmd->sta_ps_mode = psmode; ret = ath11k_wmi_cmd_send(wmi, skb, WMI_STA_POWERSAVE_MODE_CMDID); if (ret) { @@ -1269,7 +1270,7 @@ int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, u32 enable) ath11k_dbg(ar->ab, ATH11K_DBG_WMI, "WMI vdev set psmode %d vdev id %d\n", - enable, vdev_id); + psmode, vdev_id); return ret; } diff --git a/drivers/net/wireless/ath/ath11k/wmi.h b/drivers/net/wireless/ath/ath11k/wmi.h index 0584e68e7593..93e2992db862 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.h +++ b/drivers/net/wireless/ath/ath11k/wmi.h @@ -5351,7 +5351,8 @@ int ath11k_wmi_set_peer_param(struct ath11k *ar, const u8 *peer_addr, u32 vdev_id, u32 param_id, u32 param_val); int ath11k_wmi_pdev_set_param(struct ath11k *ar, u32 param_id, u32 param_value, u8 pdev_id); -int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, u32 enable); +int ath11k_wmi_pdev_set_ps_mode(struct ath11k *ar, int vdev_id, + enum wmi_sta_ps_mode psmode); int ath11k_wmi_wait_for_unified_ready(struct ath11k_base *ab); int ath11k_wmi_cmd_init(struct ath11k_base *ab); int ath11k_wmi_wait_for_service_ready(struct ath11k_base *ab); From b2beffa7d9a67b59b085616a27f1d10b1e80784f Mon Sep 17 00:00:00 2001 From: Carl Huang Date: Mon, 8 Nov 2021 14:38:26 +0200 Subject: [PATCH 0109/2048] ath11k: enable 802.11 power save mode in station mode To reduce power consumption enable 802.11 power save mode in station mode. This allows both radio and CPU to sleep more. Only enable the mode on QCA6390 and WCN6855, it's unknown how other hardware families support this feature. To test that power save mode is running run "iw dev wls1 set power_save off", check there is no NULL Data frame seen by a sniffer. And run "iw dev wls1 set power_save on" and check there is a NULL Data frame in sniffer. Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1 Signed-off-by: Carl Huang Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211108123826.8463-2-kvalo@codeaurora.org --- drivers/net/wireless/ath/ath11k/core.c | 5 ++ drivers/net/wireless/ath/ath11k/core.h | 1 + drivers/net/wireless/ath/ath11k/hw.h | 1 + drivers/net/wireless/ath/ath11k/mac.c | 87 ++++++++++++++++++++++++++ 4 files changed, 94 insertions(+) diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index 7ee2ccc49c74..bd77aedc7fe6 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -76,6 +76,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_monitor = true, .supports_shadow_regs = false, .idle_ps = false, + .supports_sta_ps = false, .cold_boot_calib = true, .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), @@ -126,6 +127,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_monitor = true, .supports_shadow_regs = false, .idle_ps = false, + .supports_sta_ps = false, .cold_boot_calib = true, .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), @@ -175,6 +177,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_monitor = false, .supports_shadow_regs = true, .idle_ps = true, + .supports_sta_ps = true, .cold_boot_calib = false, .supports_suspend = true, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), @@ -224,6 +227,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_monitor = true, .supports_shadow_regs = false, .idle_ps = false, + .supports_sta_ps = false, .cold_boot_calib = false, .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_qcn9074), @@ -273,6 +277,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_monitor = false, .supports_shadow_regs = true, .idle_ps = true, + .supports_sta_ps = true, .cold_boot_calib = false, .supports_suspend = true, .hal_desc_sz = sizeof(struct hal_rx_desc_wcn6855), diff --git a/drivers/net/wireless/ath/ath11k/core.h b/drivers/net/wireless/ath/ath11k/core.h index 31d234a51c79..2f1e10b7cc17 100644 --- a/drivers/net/wireless/ath/ath11k/core.h +++ b/drivers/net/wireless/ath/ath11k/core.h @@ -240,6 +240,7 @@ struct ath11k_vif { bool is_started; bool is_up; bool spectral_enabled; + bool ps; u32 aid; u8 bssid[ETH_ALEN]; struct cfg80211_bitrate_mask bitrate_mask; diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h index 6dcac596e3fe..2c7bd7a36ba5 100644 --- a/drivers/net/wireless/ath/ath11k/hw.h +++ b/drivers/net/wireless/ath/ath11k/hw.h @@ -170,6 +170,7 @@ struct ath11k_hw_params { bool supports_monitor; bool supports_shadow_regs; bool idle_ps; + bool supports_sta_ps; bool cold_boot_calib; bool supports_suspend; u32 hal_desc_sz; diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index a1d85a09ec35..89810b202a39 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -1049,6 +1049,83 @@ static int ath11k_mac_monitor_stop(struct ath11k *ar) return 0; } +static int ath11k_mac_vif_setup_ps(struct ath11k_vif *arvif) +{ + struct ath11k *ar = arvif->ar; + struct ieee80211_vif *vif = arvif->vif; + struct ieee80211_conf *conf = &ar->hw->conf; + enum wmi_sta_powersave_param param; + enum wmi_sta_ps_mode psmode; + int ret; + int timeout; + bool enable_ps; + + lockdep_assert_held(&arvif->ar->conf_mutex); + + if (arvif->vif->type != NL80211_IFTYPE_STATION) + return 0; + + enable_ps = arvif->ps; + + if (!arvif->is_started) { + /* mac80211 can update vif powersave state while disconnected. + * Firmware doesn't behave nicely and consumes more power than + * necessary if PS is disabled on a non-started vdev. Hence + * force-enable PS for non-running vdevs. + */ + psmode = WMI_STA_PS_MODE_ENABLED; + } else if (enable_ps) { + psmode = WMI_STA_PS_MODE_ENABLED; + param = WMI_STA_PS_PARAM_INACTIVITY_TIME; + + timeout = conf->dynamic_ps_timeout; + if (timeout == 0) { + /* firmware doesn't like 0 */ + timeout = ieee80211_tu_to_usec(vif->bss_conf.beacon_int) / 1000; + } + + ret = ath11k_wmi_set_sta_ps_param(ar, arvif->vdev_id, param, + timeout); + if (ret) { + ath11k_warn(ar->ab, "failed to set inactivity time for vdev %d: %i\n", + arvif->vdev_id, ret); + return ret; + } + } else { + psmode = WMI_STA_PS_MODE_DISABLED; + } + + ath11k_dbg(ar->ab, ATH11K_DBG_MAC, "mac vdev %d psmode %s\n", + arvif->vdev_id, psmode ? "enable" : "disable"); + + ret = ath11k_wmi_pdev_set_ps_mode(ar, arvif->vdev_id, psmode); + if (ret) { + ath11k_warn(ar->ab, "failed to set sta power save mode %d for vdev %d: %d\n", + psmode, arvif->vdev_id, ret); + return ret; + } + + return 0; +} + +static int ath11k_mac_config_ps(struct ath11k *ar) +{ + struct ath11k_vif *arvif; + int ret = 0; + + lockdep_assert_held(&ar->conf_mutex); + + list_for_each_entry(arvif, &ar->arvifs, list) { + ret = ath11k_mac_vif_setup_ps(arvif); + if (ret) { + ath11k_warn(ar->ab, "failed to setup powersave: %d\n", ret); + break; + } + } + + return ret; +} + static int ath11k_mac_op_config(struct ieee80211_hw *hw, u32 changed) { struct ath11k *ar = hw->priv; @@ -2942,6 +3019,16 @@ static void ath11k_mac_op_bss_info_changed(struct ieee80211_hw *hw, ath11k_mac_txpower_recalc(ar); } + if (changed & BSS_CHANGED_PS && + ar->ab->hw_params.supports_sta_ps) { + arvif->ps = vif->bss_conf.ps; + + ret = ath11k_mac_config_ps(ar); + if (ret) + ath11k_warn(ar->ab, "failed to setup ps on vdev %i: %d\n", + arvif->vdev_id, ret); + } + if (changed & BSS_CHANGED_MCAST_RATE && !ath11k_mac_vif_chan(arvif->vif, &def)) { band = def.chan->band; From 1a8b597ddabe7dc25aa9defd33949d455ee9cde8 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 5 Nov 2021 22:19:04 +0000 Subject: [PATCH 0110/2048] bpftool: Fix SPDX tag for Makefiles and .gitignore Bpftool is dual-licensed under GPLv2 and BSD-2-Clause. In commit 907b22365115 ("tools: bpftool: dual license all files") we made sure that all its source files were indeed covered by the two licenses, and that they had the correct SPDX tags. However, bpftool's Makefile, the Makefile for its documentation, and the .gitignore file were skipped at the time (their GPL-2.0-only tag was added later). Let's update the tags. Signed-off-by: Quentin Monnet Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Acked-by: Tobias Klauser Acked-by: Joe Stringer Acked-by: Song Liu Acked-by: Jean-Philippe Brucker Acked-by: Jesper Dangaard Brouer Acked-by: Jakub Kicinski Link: https://lore.kernel.org/bpf/20211105221904.3536-1-quentin@isovalent.com --- tools/bpf/bpftool/.gitignore | 2 +- tools/bpf/bpftool/Documentation/Makefile | 2 +- tools/bpf/bpftool/Makefile | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/bpf/bpftool/.gitignore b/tools/bpf/bpftool/.gitignore index 05ce4446b780..a736f64dc5dc 100644 --- a/tools/bpf/bpftool/.gitignore +++ b/tools/bpf/bpftool/.gitignore @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) *.d /bootstrap/ /bpftool diff --git a/tools/bpf/bpftool/Documentation/Makefile b/tools/bpf/bpftool/Documentation/Makefile index c49487905ceb..44b60784847b 100644 --- a/tools/bpf/bpftool/Documentation/Makefile +++ b/tools/bpf/bpftool/Documentation/Makefile @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) include ../../../scripts/Makefile.include include ../../../scripts/utilities.mak diff --git a/tools/bpf/bpftool/Makefile b/tools/bpf/bpftool/Makefile index c0c30e56988f..622568c7a9b8 100644 --- a/tools/bpf/bpftool/Makefile +++ b/tools/bpf/bpftool/Makefile @@ -1,4 +1,4 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) include ../../scripts/Makefile.include include ../../scripts/utilities.mak From f89315650ba34ec6c91a8bded72796980bee2a4d Mon Sep 17 00:00:00 2001 From: Mark Pashmfouroush Date: Wed, 10 Nov 2021 11:10:15 +0000 Subject: [PATCH 0111/2048] bpf: Add ingress_ifindex to bpf_sk_lookup It may be helpful to have access to the ifindex during bpf socket lookup. An example may be to scope certain socket lookup logic to specific interfaces, i.e. an interface may be made exempt from custom lookup code. Add the ifindex of the arriving connection to the bpf_sk_lookup API. Signed-off-by: Mark Pashmfouroush Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211110111016.5670-2-markpash@cloudflare.com --- include/linux/filter.h | 7 +++++-- include/uapi/linux/bpf.h | 1 + net/core/filter.c | 7 +++++++ net/ipv4/inet_hashtables.c | 8 ++++---- net/ipv4/udp.c | 8 ++++---- net/ipv6/inet6_hashtables.c | 8 ++++---- net/ipv6/udp.c | 8 ++++---- tools/include/uapi/linux/bpf.h | 1 + 8 files changed, 30 insertions(+), 18 deletions(-) diff --git a/include/linux/filter.h b/include/linux/filter.h index 24b7ed2677af..b6a216eb217a 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -1374,6 +1374,7 @@ struct bpf_sk_lookup_kern { const struct in6_addr *daddr; } v6; struct sock *selected_sk; + u32 ingress_ifindex; bool no_reuseport; }; @@ -1436,7 +1437,7 @@ extern struct static_key_false bpf_sk_lookup_enabled; static inline bool bpf_sk_lookup_run_v4(struct net *net, int protocol, const __be32 saddr, const __be16 sport, const __be32 daddr, const u16 dport, - struct sock **psk) + const int ifindex, struct sock **psk) { struct bpf_prog_array *run_array; struct sock *selected_sk = NULL; @@ -1452,6 +1453,7 @@ static inline bool bpf_sk_lookup_run_v4(struct net *net, int protocol, .v4.daddr = daddr, .sport = sport, .dport = dport, + .ingress_ifindex = ifindex, }; u32 act; @@ -1474,7 +1476,7 @@ static inline bool bpf_sk_lookup_run_v6(struct net *net, int protocol, const __be16 sport, const struct in6_addr *daddr, const u16 dport, - struct sock **psk) + const int ifindex, struct sock **psk) { struct bpf_prog_array *run_array; struct sock *selected_sk = NULL; @@ -1490,6 +1492,7 @@ static inline bool bpf_sk_lookup_run_v6(struct net *net, int protocol, .v6.daddr = daddr, .sport = sport, .dport = dport, + .ingress_ifindex = ifindex, }; u32 act; diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 509eee5f0393..6297eafdc40f 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -6316,6 +6316,7 @@ struct bpf_sk_lookup { __u32 local_ip4; /* Network byte order */ __u32 local_ip6[4]; /* Network byte order */ __u32 local_port; /* Host byte order */ + __u32 ingress_ifindex; /* The arriving interface. Determined by inet_iif. */ }; /* diff --git a/net/core/filter.c b/net/core/filter.c index 8e8d3b49c297..315a58466fc9 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -10491,6 +10491,7 @@ static bool sk_lookup_is_valid_access(int off, int size, case bpf_ctx_range_till(struct bpf_sk_lookup, local_ip6[0], local_ip6[3]): case bpf_ctx_range(struct bpf_sk_lookup, remote_port): case bpf_ctx_range(struct bpf_sk_lookup, local_port): + case bpf_ctx_range(struct bpf_sk_lookup, ingress_ifindex): bpf_ctx_record_field_size(info, sizeof(__u32)); return bpf_ctx_narrow_access_ok(off, size, sizeof(__u32)); @@ -10580,6 +10581,12 @@ static u32 sk_lookup_convert_ctx_access(enum bpf_access_type type, bpf_target_off(struct bpf_sk_lookup_kern, dport, 2, target_size)); break; + + case offsetof(struct bpf_sk_lookup, ingress_ifindex): + *insn++ = BPF_LDX_MEM(BPF_W, si->dst_reg, si->src_reg, + bpf_target_off(struct bpf_sk_lookup_kern, + ingress_ifindex, 4, target_size)); + break; } return insn - insn_buf; diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c index 75737267746f..30ab717ff1b8 100644 --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -307,7 +307,7 @@ static inline struct sock *inet_lookup_run_bpf(struct net *net, struct inet_hashinfo *hashinfo, struct sk_buff *skb, int doff, __be32 saddr, __be16 sport, - __be32 daddr, u16 hnum) + __be32 daddr, u16 hnum, const int dif) { struct sock *sk, *reuse_sk; bool no_reuseport; @@ -315,8 +315,8 @@ static inline struct sock *inet_lookup_run_bpf(struct net *net, if (hashinfo != &tcp_hashinfo) return NULL; /* only TCP is supported */ - no_reuseport = bpf_sk_lookup_run_v4(net, IPPROTO_TCP, - saddr, sport, daddr, hnum, &sk); + no_reuseport = bpf_sk_lookup_run_v4(net, IPPROTO_TCP, saddr, sport, + daddr, hnum, dif, &sk); if (no_reuseport || IS_ERR_OR_NULL(sk)) return sk; @@ -340,7 +340,7 @@ struct sock *__inet_lookup_listener(struct net *net, /* Lookup redirect from BPF */ if (static_branch_unlikely(&bpf_sk_lookup_enabled)) { result = inet_lookup_run_bpf(net, hashinfo, skb, doff, - saddr, sport, daddr, hnum); + saddr, sport, daddr, hnum, dif); if (result) goto done; } diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 2fffcf2b54f3..5fceee3de65d 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -460,7 +460,7 @@ static struct sock *udp4_lookup_run_bpf(struct net *net, struct udp_table *udptable, struct sk_buff *skb, __be32 saddr, __be16 sport, - __be32 daddr, u16 hnum) + __be32 daddr, u16 hnum, const int dif) { struct sock *sk, *reuse_sk; bool no_reuseport; @@ -468,8 +468,8 @@ static struct sock *udp4_lookup_run_bpf(struct net *net, if (udptable != &udp_table) return NULL; /* only UDP is supported */ - no_reuseport = bpf_sk_lookup_run_v4(net, IPPROTO_UDP, - saddr, sport, daddr, hnum, &sk); + no_reuseport = bpf_sk_lookup_run_v4(net, IPPROTO_UDP, saddr, sport, + daddr, hnum, dif, &sk); if (no_reuseport || IS_ERR_OR_NULL(sk)) return sk; @@ -505,7 +505,7 @@ struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr, /* Lookup redirect from BPF */ if (static_branch_unlikely(&bpf_sk_lookup_enabled)) { sk = udp4_lookup_run_bpf(net, udptable, skb, - saddr, sport, daddr, hnum); + saddr, sport, daddr, hnum, dif); if (sk) { result = sk; goto done; diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c index 67c9114835c8..4514444e96c8 100644 --- a/net/ipv6/inet6_hashtables.c +++ b/net/ipv6/inet6_hashtables.c @@ -165,7 +165,7 @@ static inline struct sock *inet6_lookup_run_bpf(struct net *net, const struct in6_addr *saddr, const __be16 sport, const struct in6_addr *daddr, - const u16 hnum) + const u16 hnum, const int dif) { struct sock *sk, *reuse_sk; bool no_reuseport; @@ -173,8 +173,8 @@ static inline struct sock *inet6_lookup_run_bpf(struct net *net, if (hashinfo != &tcp_hashinfo) return NULL; /* only TCP is supported */ - no_reuseport = bpf_sk_lookup_run_v6(net, IPPROTO_TCP, - saddr, sport, daddr, hnum, &sk); + no_reuseport = bpf_sk_lookup_run_v6(net, IPPROTO_TCP, saddr, sport, + daddr, hnum, dif, &sk); if (no_reuseport || IS_ERR_OR_NULL(sk)) return sk; @@ -198,7 +198,7 @@ struct sock *inet6_lookup_listener(struct net *net, /* Lookup redirect from BPF */ if (static_branch_unlikely(&bpf_sk_lookup_enabled)) { result = inet6_lookup_run_bpf(net, hashinfo, skb, doff, - saddr, sport, daddr, hnum); + saddr, sport, daddr, hnum, dif); if (result) goto done; } diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 12c12619ee35..ea4ea525f94a 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -195,7 +195,7 @@ static inline struct sock *udp6_lookup_run_bpf(struct net *net, const struct in6_addr *saddr, __be16 sport, const struct in6_addr *daddr, - u16 hnum) + u16 hnum, const int dif) { struct sock *sk, *reuse_sk; bool no_reuseport; @@ -203,8 +203,8 @@ static inline struct sock *udp6_lookup_run_bpf(struct net *net, if (udptable != &udp_table) return NULL; /* only UDP is supported */ - no_reuseport = bpf_sk_lookup_run_v6(net, IPPROTO_UDP, - saddr, sport, daddr, hnum, &sk); + no_reuseport = bpf_sk_lookup_run_v6(net, IPPROTO_UDP, saddr, sport, + daddr, hnum, dif, &sk); if (no_reuseport || IS_ERR_OR_NULL(sk)) return sk; @@ -240,7 +240,7 @@ struct sock *__udp6_lib_lookup(struct net *net, /* Lookup redirect from BPF */ if (static_branch_unlikely(&bpf_sk_lookup_enabled)) { sk = udp6_lookup_run_bpf(net, udptable, skb, - saddr, sport, daddr, hnum); + saddr, sport, daddr, hnum, dif); if (sk) { result = sk; goto done; diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 509eee5f0393..6297eafdc40f 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -6316,6 +6316,7 @@ struct bpf_sk_lookup { __u32 local_ip4; /* Network byte order */ __u32 local_ip6[4]; /* Network byte order */ __u32 local_port; /* Host byte order */ + __u32 ingress_ifindex; /* The arriving interface. Determined by inet_iif. */ }; /* From 8b4fd2bf1f47c3e3a63c327fca2ad5c4e2691ef8 Mon Sep 17 00:00:00 2001 From: Mark Pashmfouroush Date: Wed, 10 Nov 2021 11:10:16 +0000 Subject: [PATCH 0112/2048] selftests/bpf: Add tests for accessing ingress_ifindex in bpf_sk_lookup A new field was added to the bpf_sk_lookup data that users can access. Add tests that validate that the new ingress_ifindex field contains the right data. Signed-off-by: Mark Pashmfouroush Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211110111016.5670-3-markpash@cloudflare.com --- .../selftests/bpf/prog_tests/sk_lookup.c | 31 ++++++++++++++++++ .../selftests/bpf/progs/test_sk_lookup.c | 8 +++++ .../selftests/bpf/verifier/ctx_sk_lookup.c | 32 +++++++++++++++++++ 3 files changed, 71 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c index 6db07401bc49..57846cc7ce36 100644 --- a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c +++ b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c @@ -937,6 +937,37 @@ static void test_drop_on_lookup(struct test_sk_lookup *skel) .connect_to = { EXT_IP6, EXT_PORT }, .listen_at = { EXT_IP6, INT_PORT }, }, + /* The program will drop on success, meaning that the ifindex + * was 1. + */ + { + .desc = "TCP IPv4 drop on valid ifindex", + .lookup_prog = skel->progs.check_ifindex, + .sotype = SOCK_STREAM, + .connect_to = { EXT_IP4, EXT_PORT }, + .listen_at = { EXT_IP4, EXT_PORT }, + }, + { + .desc = "TCP IPv6 drop on valid ifindex", + .lookup_prog = skel->progs.check_ifindex, + .sotype = SOCK_STREAM, + .connect_to = { EXT_IP6, EXT_PORT }, + .listen_at = { EXT_IP6, EXT_PORT }, + }, + { + .desc = "UDP IPv4 drop on valid ifindex", + .lookup_prog = skel->progs.check_ifindex, + .sotype = SOCK_DGRAM, + .connect_to = { EXT_IP4, EXT_PORT }, + .listen_at = { EXT_IP4, EXT_PORT }, + }, + { + .desc = "UDP IPv6 drop on valid ifindex", + .lookup_prog = skel->progs.check_ifindex, + .sotype = SOCK_DGRAM, + .connect_to = { EXT_IP6, EXT_PORT }, + .listen_at = { EXT_IP6, EXT_PORT }, + }, }; const struct test *t; diff --git a/tools/testing/selftests/bpf/progs/test_sk_lookup.c b/tools/testing/selftests/bpf/progs/test_sk_lookup.c index 19d2465d9442..83b0aaa52ef7 100644 --- a/tools/testing/selftests/bpf/progs/test_sk_lookup.c +++ b/tools/testing/selftests/bpf/progs/test_sk_lookup.c @@ -84,6 +84,14 @@ int lookup_drop(struct bpf_sk_lookup *ctx) return SK_DROP; } +SEC("sk_lookup") +int check_ifindex(struct bpf_sk_lookup *ctx) +{ + if (ctx->ingress_ifindex == 1) + return SK_DROP; + return SK_PASS; +} + SEC("sk_reuseport") int reuseport_pass(struct sk_reuseport_md *ctx) { diff --git a/tools/testing/selftests/bpf/verifier/ctx_sk_lookup.c b/tools/testing/selftests/bpf/verifier/ctx_sk_lookup.c index d78627be060f..a2b006e2fd06 100644 --- a/tools/testing/selftests/bpf/verifier/ctx_sk_lookup.c +++ b/tools/testing/selftests/bpf/verifier/ctx_sk_lookup.c @@ -229,6 +229,24 @@ BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1, offsetof(struct bpf_sk_lookup, local_port)), + /* 1-byte read from ingress_ifindex field */ + BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex)), + BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex) + 1), + BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex) + 2), + BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex) + 3), + /* 2-byte read from ingress_ifindex field */ + BPF_LDX_MEM(BPF_H, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex)), + BPF_LDX_MEM(BPF_H, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex) + 2), + /* 4-byte read from ingress_ifindex field */ + BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex)), + /* 8-byte read from sk field */ BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, offsetof(struct bpf_sk_lookup, sk)), @@ -351,6 +369,20 @@ .expected_attach_type = BPF_SK_LOOKUP, .flags = F_NEEDS_EFFICIENT_UNALIGNED_ACCESS, }, +{ + "invalid 8-byte read from bpf_sk_lookup ingress_ifindex field", + .insns = { + BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, + offsetof(struct bpf_sk_lookup, ingress_ifindex)), + BPF_MOV32_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .errstr = "invalid bpf_context access", + .result = REJECT, + .prog_type = BPF_PROG_TYPE_SK_LOOKUP, + .expected_attach_type = BPF_SK_LOOKUP, + .flags = F_NEEDS_EFFICIENT_UNALIGNED_ACCESS, +}, /* invalid 1,2,4-byte reads from 8-byte fields in bpf_sk_lookup */ { "invalid 4-byte read from bpf_sk_lookup sk field", From a6ca71583137300f207343d5d950cb1c365ab911 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:17:57 -0800 Subject: [PATCH 0113/2048] libbpf: Add ability to get/set per-program load flags Add bpf_program__flags() API to retrieve prog_flags that will be (or were) supplied to BPF_PROG_LOAD command. Also add bpf_program__set_extra_flags() API to allow to set *extra* flags, in addition to those determined by program's SEC() definition. Such flags are logically OR'ed with libbpf-derived flags. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111051758.92283-2-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 14 ++++++++++++++ tools/lib/bpf/libbpf.h | 3 +++ tools/lib/bpf/libbpf.map | 2 ++ 3 files changed, 19 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index d869ebee1e27..a823b5ed705b 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -8262,6 +8262,20 @@ void bpf_program__set_expected_attach_type(struct bpf_program *prog, prog->expected_attach_type = type; } +__u32 bpf_program__flags(const struct bpf_program *prog) +{ + return prog->prog_flags; +} + +int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags) +{ + if (prog->obj->loaded) + return libbpf_err(-EBUSY); + + prog->prog_flags |= extra_flags; + return 0; +} + #define SEC_DEF(sec_pfx, ptype, atype, flags, ...) { \ .sec = sec_pfx, \ .prog_type = BPF_PROG_TYPE_##ptype, \ diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 039058763173..f69512ae8505 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -493,6 +493,9 @@ LIBBPF_API void bpf_program__set_expected_attach_type(struct bpf_program *prog, enum bpf_attach_type type); +LIBBPF_API __u32 bpf_program__flags(const struct bpf_program *prog); +LIBBPF_API int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags); + LIBBPF_API int bpf_program__set_attach_target(struct bpf_program *prog, int attach_prog_fd, const char *attach_func_name); diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index b895861a13c0..256d5d4be951 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -397,8 +397,10 @@ LIBBPF_0.6.0 { bpf_object__prev_program; bpf_prog_load_deprecated; bpf_prog_load; + bpf_program__flags; bpf_program__insn_cnt; bpf_program__insns; + bpf_program__set_extra_flags; btf__add_btf; btf__add_decl_tag; btf__raw_data; From 50dee7078b66d881c62f6177844d625f7ead6003 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:17:58 -0800 Subject: [PATCH 0114/2048] selftests/bpf: Fix bpf_prog_test_load() logic to pass extra log level After recent refactoring bpf_prog_test_load(), used across multiple selftests, lost ability to specify extra log_level 1 or 2 (for -vv and -vvv, respectively). Fix that problem by using bpf_object__load_xattr() API that supports extra log_level flags. Also restore BPF_F_TEST_RND_HI32 prog_flags by utilizing new bpf_program__set_extra_flags() API. Fixes: f87c1930ac29 ("selftests/bpf: Merge test_stub.c into testing_helpers.c") Reported-by: Alexei Starovoitov Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111051758.92283-3-andrii@kernel.org --- tools/testing/selftests/bpf/testing_helpers.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/testing_helpers.c b/tools/testing/selftests/bpf/testing_helpers.c index ef61d43adfe4..52c2f24e0898 100644 --- a/tools/testing/selftests/bpf/testing_helpers.c +++ b/tools/testing/selftests/bpf/testing_helpers.c @@ -88,6 +88,7 @@ int extra_prog_load_log_flags = 0; int bpf_prog_test_load(const char *file, enum bpf_prog_type type, struct bpf_object **pobj, int *prog_fd) { + struct bpf_object_load_attr attr = {}; struct bpf_object *obj; struct bpf_program *prog; int err; @@ -105,7 +106,11 @@ int bpf_prog_test_load(const char *file, enum bpf_prog_type type, if (type != BPF_PROG_TYPE_UNSPEC) bpf_program__set_type(prog, type); - err = bpf_object__load(obj); + bpf_program__set_extra_flags(prog, BPF_F_TEST_RND_HI32); + + attr.obj = obj; + attr.log_level = extra_prog_load_log_flags; + err = bpf_object__load_xattr(&attr); if (err) goto err_out; From 6501182c08f76e4118177b42614174547e1bd149 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:16 -0800 Subject: [PATCH 0115/2048] bpftool: Normalize compile rules to specify output file last When dealing with verbose Makefile output, it's extremely confusing when compiler invocation commands don't specify -o as the last argument. Normalize bpftool's Makefile to do just that, as most other BPF-related Makefiles are already doing that. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-2-andrii@kernel.org --- tools/bpf/bpftool/Makefile | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/tools/bpf/bpftool/Makefile b/tools/bpf/bpftool/Makefile index 622568c7a9b8..31dfef6a4121 100644 --- a/tools/bpf/bpftool/Makefile +++ b/tools/bpf/bpftool/Makefile @@ -177,7 +177,8 @@ $(OUTPUT)%.bpf.o: skeleton/%.bpf.c $(OUTPUT)vmlinux.h $(LIBBPF) -I$(if $(OUTPUT),$(OUTPUT),.) \ -I$(srctree)/tools/include/uapi/ \ -I$(LIBBPF_INCLUDE) \ - -g -O2 -Wall -target bpf -c $< -o $@ && $(LLVM_STRIP) -g $@ + -g -O2 -Wall -target bpf -c $< -o $@ + $(Q)$(LLVM_STRIP) -g $@ $(OUTPUT)%.skel.h: $(OUTPUT)%.bpf.o $(BPFTOOL_BOOTSTRAP) $(QUIET_GEN)$(BPFTOOL_BOOTSTRAP) gen skeleton $< > $@ @@ -192,10 +193,10 @@ endif CFLAGS += $(if $(BUILD_BPF_SKELS),,-DBPFTOOL_WITHOUT_SKELETONS) $(BOOTSTRAP_OUTPUT)disasm.o: $(srctree)/kernel/bpf/disasm.c - $(QUIET_CC)$(HOSTCC) $(CFLAGS) -c -MMD -o $@ $< + $(QUIET_CC)$(HOSTCC) $(CFLAGS) -c -MMD $< -o $@ $(OUTPUT)disasm.o: $(srctree)/kernel/bpf/disasm.c - $(QUIET_CC)$(CC) $(CFLAGS) -c -MMD -o $@ $< + $(QUIET_CC)$(CC) $(CFLAGS) -c -MMD $< -o $@ $(OUTPUT)feature.o: ifneq ($(feature-zlib), 1) @@ -203,17 +204,16 @@ ifneq ($(feature-zlib), 1) endif $(BPFTOOL_BOOTSTRAP): $(BOOTSTRAP_OBJS) $(LIBBPF_BOOTSTRAP) - $(QUIET_LINK)$(HOSTCC) $(CFLAGS) $(LDFLAGS) -o $@ $(BOOTSTRAP_OBJS) \ - $(LIBS_BOOTSTRAP) + $(QUIET_LINK)$(HOSTCC) $(CFLAGS) $(LDFLAGS) $(BOOTSTRAP_OBJS) $(LIBS_BOOTSTRAP) -o $@ $(OUTPUT)bpftool: $(OBJS) $(LIBBPF) - $(QUIET_LINK)$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(OBJS) $(LIBS) + $(QUIET_LINK)$(CC) $(CFLAGS) $(LDFLAGS) $(OBJS) $(LIBS) -o $@ $(BOOTSTRAP_OUTPUT)%.o: %.c $(LIBBPF_INTERNAL_HDRS) | $(BOOTSTRAP_OUTPUT) - $(QUIET_CC)$(HOSTCC) $(CFLAGS) -c -MMD -o $@ $< + $(QUIET_CC)$(HOSTCC) $(CFLAGS) -c -MMD $< -o $@ $(OUTPUT)%.o: %.c - $(QUIET_CC)$(CC) $(CFLAGS) -c -MMD -o $@ $< + $(QUIET_CC)$(CC) $(CFLAGS) -c -MMD $< -o $@ feature-detect-clean: $(call QUIET_CLEAN, feature-detect) From de29e6bbb9ee674d639cd42fe565f28757208614 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:17 -0800 Subject: [PATCH 0116/2048] selftests/bpf: Minor cleanups and normalization of Makefile Few clean ups and single-line simplifications. Also split CLEAN command into multiple $(RM) invocations as it gets dangerously close to too long argument list. Make sure that -o is used always as the last argument for saner verbose make output. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-3-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 32 ++++++++++++++-------------- 1 file changed, 16 insertions(+), 16 deletions(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 0468ea57650d..0470802c907c 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -45,10 +45,8 @@ ifneq ($(BPF_GCC),) TEST_GEN_PROGS += test_progs-bpf_gcc endif -TEST_GEN_FILES = test_lwt_ip_encap.o \ - test_tc_edt.o -TEST_FILES = xsk_prereqs.sh \ - $(wildcard progs/btf_dump_test_case_*.c) +TEST_GEN_FILES = test_lwt_ip_encap.o test_tc_edt.o +TEST_FILES = xsk_prereqs.sh $(wildcard progs/btf_dump_test_case_*.c) # Order correspond to 'make run_tests' order TEST_PROGS := test_kmod.sh \ @@ -107,7 +105,10 @@ endif OVERRIDE_TARGETS := 1 override define CLEAN $(call msg,CLEAN) - $(Q)$(RM) -r $(TEST_GEN_PROGS) $(TEST_GEN_PROGS_EXTENDED) $(TEST_GEN_FILES) $(EXTRA_CLEAN) + $(Q)$(RM) -r $(TEST_GEN_PROGS) + $(Q)$(RM) -r $(TEST_GEN_PROGS_EXTENDED) + $(Q)$(RM) -r $(TEST_GEN_FILES) + $(Q)$(RM) -r $(EXTRA_CLEAN) $(Q)$(MAKE) -C bpf_testmod clean $(Q)$(MAKE) docs-clean endef @@ -169,7 +170,7 @@ $(OUTPUT)/%:%.c $(OUTPUT)/urandom_read: urandom_read.c $(call msg,BINARY,,$@) - $(Q)$(CC) $(LDFLAGS) -o $@ $< $(LDLIBS) -Wl,--build-id=sha1 + $(Q)$(CC) $(LDFLAGS) $< $(LDLIBS) -Wl,--build-id=sha1 -o $@ $(OUTPUT)/bpf_testmod.ko: $(VMLINUX_BTF) $(wildcard bpf_testmod/Makefile bpf_testmod/*.[ch]) $(call msg,MOD,,$@) @@ -232,16 +233,16 @@ docs-clean: prefix= OUTPUT=$(OUTPUT)/ DESTDIR=$(OUTPUT)/ $@ $(BPFOBJ): $(wildcard $(BPFDIR)/*.[ch] $(BPFDIR)/Makefile) \ - ../../../include/uapi/linux/bpf.h \ + $(APIDIR)/linux/bpf.h \ | $(BUILD_DIR)/libbpf $(Q)$(MAKE) $(submake_extras) -C $(BPFDIR) OUTPUT=$(BUILD_DIR)/libbpf/ \ EXTRA_CFLAGS='-g -O0' \ DESTDIR=$(SCRATCH_DIR) prefix= all install_headers ifneq ($(BPFOBJ),$(HOST_BPFOBJ)) -$(HOST_BPFOBJ): $(wildcard $(BPFDIR)/*.[ch] $(BPFDIR)/Makefile) \ - ../../../include/uapi/linux/bpf.h \ - | $(HOST_BUILD_DIR)/libbpf +$(HOST_BPFOBJ): $(wildcard $(BPFDIR)/*.[ch] $(BPFDIR)/Makefile) \ + $(APIDIR)/linux/bpf.h \ + | $(HOST_BUILD_DIR)/libbpf $(Q)$(MAKE) $(submake_extras) -C $(BPFDIR) \ EXTRA_CFLAGS='-g -O0' \ OUTPUT=$(HOST_BUILD_DIR)/libbpf/ CC=$(HOSTCC) LD=$(HOSTLD) \ @@ -305,12 +306,12 @@ $(OUTPUT)/flow_dissector_load.o: flow_dissector_load.h # $3 - CFLAGS define CLANG_BPF_BUILD_RULE $(call msg,CLNG-BPF,$(TRUNNER_BINARY),$2) - $(Q)$(CLANG) $3 -O2 -target bpf -c $1 -o $2 -mcpu=v3 + $(Q)$(CLANG) $3 -O2 -target bpf -c $1 -mcpu=v3 -o $2 endef # Similar to CLANG_BPF_BUILD_RULE, but with disabled alu32 define CLANG_NOALU32_BPF_BUILD_RULE $(call msg,CLNG-BPF,$(TRUNNER_BINARY),$2) - $(Q)$(CLANG) $3 -O2 -target bpf -c $1 -o $2 -mcpu=v2 + $(Q)$(CLANG) $3 -O2 -target bpf -c $1 -mcpu=v2 -o $2 endef # Build BPF object using GCC define GCC_BPF_BUILD_RULE @@ -472,13 +473,12 @@ TRUNNER_TESTS_DIR := prog_tests TRUNNER_BPF_PROGS_DIR := progs TRUNNER_EXTRA_SOURCES := test_progs.c cgroup_helpers.c trace_helpers.c \ network_helpers.c testing_helpers.c \ - btf_helpers.c flow_dissector_load.h + btf_helpers.c flow_dissector_load.h TRUNNER_EXTRA_FILES := $(OUTPUT)/urandom_read $(OUTPUT)/bpf_testmod.ko \ ima_setup.sh \ $(wildcard progs/btf_dump_test_case_*.c) TRUNNER_BPF_BUILD_RULE := CLANG_BPF_BUILD_RULE -TRUNNER_BPF_CFLAGS := $(BPF_CFLAGS) $(CLANG_CFLAGS) -TRUNNER_BPF_CFLAGS += -DENABLE_ATOMICS_TESTS +TRUNNER_BPF_CFLAGS := $(BPF_CFLAGS) $(CLANG_CFLAGS) -DENABLE_ATOMICS_TESTS $(eval $(call DEFINE_TEST_RUNNER,test_progs)) # Define test_progs-no_alu32 test runner. @@ -540,7 +540,7 @@ $(OUTPUT)/bench: $(OUTPUT)/bench.o $(OUTPUT)/testing_helpers.o \ $(OUTPUT)/bench_ringbufs.o \ $(OUTPUT)/bench_bloom_filter_map.o $(call msg,BINARY,,$@) - $(Q)$(CC) $(LDFLAGS) -o $@ $(filter %.a %.o,$^) $(LDLIBS) + $(Q)$(CC) $(LDFLAGS) $(filter %.a %.o,$^) $(LDLIBS) -o $@ EXTRA_CLEAN := $(TEST_CUSTOM_PROGS) $(SCRATCH_DIR) $(HOST_SCRATCH_DIR) \ prog_tests/tests.h map_tests/tests.h verifier/tests.h \ From 957d350a8b94133d114a9b1ac3e79f1f77100681 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:18 -0800 Subject: [PATCH 0117/2048] libbpf: Turn btf_dedup_opts into OPTS-based struct btf__dedup() and struct btf_dedup_opts were added before we figured out OPTS mechanism. As such, btf_dedup_opts is non-extensible without breaking an ABI and potentially crashing user application. Unfortunately, btf__dedup() and btf_dedup_opts are short and succinct names that would be great to preserve and use going forward. So we use ___libbpf_override() macro approach, used previously for bpf_prog_load() API, to define a new btf__dedup() variant that accepts only struct btf * and struct btf_dedup_opts * arguments, and rename the old btf__dedup() implementation into btf__dedup_deprecated(). This keeps both source and binary compatibility with old and new applications. The biggest problem was struct btf_dedup_opts, which wasn't OPTS-based, and as such doesn't have `size_t sz;` as a first field. But btf__dedup() is a pretty rarely used API and I believe that the only currently known users (besides selftests) are libbpf's own bpf_linker and pahole. Neither use case actually uses options and just passes NULL. So instead of doing extra hacks, just rewrite struct btf_dedup_opts into OPTS-based one, move btf_ext argument into those opts (only bpf_linker needs to dedup btf_ext, so it's not a typical thing to specify), and drop never used `dont_resolve_fwds` option (it was never used anywhere, AFAIK, it makes BTF dedup much less useful and efficient). Just in case, for old implementation, btf__dedup_deprecated(), detect non-NULL options and error out with helpful message, to help users migrate, if there are any user playing with btf__dedup(). The last remaining piece is dedup_table_size, which is another anachronism from very early days of BTF dedup. Since then it has been reduced to the only valid value, 1, to request forced hash collisions. This is only used during testing. So instead introduce a bool flag to force collisions explicitly. This patch also adapts selftests to new btf__dedup() and btf_dedup_opts use to avoid selftests breakage. [0] Closes: https://github.com/libbpf/libbpf/issues/281 Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-4-andrii@kernel.org --- tools/lib/bpf/btf.c | 46 +++++++++++-------- tools/lib/bpf/btf.h | 20 ++++++-- tools/lib/bpf/libbpf.map | 2 + tools/lib/bpf/linker.c | 4 +- tools/testing/selftests/bpf/prog_tests/btf.c | 46 +++---------------- .../bpf/prog_tests/btf_dedup_split.c | 6 +-- 6 files changed, 58 insertions(+), 66 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index 7e4c5586bd87..fcec27622e7a 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -2846,8 +2846,7 @@ __u32 btf_ext__line_info_rec_size(const struct btf_ext *btf_ext) struct btf_dedup; -static struct btf_dedup *btf_dedup_new(struct btf *btf, struct btf_ext *btf_ext, - const struct btf_dedup_opts *opts); +static struct btf_dedup *btf_dedup_new(struct btf *btf, const struct btf_dedup_opts *opts); static void btf_dedup_free(struct btf_dedup *d); static int btf_dedup_prep(struct btf_dedup *d); static int btf_dedup_strings(struct btf_dedup *d); @@ -2994,12 +2993,17 @@ static int btf_dedup_remap_types(struct btf_dedup *d); * deduplicating structs/unions is described in greater details in comments for * `btf_dedup_is_equiv` function. */ -int btf__dedup(struct btf *btf, struct btf_ext *btf_ext, - const struct btf_dedup_opts *opts) + +DEFAULT_VERSION(btf__dedup_v0_6_0, btf__dedup, LIBBPF_0.6.0) +int btf__dedup_v0_6_0(struct btf *btf, const struct btf_dedup_opts *opts) { - struct btf_dedup *d = btf_dedup_new(btf, btf_ext, opts); + struct btf_dedup *d; int err; + if (!OPTS_VALID(opts, btf_dedup_opts)) + return libbpf_err(-EINVAL); + + d = btf_dedup_new(btf, opts); if (IS_ERR(d)) { pr_debug("btf_dedup_new failed: %ld", PTR_ERR(d)); return libbpf_err(-EINVAL); @@ -3051,6 +3055,19 @@ done: return libbpf_err(err); } +COMPAT_VERSION(bpf__dedup_deprecated, btf__dedup, LIBBPF_0.0.2) +int btf__dedup_deprecated(struct btf *btf, struct btf_ext *btf_ext, const void *unused_opts) +{ + LIBBPF_OPTS(btf_dedup_opts, opts, .btf_ext = btf_ext); + + if (unused_opts) { + pr_warn("please use new version of btf__dedup() that supports options\n"); + return libbpf_err(-ENOTSUP); + } + + return btf__dedup(btf, &opts); +} + #define BTF_UNPROCESSED_ID ((__u32)-1) #define BTF_IN_PROGRESS_ID ((__u32)-2) @@ -3163,8 +3180,7 @@ static bool btf_dedup_equal_fn(const void *k1, const void *k2, void *ctx) return k1 == k2; } -static struct btf_dedup *btf_dedup_new(struct btf *btf, struct btf_ext *btf_ext, - const struct btf_dedup_opts *opts) +static struct btf_dedup *btf_dedup_new(struct btf *btf, const struct btf_dedup_opts *opts) { struct btf_dedup *d = calloc(1, sizeof(struct btf_dedup)); hashmap_hash_fn hash_fn = btf_dedup_identity_hash_fn; @@ -3173,13 +3189,11 @@ static struct btf_dedup *btf_dedup_new(struct btf *btf, struct btf_ext *btf_ext, if (!d) return ERR_PTR(-ENOMEM); - d->opts.dont_resolve_fwds = opts && opts->dont_resolve_fwds; - /* dedup_table_size is now used only to force collisions in tests */ - if (opts && opts->dedup_table_size == 1) + if (OPTS_GET(opts, force_collisions, false)) hash_fn = btf_dedup_collision_hash_fn; d->btf = btf; - d->btf_ext = btf_ext; + d->btf_ext = OPTS_GET(opts, btf_ext, NULL); d->dedup_table = hashmap__new(hash_fn, btf_dedup_equal_fn, NULL); if (IS_ERR(d->dedup_table)) { @@ -3708,8 +3722,6 @@ static int btf_dedup_prim_type(struct btf_dedup *d, __u32 type_id) new_id = cand_id; break; } - if (d->opts.dont_resolve_fwds) - continue; if (btf_compat_enum(t, cand)) { if (btf_is_enum_fwd(t)) { /* resolve fwd to full enum */ @@ -3952,8 +3964,7 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id, return 0; /* FWD <--> STRUCT/UNION equivalence check, if enabled */ - if (!d->opts.dont_resolve_fwds - && (cand_kind == BTF_KIND_FWD || canon_kind == BTF_KIND_FWD) + if ((cand_kind == BTF_KIND_FWD || canon_kind == BTF_KIND_FWD) && cand_kind != canon_kind) { __u16 real_kind; __u16 fwd_kind; @@ -3979,10 +3990,7 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id, return btf_equal_int_tag(cand_type, canon_type); case BTF_KIND_ENUM: - if (d->opts.dont_resolve_fwds) - return btf_equal_enum(cand_type, canon_type); - else - return btf_compat_enum(cand_type, canon_type); + return btf_compat_enum(cand_type, canon_type); case BTF_KIND_FWD: case BTF_KIND_FLOAT: diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h index bc005ba3ceec..6aae4f62ee0b 100644 --- a/tools/lib/bpf/btf.h +++ b/tools/lib/bpf/btf.h @@ -245,12 +245,24 @@ LIBBPF_API int btf__add_decl_tag(struct btf *btf, const char *value, int ref_typ int component_idx); struct btf_dedup_opts { - unsigned int dedup_table_size; - bool dont_resolve_fwds; + size_t sz; + /* optional .BTF.ext info to dedup along the main BTF info */ + struct btf_ext *btf_ext; + /* force hash collisions (used for testing) */ + bool force_collisions; + size_t :0; }; +#define btf_dedup_opts__last_field force_collisions -LIBBPF_API int btf__dedup(struct btf *btf, struct btf_ext *btf_ext, - const struct btf_dedup_opts *opts); +LIBBPF_API int btf__dedup(struct btf *btf, const struct btf_dedup_opts *opts); + +LIBBPF_API int btf__dedup_v0_6_0(struct btf *btf, const struct btf_dedup_opts *opts); + +LIBBPF_DEPRECATED_SINCE(0, 7, "use btf__dedup() instead") +LIBBPF_API int btf__dedup_deprecated(struct btf *btf, struct btf_ext *btf_ext, const void *opts); +#define btf__dedup(...) ___libbpf_overload(___btf_dedup, __VA_ARGS__) +#define ___btf_dedup3(btf, btf_ext, opts) btf__dedup_deprecated(btf, btf_ext, opts) +#define ___btf_dedup2(btf, opts) btf__dedup(btf, opts) struct btf_dump; diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index 256d5d4be951..b8c8d14621d2 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -403,6 +403,8 @@ LIBBPF_0.6.0 { bpf_program__set_extra_flags; btf__add_btf; btf__add_decl_tag; + btf__dedup; + btf__dedup_deprecated; btf__raw_data; btf__type_cnt; } LIBBPF_0.5.0; diff --git a/tools/lib/bpf/linker.c b/tools/lib/bpf/linker.c index f677dccdeae4..594b206fa674 100644 --- a/tools/lib/bpf/linker.c +++ b/tools/lib/bpf/linker.c @@ -2650,6 +2650,7 @@ static int emit_elf_data_sec(struct bpf_linker *linker, const char *sec_name, static int finalize_btf(struct bpf_linker *linker) { + LIBBPF_OPTS(btf_dedup_opts, opts); struct btf *btf = linker->btf; const void *raw_data; int i, j, id, err; @@ -2686,7 +2687,8 @@ static int finalize_btf(struct bpf_linker *linker) return err; } - err = btf__dedup(linker->btf, linker->btf_ext, NULL); + opts.btf_ext = linker->btf_ext; + err = btf__dedup(linker->btf, &opts); if (err) { pr_warn("BTF dedup failed: %d\n", err); return err; diff --git a/tools/testing/selftests/bpf/prog_tests/btf.c b/tools/testing/selftests/bpf/prog_tests/btf.c index ebd1aa4d09d6..1e8b36d74df2 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf.c +++ b/tools/testing/selftests/bpf/prog_tests/btf.c @@ -6627,7 +6627,7 @@ struct btf_dedup_test { struct btf_dedup_opts opts; }; -const struct btf_dedup_test dedup_tests[] = { +static struct btf_dedup_test dedup_tests[] = { { .descr = "dedup: unused strings filtering", @@ -6647,9 +6647,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0int\0long"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: strings deduplication", @@ -6672,9 +6669,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0int\0long int"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: struct example #1", @@ -6755,9 +6749,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0a\0b\0c\0d\0int\0float\0next\0s"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: struct <-> fwd resolution w/ hash collision", @@ -6800,8 +6791,7 @@ const struct btf_dedup_test dedup_tests[] = { BTF_STR_SEC("\0s\0x"), }, .opts = { - .dont_resolve_fwds = false, - .dedup_table_size = 1, /* force hash collisions */ + .force_collisions = true, /* force hash collisions */ }, }, { @@ -6847,8 +6837,7 @@ const struct btf_dedup_test dedup_tests[] = { BTF_STR_SEC("\0s\0x"), }, .opts = { - .dont_resolve_fwds = false, - .dedup_table_size = 1, /* force hash collisions */ + .force_collisions = true, /* force hash collisions */ }, }, { @@ -6911,9 +6900,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0A\0B\0C\0D\0E\0F\0G\0H\0I\0J\0K\0L\0M\0N\0O\0P\0Q"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: no int/float duplicates", @@ -6965,9 +6951,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0int\0some other int\0float"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: enum fwd resolution", @@ -7009,9 +6992,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0e1\0e1_val\0e2\0e2_val"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: datasec and vars pass-through", @@ -7054,8 +7034,7 @@ const struct btf_dedup_test dedup_tests[] = { BTF_STR_SEC("\0.bss\0t"), }, .opts = { - .dont_resolve_fwds = false, - .dedup_table_size = 1 + .force_collisions = true }, }, { @@ -7099,9 +7078,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0t\0a1\0a2\0f\0tag"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: func/func_param tags", @@ -7152,9 +7128,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0a1\0a2\0f\0tag1\0tag2\0tag3"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: struct/struct_member tags", @@ -7200,9 +7173,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0t\0m1\0m2\0tag1\0tag2\0tag3"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, { .descr = "dedup: typedef tags", @@ -7233,9 +7203,6 @@ const struct btf_dedup_test dedup_tests[] = { }, BTF_STR_SEC("\0t\0tag1\0tag2\0tag3"), }, - .opts = { - .dont_resolve_fwds = false, - }, }, }; @@ -7293,7 +7260,7 @@ static void dump_btf_strings(const char *strs, __u32 len) static void do_test_dedup(unsigned int test_num) { - const struct btf_dedup_test *test = &dedup_tests[test_num - 1]; + struct btf_dedup_test *test = &dedup_tests[test_num - 1]; __u32 test_nr_types, expect_nr_types, test_btf_size, expect_btf_size; const struct btf_header *test_hdr, *expect_hdr; struct btf *test_btf = NULL, *expect_btf = NULL; @@ -7337,7 +7304,8 @@ static void do_test_dedup(unsigned int test_num) goto done; } - err = btf__dedup(test_btf, NULL, &test->opts); + test->opts.sz = sizeof(test->opts); + err = btf__dedup(test_btf, &test->opts); if (CHECK(err, "btf_dedup failed errno:%d", err)) { err = -1; goto done; diff --git a/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c b/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c index 64554fd33547..9d3b8d7a1537 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c @@ -92,7 +92,7 @@ struct s2 {\n\ int *f3;\n\ };\n\n", "c_dump"); - err = btf__dedup(btf2, NULL, NULL); + err = btf__dedup(btf2, NULL); if (!ASSERT_OK(err, "btf_dedup")) goto cleanup; @@ -186,7 +186,7 @@ static void test_split_fwd_resolve() { "\t'f1' type_id=7 bits_offset=0\n" "\t'f2' type_id=9 bits_offset=64"); - err = btf__dedup(btf2, NULL, NULL); + err = btf__dedup(btf2, NULL); if (!ASSERT_OK(err, "btf_dedup")) goto cleanup; @@ -283,7 +283,7 @@ static void test_split_struct_duped() { "[13] STRUCT 's3' size=8 vlen=1\n" "\t'f1' type_id=12 bits_offset=0"); - err = btf__dedup(btf2, NULL, NULL); + err = btf__dedup(btf2, NULL); if (!ASSERT_OK(err, "btf_dedup")) goto cleanup; From 6084f5dc928f2ada4331ba9eda65542e94d86bc6 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:19 -0800 Subject: [PATCH 0118/2048] libbpf: Ensure btf_dump__new() and btf_dump_opts are future-proof Change btf_dump__new() and corresponding struct btf_dump_ops structure to be extensible by using OPTS "framework" ([0]). Given we don't change the names, we use a similar approach as with bpf_prog_load(), but this time we ended up with two APIs with the same name and same number of arguments, so overloading based on number of arguments with ___libbpf_override() doesn't work. Instead, use "overloading" based on types. In this particular case, print callback has to be specified, so we detect which argument is a callback. If it's 4th (last) argument, old implementation of API is used by user code. If not, it must be 2nd, and thus new implementation is selected. The rest is handled by the same symbol versioning approach. btf_ext argument is dropped as it was never used and isn't necessary either. If in the future we'll need btf_ext, that will be added into OPTS-based struct btf_dump_opts. struct btf_dump_opts is reused for both old API and new APIs. ctx field is marked deprecated in v0.7+ and it's put at the same memory location as OPTS's sz field. Any user of new-style btf_dump__new() will have to set sz field and doesn't/shouldn't use ctx, as ctx is now passed along the callback as mandatory input argument, following the other APIs in libbpf that accept callbacks consistently. Again, this is quite ugly in implementation, but is done in the name of backwards compatibility and uniform and extensible future APIs (at the same time, sigh). And it will be gone in libbpf 1.0. [0] Closes: https://github.com/libbpf/libbpf/issues/283 Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-5-andrii@kernel.org --- tools/lib/bpf/btf.h | 51 ++++++++++++++++++++++++++++++++++++---- tools/lib/bpf/btf_dump.c | 31 +++++++++++++++++------- tools/lib/bpf/libbpf.map | 2 ++ 3 files changed, 71 insertions(+), 13 deletions(-) diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h index 6aae4f62ee0b..45310c65e865 100644 --- a/tools/lib/bpf/btf.h +++ b/tools/lib/bpf/btf.h @@ -267,15 +267,58 @@ LIBBPF_API int btf__dedup_deprecated(struct btf *btf, struct btf_ext *btf_ext, c struct btf_dump; struct btf_dump_opts { - void *ctx; + union { + size_t sz; + void *ctx; /* DEPRECATED: will be gone in v1.0 */ + }; }; typedef void (*btf_dump_printf_fn_t)(void *ctx, const char *fmt, va_list args); LIBBPF_API struct btf_dump *btf_dump__new(const struct btf *btf, - const struct btf_ext *btf_ext, - const struct btf_dump_opts *opts, - btf_dump_printf_fn_t printf_fn); + btf_dump_printf_fn_t printf_fn, + void *ctx, + const struct btf_dump_opts *opts); + +LIBBPF_API struct btf_dump *btf_dump__new_v0_6_0(const struct btf *btf, + btf_dump_printf_fn_t printf_fn, + void *ctx, + const struct btf_dump_opts *opts); + +LIBBPF_API struct btf_dump *btf_dump__new_deprecated(const struct btf *btf, + const struct btf_ext *btf_ext, + const struct btf_dump_opts *opts, + btf_dump_printf_fn_t printf_fn); + +/* Choose either btf_dump__new() or btf_dump__new_deprecated() based on the + * type of 4th argument. If it's btf_dump's print callback, use deprecated + * API; otherwise, choose the new btf_dump__new(). ___libbpf_override() + * doesn't work here because both variants have 4 input arguments. + * + * (void *) casts are necessary to avoid compilation warnings about type + * mismatches, because even though __builtin_choose_expr() only ever evaluates + * one side the other side still has to satisfy type constraints (this is + * compiler implementation limitation which might be lifted eventually, + * according to the documentation). So passing struct btf_ext in place of + * btf_dump_printf_fn_t would be generating compilation warning. Casting to + * void * avoids this issue. + * + * Also, two type compatibility checks for a function and function pointer are + * required because passing function reference into btf_dump__new() as + * btf_dump__new(..., my_callback, ...) and as btf_dump__new(..., + * &my_callback, ...) (not explicit ampersand in the latter case) actually + * differs as far as __builtin_types_compatible_p() is concerned. Thus two + * checks are combined to detect callback argument. + * + * The rest works just like in case of ___libbpf_override() usage with symbol + * versioning. + */ +#define btf_dump__new(a1, a2, a3, a4) __builtin_choose_expr( \ + __builtin_types_compatible_p(typeof(a4), btf_dump_printf_fn_t) || \ + __builtin_types_compatible_p(typeof(a4), void(void *, const char *, va_list)), \ + btf_dump__new_deprecated((void *)a1, (void *)a2, (void *)a3, (void *)a4), \ + btf_dump__new((void *)a1, (void *)a2, (void *)a3, (void *)a4)) + LIBBPF_API void btf_dump__free(struct btf_dump *d); LIBBPF_API int btf_dump__dump_type(struct btf_dump *d, __u32 id); diff --git a/tools/lib/bpf/btf_dump.c b/tools/lib/bpf/btf_dump.c index 17db62b5002e..b8cd7e4f557a 100644 --- a/tools/lib/bpf/btf_dump.c +++ b/tools/lib/bpf/btf_dump.c @@ -77,9 +77,8 @@ struct btf_dump_data { struct btf_dump { const struct btf *btf; - const struct btf_ext *btf_ext; btf_dump_printf_fn_t printf_fn; - struct btf_dump_opts opts; + void *cb_ctx; int ptr_sz; bool strip_mods; bool skip_anon_defs; @@ -138,29 +137,32 @@ static void btf_dump_printf(const struct btf_dump *d, const char *fmt, ...) va_list args; va_start(args, fmt); - d->printf_fn(d->opts.ctx, fmt, args); + d->printf_fn(d->cb_ctx, fmt, args); va_end(args); } static int btf_dump_mark_referenced(struct btf_dump *d); static int btf_dump_resize(struct btf_dump *d); -struct btf_dump *btf_dump__new(const struct btf *btf, - const struct btf_ext *btf_ext, - const struct btf_dump_opts *opts, - btf_dump_printf_fn_t printf_fn) +DEFAULT_VERSION(btf_dump__new_v0_6_0, btf_dump__new, LIBBPF_0.6.0) +struct btf_dump *btf_dump__new_v0_6_0(const struct btf *btf, + btf_dump_printf_fn_t printf_fn, + void *ctx, + const struct btf_dump_opts *opts) { struct btf_dump *d; int err; + if (!printf_fn) + return libbpf_err_ptr(-EINVAL); + d = calloc(1, sizeof(struct btf_dump)); if (!d) return libbpf_err_ptr(-ENOMEM); d->btf = btf; - d->btf_ext = btf_ext; d->printf_fn = printf_fn; - d->opts.ctx = opts ? opts->ctx : NULL; + d->cb_ctx = ctx; d->ptr_sz = btf__pointer_size(btf) ? : sizeof(void *); d->type_names = hashmap__new(str_hash_fn, str_equal_fn, NULL); @@ -186,6 +188,17 @@ err: return libbpf_err_ptr(err); } +COMPAT_VERSION(btf_dump__new_deprecated, btf_dump__new, LIBBPF_0.0.4) +struct btf_dump *btf_dump__new_deprecated(const struct btf *btf, + const struct btf_ext *btf_ext, + const struct btf_dump_opts *opts, + btf_dump_printf_fn_t printf_fn) +{ + if (!printf_fn) + return libbpf_err_ptr(-EINVAL); + return btf_dump__new_v0_6_0(btf, printf_fn, opts ? opts->ctx : NULL, opts); +} + static int btf_dump_resize(struct btf_dump *d) { int err, last_id = btf__type_cnt(d->btf) - 1; diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index b8c8d14621d2..07762d6723ef 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -407,4 +407,6 @@ LIBBPF_0.6.0 { btf__dedup_deprecated; btf__raw_data; btf__type_cnt; + btf_dump__new; + btf_dump__new_deprecated; } LIBBPF_0.5.0; From 4178893465774f91dcd49465ae6f4e3cc036b7b2 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:20 -0800 Subject: [PATCH 0119/2048] libbpf: Make perf_buffer__new() use OPTS-based interface Add new variants of perf_buffer__new() and perf_buffer__new_raw() that use OPTS-based options for future extensibility ([0]). Given all the currently used API names are best fits, re-use them and use ___libbpf_override() approach and symbol versioning to preserve ABI and source code compatibility. struct perf_buffer_opts and struct perf_buffer_raw_opts are kept as well, but they are restructured such that they are OPTS-based when used with new APIs. For struct perf_buffer_raw_opts we keep few fields intact, so we have to also preserve the memory location of them both when used as OPTS and for legacy API variants. This is achieved with anonymous padding for OPTS "incarnation" of the struct. These pads can be eventually used for new options. [0] Closes: https://github.com/libbpf/libbpf/issues/311 Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-6-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 70 +++++++++++++++++++++++++------- tools/lib/bpf/libbpf.h | 86 ++++++++++++++++++++++++++++++++++------ tools/lib/bpf/libbpf.map | 4 ++ 3 files changed, 132 insertions(+), 28 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index a823b5ed705b..7db14859b27c 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -10578,11 +10578,18 @@ error: static struct perf_buffer *__perf_buffer__new(int map_fd, size_t page_cnt, struct perf_buffer_params *p); -struct perf_buffer *perf_buffer__new(int map_fd, size_t page_cnt, - const struct perf_buffer_opts *opts) +DEFAULT_VERSION(perf_buffer__new_v0_6_0, perf_buffer__new, LIBBPF_0.6.0) +struct perf_buffer *perf_buffer__new_v0_6_0(int map_fd, size_t page_cnt, + perf_buffer_sample_fn sample_cb, + perf_buffer_lost_fn lost_cb, + void *ctx, + const struct perf_buffer_opts *opts) { struct perf_buffer_params p = {}; - struct perf_event_attr attr = { 0, }; + struct perf_event_attr attr = {}; + + if (!OPTS_VALID(opts, perf_buffer_opts)) + return libbpf_err_ptr(-EINVAL); attr.config = PERF_COUNT_SW_BPF_OUTPUT; attr.type = PERF_TYPE_SOFTWARE; @@ -10591,29 +10598,62 @@ struct perf_buffer *perf_buffer__new(int map_fd, size_t page_cnt, attr.wakeup_events = 1; p.attr = &attr; - p.sample_cb = opts ? opts->sample_cb : NULL; - p.lost_cb = opts ? opts->lost_cb : NULL; - p.ctx = opts ? opts->ctx : NULL; + p.sample_cb = sample_cb; + p.lost_cb = lost_cb; + p.ctx = ctx; return libbpf_ptr(__perf_buffer__new(map_fd, page_cnt, &p)); } -struct perf_buffer * -perf_buffer__new_raw(int map_fd, size_t page_cnt, - const struct perf_buffer_raw_opts *opts) +COMPAT_VERSION(perf_buffer__new_deprecated, perf_buffer__new, LIBBPF_0.0.4) +struct perf_buffer *perf_buffer__new_deprecated(int map_fd, size_t page_cnt, + const struct perf_buffer_opts *opts) +{ + return perf_buffer__new_v0_6_0(map_fd, page_cnt, + opts ? opts->sample_cb : NULL, + opts ? opts->lost_cb : NULL, + opts ? opts->ctx : NULL, + NULL); +} + +DEFAULT_VERSION(perf_buffer__new_raw_v0_6_0, perf_buffer__new_raw, LIBBPF_0.6.0) +struct perf_buffer *perf_buffer__new_raw_v0_6_0(int map_fd, size_t page_cnt, + struct perf_event_attr *attr, + perf_buffer_event_fn event_cb, void *ctx, + const struct perf_buffer_raw_opts *opts) { struct perf_buffer_params p = {}; - p.attr = opts->attr; - p.event_cb = opts->event_cb; - p.ctx = opts->ctx; - p.cpu_cnt = opts->cpu_cnt; - p.cpus = opts->cpus; - p.map_keys = opts->map_keys; + if (page_cnt == 0 || !attr) + return libbpf_err_ptr(-EINVAL); + + if (!OPTS_VALID(opts, perf_buffer_raw_opts)) + return libbpf_err_ptr(-EINVAL); + + p.attr = attr; + p.event_cb = event_cb; + p.ctx = ctx; + p.cpu_cnt = OPTS_GET(opts, cpu_cnt, 0); + p.cpus = OPTS_GET(opts, cpus, NULL); + p.map_keys = OPTS_GET(opts, map_keys, NULL); return libbpf_ptr(__perf_buffer__new(map_fd, page_cnt, &p)); } +COMPAT_VERSION(perf_buffer__new_raw_deprecated, perf_buffer__new_raw, LIBBPF_0.0.4) +struct perf_buffer *perf_buffer__new_raw_deprecated(int map_fd, size_t page_cnt, + const struct perf_buffer_raw_opts *opts) +{ + LIBBPF_OPTS(perf_buffer_raw_opts, inner_opts, + .cpu_cnt = opts->cpu_cnt, + .cpus = opts->cpus, + .map_keys = opts->map_keys, + ); + + return perf_buffer__new_raw_v0_6_0(map_fd, page_cnt, opts->attr, + opts->event_cb, opts->ctx, &inner_opts); +} + static struct perf_buffer *__perf_buffer__new(int map_fd, size_t page_cnt, struct perf_buffer_params *p) { diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index f69512ae8505..4ec69f224342 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -778,18 +778,52 @@ typedef void (*perf_buffer_lost_fn)(void *ctx, int cpu, __u64 cnt); /* common use perf buffer options */ struct perf_buffer_opts { - /* if specified, sample_cb is called for each sample */ - perf_buffer_sample_fn sample_cb; - /* if specified, lost_cb is called for each batch of lost samples */ - perf_buffer_lost_fn lost_cb; - /* ctx is provided to sample_cb and lost_cb */ - void *ctx; + union { + size_t sz; + struct { /* DEPRECATED: will be removed in v1.0 */ + /* if specified, sample_cb is called for each sample */ + perf_buffer_sample_fn sample_cb; + /* if specified, lost_cb is called for each batch of lost samples */ + perf_buffer_lost_fn lost_cb; + /* ctx is provided to sample_cb and lost_cb */ + void *ctx; + }; + }; }; +#define perf_buffer_opts__last_field sz +/** + * @brief **perf_buffer__new()** creates BPF perfbuf manager for a specified + * BPF_PERF_EVENT_ARRAY map + * @param map_fd FD of BPF_PERF_EVENT_ARRAY BPF map that will be used by BPF + * code to send data over to user-space + * @param page_cnt number of memory pages allocated for each per-CPU buffer + * @param sample_cb function called on each received data record + * @param lost_cb function called when record loss has occurred + * @param ctx user-provided extra context passed into *sample_cb* and *lost_cb* + * @return a new instance of struct perf_buffer on success, NULL on error with + * *errno* containing an error code + */ LIBBPF_API struct perf_buffer * perf_buffer__new(int map_fd, size_t page_cnt, + perf_buffer_sample_fn sample_cb, perf_buffer_lost_fn lost_cb, void *ctx, const struct perf_buffer_opts *opts); +LIBBPF_API struct perf_buffer * +perf_buffer__new_v0_6_0(int map_fd, size_t page_cnt, + perf_buffer_sample_fn sample_cb, perf_buffer_lost_fn lost_cb, void *ctx, + const struct perf_buffer_opts *opts); + +LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use new variant of perf_buffer__new() instead") +struct perf_buffer *perf_buffer__new_deprecated(int map_fd, size_t page_cnt, + const struct perf_buffer_opts *opts); + +#define perf_buffer__new(...) ___libbpf_overload(___perf_buffer_new, __VA_ARGS__) +#define ___perf_buffer_new6(map_fd, page_cnt, sample_cb, lost_cb, ctx, opts) \ + perf_buffer__new(map_fd, page_cnt, sample_cb, lost_cb, ctx, opts) +#define ___perf_buffer_new3(map_fd, page_cnt, opts) \ + perf_buffer__new_deprecated(map_fd, page_cnt, opts) + enum bpf_perf_event_ret { LIBBPF_PERF_EVENT_DONE = 0, LIBBPF_PERF_EVENT_ERROR = -1, @@ -803,12 +837,21 @@ typedef enum bpf_perf_event_ret /* raw perf buffer options, giving most power and control */ struct perf_buffer_raw_opts { - /* perf event attrs passed directly into perf_event_open() */ - struct perf_event_attr *attr; - /* raw event callback */ - perf_buffer_event_fn event_cb; - /* ctx is provided to event_cb */ - void *ctx; + union { + struct { + size_t sz; + long :0; + long :0; + }; + struct { /* DEPRECATED: will be removed in v1.0 */ + /* perf event attrs passed directly into perf_event_open() */ + struct perf_event_attr *attr; + /* raw event callback */ + perf_buffer_event_fn event_cb; + /* ctx is provided to event_cb */ + void *ctx; + }; + }; /* if cpu_cnt == 0, open all on all possible CPUs (up to the number of * max_entries of given PERF_EVENT_ARRAY map) */ @@ -818,11 +861,28 @@ struct perf_buffer_raw_opts { /* if cpu_cnt > 0, map_keys specify map keys to set per-CPU FDs for */ int *map_keys; }; +#define perf_buffer_raw_opts__last_field map_keys LIBBPF_API struct perf_buffer * -perf_buffer__new_raw(int map_fd, size_t page_cnt, +perf_buffer__new_raw(int map_fd, size_t page_cnt, struct perf_event_attr *attr, + perf_buffer_event_fn event_cb, void *ctx, const struct perf_buffer_raw_opts *opts); +LIBBPF_API struct perf_buffer * +perf_buffer__new_raw_v0_6_0(int map_fd, size_t page_cnt, struct perf_event_attr *attr, + perf_buffer_event_fn event_cb, void *ctx, + const struct perf_buffer_raw_opts *opts); + +LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use new variant of perf_buffer__new_raw() instead") +struct perf_buffer *perf_buffer__new_raw_deprecated(int map_fd, size_t page_cnt, + const struct perf_buffer_raw_opts *opts); + +#define perf_buffer__new_raw(...) ___libbpf_overload(___perf_buffer_new_raw, __VA_ARGS__) +#define ___perf_buffer_new_raw6(map_fd, page_cnt, attr, event_cb, ctx, opts) \ + perf_buffer__new_raw(map_fd, page_cnt, attr, event_cb, ctx, opts) +#define ___perf_buffer_new_raw3(map_fd, page_cnt, opts) \ + perf_buffer__new_raw_deprecated(map_fd, page_cnt, opts) + LIBBPF_API void perf_buffer__free(struct perf_buffer *pb); LIBBPF_API int perf_buffer__epoll_fd(const struct perf_buffer *pb); LIBBPF_API int perf_buffer__poll(struct perf_buffer *pb, int timeout_ms); diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index 07762d6723ef..9e91aa8da303 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -409,4 +409,8 @@ LIBBPF_0.6.0 { btf__type_cnt; btf_dump__new; btf_dump__new_deprecated; + perf_buffer__new; + perf_buffer__new_deprecated; + perf_buffer__new_raw; + perf_buffer__new_raw_deprecated; } LIBBPF_0.5.0; From 0b52a5f4b994c05070237271c7fac3265b640ffb Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:21 -0800 Subject: [PATCH 0120/2048] selftests/bpf: Migrate all deprecated perf_buffer uses Migrate all old-style perf_buffer__new() and perf_buffer__new_raw() calls to new v1.0+ variants. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-7-andrii@kernel.org --- tools/testing/selftests/bpf/benchs/bench_ringbufs.c | 8 ++------ tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c | 5 ++--- tools/testing/selftests/bpf/prog_tests/kfree_skb.c | 6 ++---- tools/testing/selftests/bpf/prog_tests/perf_buffer.c | 6 ++---- tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c | 7 ++----- tools/testing/selftests/bpf/test_tcpnotify_user.c | 4 +--- 6 files changed, 11 insertions(+), 25 deletions(-) diff --git a/tools/testing/selftests/bpf/benchs/bench_ringbufs.c b/tools/testing/selftests/bpf/benchs/bench_ringbufs.c index d167bffac679..52d4a2f91dbd 100644 --- a/tools/testing/selftests/bpf/benchs/bench_ringbufs.c +++ b/tools/testing/selftests/bpf/benchs/bench_ringbufs.c @@ -394,11 +394,6 @@ static void perfbuf_libbpf_setup() { struct perfbuf_libbpf_ctx *ctx = &perfbuf_libbpf_ctx; struct perf_event_attr attr; - struct perf_buffer_raw_opts pb_opts = { - .event_cb = perfbuf_process_sample_raw, - .ctx = (void *)(long)0, - .attr = &attr, - }; struct bpf_link *link; ctx->skel = perfbuf_setup_skeleton(); @@ -423,7 +418,8 @@ static void perfbuf_libbpf_setup() } ctx->perfbuf = perf_buffer__new_raw(bpf_map__fd(ctx->skel->maps.perfbuf), - args.perfbuf_sz, &pb_opts); + args.perfbuf_sz, &attr, + perfbuf_process_sample_raw, NULL, NULL); if (!ctx->perfbuf) { fprintf(stderr, "failed to create perfbuf\n"); exit(1); diff --git a/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c b/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c index 569fcc6ed660..4184c399d4c6 100644 --- a/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c +++ b/tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c @@ -85,7 +85,6 @@ void test_get_stack_raw_tp(void) const char *file_err = "./test_get_stack_rawtp_err.o"; const char *prog_name = "raw_tracepoint/sys_enter"; int i, err, prog_fd, exp_cnt = MAX_CNT_RAWTP; - struct perf_buffer_opts pb_opts = {}; struct perf_buffer *pb = NULL; struct bpf_link *link = NULL; struct timespec tv = {0, 10}; @@ -124,8 +123,8 @@ void test_get_stack_raw_tp(void) if (!ASSERT_OK_PTR(link, "attach_raw_tp")) goto close_prog; - pb_opts.sample_cb = get_stack_print_output; - pb = perf_buffer__new(bpf_map__fd(map), 8, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(map), 8, get_stack_print_output, + NULL, NULL, NULL); if (!ASSERT_OK_PTR(pb, "perf_buf__new")) goto close_prog; diff --git a/tools/testing/selftests/bpf/prog_tests/kfree_skb.c b/tools/testing/selftests/bpf/prog_tests/kfree_skb.c index 885413ed5c96..2a49f8fcde06 100644 --- a/tools/testing/selftests/bpf/prog_tests/kfree_skb.c +++ b/tools/testing/selftests/bpf/prog_tests/kfree_skb.c @@ -66,7 +66,6 @@ void serial_test_kfree_skb(void) struct bpf_map *perf_buf_map, *global_data; struct bpf_program *prog, *fentry, *fexit; struct bpf_object *obj, *obj2 = NULL; - struct perf_buffer_opts pb_opts = {}; struct perf_buffer *pb = NULL; int err, kfree_skb_fd; bool passed = false; @@ -112,9 +111,8 @@ void serial_test_kfree_skb(void) goto close_prog; /* set up perf buffer */ - pb_opts.sample_cb = on_sample; - pb_opts.ctx = &passed; - pb = perf_buffer__new(bpf_map__fd(perf_buf_map), 1, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(perf_buf_map), 1, + on_sample, NULL, &passed, NULL); if (!ASSERT_OK_PTR(pb, "perf_buf__new")) goto close_prog; diff --git a/tools/testing/selftests/bpf/prog_tests/perf_buffer.c b/tools/testing/selftests/bpf/prog_tests/perf_buffer.c index 4e32f3586a75..5fc2b3a0711e 100644 --- a/tools/testing/selftests/bpf/prog_tests/perf_buffer.c +++ b/tools/testing/selftests/bpf/prog_tests/perf_buffer.c @@ -47,7 +47,6 @@ void serial_test_perf_buffer(void) { int err, on_len, nr_on_cpus = 0, nr_cpus, i, j; int zero = 0, my_pid = getpid(); - struct perf_buffer_opts pb_opts = {}; struct test_perf_buffer *skel; cpu_set_t cpu_seen; struct perf_buffer *pb; @@ -82,9 +81,8 @@ void serial_test_perf_buffer(void) goto out_close; /* set up perf buffer */ - pb_opts.sample_cb = on_sample; - pb_opts.ctx = &cpu_seen; - pb = perf_buffer__new(bpf_map__fd(skel->maps.perf_buf_map), 1, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(skel->maps.perf_buf_map), 1, + on_sample, NULL, &cpu_seen, NULL); if (!ASSERT_OK_PTR(pb, "perf_buf__new")) goto out_close; diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c b/tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c index 3bd5904b4db5..f99386d1dc4c 100644 --- a/tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c +++ b/tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c @@ -49,7 +49,6 @@ void test_xdp_bpf2bpf(void) struct vip key4 = {.protocol = 6, .family = AF_INET}; struct bpf_program *prog; struct perf_buffer *pb = NULL; - struct perf_buffer_opts pb_opts = {}; /* Load XDP program to introspect */ pkt_skel = test_xdp__open_and_load(); @@ -86,10 +85,8 @@ void test_xdp_bpf2bpf(void) goto out; /* Set up perf buffer */ - pb_opts.sample_cb = on_sample; - pb_opts.ctx = &passed; - pb = perf_buffer__new(bpf_map__fd(ftrace_skel->maps.perf_buf_map), - 1, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(ftrace_skel->maps.perf_buf_map), 1, + on_sample, NULL, &passed, NULL); if (!ASSERT_OK_PTR(pb, "perf_buf__new")) goto out; diff --git a/tools/testing/selftests/bpf/test_tcpnotify_user.c b/tools/testing/selftests/bpf/test_tcpnotify_user.c index 63111cb082fe..4c5114765b23 100644 --- a/tools/testing/selftests/bpf/test_tcpnotify_user.c +++ b/tools/testing/selftests/bpf/test_tcpnotify_user.c @@ -72,7 +72,6 @@ int main(int argc, char **argv) { const char *file = "test_tcpnotify_kern.o"; struct bpf_map *perf_map, *global_map; - struct perf_buffer_opts pb_opts = {}; struct tcpnotify_globals g = {0}; struct perf_buffer *pb = NULL; const char *cg_path = "/foo"; @@ -117,8 +116,7 @@ int main(int argc, char **argv) return -1; } - pb_opts.sample_cb = dummyfn; - pb = perf_buffer__new(bpf_map__fd(perf_map), 8, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(perf_map), 8, dummyfn, NULL, NULL, NULL); if (!pb) goto err; From 60ba87bb6bafaaa6e8ef9a73834cf701194d1923 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:22 -0800 Subject: [PATCH 0121/2048] selftests/bpf: Update btf_dump__new() uses to v1.0+ variant Update to-be-deprecated forms of btf_dump__new(). Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-8-andrii@kernel.org --- tools/testing/selftests/bpf/btf_helpers.c | 4 +-- .../selftests/bpf/prog_tests/btf_dump.c | 33 ++++++++----------- .../selftests/bpf/prog_tests/btf_split.c | 4 +-- 3 files changed, 15 insertions(+), 26 deletions(-) diff --git a/tools/testing/selftests/bpf/btf_helpers.c b/tools/testing/selftests/bpf/btf_helpers.c index 3d1a748d09d8..acb59202486d 100644 --- a/tools/testing/selftests/bpf/btf_helpers.c +++ b/tools/testing/selftests/bpf/btf_helpers.c @@ -238,7 +238,6 @@ const char *btf_type_c_dump(const struct btf *btf) static char buf[16 * 1024]; FILE *buf_file; struct btf_dump *d = NULL; - struct btf_dump_opts opts = {}; int err, i; buf_file = fmemopen(buf, sizeof(buf) - 1, "w"); @@ -247,8 +246,7 @@ const char *btf_type_c_dump(const struct btf *btf) return NULL; } - opts.ctx = buf_file; - d = btf_dump__new(btf, NULL, &opts, btf_dump_printf); + d = btf_dump__new(btf, btf_dump_printf, buf_file, NULL); if (libbpf_get_error(d)) { fprintf(stderr, "Failed to create btf_dump instance: %ld\n", libbpf_get_error(d)); goto err_out; diff --git a/tools/testing/selftests/bpf/prog_tests/btf_dump.c b/tools/testing/selftests/bpf/prog_tests/btf_dump.c index a04961942dfa..d6272013a5a3 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_dump.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_dump.c @@ -13,25 +13,23 @@ static struct btf_dump_test_case { const char *name; const char *file; bool known_ptr_sz; - struct btf_dump_opts opts; } btf_dump_test_cases[] = { - {"btf_dump: syntax", "btf_dump_test_case_syntax", true, {}}, - {"btf_dump: ordering", "btf_dump_test_case_ordering", false, {}}, - {"btf_dump: padding", "btf_dump_test_case_padding", true, {}}, - {"btf_dump: packing", "btf_dump_test_case_packing", true, {}}, - {"btf_dump: bitfields", "btf_dump_test_case_bitfields", true, {}}, - {"btf_dump: multidim", "btf_dump_test_case_multidim", false, {}}, - {"btf_dump: namespacing", "btf_dump_test_case_namespacing", false, {}}, + {"btf_dump: syntax", "btf_dump_test_case_syntax", true}, + {"btf_dump: ordering", "btf_dump_test_case_ordering", false}, + {"btf_dump: padding", "btf_dump_test_case_padding", true}, + {"btf_dump: packing", "btf_dump_test_case_packing", true}, + {"btf_dump: bitfields", "btf_dump_test_case_bitfields", true}, + {"btf_dump: multidim", "btf_dump_test_case_multidim", false}, + {"btf_dump: namespacing", "btf_dump_test_case_namespacing", false}, }; -static int btf_dump_all_types(const struct btf *btf, - const struct btf_dump_opts *opts) +static int btf_dump_all_types(const struct btf *btf, void *ctx) { size_t type_cnt = btf__type_cnt(btf); struct btf_dump *d; int err = 0, id; - d = btf_dump__new(btf, NULL, opts, btf_dump_printf); + d = btf_dump__new(btf, btf_dump_printf, ctx, NULL); err = libbpf_get_error(d); if (err) return err; @@ -88,8 +86,7 @@ static int test_btf_dump_case(int n, struct btf_dump_test_case *t) goto done; } - t->opts.ctx = f; - err = btf_dump_all_types(btf, &t->opts); + err = btf_dump_all_types(btf, f); fclose(f); close(fd); if (CHECK(err, "btf_dump", "failure during C dumping: %d\n", err)) { @@ -137,7 +134,6 @@ static void test_btf_dump_incremental(void) { struct btf *btf = NULL; struct btf_dump *d = NULL; - struct btf_dump_opts opts; int id, err, i; dump_buf_file = open_memstream(&dump_buf, &dump_buf_sz); @@ -146,8 +142,7 @@ static void test_btf_dump_incremental(void) btf = btf__new_empty(); if (!ASSERT_OK_PTR(btf, "new_empty")) goto err_out; - opts.ctx = dump_buf_file; - d = btf_dump__new(btf, NULL, &opts, btf_dump_printf); + d = btf_dump__new(btf, btf_dump_printf, dump_buf_file, NULL); if (!ASSERT_OK(libbpf_get_error(d), "btf_dump__new")) goto err_out; @@ -815,7 +810,6 @@ static void test_btf_datasec(struct btf *btf, struct btf_dump *d, char *str, static void test_btf_dump_datasec_data(char *str) { struct btf *btf; - struct btf_dump_opts opts = { .ctx = str }; char license[4] = "GPL"; struct btf_dump *d; @@ -823,7 +817,7 @@ static void test_btf_dump_datasec_data(char *str) if (!ASSERT_OK_PTR(btf, "xdping_kern.o BTF not found")) return; - d = btf_dump__new(btf, NULL, &opts, btf_dump_snprintf); + d = btf_dump__new(btf, btf_dump_snprintf, str, NULL); if (!ASSERT_OK_PTR(d, "could not create BTF dump")) goto out; @@ -837,7 +831,6 @@ out: void test_btf_dump() { char str[STRSIZE]; - struct btf_dump_opts opts = { .ctx = str }; struct btf_dump *d; struct btf *btf; int i; @@ -857,7 +850,7 @@ void test_btf_dump() { if (!ASSERT_OK_PTR(btf, "no kernel BTF found")) return; - d = btf_dump__new(btf, NULL, &opts, btf_dump_snprintf); + d = btf_dump__new(btf, btf_dump_snprintf, str, NULL); if (!ASSERT_OK_PTR(d, "could not create BTF dump")) return; diff --git a/tools/testing/selftests/bpf/prog_tests/btf_split.c b/tools/testing/selftests/bpf/prog_tests/btf_split.c index b1ffe61f2aa9..eef1158676ed 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_split.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_split.c @@ -13,7 +13,6 @@ static void btf_dump_printf(void *ctx, const char *fmt, va_list args) } void test_btf_split() { - struct btf_dump_opts opts; struct btf_dump *d = NULL; const struct btf_type *t; struct btf *btf1, *btf2; @@ -68,8 +67,7 @@ void test_btf_split() { dump_buf_file = open_memstream(&dump_buf, &dump_buf_sz); if (!ASSERT_OK_PTR(dump_buf_file, "dump_memstream")) return; - opts.ctx = dump_buf_file; - d = btf_dump__new(btf2, NULL, &opts, btf_dump_printf); + d = btf_dump__new(btf2, btf_dump_printf, dump_buf_file, NULL); if (!ASSERT_OK_PTR(d, "btf_dump__new")) goto cleanup; for (i = 1; i < btf__type_cnt(btf2); i++) { From eda8bfa5b7c76d332ece1f24a3662ca843fd880a Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:23 -0800 Subject: [PATCH 0122/2048] tools/runqslower: Update perf_buffer__new() calls Use v1.0+ compatible variant of perf_buffer__new() call to prepare for deprecation. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-9-andrii@kernel.org --- tools/bpf/runqslower/runqslower.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/tools/bpf/runqslower/runqslower.c b/tools/bpf/runqslower/runqslower.c index d89715844952..2414cc764461 100644 --- a/tools/bpf/runqslower/runqslower.c +++ b/tools/bpf/runqslower/runqslower.c @@ -123,7 +123,6 @@ int main(int argc, char **argv) .parser = parse_arg, .doc = argp_program_doc, }; - struct perf_buffer_opts pb_opts; struct perf_buffer *pb = NULL; struct runqslower_bpf *obj; int err; @@ -165,9 +164,8 @@ int main(int argc, char **argv) printf("Tracing run queue latency higher than %llu us\n", env.min_us); printf("%-8s %-16s %-6s %14s\n", "TIME", "COMM", "PID", "LAT(us)"); - pb_opts.sample_cb = handle_event; - pb_opts.lost_cb = handle_lost_events; - pb = perf_buffer__new(bpf_map__fd(obj->maps.events), 64, &pb_opts); + pb = perf_buffer__new(bpf_map__fd(obj->maps.events), 64, + handle_event, handle_lost_events, NULL, NULL); err = libbpf_get_error(pb); if (err) { pb = NULL; From 164b04f27fbd769f57905dfddd2a8953974eeef4 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 10 Nov 2021 21:36:24 -0800 Subject: [PATCH 0123/2048] bpftool: Update btf_dump__new() and perf_buffer__new_raw() calls Use v1.0-compatible variants of btf_dump and perf_buffer "constructors". This is also a demonstration of reusing struct perf_buffer_raw_opts as OPTS-style option struct for new perf_buffer__new_raw() API. Signed-off-by: Andrii Nakryiko Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211111053624.190580-10-andrii@kernel.org --- tools/bpf/bpftool/btf.c | 2 +- tools/bpf/bpftool/gen.c | 2 +- tools/bpf/bpftool/map_perf_ring.c | 9 +++------ 3 files changed, 5 insertions(+), 8 deletions(-) diff --git a/tools/bpf/bpftool/btf.c b/tools/bpf/bpftool/btf.c index 015d2758f826..223ac7676027 100644 --- a/tools/bpf/bpftool/btf.c +++ b/tools/bpf/bpftool/btf.c @@ -418,7 +418,7 @@ static int dump_btf_c(const struct btf *btf, struct btf_dump *d; int err = 0, i; - d = btf_dump__new(btf, NULL, NULL, btf_dump_printf); + d = btf_dump__new(btf, btf_dump_printf, NULL, NULL); if (IS_ERR(d)) return PTR_ERR(d); diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c index 5c18351290f0..89f0e828bbfa 100644 --- a/tools/bpf/bpftool/gen.c +++ b/tools/bpf/bpftool/gen.c @@ -218,7 +218,7 @@ static int codegen_datasecs(struct bpf_object *obj, const char *obj_name) char sec_ident[256], map_ident[256]; int i, err = 0; - d = btf_dump__new(btf, NULL, NULL, codegen_btf_dump_printf); + d = btf_dump__new(btf, codegen_btf_dump_printf, NULL, NULL); if (IS_ERR(d)) return PTR_ERR(d); diff --git a/tools/bpf/bpftool/map_perf_ring.c b/tools/bpf/bpftool/map_perf_ring.c index b98ea702d284..6b0c410152de 100644 --- a/tools/bpf/bpftool/map_perf_ring.c +++ b/tools/bpf/bpftool/map_perf_ring.c @@ -124,7 +124,7 @@ int do_event_pipe(int argc, char **argv) .wakeup_events = 1, }; struct bpf_map_info map_info = {}; - struct perf_buffer_raw_opts opts = {}; + LIBBPF_OPTS(perf_buffer_raw_opts, opts); struct event_pipe_ctx ctx = { .all_cpus = true, .cpu = -1, @@ -190,14 +190,11 @@ int do_event_pipe(int argc, char **argv) ctx.idx = 0; } - opts.attr = &perf_attr; - opts.event_cb = print_bpf_output; - opts.ctx = &ctx; opts.cpu_cnt = ctx.all_cpus ? 0 : 1; opts.cpus = &ctx.cpu; opts.map_keys = &ctx.idx; - - pb = perf_buffer__new_raw(map_fd, MMAP_PAGE_CNT, &opts); + pb = perf_buffer__new_raw(map_fd, MMAP_PAGE_CNT, &perf_attr, + print_bpf_output, &ctx, &opts); err = libbpf_get_error(pb); if (err) { p_err("failed to create perf buffer: %s (%d)", From 8c42d2fa4eeab6c37a0b1b1aa7a2715248ef4f34 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:09 -0800 Subject: [PATCH 0124/2048] bpf: Support BTF_KIND_TYPE_TAG for btf_type_tag attributes LLVM patches ([1] for clang, [2] and [3] for BPF backend) added support for btf_type_tag attributes. This patch added support for the kernel. The main motivation for btf_type_tag is to bring kernel annotations __user, __rcu etc. to btf. With such information available in btf, bpf verifier can detect mis-usages and reject the program. For example, for __user tagged pointer, developers can then use proper helper like bpf_probe_read_user() etc. to read the data. BTF_KIND_TYPE_TAG may also useful for other tracing facility where instead of to require user to specify kernel/user address type, the kernel can detect it by itself with btf. [1] https://reviews.llvm.org/D111199 [2] https://reviews.llvm.org/D113222 [3] https://reviews.llvm.org/D113496 Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012609.1505032-1-yhs@fb.com --- include/uapi/linux/btf.h | 3 ++- kernel/bpf/btf.c | 14 +++++++++++++- tools/include/uapi/linux/btf.h | 3 ++- 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/include/uapi/linux/btf.h b/include/uapi/linux/btf.h index deb12f755f0f..b0d8fea1951d 100644 --- a/include/uapi/linux/btf.h +++ b/include/uapi/linux/btf.h @@ -43,7 +43,7 @@ struct btf_type { * "size" tells the size of the type it is describing. * * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT, - * FUNC, FUNC_PROTO, VAR and DECL_TAG. + * FUNC, FUNC_PROTO, VAR, DECL_TAG and TYPE_TAG. * "type" is a type_id referring to another type. */ union { @@ -75,6 +75,7 @@ enum { BTF_KIND_DATASEC = 15, /* Section */ BTF_KIND_FLOAT = 16, /* Floating point */ BTF_KIND_DECL_TAG = 17, /* Decl Tag */ + BTF_KIND_TYPE_TAG = 18, /* Type Tag */ NR_BTF_KINDS, BTF_KIND_MAX = NR_BTF_KINDS - 1, diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index cdb0fba65600..1dd9ba82da1e 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -282,6 +282,7 @@ static const char * const btf_kind_str[NR_BTF_KINDS] = { [BTF_KIND_DATASEC] = "DATASEC", [BTF_KIND_FLOAT] = "FLOAT", [BTF_KIND_DECL_TAG] = "DECL_TAG", + [BTF_KIND_TYPE_TAG] = "TYPE_TAG", }; const char *btf_type_str(const struct btf_type *t) @@ -418,6 +419,7 @@ static bool btf_type_is_modifier(const struct btf_type *t) case BTF_KIND_VOLATILE: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: + case BTF_KIND_TYPE_TAG: return true; } @@ -1737,6 +1739,7 @@ __btf_resolve_size(const struct btf *btf, const struct btf_type *type, case BTF_KIND_VOLATILE: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: + case BTF_KIND_TYPE_TAG: id = type->type; type = btf_type_by_id(btf, type->type); break; @@ -2345,6 +2348,8 @@ static int btf_ref_type_check_meta(struct btf_verifier_env *env, const struct btf_type *t, u32 meta_left) { + const char *value; + if (btf_type_vlen(t)) { btf_verifier_log_type(env, t, "vlen != 0"); return -EINVAL; @@ -2360,7 +2365,7 @@ static int btf_ref_type_check_meta(struct btf_verifier_env *env, return -EINVAL; } - /* typedef type must have a valid name, and other ref types, + /* typedef/type_tag type must have a valid name, and other ref types, * volatile, const, restrict, should have a null name. */ if (BTF_INFO_KIND(t->info) == BTF_KIND_TYPEDEF) { @@ -2369,6 +2374,12 @@ static int btf_ref_type_check_meta(struct btf_verifier_env *env, btf_verifier_log_type(env, t, "Invalid name"); return -EINVAL; } + } else if (BTF_INFO_KIND(t->info) == BTF_KIND_TYPE_TAG) { + value = btf_name_by_offset(env->btf, t->name_off); + if (!value || !value[0]) { + btf_verifier_log_type(env, t, "Invalid name"); + return -EINVAL; + } } else { if (t->name_off) { btf_verifier_log_type(env, t, "Invalid name"); @@ -4059,6 +4070,7 @@ static const struct btf_kind_operations * const kind_ops[NR_BTF_KINDS] = { [BTF_KIND_DATASEC] = &datasec_ops, [BTF_KIND_FLOAT] = &float_ops, [BTF_KIND_DECL_TAG] = &decl_tag_ops, + [BTF_KIND_TYPE_TAG] = &modifier_ops, }; static s32 btf_check_meta(struct btf_verifier_env *env, diff --git a/tools/include/uapi/linux/btf.h b/tools/include/uapi/linux/btf.h index deb12f755f0f..b0d8fea1951d 100644 --- a/tools/include/uapi/linux/btf.h +++ b/tools/include/uapi/linux/btf.h @@ -43,7 +43,7 @@ struct btf_type { * "size" tells the size of the type it is describing. * * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT, - * FUNC, FUNC_PROTO, VAR and DECL_TAG. + * FUNC, FUNC_PROTO, VAR, DECL_TAG and TYPE_TAG. * "type" is a type_id referring to another type. */ union { @@ -75,6 +75,7 @@ enum { BTF_KIND_DATASEC = 15, /* Section */ BTF_KIND_FLOAT = 16, /* Floating point */ BTF_KIND_DECL_TAG = 17, /* Decl Tag */ + BTF_KIND_TYPE_TAG = 18, /* Type Tag */ NR_BTF_KINDS, BTF_KIND_MAX = NR_BTF_KINDS - 1, From 2dc1e488e5cdfd937554ca81fd46ad874d244b3f Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:14 -0800 Subject: [PATCH 0125/2048] libbpf: Support BTF_KIND_TYPE_TAG Add libbpf support for BTF_KIND_TYPE_TAG. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012614.1505315-1-yhs@fb.com --- tools/lib/bpf/btf.c | 23 +++++++++++++++++++++++ tools/lib/bpf/btf.h | 9 ++++++++- tools/lib/bpf/btf_dump.c | 9 +++++++++ tools/lib/bpf/libbpf.c | 31 ++++++++++++++++++++++++++++++- tools/lib/bpf/libbpf.map | 1 + tools/lib/bpf/libbpf_internal.h | 2 ++ 6 files changed, 73 insertions(+), 2 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index fcec27622e7a..fadf089ae8fe 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -299,6 +299,7 @@ static int btf_type_size(const struct btf_type *t) case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: case BTF_KIND_FLOAT: + case BTF_KIND_TYPE_TAG: return base_size; case BTF_KIND_INT: return base_size + sizeof(__u32); @@ -349,6 +350,7 @@ static int btf_bswap_type_rest(struct btf_type *t) case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: case BTF_KIND_FLOAT: + case BTF_KIND_TYPE_TAG: return 0; case BTF_KIND_INT: *(__u32 *)(t + 1) = bswap_32(*(__u32 *)(t + 1)); @@ -649,6 +651,7 @@ int btf__align_of(const struct btf *btf, __u32 id) case BTF_KIND_VOLATILE: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: + case BTF_KIND_TYPE_TAG: return btf__align_of(btf, t->type); case BTF_KIND_ARRAY: return btf__align_of(btf, btf_array(t)->type); @@ -2235,6 +2238,22 @@ int btf__add_restrict(struct btf *btf, int ref_type_id) return btf_add_ref_kind(btf, BTF_KIND_RESTRICT, NULL, ref_type_id); } +/* + * Append new BTF_KIND_TYPE_TAG type with: + * - *value*, non-empty/non-NULL tag value; + * - *ref_type_id* - referenced type ID, it might not exist yet; + * Returns: + * - >0, type ID of newly added BTF type; + * - <0, on error. + */ +int btf__add_type_tag(struct btf *btf, const char *value, int ref_type_id) +{ + if (!value|| !value[0]) + return libbpf_err(-EINVAL); + + return btf_add_ref_kind(btf, BTF_KIND_TYPE_TAG, value, ref_type_id); +} + /* * Append new BTF_KIND_FUNC type with: * - *name*, non-empty/non-NULL name; @@ -3639,6 +3658,7 @@ static int btf_dedup_prep(struct btf_dedup *d) case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: case BTF_KIND_FLOAT: + case BTF_KIND_TYPE_TAG: h = btf_hash_common(t); break; case BTF_KIND_INT: @@ -3699,6 +3719,7 @@ static int btf_dedup_prim_type(struct btf_dedup *d, __u32 type_id) case BTF_KIND_VAR: case BTF_KIND_DATASEC: case BTF_KIND_DECL_TAG: + case BTF_KIND_TYPE_TAG: return 0; case BTF_KIND_INT: @@ -4297,6 +4318,7 @@ static int btf_dedup_ref_type(struct btf_dedup *d, __u32 type_id) case BTF_KIND_PTR: case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: + case BTF_KIND_TYPE_TAG: ref_type_id = btf_dedup_ref_type(d, t->type); if (ref_type_id < 0) return ref_type_id; @@ -4603,6 +4625,7 @@ int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ct case BTF_KIND_FUNC: case BTF_KIND_VAR: case BTF_KIND_DECL_TAG: + case BTF_KIND_TYPE_TAG: return visit(&t->type, ctx); case BTF_KIND_ARRAY: { diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h index 45310c65e865..5c73a5b0a044 100644 --- a/tools/lib/bpf/btf.h +++ b/tools/lib/bpf/btf.h @@ -227,6 +227,7 @@ LIBBPF_API int btf__add_typedef(struct btf *btf, const char *name, int ref_type_ LIBBPF_API int btf__add_volatile(struct btf *btf, int ref_type_id); LIBBPF_API int btf__add_const(struct btf *btf, int ref_type_id); LIBBPF_API int btf__add_restrict(struct btf *btf, int ref_type_id); +LIBBPF_API int btf__add_type_tag(struct btf *btf, const char *value, int ref_type_id); /* func and func_proto construction APIs */ LIBBPF_API int btf__add_func(struct btf *btf, const char *name, @@ -458,7 +459,8 @@ static inline bool btf_is_mod(const struct btf_type *t) return kind == BTF_KIND_VOLATILE || kind == BTF_KIND_CONST || - kind == BTF_KIND_RESTRICT; + kind == BTF_KIND_RESTRICT || + kind == BTF_KIND_TYPE_TAG; } static inline bool btf_is_func(const struct btf_type *t) @@ -491,6 +493,11 @@ static inline bool btf_is_decl_tag(const struct btf_type *t) return btf_kind(t) == BTF_KIND_DECL_TAG; } +static inline bool btf_is_type_tag(const struct btf_type *t) +{ + return btf_kind(t) == BTF_KIND_TYPE_TAG; +} + static inline __u8 btf_int_encoding(const struct btf_type *t) { return BTF_INT_ENCODING(*(__u32 *)(t + 1)); diff --git a/tools/lib/bpf/btf_dump.c b/tools/lib/bpf/btf_dump.c index b8cd7e4f557a..05f3e7dfec0a 100644 --- a/tools/lib/bpf/btf_dump.c +++ b/tools/lib/bpf/btf_dump.c @@ -330,6 +330,7 @@ static int btf_dump_mark_referenced(struct btf_dump *d) case BTF_KIND_FUNC: case BTF_KIND_VAR: case BTF_KIND_DECL_TAG: + case BTF_KIND_TYPE_TAG: d->type_states[t->type].referenced = 1; break; @@ -573,6 +574,7 @@ static int btf_dump_order_type(struct btf_dump *d, __u32 id, bool through_ptr) case BTF_KIND_VOLATILE: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: + case BTF_KIND_TYPE_TAG: return btf_dump_order_type(d, t->type, through_ptr); case BTF_KIND_FUNC_PROTO: { @@ -747,6 +749,7 @@ static void btf_dump_emit_type(struct btf_dump *d, __u32 id, __u32 cont_id) case BTF_KIND_VOLATILE: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: + case BTF_KIND_TYPE_TAG: btf_dump_emit_type(d, t->type, cont_id); break; case BTF_KIND_ARRAY: @@ -1167,6 +1170,7 @@ skip_mod: case BTF_KIND_CONST: case BTF_KIND_RESTRICT: case BTF_KIND_FUNC_PROTO: + case BTF_KIND_TYPE_TAG: id = t->type; break; case BTF_KIND_ARRAY: @@ -1335,6 +1339,11 @@ static void btf_dump_emit_type_chain(struct btf_dump *d, case BTF_KIND_RESTRICT: btf_dump_printf(d, " restrict"); break; + case BTF_KIND_TYPE_TAG: + btf_dump_emit_mods(d, decls); + name = btf_name_of(d, t->name_off); + btf_dump_printf(d, " __attribute__((btf_type_tag(\"%s\")))", name); + break; case BTF_KIND_ARRAY: { const struct btf_array *a = btf_array(t); const struct btf_type *next_t; diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 7db14859b27c..de7e09a6b5ec 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -197,6 +197,8 @@ enum kern_feature_id { FEAT_PERF_LINK, /* BTF_KIND_DECL_TAG support */ FEAT_BTF_DECL_TAG, + /* BTF_KIND_TYPE_TAG support */ + FEAT_BTF_TYPE_TAG, __FEAT_CNT, }; @@ -2076,6 +2078,7 @@ static const char *__btf_kind_str(__u16 kind) case BTF_KIND_DATASEC: return "datasec"; case BTF_KIND_FLOAT: return "float"; case BTF_KIND_DECL_TAG: return "decl_tag"; + case BTF_KIND_TYPE_TAG: return "type_tag"; default: return "unknown"; } } @@ -2588,8 +2591,10 @@ static bool btf_needs_sanitization(struct bpf_object *obj) bool has_float = kernel_supports(obj, FEAT_BTF_FLOAT); bool has_func = kernel_supports(obj, FEAT_BTF_FUNC); bool has_decl_tag = kernel_supports(obj, FEAT_BTF_DECL_TAG); + bool has_type_tag = kernel_supports(obj, FEAT_BTF_TYPE_TAG); - return !has_func || !has_datasec || !has_func_global || !has_float || !has_decl_tag; + return !has_func || !has_datasec || !has_func_global || !has_float || + !has_decl_tag || !has_type_tag; } static void bpf_object__sanitize_btf(struct bpf_object *obj, struct btf *btf) @@ -2599,6 +2604,7 @@ static void bpf_object__sanitize_btf(struct bpf_object *obj, struct btf *btf) bool has_float = kernel_supports(obj, FEAT_BTF_FLOAT); bool has_func = kernel_supports(obj, FEAT_BTF_FUNC); bool has_decl_tag = kernel_supports(obj, FEAT_BTF_DECL_TAG); + bool has_type_tag = kernel_supports(obj, FEAT_BTF_TYPE_TAG); struct btf_type *t; int i, j, vlen; @@ -2657,6 +2663,10 @@ static void bpf_object__sanitize_btf(struct bpf_object *obj, struct btf *btf) */ t->name_off = 0; t->info = BTF_INFO_ENC(BTF_KIND_STRUCT, 0, 0); + } else if (!has_type_tag && btf_is_type_tag(t)) { + /* replace TYPE_TAG with a CONST */ + t->name_off = 0; + t->info = BTF_INFO_ENC(BTF_KIND_CONST, 0, 0); } } } @@ -4460,6 +4470,22 @@ static int probe_kern_btf_decl_tag(void) strs, sizeof(strs))); } +static int probe_kern_btf_type_tag(void) +{ + static const char strs[] = "\0tag"; + __u32 types[] = { + /* int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + /* attr */ + BTF_TYPE_TYPE_TAG_ENC(1, 1), /* [2] */ + /* ptr */ + BTF_TYPE_ENC(0, BTF_INFO_ENC(BTF_KIND_PTR, 0, 0), 2), /* [3] */ + }; + + return probe_fd(libbpf__load_raw_btf((char *)types, sizeof(types), + strs, sizeof(strs))); +} + static int probe_kern_array_mmap(void) { struct bpf_create_map_attr attr = { @@ -4657,6 +4683,9 @@ static struct kern_feature_desc { [FEAT_BTF_DECL_TAG] = { "BTF_KIND_DECL_TAG support", probe_kern_btf_decl_tag, }, + [FEAT_BTF_TYPE_TAG] = { + "BTF_KIND_TYPE_TAG support", probe_kern_btf_type_tag, + }, }; static bool kernel_supports(const struct bpf_object *obj, enum kern_feature_id feat_id) diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index 9e91aa8da303..6a59514a48cf 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -403,6 +403,7 @@ LIBBPF_0.6.0 { bpf_program__set_extra_flags; btf__add_btf; btf__add_decl_tag; + btf__add_type_tag; btf__dedup; btf__dedup_deprecated; btf__raw_data; diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index c1e34794b829..f7ac349650a1 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -73,6 +73,8 @@ BTF_TYPE_ENC(name, BTF_INFO_ENC(BTF_KIND_FLOAT, 0, 0), sz) #define BTF_TYPE_DECL_TAG_ENC(value, type, component_idx) \ BTF_TYPE_ENC(value, BTF_INFO_ENC(BTF_KIND_DECL_TAG, 0, 0), type), (component_idx) +#define BTF_TYPE_TYPE_TAG_ENC(value, type) \ + BTF_TYPE_ENC(value, BTF_INFO_ENC(BTF_KIND_TYPE_TAG, 0, 0), type) #ifndef likely #define likely(x) __builtin_expect(!!(x), 1) From 3da5ba6f0509ace03cad38b554c89797129e90be Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:20 -0800 Subject: [PATCH 0126/2048] bpftool: Support BTF_KIND_TYPE_TAG Add bpftool support for BTF_KIND_TYPE_TAG. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012620.1505506-1-yhs@fb.com --- tools/bpf/bpftool/btf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/bpf/bpftool/btf.c b/tools/bpf/bpftool/btf.c index 223ac7676027..c7e3b0b0029e 100644 --- a/tools/bpf/bpftool/btf.c +++ b/tools/bpf/bpftool/btf.c @@ -39,6 +39,7 @@ static const char * const btf_kind_str[NR_BTF_KINDS] = { [BTF_KIND_DATASEC] = "DATASEC", [BTF_KIND_FLOAT] = "FLOAT", [BTF_KIND_DECL_TAG] = "DECL_TAG", + [BTF_KIND_TYPE_TAG] = "TYPE_TAG", }; struct btf_attach_point { @@ -142,6 +143,7 @@ static int dump_btf_type(const struct btf *btf, __u32 id, case BTF_KIND_VOLATILE: case BTF_KIND_RESTRICT: case BTF_KIND_TYPEDEF: + case BTF_KIND_TYPE_TAG: if (json_output) jsonw_uint_field(w, "type_id", t->type); else From 0dc85872203bf7b15c56c7eb228b8f3fabb17ac2 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:25 -0800 Subject: [PATCH 0127/2048] selftests/bpf: Test libbpf API function btf__add_type_tag() Add unit tests for btf__add_type_tag(). Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012625.1505748-1-yhs@fb.com --- tools/testing/selftests/bpf/btf_helpers.c | 4 +- .../selftests/bpf/prog_tests/btf_write.c | 67 +++++++++++-------- 2 files changed, 43 insertions(+), 28 deletions(-) diff --git a/tools/testing/selftests/bpf/btf_helpers.c b/tools/testing/selftests/bpf/btf_helpers.c index acb59202486d..b5941d514e17 100644 --- a/tools/testing/selftests/bpf/btf_helpers.c +++ b/tools/testing/selftests/bpf/btf_helpers.c @@ -25,11 +25,12 @@ static const char * const btf_kind_str_mapping[] = { [BTF_KIND_DATASEC] = "DATASEC", [BTF_KIND_FLOAT] = "FLOAT", [BTF_KIND_DECL_TAG] = "DECL_TAG", + [BTF_KIND_TYPE_TAG] = "TYPE_TAG", }; static const char *btf_kind_str(__u16 kind) { - if (kind > BTF_KIND_DECL_TAG) + if (kind > BTF_KIND_TYPE_TAG) return "UNKNOWN"; return btf_kind_str_mapping[kind]; } @@ -109,6 +110,7 @@ int fprintf_btf_type_raw(FILE *out, const struct btf *btf, __u32 id) case BTF_KIND_VOLATILE: case BTF_KIND_RESTRICT: case BTF_KIND_TYPEDEF: + case BTF_KIND_TYPE_TAG: fprintf(out, " type_id=%u", t->type); break; case BTF_KIND_ARRAY: { diff --git a/tools/testing/selftests/bpf/prog_tests/btf_write.c b/tools/testing/selftests/bpf/prog_tests/btf_write.c index b912eeb0b6b4..addf99c05896 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_write.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_write.c @@ -297,6 +297,16 @@ static void gen_btf(struct btf *btf) ASSERT_EQ(btf_decl_tag(t)->component_idx, 1, "tag_component_idx"); ASSERT_STREQ(btf_type_raw_dump(btf, 19), "[19] DECL_TAG 'tag2' type_id=14 component_idx=1", "raw_dump"); + + /* TYPE_TAG */ + id = btf__add_type_tag(btf, "tag1", 1); + ASSERT_EQ(id, 20, "tag_id"); + t = btf__type_by_id(btf, 20); + ASSERT_STREQ(btf__str_by_offset(btf, t->name_off), "tag1", "tag_value"); + ASSERT_EQ(btf_kind(t), BTF_KIND_TYPE_TAG, "tag_kind"); + ASSERT_EQ(t->type, 1, "tag_type"); + ASSERT_STREQ(btf_type_raw_dump(btf, 20), + "[20] TYPE_TAG 'tag1' type_id=1", "raw_dump"); } static void test_btf_add() @@ -337,7 +347,8 @@ static void test_btf_add() "[17] DATASEC 'datasec1' size=12 vlen=1\n" "\ttype_id=1 offset=4 size=8", "[18] DECL_TAG 'tag1' type_id=16 component_idx=-1", - "[19] DECL_TAG 'tag2' type_id=14 component_idx=1"); + "[19] DECL_TAG 'tag2' type_id=14 component_idx=1", + "[20] TYPE_TAG 'tag1' type_id=1"); btf__free(btf); } @@ -359,7 +370,7 @@ static void test_btf_add_btf() gen_btf(btf2); id = btf__add_btf(btf1, btf2); - if (!ASSERT_EQ(id, 20, "id")) + if (!ASSERT_EQ(id, 21, "id")) goto cleanup; VALIDATE_RAW_BTF( @@ -391,35 +402,37 @@ static void test_btf_add_btf() "\ttype_id=1 offset=4 size=8", "[18] DECL_TAG 'tag1' type_id=16 component_idx=-1", "[19] DECL_TAG 'tag2' type_id=14 component_idx=1", + "[20] TYPE_TAG 'tag1' type_id=1", /* types appended from the second BTF */ - "[20] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", - "[21] PTR '(anon)' type_id=20", - "[22] CONST '(anon)' type_id=24", - "[23] VOLATILE '(anon)' type_id=22", - "[24] RESTRICT '(anon)' type_id=23", - "[25] ARRAY '(anon)' type_id=21 index_type_id=20 nr_elems=10", - "[26] STRUCT 's1' size=8 vlen=2\n" - "\t'f1' type_id=20 bits_offset=0\n" - "\t'f2' type_id=20 bits_offset=32 bitfield_size=16", - "[27] UNION 'u1' size=8 vlen=1\n" - "\t'f1' type_id=20 bits_offset=0 bitfield_size=16", - "[28] ENUM 'e1' size=4 vlen=2\n" + "[21] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[22] PTR '(anon)' type_id=21", + "[23] CONST '(anon)' type_id=25", + "[24] VOLATILE '(anon)' type_id=23", + "[25] RESTRICT '(anon)' type_id=24", + "[26] ARRAY '(anon)' type_id=22 index_type_id=21 nr_elems=10", + "[27] STRUCT 's1' size=8 vlen=2\n" + "\t'f1' type_id=21 bits_offset=0\n" + "\t'f2' type_id=21 bits_offset=32 bitfield_size=16", + "[28] UNION 'u1' size=8 vlen=1\n" + "\t'f1' type_id=21 bits_offset=0 bitfield_size=16", + "[29] ENUM 'e1' size=4 vlen=2\n" "\t'v1' val=1\n" "\t'v2' val=2", - "[29] FWD 'struct_fwd' fwd_kind=struct", - "[30] FWD 'union_fwd' fwd_kind=union", - "[31] ENUM 'enum_fwd' size=4 vlen=0", - "[32] TYPEDEF 'typedef1' type_id=20", - "[33] FUNC 'func1' type_id=34 linkage=global", - "[34] FUNC_PROTO '(anon)' ret_type_id=20 vlen=2\n" - "\t'p1' type_id=20\n" - "\t'p2' type_id=21", - "[35] VAR 'var1' type_id=20, linkage=global-alloc", - "[36] DATASEC 'datasec1' size=12 vlen=1\n" - "\ttype_id=20 offset=4 size=8", - "[37] DECL_TAG 'tag1' type_id=35 component_idx=-1", - "[38] DECL_TAG 'tag2' type_id=33 component_idx=1"); + "[30] FWD 'struct_fwd' fwd_kind=struct", + "[31] FWD 'union_fwd' fwd_kind=union", + "[32] ENUM 'enum_fwd' size=4 vlen=0", + "[33] TYPEDEF 'typedef1' type_id=21", + "[34] FUNC 'func1' type_id=35 linkage=global", + "[35] FUNC_PROTO '(anon)' ret_type_id=21 vlen=2\n" + "\t'p1' type_id=21\n" + "\t'p2' type_id=22", + "[36] VAR 'var1' type_id=21, linkage=global-alloc", + "[37] DATASEC 'datasec1' size=12 vlen=1\n" + "\ttype_id=21 offset=4 size=8", + "[38] DECL_TAG 'tag1' type_id=36 component_idx=-1", + "[39] DECL_TAG 'tag2' type_id=34 component_idx=1", + "[40] TYPE_TAG 'tag1' type_id=21"); cleanup: btf__free(btf1); From 6aa5dabc9d0ef722905e4ca4f9751d70cf3ec8a4 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:30 -0800 Subject: [PATCH 0128/2048] selftests/bpf: Add BTF_KIND_TYPE_TAG unit tests Add BTF_KIND_TYPE_TAG unit tests. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012630.1506095-1-yhs@fb.com --- tools/testing/selftests/bpf/prog_tests/btf.c | 18 ++++++++++++++++++ tools/testing/selftests/bpf/test_btf.h | 3 +++ 2 files changed, 21 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/btf.c b/tools/testing/selftests/bpf/prog_tests/btf.c index 1e8b36d74df2..88510a2d9858 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf.c +++ b/tools/testing/selftests/bpf/prog_tests/btf.c @@ -3939,6 +3939,23 @@ static struct btf_raw_test raw_tests[] = { .btf_load_err = true, .err_str = "Invalid component_idx", }, +{ + .descr = "type_tag test #1", + .raw_types = { + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_TBD, 1), /* [2] */ + BTF_PTR_ENC(2), /* [3] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag"), + .map_type = BPF_MAP_TYPE_ARRAY, + .map_name = "tag_type_check_btf", + .key_size = sizeof(int), + .value_size = 4, + .key_type_id = 1, + .value_type_id = 1, + .max_entries = 1, +}, }; /* struct btf_raw_test raw_tests[] */ @@ -7222,6 +7239,7 @@ static int btf_type_size(const struct btf_type *t) case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: case BTF_KIND_FLOAT: + case BTF_KIND_TYPE_TAG: return base_size; case BTF_KIND_INT: return base_size + sizeof(__u32); diff --git a/tools/testing/selftests/bpf/test_btf.h b/tools/testing/selftests/bpf/test_btf.h index 32c7a57867da..128989bed8b7 100644 --- a/tools/testing/selftests/bpf/test_btf.h +++ b/tools/testing/selftests/bpf/test_btf.h @@ -72,4 +72,7 @@ #define BTF_DECL_TAG_ENC(value, type, component_idx) \ BTF_TYPE_ENC(value, BTF_INFO_ENC(BTF_KIND_DECL_TAG, 0, 0), type), (component_idx) +#define BTF_TYPE_TAG_ENC(value, type) \ + BTF_TYPE_ENC(value, BTF_INFO_ENC(BTF_KIND_TYPE_TAG, 0, 0), type) + #endif /* _TEST_BTF_H */ From 846f4826d18e660ab668eb26e83c6adf0ceb24d2 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:35 -0800 Subject: [PATCH 0129/2048] selftests/bpf: Test BTF_KIND_DECL_TAG for deduplication Add BTF_KIND_TYPE_TAG duplication unit tests. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211112012635.1506853-1-yhs@fb.com --- tools/testing/selftests/bpf/prog_tests/btf.c | 139 ++++++++++++++++++- 1 file changed, 135 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/btf.c b/tools/testing/selftests/bpf/prog_tests/btf.c index 88510a2d9858..4aa6343dc4c8 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf.c +++ b/tools/testing/selftests/bpf/prog_tests/btf.c @@ -6878,15 +6878,16 @@ static struct btf_dedup_test dedup_tests[] = { BTF_RESTRICT_ENC(8), /* [11] restrict */ BTF_FUNC_PROTO_ENC(1, 2), /* [12] func_proto */ BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 1), - BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 8), + BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 18), BTF_FUNC_ENC(NAME_TBD, 12), /* [13] func */ BTF_TYPE_FLOAT_ENC(NAME_TBD, 2), /* [14] float */ BTF_DECL_TAG_ENC(NAME_TBD, 13, -1), /* [15] decl_tag */ BTF_DECL_TAG_ENC(NAME_TBD, 13, 1), /* [16] decl_tag */ BTF_DECL_TAG_ENC(NAME_TBD, 7, -1), /* [17] decl_tag */ + BTF_TYPE_TAG_ENC(NAME_TBD, 8), /* [18] type_tag */ BTF_END_RAW, }, - BTF_STR_SEC("\0A\0B\0C\0D\0E\0F\0G\0H\0I\0J\0K\0L\0M\0N\0O\0P\0Q"), + BTF_STR_SEC("\0A\0B\0C\0D\0E\0F\0G\0H\0I\0J\0K\0L\0M\0N\0O\0P\0Q\0R"), }, .expect = { .raw_types = { @@ -6907,15 +6908,16 @@ static struct btf_dedup_test dedup_tests[] = { BTF_RESTRICT_ENC(8), /* [11] restrict */ BTF_FUNC_PROTO_ENC(1, 2), /* [12] func_proto */ BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 1), - BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 8), + BTF_FUNC_PROTO_ARG_ENC(NAME_TBD, 18), BTF_FUNC_ENC(NAME_TBD, 12), /* [13] func */ BTF_TYPE_FLOAT_ENC(NAME_TBD, 2), /* [14] float */ BTF_DECL_TAG_ENC(NAME_TBD, 13, -1), /* [15] decl_tag */ BTF_DECL_TAG_ENC(NAME_TBD, 13, 1), /* [16] decl_tag */ BTF_DECL_TAG_ENC(NAME_TBD, 7, -1), /* [17] decl_tag */ + BTF_TYPE_TAG_ENC(NAME_TBD, 8), /* [18] type_tag */ BTF_END_RAW, }, - BTF_STR_SEC("\0A\0B\0C\0D\0E\0F\0G\0H\0I\0J\0K\0L\0M\0N\0O\0P\0Q"), + BTF_STR_SEC("\0A\0B\0C\0D\0E\0F\0G\0H\0I\0J\0K\0L\0M\0N\0O\0P\0Q\0R"), }, }, { @@ -7221,6 +7223,135 @@ static struct btf_dedup_test dedup_tests[] = { BTF_STR_SEC("\0t\0tag1\0tag2\0tag3"), }, }, +{ + .descr = "dedup: btf_type_tag #1", + .input = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [5] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 5), /* [6] */ + BTF_PTR_ENC(6), /* [7] */ + /* ptr -> tag1 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [8] */ + BTF_PTR_ENC(8), /* [9] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, + .expect = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag1 -> int */ + BTF_PTR_ENC(2), /* [5] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, +}, +{ + .descr = "dedup: btf_type_tag #2", + .input = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag2 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 1), /* [5] */ + BTF_PTR_ENC(5), /* [6] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, + .expect = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag2 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 1), /* [5] */ + BTF_PTR_ENC(5), /* [6] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, +}, +{ + .descr = "dedup: btf_type_tag #3", + .input = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag1 -> tag2 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 1), /* [5] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 5), /* [6] */ + BTF_PTR_ENC(6), /* [7] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, + .expect = { + .raw_types = { + /* ptr -> tag2 -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 2), /* [3] */ + BTF_PTR_ENC(3), /* [4] */ + /* ptr -> tag1 -> tag2 -> int */ + BTF_TYPE_TAG_ENC(NAME_NTH(2), 1), /* [5] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 5), /* [6] */ + BTF_PTR_ENC(6), /* [7] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0tag2"), + }, +}, +{ + .descr = "dedup: btf_type_tag #4", + .input = { + .raw_types = { + /* ptr -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_PTR_ENC(2), /* [3] */ + /* ptr -> tag1 -> long */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 64, 8), /* [4] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 4), /* [5] */ + BTF_PTR_ENC(5), /* [6] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1"), + }, + .expect = { + .raw_types = { + /* ptr -> tag1 -> int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_PTR_ENC(2), /* [3] */ + /* ptr -> tag1 -> long */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 64, 8), /* [4] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 4), /* [5] */ + BTF_PTR_ENC(5), /* [6] */ + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1"), + }, +}, }; From 26c79fcbfa64b18ca1407a3be7ac3442aef51073 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:41 -0800 Subject: [PATCH 0130/2048] selftests/bpf: Rename progs/tag.c to progs/btf_decl_tag.c Rename progs/tag.c to progs/btf_decl_tag.c so we can introduce progs/btf_type_tag.c in the next patch. Also create a subtest for btf_decl_tag in prog_tests/btf_tag.c so we can introduce btf_type_tag subtest in the next patch. I also took opportunity to remove the check whether __has_attribute is defined or not in progs/btf_decl_tag.c since all recent clangs should already support this macro. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012641.1507144-1-yhs@fb.com --- .../selftests/bpf/prog_tests/btf_tag.c | 20 ++++++++++++------- .../bpf/progs/{tag.c => btf_decl_tag.c} | 4 ---- 2 files changed, 13 insertions(+), 11 deletions(-) rename tools/testing/selftests/bpf/progs/{tag.c => btf_decl_tag.c} (94%) diff --git a/tools/testing/selftests/bpf/prog_tests/btf_tag.c b/tools/testing/selftests/bpf/prog_tests/btf_tag.c index 91821f42714d..d15cc7a88182 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_tag.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_tag.c @@ -1,20 +1,26 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright (c) 2021 Facebook */ #include -#include "tag.skel.h" +#include "btf_decl_tag.skel.h" -void test_btf_tag(void) +static void test_btf_decl_tag(void) { - struct tag *skel; + struct btf_decl_tag *skel; - skel = tag__open_and_load(); - if (!ASSERT_OK_PTR(skel, "btf_tag")) + skel = btf_decl_tag__open_and_load(); + if (!ASSERT_OK_PTR(skel, "btf_decl_tag")) return; if (skel->rodata->skip_tests) { - printf("%s:SKIP: btf_tag attribute not supported", __func__); + printf("%s:SKIP: btf_decl_tag attribute not supported", __func__); test__skip(); } - tag__destroy(skel); + btf_decl_tag__destroy(skel); +} + +void test_btf_tag(void) +{ + if (test__start_subtest("btf_decl_tag")) + test_btf_decl_tag(); } diff --git a/tools/testing/selftests/bpf/progs/tag.c b/tools/testing/selftests/bpf/progs/btf_decl_tag.c similarity index 94% rename from tools/testing/selftests/bpf/progs/tag.c rename to tools/testing/selftests/bpf/progs/btf_decl_tag.c index 1792f4eda095..c88ccc53529a 100644 --- a/tools/testing/selftests/bpf/progs/tag.c +++ b/tools/testing/selftests/bpf/progs/btf_decl_tag.c @@ -4,10 +4,6 @@ #include #include -#ifndef __has_attribute -#define __has_attribute(x) 0 -#endif - #if __has_attribute(btf_decl_tag) #define __tag1 __attribute__((btf_decl_tag("tag1"))) #define __tag2 __attribute__((btf_decl_tag("tag2"))) From 5698a42a73a1d9cb7efd31ca1bf35daa87f5e1a9 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:46 -0800 Subject: [PATCH 0131/2048] selftests/bpf: Add a C test for btf_type_tag The following is the main btf_type_tag usage in the C test: #define __tag1 __attribute__((btf_type_tag("tag1"))) #define __tag2 __attribute__((btf_type_tag("tag2"))) struct btf_type_tag_test { int __tag1 * __tag1 __tag2 *p; } g; The bpftool raw dump with related types: [4] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED [11] STRUCT 'btf_type_tag_test' size=8 vlen=1 'p' type_id=14 bits_offset=0 [12] TYPE_TAG 'tag1' type_id=16 [13] TYPE_TAG 'tag2' type_id=12 [14] PTR '(anon)' type_id=13 [15] TYPE_TAG 'tag1' type_id=4 [16] PTR '(anon)' type_id=15 [17] VAR 'g' type_id=11, linkage=global With format C dump, we have struct btf_type_tag_test { int __attribute__((btf_type_tag("tag1"))) * __attribute__((btf_type_tag("tag1"))) __attribute__((btf_type_tag("tag2"))) *p; }; The result C code is identical to the original definition except macro's are gone. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211112012646.1508231-1-yhs@fb.com --- .../selftests/bpf/prog_tests/btf_tag.c | 24 ++++++++++++++++++ .../selftests/bpf/progs/btf_type_tag.c | 25 +++++++++++++++++++ 2 files changed, 49 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/btf_type_tag.c diff --git a/tools/testing/selftests/bpf/prog_tests/btf_tag.c b/tools/testing/selftests/bpf/prog_tests/btf_tag.c index d15cc7a88182..88d63e23e35f 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_tag.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_tag.c @@ -3,6 +3,12 @@ #include #include "btf_decl_tag.skel.h" +/* struct btf_type_tag_test is referenced in btf_type_tag.skel.h */ +struct btf_type_tag_test { + int **p; +}; +#include "btf_type_tag.skel.h" + static void test_btf_decl_tag(void) { struct btf_decl_tag *skel; @@ -19,8 +25,26 @@ static void test_btf_decl_tag(void) btf_decl_tag__destroy(skel); } +static void test_btf_type_tag(void) +{ + struct btf_type_tag *skel; + + skel = btf_type_tag__open_and_load(); + if (!ASSERT_OK_PTR(skel, "btf_type_tag")) + return; + + if (skel->rodata->skip_tests) { + printf("%s:SKIP: btf_type_tag attribute not supported", __func__); + test__skip(); + } + + btf_type_tag__destroy(skel); +} + void test_btf_tag(void) { if (test__start_subtest("btf_decl_tag")) test_btf_decl_tag(); + if (test__start_subtest("btf_type_tag")) + test_btf_type_tag(); } diff --git a/tools/testing/selftests/bpf/progs/btf_type_tag.c b/tools/testing/selftests/bpf/progs/btf_type_tag.c new file mode 100644 index 000000000000..1d488da7e920 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/btf_type_tag.c @@ -0,0 +1,25 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ +#include "vmlinux.h" +#include +#include + +#if __has_attribute(btf_type_tag) +#define __tag1 __attribute__((btf_type_tag("tag1"))) +#define __tag2 __attribute__((btf_type_tag("tag2"))) +volatile const bool skip_tests = false; +#else +#define __tag1 +#define __tag2 +volatile const bool skip_tests = true; +#endif + +struct btf_type_tag_test { + int __tag1 * __tag1 __tag2 *p; +} g; + +SEC("fentry/bpf_fentry_test1") +int BPF_PROG(sub, int x) +{ + return 0; +} From 3f1d0dc0ba290aab357083a0abfe267c8cffdc8d Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:51 -0800 Subject: [PATCH 0132/2048] selftests/bpf: Clarify llvm dependency with btf_tag selftest btf_tag selftest needs certain llvm versions (>= llvm14). Make it clear in the selftests README.rst file. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012651.1508549-1-yhs@fb.com --- tools/testing/selftests/bpf/README.rst | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/README.rst b/tools/testing/selftests/bpf/README.rst index 5e287e445f75..42ef250c7acc 100644 --- a/tools/testing/selftests/bpf/README.rst +++ b/tools/testing/selftests/bpf/README.rst @@ -204,16 +204,17 @@ __ https://reviews.llvm.org/D93563 btf_tag test and Clang version ============================== -The btf_tag selftest require LLVM support to recognize the btf_decl_tag attribute. -It was introduced in `Clang 14`__. +The btf_tag selftest requires LLVM support to recognize the btf_decl_tag and +btf_type_tag attributes. They are introduced in `Clang 14` [0_, 1_]. -Without it, the btf_tag selftest will be skipped and you will observe: +Without them, the btf_tag selftest will be skipped and you will observe: .. code-block:: console # btf_tag:SKIP -__ https://reviews.llvm.org/D111588 +.. _0: https://reviews.llvm.org/D111588 +.. _1: https://reviews.llvm.org/D111199 Clang dependencies for static linking tests =========================================== From d52f5c639dd8605d2563b77b190e278f615a2b8a Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 11 Nov 2021 17:26:56 -0800 Subject: [PATCH 0133/2048] docs/bpf: Update documentation for BTF_KIND_TYPE_TAG support Add BTF_KIND_TYPE_TAG documentation in btf.rst. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112012656.1509082-1-yhs@fb.com --- Documentation/bpf/btf.rst | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/Documentation/bpf/btf.rst b/Documentation/bpf/btf.rst index 9ad4218a751f..d0ec40d00c28 100644 --- a/Documentation/bpf/btf.rst +++ b/Documentation/bpf/btf.rst @@ -86,6 +86,7 @@ sequentially and type id is assigned to each recognized type starting from id #define BTF_KIND_DATASEC 15 /* Section */ #define BTF_KIND_FLOAT 16 /* Floating point */ #define BTF_KIND_DECL_TAG 17 /* Decl Tag */ + #define BTF_KIND_TYPE_TAG 18 /* Type Tag */ Note that the type section encodes debug info, not just pure types. ``BTF_KIND_FUNC`` is not a type, and it represents a defined subprogram. @@ -107,7 +108,7 @@ Each type contains the following common data:: * "size" tells the size of the type it is describing. * * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT, - * FUNC, FUNC_PROTO and DECL_TAG. + * FUNC, FUNC_PROTO, DECL_TAG and TYPE_TAG. * "type" is a type_id referring to another type. */ union { @@ -492,6 +493,16 @@ the attribute is applied to a ``struct``/``union`` member or a ``func`` argument, and ``btf_decl_tag.component_idx`` should be a valid index (starting from 0) pointing to a member or an argument. +2.2.17 BTF_KIND_TYPE_TAG +~~~~~~~~~~~~~~~~~~~~~~~~ + +``struct btf_type`` encoding requirement: + * ``name_off``: offset to a non-empty string + * ``info.kind_flag``: 0 + * ``info.kind``: BTF_KIND_TYPE_TAG + * ``info.vlen``: 0 + * ``type``: the type with ``btf_type_tag`` attribute + 3. BTF Kernel API ***************** From ae80b6033834342601e99f74f6a62ff5092b1cee Mon Sep 17 00:00:00 2001 From: Zekun Shen Date: Thu, 28 Oct 2021 18:37:49 -0400 Subject: [PATCH 0134/2048] ar5523: Fix null-ptr-deref with unexpected WDCMSG_TARGET_START reply Unexpected WDCMSG_TARGET_START replay can lead to null-ptr-deref when ar->tx_cmd->odata is NULL. The patch adds a null check to prevent such case. KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] ar5523_cmd+0x46a/0x581 [ar5523] ar5523_probe.cold+0x1b7/0x18da [ar5523] ? ar5523_cmd_rx_cb+0x7a0/0x7a0 [ar5523] ? __pm_runtime_set_status+0x54a/0x8f0 ? _raw_spin_trylock_bh+0x120/0x120 ? pm_runtime_barrier+0x220/0x220 ? __pm_runtime_resume+0xb1/0xf0 usb_probe_interface+0x25b/0x710 really_probe+0x209/0x5d0 driver_probe_device+0xc6/0x1b0 device_driver_attach+0xe2/0x120 I found the bug using a custome USBFuzz port. It's a research work to fuzz USB stack/drivers. I modified it to fuzz ath9k driver only, providing hand-crafted usb descriptors to QEMU. After fixing the code (fourth byte in usb packet) to WDCMSG_TARGET_START, I got the null-ptr-deref bug. I believe the bug is triggerable whenever cmd->odata is NULL. After patching, I tested with the same input and no longer see the KASAN report. This was NOT tested on a real device. Signed-off-by: Zekun Shen Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/YXsmPQ3awHFLuAj2@10-18-43-117.dynapool.wireless.nyu.edu --- drivers/net/wireless/ath/ar5523/ar5523.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/wireless/ath/ar5523/ar5523.c b/drivers/net/wireless/ath/ar5523/ar5523.c index 0e9bad33fac8..141c1b5a7b1f 100644 --- a/drivers/net/wireless/ath/ar5523/ar5523.c +++ b/drivers/net/wireless/ath/ar5523/ar5523.c @@ -153,6 +153,10 @@ static void ar5523_cmd_rx_cb(struct urb *urb) ar5523_err(ar, "Invalid reply to WDCMSG_TARGET_START"); return; } + if (!cmd->odata) { + ar5523_err(ar, "Unexpected WDCMSG_TARGET_START reply"); + return; + } memcpy(cmd->odata, hdr + 1, sizeof(u32)); cmd->olen = sizeof(u32); cmd->res = 0; From 16a2c3d5406f95ef6139de52669c60a39443f5f7 Mon Sep 17 00:00:00 2001 From: Rameshkumar Sundaram Date: Wed, 10 Nov 2021 17:10:48 +0200 Subject: [PATCH 0135/2048] ath11k: Send PPDU_STATS_CFG with proper pdev mask to firmware HTT_PPDU_STATS_CFG_PDEV_ID bit mask for target FW PPDU stats request message was set as bit 8 to 15. Bit 8 is reserved for soc stats and pdev id starts from bit 9. Hence change the bitmask as bit 9 to 15 and fill the proper pdev id in the request message. In commit 701e48a43e15 ("ath11k: add packet log support for QCA6390"), both HTT_PPDU_STATS_CFG_PDEV_ID and pdev_mask were changed, but this pdev_mask calculation is not valid for platforms which has multiple pdevs with 1 rxdma per pdev, as this is writing same value(i.e. 2) for all pdevs. Hence fixed it to consider pdev_idx as well, to make it compatible for both single and multi pd cases. Tested on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-01092-QCAHKSWPL_SILICONZ-1 Tested on: IPQ6018 hw1.0 WLAN.HK.2.5.0.1-01067-QCAHKSWPL_SILICONZ-1 Fixes: 701e48a43e15 ("ath11k: add packet log support for QCA6390") Co-developed-by: Sathishkumar Muruganandam Signed-off-by: Sathishkumar Muruganandam Signed-off-by: Rameshkumar Sundaram Signed-off-by: Jouni Malinen Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20210721212029.142388-10-jouni@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp.h | 3 ++- drivers/net/wireless/ath/ath11k/dp_tx.c | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp.h b/drivers/net/wireless/ath/ath11k/dp.h index 4794ca04f213..f524d19aca34 100644 --- a/drivers/net/wireless/ath/ath11k/dp.h +++ b/drivers/net/wireless/ath/ath11k/dp.h @@ -517,7 +517,8 @@ struct htt_ppdu_stats_cfg_cmd { } __packed; #define HTT_PPDU_STATS_CFG_MSG_TYPE GENMASK(7, 0) -#define HTT_PPDU_STATS_CFG_PDEV_ID GENMASK(15, 8) +#define HTT_PPDU_STATS_CFG_SOC_STATS BIT(8) +#define HTT_PPDU_STATS_CFG_PDEV_ID GENMASK(15, 9) #define HTT_PPDU_STATS_CFG_TLV_TYPE_BITMASK GENMASK(31, 16) enum htt_ppdu_stats_tag_type { diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c index 879fb2a9dc0c..10b76f6f710b 100644 --- a/drivers/net/wireless/ath/ath11k/dp_tx.c +++ b/drivers/net/wireless/ath/ath11k/dp_tx.c @@ -903,7 +903,7 @@ int ath11k_dp_tx_htt_h2t_ppdu_stats_req(struct ath11k *ar, u32 mask) cmd->msg = FIELD_PREP(HTT_PPDU_STATS_CFG_MSG_TYPE, HTT_H2T_MSG_TYPE_PPDU_STATS_CFG); - pdev_mask = 1 << (i + 1); + pdev_mask = 1 << (ar->pdev_idx + i); cmd->msg |= FIELD_PREP(HTT_PPDU_STATS_CFG_PDEV_ID, pdev_mask); cmd->msg |= FIELD_PREP(HTT_PPDU_STATS_CFG_TLV_TYPE_BITMASK, mask); From c802b6d7815d7c3f556efea28d0b79ef57ebcfd4 Mon Sep 17 00:00:00 2001 From: Rameshkumar Sundaram Date: Mon, 25 Oct 2021 21:33:06 +0530 Subject: [PATCH 0136/2048] ath11k: Clear auth flag only for actual association in security mode AUTH flag is needed when peer assoc command is sent from host in security mode for non-assoc cases. Firmware will handle AUTH flag when client is associating as AUTH flag will be set after key exchange. For internally provided peer assoc commands from host, there won't be any key exchange, so AUTH flag is expected to be set in host. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1-01838-QCAHKSWPL_SILICONZ-1 Co-developed-by: Lavanya Suresh Signed-off-by: Lavanya Suresh Signed-off-by: Rameshkumar Sundaram Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635177786-20854-1-git-send-email-quic_ramess@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 3 +++ drivers/net/wireless/ath/ath11k/wmi.c | 2 +- drivers/net/wireless/ath/ath11k/wmi.h | 1 + 3 files changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 89810b202a39..a62c63ffb3cf 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -2495,6 +2495,7 @@ static void ath11k_bss_assoc(struct ieee80211_hw *hw, rcu_read_unlock(); + peer_arg.is_assoc = true; ret = ath11k_wmi_send_peer_assoc_cmd(ar, &peer_arg); if (ret) { ath11k_warn(ar->ab, "failed to run peer assoc for %pM vdev %i: %d\n", @@ -3772,6 +3773,7 @@ static int ath11k_station_assoc(struct ath11k *ar, ath11k_peer_assoc_prepare(ar, vif, sta, &peer_arg, reassoc); + peer_arg.is_assoc = true; ret = ath11k_wmi_send_peer_assoc_cmd(ar, &peer_arg); if (ret) { ath11k_warn(ar->ab, "failed to run peer assoc for STA %pM vdev %i: %d\n", @@ -3983,6 +3985,7 @@ static void ath11k_sta_rc_update_wk(struct work_struct *wk) ath11k_peer_assoc_prepare(ar, arvif->vif, sta, &peer_arg, true); + peer_arg.is_assoc = false; err = ath11k_wmi_send_peer_assoc_cmd(ar, &peer_arg); if (err) ath11k_warn(ar->ab, "failed to run peer assoc for STA %pM vdev %i: %d\n", diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index b9bfe0407cd7..44057ccd0140 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -1763,7 +1763,7 @@ ath11k_wmi_copy_peer_flags(struct wmi_peer_assoc_complete_cmd *cmd, cmd->peer_flags |= WMI_PEER_AUTH; if (param->need_ptk_4_way) { cmd->peer_flags |= WMI_PEER_NEED_PTK_4_WAY; - if (!hw_crypto_disabled) + if (!hw_crypto_disabled && param->is_assoc) cmd->peer_flags &= ~WMI_PEER_AUTH; } if (param->need_gtk_2_way) diff --git a/drivers/net/wireless/ath/ath11k/wmi.h b/drivers/net/wireless/ath/ath11k/wmi.h index 93e2992db862..81d12405be54 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.h +++ b/drivers/net/wireless/ath/ath11k/wmi.h @@ -3617,6 +3617,7 @@ struct peer_assoc_params { u32 peer_he_tx_mcs_set[WMI_HOST_MAX_HE_RATE_SET]; bool twt_responder; bool twt_requester; + bool is_assoc; struct ath11k_ppe_threshold peer_ppet; }; From 314f14abdeca78de6b16f97d796a9966ce4b90ae Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Wed, 10 Nov 2021 11:23:24 -0800 Subject: [PATCH 0137/2048] bpftool: Enable libbpf's strict mode by default Otherwise, attaching with bpftool doesn't work with strict section names. Also: - Add --legacy option to switch back to pre-1.0 behavior - Print a warning when program fails to load in strict mode to point to --legacy flag - By default, don't append / to the section name; in strict mode it's relevant only for a small subset of prog types + bpftool --legacy prog loadall tools/testing/selftests/bpf/test_cgroup_link.o /sys/fs/bpf/kprobe type kprobe libbpf: failed to pin program: File exists Error: failed to pin all programs + bpftool prog loadall tools/testing/selftests/bpf/test_cgroup_link.o /sys/fs/bpf/kprobe type kprobe v1 -> v2: - strict by default (Quentin Monnet) - add more info to --legacy description (Quentin Monnet) - add bash completion (Quentin Monnet) Signed-off-by: Stanislav Fomichev Signed-off-by: Daniel Borkmann Reviewed-by: Quentin Monnet Link: https://lore.kernel.org/bpf/20211110192324.920934-1-sdf@google.com --- .../bpftool/Documentation/common_options.rst | 9 +++++ tools/bpf/bpftool/bash-completion/bpftool | 2 +- tools/bpf/bpftool/main.c | 13 +++++- tools/bpf/bpftool/main.h | 3 +- tools/bpf/bpftool/prog.c | 40 +++++++++++-------- 5 files changed, 48 insertions(+), 19 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/common_options.rst b/tools/bpf/bpftool/Documentation/common_options.rst index 05d06c74dcaa..75adf23202d8 100644 --- a/tools/bpf/bpftool/Documentation/common_options.rst +++ b/tools/bpf/bpftool/Documentation/common_options.rst @@ -20,3 +20,12 @@ Print all logs available, even debug-level information. This includes logs from libbpf as well as from the verifier, when attempting to load programs. + +-l, --legacy + Use legacy libbpf mode which has more relaxed BPF program + requirements. By default, bpftool has more strict requirements + about section names, changes pinning logic and doesn't support + some of the older non-BTF map declarations. + + See https://github.com/libbpf/libbpf/wiki/Libbpf:-the-road-to-v1.0 + for details. diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool index 88e2bcf16cca..4a1b02ff72c1 100644 --- a/tools/bpf/bpftool/bash-completion/bpftool +++ b/tools/bpf/bpftool/bash-completion/bpftool @@ -261,7 +261,7 @@ _bpftool() # Deal with options if [[ ${words[cword]} == -* ]]; then local c='--version --json --pretty --bpffs --mapcompat --debug \ - --use-loader --base-btf' + --use-loader --base-btf --legacy' COMPREPLY=( $( compgen -W "$c" -- "$cur" ) ) return 0 fi diff --git a/tools/bpf/bpftool/main.c b/tools/bpf/bpftool/main.c index 28237d7cef67..473791e87f7d 100644 --- a/tools/bpf/bpftool/main.c +++ b/tools/bpf/bpftool/main.c @@ -31,6 +31,7 @@ bool block_mount; bool verifier_logs; bool relaxed_maps; bool use_loader; +bool legacy_libbpf; struct btf *base_btf; struct hashmap *refs_table; @@ -396,6 +397,7 @@ int main(int argc, char **argv) { "debug", no_argument, NULL, 'd' }, { "use-loader", no_argument, NULL, 'L' }, { "base-btf", required_argument, NULL, 'B' }, + { "legacy", no_argument, NULL, 'l' }, { 0 } }; int opt, ret; @@ -408,7 +410,7 @@ int main(int argc, char **argv) bin_name = argv[0]; opterr = 0; - while ((opt = getopt_long(argc, argv, "VhpjfLmndB:", + while ((opt = getopt_long(argc, argv, "VhpjfLmndB:l", options, NULL)) >= 0) { switch (opt) { case 'V': @@ -454,6 +456,9 @@ int main(int argc, char **argv) case 'L': use_loader = true; break; + case 'l': + legacy_libbpf = true; + break; default: p_err("unrecognized option '%s'", argv[optind - 1]); if (json_output) @@ -463,6 +468,12 @@ int main(int argc, char **argv) } } + if (!legacy_libbpf) { + ret = libbpf_set_strict_mode(LIBBPF_STRICT_ALL); + if (ret) + p_err("failed to enable libbpf strict mode: %d", ret); + } + argc -= optind; argv += optind; if (argc < 0) diff --git a/tools/bpf/bpftool/main.h b/tools/bpf/bpftool/main.h index 383835c2604d..8d76d937a62b 100644 --- a/tools/bpf/bpftool/main.h +++ b/tools/bpf/bpftool/main.h @@ -57,7 +57,7 @@ static inline void *u64_to_ptr(__u64 ptr) #define HELP_SPEC_PROGRAM \ "PROG := { id PROG_ID | pinned FILE | tag PROG_TAG | name PROG_NAME }" #define HELP_SPEC_OPTIONS \ - "OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug}" + "OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy}" #define HELP_SPEC_MAP \ "MAP := { id MAP_ID | pinned FILE | name MAP_NAME }" #define HELP_SPEC_LINK \ @@ -90,6 +90,7 @@ extern bool block_mount; extern bool verifier_logs; extern bool relaxed_maps; extern bool use_loader; +extern bool legacy_libbpf; extern struct btf *base_btf; extern struct hashmap *refs_table; diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index dea7a49ec26e..bf85c914f2fa 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -1483,8 +1483,6 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) while (argc) { if (is_prefix(*argv, "type")) { - char *type; - NEXT_ARG(); if (common_prog_type != BPF_PROG_TYPE_UNSPEC) { @@ -1494,21 +1492,26 @@ static int load_with_options(int argc, char **argv, bool first_prog_only) if (!REQ_ARGS(1)) goto err_free_reuse_maps; - /* Put a '/' at the end of type to appease libbpf */ - type = malloc(strlen(*argv) + 2); - if (!type) { - p_err("mem alloc failed"); - goto err_free_reuse_maps; - } - *type = 0; - strcat(type, *argv); - strcat(type, "/"); + err = libbpf_prog_type_by_name(*argv, &common_prog_type, + &expected_attach_type); + if (err < 0) { + /* Put a '/' at the end of type to appease libbpf */ + char *type = malloc(strlen(*argv) + 2); - err = get_prog_type_by_name(type, &common_prog_type, - &expected_attach_type); - free(type); - if (err < 0) - goto err_free_reuse_maps; + if (!type) { + p_err("mem alloc failed"); + goto err_free_reuse_maps; + } + *type = 0; + strcat(type, *argv); + strcat(type, "/"); + + err = get_prog_type_by_name(type, &common_prog_type, + &expected_attach_type); + free(type); + if (err < 0) + goto err_free_reuse_maps; + } NEXT_ARG(); } else if (is_prefix(*argv, "map")) { @@ -1731,6 +1734,11 @@ err_unpin: else bpf_object__unpin_programs(obj, pinfile); err_close_obj: + if (!legacy_libbpf) { + p_info("Warning: bpftool is now running in libbpf strict mode and has more stringent requirements about BPF programs.\n" + "If it used to work for this object file but now doesn't, see --legacy option for more details.\n"); + } + bpf_object__close(obj); err_free_reuse_maps: for (i = 0; i < old_map_fds; i++) From 9e2ad638ae3632ef916ceb39f70e3104bf8fdc97 Mon Sep 17 00:00:00 2001 From: Song Liu Date: Fri, 12 Nov 2021 07:02:42 -0800 Subject: [PATCH 0138/2048] bpf: Extend BTF_ID_LIST_GLOBAL with parameter for number of IDs syzbot reported the following BUG w/o CONFIG_DEBUG_INFO_BTF BUG: KASAN: global-out-of-bounds in task_iter_init+0x212/0x2e7 kernel/bpf/task_iter.c:661 Read of size 4 at addr ffffffff90297404 by task swapper/0/1 CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.15.0-syzkaller #0 Hardware name: ... Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106 print_address_description.constprop.0.cold+0xf/0x309 mm/kasan/report.c:256 __kasan_report mm/kasan/report.c:442 [inline] kasan_report.cold+0x83/0xdf mm/kasan/report.c:459 task_iter_init+0x212/0x2e7 kernel/bpf/task_iter.c:661 do_one_initcall+0x103/0x650 init/main.c:1295 do_initcall_level init/main.c:1368 [inline] do_initcalls init/main.c:1384 [inline] do_basic_setup init/main.c:1403 [inline] kernel_init_freeable+0x6b1/0x73a init/main.c:1606 kernel_init+0x1a/0x1d0 init/main.c:1497 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 This is caused by hard-coded name[1] in BTF_ID_LIST_GLOBAL (w/o CONFIG_DEBUG_INFO_BTF). Fix this by adding a parameter n to BTF_ID_LIST_GLOBAL. This avoids ifdef CONFIG_DEBUG_INFO_BTF in btf.c and filter.c. Fixes: 7c7e3d31e785 ("bpf: Introduce helper bpf_find_vma") Reported-by: syzbot+e0d81ec552a21d9071aa@syzkaller.appspotmail.com Reported-by: Eric Dumazet Suggested-by: Eric Dumazet Signed-off-by: Song Liu Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211112150243.1270987-2-songliubraving@fb.com --- include/linux/btf_ids.h | 6 +++--- kernel/bpf/btf.c | 2 +- net/core/filter.c | 6 +----- 3 files changed, 5 insertions(+), 9 deletions(-) diff --git a/include/linux/btf_ids.h b/include/linux/btf_ids.h index 47d9abfbdb55..6bb42b785293 100644 --- a/include/linux/btf_ids.h +++ b/include/linux/btf_ids.h @@ -73,7 +73,7 @@ asm( \ __BTF_ID_LIST(name, local) \ extern u32 name[]; -#define BTF_ID_LIST_GLOBAL(name) \ +#define BTF_ID_LIST_GLOBAL(name, n) \ __BTF_ID_LIST(name, globl) /* The BTF_ID_LIST_SINGLE macro defines a BTF_ID_LIST with @@ -83,7 +83,7 @@ __BTF_ID_LIST(name, globl) BTF_ID_LIST(name) \ BTF_ID(prefix, typename) #define BTF_ID_LIST_GLOBAL_SINGLE(name, prefix, typename) \ - BTF_ID_LIST_GLOBAL(name) \ + BTF_ID_LIST_GLOBAL(name, 1) \ BTF_ID(prefix, typename) /* @@ -149,7 +149,7 @@ extern struct btf_id_set name; #define BTF_ID_LIST(name) static u32 name[5]; #define BTF_ID(prefix, name) #define BTF_ID_UNUSED -#define BTF_ID_LIST_GLOBAL(name) u32 name[1]; +#define BTF_ID_LIST_GLOBAL(name, n) u32 name[n]; #define BTF_ID_LIST_SINGLE(name, prefix, typename) static u32 name[1]; #define BTF_ID_LIST_GLOBAL_SINGLE(name, prefix, typename) u32 name[1]; #define BTF_SET_START(name) static struct btf_id_set name = { 0 }; diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index 1dd9ba82da1e..2a9d8a1fee1d 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -6354,7 +6354,7 @@ const struct bpf_func_proto bpf_btf_find_by_name_kind_proto = { .arg4_type = ARG_ANYTHING, }; -BTF_ID_LIST_GLOBAL(btf_task_struct_ids) +BTF_ID_LIST_GLOBAL(btf_task_struct_ids, 3) BTF_ID(struct, task_struct) BTF_ID(struct, file) BTF_ID(struct, vm_area_struct) diff --git a/net/core/filter.c b/net/core/filter.c index 315a58466fc9..46f09a8fba20 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -10611,14 +10611,10 @@ void bpf_prog_change_xdp(struct bpf_prog *prev_prog, struct bpf_prog *prog) bpf_dispatcher_change_prog(BPF_DISPATCHER_PTR(xdp), prev_prog, prog); } -#ifdef CONFIG_DEBUG_INFO_BTF -BTF_ID_LIST_GLOBAL(btf_sock_ids) +BTF_ID_LIST_GLOBAL(btf_sock_ids, MAX_BTF_SOCK_TYPE) #define BTF_SOCK_TYPE(name, type) BTF_ID(struct, type) BTF_SOCK_TYPE_xxx #undef BTF_SOCK_TYPE -#else -u32 btf_sock_ids[MAX_BTF_SOCK_TYPE]; -#endif BPF_CALL_1(bpf_skc_to_tcp6_sock, struct sock *, sk) { From d19ddb476a539fd78ad1028ae13bb38506286931 Mon Sep 17 00:00:00 2001 From: Song Liu Date: Fri, 12 Nov 2021 07:02:43 -0800 Subject: [PATCH 0139/2048] bpf: Introduce btf_tracing_ids Similar to btf_sock_ids, btf_tracing_ids provides btf ID for task_struct, file, and vm_area_struct via easy to understand format like btf_tracing_ids[BTF_TRACING_TYPE_[TASK|file|VMA]]. Suggested-by: Alexei Starovoitov Signed-off-by: Song Liu Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Link: https://lore.kernel.org/bpf/20211112150243.1270987-3-songliubraving@fb.com --- include/linux/btf_ids.h | 14 +++++++++++++- kernel/bpf/bpf_task_storage.c | 4 ++-- kernel/bpf/btf.c | 8 ++++---- kernel/bpf/stackmap.c | 2 +- kernel/bpf/task_iter.c | 12 ++++++------ kernel/bpf/verifier.c | 2 +- kernel/trace/bpf_trace.c | 4 ++-- 7 files changed, 29 insertions(+), 17 deletions(-) diff --git a/include/linux/btf_ids.h b/include/linux/btf_ids.h index 6bb42b785293..919c0fde1c51 100644 --- a/include/linux/btf_ids.h +++ b/include/linux/btf_ids.h @@ -189,6 +189,18 @@ MAX_BTF_SOCK_TYPE, extern u32 btf_sock_ids[]; #endif -extern u32 btf_task_struct_ids[]; +#define BTF_TRACING_TYPE_xxx \ + BTF_TRACING_TYPE(BTF_TRACING_TYPE_TASK, task_struct) \ + BTF_TRACING_TYPE(BTF_TRACING_TYPE_FILE, file) \ + BTF_TRACING_TYPE(BTF_TRACING_TYPE_VMA, vm_area_struct) + +enum { +#define BTF_TRACING_TYPE(name, type) name, +BTF_TRACING_TYPE_xxx +#undef BTF_TRACING_TYPE +MAX_BTF_TRACING_TYPE, +}; + +extern u32 btf_tracing_ids[]; #endif diff --git a/kernel/bpf/bpf_task_storage.c b/kernel/bpf/bpf_task_storage.c index ebfa8bc90892..bb69aea1a777 100644 --- a/kernel/bpf/bpf_task_storage.c +++ b/kernel/bpf/bpf_task_storage.c @@ -323,7 +323,7 @@ const struct bpf_func_proto bpf_task_storage_get_proto = { .ret_type = RET_PTR_TO_MAP_VALUE_OR_NULL, .arg1_type = ARG_CONST_MAP_PTR, .arg2_type = ARG_PTR_TO_BTF_ID, - .arg2_btf_id = &btf_task_struct_ids[0], + .arg2_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], .arg3_type = ARG_PTR_TO_MAP_VALUE_OR_NULL, .arg4_type = ARG_ANYTHING, }; @@ -334,5 +334,5 @@ const struct bpf_func_proto bpf_task_storage_delete_proto = { .ret_type = RET_INTEGER, .arg1_type = ARG_CONST_MAP_PTR, .arg2_type = ARG_PTR_TO_BTF_ID, - .arg2_btf_id = &btf_task_struct_ids[0], + .arg2_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], }; diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index 2a9d8a1fee1d..6b9d23be1e99 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -6354,10 +6354,10 @@ const struct bpf_func_proto bpf_btf_find_by_name_kind_proto = { .arg4_type = ARG_ANYTHING, }; -BTF_ID_LIST_GLOBAL(btf_task_struct_ids, 3) -BTF_ID(struct, task_struct) -BTF_ID(struct, file) -BTF_ID(struct, vm_area_struct) +BTF_ID_LIST_GLOBAL(btf_tracing_ids, MAX_BTF_TRACING_TYPE) +#define BTF_TRACING_TYPE(name, type) BTF_ID(struct, type) +BTF_TRACING_TYPE_xxx +#undef BTF_TRACING_TYPE /* BTF ID set registration API for modules */ diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index 1de0a1b03636..49e567209c6b 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -489,7 +489,7 @@ const struct bpf_func_proto bpf_get_task_stack_proto = { .gpl_only = false, .ret_type = RET_INTEGER, .arg1_type = ARG_PTR_TO_BTF_ID, - .arg1_btf_id = &btf_task_struct_ids[0], + .arg1_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], .arg2_type = ARG_PTR_TO_UNINIT_MEM, .arg3_type = ARG_CONST_SIZE_OR_ZERO, .arg4_type = ARG_ANYTHING, diff --git a/kernel/bpf/task_iter.c b/kernel/bpf/task_iter.c index f171479f7dd6..d94696198ef8 100644 --- a/kernel/bpf/task_iter.c +++ b/kernel/bpf/task_iter.c @@ -622,7 +622,7 @@ const struct bpf_func_proto bpf_find_vma_proto = { .func = bpf_find_vma, .ret_type = RET_INTEGER, .arg1_type = ARG_PTR_TO_BTF_ID, - .arg1_btf_id = &btf_task_struct_ids[0], + .arg1_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], .arg2_type = ARG_ANYTHING, .arg3_type = ARG_PTR_TO_FUNC, .arg4_type = ARG_PTR_TO_STACK_OR_NULL, @@ -652,19 +652,19 @@ static int __init task_iter_init(void) init_irq_work(&work->irq_work, do_mmap_read_unlock); } - task_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; + task_reg_info.ctx_arg_info[0].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_TASK]; ret = bpf_iter_reg_target(&task_reg_info); if (ret) return ret; - task_file_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; - task_file_reg_info.ctx_arg_info[1].btf_id = btf_task_struct_ids[1]; + task_file_reg_info.ctx_arg_info[0].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_TASK]; + task_file_reg_info.ctx_arg_info[1].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_FILE]; ret = bpf_iter_reg_target(&task_file_reg_info); if (ret) return ret; - task_vma_reg_info.ctx_arg_info[0].btf_id = btf_task_struct_ids[0]; - task_vma_reg_info.ctx_arg_info[1].btf_id = btf_task_struct_ids[2]; + task_vma_reg_info.ctx_arg_info[0].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_TASK]; + task_vma_reg_info.ctx_arg_info[1].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_VMA]; return bpf_iter_reg_target(&task_vma_reg_info); } late_initcall(task_iter_init); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 1aafb43f61d1..d31a031ab377 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6147,7 +6147,7 @@ static int set_find_vma_callback_state(struct bpf_verifier_env *env, callee->regs[BPF_REG_2].type = PTR_TO_BTF_ID; __mark_reg_known_zero(&callee->regs[BPF_REG_2]); callee->regs[BPF_REG_2].btf = btf_vmlinux; - callee->regs[BPF_REG_2].btf_id = btf_task_struct_ids[2]; + callee->regs[BPF_REG_2].btf_id = btf_tracing_ids[BTF_TRACING_TYPE_VMA], /* pointer to stack or null */ callee->regs[BPF_REG_3] = caller->regs[BPF_REG_4]; diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 390176a3031a..25ea521fb8f1 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -764,7 +764,7 @@ const struct bpf_func_proto bpf_get_current_task_btf_proto = { .func = bpf_get_current_task_btf, .gpl_only = true, .ret_type = RET_PTR_TO_BTF_ID, - .ret_btf_id = &btf_task_struct_ids[0], + .ret_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], }; BPF_CALL_1(bpf_task_pt_regs, struct task_struct *, task) @@ -779,7 +779,7 @@ const struct bpf_func_proto bpf_task_pt_regs_proto = { .func = bpf_task_pt_regs, .gpl_only = true, .arg1_type = ARG_PTR_TO_BTF_ID, - .arg1_btf_id = &btf_task_struct_ids[0], + .arg1_btf_id = &btf_tracing_ids[BTF_TRACING_TYPE_TASK], .ret_type = RET_PTR_TO_BTF_ID, .ret_btf_id = &bpf_task_pt_regs_ids[0], }; From 21c6ec3d5275a77348b1af0e78cbbed0ee1558d4 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Fri, 12 Nov 2021 12:48:33 -0800 Subject: [PATCH 0140/2048] selftests/bpf: Fix an unused-but-set-variable compiler warning When using clang to build selftests with LLVM=1 in make commandline, I hit the following compiler warning: xdpxceiver.c:747:6: warning: variable 'total' set but not used [-Wunused-but-set-variable] u32 total = 0; ^ This patch fixed the issue by removing that declaration and its assocatied unused operation. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211112204833.3579457-1-yhs@fb.com --- tools/testing/selftests/bpf/xdpxceiver.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/tools/testing/selftests/bpf/xdpxceiver.c b/tools/testing/selftests/bpf/xdpxceiver.c index 6c7cf8aadc79..fe7f423b8c3f 100644 --- a/tools/testing/selftests/bpf/xdpxceiver.c +++ b/tools/testing/selftests/bpf/xdpxceiver.c @@ -744,7 +744,6 @@ static void receive_pkts(struct pkt_stream *pkt_stream, struct xsk_socket_info * struct pkt *pkt = pkt_stream_get_next_rx_pkt(pkt_stream); struct xsk_umem_info *umem = xsk->umem; u32 idx_rx = 0, idx_fq = 0, rcvd, i; - u32 total = 0; int ret; while (pkt) { @@ -799,7 +798,6 @@ static void receive_pkts(struct pkt_stream *pkt_stream, struct xsk_socket_info * pthread_mutex_lock(&pacing_mutex); pkts_in_flight -= rcvd; - total += rcvd; if (pkts_in_flight < umem->num_frames) pthread_cond_signal(&pacing_cond); pthread_mutex_unlock(&pacing_mutex); From 325d956d67178af92b5b12ff950a2f93a433f2c4 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Fri, 12 Nov 2021 12:48:38 -0800 Subject: [PATCH 0141/2048] selftests/bpf: Fix a tautological-constant-out-of-range-compare compiler warning When using clang to build selftests with LLVM=1 in make commandline, I hit the following compiler warning: benchs/bench_bloom_filter_map.c:84:46: warning: result of comparison of constant 256 with expression of type '__u8' (aka 'unsigned char') is always false [-Wtautological-constant-out-of-range-compare] if (args.value_size < 2 || args.value_size > 256) { ~~~~~~~~~~~~~~~ ^ ~~~ The reason is arg.vaue_size has type __u8, so comparison "args.value_size > 256" is always false. This patch fixed the issue by doing proper comparison before assigning the value to args.value_size. The patch also fixed the same issue in two other places. Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20211112204838.3579953-1-yhs@fb.com --- .../bpf/benchs/bench_bloom_filter_map.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/tools/testing/selftests/bpf/benchs/bench_bloom_filter_map.c b/tools/testing/selftests/bpf/benchs/bench_bloom_filter_map.c index 6eeeed2913e6..5bcb8a8cdeb2 100644 --- a/tools/testing/selftests/bpf/benchs/bench_bloom_filter_map.c +++ b/tools/testing/selftests/bpf/benchs/bench_bloom_filter_map.c @@ -63,29 +63,34 @@ static const struct argp_option opts[] = { static error_t parse_arg(int key, char *arg, struct argp_state *state) { + long ret; + switch (key) { case ARG_NR_ENTRIES: - args.nr_entries = strtol(arg, NULL, 10); - if (args.nr_entries == 0) { + ret = strtol(arg, NULL, 10); + if (ret < 1 || ret > UINT_MAX) { fprintf(stderr, "Invalid nr_entries count."); argp_usage(state); } + args.nr_entries = ret; break; case ARG_NR_HASH_FUNCS: - args.nr_hash_funcs = strtol(arg, NULL, 10); - if (args.nr_hash_funcs == 0 || args.nr_hash_funcs > 15) { + ret = strtol(arg, NULL, 10); + if (ret < 1 || ret > 15) { fprintf(stderr, "The bloom filter must use 1 to 15 hash functions."); argp_usage(state); } + args.nr_hash_funcs = ret; break; case ARG_VALUE_SIZE: - args.value_size = strtol(arg, NULL, 10); - if (args.value_size < 2 || args.value_size > 256) { + ret = strtol(arg, NULL, 10); + if (ret < 2 || ret > 256) { fprintf(stderr, "Invalid value size. Must be between 2 and 256 bytes"); argp_usage(state); } + args.value_size = ret; break; default: return ARGP_ERR_UNKNOWN; From ef14102914f3f6a86d105e9fc2941d97de69c389 Mon Sep 17 00:00:00 2001 From: luo penghao Date: Thu, 11 Nov 2021 09:18:09 +0000 Subject: [PATCH 0142/2048] ipv4: drop unused assignment The assignment in the if statement will be overwritten by the following statement Reported-by: Zeal Robot Signed-off-by: luo penghao Signed-off-by: David S. Miller --- net/ipv4/igmp.c | 1 - 1 file changed, 1 deletion(-) diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index d2e2b3d18c66..2ad3c7b42d6d 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -2558,7 +2558,6 @@ int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf, msf->imsf_fmode = pmc->sfmode; psl = rtnl_dereference(pmc->sflist); if (!psl) { - len = 0; count = 0; } else { count = psl->sl_count; From 0de3521500cfdc91f49441a054ed892fc408d73f Mon Sep 17 00:00:00 2001 From: luo penghao Date: Thu, 11 Nov 2021 09:20:47 +0000 Subject: [PATCH 0143/2048] ipv4: Remove duplicate assignments there is a same action when the variable is initialized Reported-by: Zeal Robot Signed-off-by: luo penghao Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 1 - 1 file changed, 1 deletion(-) diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 9bca57ef8b83..57c1d8431386 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -672,7 +672,6 @@ struct sk_buff *ip_frag_next(struct sk_buff *skb, struct ip_frag_state *state) struct sk_buff *skb2; struct iphdr *iph; - len = state->left; /* IF: it doesn't fit, use 'mtu' - the data space left */ if (len > state->mtu) len = state->mtu; From 1274a4eb318debe33e395bb8bcf8b98e6eb4670f Mon Sep 17 00:00:00 2001 From: luo penghao Date: Thu, 11 Nov 2021 15:09:24 +0000 Subject: [PATCH 0144/2048] ipv6: Remove duplicate statements This statement is repeated with the initialization statement Reported-by: Zeal Robot Signed-off-by: luo penghao Signed-off-by: David S. Miller --- net/ipv6/exthdrs.c | 1 - 1 file changed, 1 deletion(-) diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c index 38ece3b7b839..77e34aec7e82 100644 --- a/net/ipv6/exthdrs.c +++ b/net/ipv6/exthdrs.c @@ -686,7 +686,6 @@ static int ipv6_rthdr_rcv(struct sk_buff *skb) struct net *net = dev_net(skb->dev); int accept_source_route = net->ipv6.devconf_all->accept_source_route; - idev = __in6_dev_get(skb->dev); if (idev && accept_source_route > idev->cnf.accept_source_route) accept_source_route = idev->cnf.accept_source_route; From ebbd7f64a3fbe9e0f235e39fc244ee9735e2a52a Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Nov 2021 11:46:27 +0000 Subject: [PATCH 0145/2048] bpftool: Fix memory leak in prog_dump() Following the extraction of prog_dump() from do_dump(), the struct btf allocated in prog_dump() is no longer freed on error; the struct bpf_prog_linfo is not freed at all. Make sure we release them before exiting the function. Fixes: ec2025095cf6 ("bpftool: Match several programs with same tag") Signed-off-by: Quentin Monnet Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211110114632.24537-2-quentin@isovalent.com --- tools/bpf/bpftool/prog.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index bf85c914f2fa..e47e8b06cc3d 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -709,8 +709,8 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, char func_sig[1024]; unsigned char *buf; __u32 member_len; + int fd, err = -1; ssize_t n; - int fd; if (mode == DUMP_JITED) { if (info->jited_prog_len == 0 || !info->jited_prog_insns) { @@ -749,7 +749,7 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, if (fd < 0) { p_err("can't open file %s: %s", filepath, strerror(errno)); - return -1; + goto exit_free; } n = write(fd, buf, member_len); @@ -757,7 +757,7 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, if (n != (ssize_t)member_len) { p_err("error writing output file: %s", n < 0 ? strerror(errno) : "short write"); - return -1; + goto exit_free; } if (json_output) @@ -771,7 +771,7 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, info->netns_ino, &disasm_opt); if (!name) - return -1; + goto exit_free; } if (info->nr_jited_func_lens && info->jited_func_lens) { @@ -866,9 +866,12 @@ prog_dump(struct bpf_prog_info *info, enum dump_mode mode, kernel_syms_destroy(&dd); } - btf__free(btf); + err = 0; - return 0; +exit_free: + btf__free(btf); + bpf_prog_linfo__free(prog_linfo); + return err; } static int do_dump(int argc, char **argv) From 48f5aef4c458c19ab337eed8c95a6486cc014aa3 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Nov 2021 11:46:28 +0000 Subject: [PATCH 0146/2048] bpftool: Remove inclusion of utilities.mak from Makefiles Bpftool's Makefile, and the Makefile for its documentation, both include scripts/utilities.mak, but they use none of the items defined in this file. Remove the includes. Fixes: 71bb428fe2c1 ("tools: bpf: add bpftool") Signed-off-by: Quentin Monnet Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211110114632.24537-3-quentin@isovalent.com --- tools/bpf/bpftool/Documentation/Makefile | 1 - tools/bpf/bpftool/Makefile | 1 - 2 files changed, 2 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/Makefile b/tools/bpf/bpftool/Documentation/Makefile index 44b60784847b..692e1b947490 100644 --- a/tools/bpf/bpftool/Documentation/Makefile +++ b/tools/bpf/bpftool/Documentation/Makefile @@ -1,6 +1,5 @@ # SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) include ../../../scripts/Makefile.include -include ../../../scripts/utilities.mak INSTALL ?= install RM ?= rm -f diff --git a/tools/bpf/bpftool/Makefile b/tools/bpf/bpftool/Makefile index 31dfef6a4121..1b36c11a2a26 100644 --- a/tools/bpf/bpftool/Makefile +++ b/tools/bpf/bpftool/Makefile @@ -1,6 +1,5 @@ # SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) include ../../scripts/Makefile.include -include ../../scripts/utilities.mak ifeq ($(srctree),) srctree := $(patsubst %/,%,$(dir $(CURDIR))) From 986dec18bbf41f50edc2e0aa4ac5ef8e0f64f328 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Nov 2021 11:46:30 +0000 Subject: [PATCH 0147/2048] bpftool: Fix indent in option lists in the documentation Mixed indentation levels in the lists of options in bpftool's documentation produces some unexpected results. For the "bpftool" man page, it prints a warning: $ make -C bpftool.8 GEN bpftool.8 :26: (ERROR/3) Unexpected indentation. For other pages, there is no warning, but it results in a line break appearing in the option lists in the generated man pages. RST paragraphs should have a uniform indentation level. Let's fix it. Fixes: c07ba629df97 ("tools: bpftool: Update and synchronise option list in doc and help msg") Fixes: 8cc8c6357c8f ("tools: bpftool: Document and add bash completion for -L, -B options") Signed-off-by: Quentin Monnet Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211110114632.24537-5-quentin@isovalent.com --- tools/bpf/bpftool/Documentation/bpftool-btf.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-cgroup.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-gen.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-link.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-map.rst | 6 +++--- tools/bpf/bpftool/Documentation/bpftool-prog.rst | 8 ++++---- tools/bpf/bpftool/Documentation/bpftool.rst | 6 +++--- 7 files changed, 14 insertions(+), 14 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-btf.rst b/tools/bpf/bpftool/Documentation/bpftool-btf.rst index 88b28aa7431f..4425d942dd39 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-btf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-btf.rst @@ -13,7 +13,7 @@ SYNOPSIS **bpftool** [*OPTIONS*] **btf** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | {**-d** | **--debug** } | - { **-B** | **--base-btf** } } + { **-B** | **--base-btf** } } *COMMANDS* := { **dump** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index 3e4395eede4f..13a217a2503d 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -13,7 +13,7 @@ SYNOPSIS **bpftool** [*OPTIONS*] **cgroup** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } } + { **-f** | **--bpffs** } } *COMMANDS* := { **show** | **list** | **tree** | **attach** | **detach** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-gen.rst b/tools/bpf/bpftool/Documentation/bpftool-gen.rst index 2ef2f2df0279..2a137f8a4cea 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-gen.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-gen.rst @@ -13,7 +13,7 @@ SYNOPSIS **bpftool** [*OPTIONS*] **gen** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-L** | **--use-loader** } } + { **-L** | **--use-loader** } } *COMMAND* := { **object** | **skeleton** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-link.rst b/tools/bpf/bpftool/Documentation/bpftool-link.rst index 0de90f086238..9434349636a5 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-link.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-link.rst @@ -13,7 +13,7 @@ SYNOPSIS **bpftool** [*OPTIONS*] **link** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } | { **-n** | **--nomount** } } + { **-f** | **--bpffs** } | { **-n** | **--nomount** } } *COMMANDS* := { **show** | **list** | **pin** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-map.rst b/tools/bpf/bpftool/Documentation/bpftool-map.rst index d0c4abe08aba..1445cadc15d4 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-map.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-map.rst @@ -13,11 +13,11 @@ SYNOPSIS **bpftool** [*OPTIONS*] **map** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } | { **-n** | **--nomount** } } + { **-f** | **--bpffs** } | { **-n** | **--nomount** } } *COMMANDS* := - { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** - | **delete** | **pin** | **help** } + { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** | + **delete** | **pin** | **help** } MAP COMMANDS ============= diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index 91608cb7e44a..f27265bd589b 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -13,12 +13,12 @@ SYNOPSIS **bpftool** [*OPTIONS*] **prog** *COMMAND* *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } | { **-m** | **--mapcompat** } | { **-n** | **--nomount** } | - { **-L** | **--use-loader** } } + { **-f** | **--bpffs** } | { **-m** | **--mapcompat** } | { **-n** | **--nomount** } | + { **-L** | **--use-loader** } } *COMMANDS* := - { **show** | **list** | **dump xlated** | **dump jited** | **pin** | **load** - | **loadall** | **help** } + { **show** | **list** | **dump xlated** | **dump jited** | **pin** | **load** | + **loadall** | **help** } PROG COMMANDS ============= diff --git a/tools/bpf/bpftool/Documentation/bpftool.rst b/tools/bpf/bpftool/Documentation/bpftool.rst index bb23f55bb05a..8ac86565c501 100644 --- a/tools/bpf/bpftool/Documentation/bpftool.rst +++ b/tools/bpf/bpftool/Documentation/bpftool.rst @@ -19,14 +19,14 @@ SYNOPSIS *OBJECT* := { **map** | **program** | **cgroup** | **perf** | **net** | **feature** } *OPTIONS* := { { **-V** | **--version** } | - { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } *MAP-COMMANDS* := { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** | - **delete** | **pin** | **event_pipe** | **help** } + **delete** | **pin** | **event_pipe** | **help** } *PROG-COMMANDS* := { **show** | **list** | **dump jited** | **dump xlated** | **pin** | - **load** | **attach** | **detach** | **help** } + **load** | **attach** | **detach** | **help** } *CGROUP-COMMANDS* := { **show** | **list** | **attach** | **detach** | **help** } From 3811e2753a39efb8aa5b8c133dc24f6d26f6cd96 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Nov 2021 11:46:31 +0000 Subject: [PATCH 0148/2048] bpftool: Update the lists of names for maps and prog-attach types To support the different BPF map or attach types, bpftool must remain up-to-date with the types supported by the kernel. Let's update the lists, by adding the missing Bloom filter map type and the perf_event attach type. Both missing items were found with test_bpftool_synctypes.py. Signed-off-by: Quentin Monnet Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211110114632.24537-6-quentin@isovalent.com --- tools/bpf/bpftool/Documentation/bpftool-map.rst | 2 +- tools/bpf/bpftool/bash-completion/bpftool | 3 ++- tools/bpf/bpftool/common.c | 1 + tools/bpf/bpftool/map.c | 3 ++- 4 files changed, 6 insertions(+), 3 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-map.rst b/tools/bpf/bpftool/Documentation/bpftool-map.rst index 1445cadc15d4..991d18fd84f2 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-map.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-map.rst @@ -52,7 +52,7 @@ MAP COMMANDS | | **devmap** | **devmap_hash** | **sockmap** | **cpumap** | **xskmap** | **sockhash** | | **cgroup_storage** | **reuseport_sockarray** | **percpu_cgroup_storage** | | **queue** | **stack** | **sk_storage** | **struct_ops** | **ringbuf** | **inode_storage** - | **task_storage** } +| | **task_storage** | **bloom_filter** } DESCRIPTION =========== diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool index 4a1b02ff72c1..493753a4962e 100644 --- a/tools/bpf/bpftool/bash-completion/bpftool +++ b/tools/bpf/bpftool/bash-completion/bpftool @@ -710,7 +710,8 @@ _bpftool() hash_of_maps devmap devmap_hash sockmap cpumap \ xskmap sockhash cgroup_storage reuseport_sockarray \ percpu_cgroup_storage queue stack sk_storage \ - struct_ops inode_storage task_storage ringbuf' + struct_ops ringbuf inode_storage task_storage \ + bloom_filter' COMPREPLY=( $( compgen -W "$BPFTOOL_MAP_CREATE_TYPES" -- "$cur" ) ) return 0 ;; diff --git a/tools/bpf/bpftool/common.c b/tools/bpf/bpftool/common.c index 511eccdbdfe6..fa8eb8134344 100644 --- a/tools/bpf/bpftool/common.c +++ b/tools/bpf/bpftool/common.c @@ -74,6 +74,7 @@ const char * const attach_type_name[__MAX_BPF_ATTACH_TYPE] = { [BPF_XDP] = "xdp", [BPF_SK_REUSEPORT_SELECT] = "sk_skb_reuseport_select", [BPF_SK_REUSEPORT_SELECT_OR_MIGRATE] = "sk_skb_reuseport_select_or_migrate", + [BPF_PERF_EVENT] = "perf_event", }; void p_err(const char *fmt, ...) diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index cae1f1119296..68cb121e65c4 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -53,6 +53,7 @@ const char * const map_type_name[] = { [BPF_MAP_TYPE_RINGBUF] = "ringbuf", [BPF_MAP_TYPE_INODE_STORAGE] = "inode_storage", [BPF_MAP_TYPE_TASK_STORAGE] = "task_storage", + [BPF_MAP_TYPE_BLOOM_FILTER] = "bloom_filter", }; const size_t map_type_name_size = ARRAY_SIZE(map_type_name); @@ -1477,7 +1478,7 @@ static int do_help(int argc, char **argv) " devmap | devmap_hash | sockmap | cpumap | xskmap | sockhash |\n" " cgroup_storage | reuseport_sockarray | percpu_cgroup_storage |\n" " queue | stack | sk_storage | struct_ops | ringbuf | inode_storage |\n" - " task_storage }\n" + " task_storage | bloom_filter }\n" " " HELP_SPEC_OPTIONS " |\n" " {-f|--bpffs} | {-n|--nomount} }\n" "", From b06be5651f08bc5bc305e2a3c722ddb33b783ee5 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Nov 2021 11:46:32 +0000 Subject: [PATCH 0149/2048] bpftool: Fix mixed indentation in documentation Some paragraphs in bpftool's documentation have a mix of tabs and spaces for indentation. Let's make it consistent. This patch brings no change to the text content. Signed-off-by: Quentin Monnet Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211110114632.24537-7-quentin@isovalent.com --- .../bpftool/Documentation/bpftool-cgroup.rst | 10 +-- .../bpf/bpftool/Documentation/bpftool-net.rst | 62 +++++++++---------- 2 files changed, 36 insertions(+), 36 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index 13a217a2503d..8069d37dd991 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -30,9 +30,9 @@ CGROUP COMMANDS | *PROG* := { **id** *PROG_ID* | **pinned** *FILE* | **tag** *PROG_TAG* } | *ATTACH_TYPE* := { **ingress** | **egress** | **sock_create** | **sock_ops** | **device** | | **bind4** | **bind6** | **post_bind4** | **post_bind6** | **connect4** | **connect6** | -| **getpeername4** | **getpeername6** | **getsockname4** | **getsockname6** | **sendmsg4** | -| **sendmsg6** | **recvmsg4** | **recvmsg6** | **sysctl** | **getsockopt** | **setsockopt** | -| **sock_release** } +| **getpeername4** | **getpeername6** | **getsockname4** | **getsockname6** | **sendmsg4** | +| **sendmsg6** | **recvmsg4** | **recvmsg6** | **sysctl** | **getsockopt** | **setsockopt** | +| **sock_release** } | *ATTACH_FLAGS* := { **multi** | **override** } DESCRIPTION @@ -98,9 +98,9 @@ DESCRIPTION **sendmsg6** call to sendto(2), sendmsg(2), sendmmsg(2) for an unconnected udp6 socket (since 4.18); **recvmsg4** call to recvfrom(2), recvmsg(2), recvmmsg(2) for - an unconnected udp4 socket (since 5.2); + an unconnected udp4 socket (since 5.2); **recvmsg6** call to recvfrom(2), recvmsg(2), recvmmsg(2) for - an unconnected udp6 socket (since 5.2); + an unconnected udp6 socket (since 5.2); **sysctl** sysctl access (since 5.2); **getsockopt** call to getsockopt (since 5.3); **setsockopt** call to setsockopt (since 5.3); diff --git a/tools/bpf/bpftool/Documentation/bpftool-net.rst b/tools/bpf/bpftool/Documentation/bpftool-net.rst index 1ae0375e8fea..7ec57535a7c1 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-net.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-net.rst @@ -31,44 +31,44 @@ NET COMMANDS DESCRIPTION =========== **bpftool net { show | list }** [ **dev** *NAME* ] - List bpf program attachments in the kernel networking subsystem. + List bpf program attachments in the kernel networking subsystem. - Currently, only device driver xdp attachments and tc filter - classification/action attachments are implemented, i.e., for - program types **BPF_PROG_TYPE_SCHED_CLS**, - **BPF_PROG_TYPE_SCHED_ACT** and **BPF_PROG_TYPE_XDP**. - For programs attached to a particular cgroup, e.g., - **BPF_PROG_TYPE_CGROUP_SKB**, **BPF_PROG_TYPE_CGROUP_SOCK**, - **BPF_PROG_TYPE_SOCK_OPS** and **BPF_PROG_TYPE_CGROUP_SOCK_ADDR**, - users can use **bpftool cgroup** to dump cgroup attachments. - For sk_{filter, skb, msg, reuseport} and lwt/seg6 - bpf programs, users should consult other tools, e.g., iproute2. + Currently, only device driver xdp attachments and tc filter + classification/action attachments are implemented, i.e., for + program types **BPF_PROG_TYPE_SCHED_CLS**, + **BPF_PROG_TYPE_SCHED_ACT** and **BPF_PROG_TYPE_XDP**. + For programs attached to a particular cgroup, e.g., + **BPF_PROG_TYPE_CGROUP_SKB**, **BPF_PROG_TYPE_CGROUP_SOCK**, + **BPF_PROG_TYPE_SOCK_OPS** and **BPF_PROG_TYPE_CGROUP_SOCK_ADDR**, + users can use **bpftool cgroup** to dump cgroup attachments. + For sk_{filter, skb, msg, reuseport} and lwt/seg6 + bpf programs, users should consult other tools, e.g., iproute2. - The current output will start with all xdp program attachments, followed by - all tc class/qdisc bpf program attachments. Both xdp programs and - tc programs are ordered based on ifindex number. If multiple bpf - programs attached to the same networking device through **tc filter**, - the order will be first all bpf programs attached to tc classes, then - all bpf programs attached to non clsact qdiscs, and finally all - bpf programs attached to root and clsact qdisc. + The current output will start with all xdp program attachments, followed by + all tc class/qdisc bpf program attachments. Both xdp programs and + tc programs are ordered based on ifindex number. If multiple bpf + programs attached to the same networking device through **tc filter**, + the order will be first all bpf programs attached to tc classes, then + all bpf programs attached to non clsact qdiscs, and finally all + bpf programs attached to root and clsact qdisc. **bpftool** **net attach** *ATTACH_TYPE* *PROG* **dev** *NAME* [ **overwrite** ] - Attach bpf program *PROG* to network interface *NAME* with - type specified by *ATTACH_TYPE*. Previously attached bpf program - can be replaced by the command used with **overwrite** option. - Currently, only XDP-related modes are supported for *ATTACH_TYPE*. + Attach bpf program *PROG* to network interface *NAME* with + type specified by *ATTACH_TYPE*. Previously attached bpf program + can be replaced by the command used with **overwrite** option. + Currently, only XDP-related modes are supported for *ATTACH_TYPE*. - *ATTACH_TYPE* can be of: - **xdp** - try native XDP and fallback to generic XDP if NIC driver does not support it; - **xdpgeneric** - Generic XDP. runs at generic XDP hook when packet already enters receive path as skb; - **xdpdrv** - Native XDP. runs earliest point in driver's receive path; - **xdpoffload** - Offload XDP. runs directly on NIC on each packet reception; + *ATTACH_TYPE* can be of: + **xdp** - try native XDP and fallback to generic XDP if NIC driver does not support it; + **xdpgeneric** - Generic XDP. runs at generic XDP hook when packet already enters receive path as skb; + **xdpdrv** - Native XDP. runs earliest point in driver's receive path; + **xdpoffload** - Offload XDP. runs directly on NIC on each packet reception; **bpftool** **net detach** *ATTACH_TYPE* **dev** *NAME* - Detach bpf program attached to network interface *NAME* with - type specified by *ATTACH_TYPE*. To detach bpf program, same - *ATTACH_TYPE* previously used for attach must be specified. - Currently, only XDP-related modes are supported for *ATTACH_TYPE*. + Detach bpf program attached to network interface *NAME* with + type specified by *ATTACH_TYPE*. To detach bpf program, same + *ATTACH_TYPE* previously used for attach must be specified. + Currently, only XDP-related modes are supported for *ATTACH_TYPE*. **bpftool net help** Print short help message. From e5043894b21f7d99d3db31ad06308d6c5726caa6 Mon Sep 17 00:00:00 2001 From: Hengqi Chen Date: Mon, 15 Nov 2021 09:24:36 +0800 Subject: [PATCH 0150/2048] bpftool: Use libbpf_get_error() to check error Currently, LIBBPF_STRICT_ALL mode is enabled by default for bpftool which means on error cases, some libbpf APIs would return NULL pointers. This makes IS_ERR check failed to detect such cases and result in segfault error. Use libbpf_get_error() instead like we do in libbpf itself. Signed-off-by: Hengqi Chen Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211115012436.3143318-1-hengqi.chen@gmail.com --- tools/bpf/bpftool/btf.c | 9 +++++---- tools/bpf/bpftool/gen.c | 10 ++++++---- tools/bpf/bpftool/iter.c | 7 ++++--- tools/bpf/bpftool/map.c | 10 +++++----- tools/bpf/bpftool/struct_ops.c | 14 +++++++------- 5 files changed, 27 insertions(+), 23 deletions(-) diff --git a/tools/bpf/bpftool/btf.c b/tools/bpf/bpftool/btf.c index c7e3b0b0029e..59833125ac0a 100644 --- a/tools/bpf/bpftool/btf.c +++ b/tools/bpf/bpftool/btf.c @@ -421,8 +421,9 @@ static int dump_btf_c(const struct btf *btf, int err = 0, i; d = btf_dump__new(btf, btf_dump_printf, NULL, NULL); - if (IS_ERR(d)) - return PTR_ERR(d); + err = libbpf_get_error(d); + if (err) + return err; printf("#ifndef __VMLINUX_H__\n"); printf("#define __VMLINUX_H__\n"); @@ -549,8 +550,8 @@ static int do_dump(int argc, char **argv) } btf = btf__parse_split(*argv, base ?: base_btf); - if (IS_ERR(btf)) { - err = -PTR_ERR(btf); + err = libbpf_get_error(btf); + if (err) { btf = NULL; p_err("failed to load BTF from %s: %s", *argv, strerror(err)); diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c index 89f0e828bbfa..997a2865e04a 100644 --- a/tools/bpf/bpftool/gen.c +++ b/tools/bpf/bpftool/gen.c @@ -219,8 +219,9 @@ static int codegen_datasecs(struct bpf_object *obj, const char *obj_name) int i, err = 0; d = btf_dump__new(btf, codegen_btf_dump_printf, NULL, NULL); - if (IS_ERR(d)) - return PTR_ERR(d); + err = libbpf_get_error(d); + if (err) + return err; bpf_object__for_each_map(map, obj) { /* only generate definitions for memory-mapped internal maps */ @@ -719,10 +720,11 @@ static int do_skeleton(int argc, char **argv) get_obj_name(obj_name, file); opts.object_name = obj_name; obj = bpf_object__open_mem(obj_data, file_sz, &opts); - if (IS_ERR(obj)) { + err = libbpf_get_error(obj); + if (err) { char err_buf[256]; - libbpf_strerror(PTR_ERR(obj), err_buf, sizeof(err_buf)); + libbpf_strerror(err, err_buf, sizeof(err_buf)); p_err("failed to open BPF object file: %s", err_buf); obj = NULL; goto out; diff --git a/tools/bpf/bpftool/iter.c b/tools/bpf/bpftool/iter.c index 6c0de647b8ad..f88fdc820d23 100644 --- a/tools/bpf/bpftool/iter.c +++ b/tools/bpf/bpftool/iter.c @@ -46,7 +46,8 @@ static int do_pin(int argc, char **argv) } obj = bpf_object__open(objfile); - if (IS_ERR(obj)) { + err = libbpf_get_error(obj); + if (err) { p_err("can't open objfile %s", objfile); goto close_map_fd; } @@ -64,8 +65,8 @@ static int do_pin(int argc, char **argv) } link = bpf_program__attach_iter(prog, &iter_opts); - if (IS_ERR(link)) { - err = PTR_ERR(link); + err = libbpf_get_error(link); + if (err) { p_err("attach_iter failed for program %s", bpf_program__name(prog)); goto close_obj; diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index 68cb121e65c4..25b258804f11 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -812,7 +812,7 @@ static struct btf *get_map_kv_btf(const struct bpf_map_info *info) if (info->btf_vmlinux_value_type_id) { if (!btf_vmlinux) { btf_vmlinux = libbpf_find_kernel_btf(); - if (IS_ERR(btf_vmlinux)) + if (libbpf_get_error(btf_vmlinux)) p_err("failed to get kernel btf"); } return btf_vmlinux; @@ -832,13 +832,13 @@ static struct btf *get_map_kv_btf(const struct bpf_map_info *info) static void free_map_kv_btf(struct btf *btf) { - if (!IS_ERR(btf) && btf != btf_vmlinux) + if (!libbpf_get_error(btf) && btf != btf_vmlinux) btf__free(btf); } static void free_btf_vmlinux(void) { - if (!IS_ERR(btf_vmlinux)) + if (!libbpf_get_error(btf_vmlinux)) btf__free(btf_vmlinux); } @@ -863,8 +863,8 @@ map_dump(int fd, struct bpf_map_info *info, json_writer_t *wtr, if (wtr) { btf = get_map_kv_btf(info); - if (IS_ERR(btf)) { - err = PTR_ERR(btf); + err = libbpf_get_error(btf); + if (err) { goto exit_free; } diff --git a/tools/bpf/bpftool/struct_ops.c b/tools/bpf/bpftool/struct_ops.c index 20f803dce2e4..cbdca37a53f0 100644 --- a/tools/bpf/bpftool/struct_ops.c +++ b/tools/bpf/bpftool/struct_ops.c @@ -32,7 +32,7 @@ static const struct btf *get_btf_vmlinux(void) return btf_vmlinux; btf_vmlinux = libbpf_find_kernel_btf(); - if (IS_ERR(btf_vmlinux)) + if (libbpf_get_error(btf_vmlinux)) p_err("struct_ops requires kernel CONFIG_DEBUG_INFO_BTF=y"); return btf_vmlinux; @@ -45,7 +45,7 @@ static const char *get_kern_struct_ops_name(const struct bpf_map_info *info) const char *st_ops_name; kern_btf = get_btf_vmlinux(); - if (IS_ERR(kern_btf)) + if (libbpf_get_error(kern_btf)) return ""; t = btf__type_by_id(kern_btf, info->btf_vmlinux_value_type_id); @@ -63,7 +63,7 @@ static __s32 get_map_info_type_id(void) return map_info_type_id; kern_btf = get_btf_vmlinux(); - if (IS_ERR(kern_btf)) { + if (libbpf_get_error(kern_btf)) { map_info_type_id = PTR_ERR(kern_btf); return map_info_type_id; } @@ -415,7 +415,7 @@ static int do_dump(int argc, char **argv) } kern_btf = get_btf_vmlinux(); - if (IS_ERR(kern_btf)) + if (libbpf_get_error(kern_btf)) return -1; if (!json_output) { @@ -495,7 +495,7 @@ static int do_register(int argc, char **argv) file = GET_ARG(); obj = bpf_object__open(file); - if (IS_ERR_OR_NULL(obj)) + if (libbpf_get_error(obj)) return -1; set_max_rlimit(); @@ -516,7 +516,7 @@ static int do_register(int argc, char **argv) continue; link = bpf_map__attach_struct_ops(map); - if (IS_ERR(link)) { + if (libbpf_get_error(link)) { p_err("can't register struct_ops %s: %s", bpf_map__name(map), strerror(-PTR_ERR(link))); @@ -596,7 +596,7 @@ int do_struct_ops(int argc, char **argv) err = cmd_select(cmds, argc, argv, do_help); - if (!IS_ERR(btf_vmlinux)) + if (!libbpf_get_error(btf_vmlinux)) btf__free(btf_vmlinux); return err; From 85f36923be47b6990215ad444545a6a85133a0c6 Mon Sep 17 00:00:00 2001 From: Karthikeyan Periyasamy Date: Fri, 12 Nov 2021 10:03:40 +0200 Subject: [PATCH 0151/2048] ath11k: fix fw crash due to peer get authorized before key install Firmware expects host to authorize the peer after the successful key install. But host authorize the peer before the key install, this trigger the firmware assert which leads to Q6 crash. To avoid this Q6 crash, host should authorize the peer after the key install. So introduce is_authorized in peer object to identify that peer is authorize or not. When IEEE80211_STA_CONTROL_PORT flag is unset, peer move to authorize state before the vdev up. When the same flag is set then peer move to authorize state after vdev up. So added authorise check in ath11k_bss_assoc() to handle the earlier state transition case. Also added the WMI authorize procedure in ath11k_mac_op_sta_state() to handle the non-earlier state transition case. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01492-QCAHKSWPL_SILICONZ-1 Signed-off-by: Karthikeyan Periyasamy Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636554200-12345-1-git-send-email-quic_periyasa@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 53 ++++++++++++++++++++++---- drivers/net/wireless/ath/ath11k/peer.h | 1 + 2 files changed, 47 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index a62c63ffb3cf..b82910c3f5a5 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -2474,6 +2474,8 @@ static void ath11k_bss_assoc(struct ieee80211_hw *hw, struct ath11k_vif *arvif = (void *)vif->drv_priv; struct peer_assoc_params peer_arg; struct ieee80211_sta *ap_sta; + struct ath11k_peer *peer; + bool is_auth = false; int ret; lockdep_assert_held(&ar->conf_mutex); @@ -2536,13 +2538,22 @@ static void ath11k_bss_assoc(struct ieee80211_hw *hw, "mac vdev %d up (associated) bssid %pM aid %d\n", arvif->vdev_id, bss_conf->bssid, bss_conf->aid); - /* Authorize BSS Peer */ - ret = ath11k_wmi_set_peer_param(ar, arvif->bssid, - arvif->vdev_id, - WMI_PEER_AUTHORIZE, - 1); - if (ret) - ath11k_warn(ar->ab, "Unable to authorize BSS peer: %d\n", ret); + spin_lock_bh(&ar->ab->base_lock); + + peer = ath11k_peer_find(ar->ab, arvif->vdev_id, arvif->bssid); + if (peer && peer->is_authorized) + is_auth = true; + + spin_unlock_bh(&ar->ab->base_lock); + + if (is_auth) { + ret = ath11k_wmi_set_peer_param(ar, arvif->bssid, + arvif->vdev_id, + WMI_PEER_AUTHORIZE, + 1); + if (ret) + ath11k_warn(ar->ab, "Unable to authorize BSS peer: %d\n", ret); + } ret = ath11k_wmi_send_obss_spr_cmd(ar, arvif->vdev_id, &bss_conf->he_obss_pd); @@ -4220,6 +4231,34 @@ static int ath11k_mac_op_sta_state(struct ieee80211_hw *hw, if (ret) ath11k_warn(ar->ab, "Failed to associate station: %pM\n", sta->addr); + } else if (old_state == IEEE80211_STA_ASSOC && + new_state == IEEE80211_STA_AUTHORIZED) { + spin_lock_bh(&ar->ab->base_lock); + + peer = ath11k_peer_find(ar->ab, arvif->vdev_id, sta->addr); + if (peer) + peer->is_authorized = true; + + spin_unlock_bh(&ar->ab->base_lock); + + if (vif->type == NL80211_IFTYPE_STATION && arvif->is_up) { + ret = ath11k_wmi_set_peer_param(ar, sta->addr, + arvif->vdev_id, + WMI_PEER_AUTHORIZE, + 1); + if (ret) + ath11k_warn(ar->ab, "Unable to authorize peer %pM vdev %d: %d\n", + sta->addr, arvif->vdev_id, ret); + } + } else if (old_state == IEEE80211_STA_AUTHORIZED && + new_state == IEEE80211_STA_ASSOC) { + spin_lock_bh(&ar->ab->base_lock); + + peer = ath11k_peer_find(ar->ab, arvif->vdev_id, sta->addr); + if (peer) + peer->is_authorized = false; + + spin_unlock_bh(&ar->ab->base_lock); } else if (old_state == IEEE80211_STA_ASSOC && new_state == IEEE80211_STA_AUTH && (vif->type == NL80211_IFTYPE_AP || diff --git a/drivers/net/wireless/ath/ath11k/peer.h b/drivers/net/wireless/ath/ath11k/peer.h index 619db001be8e..63fe5665badf 100644 --- a/drivers/net/wireless/ath/ath11k/peer.h +++ b/drivers/net/wireless/ath/ath11k/peer.h @@ -28,6 +28,7 @@ struct ath11k_peer { u8 ucast_keyidx; u16 sec_type; u16 sec_type_grp; + bool is_authorized; }; void ath11k_peer_unmap_event(struct ath11k_base *ab, u16 peer_id); From 4ea03443ecdac6920eb7aa3a9da2cd0b8cc6dfc8 Mon Sep 17 00:00:00 2001 From: Karthikeyan Periyasamy Date: Wed, 10 Nov 2021 21:05:57 +0530 Subject: [PATCH 0152/2048] ath11k: fix error routine when fallback of add interface fails When there is an error in add interface process from ath11k_mac_set_kickout(), the code attempts to handle a fallback for add_interface. When this fallback succeeds, the driver returns zero rather than error code. This leads to success for the non created VAP. In cleanup, driver gets remove interface callback for the non created VAP and proceeds to self peer delete request which leads to FW assert. Since it was already deleted on the fallback of add interface, return the actual error code instead of fallback return code. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-00729-QCAHKSWPL_SILICONZ-3 v2 Signed-off-by: Karthikeyan Periyasamy Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636558557-2874-1-git-send-email-quic_periyasa@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index b82910c3f5a5..5439a7918842 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -5613,7 +5613,7 @@ static int ath11k_mac_op_add_interface(struct ieee80211_hw *hw, u32 param_id, param_value; u16 nss; int i; - int ret; + int ret, fbret; int bit; vif->driver_flags |= IEEE80211_VIF_SUPPORTS_UAPSD; @@ -5816,17 +5816,17 @@ err_peer_del: if (arvif->vdev_type == WMI_VDEV_TYPE_AP) { reinit_completion(&ar->peer_delete_done); - ret = ath11k_wmi_send_peer_delete_cmd(ar, vif->addr, - arvif->vdev_id); - if (ret) { + fbret = ath11k_wmi_send_peer_delete_cmd(ar, vif->addr, + arvif->vdev_id); + if (fbret) { ath11k_warn(ar->ab, "failed to delete peer vdev_id %d addr %pM\n", arvif->vdev_id, vif->addr); goto err; } - ret = ath11k_wait_for_peer_delete_done(ar, arvif->vdev_id, - vif->addr); - if (ret) + fbret = ath11k_wait_for_peer_delete_done(ar, arvif->vdev_id, + vif->addr); + if (fbret) goto err; ar->num_peers--; From 4c375743c5fe1ef84d1dd7269dd12585957f403e Mon Sep 17 00:00:00 2001 From: Karthikeyan Periyasamy Date: Wed, 10 Nov 2021 21:36:28 +0530 Subject: [PATCH 0153/2048] ath11k: avoid unnecessary BH disable lock in STA kickout event In STA kickout event processing, the peer object is protected under spin lock BH. Release this lock after picking up the vdev_id from the peer object instead after ieee80211_report_low_ack(). This will minimize the lock hold period which will improve performance since base_lock is used across the data path. This was found in code review. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-00729-QCAHKSWPL_SILICONZ-3 v2 Signed-off-by: Karthikeyan Periyasamy Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636560388-24955-1-git-send-email-quic_periyasa@quicinc.com --- drivers/net/wireless/ath/ath11k/wmi.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 44057ccd0140..ea4bbc2d9062 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -6399,6 +6399,7 @@ static void ath11k_peer_sta_kickout_event(struct ath11k_base *ab, struct sk_buff struct ieee80211_sta *sta; struct ath11k_peer *peer; struct ath11k *ar; + u32 vdev_id; if (ath11k_pull_peer_sta_kickout_ev(ab, skb, &arg) != 0) { ath11k_warn(ab, "failed to extract peer sta kickout event"); @@ -6414,10 +6415,15 @@ static void ath11k_peer_sta_kickout_event(struct ath11k_base *ab, struct sk_buff if (!peer) { ath11k_warn(ab, "peer not found %pM\n", arg.mac_addr); + spin_unlock_bh(&ab->base_lock); goto exit; } - ar = ath11k_mac_get_ar_by_vdev_id(ab, peer->vdev_id); + vdev_id = peer->vdev_id; + + spin_unlock_bh(&ab->base_lock); + + ar = ath11k_mac_get_ar_by_vdev_id(ab, vdev_id); if (!ar) { ath11k_warn(ab, "invalid vdev id in peer sta kickout ev %d", peer->vdev_id); @@ -6438,7 +6444,6 @@ static void ath11k_peer_sta_kickout_event(struct ath11k_base *ab, struct sk_buff ieee80211_report_low_ack(sta, 10); exit: - spin_unlock_bh(&ab->base_lock); rcu_read_unlock(); } From 31aeaf547d7e3b64ba5d5442dabc530bdb9e216e Mon Sep 17 00:00:00 2001 From: Karthikeyan Periyasamy Date: Wed, 10 Nov 2021 21:51:30 +0530 Subject: [PATCH 0154/2048] ath11k: fix DMA memory free in CE pipe cleanup In CE pipe cleanup, DMA memory gets freed by the aligned address (base_addr_owner_space) which is wrong. It needs to be freed by the address (base_addr_owner_space_unaligned) returned by dma_alloc. So free the dma memory by the proper address. This was found in code review. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-00729-QCAHKSWPL_SILICONZ-3 v2 Signed-off-by: Karthikeyan Periyasamy Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636561290-18966-1-git-send-email-quic_periyasa@quicinc.com --- drivers/net/wireless/ath/ath11k/ce.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/ce.c b/drivers/net/wireless/ath/ath11k/ce.c index de8b632b058c..4249c2e94da9 100644 --- a/drivers/net/wireless/ath/ath11k/ce.c +++ b/drivers/net/wireless/ath/ath11k/ce.c @@ -953,6 +953,7 @@ int ath11k_ce_init_pipes(struct ath11k_base *ab) void ath11k_ce_free_pipes(struct ath11k_base *ab) { struct ath11k_ce_pipe *pipe; + struct ath11k_ce_ring *ce_ring; int desc_sz; int i; @@ -964,22 +965,24 @@ void ath11k_ce_free_pipes(struct ath11k_base *ab) if (pipe->src_ring) { desc_sz = ath11k_hal_ce_get_desc_size(HAL_CE_DESC_SRC); + ce_ring = pipe->src_ring; dma_free_coherent(ab->dev, pipe->src_ring->nentries * desc_sz + CE_DESC_RING_ALIGN, - pipe->src_ring->base_addr_owner_space, - pipe->src_ring->base_addr_ce_space); + ce_ring->base_addr_owner_space_unaligned, + ce_ring->base_addr_ce_space_unaligned); kfree(pipe->src_ring); pipe->src_ring = NULL; } if (pipe->dest_ring) { desc_sz = ath11k_hal_ce_get_desc_size(HAL_CE_DESC_DST); + ce_ring = pipe->dest_ring; dma_free_coherent(ab->dev, pipe->dest_ring->nentries * desc_sz + CE_DESC_RING_ALIGN, - pipe->dest_ring->base_addr_owner_space, - pipe->dest_ring->base_addr_ce_space); + ce_ring->base_addr_owner_space_unaligned, + ce_ring->base_addr_ce_space_unaligned); kfree(pipe->dest_ring); pipe->dest_ring = NULL; } @@ -987,11 +990,12 @@ void ath11k_ce_free_pipes(struct ath11k_base *ab) if (pipe->status_ring) { desc_sz = ath11k_hal_ce_get_desc_size(HAL_CE_DESC_DST_STATUS); + ce_ring = pipe->status_ring; dma_free_coherent(ab->dev, pipe->status_ring->nentries * desc_sz + CE_DESC_RING_ALIGN, - pipe->status_ring->base_addr_owner_space, - pipe->status_ring->base_addr_ce_space); + ce_ring->base_addr_owner_space_unaligned, + ce_ring->base_addr_ce_space_unaligned); kfree(pipe->status_ring); pipe->status_ring = NULL; } From 624e0a3170309eeb5b729f7a43c1ba3234325f02 Mon Sep 17 00:00:00 2001 From: Seevalamuthu Mariappan Date: Thu, 11 Nov 2021 11:22:47 +0530 Subject: [PATCH 0155/2048] ath11k: Fix 'unused-but-set-parameter' error Below compilation error is reported when built with W=1, drivers/net/wireless/ath/ath11k/mac.c:5408:22: error: parameter 'changed_flags' set but not used [-Werror,-Wunused-but-set-parameter] changed_flags is set, but left unused. So, remove unnecessary set. Compile tested only. Reported-by: kernel test robot Signed-off-by: Seevalamuthu Mariappan Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636609967-5114-1-git-send-email-quic_seevalam@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 5439a7918842..7bfe6fcdcb9a 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -5961,7 +5961,6 @@ static void ath11k_mac_op_configure_filter(struct ieee80211_hw *hw, mutex_lock(&ar->conf_mutex); - changed_flags &= SUPPORTED_FILTERS; *total_flags &= SUPPORTED_FILTERS; ar->filter_flags = *total_flags; From f187fe8e3bc65cc4d7b0916947e2d6cd65d9cd3a Mon Sep 17 00:00:00 2001 From: Venkateswara Naralasetty Date: Thu, 11 Nov 2021 20:50:02 +0530 Subject: [PATCH 0156/2048] ath11k: fix firmware crash during channel switch Currently the updated bandwidth for the peer will be configured to the firmware after channel switch from the sta_rc_update_wk. If the updated bandwidth is greater than the configured peer phymode during the peer assoc may result firmware assert. For example, initially AP is in HE40 mode and the peer phymode is configured as MODE_11AX_HE40 during peer assoc. Now user change the channel width to HE80 then, the peer bandwidth will be updated as HE80 to the firmware. This will trigger firmware assert due to peer bandwidth is greater than the peer phymode. Fix this issue by sending peer assoc command before setting the updated peer bandwith to firmware. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1 Signed-off-by: Venkateswara Naralasetty Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636644002-25446-1-git-send-email-quic_vnaralas@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 26 +++++++++++++++++++++----- 1 file changed, 21 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 7bfe6fcdcb9a..24fda9bc9209 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -3924,11 +3924,27 @@ static void ath11k_sta_rc_update_wk(struct work_struct *wk) ath11k_mac_max_he_nss(he_mcs_mask))); if (changed & IEEE80211_RC_BW_CHANGED) { - err = ath11k_wmi_set_peer_param(ar, sta->addr, arvif->vdev_id, - WMI_PEER_CHWIDTH, bw); - if (err) - ath11k_warn(ar->ab, "failed to update STA %pM peer bw %d: %d\n", - sta->addr, bw, err); + /* Send peer assoc command before set peer bandwidth param to + * avoid the mismatch between the peer phymode and the peer + * bandwidth. + */ + ath11k_peer_assoc_prepare(ar, arvif->vif, sta, &peer_arg, true); + + peer_arg.is_assoc = false; + err = ath11k_wmi_send_peer_assoc_cmd(ar, &peer_arg); + if (err) { + ath11k_warn(ar->ab, "failed to send peer assoc for STA %pM vdev %i: %d\n", + sta->addr, arvif->vdev_id, err); + } else if (wait_for_completion_timeout(&ar->peer_assoc_done, 1 * HZ)) { + err = ath11k_wmi_set_peer_param(ar, sta->addr, arvif->vdev_id, + WMI_PEER_CHWIDTH, bw); + if (err) + ath11k_warn(ar->ab, "failed to update STA %pM peer bw %d: %d\n", + sta->addr, bw, err); + } else { + ath11k_warn(ar->ab, "failed to get peer assoc conf event for %pM vdev %i\n", + sta->addr, arvif->vdev_id); + } } if (changed & IEEE80211_RC_NSS_CHANGED) { From 2c5545bfa29dd5305fa770959890a23ea39b5e69 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:01:18 +0200 Subject: [PATCH 0157/2048] ath11k: disable unused CE8 interrupts for ipq8074 Host driver doesn't need to process CE8 interrupts (used by target independently) The volume of interrupts is huge within short interval, CPU0 CPU1 CPU2 CPU3 14022188 0 0 0 GIC 71 Edge ce8 Hence disabling unused CE8 interrupt will improve CPU usage. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-2-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/ce.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath11k/ce.c b/drivers/net/wireless/ath/ath11k/ce.c index 4249c2e94da9..3bc4242c1059 100644 --- a/drivers/net/wireless/ath/ath11k/ce.c +++ b/drivers/net/wireless/ath/ath11k/ce.c @@ -77,7 +77,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = { /* CE8: target autonomous hif_memcpy */ { - .flags = CE_ATTR_FLAGS, + .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR, .src_nentries = 0, .src_sz_max = 0, .dest_nentries = 0, From 6452f0a3d5651bb7edfd9c709e78973aaa4d3bfc Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:01:26 +0200 Subject: [PATCH 0158/2048] ath11k: allocate dst ring descriptors from cacheable memory tcl_data and reo_dst rings are currently being allocated using dma_allocate_coherent() which is non cacheable. Allocating ring memory from cacheable memory area allows cached descriptor access and prefetch next descriptors to optimize CPU usage during descriptor processing on NAPI. Based on the hardware param we can enable or disable this feature for the corresponding platform. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Pradeep Kumar Chitrapu Signed-off-by: Pradeep Kumar Chitrapu Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-3-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/core.c | 5 ++++ drivers/net/wireless/ath/ath11k/dp.c | 38 ++++++++++++++++++++++---- drivers/net/wireless/ath/ath11k/dp.h | 1 + drivers/net/wireless/ath/ath11k/hal.c | 28 +++++++++++++++++-- drivers/net/wireless/ath/ath11k/hal.h | 1 + drivers/net/wireless/ath/ath11k/hw.h | 1 + 6 files changed, 67 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index bd77aedc7fe6..2ba3bf8d3fde 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -84,6 +84,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, + .alloc_cacheable_memory = true, }, { .hw_rev = ATH11K_HW_IPQ6018_HW10, @@ -135,6 +136,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, + .alloc_cacheable_memory = true, }, { .name = "qca6390 hw2.0", @@ -185,6 +187,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, + .alloc_cacheable_memory = false, }, { .name = "qcn9074 hw1.0", @@ -235,6 +238,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = true, + .alloc_cacheable_memory = true, }, { .name = "wcn6855 hw2.0", @@ -285,6 +289,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, + .alloc_cacheable_memory = false, }, }; diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c index 8baaeeb8cf82..8058b56028de 100644 --- a/drivers/net/wireless/ath/ath11k/dp.c +++ b/drivers/net/wireless/ath/ath11k/dp.c @@ -101,8 +101,11 @@ void ath11k_dp_srng_cleanup(struct ath11k_base *ab, struct dp_srng *ring) if (!ring->vaddr_unaligned) return; - dma_free_coherent(ab->dev, ring->size, ring->vaddr_unaligned, - ring->paddr_unaligned); + if (ring->cached) + kfree(ring->vaddr_unaligned); + else + dma_free_coherent(ab->dev, ring->size, ring->vaddr_unaligned, + ring->paddr_unaligned); ring->vaddr_unaligned = NULL; } @@ -222,6 +225,7 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring, int entry_sz = ath11k_hal_srng_get_entrysize(ab, type); int max_entries = ath11k_hal_srng_get_max_entries(ab, type); int ret; + bool cached = false; if (max_entries < 0 || entry_sz < 0) return -EINVAL; @@ -230,9 +234,28 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring, num_entries = max_entries; ring->size = (num_entries * entry_sz) + HAL_RING_BASE_ALIGN - 1; - ring->vaddr_unaligned = dma_alloc_coherent(ab->dev, ring->size, - &ring->paddr_unaligned, - GFP_KERNEL); + + if (ab->hw_params.alloc_cacheable_memory) { + /* Allocate the reo dst and tx completion rings from cacheable memory */ + switch (type) { + case HAL_REO_DST: + cached = true; + break; + default: + cached = false; + } + + if (cached) { + ring->vaddr_unaligned = kzalloc(ring->size, GFP_KERNEL); + ring->paddr_unaligned = virt_to_phys(ring->vaddr_unaligned); + } + } + + if (!cached) + ring->vaddr_unaligned = dma_alloc_coherent(ab->dev, ring->size, + &ring->paddr_unaligned, + GFP_KERNEL); + if (!ring->vaddr_unaligned) return -ENOMEM; @@ -292,6 +315,11 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring, return -EINVAL; } + if (cached) { + params.flags |= HAL_SRNG_FLAGS_CACHED; + ring->cached = 1; + } + ret = ath11k_hal_srng_setup(ab, type, ring_num, mac_id, ¶ms); if (ret < 0) { ath11k_warn(ab, "failed to setup srng: %d ring_id %d\n", diff --git a/drivers/net/wireless/ath/ath11k/dp.h b/drivers/net/wireless/ath/ath11k/dp.h index f524d19aca34..a4c36a9be338 100644 --- a/drivers/net/wireless/ath/ath11k/dp.h +++ b/drivers/net/wireless/ath/ath11k/dp.h @@ -64,6 +64,7 @@ struct dp_srng { dma_addr_t paddr; int size; u32 ring_id; + u8 cached; }; struct dp_rxdma_ring { diff --git a/drivers/net/wireless/ath/ath11k/hal.c b/drivers/net/wireless/ath/ath11k/hal.c index eaa0edca5576..f04edafbd0f1 100644 --- a/drivers/net/wireless/ath/ath11k/hal.c +++ b/drivers/net/wireless/ath/ath11k/hal.c @@ -627,6 +627,21 @@ u32 *ath11k_hal_srng_dst_peek(struct ath11k_base *ab, struct hal_srng *srng) return NULL; } +static void ath11k_hal_srng_prefetch_desc(struct ath11k_base *ab, + struct hal_srng *srng) +{ + u32 *desc; + + /* prefetch only if desc is available */ + desc = ath11k_hal_srng_dst_peek(ab, srng); + if (likely(desc)) { + dma_sync_single_for_cpu(ab->dev, virt_to_phys(desc), + (srng->entry_size * sizeof(u32)), + DMA_FROM_DEVICE); + prefetch(desc); + } +} + u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab, struct hal_srng *srng) { @@ -642,6 +657,10 @@ u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab, srng->u.dst_ring.tp = (srng->u.dst_ring.tp + srng->entry_size) % srng->ring_size; + /* Try to prefetch the next descriptor in the ring */ + if (srng->flags & HAL_SRNG_FLAGS_CACHED) + ath11k_hal_srng_prefetch_desc(ab, srng); + return desc; } @@ -775,11 +794,16 @@ void ath11k_hal_srng_access_begin(struct ath11k_base *ab, struct hal_srng *srng) { lockdep_assert_held(&srng->lock); - if (srng->ring_dir == HAL_SRNG_DIR_SRC) + if (srng->ring_dir == HAL_SRNG_DIR_SRC) { srng->u.src_ring.cached_tp = *(volatile u32 *)srng->u.src_ring.tp_addr; - else + } else { srng->u.dst_ring.cached_hp = *srng->u.dst_ring.hp_addr; + + /* Try to prefetch the next descriptor in the ring */ + if (srng->flags & HAL_SRNG_FLAGS_CACHED) + ath11k_hal_srng_prefetch_desc(ab, srng); + } } /* Update cached ring head/tail pointers to HW. ath11k_hal_srng_access_begin() diff --git a/drivers/net/wireless/ath/ath11k/hal.h b/drivers/net/wireless/ath/ath11k/hal.h index 35ed3a14e200..0f4f9ce74354 100644 --- a/drivers/net/wireless/ath/ath11k/hal.h +++ b/drivers/net/wireless/ath/ath11k/hal.h @@ -513,6 +513,7 @@ enum hal_srng_dir { #define HAL_SRNG_FLAGS_DATA_TLV_SWAP 0x00000020 #define HAL_SRNG_FLAGS_LOW_THRESH_INTR_EN 0x00010000 #define HAL_SRNG_FLAGS_MSI_INTR 0x00020000 +#define HAL_SRNG_FLAGS_CACHED 0x20000000 #define HAL_SRNG_FLAGS_LMAC_RING 0x80000000 #define HAL_SRNG_TLV_HDR_TAG GENMASK(9, 1) diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h index 2c7bd7a36ba5..3e64c9b94db0 100644 --- a/drivers/net/wireless/ath/ath11k/hw.h +++ b/drivers/net/wireless/ath/ath11k/hw.h @@ -178,6 +178,7 @@ struct ath11k_hw_params { u8 max_tx_ring; const struct ath11k_hw_hal_params *hal_params; bool supports_dynamic_smps_6ghz; + bool alloc_cacheable_memory; }; struct ath11k_hw_ops { From 5e76fe03dbf9f9dbc4fd454283b02594226c0718 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:03 +0200 Subject: [PATCH 0159/2048] ath11k: modify dp_rx desc access wrapper calls inline In data path, to reduce the CPU cycles spending on descriptor access wrapper function, changed those functions as static inline. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-4-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_rx.c | 114 ++++++++++++------------ 1 file changed, 59 insertions(+), 55 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index c5320847b80a..f4ae948a6dd6 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -20,13 +20,15 @@ #define ATH11K_DP_RX_FRAGMENT_TIMEOUT_MS (2 * HZ) -static u8 *ath11k_dp_rx_h_80211_hdr(struct ath11k_base *ab, struct hal_rx_desc *desc) +static inline +u8 *ath11k_dp_rx_h_80211_hdr(struct ath11k_base *ab, struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_hdr_status(desc); } -static enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline +enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_base *ab, + struct hal_rx_desc *desc) { if (!ab->hw_params.hw_ops->rx_desc_encrypt_valid(desc)) return HAL_ENCRYPT_TYPE_OPEN; @@ -34,32 +36,34 @@ static enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_bas return ab->hw_params.hw_ops->rx_desc_get_encrypt_type(desc); } -static u8 ath11k_dp_rx_h_msdu_start_decap_type(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_decap_type(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_decap_type(desc); } -static u8 ath11k_dp_rx_h_msdu_start_mesh_ctl_present(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline +u8 ath11k_dp_rx_h_msdu_start_mesh_ctl_present(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mesh_ctl(desc); } -static bool ath11k_dp_rx_h_mpdu_start_seq_ctrl_valid(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline +bool ath11k_dp_rx_h_mpdu_start_seq_ctrl_valid(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_seq_ctl_vld(desc); } -static bool ath11k_dp_rx_h_mpdu_start_fc_valid(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline bool ath11k_dp_rx_h_mpdu_start_fc_valid(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_fc_valid(desc); } -static bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab, - struct sk_buff *skb) +static inline bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab, + struct sk_buff *skb) { struct ieee80211_hdr *hdr; @@ -67,8 +71,8 @@ static bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab, return ieee80211_has_morefrags(hdr->frame_control); } -static u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab, - struct sk_buff *skb) +static inline u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab, + struct sk_buff *skb) { struct ieee80211_hdr *hdr; @@ -76,37 +80,37 @@ static u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab, return le16_to_cpu(hdr->seq_ctrl) & IEEE80211_SCTL_FRAG; } -static u16 ath11k_dp_rx_h_mpdu_start_seq_no(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u16 ath11k_dp_rx_h_mpdu_start_seq_no(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_start_seq_no(desc); } -static void *ath11k_dp_rx_get_attention(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline void *ath11k_dp_rx_get_attention(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_attention(desc); } -static bool ath11k_dp_rx_h_attn_msdu_done(struct rx_attention *attn) +static inline bool ath11k_dp_rx_h_attn_msdu_done(struct rx_attention *attn) { return !!FIELD_GET(RX_ATTENTION_INFO2_MSDU_DONE, __le32_to_cpu(attn->info2)); } -static bool ath11k_dp_rx_h_attn_l4_cksum_fail(struct rx_attention *attn) +static inline bool ath11k_dp_rx_h_attn_l4_cksum_fail(struct rx_attention *attn) { return !!FIELD_GET(RX_ATTENTION_INFO1_TCP_UDP_CKSUM_FAIL, __le32_to_cpu(attn->info1)); } -static bool ath11k_dp_rx_h_attn_ip_cksum_fail(struct rx_attention *attn) +static inline bool ath11k_dp_rx_h_attn_ip_cksum_fail(struct rx_attention *attn) { return !!FIELD_GET(RX_ATTENTION_INFO1_IP_CKSUM_FAIL, __le32_to_cpu(attn->info1)); } -static bool ath11k_dp_rx_h_attn_is_decrypted(struct rx_attention *attn) +static inline bool ath11k_dp_rx_h_attn_is_decrypted(struct rx_attention *attn) { return (FIELD_GET(RX_ATTENTION_INFO2_DCRYPT_STATUS_CODE, __le32_to_cpu(attn->info2)) == @@ -154,68 +158,68 @@ static bool ath11k_dp_rx_h_attn_msdu_len_err(struct ath11k_base *ab, return errmap & DP_RX_MPDU_ERR_MSDU_LEN; } -static u16 ath11k_dp_rx_h_msdu_start_msdu_len(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u16 ath11k_dp_rx_h_msdu_start_msdu_len(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_len(desc); } -static u8 ath11k_dp_rx_h_msdu_start_sgi(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_sgi(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_sgi(desc); } -static u8 ath11k_dp_rx_h_msdu_start_rate_mcs(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_rate_mcs(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_rate_mcs(desc); } -static u8 ath11k_dp_rx_h_msdu_start_rx_bw(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_rx_bw(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_rx_bw(desc); } -static u32 ath11k_dp_rx_h_msdu_start_freq(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u32 ath11k_dp_rx_h_msdu_start_freq(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_freq(desc); } -static u8 ath11k_dp_rx_h_msdu_start_pkt_type(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_pkt_type(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_msdu_pkt_type(desc); } -static u8 ath11k_dp_rx_h_msdu_start_nss(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_start_nss(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return hweight8(ab->hw_params.hw_ops->rx_desc_get_msdu_nss(desc)); } -static u8 ath11k_dp_rx_h_mpdu_start_tid(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_mpdu_start_tid(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_tid(desc); } -static u16 ath11k_dp_rx_h_mpdu_start_peer_id(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u16 ath11k_dp_rx_h_mpdu_start_peer_id(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_peer_id(desc); } -static u8 ath11k_dp_rx_h_msdu_end_l3pad(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline u8 ath11k_dp_rx_h_msdu_end_l3pad(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_l3_pad_bytes(desc); } -static bool ath11k_dp_rx_h_msdu_end_first_msdu(struct ath11k_base *ab, - struct hal_rx_desc *desc) +static inline bool ath11k_dp_rx_h_msdu_end_first_msdu(struct ath11k_base *ab, + struct hal_rx_desc *desc) { return ab->hw_params.hw_ops->rx_desc_get_first_msdu(desc); } @@ -233,14 +237,14 @@ static void ath11k_dp_rx_desc_end_tlv_copy(struct ath11k_base *ab, ab->hw_params.hw_ops->rx_desc_copy_attn_end_tlv(fdesc, ldesc); } -static u32 ath11k_dp_rxdesc_get_mpdulen_err(struct rx_attention *attn) +static inline u32 ath11k_dp_rxdesc_get_mpdulen_err(struct rx_attention *attn) { return FIELD_GET(RX_ATTENTION_INFO1_MPDU_LEN_ERR, __le32_to_cpu(attn->info1)); } -static u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab, - struct hal_rx_desc *rx_desc) +static inline u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab, + struct hal_rx_desc *rx_desc) { u8 *rx_pkt_hdr; @@ -249,8 +253,8 @@ static u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab, return rx_pkt_hdr; } -static bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab, - struct hal_rx_desc *rx_desc) +static inline bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab, + struct hal_rx_desc *rx_desc) { u32 tlv_tag; @@ -259,15 +263,15 @@ static bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab, return tlv_tag == HAL_RX_MPDU_START; } -static u32 ath11k_dp_rxdesc_get_ppduid(struct ath11k_base *ab, - struct hal_rx_desc *rx_desc) +static inline u32 ath11k_dp_rxdesc_get_ppduid(struct ath11k_base *ab, + struct hal_rx_desc *rx_desc) { return ab->hw_params.hw_ops->rx_desc_get_mpdu_ppdu_id(rx_desc); } -static void ath11k_dp_rxdesc_set_msdu_len(struct ath11k_base *ab, - struct hal_rx_desc *desc, - u16 len) +static inline void ath11k_dp_rxdesc_set_msdu_len(struct ath11k_base *ab, + struct hal_rx_desc *desc, + u16 len) { ab->hw_params.hw_ops->rx_desc_set_msdu_len(desc, len); } From a1775e732eb90486519de3813b83a11b7fcee2d0 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:12 +0200 Subject: [PATCH 0160/2048] ath11k: avoid additional access to ath11k_hal_srng_dst_num_free In ath11k_dp_process_rx(), after processing rx_desc from ath11k_hal_srng_dst_get_next_entry(), ath11k_hal_srng_dst_num_free() is accessed everytime because of done flag is not set. To avoid this additional access to ath11k_hal_srng_dst_num_free(), increment total_msdu_reaped only when continuation is not set and update done flag correspondingly. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-5-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_rx.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index f4ae948a6dd6..7c60f66077f5 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -2701,7 +2701,6 @@ try_again: DMA_FROM_DEVICE); num_buffs_reaped[mac_id]++; - total_msdu_reaped++; push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON, desc.info0); @@ -2728,10 +2727,15 @@ try_again: rxcb->mac_id = mac_id; __skb_queue_tail(&msdu_list, msdu); - if (total_msdu_reaped >= quota && !rxcb->is_continuation) { + if (rxcb->is_continuation) { + done = false; + } else { + total_msdu_reaped++; done = true; - break; } + + if (total_msdu_reaped >= budget) + break; } /* Hw might have updated the head pointer after we cached it. From c4d12cb37ea2e6c2b70880350d7bf1bbbd825c6c Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:18 +0200 Subject: [PATCH 0161/2048] ath11k: avoid active pdev check for each msdu The Active Pdev and CAC check are done for each msdu in ath11k_dp_rx_process_received_packets which is a overhead. To avoid this overhead, collect all msdus in a per mac msdu list and pass to function. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-6-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_rx.c | 70 ++++++++++++------------- 1 file changed, 34 insertions(+), 36 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index 7c60f66077f5..8dde2268646c 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -2600,13 +2600,11 @@ free_out: static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, struct napi_struct *napi, struct sk_buff_head *msdu_list, - int *quota, int ring_id) + int *quota, int mac_id) { - struct ath11k_skb_rxcb *rxcb; struct sk_buff *msdu; struct ath11k *ar; struct ieee80211_rx_status rx_status = {0}; - u8 mac_id; int ret; if (skb_queue_empty(msdu_list)) @@ -2614,20 +2612,20 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, rcu_read_lock(); - while (*quota && (msdu = __skb_dequeue(msdu_list))) { - rxcb = ATH11K_SKB_RXCB(msdu); - mac_id = rxcb->mac_id; - ar = ab->pdevs[mac_id].ar; - if (!rcu_dereference(ab->pdevs_active[mac_id])) { - dev_kfree_skb_any(msdu); - continue; - } + ar = ab->pdevs[mac_id].ar; + if (!rcu_dereference(ab->pdevs_active[mac_id])) { + __skb_queue_purge(msdu_list); + rcu_read_unlock(); + return; + } - if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) { - dev_kfree_skb_any(msdu); - continue; - } + if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) { + __skb_queue_purge(msdu_list); + rcu_read_unlock(); + return; + } + while ((msdu = __skb_dequeue(msdu_list))) { ret = ath11k_dp_rx_process_msdu(ar, msdu, msdu_list, &rx_status); if (ret) { ath11k_dbg(ab, ATH11K_DBG_DATA, @@ -2649,7 +2647,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, struct ath11k_dp *dp = &ab->dp; struct dp_rxdma_ring *rx_ring; int num_buffs_reaped[MAX_RADIOS] = {0}; - struct sk_buff_head msdu_list; + struct sk_buff_head msdu_list[MAX_RADIOS]; struct ath11k_skb_rxcb *rxcb; int total_msdu_reaped = 0; struct hal_srng *srng; @@ -2658,10 +2656,13 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, bool done = false; int buf_id, mac_id; struct ath11k *ar; - u32 *rx_desc; + struct hal_reo_dest_ring *desc; + enum hal_reo_dest_ring_push_reason push_reason; + u32 cookie; int i; - __skb_queue_head_init(&msdu_list); + for (i = 0; i < MAX_RADIOS; i++) + __skb_queue_head_init(&msdu_list[i]); srng = &ab->hal.srng_list[dp->reo_dst_ring[ring_id].ring_id]; @@ -2670,13 +2671,11 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, ath11k_hal_srng_access_begin(ab, srng); try_again: - while ((rx_desc = ath11k_hal_srng_dst_get_next_entry(ab, srng))) { - struct hal_reo_dest_ring desc = *(struct hal_reo_dest_ring *)rx_desc; - enum hal_reo_dest_ring_push_reason push_reason; - u32 cookie; - + while (likely(desc = + (struct hal_reo_dest_ring *)ath11k_hal_srng_dst_get_next_entry(ab, + srng))) { cookie = FIELD_GET(BUFFER_ADDR_INFO1_SW_COOKIE, - desc.buf_addr_info.info1); + desc->buf_addr_info.info1); buf_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_BUF_ID, cookie); mac_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_PDEV_ID, cookie); @@ -2703,7 +2702,7 @@ try_again: num_buffs_reaped[mac_id]++; push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON, - desc.info0); + desc->info0); if (push_reason != HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION) { dev_kfree_skb_any(msdu); @@ -2711,21 +2710,21 @@ try_again: continue; } - rxcb->is_first_msdu = !!(desc.rx_msdu_info.info0 & + rxcb->is_first_msdu = !!(desc->rx_msdu_info.info0 & RX_MSDU_DESC_INFO0_FIRST_MSDU_IN_MPDU); - rxcb->is_last_msdu = !!(desc.rx_msdu_info.info0 & + rxcb->is_last_msdu = !!(desc->rx_msdu_info.info0 & RX_MSDU_DESC_INFO0_LAST_MSDU_IN_MPDU); - rxcb->is_continuation = !!(desc.rx_msdu_info.info0 & + rxcb->is_continuation = !!(desc->rx_msdu_info.info0 & RX_MSDU_DESC_INFO0_MSDU_CONTINUATION); rxcb->peer_id = FIELD_GET(RX_MPDU_DESC_META_DATA_PEER_ID, - desc.rx_mpdu_info.meta_data); + desc->rx_mpdu_info.meta_data); rxcb->seq_no = FIELD_GET(RX_MPDU_DESC_INFO0_SEQ_NUM, - desc.rx_mpdu_info.info0); + desc->rx_mpdu_info.info0); rxcb->tid = FIELD_GET(HAL_REO_DEST_RING_INFO0_RX_QUEUE_NUM, - desc.info0); + desc->info0); rxcb->mac_id = mac_id; - __skb_queue_tail(&msdu_list, msdu); + __skb_queue_tail(&msdu_list[mac_id], msdu); if (rxcb->is_continuation) { done = false; @@ -2760,16 +2759,15 @@ try_again: if (!num_buffs_reaped[i]) continue; + ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i], + "a, i); + ar = ab->pdevs[i].ar; rx_ring = &ar->dp.rx_refill_buf_ring; ath11k_dp_rxbufs_replenish(ab, i, rx_ring, num_buffs_reaped[i], ab->hw_params.hal_params->rx_buf_rbm); } - - ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list, - "a, ring_id); - exit: return budget - quota; } From db2ecf9f0567a8f1a96f23a392cc5a30eaec4369 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:24 +0200 Subject: [PATCH 0162/2048] ath11k: remove usage quota while processing rx packets The usage of quota variable inside ath11k_dp_rx_process_received_packets() is redundant. Since we would queue only max packets to the list before calling this function so it would never exceed quota. Hence removing usage of quota variable. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-7-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_rx.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index 8dde2268646c..5a8ca9e064e9 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -2600,7 +2600,7 @@ free_out: static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, struct napi_struct *napi, struct sk_buff_head *msdu_list, - int *quota, int mac_id) + int mac_id) { struct sk_buff *msdu; struct ath11k *ar; @@ -2635,7 +2635,6 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, } ath11k_dp_rx_deliver_msdu(ar, napi, msdu, &rx_status); - (*quota)--; } rcu_read_unlock(); @@ -2652,7 +2651,6 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, int total_msdu_reaped = 0; struct hal_srng *srng; struct sk_buff *msdu; - int quota = budget; bool done = false; int buf_id, mac_id; struct ath11k *ar; @@ -2759,8 +2757,7 @@ try_again: if (!num_buffs_reaped[i]) continue; - ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i], - "a, i); + ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i], i); ar = ab->pdevs[i].ar; rx_ring = &ar->dp.rx_refill_buf_ring; @@ -2769,7 +2766,7 @@ try_again: ab->hw_params.hal_params->rx_buf_rbm); } exit: - return budget - quota; + return total_msdu_reaped; } static void ath11k_dp_rx_update_peer_stats(struct ath11k_sta *arsta, From 400588039a17a460292eb974ebba5811b8cbdb91 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:31 +0200 Subject: [PATCH 0163/2048] ath11k: add branch predictors in process_rx In datapath, add branch predictors where required in the process rx(). This protects high value rx path without having performance overhead. Also while processing rx packets, the pointer that is returned by rcu_dereference() is not dereferenced. so it is preferable to use rcu_access_pointer() here. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-8-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_rx.c | 24 +++++++++--------------- 1 file changed, 9 insertions(+), 15 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index 5a8ca9e064e9..3223d6d22628 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -2610,24 +2610,20 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, if (skb_queue_empty(msdu_list)) return; - rcu_read_lock(); - - ar = ab->pdevs[mac_id].ar; - if (!rcu_dereference(ab->pdevs_active[mac_id])) { + if (unlikely(!rcu_access_pointer(ab->pdevs_active[mac_id]))) { __skb_queue_purge(msdu_list); - rcu_read_unlock(); return; } - if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) { + ar = ab->pdevs[mac_id].ar; + if (unlikely(test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags))) { __skb_queue_purge(msdu_list); - rcu_read_unlock(); return; } while ((msdu = __skb_dequeue(msdu_list))) { ret = ath11k_dp_rx_process_msdu(ar, msdu, msdu_list, &rx_status); - if (ret) { + if (unlikely(ret)) { ath11k_dbg(ab, ATH11K_DBG_DATA, "Unable to process msdu %d", ret); dev_kfree_skb_any(msdu); @@ -2636,8 +2632,6 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab, ath11k_dp_rx_deliver_msdu(ar, napi, msdu, &rx_status); } - - rcu_read_unlock(); } int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, @@ -2682,7 +2676,7 @@ try_again: rx_ring = &ar->dp.rx_refill_buf_ring; spin_lock_bh(&rx_ring->idr_lock); msdu = idr_find(&rx_ring->bufs_idr, buf_id); - if (!msdu) { + if (unlikely(!msdu)) { ath11k_warn(ab, "frame rx with invalid buf_id %d\n", buf_id); spin_unlock_bh(&rx_ring->idr_lock); @@ -2701,8 +2695,8 @@ try_again: push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON, desc->info0); - if (push_reason != - HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION) { + if (unlikely(push_reason != + HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION)) { dev_kfree_skb_any(msdu); ab->soc_stats.hal_reo_error[dp->reo_dst_ring[ring_id].ring_id]++; continue; @@ -2741,7 +2735,7 @@ try_again: * head pointer so that we can reap complete MPDU in the current * rx processing. */ - if (!done && ath11k_hal_srng_dst_num_free(ab, srng, true)) { + if (unlikely(!done && ath11k_hal_srng_dst_num_free(ab, srng, true))) { ath11k_hal_srng_access_end(ab, srng); goto try_again; } @@ -2750,7 +2744,7 @@ try_again: spin_unlock_bh(&srng->lock); - if (!total_msdu_reaped) + if (unlikely(!total_msdu_reaped)) goto exit; for (i = 0; i < ab->num_radios; i++) { From d0e2523bfa9cb391fe966b0b6948c7e438981361 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:38 +0200 Subject: [PATCH 0164/2048] ath11k: allocate HAL_WBM2SW_RELEASE ring from cacheable memory Similar to REO destination ring, also allocate HAL_WBM2SW_RELEASE from cacheable memory so that descriptors could be prefetched during tx completion handling. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-9-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c index 8058b56028de..56b4cebbc6a5 100644 --- a/drivers/net/wireless/ath/ath11k/dp.c +++ b/drivers/net/wireless/ath/ath11k/dp.c @@ -239,6 +239,7 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring, /* Allocate the reo dst and tx completion rings from cacheable memory */ switch (type) { case HAL_REO_DST: + case HAL_WBM2SW_RELEASE: cached = true; break; default: From a8508bf7ced2e43f30b46333f09cbc79a1675616 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:02:45 +0200 Subject: [PATCH 0165/2048] ath11k: remove mod operator in dst ring processing Replace use of mod operator with a manual wrap around to avoid additional cost of using mod operation. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-10-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/hal.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/hal.c b/drivers/net/wireless/ath/ath11k/hal.c index f04edafbd0f1..7cf9e23cb922 100644 --- a/drivers/net/wireless/ath/ath11k/hal.c +++ b/drivers/net/wireless/ath/ath11k/hal.c @@ -654,8 +654,11 @@ u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab, desc = srng->ring_base_vaddr + srng->u.dst_ring.tp; - srng->u.dst_ring.tp = (srng->u.dst_ring.tp + srng->entry_size) % - srng->ring_size; + srng->u.dst_ring.tp += srng->entry_size; + + /* wrap around to start of ring*/ + if (srng->u.dst_ring.tp == srng->ring_size) + srng->u.dst_ring.tp = 0; /* Try to prefetch the next descriptor in the ring */ if (srng->flags & HAL_SRNG_FLAGS_CACHED) From cbfbed495d3289d5a0bc7c614cea639008086cfe Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:05:58 +0200 Subject: [PATCH 0166/2048] ath11k: avoid while loop in ring selection of tx completion interrupt Currently while loop is used to find the tx completion ring number and it is not required since the tx ring mask and the group id can be combined to directly fetch the ring number. Hence remove the while loop and directly get the ring number from tx mask and group id. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-11-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c index 56b4cebbc6a5..81b0b2baa461 100644 --- a/drivers/net/wireless/ath/ath11k/dp.c +++ b/drivers/net/wireless/ath/ath11k/dp.c @@ -771,13 +771,12 @@ int ath11k_dp_service_srng(struct ath11k_base *ab, const struct ath11k_hw_hal_params *hal_params; int grp_id = irq_grp->grp_id; int work_done = 0; - int i = 0, j; + int i, j; int tot_work_done = 0; - while (ab->hw_params.ring_mask->tx[grp_id] >> i) { - if (ab->hw_params.ring_mask->tx[grp_id] & BIT(i)) - ath11k_dp_tx_completion_handler(ab, i); - i++; + if (ab->hw_params.ring_mask->tx[grp_id]) { + i = __fls(ab->hw_params.ring_mask->tx[grp_id]); + ath11k_dp_tx_completion_handler(ab, i); } if (ab->hw_params.ring_mask->rx_err[grp_id]) { From bcef57ea400cc20a5389fa0e38d61063331558f8 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:06:04 +0200 Subject: [PATCH 0167/2048] ath11k: add branch predictors in dp_tx path Add branch prediction in dp_tx code path in tx and tx completion handlers. Also in ath11k_dp_tx_complete_msdu , the pointer that is returned by rcu_dereference() is not dereferenced. so it is preferable to use rcu_access_pointer() here. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Co-developed-by: Sriram R Signed-off-by: Sriram R Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-12-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_tx.c | 52 ++++++++++++------------- drivers/net/wireless/ath/ath11k/mac.c | 2 +- 2 files changed, 25 insertions(+), 29 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c index 10b76f6f710b..28a7e2c74832 100644 --- a/drivers/net/wireless/ath/ath11k/dp_tx.c +++ b/drivers/net/wireless/ath/ath11k/dp_tx.c @@ -95,11 +95,11 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif, u8 ring_selector = 0, ring_map = 0; bool tcl_ring_retry; - if (test_bit(ATH11K_FLAG_CRASH_FLUSH, &ar->ab->dev_flags)) + if (unlikely(test_bit(ATH11K_FLAG_CRASH_FLUSH, &ar->ab->dev_flags))) return -ESHUTDOWN; - if (!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) && - !ieee80211_is_data(hdr->frame_control)) + if (unlikely(!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) && + !ieee80211_is_data(hdr->frame_control))) return -ENOTSUPP; pool_id = skb_get_queue_mapping(skb) & (ATH11K_HW_MAX_QUEUES - 1); @@ -127,7 +127,7 @@ tcl_ring_sel: DP_TX_IDR_SIZE - 1, GFP_ATOMIC); spin_unlock_bh(&tx_ring->tx_idr_lock); - if (ret < 0) { + if (unlikely(ret < 0)) { if (ring_map == (BIT(ab->hw_params.max_tx_ring) - 1)) { atomic_inc(&ab->soc_stats.tx_err.misc_fail); return -ENOSPC; @@ -152,7 +152,7 @@ tcl_ring_sel: ti.meta_data_flags = arvif->tcl_metadata; } - if (ti.encap_type == HAL_TCL_ENCAP_TYPE_RAW) { + if (unlikely(ti.encap_type == HAL_TCL_ENCAP_TYPE_RAW)) { if (skb_cb->flags & ATH11K_SKB_CIPHER_SET) { ti.encrypt_type = ath11k_dp_tx_get_encrypt_type(skb_cb->cipher); @@ -173,8 +173,8 @@ tcl_ring_sel: ti.bss_ast_idx = arvif->ast_idx; ti.dscp_tid_tbl_idx = 0; - if (skb->ip_summed == CHECKSUM_PARTIAL && - ti.encap_type != HAL_TCL_ENCAP_TYPE_RAW) { + if (likely(skb->ip_summed == CHECKSUM_PARTIAL && + ti.encap_type != HAL_TCL_ENCAP_TYPE_RAW)) { ti.flags0 |= FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_IP4_CKSUM_EN, 1) | FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_UDP4_CKSUM_EN, 1) | FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_UDP6_CKSUM_EN, 1) | @@ -211,7 +211,7 @@ tcl_ring_sel: } ti.paddr = dma_map_single(ab->dev, skb->data, skb->len, DMA_TO_DEVICE); - if (dma_mapping_error(ab->dev, ti.paddr)) { + if (unlikely(dma_mapping_error(ab->dev, ti.paddr))) { atomic_inc(&ab->soc_stats.tx_err.misc_fail); ath11k_warn(ab, "failed to DMA map data Tx buffer\n"); ret = -ENOMEM; @@ -231,7 +231,7 @@ tcl_ring_sel: ath11k_hal_srng_access_begin(ab, tcl_ring); hal_tcl_desc = (void *)ath11k_hal_srng_src_get_next_entry(ab, tcl_ring); - if (!hal_tcl_desc) { + if (unlikely(!hal_tcl_desc)) { /* NOTE: It is highly unlikely we'll be running out of tcl_ring * desc because the desc is directly enqueued onto hw queue. */ @@ -245,7 +245,7 @@ tcl_ring_sel: * checking this ring earlier for each pkt tx. * Restart ring selection if some rings are not checked yet. */ - if (ring_map != (BIT(ab->hw_params.max_tx_ring) - 1) && + if (unlikely(ring_map != (BIT(ab->hw_params.max_tx_ring)) - 1) && ab->hw_params.max_tx_ring > 1) { tcl_ring_retry = true; ring_selector++; @@ -327,7 +327,7 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab, spin_lock_bh(&tx_ring->tx_idr_lock); msdu = idr_find(&tx_ring->txbuf_idr, ts->msdu_id); - if (!msdu) { + if (unlikely(!msdu)) { ath11k_warn(ab, "htt tx completion for unknown msdu_id %d\n", ts->msdu_id); spin_unlock_bh(&tx_ring->tx_idr_lock); @@ -435,16 +435,14 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar, dma_unmap_single(ab->dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE); - rcu_read_lock(); - - if (!rcu_dereference(ab->pdevs_active[ar->pdev_idx])) { + if (unlikely(!rcu_access_pointer(ab->pdevs_active[ar->pdev_idx]))) { dev_kfree_skb_any(msdu); - goto exit; + return; } - if (!skb_cb->vif) { + if (unlikely(!skb_cb->vif)) { dev_kfree_skb_any(msdu); - goto exit; + return; } info = IEEE80211_SKB_CB(msdu); @@ -465,7 +463,7 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar, (info->flags & IEEE80211_TX_CTL_NO_ACK)) info->flags |= IEEE80211_TX_STAT_NOACK_TRANSMITTED; - if (ath11k_debugfs_is_extd_tx_stats_enabled(ar)) { + if (unlikely(ath11k_debugfs_is_extd_tx_stats_enabled(ar))) { if (ts->flags & HAL_TX_STATUS_FLAGS_FIRST_MSDU) { if (ar->last_ppdu_id == 0) { ar->last_ppdu_id = ts->ppdu_id; @@ -494,9 +492,6 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar, */ ieee80211_tx_status(ar->hw, msdu); - -exit: - rcu_read_unlock(); } static inline void ath11k_dp_tx_status_parse(struct ath11k_base *ab, @@ -505,11 +500,11 @@ static inline void ath11k_dp_tx_status_parse(struct ath11k_base *ab, { ts->buf_rel_source = FIELD_GET(HAL_WBM_RELEASE_INFO0_REL_SRC_MODULE, desc->info0); - if (ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_FW && - ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_TQM) + if (unlikely(ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_FW && + ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_TQM)) return; - if (ts->buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW) + if (unlikely(ts->buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW)) return; ts->status = FIELD_GET(HAL_WBM_RELEASE_INFO0_TQM_RELEASE_REASON, @@ -556,8 +551,9 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id) ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head); } - if ((ath11k_hal_srng_dst_peek(ab, status_ring) != NULL) && - (ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head) == tx_ring->tx_status_tail)) { + if (unlikely((ath11k_hal_srng_dst_peek(ab, status_ring) != NULL) && + (ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head) == + tx_ring->tx_status_tail))) { /* TODO: Process pending tx_status messages when kfifo_is_full() */ ath11k_warn(ab, "Unable to process some of the tx_status ring desc because status_fifo is full\n"); } @@ -580,7 +576,7 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id) mac_id = FIELD_GET(DP_TX_DESC_ID_MAC_ID, desc_id); msdu_id = FIELD_GET(DP_TX_DESC_ID_MSDU_ID, desc_id); - if (ts.buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW) { + if (unlikely(ts.buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW)) { ath11k_dp_tx_process_htt_tx_complete(ab, (void *)tx_status, mac_id, msdu_id, @@ -590,7 +586,7 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id) spin_lock_bh(&tx_ring->tx_idr_lock); msdu = idr_find(&tx_ring->txbuf_idr, msdu_id); - if (!msdu) { + if (unlikely(!msdu)) { ath11k_warn(ab, "tx completion for unknown msdu_id %d\n", msdu_id); spin_unlock_bh(&tx_ring->tx_idr_lock); diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 24fda9bc9209..cec49468d615 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -5283,7 +5283,7 @@ static void ath11k_mac_op_tx(struct ieee80211_hw *hw, arsta = (struct ath11k_sta *)control->sta->drv_priv; ret = ath11k_dp_tx(ar, arvif, arsta, skb); - if (ret) { + if (unlikely(ret)) { ath11k_warn(ar->ab, "failed to transmit frame %d\n", ret); ieee80211_free_txskb(ar->hw, skb); } From be8867cb47652418e488170785bd9ffbadae3f1f Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Fri, 12 Nov 2021 11:06:11 +0200 Subject: [PATCH 0168/2048] ath11k: avoid unnecessary lock contention in tx_completion path Avoid unnecessary idr_find calls before the idr_remove calls. Because idr_remove gives the valid ptr if id is valid otherwise return NULL ptr. So removed the idr_find before idr_remove in tx completion path. Also no need to disable the bottom half preempt if it is already in the bottom half context, so modify the spin_lock_bh to spin_lock in the data tx completion path. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Co-developed-by: Karthikeyan Periyasamy Signed-off-by: Karthikeyan Periyasamy Signed-off-by: Jouni Malinen Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1630560820-21905-13-git-send-email-ppranees@codeaurora.org --- drivers/net/wireless/ath/ath11k/dp_tx.c | 32 +++++++++++-------------- 1 file changed, 14 insertions(+), 18 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c index 28a7e2c74832..88abd64e9047 100644 --- a/drivers/net/wireless/ath/ath11k/dp_tx.c +++ b/drivers/net/wireless/ath/ath11k/dp_tx.c @@ -293,20 +293,18 @@ static void ath11k_dp_tx_free_txbuf(struct ath11k_base *ab, u8 mac_id, struct sk_buff *msdu; struct ath11k_skb_cb *skb_cb; - spin_lock_bh(&tx_ring->tx_idr_lock); - msdu = idr_find(&tx_ring->txbuf_idr, msdu_id); - if (!msdu) { + spin_lock(&tx_ring->tx_idr_lock); + msdu = idr_remove(&tx_ring->txbuf_idr, msdu_id); + spin_unlock(&tx_ring->tx_idr_lock); + + if (unlikely(!msdu)) { ath11k_warn(ab, "tx completion for unknown msdu_id %d\n", msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); return; } skb_cb = ATH11K_SKB_CB(msdu); - idr_remove(&tx_ring->txbuf_idr, msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); - dma_unmap_single(ab->dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE); dev_kfree_skb_any(msdu); @@ -325,12 +323,13 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab, struct ath11k_skb_cb *skb_cb; struct ath11k *ar; - spin_lock_bh(&tx_ring->tx_idr_lock); - msdu = idr_find(&tx_ring->txbuf_idr, ts->msdu_id); + spin_lock(&tx_ring->tx_idr_lock); + msdu = idr_remove(&tx_ring->txbuf_idr, ts->msdu_id); + spin_unlock(&tx_ring->tx_idr_lock); + if (unlikely(!msdu)) { ath11k_warn(ab, "htt tx completion for unknown msdu_id %d\n", ts->msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); return; } @@ -339,9 +338,6 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab, ar = skb_cb->ar; - idr_remove(&tx_ring->txbuf_idr, ts->msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); - if (atomic_dec_and_test(&ar->dp.num_tx_pending)) wake_up(&ar->dp.tx_empty_waitq); @@ -584,16 +580,16 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id) continue; } - spin_lock_bh(&tx_ring->tx_idr_lock); - msdu = idr_find(&tx_ring->txbuf_idr, msdu_id); + spin_lock(&tx_ring->tx_idr_lock); + msdu = idr_remove(&tx_ring->txbuf_idr, msdu_id); if (unlikely(!msdu)) { ath11k_warn(ab, "tx completion for unknown msdu_id %d\n", msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); + spin_unlock(&tx_ring->tx_idr_lock); continue; } - idr_remove(&tx_ring->txbuf_idr, msdu_id); - spin_unlock_bh(&tx_ring->tx_idr_lock); + + spin_unlock(&tx_ring->tx_idr_lock); ar = ab->pdevs[mac_id].ar; From 78406044bdd0cc8987bc082b76867c63ab1c6af8 Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Wed, 13 Oct 2021 03:37:04 -0400 Subject: [PATCH 0169/2048] ath11k: enable IEEE80211_VHT_EXT_NSS_BW_CAPABLE if NSS ratio enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When NSS ratio enabled reported by firmware, SUPPORTS_VHT_EXT_NSS_BW is set in ath11k, meanwhile IEEE80211_VHT_EXT_NSS_BW_CAPABLE also need to be set, otherwise it is invalid because spec in IEEE Std 802.11™‐2020 as below. Table 9-273-Supported VHT-MCS and NSS Set subfields, it has subfield VHT Extended NSS BW Capable, its definition is: Indicates whether the STA is capable of interpreting the Extended NSS BW Support subfield of the VHT Capabilities Information field. dmesg have a message without this patch: ieee80211 phy0: copying sband (band 1) due to VHT EXT NSS BW flag It means mac80211 will set IEEE80211_VHT_EXT_NSS_BW_CAPABLE if ath11k not set it in ieee80211_register_hw(). So it is better to set it in ath11k. Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1 Signed-off-by: Wen Gong Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211013073704.15888-1-wgong@codeaurora.org --- drivers/net/wireless/ath/ath11k/mac.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index cec49468d615..659d3f387b21 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -4706,6 +4706,10 @@ ath11k_create_vht_cap(struct ath11k *ar, u32 rate_cap_tx_chainmask, vht_cap.vht_supported = 1; vht_cap.cap = ar->pdev->cap.vht_cap; + if (ar->pdev->cap.nss_ratio_enabled) + vht_cap.vht_mcs.tx_highest |= + cpu_to_le16(IEEE80211_VHT_EXT_NSS_BW_CAPABLE); + ath11k_set_vht_txbf_cap(ar, &vht_cap.cap); rxmcs_map = 0; From 1d795645e1eef97fe5d409e3dd5747a942f00e08 Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Mon, 11 Oct 2021 04:49:57 -0400 Subject: [PATCH 0170/2048] ath11k: remove return for empty tx bitrate in mac_op_sta_statistics Currently in ath11k_mac_op_sta_statistics() there is the following logic: if (!arsta->txrate.legacy && !arsta->txrate.nss) return; Unfortunately if this condition is true then the function returns without setting parameters that follow the txrate. To address this issue remove the return and instead invert the logic to set the txrate logic if (arsta->txrate.legacy || arsta->txrate.nss). The same was done also in ath10k in commit 1cd6ba8ae33e ("ath10k: remove return for NL80211_STA_INFO_TX_BITRATE"). Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1 Signed-off-by: Wen Gong Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211011084957.31024-1-wgong@codeaurora.org --- drivers/net/wireless/ath/ath11k/mac.c | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 659d3f387b21..e653a0a19d60 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -7426,21 +7426,20 @@ static void ath11k_mac_op_sta_statistics(struct ieee80211_hw *hw, sinfo->tx_duration = arsta->tx_duration; sinfo->filled |= BIT_ULL(NL80211_STA_INFO_TX_DURATION); - if (!arsta->txrate.legacy && !arsta->txrate.nss) - return; - - if (arsta->txrate.legacy) { - sinfo->txrate.legacy = arsta->txrate.legacy; - } else { - sinfo->txrate.mcs = arsta->txrate.mcs; - sinfo->txrate.nss = arsta->txrate.nss; - sinfo->txrate.bw = arsta->txrate.bw; - sinfo->txrate.he_gi = arsta->txrate.he_gi; - sinfo->txrate.he_dcm = arsta->txrate.he_dcm; - sinfo->txrate.he_ru_alloc = arsta->txrate.he_ru_alloc; + if (arsta->txrate.legacy || arsta->txrate.nss) { + if (arsta->txrate.legacy) { + sinfo->txrate.legacy = arsta->txrate.legacy; + } else { + sinfo->txrate.mcs = arsta->txrate.mcs; + sinfo->txrate.nss = arsta->txrate.nss; + sinfo->txrate.bw = arsta->txrate.bw; + sinfo->txrate.he_gi = arsta->txrate.he_gi; + sinfo->txrate.he_dcm = arsta->txrate.he_dcm; + sinfo->txrate.he_ru_alloc = arsta->txrate.he_ru_alloc; + } + sinfo->txrate.flags = arsta->txrate.flags; + sinfo->filled |= BIT_ULL(NL80211_STA_INFO_TX_BITRATE); } - sinfo->txrate.flags = arsta->txrate.flags; - sinfo->filled |= BIT_ULL(NL80211_STA_INFO_TX_BITRATE); /* TODO: Use real NF instead of default one. */ sinfo->signal = arsta->rssi_comb + ATH11K_DEFAULT_NOISE_FLOOR; From c8f2d41bbff6794329d681d108a817366aed0ba7 Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Mon, 25 Oct 2021 23:20:14 -0400 Subject: [PATCH 0171/2048] ath11k: fix the value of msecs_to_jiffies in ath11k_debugfs_fw_stats_request parameter of msecs_to_jiffies should be (3 * 1000) instead of (3 * HZ) Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1 Signed-off-by: Wen Gong Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211026032014.27010-1-quic_wgong@quicinc.com --- drivers/net/wireless/ath/ath11k/debugfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath11k/debugfs.c b/drivers/net/wireless/ath/ath11k/debugfs.c index 80afd35337a1..dba055d085be 100644 --- a/drivers/net/wireless/ath/ath11k/debugfs.c +++ b/drivers/net/wireless/ath/ath11k/debugfs.c @@ -195,7 +195,7 @@ static int ath11k_debugfs_fw_stats_request(struct ath11k *ar, * received 'update stats' event, we keep a 3 seconds timeout in case, * fw_stats_done is not marked yet */ - timeout = jiffies + msecs_to_jiffies(3 * HZ); + timeout = jiffies + msecs_to_jiffies(3 * 1000); ath11k_debugfs_fw_stats_reset(ar); From b4a0f54156ac7720de1750b6ea06657c91c52163 Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Wed, 27 Oct 2021 05:38:25 -0400 Subject: [PATCH 0172/2048] ath11k: move peer delete after vdev stop of station for QCA6390 and WCN6855 When station connect to AP, the wmi command sequence is: peer_create->vdev_start->vdev_up and sequence of station disconnect fo AP is: peer_delete->vdev_down->vdev_stop The sequence of disconnect is not opposite of connect, it caused firmware crash when it handle wmi vdev stop cmd when the AP is support TWT of 802.11 ax, because firmware need access the bss peer for vdev stop cmd. [ 390.438564] ath11k_pci 0000:05:00.0: wmi cmd send 0x6001 ret 0 [ 390.438567] ath11k_pci 0000:05:00.0: WMI peer create vdev_id 0 peer_addr c4:04:15:3b:e0:39 [ 390.472724] ath11k_pci 0000:05:00.0: mac vdev 0 start center_freq 2437 phymode 11ax-he20-2g [ 390.472731] ath11k_pci 0000:05:00.0: wmi cmd send 0x5003 ret 0 [ 390.560849] ath11k_pci 0000:05:00.0: wmi cmd send 0x5005 ret 0 [ 390.560850] ath11k_pci 0000:05:00.0: WMI mgmt vdev up id 0x0 assoc id 1 bssid c4:04:15:3b:e0:39 [ 399.432896] ath11k_pci 0000:05:00.0: WMI peer delete vdev_id 0 peer_addr c4:04:15:3b:e0:39 [ 399.432902] ath11k_pci 0000:05:00.0: wmi cmd send 0x6002 ret 0 [ 399.441380] ath11k_pci 0000:05:00.0: wmi cmd send 0x5007 ret 0 [ 399.441381] ath11k_pci 0000:05:00.0: WMI vdev down id 0x0 [ 399.454681] ath11k_pci 0000:05:00.0: wmi cmd send 0x5006 ret 0 [ 399.454682] ath11k_pci 0000:05:00.0: WMI vdev stop id 0x0 The opposite sequence of disconnect should be: vdev_down->vdev_stop->peer_delete This patch change the sequence of disconnect for station as above opposite sequence for QCA6390, firmware not crash again with this patch. Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1 Signed-off-by: Wen Gong Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211027093825.12167-1-quic_wgong@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index e653a0a19d60..34f6d37ed4d0 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -4212,6 +4212,10 @@ static int ath11k_mac_op_sta_state(struct ieee80211_hw *hw, new_state == IEEE80211_STA_NOTEXIST)) { ath11k_dp_peer_cleanup(ar, arvif->vdev_id, sta->addr); + if (ar->ab->hw_params.vdev_start_delay && + vif->type == NL80211_IFTYPE_STATION) + goto free; + ret = ath11k_peer_delete(ar, arvif->vdev_id, sta->addr); if (ret) ath11k_warn(ar->ab, "Failed to delete peer: %pM for VDEV: %d\n", @@ -4233,6 +4237,7 @@ static int ath11k_mac_op_sta_state(struct ieee80211_hw *hw, } spin_unlock_bh(&ar->ab->base_lock); +free: kfree(arsta->tx_stats); arsta->tx_stats = NULL; @@ -6616,6 +6621,19 @@ ath11k_mac_op_unassign_vif_chanctx(struct ieee80211_hw *hw, arvif->is_started = false; + if (ab->hw_params.vdev_start_delay && + arvif->vdev_type == WMI_VDEV_TYPE_STA) { + ret = ath11k_peer_delete(ar, arvif->vdev_id, arvif->bssid); + if (ret) + ath11k_warn(ar->ab, + "failed to delete peer %pM for vdev %d: %d\n", + arvif->bssid, arvif->vdev_id, ret); + else + ath11k_dbg(ar->ab, ATH11K_DBG_MAC, + "mac removed peer %pM vdev %d after vdev stop\n", + arvif->bssid, arvif->vdev_id); + } + if (ab->hw_params.vdev_start_delay && arvif->vdev_type == WMI_VDEV_TYPE_MONITOR) ath11k_wmi_vdev_down(ar, arvif->vdev_id); From 787264893c69ed091a46335dfd0f50dabb457718 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Mon, 25 Oct 2021 17:44:20 +0530 Subject: [PATCH 0173/2048] ath11k: fix FCS_ERR flag in radio tap header In radio tap header, BAD FCS flag is not updated properly because driver failed to update FCS_ERR flag in monitor mode. In rx_desc, FCS_ERR information is available in rx_attention structure and presence of this field indicates corresponding frame failed FCS check. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1 Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635164060-18423-1-git-send-email-quic_ppranees@quicinc.com --- drivers/net/wireless/ath/ath11k/dp_rx.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index 3223d6d22628..fcd7a6d27d12 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -4826,7 +4826,7 @@ static struct sk_buff * ath11k_dp_rx_mon_merg_msdus(struct ath11k *ar, u32 mac_id, struct sk_buff *head_msdu, struct sk_buff *last_msdu, - struct ieee80211_rx_status *rxs) + struct ieee80211_rx_status *rxs, bool *fcs_err) { struct ath11k_base *ab = ar->ab; struct sk_buff *msdu, *prev_buf; @@ -4836,12 +4836,17 @@ ath11k_dp_rx_mon_merg_msdus(struct ath11k *ar, u8 *dest, decap_format; struct ieee80211_hdr_3addr *wh; struct rx_attention *rx_attention; + u32 err_bitmap; if (!head_msdu) goto err_merge_fail; rx_desc = (struct hal_rx_desc *)head_msdu->data; rx_attention = ath11k_dp_rx_get_attention(ab, rx_desc); + err_bitmap = ath11k_dp_rx_h_attn_mpdu_err(rx_attention); + + if (err_bitmap & DP_RX_MPDU_ERR_FCS) + *fcs_err = true; if (ath11k_dp_rxdesc_get_mpdulen_err(rx_attention)) return NULL; @@ -4930,9 +4935,10 @@ static int ath11k_dp_rx_mon_deliver(struct ath11k *ar, u32 mac_id, struct ath11k_pdev_dp *dp = &ar->dp; struct sk_buff *mon_skb, *skb_next, *header; struct ieee80211_rx_status *rxs = &dp->rx_status; + bool fcs_err = false; mon_skb = ath11k_dp_rx_mon_merg_msdus(ar, mac_id, head_msdu, - tail_msdu, rxs); + tail_msdu, rxs, &fcs_err); if (!mon_skb) goto mon_deliver_fail; @@ -4940,6 +4946,10 @@ static int ath11k_dp_rx_mon_deliver(struct ath11k *ar, u32 mac_id, header = mon_skb; rxs->flag = 0; + + if (fcs_err) + rxs->flag = RX_FLAG_FAILED_FCS_CRC; + do { skb_next = mon_skb->next; if (!skb_next) From 9212c1b9e80a869e732769a4fe7f82d392b219be Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Mon, 25 Oct 2021 17:47:09 +0530 Subject: [PATCH 0174/2048] ath11k: send proper txpower and maxregpower values to firmware Set proper values for max_regpower, max_power, max_antenna_gain as it is because firmware will convert power values to 0.5dbm steps by multiplying it with 2. If txpower is not set, it will lead to cca stuck resulting in latency issues for QCN9074. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1-01386-QCAHKSWPL_SILICONZ-1 Co-developed-by: Lavanya Suresh Signed-off-by: Lavanya Suresh Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635164229-22880-1-git-send-email-quic_ppranees@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 12 ++++++------ drivers/net/wireless/ath/ath11k/wmi.c | 2 ++ 2 files changed, 8 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 34f6d37ed4d0..5bc82be25316 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -775,9 +775,9 @@ static int ath11k_mac_monitor_vdev_start(struct ath11k *ar, int vdev_id, arg.channel.chan_radar = !!(channel->flags & IEEE80211_CHAN_RADAR); arg.channel.min_power = 0; - arg.channel.max_power = channel->max_power * 2; - arg.channel.max_reg_power = channel->max_reg_power * 2; - arg.channel.max_antenna_gain = channel->max_antenna_gain * 2; + arg.channel.max_power = channel->max_power; + arg.channel.max_reg_power = channel->max_reg_power; + arg.channel.max_antenna_gain = channel->max_antenna_gain; arg.pref_tx_streams = ar->num_tx_chains; arg.pref_rx_streams = ar->num_rx_chains; @@ -6123,9 +6123,9 @@ ath11k_mac_vdev_start_restart(struct ath11k_vif *arvif, ath11k_phymodes[chandef->chan->band][chandef->width]; arg.channel.min_power = 0; - arg.channel.max_power = chandef->chan->max_power * 2; - arg.channel.max_reg_power = chandef->chan->max_reg_power * 2; - arg.channel.max_antenna_gain = chandef->chan->max_antenna_gain * 2; + arg.channel.max_power = chandef->chan->max_power; + arg.channel.max_reg_power = chandef->chan->max_reg_power; + arg.channel.max_antenna_gain = chandef->chan->max_antenna_gain; arg.pref_tx_streams = ar->num_tx_chains; arg.pref_rx_streams = ar->num_rx_chains; diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index ea4bbc2d9062..6248e174bd8f 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -2387,6 +2387,8 @@ int ath11k_wmi_send_scan_chan_list_cmd(struct ath11k *ar, tchan_info->reg_class_id); *reg2 |= FIELD_PREP(WMI_CHAN_REG_INFO2_ANT_MAX, tchan_info->antennamax); + *reg2 |= FIELD_PREP(WMI_CHAN_REG_INFO2_MAX_TX_PWR, + tchan_info->maxregpower); ath11k_dbg(ar->ab, ATH11K_DBG_WMI, "WMI chan scan list chan[%d] = %u, chan_info->info %8x\n", From c0b0d2e87d91ce283c8766b4b3c2ec9ac90ebf96 Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Mon, 25 Oct 2021 18:54:42 +0530 Subject: [PATCH 0175/2048] ath11k: Increment pending_mgmt_tx count before tx send invoke There is a race condition whereby the tx completion handler can be invoked before the 'num_pending_mgmt_tx" count is incremented. If that occurs, we could get warning trace indicating that 'num_pending_mgmt_tx' is 0 (because it was not yet incremented). Ideally, this trace should be seen only if mgmt tx has not happened but tx completion is received, and it is not expected in this race condition. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1-01386-QCAHKSWPL_SILICONZ-1 Co-developed-by: Lavanya Suresh Signed-off-by: Lavanya Suresh Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635168282-8845-1-git-send-email-quic_ppranees@quicinc.com --- drivers/net/wireless/ath/ath11k/mac.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 5bc82be25316..1320c8634d31 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -5202,13 +5202,15 @@ static void ath11k_mgmt_over_wmi_tx_work(struct work_struct *work) arvif = ath11k_vif_to_arvif(skb_cb->vif); if (ar->allocated_vdev_map & (1LL << arvif->vdev_id) && arvif->is_started) { + atomic_inc(&ar->num_pending_mgmt_tx); ret = ath11k_mac_mgmt_tx_wmi(ar, arvif, skb); if (ret) { + if (atomic_dec_if_positive(&ar->num_pending_mgmt_tx) < 0) + WARN_ON_ONCE(1); + ath11k_warn(ar->ab, "failed to tx mgmt frame, vdev_id %d :%d\n", arvif->vdev_id, ret); ieee80211_free_txskb(ar->hw, skb); - } else { - atomic_inc(&ar->num_pending_mgmt_tx); } } else { ath11k_warn(ar->ab, From c2262123cc49063a8ebc75525d0ab1d13a605608 Mon Sep 17 00:00:00 2001 From: Simon Wunderlich Date: Thu, 2 Sep 2021 09:16:00 +0200 Subject: [PATCH 0176/2048] batman-adv: Start new development cycle This version will contain all the (major or even only minor) changes for Linux 5.17. The version number isn't a semantic version number with major and minor information. It is just encoding the year of the expected publishing as Linux -rc1 and the number of published versions this year (starting at 0). Signed-off-by: Simon Wunderlich --- net/batman-adv/main.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/batman-adv/main.h b/net/batman-adv/main.h index 058b8f2eef65..494d1ebecac2 100644 --- a/net/batman-adv/main.h +++ b/net/batman-adv/main.h @@ -13,7 +13,7 @@ #define BATADV_DRIVER_DEVICE "batman-adv" #ifndef BATADV_SOURCE_VERSION -#define BATADV_SOURCE_VERSION "2021.3" +#define BATADV_SOURCE_VERSION "2022.0" #endif /* B.A.T.M.A.N. parameters */ From 9057d6c23e7388ee9d037fccc9a7bc8557ce277b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Linus=20L=C3=BCssing?= Date: Sun, 31 Oct 2021 22:30:12 +0100 Subject: [PATCH 0177/2048] batman-adv: allow netlink usage in unprivileged containers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently, creating a batman-adv interface in an unprivileged LXD container and attaching secondary interfaces to it with "ip" or "batctl" works fine. However all batctl debug and configuration commands fail: root@container:~# batctl originators Error received: Operation not permitted root@container:~# batctl orig_interval 1000 root@container:~# batctl orig_interval 2000 root@container:~# batctl orig_interval 1000 To fix this change the generic netlink permissions from GENL_ADMIN_PERM to GENL_UNS_ADMIN_PERM. This way a batman-adv interface is fully maintainable as root from within a user namespace, from an unprivileged container. All except one batman-adv netlink setting are per interface and do not leak information or change settings from the host system and are therefore save to retrieve or modify as root from within an unprivileged container. "batctl routing_algo" / BATADV_CMD_GET_ROUTING_ALGOS is the only exception: It provides the batman-adv kernel module wide default routing algorithm. However it is read-only from netlink and an unprivileged container is still not allowed to modify /sys/module/batman_adv/parameters/routing_algo. Instead it is advised to use the newly introduced "batctl if create routing_algo RA_NAME" / IFLA_BATADV_ALGO_NAME to set the routing algorithm on interface creation, which already works fine in an unprivileged container. Cc: Tycho Andersen Signed-off-by: Linus Lüssing Signed-off-by: Sven Eckelmann Signed-off-by: Simon Wunderlich --- net/batman-adv/netlink.c | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) diff --git a/net/batman-adv/netlink.c b/net/batman-adv/netlink.c index 29276284d281..00875e1d8c44 100644 --- a/net/batman-adv/netlink.c +++ b/net/batman-adv/netlink.c @@ -1368,21 +1368,21 @@ static const struct genl_small_ops batadv_netlink_ops[] = { { .cmd = BATADV_CMD_TP_METER, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .doit = batadv_netlink_tp_meter_start, .internal_flags = BATADV_FLAG_NEED_MESH, }, { .cmd = BATADV_CMD_TP_METER_CANCEL, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .doit = batadv_netlink_tp_meter_cancel, .internal_flags = BATADV_FLAG_NEED_MESH, }, { .cmd = BATADV_CMD_GET_ROUTING_ALGOS, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_algo_dump, }, { @@ -1397,68 +1397,68 @@ static const struct genl_small_ops batadv_netlink_ops[] = { { .cmd = BATADV_CMD_GET_TRANSTABLE_LOCAL, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_tt_local_dump, }, { .cmd = BATADV_CMD_GET_TRANSTABLE_GLOBAL, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_tt_global_dump, }, { .cmd = BATADV_CMD_GET_ORIGINATORS, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_orig_dump, }, { .cmd = BATADV_CMD_GET_NEIGHBORS, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_hardif_neigh_dump, }, { .cmd = BATADV_CMD_GET_GATEWAYS, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_gw_dump, }, { .cmd = BATADV_CMD_GET_BLA_CLAIM, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_bla_claim_dump, }, { .cmd = BATADV_CMD_GET_BLA_BACKBONE, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_bla_backbone_dump, }, { .cmd = BATADV_CMD_GET_DAT_CACHE, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_dat_cache_dump, }, { .cmd = BATADV_CMD_GET_MCAST_FLAGS, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .dumpit = batadv_mcast_flags_dump, }, { .cmd = BATADV_CMD_SET_MESH, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .doit = batadv_netlink_set_mesh, .internal_flags = BATADV_FLAG_NEED_MESH, }, { .cmd = BATADV_CMD_SET_HARDIF, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .doit = batadv_netlink_set_hardif, .internal_flags = BATADV_FLAG_NEED_MESH | BATADV_FLAG_NEED_HARDIF, @@ -1474,7 +1474,7 @@ static const struct genl_small_ops batadv_netlink_ops[] = { { .cmd = BATADV_CMD_SET_VLAN, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, - .flags = GENL_ADMIN_PERM, + .flags = GENL_UNS_ADMIN_PERM, .doit = batadv_netlink_set_vlan, .internal_flags = BATADV_FLAG_NEED_MESH | BATADV_FLAG_NEED_VLAN, From ac746c8520d9d056b6963ecca8ff1da9929d02f1 Mon Sep 17 00:00:00 2001 From: Ong Boon Leong Date: Thu, 11 Nov 2021 22:39:49 +0800 Subject: [PATCH 0178/2048] net: stmmac: enhance XDP ZC driver level switching performance The previous stmmac_xdp_set_prog() implementation uses stmmac_release() and stmmac_open() which tear down the PHY device and causes undesirable autonegotiation which causes a delay whenever AFXDP ZC is setup. This patch introduces two new functions that just sufficiently tear down DMA descriptors, buffer, NAPI process, and IRQs and reestablish them accordingly in both stmmac_xdp_release() and stammac_xdp_open(). As the results of this enhancement, we get rid of transient state introduced by the link auto-negotiation: $ ./xdpsock -i eth0 -t -z sock0@eth0:0 txonly xdp-drv pps pkts 1.00 rx 0 0 tx 634444 634560 sock0@eth0:0 txonly xdp-drv pps pkts 1.00 rx 0 0 tx 632330 1267072 sock0@eth0:0 txonly xdp-drv pps pkts 1.00 rx 0 0 tx 632438 1899584 sock0@eth0:0 txonly xdp-drv pps pkts 1.00 rx 0 0 tx 632502 2532160 Reported-by: Kurt Kanzenbach Signed-off-by: Ong Boon Leong Tested-by: Kurt Kanzenbach Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac.h | 4 +- .../net/ethernet/stmicro/stmmac/stmmac_main.c | 137 +++++++++++++++++- .../net/ethernet/stmicro/stmmac/stmmac_xdp.c | 4 +- 3 files changed, 139 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h index 43eead726886..dd7adf9b2537 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h @@ -316,8 +316,8 @@ void stmmac_set_ethtool_ops(struct net_device *netdev); void stmmac_ptp_register(struct stmmac_priv *priv); void stmmac_ptp_unregister(struct stmmac_priv *priv); -int stmmac_open(struct net_device *dev); -int stmmac_release(struct net_device *dev); +int stmmac_xdp_open(struct net_device *dev); +void stmmac_xdp_release(struct net_device *dev); int stmmac_resume(struct device *dev); int stmmac_suspend(struct device *dev); int stmmac_dvr_remove(struct device *dev); diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index d3f350c25b9b..033c35c09a54 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -3643,7 +3643,7 @@ static int stmmac_request_irq(struct net_device *dev) * 0 on success and an appropriate (-)ve integer as defined in errno.h * file on failure. */ -int stmmac_open(struct net_device *dev) +static int stmmac_open(struct net_device *dev) { struct stmmac_priv *priv = netdev_priv(dev); int mode = priv->plat->phy_interface; @@ -3767,7 +3767,7 @@ static void stmmac_fpe_stop_wq(struct stmmac_priv *priv) * Description: * This is the stop entry point of the driver. */ -int stmmac_release(struct net_device *dev) +static int stmmac_release(struct net_device *dev) { struct stmmac_priv *priv = netdev_priv(dev); u32 chan; @@ -6429,6 +6429,139 @@ void stmmac_enable_tx_queue(struct stmmac_priv *priv, u32 queue) spin_unlock_irqrestore(&ch->lock, flags); } +void stmmac_xdp_release(struct net_device *dev) +{ + struct stmmac_priv *priv = netdev_priv(dev); + u32 chan; + + /* Disable NAPI process */ + stmmac_disable_all_queues(priv); + + for (chan = 0; chan < priv->plat->tx_queues_to_use; chan++) + hrtimer_cancel(&priv->tx_queue[chan].txtimer); + + /* Free the IRQ lines */ + stmmac_free_irq(dev, REQ_IRQ_ERR_ALL, 0); + + /* Stop TX/RX DMA channels */ + stmmac_stop_all_dma(priv); + + /* Release and free the Rx/Tx resources */ + free_dma_desc_resources(priv); + + /* Disable the MAC Rx/Tx */ + stmmac_mac_set(priv, priv->ioaddr, false); + + /* set trans_start so we don't get spurious + * watchdogs during reset + */ + netif_trans_update(dev); + netif_carrier_off(dev); +} + +int stmmac_xdp_open(struct net_device *dev) +{ + struct stmmac_priv *priv = netdev_priv(dev); + u32 rx_cnt = priv->plat->rx_queues_to_use; + u32 tx_cnt = priv->plat->tx_queues_to_use; + u32 dma_csr_ch = max(rx_cnt, tx_cnt); + struct stmmac_rx_queue *rx_q; + struct stmmac_tx_queue *tx_q; + u32 buf_size; + bool sph_en; + u32 chan; + int ret; + + ret = alloc_dma_desc_resources(priv); + if (ret < 0) { + netdev_err(dev, "%s: DMA descriptors allocation failed\n", + __func__); + goto dma_desc_error; + } + + ret = init_dma_desc_rings(dev, GFP_KERNEL); + if (ret < 0) { + netdev_err(dev, "%s: DMA descriptors initialization failed\n", + __func__); + goto init_error; + } + + /* DMA CSR Channel configuration */ + for (chan = 0; chan < dma_csr_ch; chan++) + stmmac_init_chan(priv, priv->ioaddr, priv->plat->dma_cfg, chan); + + /* Adjust Split header */ + sph_en = (priv->hw->rx_csum > 0) && priv->sph; + + /* DMA RX Channel Configuration */ + for (chan = 0; chan < rx_cnt; chan++) { + rx_q = &priv->rx_queue[chan]; + + stmmac_init_rx_chan(priv, priv->ioaddr, priv->plat->dma_cfg, + rx_q->dma_rx_phy, chan); + + rx_q->rx_tail_addr = rx_q->dma_rx_phy + + (rx_q->buf_alloc_num * + sizeof(struct dma_desc)); + stmmac_set_rx_tail_ptr(priv, priv->ioaddr, + rx_q->rx_tail_addr, chan); + + if (rx_q->xsk_pool && rx_q->buf_alloc_num) { + buf_size = xsk_pool_get_rx_frame_size(rx_q->xsk_pool); + stmmac_set_dma_bfsize(priv, priv->ioaddr, + buf_size, + rx_q->queue_index); + } else { + stmmac_set_dma_bfsize(priv, priv->ioaddr, + priv->dma_buf_sz, + rx_q->queue_index); + } + + stmmac_enable_sph(priv, priv->ioaddr, sph_en, chan); + } + + /* DMA TX Channel Configuration */ + for (chan = 0; chan < tx_cnt; chan++) { + tx_q = &priv->tx_queue[chan]; + + stmmac_init_tx_chan(priv, priv->ioaddr, priv->plat->dma_cfg, + tx_q->dma_tx_phy, chan); + + tx_q->tx_tail_addr = tx_q->dma_tx_phy; + stmmac_set_tx_tail_ptr(priv, priv->ioaddr, + tx_q->tx_tail_addr, chan); + } + + /* Enable the MAC Rx/Tx */ + stmmac_mac_set(priv, priv->ioaddr, true); + + /* Start Rx & Tx DMA Channels */ + stmmac_start_all_dma(priv); + + stmmac_init_coalesce(priv); + + ret = stmmac_request_irq(dev); + if (ret) + goto irq_error; + + /* Enable NAPI process*/ + stmmac_enable_all_queues(priv); + netif_carrier_on(dev); + netif_tx_start_all_queues(dev); + + return 0; + +irq_error: + for (chan = 0; chan < priv->plat->tx_queues_to_use; chan++) + hrtimer_cancel(&priv->tx_queue[chan].txtimer); + + stmmac_hw_teardown(dev); +init_error: + free_dma_desc_resources(priv); +dma_desc_error: + return ret; +} + int stmmac_xsk_wakeup(struct net_device *dev, u32 queue, u32 flags) { struct stmmac_priv *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c index 2a616c6f7cd0..9d4d8c3dad0a 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_xdp.c @@ -119,7 +119,7 @@ int stmmac_xdp_set_prog(struct stmmac_priv *priv, struct bpf_prog *prog, need_update = !!priv->xdp_prog != !!prog; if (if_running && need_update) - stmmac_release(dev); + stmmac_xdp_release(dev); old_prog = xchg(&priv->xdp_prog, prog); if (old_prog) @@ -129,7 +129,7 @@ int stmmac_xdp_set_prog(struct stmmac_priv *priv, struct bpf_prog *prog, priv->sph = priv->sph_cap && !stmmac_xdp_is_enabled(priv); if (if_running && need_update) - stmmac_open(dev); + stmmac_xdp_open(dev); return 0; } From e99fa4230fa89f9e76073dd6af3dd85c33ba70b8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Fri, 12 Nov 2021 15:53:52 +0100 Subject: [PATCH 0179/2048] net: dsa: vsc73xxx: Make vsc73xx_remove() return void MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit vsc73xx_remove() returns zero unconditionally and no caller checks the returned value. So convert the function to return no value. Signed-off-by: Uwe Kleine-König Reviewed-by: Vladimir Oltean Signed-off-by: David S. Miller --- drivers/net/dsa/vitesse-vsc73xx-core.c | 4 +--- drivers/net/dsa/vitesse-vsc73xx.h | 2 +- 2 files changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c index a4b1447ff055..4c18f619ec02 100644 --- a/drivers/net/dsa/vitesse-vsc73xx-core.c +++ b/drivers/net/dsa/vitesse-vsc73xx-core.c @@ -1216,12 +1216,10 @@ int vsc73xx_probe(struct vsc73xx *vsc) } EXPORT_SYMBOL(vsc73xx_probe); -int vsc73xx_remove(struct vsc73xx *vsc) +void vsc73xx_remove(struct vsc73xx *vsc) { dsa_unregister_switch(vsc->ds); gpiod_set_value(vsc->reset, 1); - - return 0; } EXPORT_SYMBOL(vsc73xx_remove); diff --git a/drivers/net/dsa/vitesse-vsc73xx.h b/drivers/net/dsa/vitesse-vsc73xx.h index 30b951504e65..30b1f0a36566 100644 --- a/drivers/net/dsa/vitesse-vsc73xx.h +++ b/drivers/net/dsa/vitesse-vsc73xx.h @@ -26,5 +26,5 @@ struct vsc73xx_ops { int vsc73xx_is_addr_valid(u8 block, u8 subblock); int vsc73xx_probe(struct vsc73xx *vsc); -int vsc73xx_remove(struct vsc73xx *vsc); +void vsc73xx_remove(struct vsc73xx *vsc); void vsc73xx_shutdown(struct vsc73xx *vsc); From cb3ef7b00042479277cda7871d899378ad91f081 Mon Sep 17 00:00:00 2001 From: Harshit Mogalapalli Date: Fri, 12 Nov 2021 13:36:47 -0800 Subject: [PATCH 0180/2048] net: sched: sch_netem: Refactor code in 4-state loss generator Fixed comments to match description with variable names and refactored code to match the convention as per [1]. To match the convention mapping is done as follows: State 3 - LOST_IN_BURST_PERIOD State 4 - LOST_IN_GAP_PERIOD [1] S. Salsano, F. Ludovici, A. Ordine, "Definition of a general and intuitive loss model for packet networks and its implementation in the Netem module in the Linux kernel" Fixes: a6e2fe17eba4 ("sch_netem: replace magic numbers with enumerate") Signed-off-by: Harshit Mogalapalli Acked-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/sched/sch_netem.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index ecbb10db1111..ed4ccef5d6a8 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -208,17 +208,17 @@ static bool loss_4state(struct netem_sched_data *q) * next state and if the next packet has to be transmitted or lost. * The four states correspond to: * TX_IN_GAP_PERIOD => successfully transmitted packets within a gap period - * LOST_IN_BURST_PERIOD => isolated losses within a gap period - * LOST_IN_GAP_PERIOD => lost packets within a burst period - * TX_IN_GAP_PERIOD => successfully transmitted packets within a burst period + * LOST_IN_GAP_PERIOD => isolated losses within a gap period + * LOST_IN_BURST_PERIOD => lost packets within a burst period + * TX_IN_BURST_PERIOD => successfully transmitted packets within a burst period */ switch (clg->state) { case TX_IN_GAP_PERIOD: if (rnd < clg->a4) { - clg->state = LOST_IN_BURST_PERIOD; + clg->state = LOST_IN_GAP_PERIOD; return true; } else if (clg->a4 < rnd && rnd < clg->a1 + clg->a4) { - clg->state = LOST_IN_GAP_PERIOD; + clg->state = LOST_IN_BURST_PERIOD; return true; } else if (clg->a1 + clg->a4 < rnd) { clg->state = TX_IN_GAP_PERIOD; @@ -227,24 +227,24 @@ static bool loss_4state(struct netem_sched_data *q) break; case TX_IN_BURST_PERIOD: if (rnd < clg->a5) { - clg->state = LOST_IN_GAP_PERIOD; + clg->state = LOST_IN_BURST_PERIOD; return true; } else { clg->state = TX_IN_BURST_PERIOD; } break; - case LOST_IN_GAP_PERIOD: + case LOST_IN_BURST_PERIOD: if (rnd < clg->a3) clg->state = TX_IN_BURST_PERIOD; else if (clg->a3 < rnd && rnd < clg->a2 + clg->a3) { clg->state = TX_IN_GAP_PERIOD; } else if (clg->a2 + clg->a3 < rnd) { - clg->state = LOST_IN_GAP_PERIOD; + clg->state = LOST_IN_BURST_PERIOD; return true; } break; - case LOST_IN_BURST_PERIOD: + case LOST_IN_GAP_PERIOD: clg->state = TX_IN_GAP_PERIOD; break; } From 02d6fdecb9c38de19065f6bed8d5214556fd061d Mon Sep 17 00:00:00 2001 From: Ansuel Smith Date: Thu, 4 Nov 2021 16:00:40 +0100 Subject: [PATCH 0181/2048] regmap: allow to define reg_update_bits for no bus configuration Some device requires a special handling for reg_update_bits and can't use the normal regmap read write logic. An example is when locking is handled by the device and rmw operations requires to do atomic operations. Allow to declare a dedicated function in regmap_config for reg_update_bits in no bus configuration. Signed-off-by: Ansuel Smith Link: https://lore.kernel.org/r/20211104150040.1260-1-ansuelsmth@gmail.com Signed-off-by: Mark Brown --- drivers/base/regmap/regmap.c | 1 + include/linux/regmap.h | 7 +++++++ 2 files changed, 8 insertions(+) diff --git a/drivers/base/regmap/regmap.c b/drivers/base/regmap/regmap.c index 21a0c2562ec0..2d74f9f82aa9 100644 --- a/drivers/base/regmap/regmap.c +++ b/drivers/base/regmap/regmap.c @@ -876,6 +876,7 @@ struct regmap *__regmap_init(struct device *dev, if (!bus) { map->reg_read = config->reg_read; map->reg_write = config->reg_write; + map->reg_update_bits = config->reg_update_bits; map->defer_caching = false; goto skip_format_initialization; diff --git a/include/linux/regmap.h b/include/linux/regmap.h index e3c9a25a853a..22652e5fbc38 100644 --- a/include/linux/regmap.h +++ b/include/linux/regmap.h @@ -290,6 +290,11 @@ typedef void (*regmap_unlock)(void *); * read operation on a bus such as SPI, I2C, etc. Most of the * devices do not need this. * @reg_write: Same as above for writing. + * @reg_update_bits: Optional callback that if filled will be used to perform + * all the update_bits(rmw) operation. Should only be provided + * if the function require special handling with lock and reg + * handling and the operation cannot be represented as a simple + * update_bits operation on a bus such as SPI, I2C, etc. * @fast_io: Register IO is fast. Use a spinlock instead of a mutex * to perform locking. This field is ignored if custom lock/unlock * functions are used (see fields lock/unlock of struct regmap_config). @@ -372,6 +377,8 @@ struct regmap_config { int (*reg_read)(void *context, unsigned int reg, unsigned int *val); int (*reg_write)(void *context, unsigned int reg, unsigned int val); + int (*reg_update_bits)(void *context, unsigned int reg, + unsigned int mask, unsigned int val); bool fast_io; From cc0be1ad686fb29a4d127948486f40b17fb34b50 Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Sun, 14 Nov 2021 20:02:35 +0100 Subject: [PATCH 0182/2048] net: bridge: Slightly optimize 'find_portno()' The 'inuse' bitmap is local to this function. So we can use the non-atomic '__set_bit()' to save a few cycles. While at it, also remove some useless {}. Signed-off-by: Christophe JAILLET Acked-by: Nikolay Aleksandrov Signed-off-by: David S. Miller --- net/bridge/br_if.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c index c1183fef1f21..64b2d4fb50f5 100644 --- a/net/bridge/br_if.c +++ b/net/bridge/br_if.c @@ -397,10 +397,10 @@ static int find_portno(struct net_bridge *br) if (!inuse) return -ENOMEM; - set_bit(0, inuse); /* zero is reserved */ - list_for_each_entry(p, &br->port_list, list) { - set_bit(p->port_no, inuse); - } + __set_bit(0, inuse); /* zero is reserved */ + list_for_each_entry(p, &br->port_list, list) + __set_bit(p->port_no, inuse); + index = find_first_zero_bit(inuse, BR_MAX_PORTS); bitmap_free(inuse); From 13cae4a104d2b7205696229ba85d34cc035f8c84 Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:21 +0800 Subject: [PATCH 0183/2048] i2c: core: Allow 255 byte transfers for SMBus 3.x SMBus 3.0 increased the maximum block transfer size from 32 bytes to 255 bytes. We increase the size of struct i2c_smbus_data's block[] member. i2c_smbus_xfer() and i2c_smbus_xfer_emulated() now support 255 byte block operations, other block functions remain limited to 32 bytes for compatibility with existing callers. We allow adapters to indicate support for the larger size with I2C_FUNC_SMBUS_V3_BLOCK. Most emulated drivers should be able to use 255 byte blocks by replacing I2C_SMBUS_BLOCK_MAX with I2C_SMBUS_V3_BLOCK_MAX though some will have hardware limitations that need testing. Signed-off-by: Matt Johnston Signed-off-by: David S. Miller --- drivers/i2c/i2c-core-smbus.c | 20 +++++++++++++------- include/linux/i2c.h | 13 +++++++++++++ include/uapi/linux/i2c.h | 5 ++++- 3 files changed, 30 insertions(+), 8 deletions(-) diff --git a/drivers/i2c/i2c-core-smbus.c b/drivers/i2c/i2c-core-smbus.c index e5b2d1465e7e..743415584aba 100644 --- a/drivers/i2c/i2c-core-smbus.c +++ b/drivers/i2c/i2c-core-smbus.c @@ -303,7 +303,8 @@ static void i2c_smbus_try_get_dmabuf(struct i2c_msg *msg, u8 init_val) bool is_read = msg->flags & I2C_M_RD; unsigned char *dma_buf; - dma_buf = kzalloc(I2C_SMBUS_BLOCK_MAX + (is_read ? 2 : 3), GFP_KERNEL); + dma_buf = kzalloc(I2C_SMBUS_V3_BLOCK_MAX + (is_read ? 2 : 3), + GFP_KERNEL); if (!dma_buf) return; @@ -329,9 +330,10 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, * initialize most things with sane defaults, to keep the code below * somewhat simpler. */ - unsigned char msgbuf0[I2C_SMBUS_BLOCK_MAX+3]; - unsigned char msgbuf1[I2C_SMBUS_BLOCK_MAX+2]; + unsigned char msgbuf0[I2C_SMBUS_V3_BLOCK_MAX+3]; + unsigned char msgbuf1[I2C_SMBUS_V3_BLOCK_MAX+2]; int nmsgs = read_write == I2C_SMBUS_READ ? 2 : 1; + u16 block_max; u8 partial_pec = 0; int status; struct i2c_msg msg[2] = { @@ -350,6 +352,10 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, bool wants_pec = ((flags & I2C_CLIENT_PEC) && size != I2C_SMBUS_QUICK && size != I2C_SMBUS_I2C_BLOCK_DATA); + /* Drivers must opt in to 255 byte max block size */ + block_max = i2c_check_functionality(adapter, I2C_FUNC_SMBUS_V3_BLOCK) + ? I2C_SMBUS_V3_BLOCK_MAX : I2C_SMBUS_BLOCK_MAX; + msgbuf0[0] = command; switch (size) { case I2C_SMBUS_QUICK: @@ -399,7 +405,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, i2c_smbus_try_get_dmabuf(&msg[1], 0); } else { msg[0].len = data->block[0] + 2; - if (msg[0].len > I2C_SMBUS_BLOCK_MAX + 2) { + if (msg[0].len > block_max + 2) { dev_err(&adapter->dev, "Invalid block write size %d\n", data->block[0]); @@ -413,7 +419,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, case I2C_SMBUS_BLOCK_PROC_CALL: nmsgs = 2; /* Another special case */ read_write = I2C_SMBUS_READ; - if (data->block[0] > I2C_SMBUS_BLOCK_MAX) { + if (data->block[0] > block_max) { dev_err(&adapter->dev, "Invalid block write size %d\n", data->block[0]); @@ -430,7 +436,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, i2c_smbus_try_get_dmabuf(&msg[1], 0); break; case I2C_SMBUS_I2C_BLOCK_DATA: - if (data->block[0] > I2C_SMBUS_BLOCK_MAX) { + if (data->block[0] > block_max) { dev_err(&adapter->dev, "Invalid block %s size %d\n", read_write == I2C_SMBUS_READ ? "read" : "write", data->block[0]); @@ -498,7 +504,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, break; case I2C_SMBUS_BLOCK_DATA: case I2C_SMBUS_BLOCK_PROC_CALL: - if (msg[1].buf[0] > I2C_SMBUS_BLOCK_MAX) { + if (msg[1].buf[0] > block_max) { dev_err(&adapter->dev, "Invalid block size returned: %d\n", msg[1].buf[0]); diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 16119ac1aa97..353d6b4e7a53 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -52,6 +52,19 @@ typedef int (*i2c_slave_cb_t)(struct i2c_client *client, struct module; struct property_entry; +/* SMBus 3.0 extends the maximum block read/write size to 255 (from 32). + * The larger size is only supported by some drivers, indicated by + * the I2C_FUNC_SMBUS_V3_BLOCK functionality bit. + */ +#define I2C_SMBUS_V3_BLOCK_MAX 255 /* As specified in SMBus 3.0 standard */ + +/* Note compatibility definition in uapi header with 32 byte block */ +union i2c_smbus_data { + __u8 byte; + __u16 word; + __u8 block[I2C_SMBUS_V3_BLOCK_MAX + 1]; /* block[0] is used for length */ +}; + #if IS_ENABLED(CONFIG_I2C) /* Return the Frequency mode string based on the bus frequency */ const char *i2c_freq_mode_string(u32 bus_freq_hz); diff --git a/include/uapi/linux/i2c.h b/include/uapi/linux/i2c.h index 92326ebde350..7b7d90b50cf0 100644 --- a/include/uapi/linux/i2c.h +++ b/include/uapi/linux/i2c.h @@ -108,6 +108,7 @@ struct i2c_msg { #define I2C_FUNC_SMBUS_READ_I2C_BLOCK 0x04000000 /* I2C-like block xfer */ #define I2C_FUNC_SMBUS_WRITE_I2C_BLOCK 0x08000000 /* w/ 1-byte reg. addr. */ #define I2C_FUNC_SMBUS_HOST_NOTIFY 0x10000000 /* SMBus 2.0 or later */ +#define I2C_FUNC_SMBUS_V3_BLOCK 0x20000000 /* Device supports 255 byte block */ #define I2C_FUNC_SMBUS_BYTE (I2C_FUNC_SMBUS_READ_BYTE | \ I2C_FUNC_SMBUS_WRITE_BYTE) @@ -137,13 +138,15 @@ struct i2c_msg { /* * Data for SMBus Messages */ -#define I2C_SMBUS_BLOCK_MAX 32 /* As specified in SMBus standard */ +#define I2C_SMBUS_BLOCK_MAX 32 /* As specified in SMBus 2.0 standard */ +#ifndef __KERNEL__ union i2c_smbus_data { __u8 byte; __u16 word; __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ /* and one more for user-space compatibility */ }; +#endif /* i2c_smbus_xfer read or write markers */ #define I2C_SMBUS_READ 1 From 84a107e68b34217eff536e81a6a6f419ee0d0f7e Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:22 +0800 Subject: [PATCH 0184/2048] i2c: dev: Handle 255 byte blocks for i2c ioctl I2C_SMBUS is limited to 32 bytes due to compatibility with the 32 byte i2c_smbus_data.block I2C_RDWR allows larger transfers if sufficient sized buffers are passed. Signed-off-by: Matt Johnston Signed-off-by: David S. Miller --- drivers/i2c/i2c-dev.c | 93 ++++++++++++++++++++++++++++++------ include/uapi/linux/i2c-dev.h | 2 + include/uapi/linux/i2c.h | 2 + 3 files changed, 83 insertions(+), 14 deletions(-) diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c index bce0e8bb7852..5ee9118c0407 100644 --- a/drivers/i2c/i2c-dev.c +++ b/drivers/i2c/i2c-dev.c @@ -46,6 +46,24 @@ struct i2c_dev { struct cdev cdev; }; +/* The userspace union i2c_smbus_data for I2C_SMBUS ioctl is limited + * to 32 bytes (I2C_SMBUS_BLOCK_MAX) for compatibility. + */ +union compat_i2c_smbus_data { + __u8 byte; + __u16 word; + __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ + /* and one more for user-space compatibility */ +}; + +/* Must match i2c-dev.h definition with compat .data member */ +struct i2c_smbus_ioctl_data { + __u8 read_write; + __u8 command; + __u32 size; + union compat_i2c_smbus_data __user *data; +}; + #define I2C_MINORS (MINORMASK + 1) static LIST_HEAD(i2c_dev_list); static DEFINE_SPINLOCK(i2c_dev_list_lock); @@ -235,14 +253,17 @@ static int i2cdev_check_addr(struct i2c_adapter *adapter, unsigned int addr) static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, unsigned nmsgs, struct i2c_msg *msgs) { - u8 __user **data_ptrs; + u8 __user **data_ptrs = NULL; + u16 *orig_lens = NULL; int i, res; + res = -ENOMEM; data_ptrs = kmalloc_array(nmsgs, sizeof(u8 __user *), GFP_KERNEL); - if (data_ptrs == NULL) { - kfree(msgs); - return -ENOMEM; - } + if (data_ptrs == NULL) + goto out; + orig_lens = kmalloc_array(nmsgs, sizeof(u16), GFP_KERNEL); + if (orig_lens == NULL) + goto out; res = 0; for (i = 0; i < nmsgs; i++) { @@ -253,12 +274,30 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, } data_ptrs[i] = (u8 __user *)msgs[i].buf; - msgs[i].buf = memdup_user(data_ptrs[i], msgs[i].len); + msgs[i].buf = NULL; + if (msgs[i].len < 1) { + /* Sanity check */ + res = -EINVAL; + break; + + } + /* Allocate a larger buffer to accommodate possible 255 byte + * blocks. Read results will be dropped later + * if they are too large for the original length. + */ + orig_lens[i] = msgs[i].len; + msgs[i].buf = kmalloc(msgs[i].len + I2C_SMBUS_V3_BLOCK_MAX, + GFP_USER | __GFP_NOWARN); if (IS_ERR(msgs[i].buf)) { res = PTR_ERR(msgs[i].buf); break; } - /* memdup_user allocates with GFP_KERNEL, so DMA is ok */ + if (copy_from_user(msgs[i].buf, data_ptrs[i], msgs[i].len)) { + kfree(msgs[i].buf); + res = -EFAULT; + break; + } + /* Buffer from kmalloc, so DMA is ok */ msgs[i].flags |= I2C_M_DMA_SAFE; /* @@ -274,7 +313,7 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, */ if (msgs[i].flags & I2C_M_RECV_LEN) { if (!(msgs[i].flags & I2C_M_RD) || - msgs[i].len < 1 || msgs[i].buf[0] < 1 || + msgs[i].buf[0] < 1 || msgs[i].len < msgs[i].buf[0] + I2C_SMBUS_BLOCK_MAX) { i++; @@ -297,12 +336,16 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, res = i2c_transfer(client->adapter, msgs, nmsgs); while (i-- > 0) { if (res >= 0 && (msgs[i].flags & I2C_M_RD)) { - if (copy_to_user(data_ptrs[i], msgs[i].buf, - msgs[i].len)) + if (orig_lens[i] < msgs[i].len) + res = -EINVAL; + else if (copy_to_user(data_ptrs[i], msgs[i].buf, + msgs[i].len)) res = -EFAULT; } kfree(msgs[i].buf); } +out: + kfree(orig_lens); kfree(data_ptrs); kfree(msgs); return res; @@ -310,7 +353,7 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, u8 read_write, u8 command, u32 size, - union i2c_smbus_data __user *data) + union compat_i2c_smbus_data __user *data) { union i2c_smbus_data temp = {}; int datasize, res; @@ -371,6 +414,16 @@ static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, if (copy_from_user(&temp, data, datasize)) return -EFAULT; } + if ((size == I2C_SMBUS_BLOCK_PROC_CALL || + size == I2C_SMBUS_I2C_BLOCK_DATA || + size == I2C_SMBUS_BLOCK_DATA) && + read_write == I2C_SMBUS_WRITE && + temp.block[0] > I2C_SMBUS_BLOCK_MAX) { + /* Don't accept writes larger than the buffer size */ + dev_dbg(&client->adapter->dev, "block write is too large"); + return -EINVAL; + + } if (size == I2C_SMBUS_I2C_BLOCK_BROKEN) { /* Convert old I2C block commands to the new convention. This preserves binary compatibility. */ @@ -380,9 +433,21 @@ static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, } res = i2c_smbus_xfer(client->adapter, client->addr, client->flags, read_write, command, size, &temp); - if (!res && ((size == I2C_SMBUS_PROC_CALL) || - (size == I2C_SMBUS_BLOCK_PROC_CALL) || - (read_write == I2C_SMBUS_READ))) { + if (res) + return res; + if ((size == I2C_SMBUS_BLOCK_PROC_CALL || + size == I2C_SMBUS_I2C_BLOCK_DATA || + size == I2C_SMBUS_BLOCK_DATA) && + read_write == I2C_SMBUS_READ && + temp.block[0] > I2C_SMBUS_BLOCK_MAX) { + /* Don't accept reads larger than the buffer size */ + dev_dbg(&client->adapter->dev, "block read is too large"); + return -EINVAL; + + } + if ((size == I2C_SMBUS_PROC_CALL) || + (size == I2C_SMBUS_BLOCK_PROC_CALL) || + (read_write == I2C_SMBUS_READ)) { if (copy_to_user(data, &temp, datasize)) return -EFAULT; } diff --git a/include/uapi/linux/i2c-dev.h b/include/uapi/linux/i2c-dev.h index 1c4cec4ddd84..46ce31d42f7d 100644 --- a/include/uapi/linux/i2c-dev.h +++ b/include/uapi/linux/i2c-dev.h @@ -39,12 +39,14 @@ /* This is the structure as used in the I2C_SMBUS ioctl call */ +#ifndef __KERNEL__ struct i2c_smbus_ioctl_data { __u8 read_write; __u8 command; __u32 size; union i2c_smbus_data __user *data; }; +#endif /* This is the structure as used in the I2C_RDWR ioctl call */ struct i2c_rdwr_ioctl_data { diff --git a/include/uapi/linux/i2c.h b/include/uapi/linux/i2c.h index 7b7d90b50cf0..c3534ab1ae53 100644 --- a/include/uapi/linux/i2c.h +++ b/include/uapi/linux/i2c.h @@ -109,6 +109,8 @@ struct i2c_msg { #define I2C_FUNC_SMBUS_WRITE_I2C_BLOCK 0x08000000 /* w/ 1-byte reg. addr. */ #define I2C_FUNC_SMBUS_HOST_NOTIFY 0x10000000 /* SMBus 2.0 or later */ #define I2C_FUNC_SMBUS_V3_BLOCK 0x20000000 /* Device supports 255 byte block */ + /* Note that I2C_SMBUS ioctl only */ + /* supports a 32 byte block */ #define I2C_FUNC_SMBUS_BYTE (I2C_FUNC_SMBUS_READ_BYTE | \ I2C_FUNC_SMBUS_WRITE_BYTE) From 1b2ba1f591c9d083cc1928f8337fcbf6a749e935 Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:23 +0800 Subject: [PATCH 0185/2048] i2c: aspeed: Allow 255 byte block transfers 255 byte transfers have been tested on an AST2500 board Signed-off-by: Matt Johnston Reviewed-by: Brendan Higgins Signed-off-by: David S. Miller --- drivers/i2c/busses/i2c-aspeed.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/i2c/busses/i2c-aspeed.c b/drivers/i2c/busses/i2c-aspeed.c index 67e8b97c0c95..7395f3702fae 100644 --- a/drivers/i2c/busses/i2c-aspeed.c +++ b/drivers/i2c/busses/i2c-aspeed.c @@ -533,7 +533,7 @@ static u32 aspeed_i2c_master_irq(struct aspeed_i2c_bus *bus, u32 irq_status) msg->buf[bus->buf_index++] = recv_byte; if (msg->flags & I2C_M_RECV_LEN) { - if (unlikely(recv_byte > I2C_SMBUS_BLOCK_MAX)) { + if (unlikely(recv_byte > I2C_SMBUS_V3_BLOCK_MAX)) { bus->cmd_err = -EPROTO; aspeed_i2c_do_stop(bus); goto out_no_complete; @@ -718,7 +718,8 @@ static int aspeed_i2c_master_xfer(struct i2c_adapter *adap, static u32 aspeed_i2c_functionality(struct i2c_adapter *adap) { - return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL | I2C_FUNC_SMBUS_BLOCK_DATA; + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL | + I2C_FUNC_SMBUS_BLOCK_DATA | I2C_FUNC_SMBUS_V3_BLOCK; } #if IS_ENABLED(CONFIG_I2C_SLAVE) From 3ef2de27a05a2dc7425431d9aa780e08e56ddb7f Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:24 +0800 Subject: [PATCH 0186/2048] i2c: npcm7xx: Allow 255 byte block SMBus transfers 255 byte support has been tested on a npcm750 board Signed-off-by: Matt Johnston Reviewed-by: Tali Perry Reviewed-by: Patrick Venture Signed-off-by: David S. Miller --- drivers/i2c/busses/i2c-npcm7xx.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-npcm7xx.c b/drivers/i2c/busses/i2c-npcm7xx.c index 2ad166355ec9..6d60f65add85 100644 --- a/drivers/i2c/busses/i2c-npcm7xx.c +++ b/drivers/i2c/busses/i2c-npcm7xx.c @@ -1399,7 +1399,7 @@ static void npcm_i2c_irq_master_handler_read(struct npcm_i2c *bus) if (bus->read_block_use) { /* first byte in block protocol is the size: */ data = npcm_i2c_rd_byte(bus); - data = clamp_val(data, 1, I2C_SMBUS_BLOCK_MAX); + data = clamp_val(data, 1, I2C_SMBUS_V3_BLOCK_MAX); bus->rd_size = data + block_extra_bytes_size; bus->rd_buf[bus->rd_ind++] = data; @@ -2187,6 +2187,7 @@ static u32 npcm_i2c_functionality(struct i2c_adapter *adap) I2C_FUNC_SMBUS_EMUL | I2C_FUNC_SMBUS_BLOCK_DATA | I2C_FUNC_SMBUS_PEC | + I2C_FUNC_SMBUS_V3_BLOCK | I2C_FUNC_SLAVE; } From 0b6141eb2b14b0df03f48bd129182900078e7f90 Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:25 +0800 Subject: [PATCH 0187/2048] dt-bindings: net: New binding mctp-i2c-controller Used to define a local endpoint to communicate with MCTP peripherals attached to an I2C bus. This I2C endpoint can communicate with remote MCTP devices on the I2C bus. In the example I2C topology below (matching the second yaml example) we have MCTP devices on busses i2c1 and i2c6. MCTP-supporting busses are indicated by the 'mctp-controller' DT property on an I2C bus node. A mctp-i2c-controller I2C client DT node is placed at the top of the mux topology, since only the root I2C adapter will support I2C slave functionality. .-------. |eeprom | .------------. .------. /'-------' | adapter | | mux --@0,i2c5------' | i2c1 ----.*| --@1,i2c6--.--. |............| \'------' \ \ ......... | mctp-i2c- | \ \ \ .mctpB . | controller | \ \ '.0x30 . | | \ ......... \ '.......' | 0x50 | \ .mctpA . \ ......... '------------' '.0x1d . '.mctpC . '.......' '.0x31 . '.......' (mctpX boxes above are remote MCTP devices not included in the DT at present, they can be hotplugged/probed at runtime. A DT binding for specific fixed MCTP devices could be added later if required) Signed-off-by: Matt Johnston Reviewed-by: Rob Herring Signed-off-by: David S. Miller --- Documentation/devicetree/bindings/i2c/i2c.txt | 4 + .../bindings/net/mctp-i2c-controller.yaml | 92 +++++++++++++++++++ 2 files changed, 96 insertions(+) create mode 100644 Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml diff --git a/Documentation/devicetree/bindings/i2c/i2c.txt b/Documentation/devicetree/bindings/i2c/i2c.txt index b864916e087f..fc3dd7ec0445 100644 --- a/Documentation/devicetree/bindings/i2c/i2c.txt +++ b/Documentation/devicetree/bindings/i2c/i2c.txt @@ -95,6 +95,10 @@ wants to support one of the below features, it should adapt these bindings. - smbus-alert states that the optional SMBus-Alert feature apply to this bus. +- mctp-controller + indicates that the system is accessible via this bus as an endpoint for + MCTP over I2C transport. + Required properties (per child device) -------------------------------------- diff --git a/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml b/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml new file mode 100644 index 000000000000..afd11c9422fa --- /dev/null +++ b/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml @@ -0,0 +1,92 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/net/mctp-i2c-controller.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: MCTP I2C transport binding + +maintainers: + - Matt Johnston + +description: | + An mctp-i2c-controller defines a local MCTP endpoint on an I2C controller. + MCTP I2C is specified by DMTF DSP0237. + + An mctp-i2c-controller must be attached to an I2C adapter which supports + slave functionality. I2C busses (either directly or as subordinate mux + busses) are attached to the mctp-i2c-controller with a 'mctp-controller' + property on each used bus. Each mctp-controller I2C bus will be presented + to the host system as a separate MCTP I2C instance. + +properties: + compatible: + const: mctp-i2c-controller + + reg: + minimum: 0x40000000 + maximum: 0x4000007f + description: | + 7 bit I2C address of the local endpoint. + I2C_OWN_SLAVE_ADDRESS (1<<30) flag must be set. + +additionalProperties: false + +required: + - compatible + - reg + +examples: + - | + // Basic case of a single I2C bus + #include + + i2c { + #address-cells = <1>; + #size-cells = <0>; + mctp-controller; + + mctp@30 { + compatible = "mctp-i2c-controller"; + reg = <(0x30 | I2C_OWN_SLAVE_ADDRESS)>; + }; + }; + + - | + // Mux topology with multiple MCTP-handling busses under + // a single mctp-i2c-controller. + // i2c1 and i2c6 can have MCTP devices, i2c5 does not. + #include + + i2c1: i2c { + #address-cells = <1>; + #size-cells = <0>; + mctp-controller; + + mctp@50 { + compatible = "mctp-i2c-controller"; + reg = <(0x50 | I2C_OWN_SLAVE_ADDRESS)>; + }; + }; + + i2c-mux { + #address-cells = <1>; + #size-cells = <0>; + i2c-parent = <&i2c1>; + + i2c5: i2c@0 { + #address-cells = <1>; + #size-cells = <0>; + reg = <0>; + eeprom@33 { + reg = <0x33>; + }; + }; + + i2c6: i2c@1 { + #address-cells = <1>; + #size-cells = <0>; + reg = <1>; + mctp-controller; + }; + }; From 80be9b2c0d93dabe5dc0b4ee41da12422ed25d71 Mon Sep 17 00:00:00 2001 From: Matt Johnston Date: Mon, 15 Nov 2021 10:49:26 +0800 Subject: [PATCH 0188/2048] mctp i2c: MCTP I2C binding driver Provides MCTP network transport over an I2C bus, as specified in DMTF DSP0237. All messages between nodes are sent as SMBus Block Writes. Each I2C bus to be used for MCTP is flagged in devicetree by a 'mctp-controller' property on the bus node. Each flagged bus gets a mctpi2cX net device created based on the bus number. A 'mctp-i2c-controller' I2C client needs to be added under the adapter. In an I2C mux situation the mctp-i2c-controller node must be attached only to the root I2C bus. The I2C client will handle incoming I2C slave block write data for subordinate busses as well as its own bus. In configurations without devicetree a driver instance can be attached to a bus using the I2C slave new_device mechanism. The MCTP core will hold/release the MCTP I2C device while responses are pending (a 6 second timeout or once a socket is closed, response received etc). While held the MCTP I2C driver will lock the I2C bus so that the correct I2C mux remains selected while responses are received. (Ideally we would just lock the mux to keep the current bus selected for the response rather than a full I2C bus lock, but that isn't exposed in the I2C mux API) This driver requires I2C adapters that allow 255 byte transfers (SMBus 3.0) as the specification requires a minimum MTU of 68 bytes. Signed-off-by: Matt Johnston Signed-off-by: Jeremy Kerr Signed-off-by: David S. Miller --- drivers/net/mctp/Kconfig | 12 + drivers/net/mctp/Makefile | 1 + drivers/net/mctp/mctp-i2c.c | 982 ++++++++++++++++++++++++++++++++++++ 3 files changed, 995 insertions(+) create mode 100644 drivers/net/mctp/mctp-i2c.c diff --git a/drivers/net/mctp/Kconfig b/drivers/net/mctp/Kconfig index d8f966cedc89..b758b29c2ddf 100644 --- a/drivers/net/mctp/Kconfig +++ b/drivers/net/mctp/Kconfig @@ -3,6 +3,18 @@ if MCTP menu "MCTP Device Drivers" +config MCTP_TRANSPORT_I2C + tristate "MCTP SMBus/I2C transport" + # i2c-mux is optional, but we must build as a module if i2c-mux is a module + depends on I2C_MUX || !I2C_MUX + depends on I2C + depends on I2C_SLAVE + select MCTP_FLOWS + help + Provides a driver to access MCTP devices over SMBus/I2C transport, + from DMTF specification DSP0237. A MCTP protocol network device is + created for each I2C bus that has been assigned a mctp-i2c device. + endmenu endif diff --git a/drivers/net/mctp/Makefile b/drivers/net/mctp/Makefile index e69de29bb2d1..73dc411986a6 100644 --- a/drivers/net/mctp/Makefile +++ b/drivers/net/mctp/Makefile @@ -0,0 +1 @@ +obj-$(CONFIG_MCTP_TRANSPORT_I2C) += mctp-i2c.o diff --git a/drivers/net/mctp/mctp-i2c.c b/drivers/net/mctp/mctp-i2c.c new file mode 100644 index 000000000000..ed213b4765a1 --- /dev/null +++ b/drivers/net/mctp/mctp-i2c.c @@ -0,0 +1,982 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Management Controller Transport Protocol (MCTP) + * + * Copyright (c) 2021 Code Construct + * Copyright (c) 2021 Google + */ + +#include +#include +#include +#include +#include +#include +#include + +/* SMBus 3.0 allows 255 data bytes (plus PEC), but the + * first byte is taken for source slave address. + */ +#define MCTP_I2C_MAXBLOCK 255 +#define MCTP_I2C_MAXMTU (MCTP_I2C_MAXBLOCK - 1) +#define MCTP_I2C_MINMTU (64 + 4) +/* Allow space for address, command, byte_count, databytes, PEC */ +#define MCTP_I2C_RXBUFSZ (3 + MCTP_I2C_MAXBLOCK + 1) +#define MCTP_I2C_MINLEN 8 +#define MCTP_I2C_COMMANDCODE 0x0f +#define MCTP_I2C_TX_WORK_LEN 100 +// sufficient for 64kB at min mtu +#define MCTP_I2C_TX_QUEUE_LEN 1100 + +#define MCTP_I2C_OF_PROP "mctp-controller" + +enum { + MCTP_I2C_FLOW_STATE_NEW = 0, + MCTP_I2C_FLOW_STATE_ACTIVE, +}; + +static struct { + /* lock protects clients and also prevents adding/removing adapters + * during mctp_i2c_client probe/remove. + */ + struct mutex lock; + // list of struct mctp_i2c_client + struct list_head clients; +} mi_driver_state; + +struct mctp_i2c_client; + +// The netdev structure. One of these per I2C adapter. +struct mctp_i2c_dev { + struct net_device *ndev; + struct i2c_adapter *adapter; + struct mctp_i2c_client *client; + struct list_head list; // for mctp_i2c_client.devs + + size_t pos; + u8 buffer[MCTP_I2C_RXBUFSZ]; + + struct task_struct *tx_thread; + wait_queue_head_t tx_wq; + struct sk_buff_head tx_queue; + + // a fake entry in our tx queue to perform an unlock operation + struct sk_buff unlock_marker; + + spinlock_t flow_lock; // protects i2c_lock_count and release_count + int i2c_lock_count; + int release_count; +}; + +/* The i2c client structure. One per hardware i2c bus at the top of the + * mux tree, shared by multiple netdevs + */ +struct mctp_i2c_client { + struct i2c_client *client; + u8 lladdr; + + struct mctp_i2c_dev *sel; + struct list_head devs; + spinlock_t curr_lock; // protects sel + + struct list_head list; // for mi_driver_state.clients +}; + +// Header on the wire +struct mctp_i2c_hdr { + u8 dest_slave; + u8 command; + u8 byte_count; + u8 source_slave; +}; + +static int mctp_i2c_recv(struct mctp_i2c_dev *midev); +static int mctp_i2c_slave_cb(struct i2c_client *client, + enum i2c_slave_event event, u8 *val); + +static struct i2c_adapter *mux_root_adapter(struct i2c_adapter *adap) +{ +#if IS_ENABLED(CONFIG_I2C_MUX) + return i2c_root_adapter(&adap->dev); +#else + /* In non-mux config all i2c adapters are root adapters */ + return adap; +#endif +} + +static ssize_t mctp_current_mux_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct mctp_i2c_client *mcli = i2c_get_clientdata(to_i2c_client(dev)); + struct net_device *ndev = NULL; + unsigned long flags; + ssize_t l; + + spin_lock_irqsave(&mcli->curr_lock, flags); + if (mcli->sel) { + ndev = mcli->sel->ndev; + dev_hold(ndev); + } + spin_unlock_irqrestore(&mcli->curr_lock, flags); + l = scnprintf(buf, PAGE_SIZE, "%s\n", ndev ? ndev->name : "(none)"); + if (ndev) + dev_put(ndev); + return l; +} +static DEVICE_ATTR_RO(mctp_current_mux); + +/* Creates a new i2c slave device attached to the root adapter. + * Sets up the slave callback. + * Must be called with a client on a root adapter. + */ +static struct mctp_i2c_client *mctp_i2c_new_client(struct i2c_client *client) +{ + struct mctp_i2c_client *mcli = NULL; + struct i2c_adapter *root = NULL; + int rc; + + if (client->flags & I2C_CLIENT_TEN) { + dev_err(&client->dev, "%s failed, MCTP requires a 7-bit I2C address, addr=0x%x", + __func__, client->addr); + rc = -EINVAL; + goto err; + } + + root = mux_root_adapter(client->adapter); + if (!root) { + dev_err(&client->dev, "%s failed to find root adapter\n", __func__); + rc = -ENOENT; + goto err; + } + if (root != client->adapter) { + dev_err(&client->dev, + "A mctp-i2c-controller client cannot be placed on an I2C mux adapter.\n" + " It should be placed on the mux tree root adapter\n" + " then set mctp-controller property on adapters to attach\n"); + rc = -EINVAL; + goto err; + } + + mcli = kzalloc(sizeof(*mcli), GFP_KERNEL); + if (!mcli) { + rc = -ENOMEM; + goto err; + } + spin_lock_init(&mcli->curr_lock); + INIT_LIST_HEAD(&mcli->devs); + INIT_LIST_HEAD(&mcli->list); + mcli->lladdr = client->addr & 0xff; + mcli->client = client; + i2c_set_clientdata(client, mcli); + + rc = i2c_slave_register(mcli->client, mctp_i2c_slave_cb); + if (rc) { + dev_err(&client->dev, "%s i2c register failed %d\n", __func__, rc); + mcli->client = NULL; + i2c_set_clientdata(client, NULL); + goto err; + } + + rc = device_create_file(&client->dev, &dev_attr_mctp_current_mux); + if (rc) { + dev_err(&client->dev, "%s adding sysfs \"%s\" failed %d\n", __func__, + dev_attr_mctp_current_mux.attr.name, rc); + // continue anyway + } + + return mcli; +err: + if (mcli) { + if (mcli->client) { + device_remove_file(&mcli->client->dev, &dev_attr_mctp_current_mux); + i2c_unregister_device(mcli->client); + } + kfree(mcli); + } + return ERR_PTR(rc); +} + +static void mctp_i2c_free_client(struct mctp_i2c_client *mcli) +{ + int rc; + + WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); + WARN_ON(!list_empty(&mcli->devs)); + WARN_ON(mcli->sel); // sanity check, no locking + + device_remove_file(&mcli->client->dev, &dev_attr_mctp_current_mux); + rc = i2c_slave_unregister(mcli->client); + // leak if it fails, we can't propagate errors upwards + if (rc) + dev_err(&mcli->client->dev, "%s i2c unregister failed %d\n", __func__, rc); + else + kfree(mcli); +} + +/* Switch the mctp i2c device to receive responses. + * Call with curr_lock held + */ +static void __mctp_i2c_device_select(struct mctp_i2c_client *mcli, + struct mctp_i2c_dev *midev) +{ + assert_spin_locked(&mcli->curr_lock); + if (midev) + dev_hold(midev->ndev); + if (mcli->sel) + dev_put(mcli->sel->ndev); + mcli->sel = midev; +} + +// Switch the mctp i2c device to receive responses +static void mctp_i2c_device_select(struct mctp_i2c_client *mcli, + struct mctp_i2c_dev *midev) +{ + unsigned long flags; + + spin_lock_irqsave(&mcli->curr_lock, flags); + __mctp_i2c_device_select(mcli, midev); + spin_unlock_irqrestore(&mcli->curr_lock, flags); +} + +static int mctp_i2c_slave_cb(struct i2c_client *client, + enum i2c_slave_event event, u8 *val) +{ + struct mctp_i2c_client *mcli = i2c_get_clientdata(client); + struct mctp_i2c_dev *midev = NULL; + unsigned long flags; + int rc = 0; + + spin_lock_irqsave(&mcli->curr_lock, flags); + midev = mcli->sel; + if (midev) + dev_hold(midev->ndev); + spin_unlock_irqrestore(&mcli->curr_lock, flags); + + if (!midev) + return 0; + + switch (event) { + case I2C_SLAVE_WRITE_RECEIVED: + if (midev->pos < MCTP_I2C_RXBUFSZ) { + midev->buffer[midev->pos] = *val; + midev->pos++; + } else { + midev->ndev->stats.rx_over_errors++; + } + + break; + case I2C_SLAVE_WRITE_REQUESTED: + /* dest_slave as first byte */ + midev->buffer[0] = mcli->lladdr << 1; + midev->pos = 1; + break; + case I2C_SLAVE_STOP: + rc = mctp_i2c_recv(midev); + break; + default: + break; + } + + dev_put(midev->ndev); + return rc; +} + +// Processes incoming data that has been accumulated by the slave cb +static int mctp_i2c_recv(struct mctp_i2c_dev *midev) +{ + struct net_device *ndev = midev->ndev; + struct mctp_i2c_hdr *hdr; + struct mctp_skb_cb *cb; + struct sk_buff *skb; + u8 pec, calc_pec; + size_t recvlen; + + /* + 1 for the PEC */ + if (midev->pos < MCTP_I2C_MINLEN + 1) { + ndev->stats.rx_length_errors++; + return -EINVAL; + } + recvlen = midev->pos - 1; + + hdr = (void *)midev->buffer; + if (hdr->command != MCTP_I2C_COMMANDCODE) { + ndev->stats.rx_dropped++; + return -EINVAL; + } + + pec = midev->buffer[midev->pos - 1]; + calc_pec = i2c_smbus_pec(0, midev->buffer, recvlen); + if (pec != calc_pec) { + ndev->stats.rx_crc_errors++; + return -EINVAL; + } + + skb = netdev_alloc_skb(ndev, recvlen); + if (!skb) { + ndev->stats.rx_dropped++; + return -ENOMEM; + } + + skb->protocol = htons(ETH_P_MCTP); + skb_put_data(skb, midev->buffer, recvlen); + skb_reset_mac_header(skb); + skb_pull(skb, sizeof(struct mctp_i2c_hdr)); + skb_reset_network_header(skb); + + cb = __mctp_cb(skb); + cb->halen = 1; + cb->haddr[0] = hdr->source_slave; + + if (netif_rx(skb) == NET_RX_SUCCESS) { + ndev->stats.rx_packets++; + ndev->stats.rx_bytes += skb->len; + } else { + ndev->stats.rx_dropped++; + } + return 0; +} + +enum mctp_i2c_flow_state { + MCTP_I2C_TX_FLOW_INVALID, + MCTP_I2C_TX_FLOW_NONE, + MCTP_I2C_TX_FLOW_NEW, + MCTP_I2C_TX_FLOW_EXISTING, +}; + +static enum mctp_i2c_flow_state +mctp_i2c_get_tx_flow_state(struct mctp_i2c_dev *midev, struct sk_buff *skb) +{ + enum mctp_i2c_flow_state state; + struct mctp_sk_key *key; + struct mctp_flow *flow; + unsigned long flags; + + flow = skb_ext_find(skb, SKB_EXT_MCTP); + if (!flow) + return MCTP_I2C_TX_FLOW_NONE; + + key = flow->key; + if (!key) + return MCTP_I2C_TX_FLOW_NONE; + + spin_lock_irqsave(&key->lock, flags); + /* if the key is present but invalid, we're unlikely to be able + * to handle the flow at all; just drop now + */ + if (!key->valid) { + state = MCTP_I2C_TX_FLOW_INVALID; + + } else if (key->dev_flow_state == MCTP_I2C_FLOW_STATE_NEW) { + key->dev_flow_state = MCTP_I2C_FLOW_STATE_ACTIVE; + state = MCTP_I2C_TX_FLOW_NEW; + } else { + state = MCTP_I2C_TX_FLOW_EXISTING; + } + + spin_unlock_irqrestore(&key->lock, flags); + + return state; +} + +/* We're not contending with ourselves here; we only need to exclude other + * i2c clients from using the bus. refcounts are simply to prevent + * recursive locking. + */ +static void mctp_i2c_lock_nest(struct mctp_i2c_dev *midev) +{ + unsigned long flags; + bool lock; + + spin_lock_irqsave(&midev->flow_lock, flags); + lock = midev->i2c_lock_count == 0; + midev->i2c_lock_count++; + spin_unlock_irqrestore(&midev->flow_lock, flags); + + if (lock) + i2c_lock_bus(midev->adapter, I2C_LOCK_SEGMENT); +} + +static void mctp_i2c_unlock_nest(struct mctp_i2c_dev *midev) +{ + unsigned long flags; + bool unlock; + + spin_lock_irqsave(&midev->flow_lock, flags); + if (!WARN_ONCE(midev->i2c_lock_count == 0, "lock count underflow!")) + midev->i2c_lock_count--; + unlock = midev->i2c_lock_count == 0; + spin_unlock_irqrestore(&midev->flow_lock, flags); + + if (unlock) + i2c_unlock_bus(midev->adapter, I2C_LOCK_SEGMENT); +} + +static void mctp_i2c_xmit(struct mctp_i2c_dev *midev, struct sk_buff *skb) +{ + struct net_device_stats *stats = &midev->ndev->stats; + enum mctp_i2c_flow_state fs; + union i2c_smbus_data *data; + struct mctp_i2c_hdr *hdr; + unsigned int len; + u16 daddr; + int rc; + + fs = mctp_i2c_get_tx_flow_state(midev, skb); + + len = skb->len; + hdr = (void *)skb_mac_header(skb); + data = (void *)&hdr->byte_count; + daddr = hdr->dest_slave >> 1; + + switch (fs) { + case MCTP_I2C_TX_FLOW_NONE: + /* no flow: full lock & unlock */ + mctp_i2c_lock_nest(midev); + mctp_i2c_device_select(midev->client, midev); + rc = __i2c_smbus_xfer(midev->adapter, daddr, I2C_CLIENT_PEC, + I2C_SMBUS_WRITE, hdr->command, + I2C_SMBUS_BLOCK_DATA, data); + mctp_i2c_unlock_nest(midev); + break; + + case MCTP_I2C_TX_FLOW_NEW: + /* new flow: lock, tx, but don't unlock; that will happen + * on flow release + */ + mctp_i2c_lock_nest(midev); + mctp_i2c_device_select(midev->client, midev); + fallthrough; + + case MCTP_I2C_TX_FLOW_EXISTING: + /* existing flow: we already have the lock; just tx */ + rc = __i2c_smbus_xfer(midev->adapter, daddr, I2C_CLIENT_PEC, + I2C_SMBUS_WRITE, hdr->command, + I2C_SMBUS_BLOCK_DATA, data); + break; + + case MCTP_I2C_TX_FLOW_INVALID: + return; + } + + if (rc) { + dev_warn_ratelimited(&midev->adapter->dev, + "%s i2c_smbus_xfer failed %d", __func__, rc); + stats->tx_errors++; + } else { + stats->tx_bytes += len; + stats->tx_packets++; + } +} + +static void mctp_i2c_flow_release(struct mctp_i2c_dev *midev) +{ + unsigned long flags; + bool unlock; + + spin_lock_irqsave(&midev->flow_lock, flags); + if (midev->release_count > midev->i2c_lock_count) { + WARN_ONCE(1, "release count overflow"); + midev->release_count = midev->i2c_lock_count; + } + + midev->i2c_lock_count -= midev->release_count; + unlock = midev->i2c_lock_count == 0 && midev->release_count > 0; + midev->release_count = 0; + spin_unlock_irqrestore(&midev->flow_lock, flags); + + if (unlock) + i2c_unlock_bus(midev->adapter, I2C_LOCK_SEGMENT); +} + +static int mctp_i2c_header_create(struct sk_buff *skb, struct net_device *dev, + unsigned short type, const void *daddr, + const void *saddr, unsigned int len) +{ + struct mctp_i2c_hdr *hdr; + struct mctp_hdr *mhdr; + u8 lldst, llsrc; + + lldst = *((u8 *)daddr); + llsrc = *((u8 *)saddr); + + skb_push(skb, sizeof(struct mctp_i2c_hdr)); + skb_reset_mac_header(skb); + hdr = (void *)skb_mac_header(skb); + mhdr = mctp_hdr(skb); + hdr->dest_slave = (lldst << 1) & 0xff; + hdr->command = MCTP_I2C_COMMANDCODE; + hdr->byte_count = len + 1; + if (hdr->byte_count > MCTP_I2C_MAXBLOCK) + return -EMSGSIZE; + hdr->source_slave = ((llsrc << 1) & 0xff) | 0x01; + mhdr->ver = 0x01; + + return 0; +} + +static int mctp_i2c_tx_thread(void *data) +{ + struct mctp_i2c_dev *midev = data; + struct sk_buff *skb; + unsigned long flags; + + for (;;) { + if (kthread_should_stop()) + break; + + spin_lock_irqsave(&midev->tx_queue.lock, flags); + skb = __skb_dequeue(&midev->tx_queue); + if (netif_queue_stopped(midev->ndev)) + netif_wake_queue(midev->ndev); + spin_unlock_irqrestore(&midev->tx_queue.lock, flags); + + if (skb == &midev->unlock_marker) { + mctp_i2c_flow_release(midev); + + } else if (skb) { + mctp_i2c_xmit(midev, skb); + kfree_skb(skb); + + } else { + wait_event(midev->tx_wq, + !skb_queue_empty(&midev->tx_queue) || + kthread_should_stop()); + } + } + + return 0; +} + +static netdev_tx_t mctp_i2c_start_xmit(struct sk_buff *skb, + struct net_device *dev) +{ + struct mctp_i2c_dev *midev = netdev_priv(dev); + unsigned long flags; + + spin_lock_irqsave(&midev->tx_queue.lock, flags); + if (skb_queue_len(&midev->tx_queue) >= MCTP_I2C_TX_WORK_LEN) { + netif_stop_queue(dev); + spin_unlock_irqrestore(&midev->tx_queue.lock, flags); + netdev_err(dev, "BUG! Tx Ring full when queue awake!\n"); + return NETDEV_TX_BUSY; + } + + __skb_queue_tail(&midev->tx_queue, skb); + if (skb_queue_len(&midev->tx_queue) == MCTP_I2C_TX_WORK_LEN) + netif_stop_queue(dev); + spin_unlock_irqrestore(&midev->tx_queue.lock, flags); + + wake_up(&midev->tx_wq); + return NETDEV_TX_OK; +} + +static void mctp_i2c_release_flow(struct mctp_dev *mdev, + struct mctp_sk_key *key) + +{ + struct mctp_i2c_dev *midev = netdev_priv(mdev->dev); + unsigned long flags; + + spin_lock_irqsave(&midev->flow_lock, flags); + midev->release_count++; + spin_unlock_irqrestore(&midev->flow_lock, flags); + + /* Ensure we have a release operation queued, through the fake + * marker skb + */ + spin_lock(&midev->tx_queue.lock); + if (!midev->unlock_marker.next) + __skb_queue_tail(&midev->tx_queue, &midev->unlock_marker); + spin_unlock(&midev->tx_queue.lock); + + wake_up(&midev->tx_wq); +} + +static const struct net_device_ops mctp_i2c_ops = { + .ndo_start_xmit = mctp_i2c_start_xmit, +}; + +static const struct header_ops mctp_i2c_headops = { + .create = mctp_i2c_header_create, +}; + +static const struct mctp_netdev_ops mctp_i2c_mctp_ops = { + .release_flow = mctp_i2c_release_flow, +}; + +static void mctp_i2c_net_setup(struct net_device *dev) +{ + dev->type = ARPHRD_MCTP; + + dev->mtu = MCTP_I2C_MAXMTU; + dev->min_mtu = MCTP_I2C_MINMTU; + dev->max_mtu = MCTP_I2C_MAXMTU; + dev->tx_queue_len = MCTP_I2C_TX_QUEUE_LEN; + + dev->hard_header_len = sizeof(struct mctp_i2c_hdr); + dev->addr_len = 1; + + dev->netdev_ops = &mctp_i2c_ops; + dev->header_ops = &mctp_i2c_headops; + dev->needs_free_netdev = true; +} + +static int mctp_i2c_add_netdev(struct mctp_i2c_client *mcli, + struct i2c_adapter *adap) +{ + unsigned long flags; + struct mctp_i2c_dev *midev = NULL; + struct net_device *ndev = NULL; + struct i2c_adapter *root; + char namebuf[30]; + int rc; + + root = mux_root_adapter(adap); + if (root != mcli->client->adapter) { + dev_err(&mcli->client->dev, + "I2C adapter %s is not a child bus of %s", + mcli->client->adapter->name, root->name); + return -EINVAL; + } + + WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); + snprintf(namebuf, sizeof(namebuf), "mctpi2c%d", adap->nr); + ndev = alloc_netdev(sizeof(*midev), namebuf, NET_NAME_ENUM, mctp_i2c_net_setup); + if (!ndev) { + dev_err(&mcli->client->dev, "%s alloc netdev failed\n", __func__); + rc = -ENOMEM; + goto err; + } + dev_net_set(ndev, current->nsproxy->net_ns); + SET_NETDEV_DEV(ndev, &adap->dev); + ndev->dev_addr = &mcli->lladdr; + + midev = netdev_priv(ndev); + skb_queue_head_init(&midev->tx_queue); + INIT_LIST_HEAD(&midev->list); + midev->adapter = adap; + midev->client = mcli; + spin_lock_init(&midev->flow_lock); + midev->i2c_lock_count = 0; + midev->release_count = 0; + /* Hold references */ + get_device(&midev->adapter->dev); + get_device(&midev->client->client->dev); + midev->ndev = ndev; + init_waitqueue_head(&midev->tx_wq); + midev->tx_thread = kthread_create(mctp_i2c_tx_thread, midev, + "%s/tx", namebuf); + if (IS_ERR_OR_NULL(midev->tx_thread)) { + rc = -ENOMEM; + goto err_free; + } + + rc = mctp_register_netdev(ndev, &mctp_i2c_mctp_ops); + if (rc) { + dev_err(&mcli->client->dev, + "%s register netdev \"%s\" failed %d\n", __func__, + ndev->name, rc); + goto err_stop_kthread; + } + spin_lock_irqsave(&mcli->curr_lock, flags); + list_add(&midev->list, &mcli->devs); + // Select a device by default + if (!mcli->sel) + __mctp_i2c_device_select(mcli, midev); + spin_unlock_irqrestore(&mcli->curr_lock, flags); + + wake_up_process(midev->tx_thread); + + return 0; + +err_stop_kthread: + kthread_stop(midev->tx_thread); + +err_free: + free_netdev(ndev); + +err: + return rc; +} + +// Removes and unregisters a mctp-i2c netdev +static void mctp_i2c_free_netdev(struct mctp_i2c_dev *midev) +{ + struct mctp_i2c_client *mcli = midev->client; + unsigned long flags; + + netif_stop_queue(midev->ndev); + kthread_stop(midev->tx_thread); + skb_queue_purge(&midev->tx_queue); + + /* Release references, used only for TX which has stopped */ + put_device(&midev->adapter->dev); + put_device(&mcli->client->dev); + + /* Remove it from the parent mcli */ + spin_lock_irqsave(&mcli->curr_lock, flags); + list_del(&midev->list); + if (mcli->sel == midev) { + struct mctp_i2c_dev *first; + + first = list_first_entry_or_null(&mcli->devs, struct mctp_i2c_dev, list); + __mctp_i2c_device_select(mcli, first); + } + spin_unlock_irqrestore(&mcli->curr_lock, flags); + + /* Remove netdev. mctp_i2c_slave_cb() takes a dev_hold() so removing + * it now is safe. unregister_netdev() frees ndev and midev. + */ + mctp_unregister_netdev(midev->ndev); +} + +// Removes any netdev for adap. mcli is the parent root i2c client +static void mctp_i2c_remove_netdev(struct mctp_i2c_client *mcli, + struct i2c_adapter *adap) +{ + unsigned long flags; + struct mctp_i2c_dev *midev = NULL, *m = NULL; + + WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); + spin_lock_irqsave(&mcli->curr_lock, flags); + // list size is limited by number of MCTP netdevs on a single hardware bus + list_for_each_entry(m, &mcli->devs, list) + if (m->adapter == adap) { + midev = m; + break; + } + spin_unlock_irqrestore(&mcli->curr_lock, flags); + + if (midev) + mctp_i2c_free_netdev(midev); +} + +/* Determines whether a device is an i2c adapter. + * Optionally returns the root i2c_adapter + */ +static struct i2c_adapter *mctp_i2c_get_adapter(struct device *dev, + struct i2c_adapter **ret_root) +{ + struct i2c_adapter *root, *adap; + + if (dev->type != &i2c_adapter_type) + return NULL; + adap = to_i2c_adapter(dev); + root = mux_root_adapter(adap); + WARN_ONCE(!root, "%s failed to find root adapter for %s\n", + __func__, dev_name(dev)); + if (!root) + return NULL; + if (ret_root) + *ret_root = root; + return adap; +} + +/* Determines whether a device is an i2c adapter with the "mctp-controller" + * devicetree property set. If adap is not an OF node, returns match_no_of + */ +static bool mctp_i2c_adapter_match(struct i2c_adapter *adap, bool match_no_of) +{ + if (!adap->dev.of_node) + return match_no_of; + return of_property_read_bool(adap->dev.of_node, MCTP_I2C_OF_PROP); +} + +/* Called for each existing i2c device (adapter or client) when a + * new mctp-i2c client is probed. + */ +static int mctp_i2c_client_try_attach(struct device *dev, void *data) +{ + struct i2c_adapter *adap = NULL, *root = NULL; + struct mctp_i2c_client *mcli = data; + + adap = mctp_i2c_get_adapter(dev, &root); + if (!adap) + return 0; + if (mcli->client->adapter != root) + return 0; + // Must either have mctp-controller property on the adapter, or + // be a root adapter if it's non-devicetree + if (!mctp_i2c_adapter_match(adap, adap == root)) + return 0; + + return mctp_i2c_add_netdev(mcli, adap); +} + +static void mctp_i2c_notify_add(struct device *dev) +{ + struct mctp_i2c_client *mcli = NULL, *m = NULL; + struct i2c_adapter *root = NULL, *adap = NULL; + int rc; + + adap = mctp_i2c_get_adapter(dev, &root); + if (!adap) + return; + // Check for mctp-controller property on the adapter + if (!mctp_i2c_adapter_match(adap, false)) + return; + + /* Find an existing mcli for adap's root */ + mutex_lock(&mi_driver_state.lock); + list_for_each_entry(m, &mi_driver_state.clients, list) { + if (m->client->adapter == root) { + mcli = m; + break; + } + } + + if (mcli) { + rc = mctp_i2c_add_netdev(mcli, adap); + if (rc) + dev_warn(dev, "%s Failed adding mctp-i2c device", + __func__); + } + mutex_unlock(&mi_driver_state.lock); +} + +static void mctp_i2c_notify_del(struct device *dev) +{ + struct i2c_adapter *root = NULL, *adap = NULL; + struct mctp_i2c_client *mcli = NULL; + + adap = mctp_i2c_get_adapter(dev, &root); + if (!adap) + return; + + mutex_lock(&mi_driver_state.lock); + list_for_each_entry(mcli, &mi_driver_state.clients, list) { + if (mcli->client->adapter == root) { + mctp_i2c_remove_netdev(mcli, adap); + break; + } + } + mutex_unlock(&mi_driver_state.lock); +} + +static int mctp_i2c_probe(struct i2c_client *client) +{ + struct mctp_i2c_client *mcli = NULL; + int rc; + + /* Check for >32 byte block support required for MCTP */ + if (!i2c_check_functionality(client->adapter, I2C_FUNC_SMBUS_V3_BLOCK)) { + dev_err(&client->dev, + "%s failed, I2C bus driver does not support 255 byte block transfer\n", + __func__); + return -EOPNOTSUPP; + } + + mutex_lock(&mi_driver_state.lock); + mcli = mctp_i2c_new_client(client); + if (IS_ERR(mcli)) { + rc = PTR_ERR(mcli); + mcli = NULL; + goto out; + } else { + list_add(&mcli->list, &mi_driver_state.clients); + } + + // Add a netdev for adapters that have a 'mctp-controller' property + i2c_for_each_dev(mcli, mctp_i2c_client_try_attach); + rc = 0; +out: + mutex_unlock(&mi_driver_state.lock); + return rc; +} + +static int mctp_i2c_remove(struct i2c_client *client) +{ + struct mctp_i2c_client *mcli = i2c_get_clientdata(client); + struct mctp_i2c_dev *midev = NULL, *tmp = NULL; + + mutex_lock(&mi_driver_state.lock); + list_del(&mcli->list); + // Remove all child adapter netdevs + list_for_each_entry_safe(midev, tmp, &mcli->devs, list) + mctp_i2c_free_netdev(midev); + + mctp_i2c_free_client(mcli); + mutex_unlock(&mi_driver_state.lock); + // Callers ignore return code + return 0; +} + +/* We look for a 'mctp-controller' property on I2C busses as they are + * added/deleted, creating/removing netdevs as required. + */ +static int mctp_i2c_notifier_call(struct notifier_block *nb, + unsigned long action, void *data) +{ + struct device *dev = data; + + switch (action) { + case BUS_NOTIFY_ADD_DEVICE: + mctp_i2c_notify_add(dev); + break; + case BUS_NOTIFY_DEL_DEVICE: + mctp_i2c_notify_del(dev); + break; + } + return NOTIFY_DONE; +} + +static struct notifier_block mctp_i2c_notifier = { + .notifier_call = mctp_i2c_notifier_call, +}; + +static const struct i2c_device_id mctp_i2c_id[] = { + { "mctp-i2c", 0 }, + {}, +}; +MODULE_DEVICE_TABLE(i2c, mctp_i2c_id); + +static const struct of_device_id mctp_i2c_of_match[] = { + { .compatible = "mctp-i2c-controller" }, + {}, +}; +MODULE_DEVICE_TABLE(of, mctp_i2c_of_match); + +static struct i2c_driver mctp_i2c_driver = { + .driver = { + .name = "mctp-i2c", + .of_match_table = mctp_i2c_of_match, + }, + .probe_new = mctp_i2c_probe, + .remove = mctp_i2c_remove, + .id_table = mctp_i2c_id, +}; + +static __init int mctp_i2c_init(void) +{ + int rc; + + INIT_LIST_HEAD(&mi_driver_state.clients); + mutex_init(&mi_driver_state.lock); + pr_info("MCTP SMBus/I2C transport driver\n"); + rc = i2c_add_driver(&mctp_i2c_driver); + if (rc) + return rc; + rc = bus_register_notifier(&i2c_bus_type, &mctp_i2c_notifier); + if (rc) { + i2c_del_driver(&mctp_i2c_driver); + return rc; + } + return 0; +} + +static __exit void mctp_i2c_exit(void) +{ + int rc; + + rc = bus_unregister_notifier(&i2c_bus_type, &mctp_i2c_notifier); + if (rc) + pr_warn("%s Could not unregister notifier, %d", __func__, rc); + i2c_del_driver(&mctp_i2c_driver); +} + +module_init(mctp_i2c_init); +module_exit(mctp_i2c_exit); + +MODULE_DESCRIPTION("MCTP SMBus/I2C device"); +MODULE_LICENSE("GPL v2"); +MODULE_AUTHOR("Matt Johnston "); From 16b1c4e01c89ba07367461e0bc4cb84993c2d027 Mon Sep 17 00:00:00 2001 From: Jacky Chou Date: Mon, 15 Nov 2021 11:49:41 +0800 Subject: [PATCH 0189/2048] net: usb: ax88179_178a: add TSO feature On low-effciency embedded platforms, transmission performance is poor due to on Bulk-out with single packet. Adding TSO feature improves the transmission performance and reduces the number of interrupt caused by Bulk-out complete. Reference to module, net: usb: aqc111. Signed-off-by: Jacky Chou Signed-off-by: David S. Miller --- drivers/net/usb/ax88179_178a.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/drivers/net/usb/ax88179_178a.c b/drivers/net/usb/ax88179_178a.c index ea8aa8c33241..1a627ba4b850 100644 --- a/drivers/net/usb/ax88179_178a.c +++ b/drivers/net/usb/ax88179_178a.c @@ -1377,11 +1377,12 @@ static int ax88179_bind(struct usbnet *dev, struct usb_interface *intf) dev->mii.phy_id = 0x03; dev->mii.supports_gmii = 1; - dev->net->features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | - NETIF_F_RXCSUM; + dev->net->features |= NETIF_F_SG | NETIF_F_IP_CSUM | + NETIF_F_IPV6_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO; - dev->net->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | - NETIF_F_RXCSUM; + dev->net->hw_features |= dev->net->features; + + netif_set_gso_max_size(dev->net, 16384); /* Enable checksum offload */ *tmp = AX_RXCOE_IP | AX_RXCOE_TCP | AX_RXCOE_UDP | @@ -1526,17 +1527,19 @@ ax88179_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) { u32 tx_hdr1, tx_hdr2; int frame_size = dev->maxpacket; - int mss = skb_shinfo(skb)->gso_size; int headroom; void *ptr; tx_hdr1 = skb->len; - tx_hdr2 = mss; + tx_hdr2 = skb_shinfo(skb)->gso_size; /* Set TSO mss */ if (((skb->len + 8) % frame_size) == 0) tx_hdr2 |= 0x80008000; /* Enable padding */ headroom = skb_headroom(skb) - 8; + if ((dev->net->features & NETIF_F_SG) && skb_linearize(skb)) + return NULL; + if ((skb_header_cloned(skb) || headroom < 0) && pskb_expand_head(skb, headroom < 0 ? 8 : 0, 0, GFP_ATOMIC)) { dev_kfree_skb_any(skb); @@ -1547,6 +1550,8 @@ ax88179_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) put_unaligned_le32(tx_hdr1, ptr); put_unaligned_le32(tx_hdr2, ptr + 4); + usbnet_set_skb_tx_stats(skb, (skb_shinfo(skb)->gso_segs ?: 1), 0); + return skb; } From 9ed941178ce9894ecd0e06aebf7a3c43841817aa Mon Sep 17 00:00:00 2001 From: Guo Zhengkui Date: Mon, 15 Nov 2021 13:00:10 +0800 Subject: [PATCH 0190/2048] hinic: use ARRAY_SIZE instead of ARRAY_LEN ARRAY_SIZE defined in is safer than self-defined macros to get size of an array such as ARRAY_LEN used here. Because ARRAY_SIZE uses __must_be_array(arr) to ensure arr is really an array. Reported-by: Alejandro Colomar Signed-off-by: Guo Zhengkui Signed-off-by: David S. Miller --- .../net/ethernet/huawei/hinic/hinic_ethtool.c | 32 +++++++++---------- 1 file changed, 15 insertions(+), 17 deletions(-) diff --git a/drivers/net/ethernet/huawei/hinic/hinic_ethtool.c b/drivers/net/ethernet/huawei/hinic/hinic_ethtool.c index a85667078b72..a35a80f9a234 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_ethtool.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_ethtool.c @@ -1205,8 +1205,6 @@ static u32 hinic_get_rxfh_indir_size(struct net_device *netdev) return HINIC_RSS_INDIR_SIZE; } -#define ARRAY_LEN(arr) ((int)((int)sizeof(arr) / (int)sizeof(arr[0]))) - #define HINIC_FUNC_STAT(_stat_item) { \ .name = #_stat_item, \ .size = sizeof_field(struct hinic_vport_stats, _stat_item), \ @@ -1374,7 +1372,7 @@ static void get_drv_queue_stats(struct hinic_dev *nic_dev, u64 *data) break; hinic_txq_get_stats(&nic_dev->txqs[qid], &txq_stats); - for (j = 0; j < ARRAY_LEN(hinic_tx_queue_stats); j++, i++) { + for (j = 0; j < ARRAY_SIZE(hinic_tx_queue_stats); j++, i++) { p = (char *)&txq_stats + hinic_tx_queue_stats[j].offset; data[i] = (hinic_tx_queue_stats[j].size == @@ -1387,7 +1385,7 @@ static void get_drv_queue_stats(struct hinic_dev *nic_dev, u64 *data) break; hinic_rxq_get_stats(&nic_dev->rxqs[qid], &rxq_stats); - for (j = 0; j < ARRAY_LEN(hinic_rx_queue_stats); j++, i++) { + for (j = 0; j < ARRAY_SIZE(hinic_rx_queue_stats); j++, i++) { p = (char *)&rxq_stats + hinic_rx_queue_stats[j].offset; data[i] = (hinic_rx_queue_stats[j].size == @@ -1411,7 +1409,7 @@ static void hinic_get_ethtool_stats(struct net_device *netdev, netif_err(nic_dev, drv, netdev, "Failed to get vport stats from firmware\n"); - for (j = 0; j < ARRAY_LEN(hinic_function_stats); j++, i++) { + for (j = 0; j < ARRAY_SIZE(hinic_function_stats); j++, i++) { p = (char *)&vport_stats + hinic_function_stats[j].offset; data[i] = (hinic_function_stats[j].size == sizeof(u64)) ? *(u64 *)p : *(u32 *)p; @@ -1420,8 +1418,8 @@ static void hinic_get_ethtool_stats(struct net_device *netdev, port_stats = kzalloc(sizeof(*port_stats), GFP_KERNEL); if (!port_stats) { memset(&data[i], 0, - ARRAY_LEN(hinic_port_stats) * sizeof(*data)); - i += ARRAY_LEN(hinic_port_stats); + ARRAY_SIZE(hinic_port_stats) * sizeof(*data)); + i += ARRAY_SIZE(hinic_port_stats); goto get_drv_stats; } @@ -1430,7 +1428,7 @@ static void hinic_get_ethtool_stats(struct net_device *netdev, netif_err(nic_dev, drv, netdev, "Failed to get port stats from firmware\n"); - for (j = 0; j < ARRAY_LEN(hinic_port_stats); j++, i++) { + for (j = 0; j < ARRAY_SIZE(hinic_port_stats); j++, i++) { p = (char *)port_stats + hinic_port_stats[j].offset; data[i] = (hinic_port_stats[j].size == sizeof(u64)) ? *(u64 *)p : *(u32 *)p; @@ -1449,14 +1447,14 @@ static int hinic_get_sset_count(struct net_device *netdev, int sset) switch (sset) { case ETH_SS_TEST: - return ARRAY_LEN(hinic_test_strings); + return ARRAY_SIZE(hinic_test_strings); case ETH_SS_STATS: q_num = nic_dev->num_qps; - count = ARRAY_LEN(hinic_function_stats) + - (ARRAY_LEN(hinic_tx_queue_stats) + - ARRAY_LEN(hinic_rx_queue_stats)) * q_num; + count = ARRAY_SIZE(hinic_function_stats) + + (ARRAY_SIZE(hinic_tx_queue_stats) + + ARRAY_SIZE(hinic_rx_queue_stats)) * q_num; - count += ARRAY_LEN(hinic_port_stats); + count += ARRAY_SIZE(hinic_port_stats); return count; default: @@ -1476,27 +1474,27 @@ static void hinic_get_strings(struct net_device *netdev, memcpy(data, *hinic_test_strings, sizeof(hinic_test_strings)); return; case ETH_SS_STATS: - for (i = 0; i < ARRAY_LEN(hinic_function_stats); i++) { + for (i = 0; i < ARRAY_SIZE(hinic_function_stats); i++) { memcpy(p, hinic_function_stats[i].name, ETH_GSTRING_LEN); p += ETH_GSTRING_LEN; } - for (i = 0; i < ARRAY_LEN(hinic_port_stats); i++) { + for (i = 0; i < ARRAY_SIZE(hinic_port_stats); i++) { memcpy(p, hinic_port_stats[i].name, ETH_GSTRING_LEN); p += ETH_GSTRING_LEN; } for (i = 0; i < nic_dev->num_qps; i++) { - for (j = 0; j < ARRAY_LEN(hinic_tx_queue_stats); j++) { + for (j = 0; j < ARRAY_SIZE(hinic_tx_queue_stats); j++) { sprintf(p, hinic_tx_queue_stats[j].name, i); p += ETH_GSTRING_LEN; } } for (i = 0; i < nic_dev->num_qps; i++) { - for (j = 0; j < ARRAY_LEN(hinic_rx_queue_stats); j++) { + for (j = 0; j < ARRAY_SIZE(hinic_rx_queue_stats); j++) { sprintf(p, hinic_rx_queue_stats[j].name, i); p += ETH_GSTRING_LEN; } From 311107bdecd158870ab8556fd1289a56a6087bad Mon Sep 17 00:00:00 2001 From: Yihao Han Date: Sun, 14 Nov 2021 22:58:16 -0800 Subject: [PATCH 0191/2048] net: fddi: use swap() to make code cleaner Use the macro 'swap()' defined in 'include/linux/minmax.h' to avoid opencoding it. Signed-off-by: Yihao Han Signed-off-by: David S. Miller --- drivers/net/fddi/skfp/smt.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-) diff --git a/drivers/net/fddi/skfp/smt.c b/drivers/net/fddi/skfp/smt.c index 6b68a53f1b38..72c31f0013ad 100644 --- a/drivers/net/fddi/skfp/smt.c +++ b/drivers/net/fddi/skfp/smt.c @@ -1846,10 +1846,10 @@ void smt_swap_para(struct smt_header *sm, int len, int direction) } } + static void smt_string_swap(char *data, const char *format, int len) { const char *open_paren = NULL ; - int x ; while (len > 0 && *format) { switch (*format) { @@ -1876,19 +1876,13 @@ static void smt_string_swap(char *data, const char *format, int len) len-- ; break ; case 's' : - x = data[0] ; - data[0] = data[1] ; - data[1] = x ; + swap(data[0], data[1]) ; data += 2 ; len -= 2 ; break ; case 'l' : - x = data[0] ; - data[0] = data[3] ; - data[3] = x ; - x = data[1] ; - data[1] = data[2] ; - data[2] = x ; + swap(data[0], data[3]) ; + swap(data[1], data[2]) ; data += 4 ; len -= 4 ; break ; From 5cf46d8e741f3876222f11b0a0bcbdcddfd8754e Mon Sep 17 00:00:00 2001 From: Christophe Leroy Date: Mon, 15 Nov 2021 12:08:59 +0100 Subject: [PATCH 0192/2048] net/wan/fsl_ucc_hdlc: fix sparse warnings CHECK drivers/net/wan/fsl_ucc_hdlc.c drivers/net/wan/fsl_ucc_hdlc.c:309:57: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:309:57: expected void [noderef] __iomem * drivers/net/wan/fsl_ucc_hdlc.c:309:57: got restricted __be16 * drivers/net/wan/fsl_ucc_hdlc.c:311:46: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:311:46: expected void [noderef] __iomem * drivers/net/wan/fsl_ucc_hdlc.c:311:46: got restricted __be32 * drivers/net/wan/fsl_ucc_hdlc.c:320:57: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:320:57: expected void [noderef] __iomem * drivers/net/wan/fsl_ucc_hdlc.c:320:57: got restricted __be16 * drivers/net/wan/fsl_ucc_hdlc.c:322:46: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:322:46: expected void [noderef] __iomem * drivers/net/wan/fsl_ucc_hdlc.c:322:46: got restricted __be32 * drivers/net/wan/fsl_ucc_hdlc.c:372:29: warning: incorrect type in assignment (different base types) drivers/net/wan/fsl_ucc_hdlc.c:372:29: expected unsigned short [usertype] drivers/net/wan/fsl_ucc_hdlc.c:372:29: got restricted __be16 [usertype] drivers/net/wan/fsl_ucc_hdlc.c:379:36: warning: restricted __be16 degrades to integer drivers/net/wan/fsl_ucc_hdlc.c:402:12: warning: incorrect type in assignment (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:402:12: expected struct qe_bd [noderef] __iomem *bd drivers/net/wan/fsl_ucc_hdlc.c:402:12: got struct qe_bd *curtx_bd drivers/net/wan/fsl_ucc_hdlc.c:425:20: warning: incorrect type in assignment (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:425:20: expected struct qe_bd [noderef] __iomem *[assigned] bd drivers/net/wan/fsl_ucc_hdlc.c:425:20: got struct qe_bd *tx_bd_base drivers/net/wan/fsl_ucc_hdlc.c:427:16: error: incompatible types in comparison expression (different address spaces): drivers/net/wan/fsl_ucc_hdlc.c:427:16: struct qe_bd [noderef] __iomem * drivers/net/wan/fsl_ucc_hdlc.c:427:16: struct qe_bd * drivers/net/wan/fsl_ucc_hdlc.c:462:33: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:506:41: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:528:33: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:552:38: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:596:67: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:611:41: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:851:38: warning: incorrect type in initializer (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:854:40: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:855:40: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:858:39: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:861:37: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:866:38: warning: incorrect type in initializer (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:868:21: warning: incorrect type in argument 1 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:870:40: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:871:40: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:873:39: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:993:57: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:995:46: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:1004:57: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:1006:46: warning: incorrect type in argument 2 (different address spaces) drivers/net/wan/fsl_ucc_hdlc.c:412:35: warning: dereference of noderef expression drivers/net/wan/fsl_ucc_hdlc.c:412:35: warning: dereference of noderef expression drivers/net/wan/fsl_ucc_hdlc.c:724:29: warning: dereference of noderef expression drivers/net/wan/fsl_ucc_hdlc.c:815:21: warning: dereference of noderef expression drivers/net/wan/fsl_ucc_hdlc.c:1021:29: warning: dereference of noderef expression Most of the warnings are due to DMA memory being incorrectly handled as IO memory. Fix it by doing direct read/write and doing proper dma_rmb() / dma_wmb(). Other problems are type mismatches or lack of use of IO accessors. Cc: Rasmus Villemoes Reported-by: kernel test robot Link: https://lkml.org/lkml/2021/11/12/647 Signed-off-by: Christophe Leroy Signed-off-by: David S. Miller --- drivers/net/wan/fsl_ucc_hdlc.c | 62 ++++++++++++++++++---------------- 1 file changed, 32 insertions(+), 30 deletions(-) diff --git a/drivers/net/wan/fsl_ucc_hdlc.c b/drivers/net/wan/fsl_ucc_hdlc.c index cda1b4ce6b21..5ae2d27b5da9 100644 --- a/drivers/net/wan/fsl_ucc_hdlc.c +++ b/drivers/net/wan/fsl_ucc_hdlc.c @@ -306,9 +306,8 @@ static int uhdlc_init(struct ucc_hdlc_private *priv) else bd_status = R_E_S | R_I_S | R_W_S; - iowrite16be(bd_status, &priv->rx_bd_base[i].status); - iowrite32be(priv->dma_rx_addr + i * MAX_RX_BUF_LENGTH, - &priv->rx_bd_base[i].buf); + priv->rx_bd_base[i].status = cpu_to_be16(bd_status); + priv->rx_bd_base[i].buf = cpu_to_be32(priv->dma_rx_addr + i * MAX_RX_BUF_LENGTH); } for (i = 0; i < TX_BD_RING_LEN; i++) { @@ -317,10 +316,10 @@ static int uhdlc_init(struct ucc_hdlc_private *priv) else bd_status = T_I_S | T_TC_S | T_W_S; - iowrite16be(bd_status, &priv->tx_bd_base[i].status); - iowrite32be(priv->dma_tx_addr + i * MAX_RX_BUF_LENGTH, - &priv->tx_bd_base[i].buf); + priv->tx_bd_base[i].status = cpu_to_be16(bd_status); + priv->tx_bd_base[i].buf = cpu_to_be32(priv->dma_tx_addr + i * MAX_RX_BUF_LENGTH); } + dma_wmb(); return 0; @@ -352,10 +351,10 @@ static netdev_tx_t ucc_hdlc_tx(struct sk_buff *skb, struct net_device *dev) { hdlc_device *hdlc = dev_to_hdlc(dev); struct ucc_hdlc_private *priv = (struct ucc_hdlc_private *)hdlc->priv; - struct qe_bd __iomem *bd; + struct qe_bd *bd; u16 bd_status; unsigned long flags; - u16 *proto_head; + __be16 *proto_head; switch (dev->type) { case ARPHRD_RAWHDLC: @@ -368,14 +367,14 @@ static netdev_tx_t ucc_hdlc_tx(struct sk_buff *skb, struct net_device *dev) skb_push(skb, HDLC_HEAD_LEN); - proto_head = (u16 *)skb->data; + proto_head = (__be16 *)skb->data; *proto_head = htons(DEFAULT_HDLC_HEAD); dev->stats.tx_bytes += skb->len; break; case ARPHRD_PPP: - proto_head = (u16 *)skb->data; + proto_head = (__be16 *)skb->data; if (*proto_head != htons(DEFAULT_PPP_HEAD)) { dev->stats.tx_dropped++; dev_kfree_skb(skb); @@ -398,9 +397,10 @@ static netdev_tx_t ucc_hdlc_tx(struct sk_buff *skb, struct net_device *dev) netdev_sent_queue(dev, skb->len); spin_lock_irqsave(&priv->lock, flags); + dma_rmb(); /* Start from the next BD that should be filled */ bd = priv->curtx_bd; - bd_status = ioread16be(&bd->status); + bd_status = be16_to_cpu(bd->status); /* Save the skb pointer so we can free it later */ priv->tx_skbuff[priv->skb_curtx] = skb; @@ -415,8 +415,8 @@ static netdev_tx_t ucc_hdlc_tx(struct sk_buff *skb, struct net_device *dev) /* set bd status and length */ bd_status = (bd_status & T_W_S) | T_R_S | T_I_S | T_L_S | T_TC_S; - iowrite16be(skb->len, &bd->length); - iowrite16be(bd_status, &bd->status); + bd->length = cpu_to_be16(skb->len); + bd->status = cpu_to_be16(bd_status); /* Move to next BD in the ring */ if (!(bd_status & T_W_S)) @@ -458,8 +458,9 @@ static int hdlc_tx_done(struct ucc_hdlc_private *priv) u16 bd_status; int tx_restart = 0; + dma_rmb(); bd = priv->dirty_tx; - bd_status = ioread16be(&bd->status); + bd_status = be16_to_cpu(bd->status); /* Normal processing. */ while ((bd_status & T_R_S) == 0) { @@ -503,7 +504,7 @@ static int hdlc_tx_done(struct ucc_hdlc_private *priv) bd += 1; else bd = priv->tx_bd_base; - bd_status = ioread16be(&bd->status); + bd_status = be16_to_cpu(bd->status); } priv->dirty_tx = bd; @@ -524,8 +525,9 @@ static int hdlc_rx_done(struct ucc_hdlc_private *priv, int rx_work_limit) u16 length, howmany = 0; u8 *bdbuffer; + dma_rmb(); bd = priv->currx_bd; - bd_status = ioread16be(&bd->status); + bd_status = be16_to_cpu(bd->status); /* while there are received buffers and BD is full (~R_E) */ while (!((bd_status & (R_E_S)) || (--rx_work_limit < 0))) { @@ -549,7 +551,7 @@ static int hdlc_rx_done(struct ucc_hdlc_private *priv, int rx_work_limit) } bdbuffer = priv->rx_buffer + (priv->currx_bdnum * MAX_RX_BUF_LENGTH); - length = ioread16be(&bd->length); + length = be16_to_cpu(bd->length); switch (dev->type) { case ARPHRD_RAWHDLC: @@ -593,7 +595,7 @@ static int hdlc_rx_done(struct ucc_hdlc_private *priv, int rx_work_limit) netif_receive_skb(skb); recycle: - iowrite16be((bd_status & R_W_S) | R_E_S | R_I_S, &bd->status); + bd->status = cpu_to_be16((bd_status & R_W_S) | R_E_S | R_I_S); /* update to point at the next bd */ if (bd_status & R_W_S) { @@ -608,8 +610,9 @@ recycle: bd += 1; } - bd_status = ioread16be(&bd->status); + bd_status = be16_to_cpu(bd->status); } + dma_rmb(); priv->currx_bd = bd; return howmany; @@ -721,7 +724,7 @@ static int uhdlc_open(struct net_device *dev) /* Enable the TDM port */ if (priv->tsa) - utdm->si_regs->siglmr1_h |= (0x1 << utdm->tdm_port); + qe_setbits_8(&utdm->si_regs->siglmr1_h, 0x1 << utdm->tdm_port); priv->hdlc_busy = 1; netif_device_attach(priv->ndev); @@ -812,7 +815,7 @@ static int uhdlc_close(struct net_device *dev) (u8)QE_CR_PROTOCOL_UNSPECIFIED, 0); if (priv->tsa) - utdm->si_regs->siglmr1_h &= ~(0x1 << utdm->tdm_port); + qe_clrbits_8(&utdm->si_regs->siglmr1_h, 0x1 << utdm->tdm_port); ucc_fast_disable(priv->uccf, COMM_DIR_RX | COMM_DIR_TX); @@ -848,7 +851,7 @@ static int ucc_hdlc_attach(struct net_device *dev, unsigned short encoding, #ifdef CONFIG_PM static void store_clk_config(struct ucc_hdlc_private *priv) { - struct qe_mux *qe_mux_reg = &qe_immr->qmx; + struct qe_mux __iomem *qe_mux_reg = &qe_immr->qmx; /* store si clk */ priv->cmxsi1cr_h = ioread32be(&qe_mux_reg->cmxsi1cr_h); @@ -863,7 +866,7 @@ static void store_clk_config(struct ucc_hdlc_private *priv) static void resume_clk_config(struct ucc_hdlc_private *priv) { - struct qe_mux *qe_mux_reg = &qe_immr->qmx; + struct qe_mux __iomem *qe_mux_reg = &qe_immr->qmx; memcpy_toio(qe_mux_reg->cmxucr, priv->cmxucr, 4 * sizeof(u32)); @@ -990,9 +993,8 @@ static int uhdlc_resume(struct device *dev) else bd_status = R_E_S | R_I_S | R_W_S; - iowrite16be(bd_status, &priv->rx_bd_base[i].status); - iowrite32be(priv->dma_rx_addr + i * MAX_RX_BUF_LENGTH, - &priv->rx_bd_base[i].buf); + priv->rx_bd_base[i].status = cpu_to_be16(bd_status); + priv->rx_bd_base[i].buf = cpu_to_be32(priv->dma_rx_addr + i * MAX_RX_BUF_LENGTH); } for (i = 0; i < TX_BD_RING_LEN; i++) { @@ -1001,10 +1003,10 @@ static int uhdlc_resume(struct device *dev) else bd_status = T_I_S | T_TC_S | T_W_S; - iowrite16be(bd_status, &priv->tx_bd_base[i].status); - iowrite32be(priv->dma_tx_addr + i * MAX_RX_BUF_LENGTH, - &priv->tx_bd_base[i].buf); + priv->tx_bd_base[i].status = cpu_to_be16(bd_status); + priv->tx_bd_base[i].buf = cpu_to_be32(priv->dma_tx_addr + i * MAX_RX_BUF_LENGTH); } + dma_wmb(); /* if hdlc is busy enable TX and RX */ if (priv->hdlc_busy == 1) { @@ -1018,7 +1020,7 @@ static int uhdlc_resume(struct device *dev) /* Enable the TDM port */ if (priv->tsa) - utdm->si_regs->siglmr1_h |= (0x1 << utdm->tdm_port); + qe_setbits_8(&utdm->si_regs->siglmr1_h, 0x1 << utdm->tdm_port); } napi_enable(&priv->napi); From 34ae2c09d46a2d0abd907e139b466f798e4095a8 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Mon, 15 Nov 2021 10:00:27 +0000 Subject: [PATCH 0193/2048] net: phylink: add generic validate implementation Add a generic validate() implementation using the supported_interfaces and a bitmask of MAC pause/speed/duplex capabilities. This allows us to entirely eliminate many driver private validate() implementations. We expose the underlying phylink_get_linkmodes() function so that drivers which have special needs can still benefit from conversion. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/phy/phylink.c | 252 ++++++++++++++++++++++++++++++++++++++ include/linux/phylink.h | 31 +++++ 2 files changed, 283 insertions(+) diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index 3ad7397b8119..33462fdc7add 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -166,6 +166,258 @@ static const char *phylink_an_mode_str(unsigned int mode) return mode < ARRAY_SIZE(modestr) ? modestr[mode] : "unknown"; } +static void phylink_caps_to_linkmodes(unsigned long *linkmodes, + unsigned long caps) +{ + if (caps & MAC_SYM_PAUSE) + __set_bit(ETHTOOL_LINK_MODE_Pause_BIT, linkmodes); + + if (caps & MAC_ASYM_PAUSE) + __set_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT, linkmodes); + + if (caps & MAC_10HD) + __set_bit(ETHTOOL_LINK_MODE_10baseT_Half_BIT, linkmodes); + + if (caps & MAC_10FD) + __set_bit(ETHTOOL_LINK_MODE_10baseT_Full_BIT, linkmodes); + + if (caps & MAC_100HD) { + __set_bit(ETHTOOL_LINK_MODE_100baseT_Half_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100baseFX_Half_BIT, linkmodes); + } + + if (caps & MAC_100FD) { + __set_bit(ETHTOOL_LINK_MODE_100baseT_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100baseT1_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100baseFX_Full_BIT, linkmodes); + } + + if (caps & MAC_1000HD) + __set_bit(ETHTOOL_LINK_MODE_1000baseT_Half_BIT, linkmodes); + + if (caps & MAC_1000FD) { + __set_bit(ETHTOOL_LINK_MODE_1000baseT_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_1000baseX_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_1000baseT1_Full_BIT, linkmodes); + } + + if (caps & MAC_2500FD) { + __set_bit(ETHTOOL_LINK_MODE_2500baseT_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_2500baseX_Full_BIT, linkmodes); + } + + if (caps & MAC_5000FD) + __set_bit(ETHTOOL_LINK_MODE_5000baseT_Full_BIT, linkmodes); + + if (caps & MAC_10000FD) { + __set_bit(ETHTOOL_LINK_MODE_10000baseT_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseKX4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseKR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseR_FEC_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseCR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseSR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseLR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseLRM_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_10000baseER_Full_BIT, linkmodes); + } + + if (caps & MAC_25000FD) { + __set_bit(ETHTOOL_LINK_MODE_25000baseCR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_25000baseKR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_25000baseSR_Full_BIT, linkmodes); + } + + if (caps & MAC_40000FD) { + __set_bit(ETHTOOL_LINK_MODE_40000baseKR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_40000baseCR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_40000baseSR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_40000baseLR4_Full_BIT, linkmodes); + } + + if (caps & MAC_50000FD) { + __set_bit(ETHTOOL_LINK_MODE_50000baseCR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseKR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseSR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseKR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseSR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseCR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseLR_ER_FR_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_50000baseDR_Full_BIT, linkmodes); + } + + if (caps & MAC_56000FD) { + __set_bit(ETHTOOL_LINK_MODE_56000baseKR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_56000baseCR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_56000baseSR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_56000baseLR4_Full_BIT, linkmodes); + } + + if (caps & MAC_100000FD) { + __set_bit(ETHTOOL_LINK_MODE_100000baseKR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseSR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseKR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseSR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseCR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseLR2_ER2_FR2_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseDR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseKR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseSR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseLR_ER_FR_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseCR_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_100000baseDR_Full_BIT, linkmodes); + } + + if (caps & MAC_200000FD) { + __set_bit(ETHTOOL_LINK_MODE_200000baseKR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseSR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseLR4_ER4_FR4_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseDR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseCR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseKR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseSR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseLR2_ER2_FR2_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseDR2_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_200000baseCR2_Full_BIT, linkmodes); + } + + if (caps & MAC_400000FD) { + __set_bit(ETHTOOL_LINK_MODE_400000baseKR8_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseSR8_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseLR8_ER8_FR8_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseDR8_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseCR8_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseKR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseSR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseLR4_ER4_FR4_Full_BIT, + linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseDR4_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_400000baseCR4_Full_BIT, linkmodes); + } +} + +/** + * phylink_get_linkmodes() - get acceptable link modes + * @linkmodes: ethtool linkmode mask (must be already initialised) + * @interface: phy interface mode defined by &typedef phy_interface_t + * @mac_capabilities: bitmask of MAC capabilities + * + * Set all possible pause, speed and duplex linkmodes in @linkmodes that + * are supported by the @interface mode and @mac_capabilities. @linkmodes + * must have been initialised previously. + */ +void phylink_get_linkmodes(unsigned long *linkmodes, phy_interface_t interface, + unsigned long mac_capabilities) +{ + unsigned long caps = MAC_SYM_PAUSE | MAC_ASYM_PAUSE; + + switch (interface) { + case PHY_INTERFACE_MODE_USXGMII: + caps |= MAC_10000FD | MAC_5000FD | MAC_2500FD; + fallthrough; + + case PHY_INTERFACE_MODE_RGMII_TXID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_ID: + case PHY_INTERFACE_MODE_RGMII: + case PHY_INTERFACE_MODE_QSGMII: + case PHY_INTERFACE_MODE_SGMII: + case PHY_INTERFACE_MODE_GMII: + caps |= MAC_1000HD | MAC_1000FD; + fallthrough; + + case PHY_INTERFACE_MODE_REVRMII: + case PHY_INTERFACE_MODE_RMII: + case PHY_INTERFACE_MODE_REVMII: + case PHY_INTERFACE_MODE_MII: + caps |= MAC_10HD | MAC_10FD; + fallthrough; + + case PHY_INTERFACE_MODE_100BASEX: + caps |= MAC_100HD | MAC_100FD; + break; + + case PHY_INTERFACE_MODE_TBI: + case PHY_INTERFACE_MODE_MOCA: + case PHY_INTERFACE_MODE_RTBI: + case PHY_INTERFACE_MODE_1000BASEX: + caps |= MAC_1000HD; + fallthrough; + case PHY_INTERFACE_MODE_TRGMII: + caps |= MAC_1000FD; + break; + + case PHY_INTERFACE_MODE_2500BASEX: + caps |= MAC_2500FD; + break; + + case PHY_INTERFACE_MODE_5GBASER: + caps |= MAC_5000FD; + break; + + case PHY_INTERFACE_MODE_XGMII: + case PHY_INTERFACE_MODE_RXAUI: + case PHY_INTERFACE_MODE_XAUI: + case PHY_INTERFACE_MODE_10GBASER: + case PHY_INTERFACE_MODE_10GKR: + caps |= MAC_10000FD; + break; + + case PHY_INTERFACE_MODE_25GBASER: + caps |= MAC_25000FD; + break; + + case PHY_INTERFACE_MODE_XLGMII: + caps |= MAC_40000FD; + break; + + case PHY_INTERFACE_MODE_INTERNAL: + caps |= ~0; + break; + + case PHY_INTERFACE_MODE_NA: + case PHY_INTERFACE_MODE_MAX: + case PHY_INTERFACE_MODE_SMII: + break; + } + + phylink_caps_to_linkmodes(linkmodes, caps & mac_capabilities); +} +EXPORT_SYMBOL_GPL(phylink_get_linkmodes); + +/** + * phylink_generic_validate() - generic validate() callback implementation + * @config: a pointer to a &struct phylink_config. + * @supported: ethtool bitmask for supported link modes. + * @state: a pointer to a &struct phylink_link_state. + * + * Generic implementation of the validate() callback that MAC drivers can + * use when they pass the range of supported interfaces and MAC capabilities. + * This makes use of phylink_get_linkmodes(). + */ +void phylink_generic_validate(struct phylink_config *config, + unsigned long *supported, + struct phylink_link_state *state) +{ + __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; + + phylink_set_port_modes(mask); + phylink_set(mask, Autoneg); + phylink_get_linkmodes(mask, state->interface, config->mac_capabilities); + + linkmode_and(supported, supported, mask); + linkmode_and(state->advertising, state->advertising, mask); +} +EXPORT_SYMBOL_GPL(phylink_generic_validate); + static int phylink_validate_any(struct phylink *pl, unsigned long *supported, struct phylink_link_state *state) { diff --git a/include/linux/phylink.h b/include/linux/phylink.h index f037470b6fb3..3563820a1765 100644 --- a/include/linux/phylink.h +++ b/include/linux/phylink.h @@ -20,6 +20,29 @@ enum { MLO_AN_PHY = 0, /* Conventional PHY */ MLO_AN_FIXED, /* Fixed-link mode */ MLO_AN_INBAND, /* In-band protocol */ + + MAC_SYM_PAUSE = BIT(0), + MAC_ASYM_PAUSE = BIT(1), + MAC_10HD = BIT(2), + MAC_10FD = BIT(3), + MAC_10 = MAC_10HD | MAC_10FD, + MAC_100HD = BIT(4), + MAC_100FD = BIT(5), + MAC_100 = MAC_100HD | MAC_100FD, + MAC_1000HD = BIT(6), + MAC_1000FD = BIT(7), + MAC_1000 = MAC_1000HD | MAC_1000FD, + MAC_2500FD = BIT(8), + MAC_5000FD = BIT(9), + MAC_10000FD = BIT(10), + MAC_20000FD = BIT(11), + MAC_25000FD = BIT(12), + MAC_40000FD = BIT(13), + MAC_50000FD = BIT(14), + MAC_56000FD = BIT(15), + MAC_100000FD = BIT(16), + MAC_200000FD = BIT(17), + MAC_400000FD = BIT(18), }; static inline bool phylink_autoneg_inband(unsigned int mode) @@ -69,6 +92,7 @@ enum phylink_op_type { * if MAC link is at %MLO_AN_FIXED mode. * @supported_interfaces: bitmap describing which PHY_INTERFACE_MODE_xxx * are supported by the MAC/PCS. + * @mac_capabilities: MAC pause/speed/duplex capabilities. */ struct phylink_config { struct device *dev; @@ -79,6 +103,7 @@ struct phylink_config { void (*get_fixed_state)(struct phylink_config *config, struct phylink_link_state *state); DECLARE_PHY_INTERFACE_MASK(supported_interfaces); + unsigned long mac_capabilities; }; /** @@ -442,6 +467,12 @@ void pcs_link_up(struct phylink_pcs *pcs, unsigned int mode, phy_interface_t interface, int speed, int duplex); #endif +void phylink_get_linkmodes(unsigned long *linkmodes, phy_interface_t interface, + unsigned long mac_capabilities); +void phylink_generic_validate(struct phylink_config *config, + unsigned long *supported, + struct phylink_link_state *state); + struct phylink *phylink_create(struct phylink_config *, struct fwnode_handle *, phy_interface_t iface, const struct phylink_mac_ops *mac_ops); From 02a0988b98930491db95966fb8086072e47dabb6 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Mon, 15 Nov 2021 10:00:32 +0000 Subject: [PATCH 0194/2048] net: mvneta: use phylink_generic_validate() Convert mvneta to use phylink_generic_validate() for the bulk of its validate() implementation. This network adapter has a restriction that for 802.3z links, autonegotiation must be enabled. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/mvneta.c | 34 ++++----------------------- 1 file changed, 4 insertions(+), 30 deletions(-) diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index 5a7bdca22a63..67a644177880 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -3823,8 +3823,6 @@ static void mvneta_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - /* We only support QSGMII, SGMII, 802.3z and RGMII modes. * When in 802.3z mode, we must have AN enabled: * "Bit 2 Field InBandAnEn In-band Auto-Negotiation enable. ... @@ -3836,34 +3834,7 @@ static void mvneta_validate(struct phylink_config *config, return; } - /* Allow all the expected bits */ - phylink_set(mask, Autoneg); - phylink_set_port_modes(mask); - - /* Asymmetric pause is unsupported */ - phylink_set(mask, Pause); - - /* Half-duplex at speeds higher than 100Mbit is unsupported */ - if (state->interface != PHY_INTERFACE_MODE_2500BASEX) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - } - - if (state->interface == PHY_INTERFACE_MODE_2500BASEX) { - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - } - - if (!phy_interface_mode_is_8023z(state->interface)) { - /* 10M and 100M are only supported in non-802.3z mode */ - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); + phylink_generic_validate(config, supported, state); } static void mvneta_mac_pcs_get_state(struct phylink_config *config, @@ -5166,6 +5137,9 @@ static int mvneta_probe(struct platform_device *pdev) pp->phylink_config.dev = &dev->dev; pp->phylink_config.type = PHYLINK_NETDEV; + pp->phylink_config.mac_capabilities = MAC_SYM_PAUSE | MAC_10 | + MAC_100 | MAC_1000FD | MAC_2500FD; + phy_interface_set_rgmii(pp->phylink_config.supported_interfaces); __set_bit(PHY_INTERFACE_MODE_QSGMII, pp->phylink_config.supported_interfaces); From 5038ffea0c6c9476f1a8288b112609ac5a49f51b Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Mon, 15 Nov 2021 10:00:37 +0000 Subject: [PATCH 0195/2048] net: mvpp2: use phylink_generic_validate() Convert mvpp2 to use phylink_generic_validate() for the bulk of its validate() implementation. This network adapter has a restriction that for 802.3z links, autonegotiation must be enabled. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/marvell/mvpp2/mvpp2_main.c | 58 +++---------------- 1 file changed, 9 insertions(+), 49 deletions(-) diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c index 2b18d89d9756..df6c793f4b1b 100644 --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c @@ -6260,9 +6260,6 @@ static void mvpp2_phylink_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct mvpp2_port *port = mvpp2_phylink_to_port(config); - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - /* When in 802.3z mode, we must have AN enabled: * Bit 2 Field InBandAnEn In-band Auto-Negotiation enable. ... * When = 1 (1000BASE-X) this field must be set to 1. @@ -6271,52 +6268,7 @@ static void mvpp2_phylink_validate(struct phylink_config *config, !phylink_test(state->advertising, Autoneg)) goto empty_set; - phylink_set(mask, Autoneg); - phylink_set_port_modes(mask); - - if (port->priv->global_tx_fc) { - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - } - - switch (state->interface) { - case PHY_INTERFACE_MODE_10GBASER: - case PHY_INTERFACE_MODE_XAUI: - if (mvpp2_port_supports_xlg(port)) { - phylink_set_10g_modes(mask); - phylink_set(mask, 10000baseKR_Full); - } - break; - - case PHY_INTERFACE_MODE_RGMII: - case PHY_INTERFACE_MODE_RGMII_ID: - case PHY_INTERFACE_MODE_RGMII_RXID: - case PHY_INTERFACE_MODE_RGMII_TXID: - case PHY_INTERFACE_MODE_SGMII: - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - break; - - case PHY_INTERFACE_MODE_1000BASEX: - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - break; - - case PHY_INTERFACE_MODE_2500BASEX: - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - break; - - default: - goto empty_set; - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); + phylink_generic_validate(config, supported, state); return; empty_set: @@ -6911,12 +6863,20 @@ static int mvpp2_port_probe(struct platform_device *pdev, if (!mvpp2_use_acpi_compat_mode(port_fwnode)) { port->phylink_config.dev = &dev->dev; port->phylink_config.type = PHYLINK_NETDEV; + port->phylink_config.mac_capabilities = + MAC_2500FD | MAC_1000FD | MAC_100 | MAC_10; + + if (port->priv->global_tx_fc) + port->phylink_config.mac_capabilities |= + MAC_SYM_PAUSE | MAC_ASYM_PAUSE; if (mvpp2_port_supports_xlg(port)) { __set_bit(PHY_INTERFACE_MODE_10GBASER, port->phylink_config.supported_interfaces); __set_bit(PHY_INTERFACE_MODE_XAUI, port->phylink_config.supported_interfaces); + port->phylink_config.mac_capabilities |= + MAC_10000FD; } if (mvpp2_port_supports_rgmii(port)) From 2f6a470d6545841cf1891b87e360d3998ef024c8 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 15 Nov 2021 07:49:46 -0800 Subject: [PATCH 0196/2048] Revert "Merge branch 'mctp-i2c-driver'" This reverts commit 71812af7234f30362b43ccff33f93890ae4c0655, reversing changes made to cc0be1ad686fb29a4d127948486f40b17fb34b50. Wolfram Sang says: Please revert. Besides the driver in net, it modifies the I2C core code. This has not been acked by the I2C maintainer (in this case me). So, please don't pull this in via the net tree. The question raised here (extending SMBus calls to 255 byte) is complicated because we need ABI backwards compatibility. Link: https://lore.kernel.org/all/YZJ9H4eM%2FM7OXVN0@shikoro/ Signed-off-by: Jakub Kicinski --- Documentation/devicetree/bindings/i2c/i2c.txt | 4 - .../bindings/net/mctp-i2c-controller.yaml | 92 -- drivers/i2c/busses/i2c-aspeed.c | 5 +- drivers/i2c/busses/i2c-npcm7xx.c | 3 +- drivers/i2c/i2c-core-smbus.c | 20 +- drivers/i2c/i2c-dev.c | 93 +- drivers/net/mctp/Kconfig | 12 - drivers/net/mctp/Makefile | 1 - drivers/net/mctp/mctp-i2c.c | 982 ------------------ include/linux/i2c.h | 13 - include/uapi/linux/i2c-dev.h | 2 - include/uapi/linux/i2c.h | 7 +- 12 files changed, 25 insertions(+), 1209 deletions(-) delete mode 100644 Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml delete mode 100644 drivers/net/mctp/mctp-i2c.c diff --git a/Documentation/devicetree/bindings/i2c/i2c.txt b/Documentation/devicetree/bindings/i2c/i2c.txt index fc3dd7ec0445..b864916e087f 100644 --- a/Documentation/devicetree/bindings/i2c/i2c.txt +++ b/Documentation/devicetree/bindings/i2c/i2c.txt @@ -95,10 +95,6 @@ wants to support one of the below features, it should adapt these bindings. - smbus-alert states that the optional SMBus-Alert feature apply to this bus. -- mctp-controller - indicates that the system is accessible via this bus as an endpoint for - MCTP over I2C transport. - Required properties (per child device) -------------------------------------- diff --git a/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml b/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml deleted file mode 100644 index afd11c9422fa..000000000000 --- a/Documentation/devicetree/bindings/net/mctp-i2c-controller.yaml +++ /dev/null @@ -1,92 +0,0 @@ -# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/net/mctp-i2c-controller.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: MCTP I2C transport binding - -maintainers: - - Matt Johnston - -description: | - An mctp-i2c-controller defines a local MCTP endpoint on an I2C controller. - MCTP I2C is specified by DMTF DSP0237. - - An mctp-i2c-controller must be attached to an I2C adapter which supports - slave functionality. I2C busses (either directly or as subordinate mux - busses) are attached to the mctp-i2c-controller with a 'mctp-controller' - property on each used bus. Each mctp-controller I2C bus will be presented - to the host system as a separate MCTP I2C instance. - -properties: - compatible: - const: mctp-i2c-controller - - reg: - minimum: 0x40000000 - maximum: 0x4000007f - description: | - 7 bit I2C address of the local endpoint. - I2C_OWN_SLAVE_ADDRESS (1<<30) flag must be set. - -additionalProperties: false - -required: - - compatible - - reg - -examples: - - | - // Basic case of a single I2C bus - #include - - i2c { - #address-cells = <1>; - #size-cells = <0>; - mctp-controller; - - mctp@30 { - compatible = "mctp-i2c-controller"; - reg = <(0x30 | I2C_OWN_SLAVE_ADDRESS)>; - }; - }; - - - | - // Mux topology with multiple MCTP-handling busses under - // a single mctp-i2c-controller. - // i2c1 and i2c6 can have MCTP devices, i2c5 does not. - #include - - i2c1: i2c { - #address-cells = <1>; - #size-cells = <0>; - mctp-controller; - - mctp@50 { - compatible = "mctp-i2c-controller"; - reg = <(0x50 | I2C_OWN_SLAVE_ADDRESS)>; - }; - }; - - i2c-mux { - #address-cells = <1>; - #size-cells = <0>; - i2c-parent = <&i2c1>; - - i2c5: i2c@0 { - #address-cells = <1>; - #size-cells = <0>; - reg = <0>; - eeprom@33 { - reg = <0x33>; - }; - }; - - i2c6: i2c@1 { - #address-cells = <1>; - #size-cells = <0>; - reg = <1>; - mctp-controller; - }; - }; diff --git a/drivers/i2c/busses/i2c-aspeed.c b/drivers/i2c/busses/i2c-aspeed.c index 7395f3702fae..67e8b97c0c95 100644 --- a/drivers/i2c/busses/i2c-aspeed.c +++ b/drivers/i2c/busses/i2c-aspeed.c @@ -533,7 +533,7 @@ static u32 aspeed_i2c_master_irq(struct aspeed_i2c_bus *bus, u32 irq_status) msg->buf[bus->buf_index++] = recv_byte; if (msg->flags & I2C_M_RECV_LEN) { - if (unlikely(recv_byte > I2C_SMBUS_V3_BLOCK_MAX)) { + if (unlikely(recv_byte > I2C_SMBUS_BLOCK_MAX)) { bus->cmd_err = -EPROTO; aspeed_i2c_do_stop(bus); goto out_no_complete; @@ -718,8 +718,7 @@ static int aspeed_i2c_master_xfer(struct i2c_adapter *adap, static u32 aspeed_i2c_functionality(struct i2c_adapter *adap) { - return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL | - I2C_FUNC_SMBUS_BLOCK_DATA | I2C_FUNC_SMBUS_V3_BLOCK; + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL | I2C_FUNC_SMBUS_BLOCK_DATA; } #if IS_ENABLED(CONFIG_I2C_SLAVE) diff --git a/drivers/i2c/busses/i2c-npcm7xx.c b/drivers/i2c/busses/i2c-npcm7xx.c index 6d60f65add85..2ad166355ec9 100644 --- a/drivers/i2c/busses/i2c-npcm7xx.c +++ b/drivers/i2c/busses/i2c-npcm7xx.c @@ -1399,7 +1399,7 @@ static void npcm_i2c_irq_master_handler_read(struct npcm_i2c *bus) if (bus->read_block_use) { /* first byte in block protocol is the size: */ data = npcm_i2c_rd_byte(bus); - data = clamp_val(data, 1, I2C_SMBUS_V3_BLOCK_MAX); + data = clamp_val(data, 1, I2C_SMBUS_BLOCK_MAX); bus->rd_size = data + block_extra_bytes_size; bus->rd_buf[bus->rd_ind++] = data; @@ -2187,7 +2187,6 @@ static u32 npcm_i2c_functionality(struct i2c_adapter *adap) I2C_FUNC_SMBUS_EMUL | I2C_FUNC_SMBUS_BLOCK_DATA | I2C_FUNC_SMBUS_PEC | - I2C_FUNC_SMBUS_V3_BLOCK | I2C_FUNC_SLAVE; } diff --git a/drivers/i2c/i2c-core-smbus.c b/drivers/i2c/i2c-core-smbus.c index 743415584aba..e5b2d1465e7e 100644 --- a/drivers/i2c/i2c-core-smbus.c +++ b/drivers/i2c/i2c-core-smbus.c @@ -303,8 +303,7 @@ static void i2c_smbus_try_get_dmabuf(struct i2c_msg *msg, u8 init_val) bool is_read = msg->flags & I2C_M_RD; unsigned char *dma_buf; - dma_buf = kzalloc(I2C_SMBUS_V3_BLOCK_MAX + (is_read ? 2 : 3), - GFP_KERNEL); + dma_buf = kzalloc(I2C_SMBUS_BLOCK_MAX + (is_read ? 2 : 3), GFP_KERNEL); if (!dma_buf) return; @@ -330,10 +329,9 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, * initialize most things with sane defaults, to keep the code below * somewhat simpler. */ - unsigned char msgbuf0[I2C_SMBUS_V3_BLOCK_MAX+3]; - unsigned char msgbuf1[I2C_SMBUS_V3_BLOCK_MAX+2]; + unsigned char msgbuf0[I2C_SMBUS_BLOCK_MAX+3]; + unsigned char msgbuf1[I2C_SMBUS_BLOCK_MAX+2]; int nmsgs = read_write == I2C_SMBUS_READ ? 2 : 1; - u16 block_max; u8 partial_pec = 0; int status; struct i2c_msg msg[2] = { @@ -352,10 +350,6 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, bool wants_pec = ((flags & I2C_CLIENT_PEC) && size != I2C_SMBUS_QUICK && size != I2C_SMBUS_I2C_BLOCK_DATA); - /* Drivers must opt in to 255 byte max block size */ - block_max = i2c_check_functionality(adapter, I2C_FUNC_SMBUS_V3_BLOCK) - ? I2C_SMBUS_V3_BLOCK_MAX : I2C_SMBUS_BLOCK_MAX; - msgbuf0[0] = command; switch (size) { case I2C_SMBUS_QUICK: @@ -405,7 +399,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, i2c_smbus_try_get_dmabuf(&msg[1], 0); } else { msg[0].len = data->block[0] + 2; - if (msg[0].len > block_max + 2) { + if (msg[0].len > I2C_SMBUS_BLOCK_MAX + 2) { dev_err(&adapter->dev, "Invalid block write size %d\n", data->block[0]); @@ -419,7 +413,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, case I2C_SMBUS_BLOCK_PROC_CALL: nmsgs = 2; /* Another special case */ read_write = I2C_SMBUS_READ; - if (data->block[0] > block_max) { + if (data->block[0] > I2C_SMBUS_BLOCK_MAX) { dev_err(&adapter->dev, "Invalid block write size %d\n", data->block[0]); @@ -436,7 +430,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, i2c_smbus_try_get_dmabuf(&msg[1], 0); break; case I2C_SMBUS_I2C_BLOCK_DATA: - if (data->block[0] > block_max) { + if (data->block[0] > I2C_SMBUS_BLOCK_MAX) { dev_err(&adapter->dev, "Invalid block %s size %d\n", read_write == I2C_SMBUS_READ ? "read" : "write", data->block[0]); @@ -504,7 +498,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr, break; case I2C_SMBUS_BLOCK_DATA: case I2C_SMBUS_BLOCK_PROC_CALL: - if (msg[1].buf[0] > block_max) { + if (msg[1].buf[0] > I2C_SMBUS_BLOCK_MAX) { dev_err(&adapter->dev, "Invalid block size returned: %d\n", msg[1].buf[0]); diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c index 5ee9118c0407..bce0e8bb7852 100644 --- a/drivers/i2c/i2c-dev.c +++ b/drivers/i2c/i2c-dev.c @@ -46,24 +46,6 @@ struct i2c_dev { struct cdev cdev; }; -/* The userspace union i2c_smbus_data for I2C_SMBUS ioctl is limited - * to 32 bytes (I2C_SMBUS_BLOCK_MAX) for compatibility. - */ -union compat_i2c_smbus_data { - __u8 byte; - __u16 word; - __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ - /* and one more for user-space compatibility */ -}; - -/* Must match i2c-dev.h definition with compat .data member */ -struct i2c_smbus_ioctl_data { - __u8 read_write; - __u8 command; - __u32 size; - union compat_i2c_smbus_data __user *data; -}; - #define I2C_MINORS (MINORMASK + 1) static LIST_HEAD(i2c_dev_list); static DEFINE_SPINLOCK(i2c_dev_list_lock); @@ -253,17 +235,14 @@ static int i2cdev_check_addr(struct i2c_adapter *adapter, unsigned int addr) static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, unsigned nmsgs, struct i2c_msg *msgs) { - u8 __user **data_ptrs = NULL; - u16 *orig_lens = NULL; + u8 __user **data_ptrs; int i, res; - res = -ENOMEM; data_ptrs = kmalloc_array(nmsgs, sizeof(u8 __user *), GFP_KERNEL); - if (data_ptrs == NULL) - goto out; - orig_lens = kmalloc_array(nmsgs, sizeof(u16), GFP_KERNEL); - if (orig_lens == NULL) - goto out; + if (data_ptrs == NULL) { + kfree(msgs); + return -ENOMEM; + } res = 0; for (i = 0; i < nmsgs; i++) { @@ -274,30 +253,12 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, } data_ptrs[i] = (u8 __user *)msgs[i].buf; - msgs[i].buf = NULL; - if (msgs[i].len < 1) { - /* Sanity check */ - res = -EINVAL; - break; - - } - /* Allocate a larger buffer to accommodate possible 255 byte - * blocks. Read results will be dropped later - * if they are too large for the original length. - */ - orig_lens[i] = msgs[i].len; - msgs[i].buf = kmalloc(msgs[i].len + I2C_SMBUS_V3_BLOCK_MAX, - GFP_USER | __GFP_NOWARN); + msgs[i].buf = memdup_user(data_ptrs[i], msgs[i].len); if (IS_ERR(msgs[i].buf)) { res = PTR_ERR(msgs[i].buf); break; } - if (copy_from_user(msgs[i].buf, data_ptrs[i], msgs[i].len)) { - kfree(msgs[i].buf); - res = -EFAULT; - break; - } - /* Buffer from kmalloc, so DMA is ok */ + /* memdup_user allocates with GFP_KERNEL, so DMA is ok */ msgs[i].flags |= I2C_M_DMA_SAFE; /* @@ -313,7 +274,7 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, */ if (msgs[i].flags & I2C_M_RECV_LEN) { if (!(msgs[i].flags & I2C_M_RD) || - msgs[i].buf[0] < 1 || + msgs[i].len < 1 || msgs[i].buf[0] < 1 || msgs[i].len < msgs[i].buf[0] + I2C_SMBUS_BLOCK_MAX) { i++; @@ -336,16 +297,12 @@ static noinline int i2cdev_ioctl_rdwr(struct i2c_client *client, res = i2c_transfer(client->adapter, msgs, nmsgs); while (i-- > 0) { if (res >= 0 && (msgs[i].flags & I2C_M_RD)) { - if (orig_lens[i] < msgs[i].len) - res = -EINVAL; - else if (copy_to_user(data_ptrs[i], msgs[i].buf, - msgs[i].len)) + if (copy_to_user(data_ptrs[i], msgs[i].buf, + msgs[i].len)) res = -EFAULT; } kfree(msgs[i].buf); } -out: - kfree(orig_lens); kfree(data_ptrs); kfree(msgs); return res; @@ -353,7 +310,7 @@ out: static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, u8 read_write, u8 command, u32 size, - union compat_i2c_smbus_data __user *data) + union i2c_smbus_data __user *data) { union i2c_smbus_data temp = {}; int datasize, res; @@ -414,16 +371,6 @@ static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, if (copy_from_user(&temp, data, datasize)) return -EFAULT; } - if ((size == I2C_SMBUS_BLOCK_PROC_CALL || - size == I2C_SMBUS_I2C_BLOCK_DATA || - size == I2C_SMBUS_BLOCK_DATA) && - read_write == I2C_SMBUS_WRITE && - temp.block[0] > I2C_SMBUS_BLOCK_MAX) { - /* Don't accept writes larger than the buffer size */ - dev_dbg(&client->adapter->dev, "block write is too large"); - return -EINVAL; - - } if (size == I2C_SMBUS_I2C_BLOCK_BROKEN) { /* Convert old I2C block commands to the new convention. This preserves binary compatibility. */ @@ -433,21 +380,9 @@ static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, } res = i2c_smbus_xfer(client->adapter, client->addr, client->flags, read_write, command, size, &temp); - if (res) - return res; - if ((size == I2C_SMBUS_BLOCK_PROC_CALL || - size == I2C_SMBUS_I2C_BLOCK_DATA || - size == I2C_SMBUS_BLOCK_DATA) && - read_write == I2C_SMBUS_READ && - temp.block[0] > I2C_SMBUS_BLOCK_MAX) { - /* Don't accept reads larger than the buffer size */ - dev_dbg(&client->adapter->dev, "block read is too large"); - return -EINVAL; - - } - if ((size == I2C_SMBUS_PROC_CALL) || - (size == I2C_SMBUS_BLOCK_PROC_CALL) || - (read_write == I2C_SMBUS_READ)) { + if (!res && ((size == I2C_SMBUS_PROC_CALL) || + (size == I2C_SMBUS_BLOCK_PROC_CALL) || + (read_write == I2C_SMBUS_READ))) { if (copy_to_user(data, &temp, datasize)) return -EFAULT; } diff --git a/drivers/net/mctp/Kconfig b/drivers/net/mctp/Kconfig index b758b29c2ddf..d8f966cedc89 100644 --- a/drivers/net/mctp/Kconfig +++ b/drivers/net/mctp/Kconfig @@ -3,18 +3,6 @@ if MCTP menu "MCTP Device Drivers" -config MCTP_TRANSPORT_I2C - tristate "MCTP SMBus/I2C transport" - # i2c-mux is optional, but we must build as a module if i2c-mux is a module - depends on I2C_MUX || !I2C_MUX - depends on I2C - depends on I2C_SLAVE - select MCTP_FLOWS - help - Provides a driver to access MCTP devices over SMBus/I2C transport, - from DMTF specification DSP0237. A MCTP protocol network device is - created for each I2C bus that has been assigned a mctp-i2c device. - endmenu endif diff --git a/drivers/net/mctp/Makefile b/drivers/net/mctp/Makefile index 73dc411986a6..e69de29bb2d1 100644 --- a/drivers/net/mctp/Makefile +++ b/drivers/net/mctp/Makefile @@ -1 +0,0 @@ -obj-$(CONFIG_MCTP_TRANSPORT_I2C) += mctp-i2c.o diff --git a/drivers/net/mctp/mctp-i2c.c b/drivers/net/mctp/mctp-i2c.c deleted file mode 100644 index ed213b4765a1..000000000000 --- a/drivers/net/mctp/mctp-i2c.c +++ /dev/null @@ -1,982 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Management Controller Transport Protocol (MCTP) - * - * Copyright (c) 2021 Code Construct - * Copyright (c) 2021 Google - */ - -#include -#include -#include -#include -#include -#include -#include - -/* SMBus 3.0 allows 255 data bytes (plus PEC), but the - * first byte is taken for source slave address. - */ -#define MCTP_I2C_MAXBLOCK 255 -#define MCTP_I2C_MAXMTU (MCTP_I2C_MAXBLOCK - 1) -#define MCTP_I2C_MINMTU (64 + 4) -/* Allow space for address, command, byte_count, databytes, PEC */ -#define MCTP_I2C_RXBUFSZ (3 + MCTP_I2C_MAXBLOCK + 1) -#define MCTP_I2C_MINLEN 8 -#define MCTP_I2C_COMMANDCODE 0x0f -#define MCTP_I2C_TX_WORK_LEN 100 -// sufficient for 64kB at min mtu -#define MCTP_I2C_TX_QUEUE_LEN 1100 - -#define MCTP_I2C_OF_PROP "mctp-controller" - -enum { - MCTP_I2C_FLOW_STATE_NEW = 0, - MCTP_I2C_FLOW_STATE_ACTIVE, -}; - -static struct { - /* lock protects clients and also prevents adding/removing adapters - * during mctp_i2c_client probe/remove. - */ - struct mutex lock; - // list of struct mctp_i2c_client - struct list_head clients; -} mi_driver_state; - -struct mctp_i2c_client; - -// The netdev structure. One of these per I2C adapter. -struct mctp_i2c_dev { - struct net_device *ndev; - struct i2c_adapter *adapter; - struct mctp_i2c_client *client; - struct list_head list; // for mctp_i2c_client.devs - - size_t pos; - u8 buffer[MCTP_I2C_RXBUFSZ]; - - struct task_struct *tx_thread; - wait_queue_head_t tx_wq; - struct sk_buff_head tx_queue; - - // a fake entry in our tx queue to perform an unlock operation - struct sk_buff unlock_marker; - - spinlock_t flow_lock; // protects i2c_lock_count and release_count - int i2c_lock_count; - int release_count; -}; - -/* The i2c client structure. One per hardware i2c bus at the top of the - * mux tree, shared by multiple netdevs - */ -struct mctp_i2c_client { - struct i2c_client *client; - u8 lladdr; - - struct mctp_i2c_dev *sel; - struct list_head devs; - spinlock_t curr_lock; // protects sel - - struct list_head list; // for mi_driver_state.clients -}; - -// Header on the wire -struct mctp_i2c_hdr { - u8 dest_slave; - u8 command; - u8 byte_count; - u8 source_slave; -}; - -static int mctp_i2c_recv(struct mctp_i2c_dev *midev); -static int mctp_i2c_slave_cb(struct i2c_client *client, - enum i2c_slave_event event, u8 *val); - -static struct i2c_adapter *mux_root_adapter(struct i2c_adapter *adap) -{ -#if IS_ENABLED(CONFIG_I2C_MUX) - return i2c_root_adapter(&adap->dev); -#else - /* In non-mux config all i2c adapters are root adapters */ - return adap; -#endif -} - -static ssize_t mctp_current_mux_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct mctp_i2c_client *mcli = i2c_get_clientdata(to_i2c_client(dev)); - struct net_device *ndev = NULL; - unsigned long flags; - ssize_t l; - - spin_lock_irqsave(&mcli->curr_lock, flags); - if (mcli->sel) { - ndev = mcli->sel->ndev; - dev_hold(ndev); - } - spin_unlock_irqrestore(&mcli->curr_lock, flags); - l = scnprintf(buf, PAGE_SIZE, "%s\n", ndev ? ndev->name : "(none)"); - if (ndev) - dev_put(ndev); - return l; -} -static DEVICE_ATTR_RO(mctp_current_mux); - -/* Creates a new i2c slave device attached to the root adapter. - * Sets up the slave callback. - * Must be called with a client on a root adapter. - */ -static struct mctp_i2c_client *mctp_i2c_new_client(struct i2c_client *client) -{ - struct mctp_i2c_client *mcli = NULL; - struct i2c_adapter *root = NULL; - int rc; - - if (client->flags & I2C_CLIENT_TEN) { - dev_err(&client->dev, "%s failed, MCTP requires a 7-bit I2C address, addr=0x%x", - __func__, client->addr); - rc = -EINVAL; - goto err; - } - - root = mux_root_adapter(client->adapter); - if (!root) { - dev_err(&client->dev, "%s failed to find root adapter\n", __func__); - rc = -ENOENT; - goto err; - } - if (root != client->adapter) { - dev_err(&client->dev, - "A mctp-i2c-controller client cannot be placed on an I2C mux adapter.\n" - " It should be placed on the mux tree root adapter\n" - " then set mctp-controller property on adapters to attach\n"); - rc = -EINVAL; - goto err; - } - - mcli = kzalloc(sizeof(*mcli), GFP_KERNEL); - if (!mcli) { - rc = -ENOMEM; - goto err; - } - spin_lock_init(&mcli->curr_lock); - INIT_LIST_HEAD(&mcli->devs); - INIT_LIST_HEAD(&mcli->list); - mcli->lladdr = client->addr & 0xff; - mcli->client = client; - i2c_set_clientdata(client, mcli); - - rc = i2c_slave_register(mcli->client, mctp_i2c_slave_cb); - if (rc) { - dev_err(&client->dev, "%s i2c register failed %d\n", __func__, rc); - mcli->client = NULL; - i2c_set_clientdata(client, NULL); - goto err; - } - - rc = device_create_file(&client->dev, &dev_attr_mctp_current_mux); - if (rc) { - dev_err(&client->dev, "%s adding sysfs \"%s\" failed %d\n", __func__, - dev_attr_mctp_current_mux.attr.name, rc); - // continue anyway - } - - return mcli; -err: - if (mcli) { - if (mcli->client) { - device_remove_file(&mcli->client->dev, &dev_attr_mctp_current_mux); - i2c_unregister_device(mcli->client); - } - kfree(mcli); - } - return ERR_PTR(rc); -} - -static void mctp_i2c_free_client(struct mctp_i2c_client *mcli) -{ - int rc; - - WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); - WARN_ON(!list_empty(&mcli->devs)); - WARN_ON(mcli->sel); // sanity check, no locking - - device_remove_file(&mcli->client->dev, &dev_attr_mctp_current_mux); - rc = i2c_slave_unregister(mcli->client); - // leak if it fails, we can't propagate errors upwards - if (rc) - dev_err(&mcli->client->dev, "%s i2c unregister failed %d\n", __func__, rc); - else - kfree(mcli); -} - -/* Switch the mctp i2c device to receive responses. - * Call with curr_lock held - */ -static void __mctp_i2c_device_select(struct mctp_i2c_client *mcli, - struct mctp_i2c_dev *midev) -{ - assert_spin_locked(&mcli->curr_lock); - if (midev) - dev_hold(midev->ndev); - if (mcli->sel) - dev_put(mcli->sel->ndev); - mcli->sel = midev; -} - -// Switch the mctp i2c device to receive responses -static void mctp_i2c_device_select(struct mctp_i2c_client *mcli, - struct mctp_i2c_dev *midev) -{ - unsigned long flags; - - spin_lock_irqsave(&mcli->curr_lock, flags); - __mctp_i2c_device_select(mcli, midev); - spin_unlock_irqrestore(&mcli->curr_lock, flags); -} - -static int mctp_i2c_slave_cb(struct i2c_client *client, - enum i2c_slave_event event, u8 *val) -{ - struct mctp_i2c_client *mcli = i2c_get_clientdata(client); - struct mctp_i2c_dev *midev = NULL; - unsigned long flags; - int rc = 0; - - spin_lock_irqsave(&mcli->curr_lock, flags); - midev = mcli->sel; - if (midev) - dev_hold(midev->ndev); - spin_unlock_irqrestore(&mcli->curr_lock, flags); - - if (!midev) - return 0; - - switch (event) { - case I2C_SLAVE_WRITE_RECEIVED: - if (midev->pos < MCTP_I2C_RXBUFSZ) { - midev->buffer[midev->pos] = *val; - midev->pos++; - } else { - midev->ndev->stats.rx_over_errors++; - } - - break; - case I2C_SLAVE_WRITE_REQUESTED: - /* dest_slave as first byte */ - midev->buffer[0] = mcli->lladdr << 1; - midev->pos = 1; - break; - case I2C_SLAVE_STOP: - rc = mctp_i2c_recv(midev); - break; - default: - break; - } - - dev_put(midev->ndev); - return rc; -} - -// Processes incoming data that has been accumulated by the slave cb -static int mctp_i2c_recv(struct mctp_i2c_dev *midev) -{ - struct net_device *ndev = midev->ndev; - struct mctp_i2c_hdr *hdr; - struct mctp_skb_cb *cb; - struct sk_buff *skb; - u8 pec, calc_pec; - size_t recvlen; - - /* + 1 for the PEC */ - if (midev->pos < MCTP_I2C_MINLEN + 1) { - ndev->stats.rx_length_errors++; - return -EINVAL; - } - recvlen = midev->pos - 1; - - hdr = (void *)midev->buffer; - if (hdr->command != MCTP_I2C_COMMANDCODE) { - ndev->stats.rx_dropped++; - return -EINVAL; - } - - pec = midev->buffer[midev->pos - 1]; - calc_pec = i2c_smbus_pec(0, midev->buffer, recvlen); - if (pec != calc_pec) { - ndev->stats.rx_crc_errors++; - return -EINVAL; - } - - skb = netdev_alloc_skb(ndev, recvlen); - if (!skb) { - ndev->stats.rx_dropped++; - return -ENOMEM; - } - - skb->protocol = htons(ETH_P_MCTP); - skb_put_data(skb, midev->buffer, recvlen); - skb_reset_mac_header(skb); - skb_pull(skb, sizeof(struct mctp_i2c_hdr)); - skb_reset_network_header(skb); - - cb = __mctp_cb(skb); - cb->halen = 1; - cb->haddr[0] = hdr->source_slave; - - if (netif_rx(skb) == NET_RX_SUCCESS) { - ndev->stats.rx_packets++; - ndev->stats.rx_bytes += skb->len; - } else { - ndev->stats.rx_dropped++; - } - return 0; -} - -enum mctp_i2c_flow_state { - MCTP_I2C_TX_FLOW_INVALID, - MCTP_I2C_TX_FLOW_NONE, - MCTP_I2C_TX_FLOW_NEW, - MCTP_I2C_TX_FLOW_EXISTING, -}; - -static enum mctp_i2c_flow_state -mctp_i2c_get_tx_flow_state(struct mctp_i2c_dev *midev, struct sk_buff *skb) -{ - enum mctp_i2c_flow_state state; - struct mctp_sk_key *key; - struct mctp_flow *flow; - unsigned long flags; - - flow = skb_ext_find(skb, SKB_EXT_MCTP); - if (!flow) - return MCTP_I2C_TX_FLOW_NONE; - - key = flow->key; - if (!key) - return MCTP_I2C_TX_FLOW_NONE; - - spin_lock_irqsave(&key->lock, flags); - /* if the key is present but invalid, we're unlikely to be able - * to handle the flow at all; just drop now - */ - if (!key->valid) { - state = MCTP_I2C_TX_FLOW_INVALID; - - } else if (key->dev_flow_state == MCTP_I2C_FLOW_STATE_NEW) { - key->dev_flow_state = MCTP_I2C_FLOW_STATE_ACTIVE; - state = MCTP_I2C_TX_FLOW_NEW; - } else { - state = MCTP_I2C_TX_FLOW_EXISTING; - } - - spin_unlock_irqrestore(&key->lock, flags); - - return state; -} - -/* We're not contending with ourselves here; we only need to exclude other - * i2c clients from using the bus. refcounts are simply to prevent - * recursive locking. - */ -static void mctp_i2c_lock_nest(struct mctp_i2c_dev *midev) -{ - unsigned long flags; - bool lock; - - spin_lock_irqsave(&midev->flow_lock, flags); - lock = midev->i2c_lock_count == 0; - midev->i2c_lock_count++; - spin_unlock_irqrestore(&midev->flow_lock, flags); - - if (lock) - i2c_lock_bus(midev->adapter, I2C_LOCK_SEGMENT); -} - -static void mctp_i2c_unlock_nest(struct mctp_i2c_dev *midev) -{ - unsigned long flags; - bool unlock; - - spin_lock_irqsave(&midev->flow_lock, flags); - if (!WARN_ONCE(midev->i2c_lock_count == 0, "lock count underflow!")) - midev->i2c_lock_count--; - unlock = midev->i2c_lock_count == 0; - spin_unlock_irqrestore(&midev->flow_lock, flags); - - if (unlock) - i2c_unlock_bus(midev->adapter, I2C_LOCK_SEGMENT); -} - -static void mctp_i2c_xmit(struct mctp_i2c_dev *midev, struct sk_buff *skb) -{ - struct net_device_stats *stats = &midev->ndev->stats; - enum mctp_i2c_flow_state fs; - union i2c_smbus_data *data; - struct mctp_i2c_hdr *hdr; - unsigned int len; - u16 daddr; - int rc; - - fs = mctp_i2c_get_tx_flow_state(midev, skb); - - len = skb->len; - hdr = (void *)skb_mac_header(skb); - data = (void *)&hdr->byte_count; - daddr = hdr->dest_slave >> 1; - - switch (fs) { - case MCTP_I2C_TX_FLOW_NONE: - /* no flow: full lock & unlock */ - mctp_i2c_lock_nest(midev); - mctp_i2c_device_select(midev->client, midev); - rc = __i2c_smbus_xfer(midev->adapter, daddr, I2C_CLIENT_PEC, - I2C_SMBUS_WRITE, hdr->command, - I2C_SMBUS_BLOCK_DATA, data); - mctp_i2c_unlock_nest(midev); - break; - - case MCTP_I2C_TX_FLOW_NEW: - /* new flow: lock, tx, but don't unlock; that will happen - * on flow release - */ - mctp_i2c_lock_nest(midev); - mctp_i2c_device_select(midev->client, midev); - fallthrough; - - case MCTP_I2C_TX_FLOW_EXISTING: - /* existing flow: we already have the lock; just tx */ - rc = __i2c_smbus_xfer(midev->adapter, daddr, I2C_CLIENT_PEC, - I2C_SMBUS_WRITE, hdr->command, - I2C_SMBUS_BLOCK_DATA, data); - break; - - case MCTP_I2C_TX_FLOW_INVALID: - return; - } - - if (rc) { - dev_warn_ratelimited(&midev->adapter->dev, - "%s i2c_smbus_xfer failed %d", __func__, rc); - stats->tx_errors++; - } else { - stats->tx_bytes += len; - stats->tx_packets++; - } -} - -static void mctp_i2c_flow_release(struct mctp_i2c_dev *midev) -{ - unsigned long flags; - bool unlock; - - spin_lock_irqsave(&midev->flow_lock, flags); - if (midev->release_count > midev->i2c_lock_count) { - WARN_ONCE(1, "release count overflow"); - midev->release_count = midev->i2c_lock_count; - } - - midev->i2c_lock_count -= midev->release_count; - unlock = midev->i2c_lock_count == 0 && midev->release_count > 0; - midev->release_count = 0; - spin_unlock_irqrestore(&midev->flow_lock, flags); - - if (unlock) - i2c_unlock_bus(midev->adapter, I2C_LOCK_SEGMENT); -} - -static int mctp_i2c_header_create(struct sk_buff *skb, struct net_device *dev, - unsigned short type, const void *daddr, - const void *saddr, unsigned int len) -{ - struct mctp_i2c_hdr *hdr; - struct mctp_hdr *mhdr; - u8 lldst, llsrc; - - lldst = *((u8 *)daddr); - llsrc = *((u8 *)saddr); - - skb_push(skb, sizeof(struct mctp_i2c_hdr)); - skb_reset_mac_header(skb); - hdr = (void *)skb_mac_header(skb); - mhdr = mctp_hdr(skb); - hdr->dest_slave = (lldst << 1) & 0xff; - hdr->command = MCTP_I2C_COMMANDCODE; - hdr->byte_count = len + 1; - if (hdr->byte_count > MCTP_I2C_MAXBLOCK) - return -EMSGSIZE; - hdr->source_slave = ((llsrc << 1) & 0xff) | 0x01; - mhdr->ver = 0x01; - - return 0; -} - -static int mctp_i2c_tx_thread(void *data) -{ - struct mctp_i2c_dev *midev = data; - struct sk_buff *skb; - unsigned long flags; - - for (;;) { - if (kthread_should_stop()) - break; - - spin_lock_irqsave(&midev->tx_queue.lock, flags); - skb = __skb_dequeue(&midev->tx_queue); - if (netif_queue_stopped(midev->ndev)) - netif_wake_queue(midev->ndev); - spin_unlock_irqrestore(&midev->tx_queue.lock, flags); - - if (skb == &midev->unlock_marker) { - mctp_i2c_flow_release(midev); - - } else if (skb) { - mctp_i2c_xmit(midev, skb); - kfree_skb(skb); - - } else { - wait_event(midev->tx_wq, - !skb_queue_empty(&midev->tx_queue) || - kthread_should_stop()); - } - } - - return 0; -} - -static netdev_tx_t mctp_i2c_start_xmit(struct sk_buff *skb, - struct net_device *dev) -{ - struct mctp_i2c_dev *midev = netdev_priv(dev); - unsigned long flags; - - spin_lock_irqsave(&midev->tx_queue.lock, flags); - if (skb_queue_len(&midev->tx_queue) >= MCTP_I2C_TX_WORK_LEN) { - netif_stop_queue(dev); - spin_unlock_irqrestore(&midev->tx_queue.lock, flags); - netdev_err(dev, "BUG! Tx Ring full when queue awake!\n"); - return NETDEV_TX_BUSY; - } - - __skb_queue_tail(&midev->tx_queue, skb); - if (skb_queue_len(&midev->tx_queue) == MCTP_I2C_TX_WORK_LEN) - netif_stop_queue(dev); - spin_unlock_irqrestore(&midev->tx_queue.lock, flags); - - wake_up(&midev->tx_wq); - return NETDEV_TX_OK; -} - -static void mctp_i2c_release_flow(struct mctp_dev *mdev, - struct mctp_sk_key *key) - -{ - struct mctp_i2c_dev *midev = netdev_priv(mdev->dev); - unsigned long flags; - - spin_lock_irqsave(&midev->flow_lock, flags); - midev->release_count++; - spin_unlock_irqrestore(&midev->flow_lock, flags); - - /* Ensure we have a release operation queued, through the fake - * marker skb - */ - spin_lock(&midev->tx_queue.lock); - if (!midev->unlock_marker.next) - __skb_queue_tail(&midev->tx_queue, &midev->unlock_marker); - spin_unlock(&midev->tx_queue.lock); - - wake_up(&midev->tx_wq); -} - -static const struct net_device_ops mctp_i2c_ops = { - .ndo_start_xmit = mctp_i2c_start_xmit, -}; - -static const struct header_ops mctp_i2c_headops = { - .create = mctp_i2c_header_create, -}; - -static const struct mctp_netdev_ops mctp_i2c_mctp_ops = { - .release_flow = mctp_i2c_release_flow, -}; - -static void mctp_i2c_net_setup(struct net_device *dev) -{ - dev->type = ARPHRD_MCTP; - - dev->mtu = MCTP_I2C_MAXMTU; - dev->min_mtu = MCTP_I2C_MINMTU; - dev->max_mtu = MCTP_I2C_MAXMTU; - dev->tx_queue_len = MCTP_I2C_TX_QUEUE_LEN; - - dev->hard_header_len = sizeof(struct mctp_i2c_hdr); - dev->addr_len = 1; - - dev->netdev_ops = &mctp_i2c_ops; - dev->header_ops = &mctp_i2c_headops; - dev->needs_free_netdev = true; -} - -static int mctp_i2c_add_netdev(struct mctp_i2c_client *mcli, - struct i2c_adapter *adap) -{ - unsigned long flags; - struct mctp_i2c_dev *midev = NULL; - struct net_device *ndev = NULL; - struct i2c_adapter *root; - char namebuf[30]; - int rc; - - root = mux_root_adapter(adap); - if (root != mcli->client->adapter) { - dev_err(&mcli->client->dev, - "I2C adapter %s is not a child bus of %s", - mcli->client->adapter->name, root->name); - return -EINVAL; - } - - WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); - snprintf(namebuf, sizeof(namebuf), "mctpi2c%d", adap->nr); - ndev = alloc_netdev(sizeof(*midev), namebuf, NET_NAME_ENUM, mctp_i2c_net_setup); - if (!ndev) { - dev_err(&mcli->client->dev, "%s alloc netdev failed\n", __func__); - rc = -ENOMEM; - goto err; - } - dev_net_set(ndev, current->nsproxy->net_ns); - SET_NETDEV_DEV(ndev, &adap->dev); - ndev->dev_addr = &mcli->lladdr; - - midev = netdev_priv(ndev); - skb_queue_head_init(&midev->tx_queue); - INIT_LIST_HEAD(&midev->list); - midev->adapter = adap; - midev->client = mcli; - spin_lock_init(&midev->flow_lock); - midev->i2c_lock_count = 0; - midev->release_count = 0; - /* Hold references */ - get_device(&midev->adapter->dev); - get_device(&midev->client->client->dev); - midev->ndev = ndev; - init_waitqueue_head(&midev->tx_wq); - midev->tx_thread = kthread_create(mctp_i2c_tx_thread, midev, - "%s/tx", namebuf); - if (IS_ERR_OR_NULL(midev->tx_thread)) { - rc = -ENOMEM; - goto err_free; - } - - rc = mctp_register_netdev(ndev, &mctp_i2c_mctp_ops); - if (rc) { - dev_err(&mcli->client->dev, - "%s register netdev \"%s\" failed %d\n", __func__, - ndev->name, rc); - goto err_stop_kthread; - } - spin_lock_irqsave(&mcli->curr_lock, flags); - list_add(&midev->list, &mcli->devs); - // Select a device by default - if (!mcli->sel) - __mctp_i2c_device_select(mcli, midev); - spin_unlock_irqrestore(&mcli->curr_lock, flags); - - wake_up_process(midev->tx_thread); - - return 0; - -err_stop_kthread: - kthread_stop(midev->tx_thread); - -err_free: - free_netdev(ndev); - -err: - return rc; -} - -// Removes and unregisters a mctp-i2c netdev -static void mctp_i2c_free_netdev(struct mctp_i2c_dev *midev) -{ - struct mctp_i2c_client *mcli = midev->client; - unsigned long flags; - - netif_stop_queue(midev->ndev); - kthread_stop(midev->tx_thread); - skb_queue_purge(&midev->tx_queue); - - /* Release references, used only for TX which has stopped */ - put_device(&midev->adapter->dev); - put_device(&mcli->client->dev); - - /* Remove it from the parent mcli */ - spin_lock_irqsave(&mcli->curr_lock, flags); - list_del(&midev->list); - if (mcli->sel == midev) { - struct mctp_i2c_dev *first; - - first = list_first_entry_or_null(&mcli->devs, struct mctp_i2c_dev, list); - __mctp_i2c_device_select(mcli, first); - } - spin_unlock_irqrestore(&mcli->curr_lock, flags); - - /* Remove netdev. mctp_i2c_slave_cb() takes a dev_hold() so removing - * it now is safe. unregister_netdev() frees ndev and midev. - */ - mctp_unregister_netdev(midev->ndev); -} - -// Removes any netdev for adap. mcli is the parent root i2c client -static void mctp_i2c_remove_netdev(struct mctp_i2c_client *mcli, - struct i2c_adapter *adap) -{ - unsigned long flags; - struct mctp_i2c_dev *midev = NULL, *m = NULL; - - WARN_ON(!mutex_is_locked(&mi_driver_state.lock)); - spin_lock_irqsave(&mcli->curr_lock, flags); - // list size is limited by number of MCTP netdevs on a single hardware bus - list_for_each_entry(m, &mcli->devs, list) - if (m->adapter == adap) { - midev = m; - break; - } - spin_unlock_irqrestore(&mcli->curr_lock, flags); - - if (midev) - mctp_i2c_free_netdev(midev); -} - -/* Determines whether a device is an i2c adapter. - * Optionally returns the root i2c_adapter - */ -static struct i2c_adapter *mctp_i2c_get_adapter(struct device *dev, - struct i2c_adapter **ret_root) -{ - struct i2c_adapter *root, *adap; - - if (dev->type != &i2c_adapter_type) - return NULL; - adap = to_i2c_adapter(dev); - root = mux_root_adapter(adap); - WARN_ONCE(!root, "%s failed to find root adapter for %s\n", - __func__, dev_name(dev)); - if (!root) - return NULL; - if (ret_root) - *ret_root = root; - return adap; -} - -/* Determines whether a device is an i2c adapter with the "mctp-controller" - * devicetree property set. If adap is not an OF node, returns match_no_of - */ -static bool mctp_i2c_adapter_match(struct i2c_adapter *adap, bool match_no_of) -{ - if (!adap->dev.of_node) - return match_no_of; - return of_property_read_bool(adap->dev.of_node, MCTP_I2C_OF_PROP); -} - -/* Called for each existing i2c device (adapter or client) when a - * new mctp-i2c client is probed. - */ -static int mctp_i2c_client_try_attach(struct device *dev, void *data) -{ - struct i2c_adapter *adap = NULL, *root = NULL; - struct mctp_i2c_client *mcli = data; - - adap = mctp_i2c_get_adapter(dev, &root); - if (!adap) - return 0; - if (mcli->client->adapter != root) - return 0; - // Must either have mctp-controller property on the adapter, or - // be a root adapter if it's non-devicetree - if (!mctp_i2c_adapter_match(adap, adap == root)) - return 0; - - return mctp_i2c_add_netdev(mcli, adap); -} - -static void mctp_i2c_notify_add(struct device *dev) -{ - struct mctp_i2c_client *mcli = NULL, *m = NULL; - struct i2c_adapter *root = NULL, *adap = NULL; - int rc; - - adap = mctp_i2c_get_adapter(dev, &root); - if (!adap) - return; - // Check for mctp-controller property on the adapter - if (!mctp_i2c_adapter_match(adap, false)) - return; - - /* Find an existing mcli for adap's root */ - mutex_lock(&mi_driver_state.lock); - list_for_each_entry(m, &mi_driver_state.clients, list) { - if (m->client->adapter == root) { - mcli = m; - break; - } - } - - if (mcli) { - rc = mctp_i2c_add_netdev(mcli, adap); - if (rc) - dev_warn(dev, "%s Failed adding mctp-i2c device", - __func__); - } - mutex_unlock(&mi_driver_state.lock); -} - -static void mctp_i2c_notify_del(struct device *dev) -{ - struct i2c_adapter *root = NULL, *adap = NULL; - struct mctp_i2c_client *mcli = NULL; - - adap = mctp_i2c_get_adapter(dev, &root); - if (!adap) - return; - - mutex_lock(&mi_driver_state.lock); - list_for_each_entry(mcli, &mi_driver_state.clients, list) { - if (mcli->client->adapter == root) { - mctp_i2c_remove_netdev(mcli, adap); - break; - } - } - mutex_unlock(&mi_driver_state.lock); -} - -static int mctp_i2c_probe(struct i2c_client *client) -{ - struct mctp_i2c_client *mcli = NULL; - int rc; - - /* Check for >32 byte block support required for MCTP */ - if (!i2c_check_functionality(client->adapter, I2C_FUNC_SMBUS_V3_BLOCK)) { - dev_err(&client->dev, - "%s failed, I2C bus driver does not support 255 byte block transfer\n", - __func__); - return -EOPNOTSUPP; - } - - mutex_lock(&mi_driver_state.lock); - mcli = mctp_i2c_new_client(client); - if (IS_ERR(mcli)) { - rc = PTR_ERR(mcli); - mcli = NULL; - goto out; - } else { - list_add(&mcli->list, &mi_driver_state.clients); - } - - // Add a netdev for adapters that have a 'mctp-controller' property - i2c_for_each_dev(mcli, mctp_i2c_client_try_attach); - rc = 0; -out: - mutex_unlock(&mi_driver_state.lock); - return rc; -} - -static int mctp_i2c_remove(struct i2c_client *client) -{ - struct mctp_i2c_client *mcli = i2c_get_clientdata(client); - struct mctp_i2c_dev *midev = NULL, *tmp = NULL; - - mutex_lock(&mi_driver_state.lock); - list_del(&mcli->list); - // Remove all child adapter netdevs - list_for_each_entry_safe(midev, tmp, &mcli->devs, list) - mctp_i2c_free_netdev(midev); - - mctp_i2c_free_client(mcli); - mutex_unlock(&mi_driver_state.lock); - // Callers ignore return code - return 0; -} - -/* We look for a 'mctp-controller' property on I2C busses as they are - * added/deleted, creating/removing netdevs as required. - */ -static int mctp_i2c_notifier_call(struct notifier_block *nb, - unsigned long action, void *data) -{ - struct device *dev = data; - - switch (action) { - case BUS_NOTIFY_ADD_DEVICE: - mctp_i2c_notify_add(dev); - break; - case BUS_NOTIFY_DEL_DEVICE: - mctp_i2c_notify_del(dev); - break; - } - return NOTIFY_DONE; -} - -static struct notifier_block mctp_i2c_notifier = { - .notifier_call = mctp_i2c_notifier_call, -}; - -static const struct i2c_device_id mctp_i2c_id[] = { - { "mctp-i2c", 0 }, - {}, -}; -MODULE_DEVICE_TABLE(i2c, mctp_i2c_id); - -static const struct of_device_id mctp_i2c_of_match[] = { - { .compatible = "mctp-i2c-controller" }, - {}, -}; -MODULE_DEVICE_TABLE(of, mctp_i2c_of_match); - -static struct i2c_driver mctp_i2c_driver = { - .driver = { - .name = "mctp-i2c", - .of_match_table = mctp_i2c_of_match, - }, - .probe_new = mctp_i2c_probe, - .remove = mctp_i2c_remove, - .id_table = mctp_i2c_id, -}; - -static __init int mctp_i2c_init(void) -{ - int rc; - - INIT_LIST_HEAD(&mi_driver_state.clients); - mutex_init(&mi_driver_state.lock); - pr_info("MCTP SMBus/I2C transport driver\n"); - rc = i2c_add_driver(&mctp_i2c_driver); - if (rc) - return rc; - rc = bus_register_notifier(&i2c_bus_type, &mctp_i2c_notifier); - if (rc) { - i2c_del_driver(&mctp_i2c_driver); - return rc; - } - return 0; -} - -static __exit void mctp_i2c_exit(void) -{ - int rc; - - rc = bus_unregister_notifier(&i2c_bus_type, &mctp_i2c_notifier); - if (rc) - pr_warn("%s Could not unregister notifier, %d", __func__, rc); - i2c_del_driver(&mctp_i2c_driver); -} - -module_init(mctp_i2c_init); -module_exit(mctp_i2c_exit); - -MODULE_DESCRIPTION("MCTP SMBus/I2C device"); -MODULE_LICENSE("GPL v2"); -MODULE_AUTHOR("Matt Johnston "); diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 353d6b4e7a53..16119ac1aa97 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -52,19 +52,6 @@ typedef int (*i2c_slave_cb_t)(struct i2c_client *client, struct module; struct property_entry; -/* SMBus 3.0 extends the maximum block read/write size to 255 (from 32). - * The larger size is only supported by some drivers, indicated by - * the I2C_FUNC_SMBUS_V3_BLOCK functionality bit. - */ -#define I2C_SMBUS_V3_BLOCK_MAX 255 /* As specified in SMBus 3.0 standard */ - -/* Note compatibility definition in uapi header with 32 byte block */ -union i2c_smbus_data { - __u8 byte; - __u16 word; - __u8 block[I2C_SMBUS_V3_BLOCK_MAX + 1]; /* block[0] is used for length */ -}; - #if IS_ENABLED(CONFIG_I2C) /* Return the Frequency mode string based on the bus frequency */ const char *i2c_freq_mode_string(u32 bus_freq_hz); diff --git a/include/uapi/linux/i2c-dev.h b/include/uapi/linux/i2c-dev.h index 46ce31d42f7d..1c4cec4ddd84 100644 --- a/include/uapi/linux/i2c-dev.h +++ b/include/uapi/linux/i2c-dev.h @@ -39,14 +39,12 @@ /* This is the structure as used in the I2C_SMBUS ioctl call */ -#ifndef __KERNEL__ struct i2c_smbus_ioctl_data { __u8 read_write; __u8 command; __u32 size; union i2c_smbus_data __user *data; }; -#endif /* This is the structure as used in the I2C_RDWR ioctl call */ struct i2c_rdwr_ioctl_data { diff --git a/include/uapi/linux/i2c.h b/include/uapi/linux/i2c.h index c3534ab1ae53..92326ebde350 100644 --- a/include/uapi/linux/i2c.h +++ b/include/uapi/linux/i2c.h @@ -108,9 +108,6 @@ struct i2c_msg { #define I2C_FUNC_SMBUS_READ_I2C_BLOCK 0x04000000 /* I2C-like block xfer */ #define I2C_FUNC_SMBUS_WRITE_I2C_BLOCK 0x08000000 /* w/ 1-byte reg. addr. */ #define I2C_FUNC_SMBUS_HOST_NOTIFY 0x10000000 /* SMBus 2.0 or later */ -#define I2C_FUNC_SMBUS_V3_BLOCK 0x20000000 /* Device supports 255 byte block */ - /* Note that I2C_SMBUS ioctl only */ - /* supports a 32 byte block */ #define I2C_FUNC_SMBUS_BYTE (I2C_FUNC_SMBUS_READ_BYTE | \ I2C_FUNC_SMBUS_WRITE_BYTE) @@ -140,15 +137,13 @@ struct i2c_msg { /* * Data for SMBus Messages */ -#define I2C_SMBUS_BLOCK_MAX 32 /* As specified in SMBus 2.0 standard */ -#ifndef __KERNEL__ +#define I2C_SMBUS_BLOCK_MAX 32 /* As specified in SMBus standard */ union i2c_smbus_data { __u8 byte; __u16 word; __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ /* and one more for user-space compatibility */ }; -#endif /* i2c_smbus_xfer read or write markers */ #define I2C_SMBUS_READ 1 From 3ad4b7c81a992463c29ae130332c217607fe4452 Mon Sep 17 00:00:00 2001 From: Sean Anderson Date: Fri, 12 Nov 2021 14:04:00 -0500 Subject: [PATCH 0197/2048] net: macb: Fix several edge cases in validate There were several cases where validate() would return bogus supported modes with unusual combinations of interfaces and capabilities. For example, if state->interface was 10GBASER and the macb had HIGH_SPEED and PCS but not GIGABIT MODE, then 10/100 modes would be set anyway. In another case, SGMII could be enabled even if the mac was not a GEM (despite this being checked for later on in mac_config()). These inconsistencies make it difficult to refactor this function cleanly. There is still the open question of what exactly the requirements for SGMII and 10GBASER are, and what SGMII actually supports. If someone from Cadence (or anyone else with access to the GEM/MACB datasheet) could comment on this, it would be greatly appreciated. In particular, what is supported by Cadence vs. vendor extension/limitation? To address this, the current logic is split into three parts. First, we determine what we support, then we eliminate unsupported interfaces, and finally we set the appropriate link modes. There is still some cruft related to NA, but this can be removed in a future patch. Signed-off-by: Sean Anderson Reviewed-by: Parshuram Thombare Reviewed-by: Russell King (Oracle) Link: https://lore.kernel.org/r/20211112190400.1937855-1-sean.anderson@seco.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/cadence/macb_main.c | 114 +++++++++++++++-------- 1 file changed, 74 insertions(+), 40 deletions(-) diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c index ffce528aa00e..57c5f48d19a4 100644 --- a/drivers/net/ethernet/cadence/macb_main.c +++ b/drivers/net/ethernet/cadence/macb_main.c @@ -513,29 +513,47 @@ static void macb_validate(struct phylink_config *config, struct net_device *ndev = to_net_dev(config->dev); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; struct macb *bp = netdev_priv(ndev); + bool have_1g, have_sgmii, have_10g; - /* We only support MII, RMII, GMII, RGMII & SGMII. */ - if (state->interface != PHY_INTERFACE_MODE_NA && - state->interface != PHY_INTERFACE_MODE_MII && - state->interface != PHY_INTERFACE_MODE_RMII && - state->interface != PHY_INTERFACE_MODE_GMII && - state->interface != PHY_INTERFACE_MODE_SGMII && - state->interface != PHY_INTERFACE_MODE_10GBASER && - !phy_interface_mode_is_rgmii(state->interface)) { - linkmode_zero(supported); - return; + /* Determine what modes are supported */ + if (macb_is_gem(bp) && + (bp->caps & MACB_CAPS_GIGABIT_MODE_AVAILABLE)) { + have_1g = true; + if (bp->caps & MACB_CAPS_PCS) + have_sgmii = true; + if (bp->caps & MACB_CAPS_HIGH_SPEED) + have_10g = true; } - if (!macb_is_gem(bp) && - (state->interface == PHY_INTERFACE_MODE_GMII || - phy_interface_mode_is_rgmii(state->interface))) { + /* Eliminate unsupported modes */ + switch (state->interface) { + case PHY_INTERFACE_MODE_NA: + case PHY_INTERFACE_MODE_MII: + case PHY_INTERFACE_MODE_RMII: + break; + + case PHY_INTERFACE_MODE_GMII: + case PHY_INTERFACE_MODE_RGMII: + case PHY_INTERFACE_MODE_RGMII_ID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_TXID: + if (have_1g) + break; linkmode_zero(supported); return; - } - if (state->interface == PHY_INTERFACE_MODE_10GBASER && - !(bp->caps & MACB_CAPS_HIGH_SPEED && - bp->caps & MACB_CAPS_PCS)) { + case PHY_INTERFACE_MODE_SGMII: + if (have_sgmii) + break; + linkmode_zero(supported); + return; + + case PHY_INTERFACE_MODE_10GBASER: + if (have_10g) + break; + fallthrough; + + default: linkmode_zero(supported); return; } @@ -544,32 +562,48 @@ static void macb_validate(struct phylink_config *config, phylink_set(mask, Autoneg); phylink_set(mask, Asym_Pause); - if (bp->caps & MACB_CAPS_GIGABIT_MODE_AVAILABLE && - (state->interface == PHY_INTERFACE_MODE_NA || - state->interface == PHY_INTERFACE_MODE_10GBASER)) { - phylink_set_10g_modes(mask); - phylink_set(mask, 10000baseKR_Full); + /* And set the appropriate mask */ + switch (state->interface) { + case PHY_INTERFACE_MODE_NA: + case PHY_INTERFACE_MODE_10GBASER: + if (have_10g) { + phylink_set_10g_modes(mask); + phylink_set(mask, 10000baseKR_Full); + } if (state->interface != PHY_INTERFACE_MODE_NA) - goto out; + break; + fallthrough; + + /* FIXME: Do we actually support 10/100 for SGMII? Half duplex? */ + case PHY_INTERFACE_MODE_SGMII: + if (!have_sgmii && state->interface != PHY_INTERFACE_MODE_NA) + break; + fallthrough; + + case PHY_INTERFACE_MODE_GMII: + case PHY_INTERFACE_MODE_RGMII: + case PHY_INTERFACE_MODE_RGMII_ID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_TXID: + if (have_1g) { + phylink_set(mask, 1000baseT_Full); + phylink_set(mask, 1000baseX_Full); + + if (!(bp->caps & MACB_CAPS_NO_GIGABIT_HALF)) + phylink_set(mask, 1000baseT_Half); + } else if (state->interface != PHY_INTERFACE_MODE_NA) { + break; + } + fallthrough; + + default: + phylink_set(mask, 10baseT_Half); + phylink_set(mask, 10baseT_Full); + phylink_set(mask, 100baseT_Half); + phylink_set(mask, 100baseT_Full); + break; } - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - - if (bp->caps & MACB_CAPS_GIGABIT_MODE_AVAILABLE && - (state->interface == PHY_INTERFACE_MODE_NA || - state->interface == PHY_INTERFACE_MODE_GMII || - state->interface == PHY_INTERFACE_MODE_SGMII || - phy_interface_mode_is_rgmii(state->interface))) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - - if (!(bp->caps & MACB_CAPS_NO_GIGABIT_HALF)) - phylink_set(mask, 1000baseT_Half); - } -out: linkmode_and(supported, supported, mask); linkmode_and(state->advertising, state->advertising, mask); } From e47d0bf800e8d7f4de501987b2788c7f2ce22cd1 Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Mon, 15 Nov 2021 16:04:48 -0800 Subject: [PATCH 0198/2048] bpftool: Add current libbpf_strict mode to version output + bpftool --legacy --version bpftool v5.15.0 features: libbfd, skeletons + bpftool --version bpftool v5.15.0 features: libbfd, libbpf_strict, skeletons + bpftool --legacy --help Usage: bpftool [OPTIONS] OBJECT { COMMAND | help } bpftool batch file FILE bpftool version OBJECT := { prog | map | link | cgroup | perf | net | feature | btf | gen | struct_ops | iter } OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy} | {-V|--version} } + bpftool --help Usage: bpftool [OPTIONS] OBJECT { COMMAND | help } bpftool batch file FILE bpftool version OBJECT := { prog | map | link | cgroup | perf | net | feature | btf | gen | struct_ops | iter } OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy} | {-V|--version} } + bpftool --legacy Usage: bpftool [OPTIONS] OBJECT { COMMAND | help } bpftool batch file FILE bpftool version OBJECT := { prog | map | link | cgroup | perf | net | feature | btf | gen | struct_ops | iter } OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy} | {-V|--version} } + bpftool Usage: bpftool [OPTIONS] OBJECT { COMMAND | help } bpftool batch file FILE bpftool version OBJECT := { prog | map | link | cgroup | perf | net | feature | btf | gen | struct_ops | iter } OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy} | {-V|--version} } + bpftool --legacy version bpftool v5.15.0 features: libbfd, skeletons + bpftool version bpftool v5.15.0 features: libbfd, libbpf_strict, skeletons + bpftool --json --legacy version {"version":"5.15.0","features":{"libbfd":true,"libbpf_strict":false,"skeletons":true}} + bpftool --json version {"version":"5.15.0","features":{"libbfd":true,"libbpf_strict":true,"skeletons":true}} Suggested-by: Quentin Monnet Signed-off-by: Stanislav Fomichev Signed-off-by: Daniel Borkmann Reviewed-by: Quentin Monnet Link: https://lore.kernel.org/bpf/20211116000448.2918854-1-sdf@google.com --- tools/bpf/bpftool/main.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/tools/bpf/bpftool/main.c b/tools/bpf/bpftool/main.c index 473791e87f7d..8b71500e7cb2 100644 --- a/tools/bpf/bpftool/main.c +++ b/tools/bpf/bpftool/main.c @@ -93,6 +93,7 @@ static int do_version(int argc, char **argv) jsonw_name(json_wtr, "features"); jsonw_start_object(json_wtr); /* features */ jsonw_bool_field(json_wtr, "libbfd", has_libbfd); + jsonw_bool_field(json_wtr, "libbpf_strict", !legacy_libbpf); jsonw_bool_field(json_wtr, "skeletons", has_skeletons); jsonw_end_object(json_wtr); /* features */ @@ -106,6 +107,10 @@ static int do_version(int argc, char **argv) printf(" libbfd"); nb_features++; } + if (!legacy_libbpf) { + printf("%s libbpf_strict", nb_features++ ? "," : ""); + nb_features++; + } if (has_skeletons) printf("%s skeletons", nb_features++ ? "," : ""); printf("\n"); @@ -400,6 +405,7 @@ int main(int argc, char **argv) { "legacy", no_argument, NULL, 'l' }, { 0 } }; + bool version_requested = false; int opt, ret; last_do_help = do_help; @@ -414,7 +420,8 @@ int main(int argc, char **argv) options, NULL)) >= 0) { switch (opt) { case 'V': - return do_version(argc, argv); + version_requested = true; + break; case 'h': return do_help(argc, argv); case 'p': @@ -479,6 +486,9 @@ int main(int argc, char **argv) if (argc < 0) usage(); + if (version_requested) + return do_version(argc, argv); + ret = cmd_select(cmds, argc, argv, do_help); if (json_output) From 69a055d546156adc6f7727ec981f721d5ba9231a Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Mon, 15 Nov 2021 08:39:37 -0800 Subject: [PATCH 0199/2048] libbpf: Fix a couple of missed btf_type_tag handling in btf.c Commit 2dc1e488e5cd ("libbpf: Support BTF_KIND_TYPE_TAG") added the BTF_KIND_TYPE_TAG support. But to test vmlinux build with ... #define __user __attribute__((btf_type_tag("user"))) ... I needed to sync libbpf repo and manually copy libbpf sources to pahole. To simplify process, I used BTF_KIND_RESTRICT to simulate the BTF_KIND_TYPE_TAG with vmlinux build as "restrict" modifier is barely used in kernel. But this approach missed one case in dedup with structures where BTF_KIND_RESTRICT is handled and BTF_KIND_TYPE_TAG is not handled in btf_dedup_is_equiv(), and this will result in a pahole dedup failure. This patch fixed this issue and a selftest is added in the subsequent patch to test this scenario. The other missed handling is in btf__resolve_size(). Currently the compiler always emit like PTR->TYPE_TAG->... so in practice we don't hit the missing BTF_KIND_TYPE_TAG handling issue with compiler generated code. But lets add case BTF_KIND_TYPE_TAG in the switch statement to be future proof. Fixes: 2dc1e488e5cd ("libbpf: Support BTF_KIND_TYPE_TAG") Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211115163937.3922235-1-yhs@fb.com --- tools/lib/bpf/btf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index fadf089ae8fe..b6be579e0dc6 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -610,6 +610,7 @@ __s64 btf__resolve_size(const struct btf *btf, __u32 type_id) case BTF_KIND_RESTRICT: case BTF_KIND_VAR: case BTF_KIND_DECL_TAG: + case BTF_KIND_TYPE_TAG: type_id = t->type; break; case BTF_KIND_ARRAY: @@ -4023,6 +4024,7 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id, case BTF_KIND_PTR: case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: + case BTF_KIND_TYPE_TAG: if (cand_type->info != canon_type->info) return 0; return btf_dedup_is_equiv(d, cand_type->type, canon_type->type); From 4746158305e98c91c479539d53ef9bf8c520dd66 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Mon, 15 Nov 2021 08:39:43 -0800 Subject: [PATCH 0200/2048] selftests/bpf: Add a dedup selftest with equivalent structure types Without previous libbpf patch, the following error will occur: $ ./test_progs -t btf ... do_test_dedup:FAIL:check btf_dedup failed errno:-22#13/205 btf/dedup: btf_type_tag #5, struct:FAIL And the previous libbpf patch fixed the issue. Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211115163943.3922547-1-yhs@fb.com --- tools/testing/selftests/bpf/prog_tests/btf.c | 26 ++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/btf.c b/tools/testing/selftests/bpf/prog_tests/btf.c index 4aa6343dc4c8..f9326a13badb 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf.c +++ b/tools/testing/selftests/bpf/prog_tests/btf.c @@ -7352,6 +7352,32 @@ static struct btf_dedup_test dedup_tests[] = { BTF_STR_SEC("\0tag1"), }, }, +{ + .descr = "dedup: btf_type_tag #5, struct", + .input = { + .raw_types = { + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4), /* [3] */ + BTF_MEMBER_ENC(NAME_NTH(3), 2, BTF_MEMBER_OFFSET(0, 0)), + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [4] */ + BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4), /* [5] */ + BTF_MEMBER_ENC(NAME_NTH(3), 4, BTF_MEMBER_OFFSET(0, 0)), + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0t\0m"), + }, + .expect = { + .raw_types = { + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + BTF_TYPE_TAG_ENC(NAME_NTH(1), 1), /* [2] */ + BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4), /* [3] */ + BTF_MEMBER_ENC(NAME_NTH(3), 2, BTF_MEMBER_OFFSET(0, 0)), + BTF_END_RAW, + }, + BTF_STR_SEC("\0tag1\0t\0m"), + }, +}, }; From 4344842836e9b9a7b695dc84956cdecd83ac02e9 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Mon, 15 Nov 2021 22:58:42 +0000 Subject: [PATCH 0201/2048] bpftool: Add SPDX tags to RST documentation files Most files in the kernel repository have a SPDX tags. The files that don't have such a tag (or another license boilerplate) tend to fall under the GPL-2.0 license. In the past, bpftool's Makefile (for example) has been marked as GPL-2.0 for that reason, when in fact all bpftool is dual-licensed. To prevent a similar confusion from happening with the RST documentation files for bpftool, let's explicitly mark all files as dual-licensed. Signed-off-by: Quentin Monnet Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211115225844.33943-2-quentin@isovalent.com --- tools/bpf/bpftool/Documentation/Makefile | 2 +- tools/bpf/bpftool/Documentation/bpftool-btf.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-cgroup.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-feature.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-gen.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-iter.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-link.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-map.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-net.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-perf.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-prog.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst | 2 ++ tools/bpf/bpftool/Documentation/bpftool.rst | 2 ++ tools/bpf/bpftool/Documentation/common_options.rst | 2 ++ 14 files changed, 27 insertions(+), 1 deletion(-) diff --git a/tools/bpf/bpftool/Documentation/Makefile b/tools/bpf/bpftool/Documentation/Makefile index 692e1b947490..ac8487dcff1d 100644 --- a/tools/bpf/bpftool/Documentation/Makefile +++ b/tools/bpf/bpftool/Documentation/Makefile @@ -24,7 +24,7 @@ man: man8 man8: $(DOC_MAN8) RST2MAN_DEP := $(shell command -v rst2man 2>/dev/null) -RST2MAN_OPTS += --verbose +RST2MAN_OPTS += --verbose --strip-comments list_pages = $(sort $(basename $(filter-out $(1),$(MAN8_RST)))) see_also = $(subst " ",, \ diff --git a/tools/bpf/bpftool/Documentation/bpftool-btf.rst b/tools/bpf/bpftool/Documentation/bpftool-btf.rst index 4425d942dd39..2d2ceb7163f6 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-btf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-btf.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-btf ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index 8069d37dd991..b954faeb0f07 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-cgroup ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-feature.rst b/tools/bpf/bpftool/Documentation/bpftool-feature.rst index ab9f57ee4c3a..b1471788a15f 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-feature.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-feature.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + =============== bpftool-feature =============== diff --git a/tools/bpf/bpftool/Documentation/bpftool-gen.rst b/tools/bpf/bpftool/Documentation/bpftool-gen.rst index 2a137f8a4cea..51e2e8de5208 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-gen.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-gen.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-gen ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-iter.rst b/tools/bpf/bpftool/Documentation/bpftool-iter.rst index 471f363a725a..51914c9e8a54 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-iter.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-iter.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ============ bpftool-iter ============ diff --git a/tools/bpf/bpftool/Documentation/bpftool-link.rst b/tools/bpf/bpftool/Documentation/bpftool-link.rst index 9434349636a5..31371bcf605a 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-link.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-link.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-link ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-map.rst b/tools/bpf/bpftool/Documentation/bpftool-map.rst index 991d18fd84f2..e22c918c069c 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-map.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-map.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-map ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-net.rst b/tools/bpf/bpftool/Documentation/bpftool-net.rst index 7ec57535a7c1..6d1aa374529f 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-net.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-net.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-net ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-perf.rst b/tools/bpf/bpftool/Documentation/bpftool-perf.rst index ce52798a917d..ad554806faa2 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-perf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-perf.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-perf ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index f27265bd589b..d31148571403 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ bpftool-prog ================ diff --git a/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst b/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst index 02afc0fc14cb..77b845b5ac61 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================== bpftool-struct_ops ================== diff --git a/tools/bpf/bpftool/Documentation/bpftool.rst b/tools/bpf/bpftool/Documentation/bpftool.rst index 8ac86565c501..1248b35e67ae 100644 --- a/tools/bpf/bpftool/Documentation/bpftool.rst +++ b/tools/bpf/bpftool/Documentation/bpftool.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + ================ BPFTOOL ================ diff --git a/tools/bpf/bpftool/Documentation/common_options.rst b/tools/bpf/bpftool/Documentation/common_options.rst index 75adf23202d8..908487b9c2ad 100644 --- a/tools/bpf/bpftool/Documentation/common_options.rst +++ b/tools/bpf/bpftool/Documentation/common_options.rst @@ -1,3 +1,5 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + -h, --help Print short help message (similar to **bpftool help**). From b623181520404ef48f7421333561bd294c6c6b11 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Mon, 15 Nov 2021 22:58:43 +0000 Subject: [PATCH 0202/2048] bpftool: Update doc (use susbtitutions) and test_bpftool_synctypes.py test_bpftool_synctypes.py helps detecting inconsistencies in bpftool between the different list of types and options scattered in the sources, the documentation, and the bash completion. For options that apply to all bpftool commands, the script had a hardcoded list of values, and would use them to check whether the man pages are up-to-date. When writing the script, it felt acceptable to have this list in order to avoid to open and parse bpftool's main.h every time, and because the list of global options in bpftool doesn't change so often. However, this is prone to omissions, and we recently added a new -l|--legacy option which was described in common_options.rst, but not listed in the options summary of each manual page. The script did not complain, because it keeps comparing the hardcoded list to the (now) outdated list in the header file. To address the issue, this commit brings the following changes: - Options that are common to all bpftool commands (--json, --pretty, and --debug) are moved to a dedicated file, and used in the definition of a RST substitution. This substitution is used in the sources of all the man pages. - This list of common options is updated, with the addition of the new -l|--legacy option. - The script test_bpftool_synctypes.py is updated to compare: - Options specific to a command, found in C files, for the interactive help messages, with the same specific options from the relevant man page for that command. - Common options, checked just once: the list in main.h is compared with the new list in substitutions.rst. Signed-off-by: Quentin Monnet Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211115225844.33943-3-quentin@isovalent.com --- .../bpf/bpftool/Documentation/bpftool-btf.rst | 5 +- .../bpftool/Documentation/bpftool-cgroup.rst | 5 +- .../bpftool/Documentation/bpftool-feature.rst | 4 +- .../bpf/bpftool/Documentation/bpftool-gen.rst | 5 +- .../bpftool/Documentation/bpftool-iter.rst | 4 +- .../bpftool/Documentation/bpftool-link.rst | 5 +- .../bpf/bpftool/Documentation/bpftool-map.rst | 5 +- .../bpf/bpftool/Documentation/bpftool-net.rst | 4 +- .../bpftool/Documentation/bpftool-perf.rst | 4 +- .../bpftool/Documentation/bpftool-prog.rst | 4 +- .../Documentation/bpftool-struct_ops.rst | 4 +- tools/bpf/bpftool/Documentation/bpftool.rst | 5 +- .../bpftool/Documentation/substitutions.rst | 3 + .../selftests/bpf/test_bpftool_synctypes.py | 70 +++++++++++++++++-- 14 files changed, 102 insertions(+), 25 deletions(-) create mode 100644 tools/bpf/bpftool/Documentation/substitutions.rst diff --git a/tools/bpf/bpftool/Documentation/bpftool-btf.rst b/tools/bpf/bpftool/Documentation/bpftool-btf.rst index 2d2ceb7163f6..342716f74ec4 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-btf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-btf.rst @@ -9,13 +9,14 @@ tool for inspection of BTF data :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **btf** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | {**-d** | **--debug** } | - { **-B** | **--base-btf** } } + *OPTIONS* := { |COMMON_OPTIONS| | { **-B** | **--base-btf** } } *COMMANDS* := { **dump** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index b954faeb0f07..a17e9aa314fd 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -9,13 +9,14 @@ tool for inspection and simple manipulation of eBPF progs :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **cgroup** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } } + *OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } } *COMMANDS* := { **show** | **list** | **tree** | **attach** | **detach** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-feature.rst b/tools/bpf/bpftool/Documentation/bpftool-feature.rst index b1471788a15f..4ce9a77bc1e0 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-feature.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-feature.rst @@ -9,12 +9,14 @@ tool for inspection of eBPF-related parameters for Linux kernel or net device :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **feature** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { |COMMON_OPTIONS| } *COMMANDS* := { **probe** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-gen.rst b/tools/bpf/bpftool/Documentation/bpftool-gen.rst index 51e2e8de5208..bc276388f432 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-gen.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-gen.rst @@ -9,13 +9,14 @@ tool for BPF code-generation :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **gen** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-L** | **--use-loader** } } + *OPTIONS* := { |COMMON_OPTIONS| | { **-L** | **--use-loader** } } *COMMAND* := { **object** | **skeleton** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-iter.rst b/tools/bpf/bpftool/Documentation/bpftool-iter.rst index 51914c9e8a54..84839d488621 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-iter.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-iter.rst @@ -9,12 +9,14 @@ tool to create BPF iterators :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **iter** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { |COMMON_OPTIONS| } *COMMANDS* := { **pin** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-link.rst b/tools/bpf/bpftool/Documentation/bpftool-link.rst index 31371bcf605a..52a4eee4af54 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-link.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-link.rst @@ -9,13 +9,14 @@ tool for inspection and simple manipulation of eBPF links :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **link** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } | { **-n** | **--nomount** } } + *OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } | { **-n** | **--nomount** } } *COMMANDS* := { **show** | **list** | **pin** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-map.rst b/tools/bpf/bpftool/Documentation/bpftool-map.rst index e22c918c069c..7c188a598444 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-map.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-map.rst @@ -9,13 +9,14 @@ tool for inspection and simple manipulation of eBPF maps :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **map** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | - { **-f** | **--bpffs** } | { **-n** | **--nomount** } } + *OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } | { **-n** | **--nomount** } } *COMMANDS* := { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** | diff --git a/tools/bpf/bpftool/Documentation/bpftool-net.rst b/tools/bpf/bpftool/Documentation/bpftool-net.rst index 6d1aa374529f..f4e0a516335a 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-net.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-net.rst @@ -9,12 +9,14 @@ tool for inspection of netdev/tc related bpf prog attachments :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **net** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { |COMMON_OPTIONS| } *COMMANDS* := { **show** | **list** | **attach** | **detach** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-perf.rst b/tools/bpf/bpftool/Documentation/bpftool-perf.rst index ad554806faa2..5fea633a82f1 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-perf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-perf.rst @@ -9,12 +9,14 @@ tool for inspection of perf related bpf prog attachments :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **perf** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { |COMMON_OPTIONS| } *COMMANDS* := { **show** | **list** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index d31148571403..a2e9359e554c 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -9,12 +9,14 @@ tool for inspection and simple manipulation of eBPF progs :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **prog** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | + *OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } | { **-m** | **--mapcompat** } | { **-n** | **--nomount** } | { **-L** | **--use-loader** } } diff --git a/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst b/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst index 77b845b5ac61..ee53a122c0c7 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst @@ -9,12 +9,14 @@ tool to register/unregister/introspect BPF struct_ops :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== **bpftool** [*OPTIONS*] **struct_ops** *COMMAND* - *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { |COMMON_OPTIONS| } *COMMANDS* := { **show** | **list** | **dump** | **register** | **unregister** | **help** } diff --git a/tools/bpf/bpftool/Documentation/bpftool.rst b/tools/bpf/bpftool/Documentation/bpftool.rst index 1248b35e67ae..7084dd9fa2f8 100644 --- a/tools/bpf/bpftool/Documentation/bpftool.rst +++ b/tools/bpf/bpftool/Documentation/bpftool.rst @@ -9,6 +9,8 @@ tool for inspection and simple manipulation of eBPF programs and maps :Manual section: 8 +.. include:: substitutions.rst + SYNOPSIS ======== @@ -20,8 +22,7 @@ SYNOPSIS *OBJECT* := { **map** | **program** | **cgroup** | **perf** | **net** | **feature** } - *OPTIONS* := { { **-V** | **--version** } | - { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } } + *OPTIONS* := { { **-V** | **--version** } | |COMMON_OPTIONS| } *MAP-COMMANDS* := { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** | diff --git a/tools/bpf/bpftool/Documentation/substitutions.rst b/tools/bpf/bpftool/Documentation/substitutions.rst new file mode 100644 index 000000000000..ccf1ffa0686c --- /dev/null +++ b/tools/bpf/bpftool/Documentation/substitutions.rst @@ -0,0 +1,3 @@ +.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) + +.. |COMMON_OPTIONS| replace:: { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | { **-l** | **--legacy** } diff --git a/tools/testing/selftests/bpf/test_bpftool_synctypes.py b/tools/testing/selftests/bpf/test_bpftool_synctypes.py index be54b7335a76..3f6e562565ec 100755 --- a/tools/testing/selftests/bpf/test_bpftool_synctypes.py +++ b/tools/testing/selftests/bpf/test_bpftool_synctypes.py @@ -242,12 +242,6 @@ class FileExtractor(object): end_marker = re.compile('}\\\\n') return self.__get_description_list(start_marker, pattern, end_marker) - def default_options(self): - """ - Return the default options contained in HELP_SPEC_OPTIONS - """ - return { '-j', '--json', '-p', '--pretty', '-d', '--debug' } - def get_bashcomp_list(self, block_name): """ Search for and parse a list of type names from a variable in bash @@ -274,7 +268,56 @@ class SourceFileExtractor(FileExtractor): defined in children classes. """ def get_options(self): - return self.default_options().union(self.get_help_list_macro('HELP_SPEC_OPTIONS')) + return self.get_help_list_macro('HELP_SPEC_OPTIONS') + +class MainHeaderFileExtractor(SourceFileExtractor): + """ + An extractor for bpftool's main.h + """ + filename = os.path.join(BPFTOOL_DIR, 'main.h') + + def get_common_options(self): + """ + Parse the list of common options in main.h (options that apply to all + commands), which looks to the lists of options in other source files + but has different start and end markers: + + "OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy}" + + Return a set containing all options, such as: + + {'-p', '-d', '--legacy', '--pretty', '--debug', '--json', '-l', '-j'} + """ + start_marker = re.compile(f'"OPTIONS :=') + pattern = re.compile('([\w-]+) ?(?:\||}[ }\]"])') + end_marker = re.compile('#define') + + parser = InlineListParser(self.reader) + parser.search_block(start_marker) + return parser.parse(pattern, end_marker) + +class ManSubstitutionsExtractor(SourceFileExtractor): + """ + An extractor for substitutions.rst + """ + filename = os.path.join(BPFTOOL_DIR, 'Documentation/substitutions.rst') + + def get_common_options(self): + """ + Parse the list of common options in substitutions.rst (options that + apply to all commands). + + Return a set containing all options, such as: + + {'-p', '-d', '--legacy', '--pretty', '--debug', '--json', '-l', '-j'} + """ + start_marker = re.compile('\|COMMON_OPTIONS\| replace:: {') + pattern = re.compile('\*\*([\w/-]+)\*\*') + end_marker = re.compile('}$') + + parser = InlineListParser(self.reader) + parser.search_block(start_marker) + return parser.parse(pattern, end_marker) class ProgFileExtractor(SourceFileExtractor): """ @@ -580,6 +623,19 @@ def main(): verify(help_main_options, man_main_options, f'Comparing {source_main_info.filename} (do_help() OPTIONS) and {man_main_info.filename} (OPTIONS):') + # Compare common options (options that apply to all commands) + + main_hdr_info = MainHeaderFileExtractor() + source_common_options = main_hdr_info.get_common_options() + main_hdr_info.close() + + man_substitutions = ManSubstitutionsExtractor() + man_common_options = man_substitutions.get_common_options() + man_substitutions.close() + + verify(source_common_options, man_common_options, + f'Comparing common options from {main_hdr_info.filename} (HELP_SPEC_OPTIONS) and {man_substitutions.filename}:') + sys.exit(retval) if __name__ == "__main__": From e12cd158c8a45b3926cc2f42151384a2d7fdcec3 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Mon, 15 Nov 2021 22:58:44 +0000 Subject: [PATCH 0203/2048] selftests/bpf: Configure dir paths via env in test_bpftool_synctypes.py Script test_bpftool_synctypes.py parses a number of files in the bpftool directory (or even elsewhere in the repo) to make sure that the list of types or options in those different files are consistent. Instead of having fixed paths, let's make the directories configurable through environment variable. This should make easier in the future to run the script in a different setup, for example on an out-of-tree bpftool mirror with a different layout. Signed-off-by: Quentin Monnet Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211115225844.33943-4-quentin@isovalent.com --- .../selftests/bpf/test_bpftool_synctypes.py | 26 ++++++++++++------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/tools/testing/selftests/bpf/test_bpftool_synctypes.py b/tools/testing/selftests/bpf/test_bpftool_synctypes.py index 3f6e562565ec..6bf21e47882a 100755 --- a/tools/testing/selftests/bpf/test_bpftool_synctypes.py +++ b/tools/testing/selftests/bpf/test_bpftool_synctypes.py @@ -9,7 +9,15 @@ import os, sys LINUX_ROOT = os.path.abspath(os.path.join(__file__, os.pardir, os.pardir, os.pardir, os.pardir, os.pardir)) -BPFTOOL_DIR = os.path.join(LINUX_ROOT, 'tools/bpf/bpftool') +BPFTOOL_DIR = os.getenv('BPFTOOL_DIR', + os.path.join(LINUX_ROOT, 'tools/bpf/bpftool')) +BPFTOOL_BASHCOMP_DIR = os.getenv('BPFTOOL_BASHCOMP_DIR', + os.path.join(BPFTOOL_DIR, 'bash-completion')) +BPFTOOL_DOC_DIR = os.getenv('BPFTOOL_DOC_DIR', + os.path.join(BPFTOOL_DIR, 'Documentation')) +INCLUDE_DIR = os.getenv('INCLUDE_DIR', + os.path.join(LINUX_ROOT, 'tools/include')) + retval = 0 class BlockParser(object): @@ -300,7 +308,7 @@ class ManSubstitutionsExtractor(SourceFileExtractor): """ An extractor for substitutions.rst """ - filename = os.path.join(BPFTOOL_DIR, 'Documentation/substitutions.rst') + filename = os.path.join(BPFTOOL_DOC_DIR, 'substitutions.rst') def get_common_options(self): """ @@ -393,7 +401,7 @@ class BpfHeaderExtractor(FileExtractor): """ An extractor for the UAPI BPF header. """ - filename = os.path.join(LINUX_ROOT, 'tools/include/uapi/linux/bpf.h') + filename = os.path.join(INCLUDE_DIR, 'uapi/linux/bpf.h') def get_prog_types(self): return self.get_enum('bpf_prog_type') @@ -417,7 +425,7 @@ class ManProgExtractor(ManPageExtractor): """ An extractor for bpftool-prog.rst. """ - filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-prog.rst') + filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-prog.rst') def get_attach_types(self): return self.get_rst_list('ATTACH_TYPE') @@ -426,7 +434,7 @@ class ManMapExtractor(ManPageExtractor): """ An extractor for bpftool-map.rst. """ - filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-map.rst') + filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-map.rst') def get_map_types(self): return self.get_rst_list('TYPE') @@ -435,7 +443,7 @@ class ManCgroupExtractor(ManPageExtractor): """ An extractor for bpftool-cgroup.rst. """ - filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-cgroup.rst') + filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-cgroup.rst') def get_attach_types(self): return self.get_rst_list('ATTACH_TYPE') @@ -454,7 +462,7 @@ class BashcompExtractor(FileExtractor): """ An extractor for bpftool's bash completion file. """ - filename = os.path.join(BPFTOOL_DIR, 'bash-completion/bpftool') + filename = os.path.join(BPFTOOL_BASHCOMP_DIR, 'bpftool') def get_prog_attach_types(self): return self.get_bashcomp_list('BPFTOOL_PROG_ATTACH_TYPES') @@ -605,7 +613,7 @@ def main(): help_cmd_options = source_info.get_options() source_info.close() - man_cmd_info = ManGenericExtractor(os.path.join('Documentation', 'bpftool-' + cmd + '.rst')) + man_cmd_info = ManGenericExtractor(os.path.join(BPFTOOL_DOC_DIR, 'bpftool-' + cmd + '.rst')) man_cmd_options = man_cmd_info.get_options() man_cmd_info.close() @@ -616,7 +624,7 @@ def main(): help_main_options = source_main_info.get_options() source_main_info.close() - man_main_info = ManGenericExtractor(os.path.join('Documentation', 'bpftool.rst')) + man_main_info = ManGenericExtractor(os.path.join(BPFTOOL_DOC_DIR, 'bpftool.rst')) man_main_options = man_main_info.get_options() man_main_info.close() From dd2ac1d6d4954522c67279e42e7069b91f67f151 Mon Sep 17 00:00:00 2001 From: Zijun Hu Date: Tue, 16 Nov 2021 16:51:38 +0800 Subject: [PATCH 0204/2048] Bluetooth: hci_h4: Fix padding calculation error within h4_recv_buf() it is erroneous to calculate padding by subtracting length of type indication from skb->len, it will cause data analysis error for alignment which is greater than 1, so fixed by adding length of type indication with skb->len. Signed-off-by: Zijun Hu Signed-off-by: Marcel Holtmann --- drivers/bluetooth/hci_h4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/bluetooth/hci_h4.c b/drivers/bluetooth/hci_h4.c index 4b3b14a34794..1d0cdf023243 100644 --- a/drivers/bluetooth/hci_h4.c +++ b/drivers/bluetooth/hci_h4.c @@ -252,7 +252,7 @@ struct sk_buff *h4_recv_buf(struct hci_dev *hdev, struct sk_buff *skb, } if (!dlen) { - hu->padding = (skb->len - 1) % alignment; + hu->padding = (skb->len + 1) % alignment; hu->padding = (alignment - hu->padding) % alignment; /* No more data, complete frame */ @@ -260,7 +260,7 @@ struct sk_buff *h4_recv_buf(struct hci_dev *hdev, struct sk_buff *skb, skb = NULL; } } else { - hu->padding = (skb->len - 1) % alignment; + hu->padding = (skb->len + 1) % alignment; hu->padding = (alignment - hu->padding) % alignment; /* Complete frame */ From 3a56ef719f0b9682afb8a86d64b2399e36faa4e6 Mon Sep 17 00:00:00 2001 From: Pavel Skripkin Date: Mon, 1 Nov 2021 10:12:12 +0300 Subject: [PATCH 0205/2048] Bluetooth: stop proccessing malicious adv data Syzbot reported slab-out-of-bounds read in hci_le_adv_report_evt(). The problem was in missing validaion check. We should check if data is not malicious and we can read next data block. If we won't check ptr validness, code can read a way beyond skb->end and it can cause problems, of course. Fixes: e95beb414168 ("Bluetooth: hci_le_adv_report_evt code refactoring") Reported-and-tested-by: syzbot+e3fcb9c4f3c2a931dc40@syzkaller.appspotmail.com Signed-off-by: Pavel Skripkin Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_event.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index d4b75a6cfeee..5471fbf38873 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -5906,7 +5906,8 @@ static void hci_le_adv_report_evt(struct hci_dev *hdev, struct sk_buff *skb) struct hci_ev_le_advertising_info *ev = ptr; s8 rssi; - if (ev->length <= HCI_MAX_AD_LENGTH) { + if (ev->length <= HCI_MAX_AD_LENGTH && + ev->data + ev->length <= skb_tail_pointer(skb)) { rssi = ev->data[ev->length]; process_adv_report(hdev, ev->evt_type, &ev->bdaddr, ev->bdaddr_type, NULL, 0, rssi, @@ -5916,6 +5917,11 @@ static void hci_le_adv_report_evt(struct hci_dev *hdev, struct sk_buff *skb) } ptr += sizeof(*ev) + ev->length + 1; + + if (ptr > (void *) skb_tail_pointer(skb) - sizeof(*ev)) { + bt_dev_err(hdev, "Malicious advertising data. Stopping processing"); + break; + } } hci_dev_unlock(hdev); From a27c519a816437ec92f0ffa3adbc168c2c08725b Mon Sep 17 00:00:00 2001 From: Jackie Liu Date: Tue, 16 Nov 2021 09:17:17 +0800 Subject: [PATCH 0206/2048] Bluetooth: fix uninitialized variables notify_evt Coverity Scan report: [...] *** CID 1493985: Uninitialized variables (UNINIT) /net/bluetooth/hci_event.c: 4535 in hci_sync_conn_complete_evt() 4529 4530 /* Notify only in case of SCO over HCI transport data path which 4531 * is zero and non-zero value shall be non-HCI transport data path 4532 */ 4533 if (conn->codec.data_path == 0) { 4534 if (hdev->notify) >>> CID 1493985: Uninitialized variables (UNINIT) >>> Using uninitialized value "notify_evt" when calling "*hdev->notify". 4535 hdev->notify(hdev, notify_evt); 4536 } 4537 4538 hci_connect_cfm(conn, ev->status); 4539 if (ev->status) 4540 hci_conn_del(conn); [...] Although only btusb uses air_mode, and he only handles HCI_NOTIFY_ENABLE_SCO_CVSD and HCI_NOTIFY_ENABLE_SCO_TRANSP, there is still a very small chance that ev->air_mode is not equal to 0x2 and 0x3, but notify_evt is initialized to HCI_NOTIFY_ENABLE_SCO_CVSD or HCI_NOTIFY_ENABLE_SCO_TRANSP. the context is maybe not correct. Let us directly use the required function instead of re-initializing it, so as to restore the original logic and make the code more correct. Addresses-Coverity: ("Uninitialized variables") Fixes: f4f9fa0c07bb ("Bluetooth: Allow usb to auto-suspend when SCO use non-HCI transport") Suggested-by: Marcel Holtmann Signed-off-by: Jackie Liu Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_event.c | 23 +++++++++-------------- 1 file changed, 9 insertions(+), 14 deletions(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 5471fbf38873..9e27ff343716 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -4530,7 +4530,6 @@ static void hci_sync_conn_complete_evt(struct hci_dev *hdev, { struct hci_ev_sync_conn_complete *ev = (void *) skb->data; struct hci_conn *conn; - unsigned int notify_evt; BT_DBG("%s status 0x%2.2x", hdev->name, ev->status); @@ -4602,22 +4601,18 @@ static void hci_sync_conn_complete_evt(struct hci_dev *hdev, } bt_dev_dbg(hdev, "SCO connected with air mode: %02x", ev->air_mode); - - switch (ev->air_mode) { - case 0x02: - notify_evt = HCI_NOTIFY_ENABLE_SCO_CVSD; - break; - case 0x03: - notify_evt = HCI_NOTIFY_ENABLE_SCO_TRANSP; - break; - } - /* Notify only in case of SCO over HCI transport data path which * is zero and non-zero value shall be non-HCI transport data path */ - if (conn->codec.data_path == 0) { - if (hdev->notify) - hdev->notify(hdev, notify_evt); + if (conn->codec.data_path == 0 && hdev->notify) { + switch (ev->air_mode) { + case 0x02: + hdev->notify(hdev, HCI_NOTIFY_ENABLE_SCO_CVSD); + break; + case 0x03: + hdev->notify(hdev, HCI_NOTIFY_ENABLE_SCO_TRANSP); + break; + } } hci_connect_cfm(conn, ev->status); From 385315decf65902768a0062b4e594f59f013025e Mon Sep 17 00:00:00 2001 From: Jesse Melhuish Date: Mon, 15 Nov 2021 22:00:52 +0000 Subject: [PATCH 0207/2048] Bluetooth: Don't initialize msft/aosp when using user channel A race condition is triggered when usermode control is given to userspace before the kernel's MSFT query responds, resulting in an unexpected response to userspace's reset command. Issue can be observed in btmon: < HCI Command: Vendor (0x3f|0x001e) plen 2 #3 [hci0] 05 01 .. @ USER Open: bt_stack_manage (privileged) version 2.22 {0x0002} [hci0] < HCI Command: Reset (0x03|0x0003) plen 0 #4 [hci0] > HCI Event: Command Complete (0x0e) plen 5 #5 [hci0] Vendor (0x3f|0x001e) ncmd 1 Status: Command Disallowed (0x0c) 05 . > HCI Event: Command Complete (0x0e) plen 4 #6 [hci0] Reset (0x03|0x0003) ncmd 2 Status: Success (0x00) Reviewed-by: Abhishek Pandit-Subedi Reviewed-by: Sonny Sasaka Signed-off-by: Jesse Melhuish Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_sync.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index b794605dc882..06d08dc6ad22 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -3887,8 +3887,10 @@ setup_failed: hci_dev_test_flag(hdev, HCI_VENDOR_DIAG) && hdev->set_diag) ret = hdev->set_diag(hdev, true); - msft_do_open(hdev); - aosp_do_open(hdev); + if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL)) { + msft_do_open(hdev); + aosp_do_open(hdev); + } clear_bit(HCI_INIT, &hdev->flags); @@ -4031,8 +4033,10 @@ int hci_dev_close_sync(struct hci_dev *hdev) hci_sock_dev_event(hdev, HCI_DEV_DOWN); - aosp_do_close(hdev); - msft_do_close(hdev); + if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL)) { + aosp_do_close(hdev); + msft_do_close(hdev); + } if (hdev->flush) hdev->flush(hdev); From ebf7f6f0a6cdcc17a3da52b81e4b3a98c4005028 Mon Sep 17 00:00:00 2001 From: Tiezhu Yang Date: Fri, 5 Nov 2021 09:30:00 +0800 Subject: [PATCH 0208/2048] bpf: Change value of MAX_TAIL_CALL_CNT from 32 to 33 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the current code, the actual max tail call count is 33 which is greater than MAX_TAIL_CALL_CNT (defined as 32). The actual limit is not consistent with the meaning of MAX_TAIL_CALL_CNT and thus confusing at first glance. We can see the historical evolution from commit 04fd61ab36ec ("bpf: allow bpf programs to tail-call other bpf programs") and commit f9dabe016b63 ("bpf: Undo off-by-one in interpreter tail call count limit"). In order to avoid changing existing behavior, the actual limit is 33 now, this is reasonable. After commit 874be05f525e ("bpf, tests: Add tail call test suite"), we can see there exists failed testcase. On all archs when CONFIG_BPF_JIT_ALWAYS_ON is not set: # echo 0 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf # dmesg | grep -w FAIL Tail call error path, max count reached jited:0 ret 34 != 33 FAIL On some archs: # echo 1 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf # dmesg | grep -w FAIL Tail call error path, max count reached jited:1 ret 34 != 33 FAIL Although the above failed testcase has been fixed in commit 18935a72eb25 ("bpf/tests: Fix error in tail call limit tests"), it would still be good to change the value of MAX_TAIL_CALL_CNT from 32 to 33 to make the code more readable. The 32-bit x86 JIT was using a limit of 32, just fix the wrong comments and limit to 33 tail calls as the constant MAX_TAIL_CALL_CNT updated. For the mips64 JIT, use "ori" instead of "addiu" as suggested by Johan Almbladh. For the riscv JIT, use RV_REG_TCC directly to save one register move as suggested by Björn Töpel. For the other implementations, no function changes, it does not change the current limit 33, the new value of MAX_TAIL_CALL_CNT can reflect the actual max tail call count, the related tail call testcases in test_bpf module and selftests can work well for the interpreter and the JIT. Here are the test results on x86_64: # uname -m x86_64 # echo 0 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf test_suite=test_tail_calls # dmesg | tail -1 test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [0/8 JIT'ed] # rmmod test_bpf # echo 1 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf test_suite=test_tail_calls # dmesg | tail -1 test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [8/8 JIT'ed] # rmmod test_bpf # ./test_progs -t tailcalls #142 tailcalls:OK Summary: 1/11 PASSED, 0 SKIPPED, 0 FAILED Signed-off-by: Tiezhu Yang Signed-off-by: Daniel Borkmann Tested-by: Johan Almbladh Tested-by: Ilya Leoshkevich Acked-by: Björn Töpel Acked-by: Johan Almbladh Acked-by: Ilya Leoshkevich Link: https://lore.kernel.org/bpf/1636075800-3264-1-git-send-email-yangtiezhu@loongson.cn --- arch/arm/net/bpf_jit_32.c | 5 +++-- arch/arm64/net/bpf_jit_comp.c | 5 +++-- arch/mips/net/bpf_jit_comp32.c | 3 +-- arch/mips/net/bpf_jit_comp64.c | 2 +- arch/powerpc/net/bpf_jit_comp32.c | 4 ++-- arch/powerpc/net/bpf_jit_comp64.c | 4 ++-- arch/riscv/net/bpf_jit_comp32.c | 6 ++---- arch/riscv/net/bpf_jit_comp64.c | 7 +++---- arch/s390/net/bpf_jit_comp.c | 6 +++--- arch/sparc/net/bpf_jit_comp_64.c | 2 +- arch/x86/net/bpf_jit_comp.c | 10 +++++----- arch/x86/net/bpf_jit_comp32.c | 4 ++-- include/linux/bpf.h | 2 +- include/uapi/linux/bpf.h | 2 +- kernel/bpf/core.c | 3 ++- lib/test_bpf.c | 4 ++-- tools/include/uapi/linux/bpf.h | 2 +- 17 files changed, 35 insertions(+), 36 deletions(-) diff --git a/arch/arm/net/bpf_jit_32.c b/arch/arm/net/bpf_jit_32.c index eeb6dc0ecf46..e59b41e9ab0c 100644 --- a/arch/arm/net/bpf_jit_32.c +++ b/arch/arm/net/bpf_jit_32.c @@ -1199,7 +1199,8 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) /* tmp2[0] = array, tmp2[1] = index */ - /* if (tail_call_cnt > MAX_TAIL_CALL_CNT) + /* + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; * tail_call_cnt++; */ @@ -1208,7 +1209,7 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) tc = arm_bpf_get_reg64(tcc, tmp, ctx); emit(ARM_CMP_I(tc[0], hi), ctx); _emit(ARM_COND_EQ, ARM_CMP_I(tc[1], lo), ctx); - _emit(ARM_COND_HI, ARM_B(jmp_offset), ctx); + _emit(ARM_COND_CS, ARM_B(jmp_offset), ctx); emit(ARM_ADDS_I(tc[1], tc[1], 1), ctx); emit(ARM_ADC_I(tc[0], tc[0], 0), ctx); arm_bpf_put_reg64(tcc, tmp, ctx); diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index 86c9dc0681cc..07c12c42b751 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -287,13 +287,14 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) emit(A64_CMP(0, r3, tmp), ctx); emit(A64_B_(A64_COND_CS, jmp_offset), ctx); - /* if (tail_call_cnt > MAX_TAIL_CALL_CNT) + /* + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; * tail_call_cnt++; */ emit_a64_mov_i64(tmp, MAX_TAIL_CALL_CNT, ctx); emit(A64_CMP(1, tcc, tmp), ctx); - emit(A64_B_(A64_COND_HI, jmp_offset), ctx); + emit(A64_B_(A64_COND_CS, jmp_offset), ctx); emit(A64_ADD_I(1, tcc, tcc, 1), ctx); /* prog = array->ptrs[index]; diff --git a/arch/mips/net/bpf_jit_comp32.c b/arch/mips/net/bpf_jit_comp32.c index bd996ede12f8..044b11b65bca 100644 --- a/arch/mips/net/bpf_jit_comp32.c +++ b/arch/mips/net/bpf_jit_comp32.c @@ -1381,8 +1381,7 @@ void build_prologue(struct jit_context *ctx) * 16-byte area in the parent's stack frame. On a tail call, the * calling function jumps into the prologue after these instructions. */ - emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO, - min(MAX_TAIL_CALL_CNT + 1, 0xffff)); + emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff)); emit(ctx, sw, MIPS_R_T9, 0, MIPS_R_SP); /* diff --git a/arch/mips/net/bpf_jit_comp64.c b/arch/mips/net/bpf_jit_comp64.c index 815ade724227..6475828ffb36 100644 --- a/arch/mips/net/bpf_jit_comp64.c +++ b/arch/mips/net/bpf_jit_comp64.c @@ -552,7 +552,7 @@ void build_prologue(struct jit_context *ctx) * On a tail call, the calling function jumps into the prologue * after this instruction. */ - emit(ctx, addiu, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT + 1, 0xffff)); + emit(ctx, ori, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff)); /* === Entry-point for tail calls === */ diff --git a/arch/powerpc/net/bpf_jit_comp32.c b/arch/powerpc/net/bpf_jit_comp32.c index 0da31d41d413..8a4faa05f9e4 100644 --- a/arch/powerpc/net/bpf_jit_comp32.c +++ b/arch/powerpc/net/bpf_jit_comp32.c @@ -221,13 +221,13 @@ static int bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 o PPC_BCC(COND_GE, out); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT(PPC_RAW_CMPLWI(_R0, MAX_TAIL_CALL_CNT)); /* tail_call_cnt++; */ EMIT(PPC_RAW_ADDIC(_R0, _R0, 1)); - PPC_BCC(COND_GT, out); + PPC_BCC(COND_GE, out); /* prog = array->ptrs[index]; */ EMIT(PPC_RAW_RLWINM(_R3, b2p_index, 2, 0, 29)); diff --git a/arch/powerpc/net/bpf_jit_comp64.c b/arch/powerpc/net/bpf_jit_comp64.c index 8b5157ccfeba..8571aafcc9e1 100644 --- a/arch/powerpc/net/bpf_jit_comp64.c +++ b/arch/powerpc/net/bpf_jit_comp64.c @@ -228,12 +228,12 @@ static int bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 o PPC_BCC(COND_GE, out); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; */ PPC_BPF_LL(b2p[TMP_REG_1], 1, bpf_jit_stack_tailcallcnt(ctx)); EMIT(PPC_RAW_CMPLWI(b2p[TMP_REG_1], MAX_TAIL_CALL_CNT)); - PPC_BCC(COND_GT, out); + PPC_BCC(COND_GE, out); /* * tail_call_cnt++; diff --git a/arch/riscv/net/bpf_jit_comp32.c b/arch/riscv/net/bpf_jit_comp32.c index e6497424cbf6..529a83b85c1c 100644 --- a/arch/riscv/net/bpf_jit_comp32.c +++ b/arch/riscv/net/bpf_jit_comp32.c @@ -799,11 +799,10 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) emit_bcc(BPF_JGE, lo(idx_reg), RV_REG_T1, off, ctx); /* - * temp_tcc = tcc - 1; - * if (tcc < 0) + * if (--tcc < 0) * goto out; */ - emit(rv_addi(RV_REG_T1, RV_REG_TCC, -1), ctx); + emit(rv_addi(RV_REG_TCC, RV_REG_TCC, -1), ctx); off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); emit_bcc(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx); @@ -829,7 +828,6 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) if (is_12b_check(off, insn)) return -1; emit(rv_lw(RV_REG_T0, off, RV_REG_T0), ctx); - emit(rv_addi(RV_REG_TCC, RV_REG_T1, 0), ctx); /* Epilogue jumps to *(t0 + 4). */ __build_epilogue(true, ctx); return 0; diff --git a/arch/riscv/net/bpf_jit_comp64.c b/arch/riscv/net/bpf_jit_comp64.c index f2a779c7e225..603630b6f3c5 100644 --- a/arch/riscv/net/bpf_jit_comp64.c +++ b/arch/riscv/net/bpf_jit_comp64.c @@ -327,12 +327,12 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); emit_branch(BPF_JGE, RV_REG_A2, RV_REG_T1, off, ctx); - /* if (TCC-- < 0) + /* if (--TCC < 0) * goto out; */ - emit_addi(RV_REG_T1, tcc, -1, ctx); + emit_addi(RV_REG_TCC, tcc, -1, ctx); off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); - emit_branch(BPF_JSLT, tcc, RV_REG_ZERO, off, ctx); + emit_branch(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx); /* prog = array->ptrs[index]; * if (!prog) @@ -352,7 +352,6 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) if (is_12b_check(off, insn)) return -1; emit_ld(RV_REG_T3, off, RV_REG_T2, ctx); - emit_mv(RV_REG_TCC, RV_REG_T1, ctx); __build_epilogue(true, ctx); return 0; } diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c index 233cc9bcd652..9ff2bd83aad7 100644 --- a/arch/s390/net/bpf_jit_comp.c +++ b/arch/s390/net/bpf_jit_comp.c @@ -1369,7 +1369,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, jit->prg); /* - * if (tail_call_cnt++ > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ @@ -1381,9 +1381,9 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, EMIT4_IMM(0xa7080000, REG_W0, 1); /* laal %w1,%w0,off(%r15) */ EMIT6_DISP_LH(0xeb000000, 0x00fa, REG_W1, REG_W0, REG_15, off); - /* clij %w1,MAX_TAIL_CALL_CNT,0x2,out */ + /* clij %w1,MAX_TAIL_CALL_CNT-1,0x2,out */ patch_2_clij = jit->prg; - EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT, + EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT - 1, 2, jit->prg); /* diff --git a/arch/sparc/net/bpf_jit_comp_64.c b/arch/sparc/net/bpf_jit_comp_64.c index 9a2f20cbd48b..0bfe1c72a0c9 100644 --- a/arch/sparc/net/bpf_jit_comp_64.c +++ b/arch/sparc/net/bpf_jit_comp_64.c @@ -867,7 +867,7 @@ static void emit_tail_call(struct jit_ctx *ctx) emit(LD32 | IMMED | RS1(SP) | S13(off) | RD(tmp), ctx); emit_cmpi(tmp, MAX_TAIL_CALL_CNT, ctx); #define OFFSET2 13 - emit_branch(BGU, ctx->idx, ctx->idx + OFFSET2, ctx); + emit_branch(BGEU, ctx->idx, ctx->idx + OFFSET2, ctx); emit_nop(ctx); emit_alu_K(ADD, tmp, 1, ctx); diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 726700fabca6..631847907786 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -412,7 +412,7 @@ static void emit_indirect_jump(u8 **pprog, int reg, u8 *ip) * ... bpf_tail_call(void *ctx, struct bpf_array *array, u64 index) ... * if (index >= array->map.max_entries) * goto out; - * if (++tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; * prog = array->ptrs[index]; * if (prog == NULL) @@ -446,14 +446,14 @@ static void emit_bpf_tail_call_indirect(u8 **pprog, bool *callee_regs_used, EMIT2(X86_JBE, offset); /* jbe out */ /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT2_off32(0x8B, 0x85, tcc_off); /* mov eax, dword ptr [rbp - tcc_off] */ EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT); /* cmp eax, MAX_TAIL_CALL_CNT */ offset = ctx->tail_call_indirect_label - (prog + 2 - start); - EMIT2(X86_JA, offset); /* ja out */ + EMIT2(X86_JAE, offset); /* jae out */ EMIT3(0x83, 0xC0, 0x01); /* add eax, 1 */ EMIT2_off32(0x89, 0x85, tcc_off); /* mov dword ptr [rbp - tcc_off], eax */ @@ -504,14 +504,14 @@ static void emit_bpf_tail_call_direct(struct bpf_jit_poke_descriptor *poke, int offset; /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT2_off32(0x8B, 0x85, tcc_off); /* mov eax, dword ptr [rbp - tcc_off] */ EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT); /* cmp eax, MAX_TAIL_CALL_CNT */ offset = ctx->tail_call_direct_label - (prog + 2 - start); - EMIT2(X86_JA, offset); /* ja out */ + EMIT2(X86_JAE, offset); /* jae out */ EMIT3(0x83, 0xC0, 0x01); /* add eax, 1 */ EMIT2_off32(0x89, 0x85, tcc_off); /* mov dword ptr [rbp - tcc_off], eax */ diff --git a/arch/x86/net/bpf_jit_comp32.c b/arch/x86/net/bpf_jit_comp32.c index da9b7cfa4632..429a89c5468b 100644 --- a/arch/x86/net/bpf_jit_comp32.c +++ b/arch/x86/net/bpf_jit_comp32.c @@ -1323,7 +1323,7 @@ static void emit_bpf_tail_call(u8 **pprog, u8 *ip) EMIT2(IA32_JBE, jmp_label(jmp_label1, 2)); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ lo = (u32)MAX_TAIL_CALL_CNT; @@ -1337,7 +1337,7 @@ static void emit_bpf_tail_call(u8 **pprog, u8 *ip) /* cmp ecx,lo */ EMIT3(0x83, add_1reg(0xF8, IA32_ECX), lo); - /* ja out */ + /* jae out */ EMIT2(IA32_JAE, jmp_label(jmp_label1, 2)); /* add eax,0x1 */ diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 56098c866704..cc7a0c36e7df 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -1081,7 +1081,7 @@ struct bpf_array { }; #define BPF_COMPLEXITY_LIMIT_INSNS 1000000 /* yes. 1M insns */ -#define MAX_TAIL_CALL_CNT 32 +#define MAX_TAIL_CALL_CNT 33 #define BPF_F_ACCESS_MASK (BPF_F_RDONLY | \ BPF_F_RDONLY_PROG | \ diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 6297eafdc40f..a69e4b04ffeb 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -1744,7 +1744,7 @@ union bpf_attr { * if the maximum number of tail calls has been reached for this * chain of programs. This limit is defined in the kernel by the * macro **MAX_TAIL_CALL_CNT** (not accessible to user space), - * which is currently set to 32. + * which is currently set to 33. * Return * 0 on success, or a negative error in case of failure. * diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 2405e39d800f..b52dc845ecea 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1574,7 +1574,8 @@ select_insn: if (unlikely(index >= array->map.max_entries)) goto out; - if (unlikely(tail_call_cnt > MAX_TAIL_CALL_CNT)) + + if (unlikely(tail_call_cnt >= MAX_TAIL_CALL_CNT)) goto out; tail_call_cnt++; diff --git a/lib/test_bpf.c b/lib/test_bpf.c index adae39567264..0c5cb2d6436a 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -14683,7 +14683,7 @@ static struct tail_call_test tail_call_tests[] = { BPF_EXIT_INSN(), }, .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, - .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, + .result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS, }, { "Tail call count preserved across function calls", @@ -14705,7 +14705,7 @@ static struct tail_call_test tail_call_tests[] = { }, .stack_depth = 8, .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, - .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, + .result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS, }, { "Tail call error path, NULL target", diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 6297eafdc40f..a69e4b04ffeb 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -1744,7 +1744,7 @@ union bpf_attr { * if the maximum number of tail calls has been reached for this * chain of programs. This limit is defined in the kernel by the * macro **MAX_TAIL_CALL_CNT** (not accessible to user space), - * which is currently set to 32. + * which is currently set to 33. * Return * 0 on success, or a negative error in case of failure. * From d519f350967a60b85a574ad8aeac43f2b4384746 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:30 -0800 Subject: [PATCH 0209/2048] tcp: minor optimization in tcp_add_backlog() If packet is going to be coalesced, sk_sndbuf/sk_rcvbuf values are not used. Defer their access to the point we need them. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 13d868c43284..82a9e1b75405 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1800,8 +1800,7 @@ int tcp_v4_early_demux(struct sk_buff *skb) bool tcp_add_backlog(struct sock *sk, struct sk_buff *skb) { - u32 limit = READ_ONCE(sk->sk_rcvbuf) + READ_ONCE(sk->sk_sndbuf); - u32 tail_gso_size, tail_gso_segs; + u32 limit, tail_gso_size, tail_gso_segs; struct skb_shared_info *shinfo; const struct tcphdr *th; struct tcphdr *thtail; @@ -1909,7 +1908,7 @@ no_coalesce: * to reduce memory overhead, so add a little headroom here. * Few sockets backlog are possibly concurrently non empty. */ - limit += 64*1024; + limit = READ_ONCE(sk->sk_rcvbuf) + READ_ONCE(sk->sk_sndbuf) + 64*1024; if (unlikely(sk_add_backlog(sk, skb, limit))) { bh_unlock_sock(sk); From 283c6b54bca13313a4f437719f600a3ad2135847 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:31 -0800 Subject: [PATCH 0210/2048] tcp: remove dead code in __tcp_v6_send_check() For some reason, I forgot to change __tcp_v6_send_check() at the same time I removed (ip_summed == CHECKSUM_PARTIAL) check in __tcp_v4_send_check() Fixes: 98be9b12096f ("tcp: remove dead code after CHECKSUM_PARTIAL adoption") Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/ip6_checksum.h | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/include/net/ip6_checksum.h b/include/net/ip6_checksum.h index b3f4eaa88672..ea681910b7a3 100644 --- a/include/net/ip6_checksum.h +++ b/include/net/ip6_checksum.h @@ -65,15 +65,9 @@ static inline void __tcp_v6_send_check(struct sk_buff *skb, { struct tcphdr *th = tcp_hdr(skb); - if (skb->ip_summed == CHECKSUM_PARTIAL) { - th->check = ~tcp_v6_check(skb->len, saddr, daddr, 0); - skb->csum_start = skb_transport_header(skb) - skb->head; - skb->csum_offset = offsetof(struct tcphdr, check); - } else { - th->check = tcp_v6_check(skb->len, saddr, daddr, - csum_partial(th, th->doff << 2, - skb->csum)); - } + th->check = ~tcp_v6_check(skb->len, saddr, daddr, 0); + skb->csum_start = skb_transport_header(skb) - skb->head; + skb->csum_offset = offsetof(struct tcphdr, check); } static inline void tcp_v6_gso_csum_prep(struct sk_buff *skb) From 373544020024668ea552a7699c9c9f100b6bc9d9 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:32 -0800 Subject: [PATCH 0211/2048] tcp: small optimization in tcp_v6_send_check() For TCP flows, inet6_sk(sk)->saddr has the same value than sk->sk_v6_rcv_saddr. Using sk->sk_v6_rcv_saddr increases data locality. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/tcp_ipv6.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 551fce49841d..1f1a89f096de 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1893,9 +1893,7 @@ static struct timewait_sock_ops tcp6_timewait_sock_ops = { INDIRECT_CALLABLE_SCOPE void tcp_v6_send_check(struct sock *sk, struct sk_buff *skb) { - struct ipv6_pinfo *np = inet6_sk(sk); - - __tcp_v6_send_check(skb, &np->saddr, &sk->sk_v6_daddr); + __tcp_v6_send_check(skb, &sk->sk_v6_rcv_saddr, &sk->sk_v6_daddr); } const struct inet_connection_sock_af_ops ipv6_specific = { From 42f67eea3ba36cef2dce2e853de6ddcb2e89eb39 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:33 -0800 Subject: [PATCH 0212/2048] net: use sk_is_tcp() in more places Move sk_is_tcp() to include/net/sock.h and use it where we can. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skmsg.h | 6 ------ include/net/sock.h | 5 +++++ net/core/skbuff.c | 6 ++---- net/core/sock.c | 6 ++---- 4 files changed, 9 insertions(+), 14 deletions(-) diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h index 584d94be9c8b..18a717fe62eb 100644 --- a/include/linux/skmsg.h +++ b/include/linux/skmsg.h @@ -507,12 +507,6 @@ static inline bool sk_psock_strp_enabled(struct sk_psock *psock) return !!psock->saved_data_ready; } -static inline bool sk_is_tcp(const struct sock *sk) -{ - return sk->sk_type == SOCK_STREAM && - sk->sk_protocol == IPPROTO_TCP; -} - static inline bool sk_is_udp(const struct sock *sk) { return sk->sk_type == SOCK_DGRAM && diff --git a/include/net/sock.h b/include/net/sock.h index b32906e1ab55..5bdeffdea5ec 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2638,6 +2638,11 @@ static inline void skb_setup_tx_timestamp(struct sk_buff *skb, __u16 tsflags) &skb_shinfo(skb)->tskey); } +static inline bool sk_is_tcp(const struct sock *sk) +{ + return sk->sk_type == SOCK_STREAM && sk->sk_protocol == IPPROTO_TCP; +} + /** * sk_eat_skb - Release a skb if it is no longer needed * @sk: socket to eat this skb from diff --git a/net/core/skbuff.c b/net/core/skbuff.c index ba2f38246f07..d57796f38a0b 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -4849,8 +4849,7 @@ static void __skb_complete_tx_timestamp(struct sk_buff *skb, serr->header.h4.iif = skb->dev ? skb->dev->ifindex : 0; if (sk->sk_tsflags & SOF_TIMESTAMPING_OPT_ID) { serr->ee.ee_data = skb_shinfo(skb)->tskey; - if (sk->sk_protocol == IPPROTO_TCP && - sk->sk_type == SOCK_STREAM) + if (sk_is_tcp(sk)) serr->ee.ee_data -= sk->sk_tskey; } @@ -4919,8 +4918,7 @@ void __skb_tstamp_tx(struct sk_buff *orig_skb, if (tsonly) { #ifdef CONFIG_INET if ((sk->sk_tsflags & SOF_TIMESTAMPING_OPT_STATS) && - sk->sk_protocol == IPPROTO_TCP && - sk->sk_type == SOCK_STREAM) { + sk_is_tcp(sk)) { skb = tcp_get_timestamping_opt_stats(sk, orig_skb, ack_skb); opt_stats = true; diff --git a/net/core/sock.c b/net/core/sock.c index 8f2b2f2c0e7b..0be8e43f44b9 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -872,8 +872,7 @@ int sock_set_timestamping(struct sock *sk, int optname, if (val & SOF_TIMESTAMPING_OPT_ID && !(sk->sk_tsflags & SOF_TIMESTAMPING_OPT_ID)) { - if (sk->sk_protocol == IPPROTO_TCP && - sk->sk_type == SOCK_STREAM) { + if (sk_is_tcp(sk)) { if ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)) return -EINVAL; @@ -1370,8 +1369,7 @@ set_sndbuf: case SO_ZEROCOPY: if (sk->sk_family == PF_INET || sk->sk_family == PF_INET6) { - if (!((sk->sk_type == SOCK_STREAM && - sk->sk_protocol == IPPROTO_TCP) || + if (!(sk_is_tcp(sk) || (sk->sk_type == SOCK_DGRAM && sk->sk_protocol == IPPROTO_UDP))) ret = -ENOTSUPP; From d0d598ca86bd9e595f16a39097707c90841afe80 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:34 -0800 Subject: [PATCH 0213/2048] net: remove sk_route_forced_caps We were only using one bit, and we can replace it by sk_is_tcp() Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 3 --- net/core/sock.c | 4 +++- net/ipv4/tcp.c | 1 - 3 files changed, 3 insertions(+), 5 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 5bdeffdea5ec..ebad629dd9ed 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -285,8 +285,6 @@ struct bpf_local_storage; * @sk_no_check_rx: allow zero checksum in RX packets * @sk_route_caps: route capabilities (e.g. %NETIF_F_TSO) * @sk_route_nocaps: forbidden route capabilities (e.g NETIF_F_GSO_MASK) - * @sk_route_forced_caps: static, forced route capabilities - * (set in tcp_init_sock()) * @sk_gso_type: GSO type (e.g. %SKB_GSO_TCPV4) * @sk_gso_max_size: Maximum GSO segment size to build * @sk_gso_max_segs: Maximum number of GSO segments @@ -461,7 +459,6 @@ struct sock { struct page_frag sk_frag; netdev_features_t sk_route_caps; netdev_features_t sk_route_nocaps; - netdev_features_t sk_route_forced_caps; int sk_gso_type; unsigned int sk_gso_max_size; gfp_t sk_allocation; diff --git a/net/core/sock.c b/net/core/sock.c index 0be8e43f44b9..257b5fa60480 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2244,7 +2244,9 @@ void sk_setup_caps(struct sock *sk, struct dst_entry *dst) u32 max_segs = 1; sk_dst_set(sk, dst); - sk->sk_route_caps = dst->dev->features | sk->sk_route_forced_caps; + sk->sk_route_caps = dst->dev->features; + if (sk_is_tcp(sk)) + sk->sk_route_caps |= NETIF_F_GSO; if (sk->sk_route_caps & NETIF_F_GSO) sk->sk_route_caps |= NETIF_F_GSO_SOFTWARE; sk->sk_route_caps &= ~sk->sk_route_nocaps; diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index b7796b4cf0a0..4fa4b29260bd 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -456,7 +456,6 @@ void tcp_init_sock(struct sock *sk) WRITE_ONCE(sk->sk_rcvbuf, sock_net(sk)->ipv4.sysctl_tcp_rmem[1]); sk_sockets_allocated_inc(sk); - sk->sk_route_forced_caps = NETIF_F_GSO; } EXPORT_SYMBOL(tcp_init_sock); From aba546565b613e74b84b8261999ea82b5561d3f1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:35 -0800 Subject: [PATCH 0214/2048] net: remove sk_route_nocaps Instead of using a full netdev_features_t, we can use a single bit, as sk_route_nocaps is only used to remove NETIF_F_GSO_MASK from sk->sk_route_cap. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 11 +++++------ net/core/sock.c | 3 ++- net/ipv4/tcp_ipv4.c | 4 ++-- net/ipv4/tcp_output.c | 2 +- net/ipv6/ip6_output.c | 2 +- 5 files changed, 11 insertions(+), 11 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index ebad629dd9ed..985ddcd33504 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -284,7 +284,7 @@ struct bpf_local_storage; * @sk_no_check_tx: %SO_NO_CHECK setting, set checksum in TX packets * @sk_no_check_rx: allow zero checksum in RX packets * @sk_route_caps: route capabilities (e.g. %NETIF_F_TSO) - * @sk_route_nocaps: forbidden route capabilities (e.g NETIF_F_GSO_MASK) + * @sk_gso_disabled: if set, NETIF_F_GSO_MASK is forbidden. * @sk_gso_type: GSO type (e.g. %SKB_GSO_TCPV4) * @sk_gso_max_size: Maximum GSO segment size to build * @sk_gso_max_segs: Maximum number of GSO segments @@ -458,7 +458,6 @@ struct sock { unsigned long sk_max_pacing_rate; struct page_frag sk_frag; netdev_features_t sk_route_caps; - netdev_features_t sk_route_nocaps; int sk_gso_type; unsigned int sk_gso_max_size; gfp_t sk_allocation; @@ -468,7 +467,7 @@ struct sock { * Because of non atomicity rules, all * changes are protected by socket lock. */ - u8 sk_padding : 1, + u8 sk_gso_disabled : 1, sk_kern_sock : 1, sk_no_check_tx : 1, sk_no_check_rx : 1, @@ -2121,10 +2120,10 @@ static inline bool sk_can_gso(const struct sock *sk) void sk_setup_caps(struct sock *sk, struct dst_entry *dst); -static inline void sk_nocaps_add(struct sock *sk, netdev_features_t flags) +static inline void sk_gso_disable(struct sock *sk) { - sk->sk_route_nocaps |= flags; - sk->sk_route_caps &= ~flags; + sk->sk_gso_disabled = 1; + sk->sk_route_caps &= ~NETIF_F_GSO_MASK; } static inline int skb_do_copy_data_nocache(struct sock *sk, struct sk_buff *skb, diff --git a/net/core/sock.c b/net/core/sock.c index 257b5fa60480..99738e14224c 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2249,7 +2249,8 @@ void sk_setup_caps(struct sock *sk, struct dst_entry *dst) sk->sk_route_caps |= NETIF_F_GSO; if (sk->sk_route_caps & NETIF_F_GSO) sk->sk_route_caps |= NETIF_F_GSO_SOFTWARE; - sk->sk_route_caps &= ~sk->sk_route_nocaps; + if (unlikely(sk->sk_gso_disabled)) + sk->sk_route_caps &= ~NETIF_F_GSO_MASK; if (sk_can_gso(sk)) { if (dst->header_len && !xfrm_dst_offload_ok(dst)) { sk->sk_route_caps &= ~NETIF_F_GSO_MASK; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 82a9e1b75405..5ad81bfb27b2 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1182,7 +1182,7 @@ int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, if (!md5sig) return -ENOMEM; - sk_nocaps_add(sk, NETIF_F_GSO_MASK); + sk_gso_disable(sk); INIT_HLIST_HEAD(&md5sig->head); rcu_assign_pointer(tp->md5sig_info, md5sig); } @@ -1620,7 +1620,7 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb, */ tcp_md5_do_add(newsk, addr, AF_INET, 32, l3index, key->flags, key->key, key->keylen, GFP_ATOMIC); - sk_nocaps_add(newsk, NETIF_F_GSO_MASK); + sk_gso_disable(newsk); } #endif diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 2e6e5a70168e..5079832af5c1 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -1359,7 +1359,7 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, #ifdef CONFIG_TCP_MD5SIG /* Calculate the MD5 hash, as we have all we need now */ if (md5) { - sk_nocaps_add(sk, NETIF_F_GSO_MASK); + sk_gso_disable(sk); tp->af_specific->calc_md5_hash(opts.hash_location, md5, sk, skb); } diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 2f044a49afa8..007e433d4d4d 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -977,7 +977,7 @@ slow_path: fail_toobig: if (skb->sk && dst_allfrag(skb_dst(skb))) - sk_nocaps_add(skb->sk, NETIF_F_GSO_MASK); + sk_gso_disable(skb->sk); icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu); err = -EMSGSIZE; From 1b31debca83284486cd736757b5f26d51719ef80 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:36 -0800 Subject: [PATCH 0215/2048] ipv6: shrink struct ipcm6_cookie gso_size can be moved after tclass, to use an existing hole. (8 bytes saved on 64bit arches) Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/ipv6.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/net/ipv6.h b/include/net/ipv6.h index c19bf51ded1d..53ac7707ca70 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -345,9 +345,9 @@ struct ipcm6_cookie { struct sockcm_cookie sockc; __s16 hlimit; __s16 tclass; + __u16 gso_size; __s8 dontfrag; struct ipv6_txoptions *opt; - __u16 gso_size; }; static inline void ipcm6_init(struct ipcm6_cookie *ipc6) From 1ace2b4d2b4e1db8fc62d872ab54ab48f6215ecd Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:37 -0800 Subject: [PATCH 0216/2048] net: shrink struct sock by 8 bytes Move sk_bind_phc next to sk_peer_lock to fill a hole. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/net/sock.h b/include/net/sock.h index 985ddcd33504..2333ab081789 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -489,6 +489,7 @@ struct sock { u16 sk_busy_poll_budget; #endif spinlock_t sk_peer_lock; + int sk_bind_phc; struct pid *sk_peer_pid; const struct cred *sk_peer_cred; @@ -498,7 +499,6 @@ struct sock { seqlock_t sk_stamp_seq; #endif u16 sk_tsflags; - int sk_bind_phc; u8 sk_shutdown; u32 sk_tskey; atomic_t sk_zckey; From 6c302e799a0d4be1362f505453b714fe05d91f2a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:38 -0800 Subject: [PATCH 0217/2048] net: forward_alloc_get depends on CONFIG_MPTCP (struct proto)->sk_forward_alloc is currently only used by MPTCP. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 2333ab081789..cb97c448472a 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1206,7 +1206,9 @@ struct proto { unsigned int inuse_idx; #endif +#if IS_ENABLED(CONFIG_MPTCP) int (*forward_alloc_get)(const struct sock *sk); +#endif bool (*stream_memory_free)(const struct sock *sk, int wake); bool (*sock_is_readable)(struct sock *sk); @@ -1295,10 +1297,11 @@ INDIRECT_CALLABLE_DECLARE(bool tcp_stream_memory_free(const struct sock *sk, int static inline int sk_forward_alloc_get(const struct sock *sk) { - if (!sk->sk_prot->forward_alloc_get) - return sk->sk_forward_alloc; - - return sk->sk_prot->forward_alloc_get(sk); +#if IS_ENABLED(CONFIG_MPTCP) + if (sk->sk_prot->forward_alloc_get) + return sk->sk_prot->forward_alloc_get(sk); +#endif + return sk->sk_forward_alloc; } static inline bool __sk_stream_memory_free(const struct sock *sk, int wake) From 91b6d325635617540b6a1646ddb138bb17cbd569 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:39 -0800 Subject: [PATCH 0218/2048] net: cache align tcp_memory_allocated, tcp_sockets_allocated tcp_memory_allocated and tcp_sockets_allocated often share a common cache line, source of false sharing. Also take care of udp_memory_allocated and mptcp_sockets_allocated. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 4 ++-- net/ipv4/udp.c | 2 +- net/mptcp/protocol.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 4fa4b29260bd..862e8cb8dda5 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -292,7 +292,7 @@ EXPORT_PER_CPU_SYMBOL_GPL(tcp_orphan_count); long sysctl_tcp_mem[3] __read_mostly; EXPORT_SYMBOL(sysctl_tcp_mem); -atomic_long_t tcp_memory_allocated; /* Current allocated memory. */ +atomic_long_t tcp_memory_allocated ____cacheline_aligned_in_smp; /* Current allocated memory. */ EXPORT_SYMBOL(tcp_memory_allocated); #if IS_ENABLED(CONFIG_SMC) @@ -303,7 +303,7 @@ EXPORT_SYMBOL(tcp_have_smc); /* * Current number of TCP sockets. */ -struct percpu_counter tcp_sockets_allocated; +struct percpu_counter tcp_sockets_allocated ____cacheline_aligned_in_smp; EXPORT_SYMBOL(tcp_sockets_allocated); /* diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 0e2f1c05da28..7101e6d892d6 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -122,7 +122,7 @@ EXPORT_SYMBOL(udp_table); long sysctl_udp_mem[3] __read_mostly; EXPORT_SYMBOL(sysctl_udp_mem); -atomic_long_t udp_memory_allocated; +atomic_long_t udp_memory_allocated ____cacheline_aligned_in_smp; EXPORT_SYMBOL(udp_memory_allocated); #define MAX_UDP_PORTS 65536 diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index b7e32e316738..6db93da59843 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -48,7 +48,7 @@ enum { MPTCP_CMSG_TS = BIT(0), }; -static struct percpu_counter mptcp_sockets_allocated; +static struct percpu_counter mptcp_sockets_allocated ____cacheline_aligned_in_smp; static void __mptcp_destroy_sock(struct sock *sk); static void __mptcp_check_send_data_fin(struct sock *sk); From 93afcfd1db35882921b2521a637c78755c27b02c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:40 -0800 Subject: [PATCH 0219/2048] tcp: small optimization in tcp recvmsg() When reading large chunks of data, incoming packets might be added to the backlog from BH. tcp recvmsg() detects the backlog queue is not empty, and uses a release_sock()/lock_sock() pair to process this backlog. We now have __sk_flush_backlog() to perform this a bit faster. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 862e8cb8dda5..24d77a32c9cb 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2409,8 +2409,7 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, if (copied >= target) { /* Do not sleep, just process backlog. */ - release_sock(sk); - lock_sock(sk); + __sk_flush_backlog(sk); } else { sk_wait_data(sk, &timeo, last); } From d2489c7b6d7d5ed4b32b56703c57c47bfbfe7fa5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:41 -0800 Subject: [PATCH 0220/2048] tcp: add RETPOLINE mitigation to sk_backlog_rcv Use INDIRECT_CALL_INET() to avoid an indirect call when/if CONFIG_RETPOLINE=y Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 8 +++++++- net/core/sock.c | 5 ++++- net/ipv6/tcp_ipv6.c | 5 +++-- 3 files changed, 14 insertions(+), 4 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index cb97c448472a..2d40fe4c7718 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1018,12 +1018,18 @@ static inline __must_check int sk_add_backlog(struct sock *sk, struct sk_buff *s int __sk_backlog_rcv(struct sock *sk, struct sk_buff *skb); +INDIRECT_CALLABLE_DECLARE(int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)); +INDIRECT_CALLABLE_DECLARE(int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)); + static inline int sk_backlog_rcv(struct sock *sk, struct sk_buff *skb) { if (sk_memalloc_socks() && skb_pfmemalloc(skb)) return __sk_backlog_rcv(sk, skb); - return sk->sk_backlog_rcv(sk, skb); + return INDIRECT_CALL_INET(sk->sk_backlog_rcv, + tcp_v6_do_rcv, + tcp_v4_do_rcv, + sk, skb); } static inline void sk_incoming_cpu_update(struct sock *sk) diff --git a/net/core/sock.c b/net/core/sock.c index 99738e14224c..c57d9883f62c 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -327,7 +327,10 @@ int __sk_backlog_rcv(struct sock *sk, struct sk_buff *skb) BUG_ON(!sock_flag(sk, SOCK_MEMALLOC)); noreclaim_flag = memalloc_noreclaim_save(); - ret = sk->sk_backlog_rcv(sk, skb); + ret = INDIRECT_CALL_INET(sk->sk_backlog_rcv, + tcp_v6_do_rcv, + tcp_v4_do_rcv, + sk, skb); memalloc_noreclaim_restore(noreclaim_flag); return ret; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 1f1a89f096de..f41f14b70123 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -72,7 +72,7 @@ static void tcp_v6_send_reset(const struct sock *sk, struct sk_buff *skb); static void tcp_v6_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb, struct request_sock *req); -static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb); +INDIRECT_CALLABLE_SCOPE int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb); static const struct inet_connection_sock_af_ops ipv6_mapped; const struct inet_connection_sock_af_ops ipv6_specific; @@ -1466,7 +1466,8 @@ INDIRECT_CALLABLE_DECLARE(struct dst_entry *ipv4_dst_check(struct dst_entry *, * This is because we cannot sleep with the original spinlock * held. */ -static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) +INDIRECT_CALLABLE_SCOPE +int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) { struct ipv6_pinfo *np = tcp_inet6_sk(sk); struct sk_buff *opt_skb = NULL; From 0307a0b74b3af6ecb1c8b7f727376130b15bbf44 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:42 -0800 Subject: [PATCH 0221/2048] tcp: annotate data-races on tp->segs_in and tp->data_segs_in tcp_segs_in() can be called from BH, while socket spinlock is held but socket owned by user, eventually reading these fields from tcp_get_info() Found by code inspection, no need to backport this patch to older kernels. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/tcp.h | 8 ++++++-- net/ipv4/tcp.c | 6 ++++-- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/include/net/tcp.h b/include/net/tcp.h index 4da22b41bde6..05c81677aaf7 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2172,9 +2172,13 @@ static inline void tcp_segs_in(struct tcp_sock *tp, const struct sk_buff *skb) u16 segs_in; segs_in = max_t(u16, 1, skb_shinfo(skb)->gso_segs); - tp->segs_in += segs_in; + + /* We update these fields while other threads might + * read them from tcp_get_info() + */ + WRITE_ONCE(tp->segs_in, tp->segs_in + segs_in); if (skb->len > tcp_hdrlen(skb)) - tp->data_segs_in += segs_in; + WRITE_ONCE(tp->data_segs_in, tp->data_segs_in + segs_in); } /* diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 24d77a32c9cb..267b2b18f048 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -3769,10 +3769,12 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) tcp_get_info_chrono_stats(tp, info); info->tcpi_segs_out = tp->segs_out; - info->tcpi_segs_in = tp->segs_in; + + /* segs_in and data_segs_in can be updated from tcp_segs_in() from BH */ + info->tcpi_segs_in = READ_ONCE(tp->segs_in); + info->tcpi_data_segs_in = READ_ONCE(tp->data_segs_in); info->tcpi_min_rtt = tcp_min_rtt(tp); - info->tcpi_data_segs_in = tp->data_segs_in; info->tcpi_data_segs_out = tp->data_segs_out; info->tcpi_delivery_rate_app_limited = tp->rate_app_limited ? 1 : 0; From 7b6a893a5991f5e8a56795155ae86333b03080b7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:43 -0800 Subject: [PATCH 0222/2048] tcp: annotate races around tp->urg_data tcp_poll() and tcp_ioctl() are reading tp->urg_data without socket lock owned. Also, it is faster to first check tp->urg_data in tcp_poll(), then tp->urg_seq == tp->copied_seq, because tp->urg_seq is located in a different/cold cache line. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 17 +++++++++-------- net/ipv4/tcp_input.c | 4 ++-- 2 files changed, 11 insertions(+), 10 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 267b2b18f048..313cf648c349 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -545,10 +545,11 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait) if (state != TCP_SYN_SENT && (state != TCP_SYN_RECV || rcu_access_pointer(tp->fastopen_rsk))) { int target = sock_rcvlowat(sk, 0, INT_MAX); + u16 urg_data = READ_ONCE(tp->urg_data); - if (READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq) && - !sock_flag(sk, SOCK_URGINLINE) && - tp->urg_data) + if (urg_data && + READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq) && + !sock_flag(sk, SOCK_URGINLINE)) target++; if (tcp_stream_is_readable(sk, target)) @@ -573,7 +574,7 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait) } else mask |= EPOLLOUT | EPOLLWRNORM; - if (tp->urg_data & TCP_URG_VALID) + if (urg_data & TCP_URG_VALID) mask |= EPOLLPRI; } else if (state == TCP_SYN_SENT && inet_sk(sk)->defer_connect) { /* Active TCP fastopen socket with defer_connect @@ -607,7 +608,7 @@ int tcp_ioctl(struct sock *sk, int cmd, unsigned long arg) unlock_sock_fast(sk, slow); break; case SIOCATMARK: - answ = tp->urg_data && + answ = READ_ONCE(tp->urg_data) && READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq); break; case SIOCOUTQ: @@ -1465,7 +1466,7 @@ static int tcp_recv_urg(struct sock *sk, struct msghdr *msg, int len, int flags) char c = tp->urg_data; if (!(flags & MSG_PEEK)) - tp->urg_data = TCP_URG_READ; + WRITE_ONCE(tp->urg_data, TCP_URG_READ); /* Read urgent data. */ msg->msg_flags |= MSG_OOB; @@ -2465,7 +2466,7 @@ found_ok_skb: skip_copy: if (tp->urg_data && after(tp->copied_seq, tp->urg_seq)) { - tp->urg_data = 0; + WRITE_ONCE(tp->urg_data, 0); tcp_fast_path_check(sk); } @@ -2959,7 +2960,7 @@ int tcp_disconnect(struct sock *sk, int flags) tcp_clear_xmit_timers(sk); __skb_queue_purge(&sk->sk_receive_queue); WRITE_ONCE(tp->copied_seq, tp->rcv_nxt); - tp->urg_data = 0; + WRITE_ONCE(tp->urg_data, 0); tcp_write_queue_purge(sk); tcp_fastopen_active_disable_ofo_check(sk); skb_rbtree_purge(&tp->out_of_order_queue); diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 246ab7b5e857..5ee07a337652 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5591,7 +5591,7 @@ static void tcp_check_urg(struct sock *sk, const struct tcphdr *th) } } - tp->urg_data = TCP_URG_NOTYET; + WRITE_ONCE(tp->urg_data, TCP_URG_NOTYET); WRITE_ONCE(tp->urg_seq, ptr); /* Disable header prediction. */ @@ -5617,7 +5617,7 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, const struct tcphdr *t u8 tmp; if (skb_copy_bits(skb, ptr, &tmp, 1)) BUG(); - tp->urg_data = TCP_URG_VALID | tmp; + WRITE_ONCE(tp->urg_data, TCP_URG_VALID | tmp); if (!sock_flag(sk, SOCK_DEAD)) sk->sk_data_ready(sk); } From b96c51bd3bd826a3391cbf4b1281a1e0bf9df90a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:44 -0800 Subject: [PATCH 0223/2048] tcp: tp->urg_data is unlikely to be set Use some unlikely() hints in the fast path. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 10 +++++----- net/ipv4/tcp_input.c | 4 ++-- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 313cf648c349..9175e0d729f5 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -547,7 +547,7 @@ __poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait) int target = sock_rcvlowat(sk, 0, INT_MAX); u16 urg_data = READ_ONCE(tp->urg_data); - if (urg_data && + if (unlikely(urg_data) && READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq) && !sock_flag(sk, SOCK_URGINLINE)) target++; @@ -1633,7 +1633,7 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc, len = skb->len - offset; /* Stop reading if we hit a patch of urgent data */ - if (tp->urg_data) { + if (unlikely(tp->urg_data)) { u32 urg_offset = tp->urg_seq - seq; if (urg_offset < len) len = urg_offset; @@ -2326,7 +2326,7 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, u32 offset; /* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */ - if (tp->urg_data && tp->urg_seq == *seq) { + if (unlikely(tp->urg_data) && tp->urg_seq == *seq) { if (copied) break; if (signal_pending(current)) { @@ -2431,7 +2431,7 @@ found_ok_skb: used = len; /* Do we have urgent data here? */ - if (tp->urg_data) { + if (unlikely(tp->urg_data)) { u32 urg_offset = tp->urg_seq - *seq; if (urg_offset < used) { if (!urg_offset) { @@ -2465,7 +2465,7 @@ found_ok_skb: tcp_rcv_space_adjust(sk); skip_copy: - if (tp->urg_data && after(tp->copied_seq, tp->urg_seq)) { + if (unlikely(tp->urg_data) && after(tp->copied_seq, tp->urg_seq)) { WRITE_ONCE(tp->urg_data, 0); tcp_fast_path_check(sk); } diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 5ee07a337652..3658b9c3dd2b 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5604,11 +5604,11 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, const struct tcphdr *t struct tcp_sock *tp = tcp_sk(sk); /* Check if we get a new urgent pointer - normally not. */ - if (th->urg) + if (unlikely(th->urg)) tcp_check_urg(sk, th); /* Do we wait for any urgent data? - normally not... */ - if (tp->urg_data == TCP_URG_NOTYET) { + if (unlikely(tp->urg_data == TCP_URG_NOTYET)) { u32 ptr = tp->urg_seq - ntohl(th->seq) + (th->doff * 4) - th->syn; From 3df684c1a3d08a4f649689053a3d527b3b5fda9e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:45 -0800 Subject: [PATCH 0224/2048] tcp: avoid indirect calls to sock_rfree TCP uses sk_eat_skb() when skbs can be removed from receive queue. However, the call to skb_orphan() from __kfree_skb() incurs an indirect call so sock_rfee(), which is more expensive than a direct call, especially for CONFIG_RETPOLINE=y. Add tcp_eat_recv_skb() function to make the call before __kfree_skb(). Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 9175e0d729f5..4e7011672aa9 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1580,6 +1580,16 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied) tcp_send_ack(sk); } +static void tcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb) +{ + if (likely(skb->destructor == sock_rfree)) { + sock_rfree(skb); + skb->destructor = NULL; + skb->sk = NULL; + } + sk_eat_skb(sk, skb); +} + static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off) { struct sk_buff *skb; @@ -1599,7 +1609,7 @@ static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off) * splitted a fat GRO packet, while we released socket lock * in skb_splice_bits() */ - sk_eat_skb(sk, skb); + tcp_eat_recv_skb(sk, skb); } return NULL; } @@ -1665,11 +1675,11 @@ int tcp_read_sock(struct sock *sk, read_descriptor_t *desc, continue; } if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) { - sk_eat_skb(sk, skb); + tcp_eat_recv_skb(sk, skb); ++seq; break; } - sk_eat_skb(sk, skb); + tcp_eat_recv_skb(sk, skb); if (!desc->count) break; WRITE_ONCE(tp->copied_seq, seq); @@ -2481,14 +2491,14 @@ skip_copy: if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) goto found_fin_ok; if (!(flags & MSG_PEEK)) - sk_eat_skb(sk, skb); + tcp_eat_recv_skb(sk, skb); continue; found_fin_ok: /* Process the FIN. */ WRITE_ONCE(*seq, *seq + 1); if (!(flags & MSG_PEEK)) - sk_eat_skb(sk, skb); + tcp_eat_recv_skb(sk, skb); break; } while (len > 0); From f35f821935d8df76f9c92e2431a225bdff938169 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:46 -0800 Subject: [PATCH 0225/2048] tcp: defer skb freeing after socket lock is released tcp recvmsg() (or rx zerocopy) spends a fair amount of time freeing skbs after their payload has been consumed. A typical ~64KB GRO packet has to release ~45 page references, eventually going to page allocator for each of them. Currently, this freeing is performed while socket lock is held, meaning that there is a high chance that BH handler has to queue incoming packets to tcp socket backlog. This can cause additional latencies, because the user thread has to process the backlog at release_sock() time, and while doing so, additional frames can be added by BH handler. This patch adds logic to defer these frees after socket lock is released, or directly from BH handler if possible. Being able to free these skbs from BH handler helps a lot, because this avoids the usual alloc/free assymetry, when BH handler and user thread do not run on same cpu or NUMA node. One cpu can now be fully utilized for the kernel->user copy, and another cpu is handling BH processing and skb/page allocs/frees (assuming RFS is not forcing use of a single CPU) Tested: 100Gbit NIC Max throughput for one TCP_STREAM flow, over 10 runs MTU : 1500 Before: 55 Gbit After: 66 Gbit MTU : 4096+(headers) Before: 82 Gbit After: 95 Gbit Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 2 ++ include/net/sock.h | 3 +++ include/net/tcp.h | 10 ++++++++++ net/ipv4/tcp.c | 27 +++++++++++++++++++++++++-- net/ipv4/tcp_ipv4.c | 1 + net/ipv6/tcp_ipv6.c | 1 + 6 files changed, 42 insertions(+), 2 deletions(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 686a666d073d..b8b806512e16 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #if IS_ENABLED(CONFIG_NF_CONNTRACK) @@ -743,6 +744,7 @@ struct sk_buff { }; struct rb_node rbnode; /* used in netem, ip4 defrag, and tcp stack */ struct list_head list; + struct llist_node ll_node; }; union { diff --git a/include/net/sock.h b/include/net/sock.h index 2d40fe4c7718..2578d1f455a7 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -63,6 +63,7 @@ #include #include #include +#include #include #include #include @@ -408,6 +409,8 @@ struct sock { struct sk_buff *head; struct sk_buff *tail; } sk_backlog; + struct llist_head defer_list; + #define sk_rmem_alloc sk_backlog.rmem_alloc int sk_forward_alloc; diff --git a/include/net/tcp.h b/include/net/tcp.h index 05c81677aaf7..44e442bf23f9 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1368,6 +1368,16 @@ static inline bool tcp_checksum_complete(struct sk_buff *skb) } bool tcp_add_backlog(struct sock *sk, struct sk_buff *skb); + +void __sk_defer_free_flush(struct sock *sk); + +static inline void sk_defer_free_flush(struct sock *sk) +{ + if (llist_empty(&sk->defer_list)) + return; + __sk_defer_free_flush(sk); +} + int tcp_filter(struct sock *sk, struct sk_buff *skb); void tcp_set_state(struct sock *sk, int state); void tcp_done(struct sock *sk); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 4e7011672aa9..33cd9a1c199c 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1580,14 +1580,34 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied) tcp_send_ack(sk); } +void __sk_defer_free_flush(struct sock *sk) +{ + struct llist_node *head; + struct sk_buff *skb, *n; + + head = llist_del_all(&sk->defer_list); + llist_for_each_entry_safe(skb, n, head, ll_node) { + prefetch(n); + skb_mark_not_on_list(skb); + __kfree_skb(skb); + } +} +EXPORT_SYMBOL(__sk_defer_free_flush); + static void tcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb) { + __skb_unlink(skb, &sk->sk_receive_queue); if (likely(skb->destructor == sock_rfree)) { sock_rfree(skb); skb->destructor = NULL; skb->sk = NULL; + if (!skb_queue_empty(&sk->sk_receive_queue) || + !llist_empty(&sk->defer_list)) { + llist_add(&skb->ll_node, &sk->defer_list); + return; + } } - sk_eat_skb(sk, skb); + __kfree_skb(skb); } static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off) @@ -2422,6 +2442,7 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, /* Do not sleep, just process backlog. */ __sk_flush_backlog(sk); } else { + sk_defer_free_flush(sk); sk_wait_data(sk, &timeo, last); } @@ -2540,6 +2561,7 @@ int tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int nonblock, ret = tcp_recvmsg_locked(sk, msg, len, nonblock, flags, &tss, &cmsg_flags); release_sock(sk); + sk_defer_free_flush(sk); if (cmsg_flags && ret >= 0) { if (cmsg_flags & TCP_CMSG_TS) @@ -3065,7 +3087,7 @@ int tcp_disconnect(struct sock *sk, int flags) sk->sk_frag.page = NULL; sk->sk_frag.offset = 0; } - + sk_defer_free_flush(sk); sk_error_report(sk); return 0; } @@ -4194,6 +4216,7 @@ static int do_tcp_getsockopt(struct sock *sk, int level, err = BPF_CGROUP_RUN_PROG_GETSOCKOPT_KERN(sk, level, optname, &zc, &len, err); release_sock(sk); + sk_defer_free_flush(sk); if (len >= offsetofend(struct tcp_zerocopy_receive, msg_flags)) goto zerocopy_rcv_cmsg; switch (len) { diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 5ad81bfb27b2..3dd19a2bf06c 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2102,6 +2102,7 @@ process: sk_incoming_cpu_update(sk); + sk_defer_free_flush(sk); bh_lock_sock_nested(sk); tcp_segs_in(tcp_sk(sk), skb); ret = 0; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index f41f14b70123..3b7d6ede1364 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1758,6 +1758,7 @@ process: sk_incoming_cpu_update(sk); + sk_defer_free_flush(sk); bh_lock_sock_nested(sk); tcp_segs_in(tcp_sk(sk), skb); ret = 0; From 8bd172b787298124ef75c0e466101107c036d54d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:47 -0800 Subject: [PATCH 0226/2048] tcp: check local var (timeo) before socket fields in one test Testing timeo before sk_err/sk_state/sk_shutdown makes more sense. Modern applications use non-blocking IO, while a socket is terminated only once during its life time. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 33cd9a1c199c..7b1886103556 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2399,10 +2399,10 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, break; if (copied) { - if (sk->sk_err || + if (!timeo || + sk->sk_err || sk->sk_state == TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN) || - !timeo || signal_pending(current)) break; } else { From 29fbc26e6dfc7be351c23261938de3f93f5cde57 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:48 -0800 Subject: [PATCH 0227/2048] tcp: do not call tcp_cleanup_rbuf() if we have a backlog Under pressure, tcp recvmsg() has logic to process the socket backlog, but calls tcp_cleanup_rbuf() right before. Avoiding sending ACK right before processing new segments makes a lot of sense, as this decrease the number of ACK packets, with no impact on effective ACK clocking. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 7b1886103556..d1949fdb1462 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2436,12 +2436,11 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, } } - tcp_cleanup_rbuf(sk, copied); - if (copied >= target) { /* Do not sleep, just process backlog. */ __sk_flush_backlog(sk); } else { + tcp_cleanup_rbuf(sk, copied); sk_defer_free_flush(sk); sk_wait_data(sk, &timeo, last); } From 43f51df4172955971ef5498f09308a9dc0291766 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 11:02:49 -0800 Subject: [PATCH 0228/2048] net: move early demux fields close to sk_refcnt sk_rx_dst/sk_rx_dst_ifindex/sk_rx_dst_cookie are read in early demux, and currently spans two cache lines. Moving them close to sk_refcnt makes more sense, as only one cache line is needed. New layout for this hot cache line is : struct sock { struct sock_common __sk_common; /* 0 0x88 */ /* --- cacheline 2 boundary (128 bytes) was 8 bytes ago --- */ struct dst_entry * sk_rx_dst; /* 0x88 0x8 */ int sk_rx_dst_ifindex; /* 0x90 0x4 */ u32 sk_rx_dst_cookie; /* 0x94 0x4 */ socket_lock_t sk_lock; /* 0x98 0x20 */ atomic_t sk_drops; /* 0xb8 0x4 */ int sk_rcvlowat; /* 0xbc 0x4 */ /* --- cacheline 3 boundary (192 bytes) --- */ Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 2578d1f455a7..95cc03bd3fac 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -390,6 +390,11 @@ struct sock { #define sk_flags __sk_common.skc_flags #define sk_rxhash __sk_common.skc_rxhash + /* early demux fields */ + struct dst_entry *sk_rx_dst; + int sk_rx_dst_ifindex; + u32 sk_rx_dst_cookie; + socket_lock_t sk_lock; atomic_t sk_drops; int sk_rcvlowat; @@ -432,9 +437,6 @@ struct sock { #ifdef CONFIG_XFRM struct xfrm_policy __rcu *sk_policy[2]; #endif - struct dst_entry *sk_rx_dst; - int sk_rx_dst_ifindex; - u32 sk_rx_dst_cookie; struct dst_entry __rcu *sk_dst_cache; atomic_t sk_omem_alloc; From 4721031c3559db8eae61df305f10c00099a7c1d0 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:05:51 -0800 Subject: [PATCH 0229/2048] net: move gro definitions to include/net/gro.h include/linux/netdevice.h became too big, move gro stuff into include/net/gro.h Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- .../net/ethernet/broadcom/bnx2x/bnx2x_cmn.c | 1 + drivers/net/ethernet/broadcom/bnxt/bnxt.c | 1 + .../net/ethernet/hisilicon/hns3/hns3_enet.c | 1 + .../net/ethernet/mellanox/mlx5/core/en_rx.c | 1 + drivers/net/ethernet/qlogic/qede/qede_fp.c | 1 + drivers/net/geneve.c | 1 + drivers/net/vxlan.c | 1 + include/linux/netdevice.h | 348 --------------- include/net/gro.h | 396 +++++++++++++++++- include/net/ip.h | 8 - include/net/ip6_checksum.h | 8 - include/net/udp.h | 24 -- net/core/skbuff.c | 1 + net/ipv4/af_inet.c | 1 + net/ipv4/esp4_offload.c | 1 + net/ipv4/fou.c | 1 + net/ipv4/gre_offload.c | 1 + net/ipv4/tcp_offload.c | 1 + net/ipv4/udp_offload.c | 1 + net/ipv6/esp6_offload.c | 1 + net/ipv6/tcpv6_offload.c | 1 + net/ipv6/udp_offload.c | 1 + 22 files changed, 411 insertions(+), 390 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c index e8e8c2d593c5..54a2334dee56 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index c04ea83188e2..c057b1df86a9 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index 9ccebbaa0d69..13835a37b3a2 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index 96967b0a2441..e384f6458c06 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include #include "en.h" diff --git a/drivers/net/ethernet/qlogic/qede/qede_fp.c b/drivers/net/ethernet/qlogic/qede/qede_fp.c index 065e9004598e..e113fbd56e86 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_fp.c +++ b/drivers/net/ethernet/qlogic/qede/qede_fp.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index 1ab94b5f9bbf..9d26d1b965d2 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #define GENEVE_NETDEV_VER "0.6" diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index 141635a35c28..563f86de0e0d 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 3ec42495a43a..d95c9839ce90 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2520,109 +2520,6 @@ static inline void netif_napi_del(struct napi_struct *napi) synchronize_net(); } -struct napi_gro_cb { - /* Virtual address of skb_shinfo(skb)->frags[0].page + offset. */ - void *frag0; - - /* Length of frag0. */ - unsigned int frag0_len; - - /* This indicates where we are processing relative to skb->data. */ - int data_offset; - - /* This is non-zero if the packet cannot be merged with the new skb. */ - u16 flush; - - /* Save the IP ID here and check when we get to the transport layer */ - u16 flush_id; - - /* Number of segments aggregated. */ - u16 count; - - /* Start offset for remote checksum offload */ - u16 gro_remcsum_start; - - /* jiffies when first packet was created/queued */ - unsigned long age; - - /* Used in ipv6_gro_receive() and foo-over-udp */ - u16 proto; - - /* This is non-zero if the packet may be of the same flow. */ - u8 same_flow:1; - - /* Used in tunnel GRO receive */ - u8 encap_mark:1; - - /* GRO checksum is valid */ - u8 csum_valid:1; - - /* Number of checksums via CHECKSUM_UNNECESSARY */ - u8 csum_cnt:3; - - /* Free the skb? */ - u8 free:2; -#define NAPI_GRO_FREE 1 -#define NAPI_GRO_FREE_STOLEN_HEAD 2 - - /* Used in foo-over-udp, set in udp[46]_gro_receive */ - u8 is_ipv6:1; - - /* Used in GRE, set in fou/gue_gro_receive */ - u8 is_fou:1; - - /* Used to determine if flush_id can be ignored */ - u8 is_atomic:1; - - /* Number of gro_receive callbacks this packet already went through */ - u8 recursion_counter:4; - - /* GRO is done by frag_list pointer chaining. */ - u8 is_flist:1; - - /* used to support CHECKSUM_COMPLETE for tunneling protocols */ - __wsum csum; - - /* used in skb_gro_receive() slow path */ - struct sk_buff *last; -}; - -#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb) - -#define GRO_RECURSION_LIMIT 15 -static inline int gro_recursion_inc_test(struct sk_buff *skb) -{ - return ++NAPI_GRO_CB(skb)->recursion_counter == GRO_RECURSION_LIMIT; -} - -typedef struct sk_buff *(*gro_receive_t)(struct list_head *, struct sk_buff *); -static inline struct sk_buff *call_gro_receive(gro_receive_t cb, - struct list_head *head, - struct sk_buff *skb) -{ - if (unlikely(gro_recursion_inc_test(skb))) { - NAPI_GRO_CB(skb)->flush |= 1; - return NULL; - } - - return cb(head, skb); -} - -typedef struct sk_buff *(*gro_receive_sk_t)(struct sock *, struct list_head *, - struct sk_buff *); -static inline struct sk_buff *call_gro_receive_sk(gro_receive_sk_t cb, - struct sock *sk, - struct list_head *head, - struct sk_buff *skb) -{ - if (unlikely(gro_recursion_inc_test(skb))) { - NAPI_GRO_CB(skb)->flush |= 1; - return NULL; - } - - return cb(sk, head, skb); -} - struct packet_type { __be16 type; /* This is really htons(ether_type). */ bool ignore_outgoing; @@ -3008,251 +2905,6 @@ int dev_restart(struct net_device *dev); int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb); int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb); -static inline unsigned int skb_gro_offset(const struct sk_buff *skb) -{ - return NAPI_GRO_CB(skb)->data_offset; -} - -static inline unsigned int skb_gro_len(const struct sk_buff *skb) -{ - return skb->len - NAPI_GRO_CB(skb)->data_offset; -} - -static inline void skb_gro_pull(struct sk_buff *skb, unsigned int len) -{ - NAPI_GRO_CB(skb)->data_offset += len; -} - -static inline void *skb_gro_header_fast(struct sk_buff *skb, - unsigned int offset) -{ - return NAPI_GRO_CB(skb)->frag0 + offset; -} - -static inline int skb_gro_header_hard(struct sk_buff *skb, unsigned int hlen) -{ - return NAPI_GRO_CB(skb)->frag0_len < hlen; -} - -static inline void skb_gro_frag0_invalidate(struct sk_buff *skb) -{ - NAPI_GRO_CB(skb)->frag0 = NULL; - NAPI_GRO_CB(skb)->frag0_len = 0; -} - -static inline void *skb_gro_header_slow(struct sk_buff *skb, unsigned int hlen, - unsigned int offset) -{ - if (!pskb_may_pull(skb, hlen)) - return NULL; - - skb_gro_frag0_invalidate(skb); - return skb->data + offset; -} - -static inline void *skb_gro_network_header(struct sk_buff *skb) -{ - return (NAPI_GRO_CB(skb)->frag0 ?: skb->data) + - skb_network_offset(skb); -} - -static inline void skb_gro_postpull_rcsum(struct sk_buff *skb, - const void *start, unsigned int len) -{ - if (NAPI_GRO_CB(skb)->csum_valid) - NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum, - csum_partial(start, len, 0)); -} - -/* GRO checksum functions. These are logical equivalents of the normal - * checksum functions (in skbuff.h) except that they operate on the GRO - * offsets and fields in sk_buff. - */ - -__sum16 __skb_gro_checksum_complete(struct sk_buff *skb); - -static inline bool skb_at_gro_remcsum_start(struct sk_buff *skb) -{ - return (NAPI_GRO_CB(skb)->gro_remcsum_start == skb_gro_offset(skb)); -} - -static inline bool __skb_gro_checksum_validate_needed(struct sk_buff *skb, - bool zero_okay, - __sum16 check) -{ - return ((skb->ip_summed != CHECKSUM_PARTIAL || - skb_checksum_start_offset(skb) < - skb_gro_offset(skb)) && - !skb_at_gro_remcsum_start(skb) && - NAPI_GRO_CB(skb)->csum_cnt == 0 && - (!zero_okay || check)); -} - -static inline __sum16 __skb_gro_checksum_validate_complete(struct sk_buff *skb, - __wsum psum) -{ - if (NAPI_GRO_CB(skb)->csum_valid && - !csum_fold(csum_add(psum, NAPI_GRO_CB(skb)->csum))) - return 0; - - NAPI_GRO_CB(skb)->csum = psum; - - return __skb_gro_checksum_complete(skb); -} - -static inline void skb_gro_incr_csum_unnecessary(struct sk_buff *skb) -{ - if (NAPI_GRO_CB(skb)->csum_cnt > 0) { - /* Consume a checksum from CHECKSUM_UNNECESSARY */ - NAPI_GRO_CB(skb)->csum_cnt--; - } else { - /* Update skb for CHECKSUM_UNNECESSARY and csum_level when we - * verified a new top level checksum or an encapsulated one - * during GRO. This saves work if we fallback to normal path. - */ - __skb_incr_checksum_unnecessary(skb); - } -} - -#define __skb_gro_checksum_validate(skb, proto, zero_okay, check, \ - compute_pseudo) \ -({ \ - __sum16 __ret = 0; \ - if (__skb_gro_checksum_validate_needed(skb, zero_okay, check)) \ - __ret = __skb_gro_checksum_validate_complete(skb, \ - compute_pseudo(skb, proto)); \ - if (!__ret) \ - skb_gro_incr_csum_unnecessary(skb); \ - __ret; \ -}) - -#define skb_gro_checksum_validate(skb, proto, compute_pseudo) \ - __skb_gro_checksum_validate(skb, proto, false, 0, compute_pseudo) - -#define skb_gro_checksum_validate_zero_check(skb, proto, check, \ - compute_pseudo) \ - __skb_gro_checksum_validate(skb, proto, true, check, compute_pseudo) - -#define skb_gro_checksum_simple_validate(skb) \ - __skb_gro_checksum_validate(skb, 0, false, 0, null_compute_pseudo) - -static inline bool __skb_gro_checksum_convert_check(struct sk_buff *skb) -{ - return (NAPI_GRO_CB(skb)->csum_cnt == 0 && - !NAPI_GRO_CB(skb)->csum_valid); -} - -static inline void __skb_gro_checksum_convert(struct sk_buff *skb, - __wsum pseudo) -{ - NAPI_GRO_CB(skb)->csum = ~pseudo; - NAPI_GRO_CB(skb)->csum_valid = 1; -} - -#define skb_gro_checksum_try_convert(skb, proto, compute_pseudo) \ -do { \ - if (__skb_gro_checksum_convert_check(skb)) \ - __skb_gro_checksum_convert(skb, \ - compute_pseudo(skb, proto)); \ -} while (0) - -struct gro_remcsum { - int offset; - __wsum delta; -}; - -static inline void skb_gro_remcsum_init(struct gro_remcsum *grc) -{ - grc->offset = 0; - grc->delta = 0; -} - -static inline void *skb_gro_remcsum_process(struct sk_buff *skb, void *ptr, - unsigned int off, size_t hdrlen, - int start, int offset, - struct gro_remcsum *grc, - bool nopartial) -{ - __wsum delta; - size_t plen = hdrlen + max_t(size_t, offset + sizeof(u16), start); - - BUG_ON(!NAPI_GRO_CB(skb)->csum_valid); - - if (!nopartial) { - NAPI_GRO_CB(skb)->gro_remcsum_start = off + hdrlen + start; - return ptr; - } - - ptr = skb_gro_header_fast(skb, off); - if (skb_gro_header_hard(skb, off + plen)) { - ptr = skb_gro_header_slow(skb, off + plen, off); - if (!ptr) - return NULL; - } - - delta = remcsum_adjust(ptr + hdrlen, NAPI_GRO_CB(skb)->csum, - start, offset); - - /* Adjust skb->csum since we changed the packet */ - NAPI_GRO_CB(skb)->csum = csum_add(NAPI_GRO_CB(skb)->csum, delta); - - grc->offset = off + hdrlen + offset; - grc->delta = delta; - - return ptr; -} - -static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb, - struct gro_remcsum *grc) -{ - void *ptr; - size_t plen = grc->offset + sizeof(u16); - - if (!grc->delta) - return; - - ptr = skb_gro_header_fast(skb, grc->offset); - if (skb_gro_header_hard(skb, grc->offset + sizeof(u16))) { - ptr = skb_gro_header_slow(skb, plen, grc->offset); - if (!ptr) - return; - } - - remcsum_unadjust((__sum16 *)ptr, grc->delta); -} - -#ifdef CONFIG_XFRM_OFFLOAD -static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush) -{ - if (PTR_ERR(pp) != -EINPROGRESS) - NAPI_GRO_CB(skb)->flush |= flush; -} -static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, - struct sk_buff *pp, - int flush, - struct gro_remcsum *grc) -{ - if (PTR_ERR(pp) != -EINPROGRESS) { - NAPI_GRO_CB(skb)->flush |= flush; - skb_gro_remcsum_cleanup(skb, grc); - skb->remcsum_offload = 0; - } -} -#else -static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush) -{ - NAPI_GRO_CB(skb)->flush |= flush; -} -static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, - struct sk_buff *pp, - int flush, - struct gro_remcsum *grc) -{ - NAPI_GRO_CB(skb)->flush |= flush; - skb_gro_remcsum_cleanup(skb, grc); - skb->remcsum_offload = 0; -} -#endif static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev, unsigned short type, diff --git a/include/net/gro.h b/include/net/gro.h index 01edaf3fdda0..1ffbe74b2e35 100644 --- a/include/net/gro.h +++ b/include/net/gro.h @@ -4,9 +4,367 @@ #define _NET_IPV6_GRO_H #include +#include +#include +#include +#include -struct list_head; -struct sk_buff; +struct napi_gro_cb { + /* Virtual address of skb_shinfo(skb)->frags[0].page + offset. */ + void *frag0; + + /* Length of frag0. */ + unsigned int frag0_len; + + /* This indicates where we are processing relative to skb->data. */ + int data_offset; + + /* This is non-zero if the packet cannot be merged with the new skb. */ + u16 flush; + + /* Save the IP ID here and check when we get to the transport layer */ + u16 flush_id; + + /* Number of segments aggregated. */ + u16 count; + + /* Start offset for remote checksum offload */ + u16 gro_remcsum_start; + + /* jiffies when first packet was created/queued */ + unsigned long age; + + /* Used in ipv6_gro_receive() and foo-over-udp */ + u16 proto; + + /* This is non-zero if the packet may be of the same flow. */ + u8 same_flow:1; + + /* Used in tunnel GRO receive */ + u8 encap_mark:1; + + /* GRO checksum is valid */ + u8 csum_valid:1; + + /* Number of checksums via CHECKSUM_UNNECESSARY */ + u8 csum_cnt:3; + + /* Free the skb? */ + u8 free:2; +#define NAPI_GRO_FREE 1 +#define NAPI_GRO_FREE_STOLEN_HEAD 2 + + /* Used in foo-over-udp, set in udp[46]_gro_receive */ + u8 is_ipv6:1; + + /* Used in GRE, set in fou/gue_gro_receive */ + u8 is_fou:1; + + /* Used to determine if flush_id can be ignored */ + u8 is_atomic:1; + + /* Number of gro_receive callbacks this packet already went through */ + u8 recursion_counter:4; + + /* GRO is done by frag_list pointer chaining. */ + u8 is_flist:1; + + /* used to support CHECKSUM_COMPLETE for tunneling protocols */ + __wsum csum; + + /* used in skb_gro_receive() slow path */ + struct sk_buff *last; +}; + +#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb) + +#define GRO_RECURSION_LIMIT 15 +static inline int gro_recursion_inc_test(struct sk_buff *skb) +{ + return ++NAPI_GRO_CB(skb)->recursion_counter == GRO_RECURSION_LIMIT; +} + +typedef struct sk_buff *(*gro_receive_t)(struct list_head *, struct sk_buff *); +static inline struct sk_buff *call_gro_receive(gro_receive_t cb, + struct list_head *head, + struct sk_buff *skb) +{ + if (unlikely(gro_recursion_inc_test(skb))) { + NAPI_GRO_CB(skb)->flush |= 1; + return NULL; + } + + return cb(head, skb); +} + +typedef struct sk_buff *(*gro_receive_sk_t)(struct sock *, struct list_head *, + struct sk_buff *); +static inline struct sk_buff *call_gro_receive_sk(gro_receive_sk_t cb, + struct sock *sk, + struct list_head *head, + struct sk_buff *skb) +{ + if (unlikely(gro_recursion_inc_test(skb))) { + NAPI_GRO_CB(skb)->flush |= 1; + return NULL; + } + + return cb(sk, head, skb); +} + +static inline unsigned int skb_gro_offset(const struct sk_buff *skb) +{ + return NAPI_GRO_CB(skb)->data_offset; +} + +static inline unsigned int skb_gro_len(const struct sk_buff *skb) +{ + return skb->len - NAPI_GRO_CB(skb)->data_offset; +} + +static inline void skb_gro_pull(struct sk_buff *skb, unsigned int len) +{ + NAPI_GRO_CB(skb)->data_offset += len; +} + +static inline void *skb_gro_header_fast(struct sk_buff *skb, + unsigned int offset) +{ + return NAPI_GRO_CB(skb)->frag0 + offset; +} + +static inline int skb_gro_header_hard(struct sk_buff *skb, unsigned int hlen) +{ + return NAPI_GRO_CB(skb)->frag0_len < hlen; +} + +static inline void skb_gro_frag0_invalidate(struct sk_buff *skb) +{ + NAPI_GRO_CB(skb)->frag0 = NULL; + NAPI_GRO_CB(skb)->frag0_len = 0; +} + +static inline void *skb_gro_header_slow(struct sk_buff *skb, unsigned int hlen, + unsigned int offset) +{ + if (!pskb_may_pull(skb, hlen)) + return NULL; + + skb_gro_frag0_invalidate(skb); + return skb->data + offset; +} + +static inline void *skb_gro_network_header(struct sk_buff *skb) +{ + return (NAPI_GRO_CB(skb)->frag0 ?: skb->data) + + skb_network_offset(skb); +} + +static inline __wsum inet_gro_compute_pseudo(struct sk_buff *skb, int proto) +{ + const struct iphdr *iph = skb_gro_network_header(skb); + + return csum_tcpudp_nofold(iph->saddr, iph->daddr, + skb_gro_len(skb), proto, 0); +} + +static inline void skb_gro_postpull_rcsum(struct sk_buff *skb, + const void *start, unsigned int len) +{ + if (NAPI_GRO_CB(skb)->csum_valid) + NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum, + csum_partial(start, len, 0)); +} + +/* GRO checksum functions. These are logical equivalents of the normal + * checksum functions (in skbuff.h) except that they operate on the GRO + * offsets and fields in sk_buff. + */ + +__sum16 __skb_gro_checksum_complete(struct sk_buff *skb); + +static inline bool skb_at_gro_remcsum_start(struct sk_buff *skb) +{ + return (NAPI_GRO_CB(skb)->gro_remcsum_start == skb_gro_offset(skb)); +} + +static inline bool __skb_gro_checksum_validate_needed(struct sk_buff *skb, + bool zero_okay, + __sum16 check) +{ + return ((skb->ip_summed != CHECKSUM_PARTIAL || + skb_checksum_start_offset(skb) < + skb_gro_offset(skb)) && + !skb_at_gro_remcsum_start(skb) && + NAPI_GRO_CB(skb)->csum_cnt == 0 && + (!zero_okay || check)); +} + +static inline __sum16 __skb_gro_checksum_validate_complete(struct sk_buff *skb, + __wsum psum) +{ + if (NAPI_GRO_CB(skb)->csum_valid && + !csum_fold(csum_add(psum, NAPI_GRO_CB(skb)->csum))) + return 0; + + NAPI_GRO_CB(skb)->csum = psum; + + return __skb_gro_checksum_complete(skb); +} + +static inline void skb_gro_incr_csum_unnecessary(struct sk_buff *skb) +{ + if (NAPI_GRO_CB(skb)->csum_cnt > 0) { + /* Consume a checksum from CHECKSUM_UNNECESSARY */ + NAPI_GRO_CB(skb)->csum_cnt--; + } else { + /* Update skb for CHECKSUM_UNNECESSARY and csum_level when we + * verified a new top level checksum or an encapsulated one + * during GRO. This saves work if we fallback to normal path. + */ + __skb_incr_checksum_unnecessary(skb); + } +} + +#define __skb_gro_checksum_validate(skb, proto, zero_okay, check, \ + compute_pseudo) \ +({ \ + __sum16 __ret = 0; \ + if (__skb_gro_checksum_validate_needed(skb, zero_okay, check)) \ + __ret = __skb_gro_checksum_validate_complete(skb, \ + compute_pseudo(skb, proto)); \ + if (!__ret) \ + skb_gro_incr_csum_unnecessary(skb); \ + __ret; \ +}) + +#define skb_gro_checksum_validate(skb, proto, compute_pseudo) \ + __skb_gro_checksum_validate(skb, proto, false, 0, compute_pseudo) + +#define skb_gro_checksum_validate_zero_check(skb, proto, check, \ + compute_pseudo) \ + __skb_gro_checksum_validate(skb, proto, true, check, compute_pseudo) + +#define skb_gro_checksum_simple_validate(skb) \ + __skb_gro_checksum_validate(skb, 0, false, 0, null_compute_pseudo) + +static inline bool __skb_gro_checksum_convert_check(struct sk_buff *skb) +{ + return (NAPI_GRO_CB(skb)->csum_cnt == 0 && + !NAPI_GRO_CB(skb)->csum_valid); +} + +static inline void __skb_gro_checksum_convert(struct sk_buff *skb, + __wsum pseudo) +{ + NAPI_GRO_CB(skb)->csum = ~pseudo; + NAPI_GRO_CB(skb)->csum_valid = 1; +} + +#define skb_gro_checksum_try_convert(skb, proto, compute_pseudo) \ +do { \ + if (__skb_gro_checksum_convert_check(skb)) \ + __skb_gro_checksum_convert(skb, \ + compute_pseudo(skb, proto)); \ +} while (0) + +struct gro_remcsum { + int offset; + __wsum delta; +}; + +static inline void skb_gro_remcsum_init(struct gro_remcsum *grc) +{ + grc->offset = 0; + grc->delta = 0; +} + +static inline void *skb_gro_remcsum_process(struct sk_buff *skb, void *ptr, + unsigned int off, size_t hdrlen, + int start, int offset, + struct gro_remcsum *grc, + bool nopartial) +{ + __wsum delta; + size_t plen = hdrlen + max_t(size_t, offset + sizeof(u16), start); + + BUG_ON(!NAPI_GRO_CB(skb)->csum_valid); + + if (!nopartial) { + NAPI_GRO_CB(skb)->gro_remcsum_start = off + hdrlen + start; + return ptr; + } + + ptr = skb_gro_header_fast(skb, off); + if (skb_gro_header_hard(skb, off + plen)) { + ptr = skb_gro_header_slow(skb, off + plen, off); + if (!ptr) + return NULL; + } + + delta = remcsum_adjust(ptr + hdrlen, NAPI_GRO_CB(skb)->csum, + start, offset); + + /* Adjust skb->csum since we changed the packet */ + NAPI_GRO_CB(skb)->csum = csum_add(NAPI_GRO_CB(skb)->csum, delta); + + grc->offset = off + hdrlen + offset; + grc->delta = delta; + + return ptr; +} + +static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb, + struct gro_remcsum *grc) +{ + void *ptr; + size_t plen = grc->offset + sizeof(u16); + + if (!grc->delta) + return; + + ptr = skb_gro_header_fast(skb, grc->offset); + if (skb_gro_header_hard(skb, grc->offset + sizeof(u16))) { + ptr = skb_gro_header_slow(skb, plen, grc->offset); + if (!ptr) + return; + } + + remcsum_unadjust((__sum16 *)ptr, grc->delta); +} + +#ifdef CONFIG_XFRM_OFFLOAD +static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush) +{ + if (PTR_ERR(pp) != -EINPROGRESS) + NAPI_GRO_CB(skb)->flush |= flush; +} +static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, + struct sk_buff *pp, + int flush, + struct gro_remcsum *grc) +{ + if (PTR_ERR(pp) != -EINPROGRESS) { + NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_remcsum_cleanup(skb, grc); + skb->remcsum_offload = 0; + } +} +#else +static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush) +{ + NAPI_GRO_CB(skb)->flush |= flush; +} +static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, + struct sk_buff *pp, + int flush, + struct gro_remcsum *grc) +{ + NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_remcsum_cleanup(skb, grc); + skb->remcsum_offload = 0; +} +#endif INDIRECT_CALLABLE_DECLARE(struct sk_buff *ipv6_gro_receive(struct list_head *, struct sk_buff *)); @@ -15,6 +373,14 @@ INDIRECT_CALLABLE_DECLARE(struct sk_buff *inet_gro_receive(struct list_head *, struct sk_buff *)); INDIRECT_CALLABLE_DECLARE(int inet_gro_complete(struct sk_buff *, int)); +INDIRECT_CALLABLE_DECLARE(struct sk_buff *udp4_gro_receive(struct list_head *, + struct sk_buff *)); +INDIRECT_CALLABLE_DECLARE(int udp4_gro_complete(struct sk_buff *, int)); + +INDIRECT_CALLABLE_DECLARE(struct sk_buff *udp6_gro_receive(struct list_head *, + struct sk_buff *)); +INDIRECT_CALLABLE_DECLARE(int udp6_gro_complete(struct sk_buff *, int)); + #define indirect_call_gro_receive_inet(cb, f2, f1, head, skb) \ ({ \ unlikely(gro_recursion_inc_test(skb)) ? \ @@ -22,4 +388,30 @@ INDIRECT_CALLABLE_DECLARE(int inet_gro_complete(struct sk_buff *, int)); INDIRECT_CALL_INET(cb, f2, f1, head, skb); \ }) +struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb, + struct udphdr *uh, struct sock *sk); +int udp_gro_complete(struct sk_buff *skb, int nhoff, udp_lookup_t lookup); + +static inline struct udphdr *udp_gro_udphdr(struct sk_buff *skb) +{ + struct udphdr *uh; + unsigned int hlen, off; + + off = skb_gro_offset(skb); + hlen = off + sizeof(*uh); + uh = skb_gro_header_fast(skb, off); + if (skb_gro_header_hard(skb, hlen)) + uh = skb_gro_header_slow(skb, hlen, off); + + return uh; +} + +static inline __wsum ip6_gro_compute_pseudo(struct sk_buff *skb, int proto) +{ + const struct ipv6hdr *iph = skb_gro_network_header(skb); + + return ~csum_unfold(csum_ipv6_magic(&iph->saddr, &iph->daddr, + skb_gro_len(skb), proto, 0)); +} + #endif /* _NET_IPV6_GRO_H */ diff --git a/include/net/ip.h b/include/net/ip.h index b71e88507c4a..7d1088888c10 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -568,14 +568,6 @@ static inline void iph_to_flow_copy_v4addrs(struct flow_keys *flow, flow->control.addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS; } -static inline __wsum inet_gro_compute_pseudo(struct sk_buff *skb, int proto) -{ - const struct iphdr *iph = skb_gro_network_header(skb); - - return csum_tcpudp_nofold(iph->saddr, iph->daddr, - skb_gro_len(skb), proto, 0); -} - /* * Map a multicast IP onto multicast MAC for type ethernet. */ diff --git a/include/net/ip6_checksum.h b/include/net/ip6_checksum.h index ea681910b7a3..c8a96b888277 100644 --- a/include/net/ip6_checksum.h +++ b/include/net/ip6_checksum.h @@ -43,14 +43,6 @@ static inline __wsum ip6_compute_pseudo(struct sk_buff *skb, int proto) skb->len, proto, 0)); } -static inline __wsum ip6_gro_compute_pseudo(struct sk_buff *skb, int proto) -{ - const struct ipv6hdr *iph = skb_gro_network_header(skb); - - return ~csum_unfold(csum_ipv6_magic(&iph->saddr, &iph->daddr, - skb_gro_len(skb), proto, 0)); -} - static __inline__ __sum16 tcp_v6_check(int len, const struct in6_addr *saddr, const struct in6_addr *daddr, diff --git a/include/net/udp.h b/include/net/udp.h index 909ecf447e0f..f1c2a88c9005 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -167,36 +167,12 @@ static inline void udp_csum_pull_header(struct sk_buff *skb) typedef struct sock *(*udp_lookup_t)(const struct sk_buff *skb, __be16 sport, __be16 dport); -INDIRECT_CALLABLE_DECLARE(struct sk_buff *udp4_gro_receive(struct list_head *, - struct sk_buff *)); -INDIRECT_CALLABLE_DECLARE(int udp4_gro_complete(struct sk_buff *, int)); -INDIRECT_CALLABLE_DECLARE(struct sk_buff *udp6_gro_receive(struct list_head *, - struct sk_buff *)); -INDIRECT_CALLABLE_DECLARE(int udp6_gro_complete(struct sk_buff *, int)); INDIRECT_CALLABLE_DECLARE(void udp_v6_early_demux(struct sk_buff *)); INDIRECT_CALLABLE_DECLARE(int udpv6_rcv(struct sk_buff *)); -struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb, - struct udphdr *uh, struct sock *sk); -int udp_gro_complete(struct sk_buff *skb, int nhoff, udp_lookup_t lookup); - struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb, netdev_features_t features, bool is_ipv6); -static inline struct udphdr *udp_gro_udphdr(struct sk_buff *skb) -{ - struct udphdr *uh; - unsigned int hlen, off; - - off = skb_gro_offset(skb); - hlen = off + sizeof(*uh); - uh = skb_gro_header_fast(skb, off); - if (skb_gro_header_hard(skb, hlen)) - uh = skb_gro_header_slow(skb, hlen, off); - - return uh; -} - /* hash routines shared between UDPv4/6 and UDP-Litev4/6 */ static inline int udp_lib_hash(struct sock *sk) { diff --git a/net/core/skbuff.c b/net/core/skbuff.c index d57796f38a0b..5b9472bdceff 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -64,6 +64,7 @@ #include #include +#include #include #include #include diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 0189e3cd4a7d..6d52b6491255 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -99,6 +99,7 @@ #include #include #include +#include #include #include #include diff --git a/net/ipv4/esp4_offload.c b/net/ipv4/esp4_offload.c index 8e4e9aa12130..d87f02a6e934 100644 --- a/net/ipv4/esp4_offload.c +++ b/net/ipv4/esp4_offload.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #include diff --git a/net/ipv4/fou.c b/net/ipv4/fou.c index 8fcbc6258ec5..b56d6b40c0a2 100644 --- a/net/ipv4/fou.c +++ b/net/ipv4/fou.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include #include diff --git a/net/ipv4/gre_offload.c b/net/ipv4/gre_offload.c index 1121a9d5fed9..740298dac7d3 100644 --- a/net/ipv4/gre_offload.c +++ b/net/ipv4/gre_offload.c @@ -10,6 +10,7 @@ #include #include #include +#include static struct sk_buff *gre_gso_segment(struct sk_buff *skb, netdev_features_t features) diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c index fc61cd3fea65..30abde86db45 100644 --- a/net/ipv4/tcp_offload.c +++ b/net/ipv4/tcp_offload.c @@ -8,6 +8,7 @@ #include #include +#include #include #include diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c index 86d32a1e62ac..7fbf9975e8c0 100644 --- a/net/ipv4/udp_offload.c +++ b/net/ipv4/udp_offload.c @@ -7,6 +7,7 @@ */ #include +#include #include #include #include diff --git a/net/ipv6/esp6_offload.c b/net/ipv6/esp6_offload.c index a349d4798077..ba5e81cd569c 100644 --- a/net/ipv6/esp6_offload.c +++ b/net/ipv6/esp6_offload.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #include diff --git a/net/ipv6/tcpv6_offload.c b/net/ipv6/tcpv6_offload.c index 1796856bc24f..39db5a226855 100644 --- a/net/ipv6/tcpv6_offload.c +++ b/net/ipv6/tcpv6_offload.c @@ -7,6 +7,7 @@ */ #include #include +#include #include #include #include diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c index b3d9ed96e5ea..50a8a65fad23 100644 --- a/net/ipv6/udp_offload.c +++ b/net/ipv6/udp_offload.c @@ -13,6 +13,7 @@ #include #include #include "ip6_offload.h" +#include static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb, netdev_features_t features) From 0b935d7f8c07bf0a192712bdbf76dbf45ef8b115 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:05:52 -0800 Subject: [PATCH 0230/2048] net: gro: move skb_gro_receive_list to udp_offload.c This helper is used once, no need to keep it in fat net/core/skbuff.c Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 - net/core/skbuff.c | 26 -------------------------- net/ipv4/udp_offload.c | 27 +++++++++++++++++++++++++++ 3 files changed, 27 insertions(+), 27 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index d95c9839ce90..ce6ee1453dbc 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2903,7 +2903,6 @@ struct net_device *dev_get_by_napi_id(unsigned int napi_id); int netdev_get_name(struct net *net, char *name, int ifindex); int dev_restart(struct net_device *dev); int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb); -int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb); static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev, diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 5b9472bdceff..8560d50c960b 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -3920,32 +3920,6 @@ err_linearize: } EXPORT_SYMBOL_GPL(skb_segment_list); -int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb) -{ - if (unlikely(p->len + skb->len >= 65536)) - return -E2BIG; - - if (NAPI_GRO_CB(p)->last == p) - skb_shinfo(p)->frag_list = skb; - else - NAPI_GRO_CB(p)->last->next = skb; - - skb_pull(skb, skb_gro_offset(skb)); - - NAPI_GRO_CB(p)->last = skb; - NAPI_GRO_CB(p)->count++; - p->data_len += skb->len; - - /* sk owenrship - if any - completely transferred to the aggregated packet */ - skb->destructor = NULL; - p->truesize += skb->truesize; - p->len += skb->len; - - NAPI_GRO_CB(skb)->same_flow = 1; - - return 0; -} - /** * skb_segment - Perform protocol segmentation on skb. * @head_skb: buffer to segment diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c index 7fbf9975e8c0..cbeb8965d1b7 100644 --- a/net/ipv4/udp_offload.c +++ b/net/ipv4/udp_offload.c @@ -425,6 +425,33 @@ out: return segs; } +static int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb) +{ + if (unlikely(p->len + skb->len >= 65536)) + return -E2BIG; + + if (NAPI_GRO_CB(p)->last == p) + skb_shinfo(p)->frag_list = skb; + else + NAPI_GRO_CB(p)->last->next = skb; + + skb_pull(skb, skb_gro_offset(skb)); + + NAPI_GRO_CB(p)->last = skb; + NAPI_GRO_CB(p)->count++; + p->data_len += skb->len; + + /* sk owenrship - if any - completely transferred to the aggregated packet */ + skb->destructor = NULL; + p->truesize += skb->truesize; + p->len += skb->len; + + NAPI_GRO_CB(skb)->same_flow = 1; + + return 0; +} + + #define UDP_GRO_CNT_MAX 64 static struct sk_buff *udp_gro_receive_segment(struct list_head *head, struct sk_buff *skb) From e456a18a390b96f22b0de2acd4d0f49c72ed2280 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:05:53 -0800 Subject: [PATCH 0231/2048] net: gro: move skb_gro_receive into net/core/gro.c net/core/gro.c will contain all core gro functions, to shrink net/core/skbuff.c and net/core/dev.c Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 - include/net/gro.h | 2 + net/core/Makefile | 2 +- net/core/gro.c | 118 ++++++++++++++++++++++++++++++++++++++ net/core/skbuff.c | 117 ------------------------------------- 5 files changed, 121 insertions(+), 119 deletions(-) create mode 100644 net/core/gro.c diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index ce6ee1453dbc..93d397db9ec4 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2902,7 +2902,6 @@ struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex); struct net_device *dev_get_by_napi_id(unsigned int napi_id); int netdev_get_name(struct net *net, char *name, int ifindex); int dev_restart(struct net_device *dev); -int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb); static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev, diff --git a/include/net/gro.h b/include/net/gro.h index 1ffbe74b2e35..f988bf3440f8 100644 --- a/include/net/gro.h +++ b/include/net/gro.h @@ -414,4 +414,6 @@ static inline __wsum ip6_gro_compute_pseudo(struct sk_buff *skb, int proto) skb_gro_len(skb), proto, 0)); } +int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb); + #endif /* _NET_IPV6_GRO_H */ diff --git a/net/core/Makefile b/net/core/Makefile index 4268846f2f47..6bdcb2cafed8 100644 --- a/net/core/Makefile +++ b/net/core/Makefile @@ -11,7 +11,7 @@ obj-$(CONFIG_SYSCTL) += sysctl_net_core.o obj-y += dev.o dev_addr_lists.o dst.o netevent.o \ neighbour.o rtnetlink.o utils.o link_watch.o filter.o \ sock_diag.o dev_ioctl.o tso.o sock_reuseport.o \ - fib_notifier.o xdp.o flow_offload.o + fib_notifier.o xdp.o flow_offload.o gro.o obj-y += net-sysfs.o obj-$(CONFIG_PAGE_POOL) += page_pool.o diff --git a/net/core/gro.c b/net/core/gro.c new file mode 100644 index 000000000000..91a74c4da9ff --- /dev/null +++ b/net/core/gro.c @@ -0,0 +1,118 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +#include + +int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb) +{ + struct skb_shared_info *pinfo, *skbinfo = skb_shinfo(skb); + unsigned int offset = skb_gro_offset(skb); + unsigned int headlen = skb_headlen(skb); + unsigned int len = skb_gro_len(skb); + unsigned int delta_truesize; + unsigned int new_truesize; + struct sk_buff *lp; + + if (unlikely(p->len + len >= 65536 || NAPI_GRO_CB(skb)->flush)) + return -E2BIG; + + lp = NAPI_GRO_CB(p)->last; + pinfo = skb_shinfo(lp); + + if (headlen <= offset) { + skb_frag_t *frag; + skb_frag_t *frag2; + int i = skbinfo->nr_frags; + int nr_frags = pinfo->nr_frags + i; + + if (nr_frags > MAX_SKB_FRAGS) + goto merge; + + offset -= headlen; + pinfo->nr_frags = nr_frags; + skbinfo->nr_frags = 0; + + frag = pinfo->frags + nr_frags; + frag2 = skbinfo->frags + i; + do { + *--frag = *--frag2; + } while (--i); + + skb_frag_off_add(frag, offset); + skb_frag_size_sub(frag, offset); + + /* all fragments truesize : remove (head size + sk_buff) */ + new_truesize = SKB_TRUESIZE(skb_end_offset(skb)); + delta_truesize = skb->truesize - new_truesize; + + skb->truesize = new_truesize; + skb->len -= skb->data_len; + skb->data_len = 0; + + NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE; + goto done; + } else if (skb->head_frag) { + int nr_frags = pinfo->nr_frags; + skb_frag_t *frag = pinfo->frags + nr_frags; + struct page *page = virt_to_head_page(skb->head); + unsigned int first_size = headlen - offset; + unsigned int first_offset; + + if (nr_frags + 1 + skbinfo->nr_frags > MAX_SKB_FRAGS) + goto merge; + + first_offset = skb->data - + (unsigned char *)page_address(page) + + offset; + + pinfo->nr_frags = nr_frags + 1 + skbinfo->nr_frags; + + __skb_frag_set_page(frag, page); + skb_frag_off_set(frag, first_offset); + skb_frag_size_set(frag, first_size); + + memcpy(frag + 1, skbinfo->frags, sizeof(*frag) * skbinfo->nr_frags); + /* We dont need to clear skbinfo->nr_frags here */ + + new_truesize = SKB_DATA_ALIGN(sizeof(struct sk_buff)); + delta_truesize = skb->truesize - new_truesize; + skb->truesize = new_truesize; + NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE_STOLEN_HEAD; + goto done; + } + +merge: + /* sk owenrship - if any - completely transferred to the aggregated packet */ + skb->destructor = NULL; + delta_truesize = skb->truesize; + if (offset > headlen) { + unsigned int eat = offset - headlen; + + skb_frag_off_add(&skbinfo->frags[0], eat); + skb_frag_size_sub(&skbinfo->frags[0], eat); + skb->data_len -= eat; + skb->len -= eat; + offset = headlen; + } + + __skb_pull(skb, offset); + + if (NAPI_GRO_CB(p)->last == p) + skb_shinfo(p)->frag_list = skb; + else + NAPI_GRO_CB(p)->last->next = skb; + NAPI_GRO_CB(p)->last = skb; + __skb_header_release(skb); + lp = p; + +done: + NAPI_GRO_CB(p)->count++; + p->data_len += len; + p->truesize += delta_truesize; + p->len += len; + if (lp != p) { + lp->data_len += len; + lp->truesize += delta_truesize; + lp->len += len; + } + NAPI_GRO_CB(skb)->same_flow = 1; + return 0; +} diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 8560d50c960b..73720bc1daa3 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -64,7 +64,6 @@ #include #include -#include #include #include #include @@ -4272,122 +4271,6 @@ err: } EXPORT_SYMBOL_GPL(skb_segment); -int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb) -{ - struct skb_shared_info *pinfo, *skbinfo = skb_shinfo(skb); - unsigned int offset = skb_gro_offset(skb); - unsigned int headlen = skb_headlen(skb); - unsigned int len = skb_gro_len(skb); - unsigned int delta_truesize; - unsigned int new_truesize; - struct sk_buff *lp; - - if (unlikely(p->len + len >= 65536 || NAPI_GRO_CB(skb)->flush)) - return -E2BIG; - - lp = NAPI_GRO_CB(p)->last; - pinfo = skb_shinfo(lp); - - if (headlen <= offset) { - skb_frag_t *frag; - skb_frag_t *frag2; - int i = skbinfo->nr_frags; - int nr_frags = pinfo->nr_frags + i; - - if (nr_frags > MAX_SKB_FRAGS) - goto merge; - - offset -= headlen; - pinfo->nr_frags = nr_frags; - skbinfo->nr_frags = 0; - - frag = pinfo->frags + nr_frags; - frag2 = skbinfo->frags + i; - do { - *--frag = *--frag2; - } while (--i); - - skb_frag_off_add(frag, offset); - skb_frag_size_sub(frag, offset); - - /* all fragments truesize : remove (head size + sk_buff) */ - new_truesize = SKB_TRUESIZE(skb_end_offset(skb)); - delta_truesize = skb->truesize - new_truesize; - - skb->truesize = new_truesize; - skb->len -= skb->data_len; - skb->data_len = 0; - - NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE; - goto done; - } else if (skb->head_frag) { - int nr_frags = pinfo->nr_frags; - skb_frag_t *frag = pinfo->frags + nr_frags; - struct page *page = virt_to_head_page(skb->head); - unsigned int first_size = headlen - offset; - unsigned int first_offset; - - if (nr_frags + 1 + skbinfo->nr_frags > MAX_SKB_FRAGS) - goto merge; - - first_offset = skb->data - - (unsigned char *)page_address(page) + - offset; - - pinfo->nr_frags = nr_frags + 1 + skbinfo->nr_frags; - - __skb_frag_set_page(frag, page); - skb_frag_off_set(frag, first_offset); - skb_frag_size_set(frag, first_size); - - memcpy(frag + 1, skbinfo->frags, sizeof(*frag) * skbinfo->nr_frags); - /* We dont need to clear skbinfo->nr_frags here */ - - new_truesize = SKB_DATA_ALIGN(sizeof(struct sk_buff)); - delta_truesize = skb->truesize - new_truesize; - skb->truesize = new_truesize; - NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE_STOLEN_HEAD; - goto done; - } - -merge: - /* sk owenrship - if any - completely transferred to the aggregated packet */ - skb->destructor = NULL; - delta_truesize = skb->truesize; - if (offset > headlen) { - unsigned int eat = offset - headlen; - - skb_frag_off_add(&skbinfo->frags[0], eat); - skb_frag_size_sub(&skbinfo->frags[0], eat); - skb->data_len -= eat; - skb->len -= eat; - offset = headlen; - } - - __skb_pull(skb, offset); - - if (NAPI_GRO_CB(p)->last == p) - skb_shinfo(p)->frag_list = skb; - else - NAPI_GRO_CB(p)->last->next = skb; - NAPI_GRO_CB(p)->last = skb; - __skb_header_release(skb); - lp = p; - -done: - NAPI_GRO_CB(p)->count++; - p->data_len += len; - p->truesize += delta_truesize; - p->len += len; - if (lp != p) { - lp->data_len += len; - lp->truesize += delta_truesize; - lp->len += len; - } - NAPI_GRO_CB(skb)->same_flow = 1; - return 0; -} - #ifdef CONFIG_SKB_EXTENSIONS #define SKB_EXT_ALIGN_VALUE 8 #define SKB_EXT_CHUNKSIZEOF(x) (ALIGN((sizeof(x)), SKB_EXT_ALIGN_VALUE) / SKB_EXT_ALIGN_VALUE) From 587652bbdd06ab38a4c1b85e40f933d2cf4a1147 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:05:54 -0800 Subject: [PATCH 0232/2048] net: gro: populate net/core/gro.c Move gro code and data from net/core/dev.c to net/core/gro.c to ease maintenance. gro_normal_list() and gro_normal_one() are inlined because they are called from both files. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 + include/net/gro.h | 22 ++ net/core/dev.c | 668 +------------------------------------- net/core/gro.c | 648 ++++++++++++++++++++++++++++++++++++ 4 files changed, 672 insertions(+), 667 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 93d397db9ec4..31a7e6b27681 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -3657,6 +3657,7 @@ int netif_rx_ni(struct sk_buff *skb); int netif_rx_any_context(struct sk_buff *skb); int netif_receive_skb(struct sk_buff *skb); int netif_receive_skb_core(struct sk_buff *skb); +void netif_receive_skb_list_internal(struct list_head *head); void netif_receive_skb_list(struct list_head *head); gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb); void napi_gro_flush(struct napi_struct *napi, bool flush_old); diff --git a/include/net/gro.h b/include/net/gro.h index f988bf3440f8..d0e7df691a80 100644 --- a/include/net/gro.h +++ b/include/net/gro.h @@ -416,4 +416,26 @@ static inline __wsum ip6_gro_compute_pseudo(struct sk_buff *skb, int proto) int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb); +/* Pass the currently batched GRO_NORMAL SKBs up to the stack. */ +static inline void gro_normal_list(struct napi_struct *napi) +{ + if (!napi->rx_count) + return; + netif_receive_skb_list_internal(&napi->rx_list); + INIT_LIST_HEAD(&napi->rx_list); + napi->rx_count = 0; +} + +/* Queue one GRO_NORMAL SKB up for list processing. If batch size exceeded, + * pass the whole batch up to the stack. + */ +static inline void gro_normal_one(struct napi_struct *napi, struct sk_buff *skb, int segs) +{ + list_add_tail(&skb->list, &napi->rx_list); + napi->rx_count += segs; + if (napi->rx_count >= gro_normal_batch) + gro_normal_list(napi); +} + + #endif /* _NET_IPV6_GRO_H */ diff --git a/net/core/dev.c b/net/core/dev.c index 15ac064b5562..92c9258cbf28 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -153,16 +153,10 @@ #include "net-sysfs.h" -#define MAX_GRO_SKBS 8 - -/* This should be increased if a protocol with a bigger head is added. */ -#define GRO_MAX_HEAD (MAX_HEADER + 128) static DEFINE_SPINLOCK(ptype_lock); -static DEFINE_SPINLOCK(offload_lock); struct list_head ptype_base[PTYPE_HASH_SIZE] __read_mostly; struct list_head ptype_all __read_mostly; /* Taps */ -static struct list_head offload_base __read_mostly; static int netif_rx_internal(struct sk_buff *skb); static int call_netdevice_notifiers_info(unsigned long val, @@ -604,84 +598,6 @@ void dev_remove_pack(struct packet_type *pt) EXPORT_SYMBOL(dev_remove_pack); -/** - * dev_add_offload - register offload handlers - * @po: protocol offload declaration - * - * Add protocol offload handlers to the networking stack. The passed - * &proto_offload is linked into kernel lists and may not be freed until - * it has been removed from the kernel lists. - * - * This call does not sleep therefore it can not - * guarantee all CPU's that are in middle of receiving packets - * will see the new offload handlers (until the next received packet). - */ -void dev_add_offload(struct packet_offload *po) -{ - struct packet_offload *elem; - - spin_lock(&offload_lock); - list_for_each_entry(elem, &offload_base, list) { - if (po->priority < elem->priority) - break; - } - list_add_rcu(&po->list, elem->list.prev); - spin_unlock(&offload_lock); -} -EXPORT_SYMBOL(dev_add_offload); - -/** - * __dev_remove_offload - remove offload handler - * @po: packet offload declaration - * - * Remove a protocol offload handler that was previously added to the - * kernel offload handlers by dev_add_offload(). The passed &offload_type - * is removed from the kernel lists and can be freed or reused once this - * function returns. - * - * The packet type might still be in use by receivers - * and must not be freed until after all the CPU's have gone - * through a quiescent state. - */ -static void __dev_remove_offload(struct packet_offload *po) -{ - struct list_head *head = &offload_base; - struct packet_offload *po1; - - spin_lock(&offload_lock); - - list_for_each_entry(po1, head, list) { - if (po == po1) { - list_del_rcu(&po->list); - goto out; - } - } - - pr_warn("dev_remove_offload: %p not found\n", po); -out: - spin_unlock(&offload_lock); -} - -/** - * dev_remove_offload - remove packet offload handler - * @po: packet offload declaration - * - * Remove a packet offload handler that was previously added to the kernel - * offload handlers by dev_add_offload(). The passed &offload_type is - * removed from the kernel lists and can be freed or reused once this - * function returns. - * - * This call sleeps to guarantee that no CPU is looking at the packet - * type after return. - */ -void dev_remove_offload(struct packet_offload *po) -{ - __dev_remove_offload(po); - - synchronize_net(); -} -EXPORT_SYMBOL(dev_remove_offload); - /******************************************************************************* * * Device Interface Subroutines @@ -3315,40 +3231,6 @@ __be16 skb_network_protocol(struct sk_buff *skb, int *depth) return __vlan_get_protocol(skb, type, depth); } -/** - * skb_mac_gso_segment - mac layer segmentation handler. - * @skb: buffer to segment - * @features: features for the output path (see dev->features) - */ -struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb, - netdev_features_t features) -{ - struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT); - struct packet_offload *ptype; - int vlan_depth = skb->mac_len; - __be16 type = skb_network_protocol(skb, &vlan_depth); - - if (unlikely(!type)) - return ERR_PTR(-EINVAL); - - __skb_pull(skb, vlan_depth); - - rcu_read_lock(); - list_for_each_entry_rcu(ptype, &offload_base, list) { - if (ptype->type == type && ptype->callbacks.gso_segment) { - segs = ptype->callbacks.gso_segment(skb, features); - break; - } - } - rcu_read_unlock(); - - __skb_push(skb, skb->data - skb_mac_header(skb)); - - return segs; -} -EXPORT_SYMBOL(skb_mac_gso_segment); - - /* openvswitch calls this on rx path, so we need a different check. */ static inline bool skb_needs_check(struct sk_buff *skb, bool tx_path) @@ -4320,8 +4202,6 @@ int dev_weight_rx_bias __read_mostly = 1; /* bias for backlog weight */ int dev_weight_tx_bias __read_mostly = 1; /* bias for output_queue quota */ int dev_rx_weight __read_mostly = 64; int dev_tx_weight __read_mostly = 64; -/* Maximum number of GRO_NORMAL skbs to batch up for list-RX */ -int gro_normal_batch __read_mostly = 8; /* Called with irq disabled */ static inline void ____napi_schedule(struct softnet_data *sd, @@ -5664,7 +5544,7 @@ static int netif_receive_skb_internal(struct sk_buff *skb) return ret; } -static void netif_receive_skb_list_internal(struct list_head *head) +void netif_receive_skb_list_internal(struct list_head *head) { struct sk_buff *skb, *next; struct list_head sublist; @@ -5842,550 +5722,6 @@ static void flush_all_backlogs(void) cpus_read_unlock(); } -/* Pass the currently batched GRO_NORMAL SKBs up to the stack. */ -static void gro_normal_list(struct napi_struct *napi) -{ - if (!napi->rx_count) - return; - netif_receive_skb_list_internal(&napi->rx_list); - INIT_LIST_HEAD(&napi->rx_list); - napi->rx_count = 0; -} - -/* Queue one GRO_NORMAL SKB up for list processing. If batch size exceeded, - * pass the whole batch up to the stack. - */ -static void gro_normal_one(struct napi_struct *napi, struct sk_buff *skb, int segs) -{ - list_add_tail(&skb->list, &napi->rx_list); - napi->rx_count += segs; - if (napi->rx_count >= gro_normal_batch) - gro_normal_list(napi); -} - -static void napi_gro_complete(struct napi_struct *napi, struct sk_buff *skb) -{ - struct packet_offload *ptype; - __be16 type = skb->protocol; - struct list_head *head = &offload_base; - int err = -ENOENT; - - BUILD_BUG_ON(sizeof(struct napi_gro_cb) > sizeof(skb->cb)); - - if (NAPI_GRO_CB(skb)->count == 1) { - skb_shinfo(skb)->gso_size = 0; - goto out; - } - - rcu_read_lock(); - list_for_each_entry_rcu(ptype, head, list) { - if (ptype->type != type || !ptype->callbacks.gro_complete) - continue; - - err = INDIRECT_CALL_INET(ptype->callbacks.gro_complete, - ipv6_gro_complete, inet_gro_complete, - skb, 0); - break; - } - rcu_read_unlock(); - - if (err) { - WARN_ON(&ptype->list == head); - kfree_skb(skb); - return; - } - -out: - gro_normal_one(napi, skb, NAPI_GRO_CB(skb)->count); -} - -static void __napi_gro_flush_chain(struct napi_struct *napi, u32 index, - bool flush_old) -{ - struct list_head *head = &napi->gro_hash[index].list; - struct sk_buff *skb, *p; - - list_for_each_entry_safe_reverse(skb, p, head, list) { - if (flush_old && NAPI_GRO_CB(skb)->age == jiffies) - return; - skb_list_del_init(skb); - napi_gro_complete(napi, skb); - napi->gro_hash[index].count--; - } - - if (!napi->gro_hash[index].count) - __clear_bit(index, &napi->gro_bitmask); -} - -/* napi->gro_hash[].list contains packets ordered by age. - * youngest packets at the head of it. - * Complete skbs in reverse order to reduce latencies. - */ -void napi_gro_flush(struct napi_struct *napi, bool flush_old) -{ - unsigned long bitmask = napi->gro_bitmask; - unsigned int i, base = ~0U; - - while ((i = ffs(bitmask)) != 0) { - bitmask >>= i; - base += i; - __napi_gro_flush_chain(napi, base, flush_old); - } -} -EXPORT_SYMBOL(napi_gro_flush); - -static void gro_list_prepare(const struct list_head *head, - const struct sk_buff *skb) -{ - unsigned int maclen = skb->dev->hard_header_len; - u32 hash = skb_get_hash_raw(skb); - struct sk_buff *p; - - list_for_each_entry(p, head, list) { - unsigned long diffs; - - NAPI_GRO_CB(p)->flush = 0; - - if (hash != skb_get_hash_raw(p)) { - NAPI_GRO_CB(p)->same_flow = 0; - continue; - } - - diffs = (unsigned long)p->dev ^ (unsigned long)skb->dev; - diffs |= skb_vlan_tag_present(p) ^ skb_vlan_tag_present(skb); - if (skb_vlan_tag_present(p)) - diffs |= skb_vlan_tag_get(p) ^ skb_vlan_tag_get(skb); - diffs |= skb_metadata_differs(p, skb); - if (maclen == ETH_HLEN) - diffs |= compare_ether_header(skb_mac_header(p), - skb_mac_header(skb)); - else if (!diffs) - diffs = memcmp(skb_mac_header(p), - skb_mac_header(skb), - maclen); - - /* in most common scenarions 'slow_gro' is 0 - * otherwise we are already on some slower paths - * either skip all the infrequent tests altogether or - * avoid trying too hard to skip each of them individually - */ - if (!diffs && unlikely(skb->slow_gro | p->slow_gro)) { -#if IS_ENABLED(CONFIG_SKB_EXTENSIONS) && IS_ENABLED(CONFIG_NET_TC_SKB_EXT) - struct tc_skb_ext *skb_ext; - struct tc_skb_ext *p_ext; -#endif - - diffs |= p->sk != skb->sk; - diffs |= skb_metadata_dst_cmp(p, skb); - diffs |= skb_get_nfct(p) ^ skb_get_nfct(skb); - -#if IS_ENABLED(CONFIG_SKB_EXTENSIONS) && IS_ENABLED(CONFIG_NET_TC_SKB_EXT) - skb_ext = skb_ext_find(skb, TC_SKB_EXT); - p_ext = skb_ext_find(p, TC_SKB_EXT); - - diffs |= (!!p_ext) ^ (!!skb_ext); - if (!diffs && unlikely(skb_ext)) - diffs |= p_ext->chain ^ skb_ext->chain; -#endif - } - - NAPI_GRO_CB(p)->same_flow = !diffs; - } -} - -static inline void skb_gro_reset_offset(struct sk_buff *skb, u32 nhoff) -{ - const struct skb_shared_info *pinfo = skb_shinfo(skb); - const skb_frag_t *frag0 = &pinfo->frags[0]; - - NAPI_GRO_CB(skb)->data_offset = 0; - NAPI_GRO_CB(skb)->frag0 = NULL; - NAPI_GRO_CB(skb)->frag0_len = 0; - - if (!skb_headlen(skb) && pinfo->nr_frags && - !PageHighMem(skb_frag_page(frag0)) && - (!NET_IP_ALIGN || !((skb_frag_off(frag0) + nhoff) & 3))) { - NAPI_GRO_CB(skb)->frag0 = skb_frag_address(frag0); - NAPI_GRO_CB(skb)->frag0_len = min_t(unsigned int, - skb_frag_size(frag0), - skb->end - skb->tail); - } -} - -static void gro_pull_from_frag0(struct sk_buff *skb, int grow) -{ - struct skb_shared_info *pinfo = skb_shinfo(skb); - - BUG_ON(skb->end - skb->tail < grow); - - memcpy(skb_tail_pointer(skb), NAPI_GRO_CB(skb)->frag0, grow); - - skb->data_len -= grow; - skb->tail += grow; - - skb_frag_off_add(&pinfo->frags[0], grow); - skb_frag_size_sub(&pinfo->frags[0], grow); - - if (unlikely(!skb_frag_size(&pinfo->frags[0]))) { - skb_frag_unref(skb, 0); - memmove(pinfo->frags, pinfo->frags + 1, - --pinfo->nr_frags * sizeof(pinfo->frags[0])); - } -} - -static void gro_flush_oldest(struct napi_struct *napi, struct list_head *head) -{ - struct sk_buff *oldest; - - oldest = list_last_entry(head, struct sk_buff, list); - - /* We are called with head length >= MAX_GRO_SKBS, so this is - * impossible. - */ - if (WARN_ON_ONCE(!oldest)) - return; - - /* Do not adjust napi->gro_hash[].count, caller is adding a new - * SKB to the chain. - */ - skb_list_del_init(oldest); - napi_gro_complete(napi, oldest); -} - -static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb) -{ - u32 bucket = skb_get_hash_raw(skb) & (GRO_HASH_BUCKETS - 1); - struct gro_list *gro_list = &napi->gro_hash[bucket]; - struct list_head *head = &offload_base; - struct packet_offload *ptype; - __be16 type = skb->protocol; - struct sk_buff *pp = NULL; - enum gro_result ret; - int same_flow; - int grow; - - if (netif_elide_gro(skb->dev)) - goto normal; - - gro_list_prepare(&gro_list->list, skb); - - rcu_read_lock(); - list_for_each_entry_rcu(ptype, head, list) { - if (ptype->type != type || !ptype->callbacks.gro_receive) - continue; - - skb_set_network_header(skb, skb_gro_offset(skb)); - skb_reset_mac_len(skb); - NAPI_GRO_CB(skb)->same_flow = 0; - NAPI_GRO_CB(skb)->flush = skb_is_gso(skb) || skb_has_frag_list(skb); - NAPI_GRO_CB(skb)->free = 0; - NAPI_GRO_CB(skb)->encap_mark = 0; - NAPI_GRO_CB(skb)->recursion_counter = 0; - NAPI_GRO_CB(skb)->is_fou = 0; - NAPI_GRO_CB(skb)->is_atomic = 1; - NAPI_GRO_CB(skb)->gro_remcsum_start = 0; - - /* Setup for GRO checksum validation */ - switch (skb->ip_summed) { - case CHECKSUM_COMPLETE: - NAPI_GRO_CB(skb)->csum = skb->csum; - NAPI_GRO_CB(skb)->csum_valid = 1; - NAPI_GRO_CB(skb)->csum_cnt = 0; - break; - case CHECKSUM_UNNECESSARY: - NAPI_GRO_CB(skb)->csum_cnt = skb->csum_level + 1; - NAPI_GRO_CB(skb)->csum_valid = 0; - break; - default: - NAPI_GRO_CB(skb)->csum_cnt = 0; - NAPI_GRO_CB(skb)->csum_valid = 0; - } - - pp = INDIRECT_CALL_INET(ptype->callbacks.gro_receive, - ipv6_gro_receive, inet_gro_receive, - &gro_list->list, skb); - break; - } - rcu_read_unlock(); - - if (&ptype->list == head) - goto normal; - - if (PTR_ERR(pp) == -EINPROGRESS) { - ret = GRO_CONSUMED; - goto ok; - } - - same_flow = NAPI_GRO_CB(skb)->same_flow; - ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED; - - if (pp) { - skb_list_del_init(pp); - napi_gro_complete(napi, pp); - gro_list->count--; - } - - if (same_flow) - goto ok; - - if (NAPI_GRO_CB(skb)->flush) - goto normal; - - if (unlikely(gro_list->count >= MAX_GRO_SKBS)) - gro_flush_oldest(napi, &gro_list->list); - else - gro_list->count++; - - NAPI_GRO_CB(skb)->count = 1; - NAPI_GRO_CB(skb)->age = jiffies; - NAPI_GRO_CB(skb)->last = skb; - skb_shinfo(skb)->gso_size = skb_gro_len(skb); - list_add(&skb->list, &gro_list->list); - ret = GRO_HELD; - -pull: - grow = skb_gro_offset(skb) - skb_headlen(skb); - if (grow > 0) - gro_pull_from_frag0(skb, grow); -ok: - if (gro_list->count) { - if (!test_bit(bucket, &napi->gro_bitmask)) - __set_bit(bucket, &napi->gro_bitmask); - } else if (test_bit(bucket, &napi->gro_bitmask)) { - __clear_bit(bucket, &napi->gro_bitmask); - } - - return ret; - -normal: - ret = GRO_NORMAL; - goto pull; -} - -struct packet_offload *gro_find_receive_by_type(__be16 type) -{ - struct list_head *offload_head = &offload_base; - struct packet_offload *ptype; - - list_for_each_entry_rcu(ptype, offload_head, list) { - if (ptype->type != type || !ptype->callbacks.gro_receive) - continue; - return ptype; - } - return NULL; -} -EXPORT_SYMBOL(gro_find_receive_by_type); - -struct packet_offload *gro_find_complete_by_type(__be16 type) -{ - struct list_head *offload_head = &offload_base; - struct packet_offload *ptype; - - list_for_each_entry_rcu(ptype, offload_head, list) { - if (ptype->type != type || !ptype->callbacks.gro_complete) - continue; - return ptype; - } - return NULL; -} -EXPORT_SYMBOL(gro_find_complete_by_type); - -static gro_result_t napi_skb_finish(struct napi_struct *napi, - struct sk_buff *skb, - gro_result_t ret) -{ - switch (ret) { - case GRO_NORMAL: - gro_normal_one(napi, skb, 1); - break; - - case GRO_MERGED_FREE: - if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD) - napi_skb_free_stolen_head(skb); - else if (skb->fclone != SKB_FCLONE_UNAVAILABLE) - __kfree_skb(skb); - else - __kfree_skb_defer(skb); - break; - - case GRO_HELD: - case GRO_MERGED: - case GRO_CONSUMED: - break; - } - - return ret; -} - -gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb) -{ - gro_result_t ret; - - skb_mark_napi_id(skb, napi); - trace_napi_gro_receive_entry(skb); - - skb_gro_reset_offset(skb, 0); - - ret = napi_skb_finish(napi, skb, dev_gro_receive(napi, skb)); - trace_napi_gro_receive_exit(ret); - - return ret; -} -EXPORT_SYMBOL(napi_gro_receive); - -static void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb) -{ - if (unlikely(skb->pfmemalloc)) { - consume_skb(skb); - return; - } - __skb_pull(skb, skb_headlen(skb)); - /* restore the reserve we had after netdev_alloc_skb_ip_align() */ - skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN - skb_headroom(skb)); - __vlan_hwaccel_clear_tag(skb); - skb->dev = napi->dev; - skb->skb_iif = 0; - - /* eth_type_trans() assumes pkt_type is PACKET_HOST */ - skb->pkt_type = PACKET_HOST; - - skb->encapsulation = 0; - skb_shinfo(skb)->gso_type = 0; - skb->truesize = SKB_TRUESIZE(skb_end_offset(skb)); - if (unlikely(skb->slow_gro)) { - skb_orphan(skb); - skb_ext_reset(skb); - nf_reset_ct(skb); - skb->slow_gro = 0; - } - - napi->skb = skb; -} - -struct sk_buff *napi_get_frags(struct napi_struct *napi) -{ - struct sk_buff *skb = napi->skb; - - if (!skb) { - skb = napi_alloc_skb(napi, GRO_MAX_HEAD); - if (skb) { - napi->skb = skb; - skb_mark_napi_id(skb, napi); - } - } - return skb; -} -EXPORT_SYMBOL(napi_get_frags); - -static gro_result_t napi_frags_finish(struct napi_struct *napi, - struct sk_buff *skb, - gro_result_t ret) -{ - switch (ret) { - case GRO_NORMAL: - case GRO_HELD: - __skb_push(skb, ETH_HLEN); - skb->protocol = eth_type_trans(skb, skb->dev); - if (ret == GRO_NORMAL) - gro_normal_one(napi, skb, 1); - break; - - case GRO_MERGED_FREE: - if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD) - napi_skb_free_stolen_head(skb); - else - napi_reuse_skb(napi, skb); - break; - - case GRO_MERGED: - case GRO_CONSUMED: - break; - } - - return ret; -} - -/* Upper GRO stack assumes network header starts at gro_offset=0 - * Drivers could call both napi_gro_frags() and napi_gro_receive() - * We copy ethernet header into skb->data to have a common layout. - */ -static struct sk_buff *napi_frags_skb(struct napi_struct *napi) -{ - struct sk_buff *skb = napi->skb; - const struct ethhdr *eth; - unsigned int hlen = sizeof(*eth); - - napi->skb = NULL; - - skb_reset_mac_header(skb); - skb_gro_reset_offset(skb, hlen); - - if (unlikely(skb_gro_header_hard(skb, hlen))) { - eth = skb_gro_header_slow(skb, hlen, 0); - if (unlikely(!eth)) { - net_warn_ratelimited("%s: dropping impossible skb from %s\n", - __func__, napi->dev->name); - napi_reuse_skb(napi, skb); - return NULL; - } - } else { - eth = (const struct ethhdr *)skb->data; - gro_pull_from_frag0(skb, hlen); - NAPI_GRO_CB(skb)->frag0 += hlen; - NAPI_GRO_CB(skb)->frag0_len -= hlen; - } - __skb_pull(skb, hlen); - - /* - * This works because the only protocols we care about don't require - * special handling. - * We'll fix it up properly in napi_frags_finish() - */ - skb->protocol = eth->h_proto; - - return skb; -} - -gro_result_t napi_gro_frags(struct napi_struct *napi) -{ - gro_result_t ret; - struct sk_buff *skb = napi_frags_skb(napi); - - trace_napi_gro_frags_entry(skb); - - ret = napi_frags_finish(napi, skb, dev_gro_receive(napi, skb)); - trace_napi_gro_frags_exit(ret); - - return ret; -} -EXPORT_SYMBOL(napi_gro_frags); - -/* Compute the checksum from gro_offset and return the folded value - * after adding in any pseudo checksum. - */ -__sum16 __skb_gro_checksum_complete(struct sk_buff *skb) -{ - __wsum wsum; - __sum16 sum; - - wsum = skb_checksum(skb, skb_gro_offset(skb), skb_gro_len(skb), 0); - - /* NAPI_GRO_CB(skb)->csum holds pseudo checksum */ - sum = csum_fold(csum_add(NAPI_GRO_CB(skb)->csum, wsum)); - /* See comments in __skb_checksum_complete(). */ - if (likely(!sum)) { - if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE) && - !skb->csum_complete_sw) - netdev_rx_csum_fault(skb->dev, skb); - } - - NAPI_GRO_CB(skb)->csum = wsum; - NAPI_GRO_CB(skb)->csum_valid = 1; - - return sum; -} -EXPORT_SYMBOL(__skb_gro_checksum_complete); - static void net_rps_send_ipi(struct softnet_data *remsd) { #ifdef CONFIG_RPS @@ -11640,8 +10976,6 @@ static int __init net_dev_init(void) for (i = 0; i < PTYPE_HASH_SIZE; i++) INIT_LIST_HEAD(&ptype_base[i]); - INIT_LIST_HEAD(&offload_base); - if (register_pernet_subsys(&netdev_net_ops)) goto out; diff --git a/net/core/gro.c b/net/core/gro.c index 91a74c4da9ff..8ec8b44596da 100644 --- a/net/core/gro.c +++ b/net/core/gro.c @@ -1,5 +1,129 @@ // SPDX-License-Identifier: GPL-2.0-or-later #include +#include +#include +#include + +#define MAX_GRO_SKBS 8 + +/* This should be increased if a protocol with a bigger head is added. */ +#define GRO_MAX_HEAD (MAX_HEADER + 128) + +static DEFINE_SPINLOCK(offload_lock); +static struct list_head offload_base __read_mostly = LIST_HEAD_INIT(offload_base); +/* Maximum number of GRO_NORMAL skbs to batch up for list-RX */ +int gro_normal_batch __read_mostly = 8; + +/** + * dev_add_offload - register offload handlers + * @po: protocol offload declaration + * + * Add protocol offload handlers to the networking stack. The passed + * &proto_offload is linked into kernel lists and may not be freed until + * it has been removed from the kernel lists. + * + * This call does not sleep therefore it can not + * guarantee all CPU's that are in middle of receiving packets + * will see the new offload handlers (until the next received packet). + */ +void dev_add_offload(struct packet_offload *po) +{ + struct packet_offload *elem; + + spin_lock(&offload_lock); + list_for_each_entry(elem, &offload_base, list) { + if (po->priority < elem->priority) + break; + } + list_add_rcu(&po->list, elem->list.prev); + spin_unlock(&offload_lock); +} +EXPORT_SYMBOL(dev_add_offload); + +/** + * __dev_remove_offload - remove offload handler + * @po: packet offload declaration + * + * Remove a protocol offload handler that was previously added to the + * kernel offload handlers by dev_add_offload(). The passed &offload_type + * is removed from the kernel lists and can be freed or reused once this + * function returns. + * + * The packet type might still be in use by receivers + * and must not be freed until after all the CPU's have gone + * through a quiescent state. + */ +static void __dev_remove_offload(struct packet_offload *po) +{ + struct list_head *head = &offload_base; + struct packet_offload *po1; + + spin_lock(&offload_lock); + + list_for_each_entry(po1, head, list) { + if (po == po1) { + list_del_rcu(&po->list); + goto out; + } + } + + pr_warn("dev_remove_offload: %p not found\n", po); +out: + spin_unlock(&offload_lock); +} + +/** + * dev_remove_offload - remove packet offload handler + * @po: packet offload declaration + * + * Remove a packet offload handler that was previously added to the kernel + * offload handlers by dev_add_offload(). The passed &offload_type is + * removed from the kernel lists and can be freed or reused once this + * function returns. + * + * This call sleeps to guarantee that no CPU is looking at the packet + * type after return. + */ +void dev_remove_offload(struct packet_offload *po) +{ + __dev_remove_offload(po); + + synchronize_net(); +} +EXPORT_SYMBOL(dev_remove_offload); + +/** + * skb_mac_gso_segment - mac layer segmentation handler. + * @skb: buffer to segment + * @features: features for the output path (see dev->features) + */ +struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb, + netdev_features_t features) +{ + struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT); + struct packet_offload *ptype; + int vlan_depth = skb->mac_len; + __be16 type = skb_network_protocol(skb, &vlan_depth); + + if (unlikely(!type)) + return ERR_PTR(-EINVAL); + + __skb_pull(skb, vlan_depth); + + rcu_read_lock(); + list_for_each_entry_rcu(ptype, &offload_base, list) { + if (ptype->type == type && ptype->callbacks.gso_segment) { + segs = ptype->callbacks.gso_segment(skb, features); + break; + } + } + rcu_read_unlock(); + + __skb_push(skb, skb->data - skb_mac_header(skb)); + + return segs; +} +EXPORT_SYMBOL(skb_mac_gso_segment); int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb) { @@ -116,3 +240,527 @@ done: NAPI_GRO_CB(skb)->same_flow = 1; return 0; } + + +static void napi_gro_complete(struct napi_struct *napi, struct sk_buff *skb) +{ + struct packet_offload *ptype; + __be16 type = skb->protocol; + struct list_head *head = &offload_base; + int err = -ENOENT; + + BUILD_BUG_ON(sizeof(struct napi_gro_cb) > sizeof(skb->cb)); + + if (NAPI_GRO_CB(skb)->count == 1) { + skb_shinfo(skb)->gso_size = 0; + goto out; + } + + rcu_read_lock(); + list_for_each_entry_rcu(ptype, head, list) { + if (ptype->type != type || !ptype->callbacks.gro_complete) + continue; + + err = INDIRECT_CALL_INET(ptype->callbacks.gro_complete, + ipv6_gro_complete, inet_gro_complete, + skb, 0); + break; + } + rcu_read_unlock(); + + if (err) { + WARN_ON(&ptype->list == head); + kfree_skb(skb); + return; + } + +out: + gro_normal_one(napi, skb, NAPI_GRO_CB(skb)->count); +} + +static void __napi_gro_flush_chain(struct napi_struct *napi, u32 index, + bool flush_old) +{ + struct list_head *head = &napi->gro_hash[index].list; + struct sk_buff *skb, *p; + + list_for_each_entry_safe_reverse(skb, p, head, list) { + if (flush_old && NAPI_GRO_CB(skb)->age == jiffies) + return; + skb_list_del_init(skb); + napi_gro_complete(napi, skb); + napi->gro_hash[index].count--; + } + + if (!napi->gro_hash[index].count) + __clear_bit(index, &napi->gro_bitmask); +} + +/* napi->gro_hash[].list contains packets ordered by age. + * youngest packets at the head of it. + * Complete skbs in reverse order to reduce latencies. + */ +void napi_gro_flush(struct napi_struct *napi, bool flush_old) +{ + unsigned long bitmask = napi->gro_bitmask; + unsigned int i, base = ~0U; + + while ((i = ffs(bitmask)) != 0) { + bitmask >>= i; + base += i; + __napi_gro_flush_chain(napi, base, flush_old); + } +} +EXPORT_SYMBOL(napi_gro_flush); + +static void gro_list_prepare(const struct list_head *head, + const struct sk_buff *skb) +{ + unsigned int maclen = skb->dev->hard_header_len; + u32 hash = skb_get_hash_raw(skb); + struct sk_buff *p; + + list_for_each_entry(p, head, list) { + unsigned long diffs; + + NAPI_GRO_CB(p)->flush = 0; + + if (hash != skb_get_hash_raw(p)) { + NAPI_GRO_CB(p)->same_flow = 0; + continue; + } + + diffs = (unsigned long)p->dev ^ (unsigned long)skb->dev; + diffs |= skb_vlan_tag_present(p) ^ skb_vlan_tag_present(skb); + if (skb_vlan_tag_present(p)) + diffs |= skb_vlan_tag_get(p) ^ skb_vlan_tag_get(skb); + diffs |= skb_metadata_differs(p, skb); + if (maclen == ETH_HLEN) + diffs |= compare_ether_header(skb_mac_header(p), + skb_mac_header(skb)); + else if (!diffs) + diffs = memcmp(skb_mac_header(p), + skb_mac_header(skb), + maclen); + + /* in most common scenarions 'slow_gro' is 0 + * otherwise we are already on some slower paths + * either skip all the infrequent tests altogether or + * avoid trying too hard to skip each of them individually + */ + if (!diffs && unlikely(skb->slow_gro | p->slow_gro)) { +#if IS_ENABLED(CONFIG_SKB_EXTENSIONS) && IS_ENABLED(CONFIG_NET_TC_SKB_EXT) + struct tc_skb_ext *skb_ext; + struct tc_skb_ext *p_ext; +#endif + + diffs |= p->sk != skb->sk; + diffs |= skb_metadata_dst_cmp(p, skb); + diffs |= skb_get_nfct(p) ^ skb_get_nfct(skb); + +#if IS_ENABLED(CONFIG_SKB_EXTENSIONS) && IS_ENABLED(CONFIG_NET_TC_SKB_EXT) + skb_ext = skb_ext_find(skb, TC_SKB_EXT); + p_ext = skb_ext_find(p, TC_SKB_EXT); + + diffs |= (!!p_ext) ^ (!!skb_ext); + if (!diffs && unlikely(skb_ext)) + diffs |= p_ext->chain ^ skb_ext->chain; +#endif + } + + NAPI_GRO_CB(p)->same_flow = !diffs; + } +} + +static inline void skb_gro_reset_offset(struct sk_buff *skb, u32 nhoff) +{ + const struct skb_shared_info *pinfo = skb_shinfo(skb); + const skb_frag_t *frag0 = &pinfo->frags[0]; + + NAPI_GRO_CB(skb)->data_offset = 0; + NAPI_GRO_CB(skb)->frag0 = NULL; + NAPI_GRO_CB(skb)->frag0_len = 0; + + if (!skb_headlen(skb) && pinfo->nr_frags && + !PageHighMem(skb_frag_page(frag0)) && + (!NET_IP_ALIGN || !((skb_frag_off(frag0) + nhoff) & 3))) { + NAPI_GRO_CB(skb)->frag0 = skb_frag_address(frag0); + NAPI_GRO_CB(skb)->frag0_len = min_t(unsigned int, + skb_frag_size(frag0), + skb->end - skb->tail); + } +} + +static void gro_pull_from_frag0(struct sk_buff *skb, int grow) +{ + struct skb_shared_info *pinfo = skb_shinfo(skb); + + BUG_ON(skb->end - skb->tail < grow); + + memcpy(skb_tail_pointer(skb), NAPI_GRO_CB(skb)->frag0, grow); + + skb->data_len -= grow; + skb->tail += grow; + + skb_frag_off_add(&pinfo->frags[0], grow); + skb_frag_size_sub(&pinfo->frags[0], grow); + + if (unlikely(!skb_frag_size(&pinfo->frags[0]))) { + skb_frag_unref(skb, 0); + memmove(pinfo->frags, pinfo->frags + 1, + --pinfo->nr_frags * sizeof(pinfo->frags[0])); + } +} + +static void gro_flush_oldest(struct napi_struct *napi, struct list_head *head) +{ + struct sk_buff *oldest; + + oldest = list_last_entry(head, struct sk_buff, list); + + /* We are called with head length >= MAX_GRO_SKBS, so this is + * impossible. + */ + if (WARN_ON_ONCE(!oldest)) + return; + + /* Do not adjust napi->gro_hash[].count, caller is adding a new + * SKB to the chain. + */ + skb_list_del_init(oldest); + napi_gro_complete(napi, oldest); +} + +static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb) +{ + u32 bucket = skb_get_hash_raw(skb) & (GRO_HASH_BUCKETS - 1); + struct gro_list *gro_list = &napi->gro_hash[bucket]; + struct list_head *head = &offload_base; + struct packet_offload *ptype; + __be16 type = skb->protocol; + struct sk_buff *pp = NULL; + enum gro_result ret; + int same_flow; + int grow; + + if (netif_elide_gro(skb->dev)) + goto normal; + + gro_list_prepare(&gro_list->list, skb); + + rcu_read_lock(); + list_for_each_entry_rcu(ptype, head, list) { + if (ptype->type != type || !ptype->callbacks.gro_receive) + continue; + + skb_set_network_header(skb, skb_gro_offset(skb)); + skb_reset_mac_len(skb); + NAPI_GRO_CB(skb)->same_flow = 0; + NAPI_GRO_CB(skb)->flush = skb_is_gso(skb) || skb_has_frag_list(skb); + NAPI_GRO_CB(skb)->free = 0; + NAPI_GRO_CB(skb)->encap_mark = 0; + NAPI_GRO_CB(skb)->recursion_counter = 0; + NAPI_GRO_CB(skb)->is_fou = 0; + NAPI_GRO_CB(skb)->is_atomic = 1; + NAPI_GRO_CB(skb)->gro_remcsum_start = 0; + + /* Setup for GRO checksum validation */ + switch (skb->ip_summed) { + case CHECKSUM_COMPLETE: + NAPI_GRO_CB(skb)->csum = skb->csum; + NAPI_GRO_CB(skb)->csum_valid = 1; + NAPI_GRO_CB(skb)->csum_cnt = 0; + break; + case CHECKSUM_UNNECESSARY: + NAPI_GRO_CB(skb)->csum_cnt = skb->csum_level + 1; + NAPI_GRO_CB(skb)->csum_valid = 0; + break; + default: + NAPI_GRO_CB(skb)->csum_cnt = 0; + NAPI_GRO_CB(skb)->csum_valid = 0; + } + + pp = INDIRECT_CALL_INET(ptype->callbacks.gro_receive, + ipv6_gro_receive, inet_gro_receive, + &gro_list->list, skb); + break; + } + rcu_read_unlock(); + + if (&ptype->list == head) + goto normal; + + if (PTR_ERR(pp) == -EINPROGRESS) { + ret = GRO_CONSUMED; + goto ok; + } + + same_flow = NAPI_GRO_CB(skb)->same_flow; + ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED; + + if (pp) { + skb_list_del_init(pp); + napi_gro_complete(napi, pp); + gro_list->count--; + } + + if (same_flow) + goto ok; + + if (NAPI_GRO_CB(skb)->flush) + goto normal; + + if (unlikely(gro_list->count >= MAX_GRO_SKBS)) + gro_flush_oldest(napi, &gro_list->list); + else + gro_list->count++; + + NAPI_GRO_CB(skb)->count = 1; + NAPI_GRO_CB(skb)->age = jiffies; + NAPI_GRO_CB(skb)->last = skb; + skb_shinfo(skb)->gso_size = skb_gro_len(skb); + list_add(&skb->list, &gro_list->list); + ret = GRO_HELD; + +pull: + grow = skb_gro_offset(skb) - skb_headlen(skb); + if (grow > 0) + gro_pull_from_frag0(skb, grow); +ok: + if (gro_list->count) { + if (!test_bit(bucket, &napi->gro_bitmask)) + __set_bit(bucket, &napi->gro_bitmask); + } else if (test_bit(bucket, &napi->gro_bitmask)) { + __clear_bit(bucket, &napi->gro_bitmask); + } + + return ret; + +normal: + ret = GRO_NORMAL; + goto pull; +} + +struct packet_offload *gro_find_receive_by_type(__be16 type) +{ + struct list_head *offload_head = &offload_base; + struct packet_offload *ptype; + + list_for_each_entry_rcu(ptype, offload_head, list) { + if (ptype->type != type || !ptype->callbacks.gro_receive) + continue; + return ptype; + } + return NULL; +} +EXPORT_SYMBOL(gro_find_receive_by_type); + +struct packet_offload *gro_find_complete_by_type(__be16 type) +{ + struct list_head *offload_head = &offload_base; + struct packet_offload *ptype; + + list_for_each_entry_rcu(ptype, offload_head, list) { + if (ptype->type != type || !ptype->callbacks.gro_complete) + continue; + return ptype; + } + return NULL; +} +EXPORT_SYMBOL(gro_find_complete_by_type); + +static gro_result_t napi_skb_finish(struct napi_struct *napi, + struct sk_buff *skb, + gro_result_t ret) +{ + switch (ret) { + case GRO_NORMAL: + gro_normal_one(napi, skb, 1); + break; + + case GRO_MERGED_FREE: + if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD) + napi_skb_free_stolen_head(skb); + else if (skb->fclone != SKB_FCLONE_UNAVAILABLE) + __kfree_skb(skb); + else + __kfree_skb_defer(skb); + break; + + case GRO_HELD: + case GRO_MERGED: + case GRO_CONSUMED: + break; + } + + return ret; +} + +gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb) +{ + gro_result_t ret; + + skb_mark_napi_id(skb, napi); + trace_napi_gro_receive_entry(skb); + + skb_gro_reset_offset(skb, 0); + + ret = napi_skb_finish(napi, skb, dev_gro_receive(napi, skb)); + trace_napi_gro_receive_exit(ret); + + return ret; +} +EXPORT_SYMBOL(napi_gro_receive); + +static void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb) +{ + if (unlikely(skb->pfmemalloc)) { + consume_skb(skb); + return; + } + __skb_pull(skb, skb_headlen(skb)); + /* restore the reserve we had after netdev_alloc_skb_ip_align() */ + skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN - skb_headroom(skb)); + __vlan_hwaccel_clear_tag(skb); + skb->dev = napi->dev; + skb->skb_iif = 0; + + /* eth_type_trans() assumes pkt_type is PACKET_HOST */ + skb->pkt_type = PACKET_HOST; + + skb->encapsulation = 0; + skb_shinfo(skb)->gso_type = 0; + skb->truesize = SKB_TRUESIZE(skb_end_offset(skb)); + if (unlikely(skb->slow_gro)) { + skb_orphan(skb); + skb_ext_reset(skb); + nf_reset_ct(skb); + skb->slow_gro = 0; + } + + napi->skb = skb; +} + +struct sk_buff *napi_get_frags(struct napi_struct *napi) +{ + struct sk_buff *skb = napi->skb; + + if (!skb) { + skb = napi_alloc_skb(napi, GRO_MAX_HEAD); + if (skb) { + napi->skb = skb; + skb_mark_napi_id(skb, napi); + } + } + return skb; +} +EXPORT_SYMBOL(napi_get_frags); + +static gro_result_t napi_frags_finish(struct napi_struct *napi, + struct sk_buff *skb, + gro_result_t ret) +{ + switch (ret) { + case GRO_NORMAL: + case GRO_HELD: + __skb_push(skb, ETH_HLEN); + skb->protocol = eth_type_trans(skb, skb->dev); + if (ret == GRO_NORMAL) + gro_normal_one(napi, skb, 1); + break; + + case GRO_MERGED_FREE: + if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD) + napi_skb_free_stolen_head(skb); + else + napi_reuse_skb(napi, skb); + break; + + case GRO_MERGED: + case GRO_CONSUMED: + break; + } + + return ret; +} + +/* Upper GRO stack assumes network header starts at gro_offset=0 + * Drivers could call both napi_gro_frags() and napi_gro_receive() + * We copy ethernet header into skb->data to have a common layout. + */ +static struct sk_buff *napi_frags_skb(struct napi_struct *napi) +{ + struct sk_buff *skb = napi->skb; + const struct ethhdr *eth; + unsigned int hlen = sizeof(*eth); + + napi->skb = NULL; + + skb_reset_mac_header(skb); + skb_gro_reset_offset(skb, hlen); + + if (unlikely(skb_gro_header_hard(skb, hlen))) { + eth = skb_gro_header_slow(skb, hlen, 0); + if (unlikely(!eth)) { + net_warn_ratelimited("%s: dropping impossible skb from %s\n", + __func__, napi->dev->name); + napi_reuse_skb(napi, skb); + return NULL; + } + } else { + eth = (const struct ethhdr *)skb->data; + gro_pull_from_frag0(skb, hlen); + NAPI_GRO_CB(skb)->frag0 += hlen; + NAPI_GRO_CB(skb)->frag0_len -= hlen; + } + __skb_pull(skb, hlen); + + /* + * This works because the only protocols we care about don't require + * special handling. + * We'll fix it up properly in napi_frags_finish() + */ + skb->protocol = eth->h_proto; + + return skb; +} + +gro_result_t napi_gro_frags(struct napi_struct *napi) +{ + gro_result_t ret; + struct sk_buff *skb = napi_frags_skb(napi); + + trace_napi_gro_frags_entry(skb); + + ret = napi_frags_finish(napi, skb, dev_gro_receive(napi, skb)); + trace_napi_gro_frags_exit(ret); + + return ret; +} +EXPORT_SYMBOL(napi_gro_frags); + +/* Compute the checksum from gro_offset and return the folded value + * after adding in any pseudo checksum. + */ +__sum16 __skb_gro_checksum_complete(struct sk_buff *skb) +{ + __wsum wsum; + __sum16 sum; + + wsum = skb_checksum(skb, skb_gro_offset(skb), skb_gro_len(skb), 0); + + /* NAPI_GRO_CB(skb)->csum holds pseudo checksum */ + sum = csum_fold(csum_add(NAPI_GRO_CB(skb)->csum, wsum)); + /* See comments in __skb_checksum_complete(). */ + if (likely(!sum)) { + if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE) && + !skb->csum_complete_sw) + netdev_rx_csum_fault(skb->dev, skb); + } + + NAPI_GRO_CB(skb)->csum = wsum; + NAPI_GRO_CB(skb)->csum_valid = 1; + + return sum; +} +EXPORT_SYMBOL(__skb_gro_checksum_complete); From 2a12ae5d433df3d3c3f1a930799ec09cb2b8058f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:11:47 -0800 Subject: [PATCH 0233/2048] net: inline sock_prot_inuse_add() sock_prot_inuse_add() is very small, we can inline it. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 14 +++++++++++--- net/core/sock.c | 11 ----------- 2 files changed, 11 insertions(+), 14 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 95cc03bd3fac..5a1e1df3cefd 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1429,13 +1429,21 @@ proto_memory_pressure(struct proto *prot) #ifdef CONFIG_PROC_FS +#define PROTO_INUSE_NR 64 /* should be enough for the first time */ +struct prot_inuse { + int val[PROTO_INUSE_NR]; +}; /* Called with local bh disabled */ -void sock_prot_inuse_add(struct net *net, struct proto *prot, int inc); +static inline void sock_prot_inuse_add(const struct net *net, + const struct proto *prot, int val) +{ + __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); +} int sock_prot_inuse_get(struct net *net, struct proto *proto); int sock_inuse_get(struct net *net); #else -static inline void sock_prot_inuse_add(struct net *net, struct proto *prot, - int inc) +static inline void sock_prot_inuse_add(const struct net *net, + const struct proto *prot, int val) { } #endif diff --git a/net/core/sock.c b/net/core/sock.c index c57d9883f62c..c0f9bdd5cf68 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3536,19 +3536,8 @@ void sk_get_meminfo(const struct sock *sk, u32 *mem) } #ifdef CONFIG_PROC_FS -#define PROTO_INUSE_NR 64 /* should be enough for the first time */ -struct prot_inuse { - int val[PROTO_INUSE_NR]; -}; - static DECLARE_BITMAP(proto_inuse_idx, PROTO_INUSE_NR); -void sock_prot_inuse_add(struct net *net, struct proto *prot, int val) -{ - __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); -} -EXPORT_SYMBOL_GPL(sock_prot_inuse_add); - int sock_prot_inuse_get(struct net *net, struct proto *prot) { int cpu, idx = prot->inuse_idx; From d477eb9004845cb2dc92ad5eed79a437738a868a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:11:48 -0800 Subject: [PATCH 0234/2048] net: make sock_inuse_add() available MPTCP hard codes it, let us instead provide this helper. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 10 ++++++++++ net/core/sock.c | 10 ---------- net/mptcp/subflow.c | 4 +--- 3 files changed, 11 insertions(+), 13 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 5a1e1df3cefd..c4c981a51797 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1439,6 +1439,12 @@ static inline void sock_prot_inuse_add(const struct net *net, { __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); } + +static inline void sock_inuse_add(const struct net *net, int val) +{ + this_cpu_add(*net->core.sock_inuse, val); +} + int sock_prot_inuse_get(struct net *net, struct proto *proto); int sock_inuse_get(struct net *net); #else @@ -1446,6 +1452,10 @@ static inline void sock_prot_inuse_add(const struct net *net, const struct proto *prot, int val) { } + +static inline void sock_inuse_add(const struct net *net, int val) +{ +} #endif diff --git a/net/core/sock.c b/net/core/sock.c index c0f9bdd5cf68..a9bd22b883b9 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -144,8 +144,6 @@ static DEFINE_MUTEX(proto_list_mutex); static LIST_HEAD(proto_list); -static void sock_inuse_add(struct net *net, int val); - /** * sk_ns_capable - General socket capability test * @sk: Socket to use a capability on or through @@ -3550,11 +3548,6 @@ int sock_prot_inuse_get(struct net *net, struct proto *prot) } EXPORT_SYMBOL_GPL(sock_prot_inuse_get); -static void sock_inuse_add(struct net *net, int val) -{ - this_cpu_add(*net->core.sock_inuse, val); -} - int sock_inuse_get(struct net *net) { int cpu, res = 0; @@ -3633,9 +3626,6 @@ static inline void release_proto_idx(struct proto *prot) { } -static void sock_inuse_add(struct net *net, int val) -{ -} #endif static void tw_prot_cleanup(struct timewait_sock_ops *twsk_prot) diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 6172f380dfb7..49787a1d7b34 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1534,9 +1534,7 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock) */ sf->sk->sk_net_refcnt = 1; get_net(net); -#ifdef CONFIG_PROC_FS - this_cpu_add(*net->core.sock_inuse, 1); -#endif + sock_inuse_add(net, 1); err = tcp_set_ulp(sf->sk, "mptcp"); release_sock(sf->sk); From 4199bae10c49e24bc2c5d8c06a68820d56640000 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:11:49 -0800 Subject: [PATCH 0235/2048] net: merge net->core.prot_inuse and net->core.sock_inuse net->core.sock_inuse is a per cpu variable (int), while net->core.prot_inuse is another per cpu variable of 64 integers. per cpu allocator tend to place them in very different places. Grouping them together makes sense, since it makes updates potentially faster, if hitting the same cache line. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/netns/core.h | 1 - include/net/sock.h | 3 ++- net/core/sock.c | 12 +----------- 3 files changed, 3 insertions(+), 13 deletions(-) diff --git a/include/net/netns/core.h b/include/net/netns/core.h index 36c2d998a43c..552bc25b1933 100644 --- a/include/net/netns/core.h +++ b/include/net/netns/core.h @@ -12,7 +12,6 @@ struct netns_core { int sysctl_somaxconn; #ifdef CONFIG_PROC_FS - int __percpu *sock_inuse; struct prot_inuse __percpu *prot_inuse; #endif }; diff --git a/include/net/sock.h b/include/net/sock.h index c4c981a51797..5589312531df 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1431,6 +1431,7 @@ proto_memory_pressure(struct proto *prot) #ifdef CONFIG_PROC_FS #define PROTO_INUSE_NR 64 /* should be enough for the first time */ struct prot_inuse { + int all; int val[PROTO_INUSE_NR]; }; /* Called with local bh disabled */ @@ -1442,7 +1443,7 @@ static inline void sock_prot_inuse_add(const struct net *net, static inline void sock_inuse_add(const struct net *net, int val) { - this_cpu_add(*net->core.sock_inuse, val); + this_cpu_add(net->core.prot_inuse->all, val); } int sock_prot_inuse_get(struct net *net, struct proto *proto); diff --git a/net/core/sock.c b/net/core/sock.c index a9bd22b883b9..d7fc8b5e2569 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3553,7 +3553,7 @@ int sock_inuse_get(struct net *net) int cpu, res = 0; for_each_possible_cpu(cpu) - res += *per_cpu_ptr(net->core.sock_inuse, cpu); + res += per_cpu_ptr(net->core.prot_inuse, cpu)->all; return res; } @@ -3565,22 +3565,12 @@ static int __net_init sock_inuse_init_net(struct net *net) net->core.prot_inuse = alloc_percpu(struct prot_inuse); if (net->core.prot_inuse == NULL) return -ENOMEM; - - net->core.sock_inuse = alloc_percpu(int); - if (net->core.sock_inuse == NULL) - goto out; - return 0; - -out: - free_percpu(net->core.prot_inuse); - return -ENOMEM; } static void __net_exit sock_inuse_exit_net(struct net *net) { free_percpu(net->core.prot_inuse); - free_percpu(net->core.sock_inuse); } static struct pernet_operations net_inuse_ops = { From b3cb764aa1d753cf6a58858f9e2097ba71e8100b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:11:50 -0800 Subject: [PATCH 0236/2048] net: drop nopreempt requirement on sock_prot_inuse_add() This is distracting really, let's make this simpler, because many callers had to take care of this by themselves, even if on x86 this adds more code than really needed. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 4 ++-- net/ieee802154/socket.c | 4 ++-- net/ipv4/raw.c | 2 +- net/ipv6/ipv6_sockglue.c | 8 ++++---- net/netlink/af_netlink.c | 4 ---- net/packet/af_packet.c | 4 ---- net/sctp/socket.c | 5 ----- net/smc/af_smc.c | 2 +- net/unix/af_unix.c | 4 ---- net/xdp/xsk.c | 4 ---- 10 files changed, 10 insertions(+), 31 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 5589312531df..f09c0c4736c4 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1434,11 +1434,11 @@ struct prot_inuse { int all; int val[PROTO_INUSE_NR]; }; -/* Called with local bh disabled */ + static inline void sock_prot_inuse_add(const struct net *net, const struct proto *prot, int val) { - __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); + this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); } static inline void sock_inuse_add(const struct net *net, int val) diff --git a/net/ieee802154/socket.c b/net/ieee802154/socket.c index 7bb9ef35c570..3b2366a88c3c 100644 --- a/net/ieee802154/socket.c +++ b/net/ieee802154/socket.c @@ -174,8 +174,8 @@ static int raw_hash(struct sock *sk) { write_lock_bh(&raw_lock); sk_add_node(sk, &raw_head); - sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&raw_lock); + sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); return 0; } @@ -453,8 +453,8 @@ static int dgram_hash(struct sock *sk) { write_lock_bh(&dgram_lock); sk_add_node(sk, &dgram_head); - sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&dgram_lock); + sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); return 0; } diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index bb446e60cf58..3910bb08a986 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -99,8 +99,8 @@ int raw_hash_sk(struct sock *sk) write_lock_bh(&h->lock); sk_add_node(sk, head); - sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&h->lock); + sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); return 0; } diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c index 41efca817db4..04683ad6d3df 100644 --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -471,10 +471,10 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname, if (sk->sk_protocol == IPPROTO_TCP) { struct inet_connection_sock *icsk = inet_csk(sk); - local_bh_disable(); + sock_prot_inuse_add(net, sk->sk_prot, -1); sock_prot_inuse_add(net, &tcp_prot, 1); - local_bh_enable(); + sk->sk_prot = &tcp_prot; icsk->icsk_af_ops = &ipv4_specific; sk->sk_socket->ops = &inet_stream_ops; @@ -485,10 +485,10 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname, if (sk->sk_protocol == IPPROTO_UDPLITE) prot = &udplite_prot; - local_bh_disable(); + sock_prot_inuse_add(net, sk->sk_prot, -1); sock_prot_inuse_add(net, prot, 1); - local_bh_enable(); + sk->sk_prot = prot; sk->sk_socket->ops = &inet_dgram_ops; sk->sk_family = PF_INET; diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 4c575324a985..1a19d179e913 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -707,9 +707,7 @@ static int netlink_create(struct net *net, struct socket *sock, int protocol, if (err < 0) goto out_module; - local_bh_disable(); sock_prot_inuse_add(net, &netlink_proto, 1); - local_bh_enable(); nlk = nlk_sk(sock->sk); nlk->module = module; @@ -809,9 +807,7 @@ static int netlink_release(struct socket *sock) netlink_table_ungrab(); } - local_bh_disable(); sock_prot_inuse_add(sock_net(sk), &netlink_proto, -1); - local_bh_enable(); call_rcu(&nlk->rcu, deferred_put_nlk_sk); return 0; } diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 46943a18a10d..a1ffdb48cc47 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -3102,9 +3102,7 @@ static int packet_release(struct socket *sock) sk_del_node_init_rcu(sk); mutex_unlock(&net->packet.sklist_lock); - preempt_disable(); sock_prot_inuse_add(net, sk->sk_prot, -1); - preempt_enable(); spin_lock(&po->bind_lock); unregister_prot_hook(sk, false); @@ -3368,9 +3366,7 @@ static int packet_create(struct net *net, struct socket *sock, int protocol, sk_add_node_tail_rcu(sk, &net->packet.sklist); mutex_unlock(&net->packet.sklist_lock); - preempt_disable(); sock_prot_inuse_add(net, &packet_proto, 1); - preempt_enable(); return 0; out2: diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 33391254fa82..055a6d3ec6e2 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -5068,12 +5068,9 @@ static int sctp_init_sock(struct sock *sk) SCTP_DBG_OBJCNT_INC(sock); - local_bh_disable(); sk_sockets_allocated_inc(sk); sock_prot_inuse_add(net, sk->sk_prot, 1); - local_bh_enable(); - return 0; } @@ -5099,10 +5096,8 @@ static void sctp_destroy_sock(struct sock *sk) list_del(&sp->auto_asconf_list); } sctp_endpoint_free(sp->ep); - local_bh_disable(); sk_sockets_allocated_dec(sk); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); - local_bh_enable(); } /* Triggered when there are no references on the socket anymore */ diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index 59284da9116d..ff5cd0c30741 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -89,8 +89,8 @@ int smc_hash_sk(struct sock *sk) write_lock_bh(&h->lock); sk_add_node(sk, head); - sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&h->lock); + sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); return 0; } diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 78e08e82c08c..54e5553a150e 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -522,9 +522,7 @@ static void unix_sock_destructor(struct sock *sk) unix_release_addr(u->addr); atomic_long_dec(&unix_nr_socks); - local_bh_disable(); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); - local_bh_enable(); #ifdef UNIX_REFCNT_DEBUG pr_debug("UNIX %p is destroyed, %ld are still alive.\n", sk, atomic_long_read(&unix_nr_socks)); @@ -889,9 +887,7 @@ static struct sock *unix_create1(struct net *net, struct socket *sock, int kern, memset(&u->scm_stat, 0, sizeof(struct scm_stat)); unix_insert_socket(unix_sockets_unbound(sk), sk); - local_bh_disable(); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); - local_bh_enable(); return sk; diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index f16074eb53c7..28ef3f4465ae 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -794,9 +794,7 @@ static int xsk_release(struct socket *sock) sk_del_node_init_rcu(sk); mutex_unlock(&net->xdp.lock); - local_bh_disable(); sock_prot_inuse_add(net, sk->sk_prot, -1); - local_bh_enable(); xsk_delete_from_maps(xs); mutex_lock(&xs->mutex); @@ -1396,9 +1394,7 @@ static int xsk_create(struct net *net, struct socket *sock, int protocol, sk_add_node_rcu(sk, &net->xdp.list); mutex_unlock(&net->xdp.lock); - local_bh_disable(); sock_prot_inuse_add(net, &xsk_proto, 1); - local_bh_enable(); return 0; } From d41bc48bfab2076f7db88d079a3a3203dd9c4a54 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Mon, 15 Nov 2021 17:30:41 -0800 Subject: [PATCH 0237/2048] selftests/bpf: Add uprobe triggering overhead benchmarks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add benchmark to measure overhead of uprobes and uretprobes. Also have a baseline (no uprobe attached) benchmark. On my dev machine, baseline benchmark can trigger 130M user_target() invocations. When uprobe is attached, this falls to just 700K. With uretprobe, we get down to 520K: $ sudo ./bench trig-uprobe-base -a Summary: hits 131.289 ± 2.872M/s # UPROBE $ sudo ./bench -a trig-uprobe-without-nop Summary: hits 0.729 ± 0.007M/s $ sudo ./bench -a trig-uprobe-with-nop Summary: hits 1.798 ± 0.017M/s # URETPROBE $ sudo ./bench -a trig-uretprobe-without-nop Summary: hits 0.508 ± 0.012M/s $ sudo ./bench -a trig-uretprobe-with-nop Summary: hits 0.883 ± 0.008M/s So there is almost 2.5x performance difference between probing nop vs non-nop instruction for entry uprobe. And 1.7x difference for uretprobe. This means that non-nop uprobe overhead is around 1.4 microseconds for uprobe and 2 microseconds for non-nop uretprobe. For nop variants, uprobe and uretprobe overhead is down to 0.556 and 1.13 microseconds, respectively. For comparison, just doing a very low-overhead syscall (with no BPF programs attached anywhere) gives: $ sudo ./bench trig-base -a Summary: hits 4.830 ± 0.036M/s So uprobes are about 2.67x slower than pure context switch. Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211116013041.4072571-1-andrii@kernel.org --- tools/testing/selftests/bpf/Makefile | 4 +- tools/testing/selftests/bpf/bench.c | 10 ++ .../selftests/bpf/benchs/bench_trigger.c | 146 ++++++++++++++++++ .../selftests/bpf/progs/trigger_bench.c | 7 + 4 files changed, 166 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 0470802c907c..35684d61aaeb 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -533,7 +533,9 @@ $(OUTPUT)/bench_ringbufs.o: $(OUTPUT)/ringbuf_bench.skel.h \ $(OUTPUT)/bench_bloom_filter_map.o: $(OUTPUT)/bloom_filter_bench.skel.h $(OUTPUT)/bench.o: bench.h testing_helpers.h $(BPFOBJ) $(OUTPUT)/bench: LDLIBS += -lm -$(OUTPUT)/bench: $(OUTPUT)/bench.o $(OUTPUT)/testing_helpers.o \ +$(OUTPUT)/bench: $(OUTPUT)/bench.o \ + $(OUTPUT)/testing_helpers.o \ + $(OUTPUT)/trace_helpers.o \ $(OUTPUT)/bench_count.o \ $(OUTPUT)/bench_rename.o \ $(OUTPUT)/bench_trigger.o \ diff --git a/tools/testing/selftests/bpf/bench.c b/tools/testing/selftests/bpf/bench.c index cc4722f693e9..c75e7ee28746 100644 --- a/tools/testing/selftests/bpf/bench.c +++ b/tools/testing/selftests/bpf/bench.c @@ -359,6 +359,11 @@ extern const struct bench bench_trig_kprobe; extern const struct bench bench_trig_fentry; extern const struct bench bench_trig_fentry_sleep; extern const struct bench bench_trig_fmodret; +extern const struct bench bench_trig_uprobe_base; +extern const struct bench bench_trig_uprobe_with_nop; +extern const struct bench bench_trig_uretprobe_with_nop; +extern const struct bench bench_trig_uprobe_without_nop; +extern const struct bench bench_trig_uretprobe_without_nop; extern const struct bench bench_rb_libbpf; extern const struct bench bench_rb_custom; extern const struct bench bench_pb_libbpf; @@ -385,6 +390,11 @@ static const struct bench *benchs[] = { &bench_trig_fentry, &bench_trig_fentry_sleep, &bench_trig_fmodret, + &bench_trig_uprobe_base, + &bench_trig_uprobe_with_nop, + &bench_trig_uretprobe_with_nop, + &bench_trig_uprobe_without_nop, + &bench_trig_uretprobe_without_nop, &bench_rb_libbpf, &bench_rb_custom, &bench_pb_libbpf, diff --git a/tools/testing/selftests/bpf/benchs/bench_trigger.c b/tools/testing/selftests/bpf/benchs/bench_trigger.c index f41a491a8cc0..049a5ad56f65 100644 --- a/tools/testing/selftests/bpf/benchs/bench_trigger.c +++ b/tools/testing/selftests/bpf/benchs/bench_trigger.c @@ -2,6 +2,7 @@ /* Copyright (c) 2020 Facebook */ #include "bench.h" #include "trigger_bench.skel.h" +#include "trace_helpers.h" /* BPF triggering benchmarks */ static struct trigger_ctx { @@ -107,6 +108,101 @@ static void *trigger_consumer(void *input) return NULL; } +/* make sure call is not inlined and not avoided by compiler, so __weak and + * inline asm volatile in the body of the function + * + * There is a performance difference between uprobing at nop location vs other + * instructions. So use two different targets, one of which starts with nop + * and another doesn't. + * + * GCC doesn't generate stack setup preample for these functions due to them + * having no input arguments and doing nothing in the body. + */ +__weak void uprobe_target_with_nop(void) +{ + asm volatile ("nop"); +} + +__weak void uprobe_target_without_nop(void) +{ + asm volatile (""); +} + +static void *uprobe_base_producer(void *input) +{ + while (true) { + uprobe_target_with_nop(); + atomic_inc(&base_hits.value); + } + return NULL; +} + +static void *uprobe_producer_with_nop(void *input) +{ + while (true) + uprobe_target_with_nop(); + return NULL; +} + +static void *uprobe_producer_without_nop(void *input) +{ + while (true) + uprobe_target_without_nop(); + return NULL; +} + +static void usetup(bool use_retprobe, bool use_nop) +{ + size_t uprobe_offset; + ssize_t base_addr; + struct bpf_link *link; + + setup_libbpf(); + + ctx.skel = trigger_bench__open_and_load(); + if (!ctx.skel) { + fprintf(stderr, "failed to open skeleton\n"); + exit(1); + } + + base_addr = get_base_addr(); + if (use_nop) + uprobe_offset = get_uprobe_offset(&uprobe_target_with_nop, base_addr); + else + uprobe_offset = get_uprobe_offset(&uprobe_target_without_nop, base_addr); + + link = bpf_program__attach_uprobe(ctx.skel->progs.bench_trigger_uprobe, + use_retprobe, + -1 /* all PIDs */, + "/proc/self/exe", + uprobe_offset); + if (!link) { + fprintf(stderr, "failed to attach uprobe!\n"); + exit(1); + } + ctx.skel->links.bench_trigger_uprobe = link; +} + +static void uprobe_setup_with_nop() +{ + usetup(false, true); +} + +static void uretprobe_setup_with_nop() +{ + usetup(true, true); +} + +static void uprobe_setup_without_nop() +{ + usetup(false, false); +} + +static void uretprobe_setup_without_nop() +{ + usetup(true, false); +} + const struct bench bench_trig_base = { .name = "trig-base", .validate = trigger_validate, @@ -182,3 +278,53 @@ const struct bench bench_trig_fmodret = { .report_progress = hits_drops_report_progress, .report_final = hits_drops_report_final, }; + +const struct bench bench_trig_uprobe_base = { + .name = "trig-uprobe-base", + .setup = NULL, /* no uprobe/uretprobe is attached */ + .producer_thread = uprobe_base_producer, + .consumer_thread = trigger_consumer, + .measure = trigger_base_measure, + .report_progress = hits_drops_report_progress, + .report_final = hits_drops_report_final, +}; + +const struct bench bench_trig_uprobe_with_nop = { + .name = "trig-uprobe-with-nop", + .setup = uprobe_setup_with_nop, + .producer_thread = uprobe_producer_with_nop, + .consumer_thread = trigger_consumer, + .measure = trigger_measure, + .report_progress = hits_drops_report_progress, + .report_final = hits_drops_report_final, +}; + +const struct bench bench_trig_uretprobe_with_nop = { + .name = "trig-uretprobe-with-nop", + .setup = uretprobe_setup_with_nop, + .producer_thread = uprobe_producer_with_nop, + .consumer_thread = trigger_consumer, + .measure = trigger_measure, + .report_progress = hits_drops_report_progress, + .report_final = hits_drops_report_final, +}; + +const struct bench bench_trig_uprobe_without_nop = { + .name = "trig-uprobe-without-nop", + .setup = uprobe_setup_without_nop, + .producer_thread = uprobe_producer_without_nop, + .consumer_thread = trigger_consumer, + .measure = trigger_measure, + .report_progress = hits_drops_report_progress, + .report_final = hits_drops_report_final, +}; + +const struct bench bench_trig_uretprobe_without_nop = { + .name = "trig-uretprobe-without-nop", + .setup = uretprobe_setup_without_nop, + .producer_thread = uprobe_producer_without_nop, + .consumer_thread = trigger_consumer, + .measure = trigger_measure, + .report_progress = hits_drops_report_progress, + .report_final = hits_drops_report_final, +}; diff --git a/tools/testing/selftests/bpf/progs/trigger_bench.c b/tools/testing/selftests/bpf/progs/trigger_bench.c index 9a4d09590b3d..2098f3f27f18 100644 --- a/tools/testing/selftests/bpf/progs/trigger_bench.c +++ b/tools/testing/selftests/bpf/progs/trigger_bench.c @@ -52,3 +52,10 @@ int bench_trigger_fmodret(void *ctx) __sync_add_and_fetch(&hits, 1); return -22; } + +SEC("uprobe/self/uprobe_target") +int bench_trigger_uprobe(void *ctx) +{ + __sync_add_and_fetch(&hits, 1); + return 0; +} From be6c5ba2b00a24883162aae90a59429441b97619 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sun, 14 Nov 2021 19:05:17 -0800 Subject: [PATCH 0238/2048] Bluetooth: btmrvl_main: repair a non-kernel-doc comment Do not use "/**" to begin a non-kernel-doc comment. Fixes this build warning: drivers/bluetooth/btmrvl_main.c:2: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst Signed-off-by: Randy Dunlap Reported-by: kernel test robot Cc: Marcel Holtmann Cc: Johan Hedberg Cc: Luiz Augusto von Dentz Cc: linux-bluetooth@vger.kernel.org Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btmrvl_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/bluetooth/btmrvl_main.c b/drivers/bluetooth/btmrvl_main.c index 5ccbe4d459d0..181338f60530 100644 --- a/drivers/bluetooth/btmrvl_main.c +++ b/drivers/bluetooth/btmrvl_main.c @@ -1,4 +1,4 @@ -/** +/* * Marvell Bluetooth driver * * Copyright (C) 2009, Marvell International Ltd. From 2bd1b237616bd91a3f4f0cd94dc53cd6cba7aff9 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Thu, 11 Nov 2021 16:48:42 -0800 Subject: [PATCH 0239/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_DISCOVERABLE to use cmd_sync This makes MGMT_OP_SET_DISCOVERABLE use hci_cmd_sync_queue instead of use a dedicated discoverable_update work. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 2 - include/net/bluetooth/hci_sync.h | 3 ++ net/bluetooth/hci_request.c | 12 ----- net/bluetooth/hci_sync.c | 90 +++++++++++++++++++++++++++++++- net/bluetooth/mgmt.c | 43 +++++++-------- 5 files changed, 112 insertions(+), 38 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index b5f061882c10..5f57ff81b7b8 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -496,7 +496,6 @@ struct hci_dev { struct work_struct bg_scan_update; struct work_struct scan_update; struct work_struct connectable_update; - struct work_struct discoverable_update; struct delayed_work le_scan_disable; struct delayed_work le_scan_restart; @@ -1828,7 +1827,6 @@ void mgmt_new_conn_param(struct hci_dev *hdev, bdaddr_t *bdaddr, void mgmt_smp_complete(struct hci_conn *conn, bool complete); bool mgmt_get_connectable(struct hci_dev *hdev); void mgmt_set_connectable_complete(struct hci_dev *hdev, u8 status); -void mgmt_set_discoverable_complete(struct hci_dev *hdev, u8 status); u8 mgmt_get_adv_discov_flags(struct hci_dev *hdev); void mgmt_advertising_added(struct sock *sk, struct hci_dev *hdev, u8 instance); diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 00b13e8ca800..d335c0ce8c5d 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -90,6 +90,9 @@ int hci_dev_close_sync(struct hci_dev *hdev); int hci_powered_update_sync(struct hci_dev *hdev); int hci_set_powered_sync(struct hci_dev *hdev, u8 val); +int hci_update_discoverable_sync(struct hci_dev *hdev); +int hci_update_discoverable(struct hci_dev *hdev); + int hci_start_discovery_sync(struct hci_dev *hdev); int hci_stop_discovery_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 8aa6e1840c9a..9f355d8c34f2 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -2131,16 +2131,6 @@ static int discoverable_update(struct hci_request *req, unsigned long opt) return 0; } -static void discoverable_update_work(struct work_struct *work) -{ - struct hci_dev *hdev = container_of(work, struct hci_dev, - discoverable_update); - u8 status; - - hci_req_sync(hdev, discoverable_update, 0, HCI_CMD_TIMEOUT, &status); - mgmt_set_discoverable_complete(hdev, status); -} - void __hci_abort_conn(struct hci_request *req, struct hci_conn *conn, u8 reason) { @@ -2852,7 +2842,6 @@ void hci_request_setup(struct hci_dev *hdev) INIT_WORK(&hdev->bg_scan_update, bg_scan_update); INIT_WORK(&hdev->scan_update, scan_update_work); INIT_WORK(&hdev->connectable_update, connectable_update_work); - INIT_WORK(&hdev->discoverable_update, discoverable_update_work); INIT_DELAYED_WORK(&hdev->discov_off, discov_off); INIT_DELAYED_WORK(&hdev->le_scan_disable, le_scan_disable_work); INIT_DELAYED_WORK(&hdev->le_scan_restart, le_scan_restart_work); @@ -2868,7 +2857,6 @@ void hci_request_cancel_all(struct hci_dev *hdev) cancel_work_sync(&hdev->bg_scan_update); cancel_work_sync(&hdev->scan_update); cancel_work_sync(&hdev->connectable_update); - cancel_work_sync(&hdev->discoverable_update); cancel_delayed_work_sync(&hdev->discov_off); cancel_delayed_work_sync(&hdev->le_scan_disable); cancel_delayed_work_sync(&hdev->le_scan_restart); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 06d08dc6ad22..7be9f8a7340d 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -1699,7 +1699,6 @@ static int hci_resume_advertising_sync(struct hci_dev *hdev) hdev->advertising_paused = false; if (hdev->advertising_old_state) { hci_dev_set_flag(hdev, HCI_ADVERTISING); - queue_work(hdev->req_workqueue, &hdev->discoverable_update); hdev->advertising_old_state = 0; } @@ -4392,6 +4391,95 @@ int hci_set_powered_sync(struct hci_dev *hdev, u8 val) return hci_power_off_sync(hdev); } +static int hci_write_iac_sync(struct hci_dev *hdev) +{ + struct hci_cp_write_current_iac_lap cp; + + if (!hci_dev_test_flag(hdev, HCI_DISCOVERABLE)) + return 0; + + memset(&cp, 0, sizeof(cp)); + + if (hci_dev_test_flag(hdev, HCI_LIMITED_DISCOVERABLE)) { + /* Limited discoverable mode */ + cp.num_iac = min_t(u8, hdev->num_iac, 2); + cp.iac_lap[0] = 0x00; /* LIAC */ + cp.iac_lap[1] = 0x8b; + cp.iac_lap[2] = 0x9e; + cp.iac_lap[3] = 0x33; /* GIAC */ + cp.iac_lap[4] = 0x8b; + cp.iac_lap[5] = 0x9e; + } else { + /* General discoverable mode */ + cp.num_iac = 1; + cp.iac_lap[0] = 0x33; /* GIAC */ + cp.iac_lap[1] = 0x8b; + cp.iac_lap[2] = 0x9e; + } + + return __hci_cmd_sync_status(hdev, HCI_OP_WRITE_CURRENT_IAC_LAP, + (cp.num_iac * 3) + 1, &cp, + HCI_CMD_TIMEOUT); +} + +int hci_update_discoverable_sync(struct hci_dev *hdev) +{ + int err = 0; + + if (hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) { + err = hci_write_iac_sync(hdev); + if (err) + return err; + + err = hci_update_scan_sync(hdev); + if (err) + return err; + + err = hci_update_class_sync(hdev); + if (err) + return err; + } + + /* Advertising instances don't use the global discoverable setting, so + * only update AD if advertising was enabled using Set Advertising. + */ + if (hci_dev_test_flag(hdev, HCI_ADVERTISING)) { + err = hci_update_adv_data_sync(hdev, 0x00); + if (err) + return err; + + /* Discoverable mode affects the local advertising + * address in limited privacy mode. + */ + if (hci_dev_test_flag(hdev, HCI_LIMITED_PRIVACY)) { + if (ext_adv_capable(hdev)) + err = hci_start_ext_adv_sync(hdev, 0x00); + else + err = hci_enable_advertising_sync(hdev); + } + } + + return err; +} + +static int update_discoverable_sync(struct hci_dev *hdev, void *data) +{ + return hci_update_discoverable_sync(hdev); +} + +int hci_update_discoverable(struct hci_dev *hdev) +{ + /* Only queue if it would have any effect */ + if (hdev_is_powered(hdev) && + hci_dev_test_flag(hdev, HCI_ADVERTISING) && + hci_dev_test_flag(hdev, HCI_DISCOVERABLE) && + hci_dev_test_flag(hdev, HCI_LIMITED_PRIVACY)) + return hci_cmd_sync_queue(hdev, update_discoverable_sync, NULL, + NULL); + + return 0; +} + static int hci_inquiry_sync(struct hci_dev *hdev, u8 length) { const u8 giac[3] = { 0x33, 0x8b, 0x9e }; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 06384d761928..2697e1ef1964 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1370,23 +1370,20 @@ static u8 mgmt_le_support(struct hci_dev *hdev) return MGMT_STATUS_SUCCESS; } -void mgmt_set_discoverable_complete(struct hci_dev *hdev, u8 status) +static void mgmt_set_discoverable_complete(struct hci_dev *hdev, void *data, + int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; - bt_dev_dbg(hdev, "status 0x%02x", status); + bt_dev_dbg(hdev, "err %d", err); hci_dev_lock(hdev); - cmd = pending_find(MGMT_OP_SET_DISCOVERABLE, hdev); - if (!cmd) - goto unlock; - - if (status) { - u8 mgmt_err = mgmt_status(status); + if (err) { + u8 mgmt_err = mgmt_status(err); mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, mgmt_err); hci_dev_clear_flag(hdev, HCI_LIMITED_DISCOVERABLE); - goto remove_cmd; + goto done; } if (hci_dev_test_flag(hdev, HCI_DISCOVERABLE) && @@ -1398,13 +1395,18 @@ void mgmt_set_discoverable_complete(struct hci_dev *hdev, u8 status) send_settings_rsp(cmd->sk, MGMT_OP_SET_DISCOVERABLE, hdev); new_settings(hdev, cmd->sk); -remove_cmd: - mgmt_pending_remove(cmd); - -unlock: +done: + mgmt_pending_free(cmd); hci_dev_unlock(hdev); } +static int set_discoverable_sync(struct hci_dev *hdev, void *data) +{ + BT_DBG("%s", hdev->name); + + return hci_update_discoverable_sync(hdev); +} + static int set_discoverable(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -1503,7 +1505,7 @@ static int set_discoverable(struct sock *sk, struct hci_dev *hdev, void *data, goto failed; } - cmd = mgmt_pending_add(sk, MGMT_OP_SET_DISCOVERABLE, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_SET_DISCOVERABLE, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; @@ -1527,8 +1529,8 @@ static int set_discoverable(struct sock *sk, struct hci_dev *hdev, void *data, else hci_dev_clear_flag(hdev, HCI_LIMITED_DISCOVERABLE); - queue_work(hdev->req_workqueue, &hdev->discoverable_update); - err = 0; + err = hci_cmd_sync_queue(hdev, set_discoverable_sync, cmd, + mgmt_set_discoverable_complete); failed: hci_dev_unlock(hdev); @@ -1677,12 +1679,7 @@ static int set_bondable(struct sock *sk, struct hci_dev *hdev, void *data, /* In limited privacy mode the change of bondable mode * may affect the local advertising address. */ - if (hdev_is_powered(hdev) && - hci_dev_test_flag(hdev, HCI_ADVERTISING) && - hci_dev_test_flag(hdev, HCI_DISCOVERABLE) && - hci_dev_test_flag(hdev, HCI_LIMITED_PRIVACY)) - queue_work(hdev->req_workqueue, - &hdev->discoverable_update); + hci_update_discoverable(hdev); err = new_settings(hdev, sk); } From f056a65783cce9c1279c1635d92768ce5962e4d6 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Thu, 11 Nov 2021 16:48:43 -0800 Subject: [PATCH 0240/2048] Bluetooth: hci_sync: Convert MGMT_OP_SET_CONNECTABLE to use cmd_sync This makes MGMT_OP_SET_CONNEABLE use hci_cmd_sync_queue instead of use a dedicated connetable_update work. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 2 -- include/net/bluetooth/hci_sync.h | 2 ++ net/bluetooth/hci_request.c | 43 -------------------------------- net/bluetooth/hci_sync.c | 31 +++++++++++++++++++++++ net/bluetooth/mgmt.c | 36 +++++++++++++------------- 5 files changed, 52 insertions(+), 62 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 5f57ff81b7b8..acb46ae27b5a 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -495,7 +495,6 @@ struct hci_dev { struct work_struct discov_update; struct work_struct bg_scan_update; struct work_struct scan_update; - struct work_struct connectable_update; struct delayed_work le_scan_disable; struct delayed_work le_scan_restart; @@ -1826,7 +1825,6 @@ void mgmt_new_conn_param(struct hci_dev *hdev, bdaddr_t *bdaddr, u16 max_interval, u16 latency, u16 timeout); void mgmt_smp_complete(struct hci_conn *conn, bool complete); bool mgmt_get_connectable(struct hci_dev *hdev); -void mgmt_set_connectable_complete(struct hci_dev *hdev, u8 status); u8 mgmt_get_adv_discov_flags(struct hci_dev *hdev); void mgmt_advertising_added(struct sock *sk, struct hci_dev *hdev, u8 instance); diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index d335c0ce8c5d..0336c1bc5d25 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -93,6 +93,8 @@ int hci_set_powered_sync(struct hci_dev *hdev, u8 val); int hci_update_discoverable_sync(struct hci_dev *hdev); int hci_update_discoverable(struct hci_dev *hdev); +int hci_update_connectable_sync(struct hci_dev *hdev); + int hci_start_discovery_sync(struct hci_dev *hdev); int hci_stop_discovery_sync(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 9f355d8c34f2..98bf425eaf1b 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -1987,47 +1987,6 @@ static void scan_update_work(struct work_struct *work) hci_req_sync(hdev, update_scan, 0, HCI_CMD_TIMEOUT, NULL); } -static int connectable_update(struct hci_request *req, unsigned long opt) -{ - struct hci_dev *hdev = req->hdev; - - hci_dev_lock(hdev); - - __hci_req_update_scan(req); - - /* If BR/EDR is not enabled and we disable advertising as a - * by-product of disabling connectable, we need to update the - * advertising flags. - */ - if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) - __hci_req_update_adv_data(req, hdev->cur_adv_instance); - - /* Update the advertising parameters if necessary */ - if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || - !list_empty(&hdev->adv_instances)) { - if (ext_adv_capable(hdev)) - __hci_req_start_ext_adv(req, hdev->cur_adv_instance); - else - __hci_req_enable_advertising(req); - } - - __hci_update_background_scan(req); - - hci_dev_unlock(hdev); - - return 0; -} - -static void connectable_update_work(struct work_struct *work) -{ - struct hci_dev *hdev = container_of(work, struct hci_dev, - connectable_update); - u8 status; - - hci_req_sync(hdev, connectable_update, 0, HCI_CMD_TIMEOUT, &status); - mgmt_set_connectable_complete(hdev, status); -} - static u8 get_service_classes(struct hci_dev *hdev) { struct bt_uuid *uuid; @@ -2841,7 +2800,6 @@ void hci_request_setup(struct hci_dev *hdev) INIT_WORK(&hdev->discov_update, discov_update); INIT_WORK(&hdev->bg_scan_update, bg_scan_update); INIT_WORK(&hdev->scan_update, scan_update_work); - INIT_WORK(&hdev->connectable_update, connectable_update_work); INIT_DELAYED_WORK(&hdev->discov_off, discov_off); INIT_DELAYED_WORK(&hdev->le_scan_disable, le_scan_disable_work); INIT_DELAYED_WORK(&hdev->le_scan_restart, le_scan_restart_work); @@ -2856,7 +2814,6 @@ void hci_request_cancel_all(struct hci_dev *hdev) cancel_work_sync(&hdev->discov_update); cancel_work_sync(&hdev->bg_scan_update); cancel_work_sync(&hdev->scan_update); - cancel_work_sync(&hdev->connectable_update); cancel_delayed_work_sync(&hdev->discov_off); cancel_delayed_work_sync(&hdev->le_scan_disable); cancel_delayed_work_sync(&hdev->le_scan_restart); diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 7be9f8a7340d..ad86caf41f91 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -4480,6 +4480,37 @@ int hci_update_discoverable(struct hci_dev *hdev) return 0; } +int hci_update_connectable_sync(struct hci_dev *hdev) +{ + int err; + + err = hci_update_scan_sync(hdev); + if (err) + return err; + + /* If BR/EDR is not enabled and we disable advertising as a + * by-product of disabling connectable, we need to update the + * advertising flags. + */ + if (!hci_dev_test_flag(hdev, HCI_BREDR_ENABLED)) + err = hci_update_adv_data_sync(hdev, hdev->cur_adv_instance); + + /* Update the advertising parameters if necessary */ + if (hci_dev_test_flag(hdev, HCI_ADVERTISING) || + !list_empty(&hdev->adv_instances)) { + if (ext_adv_capable(hdev)) + err = hci_start_ext_adv_sync(hdev, + hdev->cur_adv_instance); + else + err = hci_enable_advertising_sync(hdev); + + if (err) + return err; + } + + return hci_update_passive_scan_sync(hdev); +} + static int hci_inquiry_sync(struct hci_dev *hdev, u8 length) { const u8 giac[3] = { 0x33, 0x8b, 0x9e }; diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 2697e1ef1964..f8f74d344297 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -1537,31 +1537,26 @@ failed: return err; } -void mgmt_set_connectable_complete(struct hci_dev *hdev, u8 status) +static void mgmt_set_connectable_complete(struct hci_dev *hdev, void *data, + int err) { - struct mgmt_pending_cmd *cmd; + struct mgmt_pending_cmd *cmd = data; - bt_dev_dbg(hdev, "status 0x%02x", status); + bt_dev_dbg(hdev, "err %d", err); hci_dev_lock(hdev); - cmd = pending_find(MGMT_OP_SET_CONNECTABLE, hdev); - if (!cmd) - goto unlock; - - if (status) { - u8 mgmt_err = mgmt_status(status); + if (err) { + u8 mgmt_err = mgmt_status(err); mgmt_cmd_status(cmd->sk, cmd->index, cmd->opcode, mgmt_err); - goto remove_cmd; + goto done; } send_settings_rsp(cmd->sk, MGMT_OP_SET_CONNECTABLE, hdev); new_settings(hdev, cmd->sk); -remove_cmd: - mgmt_pending_remove(cmd); - -unlock: +done: + mgmt_pending_free(cmd); hci_dev_unlock(hdev); } @@ -1594,6 +1589,13 @@ static int set_connectable_update_settings(struct hci_dev *hdev, return 0; } +static int set_connectable_sync(struct hci_dev *hdev, void *data) +{ + BT_DBG("%s", hdev->name); + + return hci_update_connectable_sync(hdev); +} + static int set_connectable(struct sock *sk, struct hci_dev *hdev, void *data, u16 len) { @@ -1626,7 +1628,7 @@ static int set_connectable(struct sock *sk, struct hci_dev *hdev, void *data, goto failed; } - cmd = mgmt_pending_add(sk, MGMT_OP_SET_CONNECTABLE, hdev, data, len); + cmd = mgmt_pending_new(sk, MGMT_OP_SET_CONNECTABLE, hdev, data, len); if (!cmd) { err = -ENOMEM; goto failed; @@ -1643,8 +1645,8 @@ static int set_connectable(struct sock *sk, struct hci_dev *hdev, void *data, hci_dev_clear_flag(hdev, HCI_CONNECTABLE); } - queue_work(hdev->req_workqueue, &hdev->connectable_update); - err = 0; + err = hci_cmd_sync_queue(hdev, set_connectable_sync, cmd, + mgmt_set_connectable_complete); failed: hci_dev_unlock(hdev); From 9482c5074a7d0bec682cc35a12d687f185684f31 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Thu, 11 Nov 2021 16:48:44 -0800 Subject: [PATCH 0241/2048] Bluetooth: hci_request: Remove bg_scan_update work This work is no longer necessary since all the code using it has been converted to use hci_passive_scan/hci_passive_scan_sync. Signed-off-by: Luiz Augusto von Dentz Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci_core.h | 1 - net/bluetooth/hci_request.c | 107 ------------------------------- net/bluetooth/hci_request.h | 5 -- 3 files changed, 113 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index acb46ae27b5a..2560cfe80db8 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -493,7 +493,6 @@ struct hci_dev { struct work_struct tx_work; struct work_struct discov_update; - struct work_struct bg_scan_update; struct work_struct scan_update; struct delayed_work le_scan_disable; struct delayed_work le_scan_restart; diff --git a/net/bluetooth/hci_request.c b/net/bluetooth/hci_request.c index 98bf425eaf1b..8b3205e4b23e 100644 --- a/net/bluetooth/hci_request.c +++ b/net/bluetooth/hci_request.c @@ -368,82 +368,6 @@ static bool __hci_update_interleaved_scan(struct hci_dev *hdev) return false; } -/* This function controls the background scanning based on hdev->pend_le_conns - * list. If there are pending LE connection we start the background scanning, - * otherwise we stop it. - * - * This function requires the caller holds hdev->lock. - */ -static void __hci_update_background_scan(struct hci_request *req) -{ - struct hci_dev *hdev = req->hdev; - - if (!test_bit(HCI_UP, &hdev->flags) || - test_bit(HCI_INIT, &hdev->flags) || - hci_dev_test_flag(hdev, HCI_SETUP) || - hci_dev_test_flag(hdev, HCI_CONFIG) || - hci_dev_test_flag(hdev, HCI_AUTO_OFF) || - hci_dev_test_flag(hdev, HCI_UNREGISTER)) - return; - - /* No point in doing scanning if LE support hasn't been enabled */ - if (!hci_dev_test_flag(hdev, HCI_LE_ENABLED)) - return; - - /* If discovery is active don't interfere with it */ - if (hdev->discovery.state != DISCOVERY_STOPPED) - return; - - /* Reset RSSI and UUID filters when starting background scanning - * since these filters are meant for service discovery only. - * - * The Start Discovery and Start Service Discovery operations - * ensure to set proper values for RSSI threshold and UUID - * filter list. So it is safe to just reset them here. - */ - hci_discovery_filter_clear(hdev); - - bt_dev_dbg(hdev, "ADV monitoring is %s", - hci_is_adv_monitoring(hdev) ? "on" : "off"); - - if (list_empty(&hdev->pend_le_conns) && - list_empty(&hdev->pend_le_reports) && - !hci_is_adv_monitoring(hdev)) { - /* If there is no pending LE connections or devices - * to be scanned for or no ADV monitors, we should stop the - * background scanning. - */ - - /* If controller is not scanning we are done. */ - if (!hci_dev_test_flag(hdev, HCI_LE_SCAN)) - return; - - hci_req_add_le_scan_disable(req, false); - - bt_dev_dbg(hdev, "stopping background scanning"); - } else { - /* If there is at least one pending LE connection, we should - * keep the background scan running. - */ - - /* If controller is connecting, we should not start scanning - * since some controllers are not able to scan and connect at - * the same time. - */ - if (hci_lookup_le_connect(hdev)) - return; - - /* If controller is currently scanning, we stop it to ensure we - * don't miss any advertising (due to duplicates filter). - */ - if (hci_dev_test_flag(hdev, HCI_LE_SCAN)) - hci_req_add_le_scan_disable(req, false); - - hci_req_add_le_passive_scan(req); - bt_dev_dbg(hdev, "starting background scanning"); - } -} - void __hci_req_update_name(struct hci_request *req) { struct hci_dev *hdev = req->hdev; @@ -2183,35 +2107,6 @@ int hci_abort_conn(struct hci_conn *conn, u8 reason) return 0; } -static int update_bg_scan(struct hci_request *req, unsigned long opt) -{ - hci_dev_lock(req->hdev); - __hci_update_background_scan(req); - hci_dev_unlock(req->hdev); - return 0; -} - -static void bg_scan_update(struct work_struct *work) -{ - struct hci_dev *hdev = container_of(work, struct hci_dev, - bg_scan_update); - struct hci_conn *conn; - u8 status; - int err; - - err = hci_req_sync(hdev, update_bg_scan, 0, HCI_CMD_TIMEOUT, &status); - if (!err) - return; - - hci_dev_lock(hdev); - - conn = hci_conn_hash_lookup_state(hdev, LE_LINK, BT_CONNECT); - if (conn) - hci_le_conn_failed(conn, status); - - hci_dev_unlock(hdev); -} - static int le_scan_disable(struct hci_request *req, unsigned long opt) { hci_req_add_le_scan_disable(req, false); @@ -2798,7 +2693,6 @@ int __hci_req_hci_power_on(struct hci_dev *hdev) void hci_request_setup(struct hci_dev *hdev) { INIT_WORK(&hdev->discov_update, discov_update); - INIT_WORK(&hdev->bg_scan_update, bg_scan_update); INIT_WORK(&hdev->scan_update, scan_update_work); INIT_DELAYED_WORK(&hdev->discov_off, discov_off); INIT_DELAYED_WORK(&hdev->le_scan_disable, le_scan_disable_work); @@ -2812,7 +2706,6 @@ void hci_request_cancel_all(struct hci_dev *hdev) hci_req_sync_cancel(hdev, ENODEV); cancel_work_sync(&hdev->discov_update); - cancel_work_sync(&hdev->bg_scan_update); cancel_work_sync(&hdev->scan_update); cancel_delayed_work_sync(&hdev->discov_off); cancel_delayed_work_sync(&hdev->le_scan_disable); diff --git a/net/bluetooth/hci_request.h b/net/bluetooth/hci_request.h index ba75c2da70f4..5f8e8846ec74 100644 --- a/net/bluetooth/hci_request.h +++ b/net/bluetooth/hci_request.h @@ -127,10 +127,5 @@ int hci_abort_conn(struct hci_conn *conn, u8 reason); void __hci_abort_conn(struct hci_request *req, struct hci_conn *conn, u8 reason); -static inline void hci_update_background_scan(struct hci_dev *hdev) -{ - queue_work(hdev->req_workqueue, &hdev->bg_scan_update); -} - void hci_request_setup(struct hci_dev *hdev); void hci_request_cancel_all(struct hci_dev *hdev); From 0f281a5e5b673698c246da5d7bf0d5fa427c47c5 Mon Sep 17 00:00:00 2001 From: Archie Pusaka Date: Thu, 11 Nov 2021 13:20:53 +0800 Subject: [PATCH 0242/2048] Bluetooth: Ignore HCI_ERROR_CANCELLED_BY_HOST on adv set terminated event This event is received when the controller stops advertising, specifically for these three reasons: (a) Connection is successfully created (success). (b) Timeout is reached (error). (c) Number of advertising events is reached (error). (*) This event is NOT generated when the host stops the advertisement. Refer to the BT spec ver 5.3 vol 4 part E sec 7.7.65.18. Note that the section was revised from BT spec ver 5.0 vol 2 part E sec 7.7.65.18 which was ambiguous about (*). Some chips (e.g. RTL8822CE) send this event when the host stops the advertisement with status = HCI_ERROR_CANCELLED_BY_HOST (due to (*) above). This is treated as an error and the advertisement will be removed and userspace will be informed via MGMT event. On suspend, we are supposed to temporarily disable advertisements, and continue advertising on resume. However, due to the behavior above, the advertisements are removed instead. This patch returns early if HCI_ERROR_CANCELLED_BY_HOST is received. Btmon snippet of the unexpected behavior: @ MGMT Command: Remove Advertising (0x003f) plen 1 Instance: 1 < HCI Command: LE Set Extended Advertising Enable (0x08|0x0039) plen 6 Extended advertising: Disabled (0x00) Number of sets: 1 (0x01) Entry 0 Handle: 0x01 Duration: 0 ms (0x00) Max ext adv events: 0 > HCI Event: LE Meta Event (0x3e) plen 6 LE Advertising Set Terminated (0x12) Status: Operation Cancelled by Host (0x44) Handle: 1 Connection handle: 0 Number of completed extended advertising events: 5 > HCI Event: Command Complete (0x0e) plen 4 LE Set Extended Advertising Enable (0x08|0x0039) ncmd 2 Status: Success (0x00) Signed-off-by: Archie Pusaka Signed-off-by: Marcel Holtmann --- include/net/bluetooth/hci.h | 1 + net/bluetooth/hci_event.c | 12 ++++++++++++ 2 files changed, 13 insertions(+) diff --git a/include/net/bluetooth/hci.h b/include/net/bluetooth/hci.h index 63065bc01b76..84db6b275231 100644 --- a/include/net/bluetooth/hci.h +++ b/include/net/bluetooth/hci.h @@ -566,6 +566,7 @@ enum { #define HCI_ERROR_INVALID_LL_PARAMS 0x1e #define HCI_ERROR_UNSPECIFIED 0x1f #define HCI_ERROR_ADVERTISING_TIMEOUT 0x3c +#define HCI_ERROR_CANCELLED_BY_HOST 0x44 /* Flow control modes */ #define HCI_FLOW_CTL_MODE_PACKET_BASED 0x00 diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 9e27ff343716..8eae38129a6e 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -5533,6 +5533,18 @@ static void hci_le_ext_adv_term_evt(struct hci_dev *hdev, struct sk_buff *skb) adv = hci_find_adv_instance(hdev, ev->handle); + /* The Bluetooth Core 5.3 specification clearly states that this event + * shall not be sent when the Host disables the advertising set. So in + * case of HCI_ERROR_CANCELLED_BY_HOST, just ignore the event. + * + * When the Host disables an advertising set, all cleanup is done via + * its command callback and not needed to be duplicated here. + */ + if (ev->status == HCI_ERROR_CANCELLED_BY_HOST) { + bt_dev_warn_ratelimited(hdev, "Unexpected advertising set terminated event"); + return; + } + if (ev->status) { if (!adv) return; From 1f9d56574334590b836695cfec78dad029f36dba Mon Sep 17 00:00:00 2001 From: Archie Pusaka Date: Thu, 11 Nov 2021 13:20:54 +0800 Subject: [PATCH 0243/2048] Bluetooth: Attempt to clear HCI_LE_ADV on adv set terminated error event We should clear the flag if the adv instance removed due to receiving this error status is the last one we have. Signed-off-by: Archie Pusaka Reviewed-by: Miao-chen Chou Signed-off-by: Marcel Holtmann --- net/bluetooth/hci_event.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 8eae38129a6e..efc5458b1345 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -5527,7 +5527,7 @@ static void hci_le_ext_adv_term_evt(struct hci_dev *hdev, struct sk_buff *skb) { struct hci_evt_le_ext_adv_set_term *ev = (void *) skb->data; struct hci_conn *conn; - struct adv_info *adv; + struct adv_info *adv, *n; BT_DBG("%s status 0x%2.2x", hdev->name, ev->status); @@ -5553,6 +5553,13 @@ static void hci_le_ext_adv_term_evt(struct hci_dev *hdev, struct sk_buff *skb) hci_remove_adv_instance(hdev, ev->handle); mgmt_advertising_removed(NULL, hdev, ev->handle); + list_for_each_entry_safe(adv, n, &hdev->adv_instances, list) { + if (adv->enabled) + return; + } + + /* We are no longer advertising, clear HCI_LE_ADV */ + hci_dev_clear_flag(hdev, HCI_LE_ADV); return; } From 28491d7ef4af471841e454f8c1f77384f93c6fef Mon Sep 17 00:00:00 2001 From: "mark-yw.chen" Date: Fri, 5 Nov 2021 02:26:05 +0800 Subject: [PATCH 0244/2048] Bluetooth: btusb: enable Mediatek to support AOSP extension This patch enables AOSP extension for Mediatek Chip (MT7921 & MT7922). Signed-off-by: mark-yw.chen Signed-off-by: Marcel Holtmann --- drivers/bluetooth/btusb.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index 46d892bbde62..30a057e1d4e3 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2565,6 +2565,7 @@ static int btusb_mtk_setup(struct hci_dev *hdev) } hci_set_msft_opcode(hdev, 0xFD30); + hci_set_aosp_capable(hdev); goto done; default: bt_dev_err(hdev, "Unsupported hardware variant (%08x)", From c2c60ea37e5b6be58c9dd7aff0b2e86ba0f18e0b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:23:01 -0800 Subject: [PATCH 0245/2048] once: use __section(".data.once") .data.once contains nicely packed bool variables. It is used already by DO_ONCE_LITE(). Using it also in DO_ONCE() removes holes in .data section. Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/linux/once.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/once.h b/include/linux/once.h index d361fb14ac3a..f54523052bbc 100644 --- a/include/linux/once.h +++ b/include/linux/once.h @@ -38,7 +38,7 @@ void __do_once_done(bool *done, struct static_key_true *once_key, #define DO_ONCE(func, ...) \ ({ \ bool ___ret = false; \ - static bool ___done = false; \ + static bool __section(".data.once") ___done = false; \ static DEFINE_STATIC_KEY_TRUE(___once_key); \ if (static_branch_unlikely(&___once_key)) { \ unsigned long ___flags; \ From 7071732c26fe2cf141185ed16a8a85d02495ae8c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:23:02 -0800 Subject: [PATCH 0246/2048] net: use .data.once section in netdev_level_once() Same rationale than prior patch : using the dedicated section avoid holes and pack all these bool values. Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/linux/netdevice.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 31a7e6b27681..dd328364dfe9 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -4942,7 +4942,7 @@ void netdev_info(const struct net_device *dev, const char *format, ...); #define netdev_level_once(level, dev, fmt, ...) \ do { \ - static bool __print_once __read_mostly; \ + static bool __section(".data.once") __print_once; \ \ if (!__print_once) { \ __print_once = true; \ From 49ecc2e9c3abd269951972fa8b23a4d081111b80 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 15 Nov 2021 09:23:03 -0800 Subject: [PATCH 0247/2048] net: align static siphash keys siphash keys use 16 bytes. Define siphash_aligned_key_t macro so that we can make sure they are not crossing a cache line boundary. Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/linux/siphash.h | 2 ++ net/core/flow_dissector.c | 2 +- net/core/secure_seq.c | 4 ++-- net/ipv4/route.c | 2 +- net/ipv4/syncookies.c | 2 +- net/ipv6/route.c | 2 +- net/ipv6/syncookies.c | 2 +- net/netfilter/nf_conntrack_core.c | 4 ++-- net/netfilter/nf_conntrack_expect.c | 2 +- net/netfilter/nf_conntrack_netlink.c | 2 +- net/netfilter/nf_nat_core.c | 2 +- 11 files changed, 14 insertions(+), 12 deletions(-) diff --git a/include/linux/siphash.h b/include/linux/siphash.h index bf21591a9e5e..3f7427b9e935 100644 --- a/include/linux/siphash.h +++ b/include/linux/siphash.h @@ -21,6 +21,8 @@ typedef struct { u64 key[2]; } siphash_key_t; +#define siphash_aligned_key_t siphash_key_t __aligned(16) + static inline bool siphash_key_is_zero(const siphash_key_t *key) { return !(key->key[0] | key->key[1]); diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c index 3255f57f5131..257976cb55ce 100644 --- a/net/core/flow_dissector.c +++ b/net/core/flow_dissector.c @@ -1460,7 +1460,7 @@ out_bad: } EXPORT_SYMBOL(__skb_flow_dissect); -static siphash_key_t hashrnd __read_mostly; +static siphash_aligned_key_t hashrnd; static __always_inline void __flow_hash_secret_init(void) { net_get_random_once(&hashrnd, sizeof(hashrnd)); diff --git a/net/core/secure_seq.c b/net/core/secure_seq.c index b5bc680d4755..9b8443774449 100644 --- a/net/core/secure_seq.c +++ b/net/core/secure_seq.c @@ -19,8 +19,8 @@ #include #include -static siphash_key_t net_secret __read_mostly; -static siphash_key_t ts_secret __read_mostly; +static siphash_aligned_key_t net_secret; +static siphash_aligned_key_t ts_secret; static __always_inline void net_secret_init(void) { diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 0b4103b1e622..243a0c52be42 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -602,7 +602,7 @@ static void fnhe_remove_oldest(struct fnhe_hash_bucket *hash) static u32 fnhe_hashfun(__be32 daddr) { - static siphash_key_t fnhe_hash_key __read_mostly; + static siphash_aligned_key_t fnhe_hash_key; u64 hval; net_get_random_once(&fnhe_hash_key, sizeof(fnhe_hash_key)); diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c index 8696dc343ad2..2cb3b852d148 100644 --- a/net/ipv4/syncookies.c +++ b/net/ipv4/syncookies.c @@ -14,7 +14,7 @@ #include #include -static siphash_key_t syncookie_secret[2] __read_mostly; +static siphash_aligned_key_t syncookie_secret[2]; #define COOKIEBITS 24 /* Upper bits store count */ #define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 3ae25b8ffbd6..5e8f2f15607d 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1485,7 +1485,7 @@ static void rt6_exception_remove_oldest(struct rt6_exception_bucket *bucket) static u32 rt6_exception_hash(const struct in6_addr *dst, const struct in6_addr *src) { - static siphash_key_t rt6_exception_key __read_mostly; + static siphash_aligned_key_t rt6_exception_key; struct { struct in6_addr dst; struct in6_addr src; diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index e8cfb9e997bf..d1b61d00368e 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -20,7 +20,7 @@ #define COOKIEBITS 24 /* Upper bits store count */ #define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1) -static siphash_key_t syncookie6_secret[2] __read_mostly; +static siphash_aligned_key_t syncookie6_secret[2]; /* RFC 2460, Section 8.3: * [ipv6 tcp] MSS must be computed as the maximum packet size minus 60 [..] diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index 770a63103c7a..054ee9d25efe 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -189,7 +189,7 @@ EXPORT_SYMBOL_GPL(nf_conntrack_htable_size); unsigned int nf_conntrack_max __read_mostly; EXPORT_SYMBOL_GPL(nf_conntrack_max); seqcount_spinlock_t nf_conntrack_generation __read_mostly; -static siphash_key_t nf_conntrack_hash_rnd __read_mostly; +static siphash_aligned_key_t nf_conntrack_hash_rnd; static u32 hash_conntrack_raw(const struct nf_conntrack_tuple *tuple, unsigned int zoneid, @@ -482,7 +482,7 @@ EXPORT_SYMBOL_GPL(nf_ct_invert_tuple); */ u32 nf_ct_get_id(const struct nf_conn *ct) { - static __read_mostly siphash_key_t ct_id_seed; + static siphash_aligned_key_t ct_id_seed; unsigned long a, b, c, d; net_get_random_once(&ct_id_seed, sizeof(ct_id_seed)); diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c index f562eeef4234..1e89b595ecd0 100644 --- a/net/netfilter/nf_conntrack_expect.c +++ b/net/netfilter/nf_conntrack_expect.c @@ -41,7 +41,7 @@ EXPORT_SYMBOL_GPL(nf_ct_expect_hash); unsigned int nf_ct_expect_max __read_mostly; static struct kmem_cache *nf_ct_expect_cachep __read_mostly; -static siphash_key_t nf_ct_expect_hashrnd __read_mostly; +static siphash_aligned_key_t nf_ct_expect_hashrnd; /* nf_conntrack_expect helper functions */ void nf_ct_unlink_expect_report(struct nf_conntrack_expect *exp, diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index f1e5443fe7c7..3d6c8da3de1f 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -2997,7 +2997,7 @@ static const union nf_inet_addr any_addr; static __be32 nf_expect_get_id(const struct nf_conntrack_expect *exp) { - static __read_mostly siphash_key_t exp_id_seed; + static siphash_aligned_key_t exp_id_seed; unsigned long a, b, c, d; net_get_random_once(&exp_id_seed, sizeof(exp_id_seed)); diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c index 4d50d51db796..ab9f6c75524d 100644 --- a/net/netfilter/nf_nat_core.c +++ b/net/netfilter/nf_nat_core.c @@ -34,7 +34,7 @@ static unsigned int nat_net_id __read_mostly; static struct hlist_head *nf_nat_bysource __read_mostly; static unsigned int nf_nat_htable_size __read_mostly; -static siphash_key_t nf_nat_hash_rnd __read_mostly; +static siphash_aligned_key_t nf_nat_hash_rnd; struct nf_nat_lookup_hook_priv { struct nf_hook_entries __rcu *entries; From 4b5f82f6aaef3fa95cce52deb8510f55ddda6a71 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Mon, 15 Nov 2021 21:17:56 +0100 Subject: [PATCH 0248/2048] r8169: enable ASPM L1/L1.1 from RTL8168h With newer chip versions ASPM-related issues seem to occur only if L1.2 is enabled. I have a test system with RTL8168h that gives a number of rx_missed errors when running iperf and L1.2 is enabled. With L1.2 disabled (and L1 + L1.1 active) everything is fine. See also [0]. Can't test this, but L1 + L1.1 being active should be sufficient to reach higher package power saving states. [0] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1942830 Signed-off-by: Heiner Kallweit Link: https://lore.kernel.org/r/36feb8c4-a0b6-422a-899c-e61f2e869dfe@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/realtek/r8169_main.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index bbe21db20417..d87b48432ca3 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -5271,12 +5271,6 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (rc) return rc; - /* Disable ASPM L1 as that cause random device stop working - * problems as well as full system hangs for some PCIe devices users. - */ - rc = pci_disable_link_state(pdev, PCIE_LINK_STATE_L1); - tp->aspm_manageable = !rc; - /* enable device (incl. PCI PM wakeup and hotplug setup) */ rc = pcim_enable_device(pdev); if (rc < 0) { @@ -5319,6 +5313,17 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) tp->mac_version = chipset; + /* Disable ASPM L1 as that cause random device stop working + * problems as well as full system hangs for some PCIe devices users. + * Chips from RTL8168h partially have issues with L1.2, but seem + * to work fine with L1 and L1.1. + */ + if (tp->mac_version >= RTL_GIGA_MAC_VER_45) + rc = pci_disable_link_state(pdev, PCIE_LINK_STATE_L1_2); + else + rc = pci_disable_link_state(pdev, PCIE_LINK_STATE_L1); + tp->aspm_manageable = !rc; + tp->dash_type = rtl_check_dash(tp); tp->cp_cmd = RTL_R16(tp, CPlusCmd) & CPCMD_MASK; From 2d6600c754f8e4e52c392c3b5aecc554145e2524 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Mon, 15 Nov 2021 21:51:14 +0100 Subject: [PATCH 0249/2048] r8169: disable detection of chip versions 49 and 50 It seems these chip versions never made it to the wild. Therefore disable detection and if nobody complains remove support completely later. Signed-off-by: Heiner Kallweit Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/realtek/r8169_main.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index d87b48432ca3..75adf9707409 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -1978,8 +1978,11 @@ static enum mac_version rtl8169_get_mac_version(u16 xid, bool gmii) /* 8168EP family. */ { 0x7cf, 0x502, RTL_GIGA_MAC_VER_51 }, - { 0x7cf, 0x501, RTL_GIGA_MAC_VER_50 }, - { 0x7cf, 0x500, RTL_GIGA_MAC_VER_49 }, + /* It seems this chip version never made it to + * the wild. Let's disable detection. + * { 0x7cf, 0x501, RTL_GIGA_MAC_VER_50 }, + * { 0x7cf, 0x500, RTL_GIGA_MAC_VER_49 }, + */ /* 8168H family. */ { 0x7cf, 0x541, RTL_GIGA_MAC_VER_46 }, From 6c8a5cf97c3f9d4c1a322c095ca60acf691d0226 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Mon, 15 Nov 2021 21:51:52 +0100 Subject: [PATCH 0250/2048] r8169: disable detection of chip version 45 It seems this chip version never made it to the wild. Therefore disable detection and if nobody complains remove support completely later. Signed-off-by: Heiner Kallweit Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/realtek/r8169_main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 75adf9707409..d6a456cce85c 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -1986,7 +1986,10 @@ static enum mac_version rtl8169_get_mac_version(u16 xid, bool gmii) /* 8168H family. */ { 0x7cf, 0x541, RTL_GIGA_MAC_VER_46 }, - { 0x7cf, 0x540, RTL_GIGA_MAC_VER_45 }, + /* It seems this chip version never made it to + * the wild. Let's disable detection. + * { 0x7cf, 0x540, RTL_GIGA_MAC_VER_45 }, + */ /* 8168G family. */ { 0x7cf, 0x5c8, RTL_GIGA_MAC_VER_44 }, From 364ef1f3785785404473e02c6a020dae9d44fc4c Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Mon, 15 Nov 2021 21:52:35 +0100 Subject: [PATCH 0251/2048] r8169: disable detection of chip version 41 It seems this chip version never made it to the wild. Therefore disable detection and if nobody complains remove support completely later. Signed-off-by: Heiner Kallweit Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/realtek/r8169_main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index d6a456cce85c..e896e5eca804 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -1994,7 +1994,10 @@ static enum mac_version rtl8169_get_mac_version(u16 xid, bool gmii) /* 8168G family. */ { 0x7cf, 0x5c8, RTL_GIGA_MAC_VER_44 }, { 0x7cf, 0x509, RTL_GIGA_MAC_VER_42 }, - { 0x7cf, 0x4c1, RTL_GIGA_MAC_VER_41 }, + /* It seems this chip version never made it to + * the wild. Let's disable detection. + * { 0x7cf, 0x4c1, RTL_GIGA_MAC_VER_41 }, + */ { 0x7cf, 0x4c0, RTL_GIGA_MAC_VER_40 }, /* 8168F family. */ From b9241f54138ca5af4d3c5ca6db56be83d7491508 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Mon, 15 Nov 2021 17:11:17 +0000 Subject: [PATCH 0252/2048] net: document SMII and correct phylink's new validation mechanism SMII has not been documented in the kernel, but information on this PHY interface mode has been recently found. Document it, and correct the recently introduced phylink handling for this interface mode. Signed-off-by: Russell King (Oracle) Reviewed-by: Andrew Lunn Link: https://lore.kernel.org/r/E1mmfVl-0075nP-14@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- Documentation/networking/phy.rst | 5 +++++ drivers/net/phy/phylink.c | 2 +- include/linux/phy.h | 2 +- 3 files changed, 7 insertions(+), 2 deletions(-) diff --git a/Documentation/networking/phy.rst b/Documentation/networking/phy.rst index 571ba08386e7..d43da709bf40 100644 --- a/Documentation/networking/phy.rst +++ b/Documentation/networking/phy.rst @@ -237,6 +237,11 @@ negotiation results. Some of the interface modes are described below: +``PHY_INTERFACE_MODE_SMII`` + This is serial MII, clocked at 125MHz, supporting 100M and 10M speeds. + Some details can be found in + https://opencores.org/ocsvn/smii/smii/trunk/doc/SMII.pdf + ``PHY_INTERFACE_MODE_1000BASEX`` This defines the 1000BASE-X single-lane serdes link as defined by the 802.3 standard section 36. The link operates at a fixed bit rate of diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index 33462fdc7add..f7156b6868e7 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -336,6 +336,7 @@ void phylink_get_linkmodes(unsigned long *linkmodes, phy_interface_t interface, case PHY_INTERFACE_MODE_REVRMII: case PHY_INTERFACE_MODE_RMII: + case PHY_INTERFACE_MODE_SMII: case PHY_INTERFACE_MODE_REVMII: case PHY_INTERFACE_MODE_MII: caps |= MAC_10HD | MAC_10FD; @@ -385,7 +386,6 @@ void phylink_get_linkmodes(unsigned long *linkmodes, phy_interface_t interface, case PHY_INTERFACE_MODE_NA: case PHY_INTERFACE_MODE_MAX: - case PHY_INTERFACE_MODE_SMII: break; } diff --git a/include/linux/phy.h b/include/linux/phy.h index 96e43fbb2dd8..1e57cdd95da3 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -99,7 +99,7 @@ extern const int phy_10gbit_features_array[1]; * @PHY_INTERFACE_MODE_RGMII_RXID: RGMII with Internal RX delay * @PHY_INTERFACE_MODE_RGMII_TXID: RGMII with Internal RX delay * @PHY_INTERFACE_MODE_RTBI: Reduced TBI - * @PHY_INTERFACE_MODE_SMII: ??? MII + * @PHY_INTERFACE_MODE_SMII: Serial MII * @PHY_INTERFACE_MODE_XGMII: 10 gigabit media-independent interface * @PHY_INTERFACE_MODE_XLGMII:40 gigabit media-independent interface * @PHY_INTERFACE_MODE_MOCA: Multimedia over Coax From bc541621f8ba18cc8aeed4c48b63aec181d6c20f Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Tue, 14 Sep 2021 23:26:17 -0700 Subject: [PATCH 0253/2048] net/mlx5e: Support ethtool cq mode Add support for ethtool coalesce cq mode set and get. Signed-off-by: Saeed Mahameed Reviewed-by: Tariq Toukan --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 7 ++- .../ethernet/mellanox/mlx5/core/en_ethtool.c | 49 ++++++++++++++++--- .../net/ethernet/mellanox/mlx5/core/en_rep.c | 4 +- .../mellanox/mlx5/core/ipoib/ethtool.c | 4 +- 4 files changed, 52 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index f0ac6b0d9653..48b12ee44b8d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -1148,9 +1148,12 @@ void mlx5e_ethtool_get_channels(struct mlx5e_priv *priv, int mlx5e_ethtool_set_channels(struct mlx5e_priv *priv, struct ethtool_channels *ch); int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv, - struct ethtool_coalesce *coal); + struct ethtool_coalesce *coal, + struct kernel_ethtool_coalesce *kernel_coal); int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, - struct ethtool_coalesce *coal); + struct ethtool_coalesce *coal, + struct kernel_ethtool_coalesce *kernel_coal, + struct netlink_ext_ack *extack); int mlx5e_ethtool_get_link_ksettings(struct mlx5e_priv *priv, struct ethtool_link_ksettings *link_ksettings); int mlx5e_ethtool_set_link_ksettings(struct mlx5e_priv *priv, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c index c2ea5fad48dd..45bdfcb3dcc7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c @@ -511,7 +511,8 @@ static int mlx5e_set_channels(struct net_device *dev, } int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv, - struct ethtool_coalesce *coal) + struct ethtool_coalesce *coal, + struct kernel_ethtool_coalesce *kernel_coal) { struct dim_cq_moder *rx_moder, *tx_moder; @@ -528,6 +529,11 @@ int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv, coal->tx_max_coalesced_frames = tx_moder->pkts; coal->use_adaptive_tx_coalesce = priv->channels.params.tx_dim_enabled; + kernel_coal->use_cqe_mode_rx = + MLX5E_GET_PFLAG(&priv->channels.params, MLX5E_PFLAG_RX_CQE_BASED_MODER); + kernel_coal->use_cqe_mode_tx = + MLX5E_GET_PFLAG(&priv->channels.params, MLX5E_PFLAG_TX_CQE_BASED_MODER); + return 0; } @@ -538,7 +544,7 @@ static int mlx5e_get_coalesce(struct net_device *netdev, { struct mlx5e_priv *priv = netdev_priv(netdev); - return mlx5e_ethtool_get_coalesce(priv, coal); + return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal); } #define MLX5E_MAX_COAL_TIME MLX5_MAX_CQ_PERIOD @@ -578,14 +584,26 @@ mlx5e_set_priv_channels_rx_coalesce(struct mlx5e_priv *priv, struct ethtool_coal } } +/* convert a boolean value of cq_mode to mlx5 period mode + * true : MLX5_CQ_PERIOD_MODE_START_FROM_CQE + * false : MLX5_CQ_PERIOD_MODE_START_FROM_EQE + */ +static int cqe_mode_to_period_mode(bool val) +{ + return val ? MLX5_CQ_PERIOD_MODE_START_FROM_CQE : MLX5_CQ_PERIOD_MODE_START_FROM_EQE; +} + int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, - struct ethtool_coalesce *coal) + struct ethtool_coalesce *coal, + struct kernel_ethtool_coalesce *kernel_coal, + struct netlink_ext_ack *extack) { struct dim_cq_moder *rx_moder, *tx_moder; struct mlx5_core_dev *mdev = priv->mdev; struct mlx5e_params new_params; bool reset_rx, reset_tx; bool reset = true; + u8 cq_period_mode; int err = 0; if (!MLX5_CAP_GEN(mdev, cq_moderation)) @@ -605,6 +623,12 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, return -ERANGE; } + if ((kernel_coal->use_cqe_mode_rx || kernel_coal->use_cqe_mode_tx) && + !MLX5_CAP_GEN(priv->mdev, cq_period_start_from_cqe)) { + NL_SET_ERR_MSG_MOD(extack, "cqe_mode_rx/tx is not supported on this device"); + return -EOPNOTSUPP; + } + mutex_lock(&priv->state_lock); new_params = priv->channels.params; @@ -621,6 +645,18 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, reset_rx = !!coal->use_adaptive_rx_coalesce != priv->channels.params.rx_dim_enabled; reset_tx = !!coal->use_adaptive_tx_coalesce != priv->channels.params.tx_dim_enabled; + cq_period_mode = cqe_mode_to_period_mode(kernel_coal->use_cqe_mode_rx); + if (cq_period_mode != rx_moder->cq_period_mode) { + mlx5e_set_rx_cq_mode_params(&new_params, cq_period_mode); + reset_rx = true; + } + + cq_period_mode = cqe_mode_to_period_mode(kernel_coal->use_cqe_mode_tx); + if (cq_period_mode != tx_moder->cq_period_mode) { + mlx5e_set_tx_cq_mode_params(&new_params, cq_period_mode); + reset_tx = true; + } + if (reset_rx) { u8 mode = MLX5E_GET_PFLAG(&new_params, MLX5E_PFLAG_RX_CQE_BASED_MODER); @@ -656,9 +692,9 @@ static int mlx5e_set_coalesce(struct net_device *netdev, struct kernel_ethtool_coalesce *kernel_coal, struct netlink_ext_ack *extack) { - struct mlx5e_priv *priv = netdev_priv(netdev); + struct mlx5e_priv *priv = netdev_priv(netdev); - return mlx5e_ethtool_set_coalesce(priv, coal); + return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack); } static void ptys2ethtool_supported_link(struct mlx5_core_dev *mdev, @@ -2358,7 +2394,8 @@ static void mlx5e_get_rmon_stats(struct net_device *netdev, const struct ethtool_ops mlx5e_ethtool_ops = { .supported_coalesce_params = ETHTOOL_COALESCE_USECS | ETHTOOL_COALESCE_MAX_FRAMES | - ETHTOOL_COALESCE_USE_ADAPTIVE, + ETHTOOL_COALESCE_USE_ADAPTIVE | + ETHTOOL_COALESCE_USE_CQE, .get_drvinfo = mlx5e_get_drvinfo, .get_link = ethtool_op_get_link, .get_link_ext_state = mlx5e_get_link_ext_state, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index e58a9ec42553..8c81aeba07db 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -258,7 +258,7 @@ static int mlx5e_rep_get_coalesce(struct net_device *netdev, { struct mlx5e_priv *priv = netdev_priv(netdev); - return mlx5e_ethtool_get_coalesce(priv, coal); + return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal); } static int mlx5e_rep_set_coalesce(struct net_device *netdev, @@ -268,7 +268,7 @@ static int mlx5e_rep_set_coalesce(struct net_device *netdev, { struct mlx5e_priv *priv = netdev_priv(netdev); - return mlx5e_ethtool_set_coalesce(priv, coal); + return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack); } static u32 mlx5e_rep_get_rxfh_key_size(struct net_device *netdev) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c index 962d41418ce7..f23e33ac9c6b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c @@ -105,7 +105,7 @@ static int mlx5i_set_coalesce(struct net_device *netdev, { struct mlx5e_priv *priv = mlx5i_epriv(netdev); - return mlx5e_ethtool_set_coalesce(priv, coal); + return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack); } static int mlx5i_get_coalesce(struct net_device *netdev, @@ -115,7 +115,7 @@ static int mlx5i_get_coalesce(struct net_device *netdev, { struct mlx5e_priv *priv = mlx5i_epriv(netdev); - return mlx5e_ethtool_get_coalesce(priv, coal); + return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal); } static int mlx5i_get_ts_info(struct net_device *netdev, From aef0f8c67d752ed68cc18582711ac6accf0ad9b6 Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Wed, 3 Nov 2021 14:01:05 -0700 Subject: [PATCH 0254/2048] net/mlx5: Fix format-security build warnings Treat the string as an argument to avoid this. drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c:482:5: error: format string is not a string literal (potentially insecure) name); ^~~~ drivers/net/ethernet/mellanox/mlx5/core/en_stats.c:2079:4: error: format string is not a string literal (potentially insecure) ptp_ch_stats_desc[i].format); ^~~~~~~~~~~~~~~~~~~~~~~~~~~ Signed-off-by: Saeed Mahameed Reviewed-by: Shay Drory Reviewed-by: Moshe Shemesh --- drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 2 +- drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index 2a9bfc3ffa2e..3c91a11e27ad 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -2076,7 +2076,7 @@ static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(ptp) for (i = 0; i < NUM_PTP_CH_STATS; i++) sprintf(data + (idx++) * ETH_GSTRING_LEN, - ptp_ch_stats_desc[i].format); + "%s", ptp_ch_stats_desc[i].format); if (priv->tx_ptp_opened) { for (tc = 0; tc < priv->max_opened_tc; tc++) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c index 830444f927d4..19bf2b66707d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c @@ -479,7 +479,7 @@ irq_pool_alloc(struct mlx5_core_dev *dev, int start, int size, char *name, pool->xa_num_irqs.max = start + size - 1; if (name) snprintf(pool->name, MLX5_MAX_IRQ_NAME - MLX5_MAX_IRQ_IDX_CHARS, - name); + "%s", name); pool->min_threshold = min_threshold * MLX5_EQ_REFS_PER_IRQ; pool->max_threshold = max_threshold * MLX5_EQ_REFS_PER_IRQ; mlx5_core_dbg(dev, "pool->name = %s, pool->size = %d, pool->start = %d", From f28a14c1dcb0dbdae719a05b71ed30ce31221e75 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Tue, 9 Nov 2021 15:44:58 +0200 Subject: [PATCH 0255/2048] net/mlx5: Avoid printing health buffer when firmware is unavailable Use firmware version field as an indication to health buffer's sanity. When firmware version is 0xFFFFFFFF, deduce that firmware is unavailable and avoid printing the health buffer to dmesg as it doesn't provide debug info. Signed-off-by: Aya Levin Reviewed-by: Gal Pressman Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/health.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c index 64f1abc4dc36..75121bc1eaa5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/health.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c @@ -420,6 +420,11 @@ static void print_health_info(struct mlx5_core_dev *dev) if (!ioread8(&h->synd)) return; + if (ioread32be(&h->fw_ver) == 0xFFFFFFFF) { + mlx5_log(dev, LOGLEVEL_ERR, "PCI slot is unavailable\n"); + return; + } + rfr_severity = ioread8(&h->rfr_severity); severity = mlx5_health_get_severity(rfr_severity); mlx5_log(dev, severity, "Health issue observed, %s, severity(%d) %s:\n", From 2c0e5cf5206ecd5da3c6bc5799671c2172713d71 Mon Sep 17 00:00:00 2001 From: Paul Blakey Date: Mon, 5 Jul 2021 11:31:47 +0300 Subject: [PATCH 0256/2048] net/mlx5e: Refactor mod header management API For all mod hdr related functions to reside in a single self contained component (mod_hdr.c), refactor alloc() and add get_id() so that user won't rely on internal implementation, and move both to mod_hdr component. Rename the prefix to mlx5e_mod_hdr_* as other mod hdr functions. Signed-off-by: Paul Blakey Reviewed-by: Oz Shlomo Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/en/mod_hdr.c | 47 ++++++++++ .../ethernet/mellanox/mlx5/core/en/mod_hdr.h | 13 +++ .../mellanox/mlx5/core/en/tc/sample.c | 5 +- .../ethernet/mellanox/mlx5/core/en/tc_ct.c | 25 ++---- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 90 ++++--------------- .../net/ethernet/mellanox/mlx5/core/en_tc.h | 5 -- .../mellanox/mlx5/core/esw/indir_table.c | 5 +- 7 files changed, 90 insertions(+), 100 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c index 7edde4d536fd..19d05fb4aab2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c @@ -155,3 +155,50 @@ struct mlx5_modify_hdr *mlx5e_mod_hdr_get(struct mlx5e_mod_hdr_handle *mh) return mh->modify_hdr; } +char * +mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace, + struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts) +{ + int new_num_actions, max_hw_actions; + size_t new_sz, old_sz; + void *ret; + + if (mod_hdr_acts->num_actions < mod_hdr_acts->max_actions) + goto out; + + max_hw_actions = mlx5e_mod_hdr_max_actions(mdev, namespace); + new_num_actions = min(max_hw_actions, + mod_hdr_acts->actions ? + mod_hdr_acts->max_actions * 2 : 1); + if (mod_hdr_acts->max_actions == new_num_actions) + return ERR_PTR(-ENOSPC); + + new_sz = MLX5_MH_ACT_SZ * new_num_actions; + old_sz = mod_hdr_acts->max_actions * MLX5_MH_ACT_SZ; + + ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL); + if (!ret) + return ERR_PTR(-ENOMEM); + + memset(ret + old_sz, 0, new_sz - old_sz); + mod_hdr_acts->actions = ret; + mod_hdr_acts->max_actions = new_num_actions; + +out: + return mod_hdr_acts->actions + (mod_hdr_acts->num_actions * MLX5_MH_ACT_SZ); +} + +void +mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts) +{ + kfree(mod_hdr_acts->actions); + mod_hdr_acts->actions = NULL; + mod_hdr_acts->num_actions = 0; + mod_hdr_acts->max_actions = 0; +} + +char * +mlx5e_mod_hdr_get_item(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts, int pos) +{ + return mod_hdr_acts->actions + (pos * MLX5_MH_ACT_SZ); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h index 33b23d8f9182..b8cd1a7a31be 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h @@ -15,6 +15,11 @@ struct mlx5e_tc_mod_hdr_acts { void *actions; }; +char *mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace, + struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); +void mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); +char *mlx5e_mod_hdr_get_item(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts, int pos); + struct mlx5e_mod_hdr_handle * mlx5e_mod_hdr_attach(struct mlx5_core_dev *mdev, struct mod_hdr_tbl *tbl, @@ -28,4 +33,12 @@ struct mlx5_modify_hdr *mlx5e_mod_hdr_get(struct mlx5e_mod_hdr_handle *mh); void mlx5e_mod_hdr_tbl_init(struct mod_hdr_tbl *tbl); void mlx5e_mod_hdr_tbl_destroy(struct mod_hdr_tbl *tbl); +static inline int mlx5e_mod_hdr_max_actions(struct mlx5_core_dev *mdev, int namespace) +{ + if (namespace == MLX5_FLOW_NAMESPACE_FDB) /* FDB offloading */ + return MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, max_modify_header_actions); + else /* namespace is MLX5_FLOW_NAMESPACE_KERNEL - NIC offloading */ + return MLX5_CAP_FLOWTABLE_NIC_RX(mdev, max_modify_header_actions); +} + #endif /* __MLX5E_EN_MOD_HDR_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c index df6888c4793c..ff4b4f8a5a9d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c @@ -5,6 +5,7 @@ #include #include "en/mapping.h" #include "en/tc/post_act.h" +#include "en/mod_hdr.h" #include "sample.h" #include "eswitch.h" #include "en_tc.h" @@ -255,12 +256,12 @@ sample_modify_hdr_get(struct mlx5_core_dev *mdev, u32 obj_id, goto err_modify_hdr; } - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); return modify_hdr; err_modify_hdr: err_post_act: - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); err_set_regc0: return ERR_PTR(err); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index c1c6e74c79c4..89065fac7590 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -609,22 +609,15 @@ mlx5_tc_ct_entry_create_nat(struct mlx5_tc_ct_priv *ct_priv, struct flow_action *flow_action = &flow_rule->action; struct mlx5_core_dev *mdev = ct_priv->dev; struct flow_action_entry *act; - size_t action_size; char *modact; int err, i; - action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto); - flow_action_for_each(i, act, flow_action) { switch (act->id) { case FLOW_ACTION_MANGLE: { - err = alloc_mod_hdr_actions(mdev, ct_priv->ns_type, - mod_acts); - if (err) - return err; - - modact = mod_acts->actions + - mod_acts->num_actions * action_size; + modact = mlx5e_mod_hdr_alloc(mdev, ct_priv->ns_type, mod_acts); + if (IS_ERR(modact)) + return PTR_ERR(modact); err = mlx5_tc_ct_parse_mangle_to_mod_act(act, modact); if (err) @@ -706,11 +699,11 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv, attr->modify_hdr = mlx5e_mod_hdr_get(*mh); } - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); return 0; err_mapping: - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); mlx5_put_label_mapping(ct_priv, attr->ct_attr.ct_labels_id); return err; } @@ -1445,7 +1438,7 @@ static int tc_ct_pre_ct_add_rules(struct mlx5_ct_ft *ct_ft, } pre_ct->miss_rule = rule; - dealloc_mod_hdr_actions(&pre_mod_acts); + mlx5e_mod_hdr_dealloc(&pre_mod_acts); kvfree(spec); return 0; @@ -1454,7 +1447,7 @@ err_miss_rule: err_flow_rule: mlx5_modify_header_dealloc(dev, pre_ct->modify_hdr); err_mapping: - dealloc_mod_hdr_actions(&pre_mod_acts); + mlx5e_mod_hdr_dealloc(&pre_mod_acts); kvfree(spec); return err; } @@ -1850,14 +1843,14 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv, } attr->ct_attr.ct_flow = ct_flow; - dealloc_mod_hdr_actions(&pre_mod_acts); + mlx5e_mod_hdr_dealloc(&pre_mod_acts); return ct_flow->pre_ct_rule; err_insert_orig: mlx5_modify_header_dealloc(priv->mdev, pre_ct_attr->modify_hdr); err_mapping: - dealloc_mod_hdr_actions(&pre_mod_acts); + mlx5e_mod_hdr_dealloc(&pre_mod_acts); mlx5_chains_put_chain_mapping(ct_priv->chains, ct_flow->chain_mapping); err_get_chain: kfree(ct_flow->pre_ct_attr); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 835caa1c7b74..e620100eabe0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -71,7 +71,6 @@ #include "lag/mp.h" #define nic_chains(priv) ((priv)->fs.tc.chains) -#define MLX5_MH_ACT_SZ MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto) #define MLX5E_TC_TABLE_NUM_GROUPS 4 #define MLX5E_TC_TABLE_MAX_GROUP_SIZE BIT(18) @@ -209,12 +208,9 @@ mlx5e_tc_match_to_reg_set_and_get_id(struct mlx5_core_dev *mdev, char *modact; int err; - err = alloc_mod_hdr_actions(mdev, ns, mod_hdr_acts); - if (err) - return err; - - modact = mod_hdr_acts->actions + - (mod_hdr_acts->num_actions * MLX5_MH_ACT_SZ); + modact = mlx5e_mod_hdr_alloc(mdev, ns, mod_hdr_acts); + if (IS_ERR(modact)) + return PTR_ERR(modact); /* Firmware has 5bit length field and 0 means 32bits */ if (mlen == 32) @@ -333,7 +329,7 @@ void mlx5e_tc_match_to_reg_mod_hdr_change(struct mlx5_core_dev *mdev, int mlen = mlx5e_tc_attr_to_reg_mappings[type].mlen; char *modact; - modact = mod_hdr_acts->actions + (act_id * MLX5_MH_ACT_SZ); + modact = mlx5e_mod_hdr_get_item(mod_hdr_acts, act_id); /* Firmware has 5bit length field and 0 means 32bits */ if (mlen == 32) @@ -1076,7 +1072,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) { err = mlx5e_attach_mod_hdr(priv, flow, parse_attr); - dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts); + mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts); if (err) return err; } @@ -1623,7 +1619,7 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv, mlx5_tc_ct_match_del(get_ct_priv(priv), &flow->attr->ct_attr); if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) { - dealloc_mod_hdr_actions(&attr->parse_attr->mod_hdr_acts); + mlx5e_mod_hdr_dealloc(&attr->parse_attr->mod_hdr_acts); if (vf_tun && attr->modify_hdr) mlx5_modify_header_dealloc(priv->mdev, attr->modify_hdr); else @@ -2766,13 +2762,12 @@ static int offload_pedit_fields(struct mlx5e_priv *priv, struct netlink_ext_ack *extack) { struct pedit_headers *set_masks, *add_masks, *set_vals, *add_vals; - int i, action_size, first, last, next_z; void *headers_c, *headers_v, *action, *vals_p; u32 *s_masks_p, *a_masks_p, s_mask, a_mask; struct mlx5e_tc_mod_hdr_acts *mod_acts; - struct mlx5_fields *f; unsigned long mask, field_mask; - int err; + int i, first, last, next_z; + struct mlx5_fields *f; u8 cmd; mod_acts = &parse_attr->mod_hdr_acts; @@ -2784,8 +2779,6 @@ static int offload_pedit_fields(struct mlx5e_priv *priv, set_vals = &hdrs[0].vals; add_vals = &hdrs[1].vals; - action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto); - for (i = 0; i < ARRAY_SIZE(fields); i++) { bool skip; @@ -2853,18 +2846,16 @@ static int offload_pedit_fields(struct mlx5e_priv *priv, return -EOPNOTSUPP; } - err = alloc_mod_hdr_actions(priv->mdev, namespace, mod_acts); - if (err) { + action = mlx5e_mod_hdr_alloc(priv->mdev, namespace, mod_acts); + if (IS_ERR(action)) { NL_SET_ERR_MSG_MOD(extack, "too many pedit actions, can't offload"); mlx5_core_warn(priv->mdev, "mlx5: parsed %d pedit actions, can't do more\n", mod_acts->num_actions); - return err; + return PTR_ERR(action); } - action = mod_acts->actions + - (mod_acts->num_actions * action_size); MLX5_SET(set_action_in, action, action_type, cmd); MLX5_SET(set_action_in, action, field, f->field); @@ -2894,57 +2885,6 @@ static int offload_pedit_fields(struct mlx5e_priv *priv, return 0; } -static int mlx5e_flow_namespace_max_modify_action(struct mlx5_core_dev *mdev, - int namespace) -{ - if (namespace == MLX5_FLOW_NAMESPACE_FDB) /* FDB offloading */ - return MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, max_modify_header_actions); - else /* namespace is MLX5_FLOW_NAMESPACE_KERNEL - NIC offloading */ - return MLX5_CAP_FLOWTABLE_NIC_RX(mdev, max_modify_header_actions); -} - -int alloc_mod_hdr_actions(struct mlx5_core_dev *mdev, - int namespace, - struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts) -{ - int action_size, new_num_actions, max_hw_actions; - size_t new_sz, old_sz; - void *ret; - - if (mod_hdr_acts->num_actions < mod_hdr_acts->max_actions) - return 0; - - action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto); - - max_hw_actions = mlx5e_flow_namespace_max_modify_action(mdev, - namespace); - new_num_actions = min(max_hw_actions, - mod_hdr_acts->actions ? - mod_hdr_acts->max_actions * 2 : 1); - if (mod_hdr_acts->max_actions == new_num_actions) - return -ENOSPC; - - new_sz = action_size * new_num_actions; - old_sz = mod_hdr_acts->max_actions * action_size; - ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL); - if (!ret) - return -ENOMEM; - - memset(ret + old_sz, 0, new_sz - old_sz); - mod_hdr_acts->actions = ret; - mod_hdr_acts->max_actions = new_num_actions; - - return 0; -} - -void dealloc_mod_hdr_actions(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts) -{ - kfree(mod_hdr_acts->actions); - mod_hdr_acts->actions = NULL; - mod_hdr_acts->num_actions = 0; - mod_hdr_acts->max_actions = 0; -} - static const struct pedit_headers zero_masks = {}; static int @@ -2967,7 +2907,7 @@ parse_pedit_to_modify_hdr(struct mlx5e_priv *priv, goto out_err; } - if (!mlx5e_flow_namespace_max_modify_action(priv->mdev, namespace)) { + if (!mlx5e_mod_hdr_max_actions(priv->mdev, namespace)) { NL_SET_ERR_MSG_MOD(extack, "The pedit offload action is not supported"); goto out_err; @@ -3060,7 +3000,7 @@ static int alloc_tc_pedit_action(struct mlx5e_priv *priv, int namespace, return 0; out_dealloc_parsed_actions: - dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts); + mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts); return err; } @@ -3489,7 +3429,7 @@ actions_prepare_mod_hdr_actions(struct mlx5e_priv *priv, return 0; attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; - dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts); + mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts); if (ns_type != MLX5_FLOW_NAMESPACE_FDB) return 0; @@ -4708,7 +4648,7 @@ mlx5e_add_nic_flow(struct mlx5e_priv *priv, err_free: flow_flag_set(flow, FAILED); - dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts); + mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts); mlx5e_flow_put(priv, flow); out: return err; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h index fdb222793027..eb042f0f5a41 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h @@ -247,11 +247,6 @@ int mlx5e_tc_add_flow_mod_hdr(struct mlx5e_priv *priv, struct mlx5e_tc_flow_parse_attr *parse_attr, struct mlx5e_tc_flow *flow); -int alloc_mod_hdr_actions(struct mlx5_core_dev *mdev, - int namespace, - struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); -void dealloc_mod_hdr_actions(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); - struct mlx5e_tc_flow; u32 mlx5e_tc_get_flow_tun_id(struct mlx5e_tc_flow *flow); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c index 425c91814b34..c275fe028b6d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c @@ -14,6 +14,7 @@ #include "fs_core.h" #include "esw/indir_table.h" #include "lib/fs_chains.h" +#include "en/mod_hdr.h" #define MLX5_ESW_INDIR_TABLE_SIZE 128 #define MLX5_ESW_INDIR_TABLE_RECIRC_IDX_MAX (MLX5_ESW_INDIR_TABLE_SIZE - 2) @@ -226,7 +227,7 @@ static int mlx5_esw_indir_table_rule_get(struct mlx5_eswitch *esw, goto err_handle; } - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); rule->handle = handle; rule->vni = esw_attr->rx_tun_attr->vni; rule->mh = flow_act.modify_hdr; @@ -243,7 +244,7 @@ err_table: mlx5_modify_header_dealloc(esw->dev, flow_act.modify_hdr); err_mod_hdr_alloc: err_mod_hdr_regc1: - dealloc_mod_hdr_actions(&mod_acts); + mlx5e_mod_hdr_dealloc(&mod_acts); err_mod_hdr_regc0: err_ethertype: kfree(rule); From 1cfd3490f2780edb3f1b2f3f64c5c80c67c64075 Mon Sep 17 00:00:00 2001 From: Paul Blakey Date: Wed, 25 Aug 2021 16:46:41 +0300 Subject: [PATCH 0257/2048] net/mlx5: CT: Allow static allocation of mod headers As each CT rule uses at least 4 modify header actions, each rule causes at least 3 reallocations by the mod header actions api. Allow initial static allocation of the mod acts array, and use it for CT rules. If the static allocation is exceeded go back to dynamic allocation. Signed-off-by: Saeed Mahameed Signed-off-by: Paul Blakey Reviewed-by: Oz Shlomo Reviewed-by: Roi Dayan --- .../ethernet/mellanox/mlx5/core/en/mod_hdr.c | 17 ++++++++++++++--- .../ethernet/mellanox/mlx5/core/en/mod_hdr.h | 13 +++++++++++++ .../net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 9 ++++++++- 3 files changed, 35 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c index 19d05fb4aab2..17325c5d6516 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c @@ -176,11 +176,20 @@ mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace, new_sz = MLX5_MH_ACT_SZ * new_num_actions; old_sz = mod_hdr_acts->max_actions * MLX5_MH_ACT_SZ; - ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL); + if (mod_hdr_acts->is_static) { + ret = kzalloc(new_sz, GFP_KERNEL); + if (ret) { + memcpy(ret, mod_hdr_acts->actions, old_sz); + mod_hdr_acts->is_static = false; + } + } else { + ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL); + if (ret) + memset(ret + old_sz, 0, new_sz - old_sz); + } if (!ret) return ERR_PTR(-ENOMEM); - memset(ret + old_sz, 0, new_sz - old_sz); mod_hdr_acts->actions = ret; mod_hdr_acts->max_actions = new_num_actions; @@ -191,7 +200,9 @@ out: void mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts) { - kfree(mod_hdr_acts->actions); + if (!mod_hdr_acts->is_static) + kfree(mod_hdr_acts->actions); + mod_hdr_acts->actions = NULL; mod_hdr_acts->num_actions = 0; mod_hdr_acts->max_actions = 0; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h index b8cd1a7a31be..b8dac418d0a5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h @@ -7,14 +7,27 @@ #include #include +#define MLX5_MH_ACT_SZ MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto) + struct mlx5e_mod_hdr_handle; struct mlx5e_tc_mod_hdr_acts { int num_actions; int max_actions; + bool is_static; void *actions; }; +#define DECLARE_MOD_HDR_ACTS_ACTIONS(name, len) \ + u8 name[len][MLX5_MH_ACT_SZ] = {} + +#define DECLARE_MOD_HDR_ACTS(name, acts_arr) \ + struct mlx5e_tc_mod_hdr_acts name = { \ + .max_actions = ARRAY_SIZE(acts_arr), \ + .is_static = true, \ + .actions = acts_arr, \ + } + char *mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace, struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); void mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index 89065fac7590..f89a4c7a4f71 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -36,6 +36,12 @@ #define MLX5_CT_LABELS_BITS (mlx5e_tc_attr_to_reg_mappings[LABELS_TO_REG].mlen) #define MLX5_CT_LABELS_MASK GENMASK(MLX5_CT_LABELS_BITS - 1, 0) +/* Statically allocate modify actions for + * ipv6 and port nat (5) + tuple fields (4) + nic mode zone restore (1) = 10. + * This will be increased dynamically if needed (for the ipv6 snat + dnat). + */ +#define MLX5_CT_MIN_MOD_ACTS 10 + #define ct_dbg(fmt, args...)\ netdev_dbg(ct_priv->netdev, "ct_debug: " fmt "\n", ##args) @@ -645,7 +651,8 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv, struct mlx5e_mod_hdr_handle **mh, u8 zone_restore_id, bool nat) { - struct mlx5e_tc_mod_hdr_acts mod_acts = {}; + DECLARE_MOD_HDR_ACTS_ACTIONS(actions_arr, MLX5_CT_MIN_MOD_ACTS); + DECLARE_MOD_HDR_ACTS(mod_acts, actions_arr); struct flow_action_entry *meta; u16 ct_state = 0; int err; From 0164a9bd9d63c56925cf0aae2731730778fae3f6 Mon Sep 17 00:00:00 2001 From: Yihao Han Date: Tue, 2 Nov 2021 23:21:09 -0700 Subject: [PATCH 0258/2048] net/mlx5: TC, using swap() instead of tmp variable swap() was used instead of the tmp variable to swap values Signed-off-by: Yihao Han Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index f89a4c7a4f71..9a31f45a9d9b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -907,12 +907,9 @@ mlx5_tc_ct_shared_counter_get(struct mlx5_tc_ct_priv *ct_priv, struct mlx5_ct_tuple rev_tuple = entry->tuple; struct mlx5_ct_counter *shared_counter; struct mlx5_ct_entry *rev_entry; - __be16 tmp_port; /* get the reversed tuple */ - tmp_port = rev_tuple.port.src; - rev_tuple.port.src = rev_tuple.port.dst; - rev_tuple.port.dst = tmp_port; + swap(rev_tuple.port.src, rev_tuple.port.dst); if (rev_tuple.addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) { __be32 tmp_addr = rev_tuple.ip.src_v4; From 972fe492e847c7ed9679a4b4aa79f9fe06b9ae7b Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Mon, 1 Nov 2021 18:02:00 +0200 Subject: [PATCH 0259/2048] net/mlx5e: TC, Destroy nic flow counter if exists Counter is only added if counter flag exists. So check the counter fag exists for deleting the counter. This is the same as in add/del fdb flow. Signed-off-by: Roi Dayan Reviewed-by: Vlad Buslov Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index e620100eabe0..3e542b030fc1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -1133,7 +1133,8 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) mlx5e_detach_mod_hdr(priv, flow); - mlx5_fc_destroy(priv->mdev, attr->counter); + if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) + mlx5_fc_destroy(priv->mdev, attr->counter); if (flow_flag_test(flow, HAIRPIN)) mlx5e_hairpin_flow_del(priv, flow); From 88d974860412f2ca337086aa25e12b1094d55c6c Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Mon, 1 Nov 2021 18:13:02 +0200 Subject: [PATCH 0260/2048] net/mlx5e: TC, Move kfree() calls after destroying all resources When deleting fdb/nic flow rules first release all resources and then call the kfree() calls instead of sparse them around the function. Signed-off-by: Roi Dayan Reviewed-by: Vlad Buslov Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 3e542b030fc1..aa4da8d1e252 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -1128,8 +1128,6 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, } mutex_unlock(&priv->fs.tc.t_lock); - kvfree(attr->parse_attr); - if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) mlx5e_detach_mod_hdr(priv, flow); @@ -1139,6 +1137,7 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, if (flow_flag_test(flow, HAIRPIN)) mlx5e_hairpin_flow_del(priv, flow); + kvfree(attr->parse_attr); kfree(flow->attr); } @@ -1626,9 +1625,6 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv, else mlx5e_detach_mod_hdr(priv, flow); } - kfree(attr->sample_attr); - kvfree(attr->parse_attr); - kvfree(attr->esw_attr->rx_tun_attr); if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) mlx5_fc_destroy(esw_attr->counter_dev, attr->counter); @@ -1642,6 +1638,9 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv, if (flow_flag_test(flow, L3_TO_L2_DECAP)) mlx5e_detach_decap(priv, flow); + kfree(attr->sample_attr); + kvfree(attr->esw_attr->rx_tun_attr); + kvfree(attr->parse_attr); kfree(flow->attr); } From fc3a879aea35fba713ae2d0fd42f2f2e7eaac1d9 Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Wed, 10 Nov 2021 16:19:41 +0200 Subject: [PATCH 0261/2048] net/mlx5e: TC, Move comment about mod header flag to correct place Move the comment to the correct place where the driver actually removes the flag and not in the check that maybe pedit actions exists. Signed-off-by: Roi Dayan Reviewed-by: Maor Dickman Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index aa4da8d1e252..686bb2e08e9e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -3424,10 +3424,10 @@ actions_prepare_mod_hdr_actions(struct mlx5e_priv *priv, if (err) return err; - /* In case all pedit actions are skipped, remove the MOD_HDR flag. */ if (parse_attr->mod_hdr_acts.num_actions > 0) return 0; + /* In case all pedit actions are skipped, remove the MOD_HDR flag. */ attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts); From 819c319c8c919307fa764ff02a228c06e727501c Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Tue, 26 Oct 2021 17:08:24 +0800 Subject: [PATCH 0262/2048] net/mlx5e: Specify out ifindex when looking up decap route There is a use case that the local and remote VTEPs are in the same host. Currently, the out ifindex is not specified when looking up the decap route for offloads. So in this case, a local route is returned and the route dev is lo. Actual tunnel interface can be created with a parameter "dev" [1], which specifies the physical device to use for tunnel endpoint communication. Pass this parameter to driver when looking up decap route for offloads. So that a unicast route will be returned. [1] ip link add name vxlan1 type vxlan id 100 dev enp4s0f0 remote 1.1.1.1 dstport 4789 Signed-off-by: Chris Mi Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/en/tc_tun.c | 23 ++++++++++--------- .../ethernet/mellanox/mlx5/core/en/tc_tun.h | 3 ++- .../mellanox/mlx5/core/en/tc_tun_encap.c | 4 ++-- 3 files changed, 16 insertions(+), 14 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c index a5e450973225..33815246fead 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c @@ -103,7 +103,7 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv, } static int mlx5e_route_lookup_ipv4_get(struct mlx5e_priv *priv, - struct net_device *mirred_dev, + struct net_device *dev, struct mlx5e_tc_tun_route_attr *attr) { struct net_device *route_dev; @@ -122,13 +122,13 @@ static int mlx5e_route_lookup_ipv4_get(struct mlx5e_priv *priv, uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH); attr->fl.fl4.flowi4_oif = uplink_dev->ifindex; } else { - struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(mirred_dev); + struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(dev); if (tunnel && tunnel->get_remote_ifindex) - attr->fl.fl4.flowi4_oif = tunnel->get_remote_ifindex(mirred_dev); + attr->fl.fl4.flowi4_oif = tunnel->get_remote_ifindex(dev); } - rt = ip_route_output_key(dev_net(mirred_dev), &attr->fl.fl4); + rt = ip_route_output_key(dev_net(dev), &attr->fl.fl4); if (IS_ERR(rt)) return PTR_ERR(rt); @@ -440,10 +440,10 @@ release_neigh: #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6) static int mlx5e_route_lookup_ipv6_get(struct mlx5e_priv *priv, - struct net_device *mirred_dev, + struct net_device *dev, struct mlx5e_tc_tun_route_attr *attr) { - struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(mirred_dev); + struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(dev); struct net_device *route_dev; struct net_device *out_dev; struct dst_entry *dst; @@ -451,8 +451,8 @@ static int mlx5e_route_lookup_ipv6_get(struct mlx5e_priv *priv, int ret; if (tunnel && tunnel->get_remote_ifindex) - attr->fl.fl6.flowi6_oif = tunnel->get_remote_ifindex(mirred_dev); - dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(mirred_dev), NULL, &attr->fl.fl6, + attr->fl.fl6.flowi6_oif = tunnel->get_remote_ifindex(dev); + dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(dev), NULL, &attr->fl.fl6, NULL); if (IS_ERR(dst)) return PTR_ERR(dst); @@ -708,7 +708,8 @@ release_neigh: int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv, struct mlx5_flow_spec *spec, - struct mlx5_flow_attr *flow_attr) + struct mlx5_flow_attr *flow_attr, + struct net_device *filter_dev) { struct mlx5_esw_flow_attr *esw_attr = flow_attr->esw_attr; struct mlx5e_tc_int_port *int_port; @@ -720,14 +721,14 @@ int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv, /* Addresses are swapped for decap */ attr.fl.fl4.saddr = esw_attr->rx_tun_attr->dst_ip.v4; attr.fl.fl4.daddr = esw_attr->rx_tun_attr->src_ip.v4; - err = mlx5e_route_lookup_ipv4_get(priv, priv->netdev, &attr); + err = mlx5e_route_lookup_ipv4_get(priv, filter_dev, &attr); } #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6) else if (flow_attr->tun_ip_version == 6) { /* Addresses are swapped for decap */ attr.fl.fl6.saddr = esw_attr->rx_tun_attr->dst_ip.v6; attr.fl.fl6.daddr = esw_attr->rx_tun_attr->src_ip.v6; - err = mlx5e_route_lookup_ipv6_get(priv, priv->netdev, &attr); + err = mlx5e_route_lookup_ipv6_get(priv, filter_dev, &attr); } #endif else diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h index aa092eaeaec3..b38f693bbb52 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h @@ -94,7 +94,8 @@ mlx5e_tc_tun_update_header_ipv6(struct mlx5e_priv *priv, #endif int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv, struct mlx5_flow_spec *spec, - struct mlx5_flow_attr *attr); + struct mlx5_flow_attr *attr, + struct net_device *filter_dev); bool mlx5e_tc_tun_device_to_offload(struct mlx5e_priv *priv, struct net_device *netdev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c index 660cca73c36c..de16bbc08679 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c @@ -1153,7 +1153,7 @@ int mlx5e_attach_decap_route(struct mlx5e_priv *priv, tbl_time_before = mlx5e_route_tbl_get_last_update(priv); tbl_time_after = tbl_time_before; - err = mlx5e_tc_tun_route_lookup(priv, &parse_attr->spec, attr); + err = mlx5e_tc_tun_route_lookup(priv, &parse_attr->spec, attr, parse_attr->filter_dev); if (err || !esw_attr->rx_tun_attr->decap_vport) goto out; @@ -1474,7 +1474,7 @@ static void mlx5e_reoffload_decap(struct mlx5e_priv *priv, parse_attr = attr->parse_attr; spec = &parse_attr->spec; - err = mlx5e_tc_tun_route_lookup(priv, spec, attr); + err = mlx5e_tc_tun_route_lookup(priv, spec, attr, parse_attr->filter_dev); if (err) { mlx5_core_warn(priv->mdev, "Failed to lookup route for flow, %d\n", err); From fcf8ec54b0477293ecf5a4a01bfe88d5dea6c8c0 Mon Sep 17 00:00:00 2001 From: Parav Pandit Date: Wed, 20 Oct 2021 07:56:01 +0300 Subject: [PATCH 0263/2048] net/mlx5: E-switch, Remove vport enabled check An eswitch vport of the devlink port is always enabled before a devlink port is registered. And a eswitch vport is always disabled after a devlink port is unregistered. Hence avoid the vport enabled check in the devlink callback routine. Such check is only applicable in the legacy SR-IOV callbacks. Signed-off-by: Parav Pandit Reviewed-by: Sunil Sudhakar Rani Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/eswitch.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index ec136b499204..b039f8b07d31 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1704,7 +1704,6 @@ int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port, { struct mlx5_eswitch *esw; struct mlx5_vport *vport; - int err = -EOPNOTSUPP; u16 vport_num; esw = mlx5_devlink_eswitch_get(port->devlink); @@ -1722,13 +1721,10 @@ int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port, } mutex_lock(&esw->state_lock); - if (vport->enabled) { - ether_addr_copy(hw_addr, vport->info.mac); - *hw_addr_len = ETH_ALEN; - err = 0; - } + ether_addr_copy(hw_addr, vport->info.mac); + *hw_addr_len = ETH_ALEN; mutex_unlock(&esw->state_lock); - return err; + return 0; } int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, @@ -1737,8 +1733,8 @@ int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, { struct mlx5_eswitch *esw; struct mlx5_vport *vport; - int err = -EOPNOTSUPP; u16 vport_num; + int err; esw = mlx5_devlink_eswitch_get(port->devlink); if (IS_ERR(esw)) { @@ -1758,10 +1754,7 @@ int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, } mutex_lock(&esw->state_lock); - if (vport->enabled) - err = mlx5_esw_set_vport_mac_locked(esw, vport, hw_addr); - else - NL_SET_ERR_MSG_MOD(extack, "Eswitch vport is disabled"); + err = mlx5_esw_set_vport_mac_locked(esw, vport, hw_addr); mutex_unlock(&esw->state_lock); return err; } From b22fd4381d15ae66e42a7fab6eedfbb72d714ebb Mon Sep 17 00:00:00 2001 From: Parav Pandit Date: Thu, 21 Oct 2021 18:17:52 +0300 Subject: [PATCH 0264/2048] net/mlx5: E-switch, Reuse mlx5_eswitch_set_vport_mac mlx5_eswitch_set_vport_mac() routine already does necessary checks which are duplicated in implementation of mlx5_devlink_port_function_hw_addr_set(). Hence, reuse mlx5_eswitch_set_vport_mac() and cut down the code. Signed-off-by: Parav Pandit Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 12 +----------- 1 file changed, 1 insertion(+), 11 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index b039f8b07d31..c0526fc27ad6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1732,9 +1732,7 @@ int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, struct netlink_ext_ack *extack) { struct mlx5_eswitch *esw; - struct mlx5_vport *vport; u16 vport_num; - int err; esw = mlx5_devlink_eswitch_get(port->devlink); if (IS_ERR(esw)) { @@ -1747,16 +1745,8 @@ int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, NL_SET_ERR_MSG_MOD(extack, "Port doesn't support set hw_addr"); return -EINVAL; } - vport = mlx5_eswitch_get_vport(esw, vport_num); - if (IS_ERR(vport)) { - NL_SET_ERR_MSG_MOD(extack, "Invalid port"); - return PTR_ERR(vport); - } - mutex_lock(&esw->state_lock); - err = mlx5_esw_set_vport_mac_locked(esw, vport, hw_addr); - mutex_unlock(&esw->state_lock); - return err; + return mlx5_eswitch_set_vport_mac(esw, vport_num, hw_addr); } int mlx5_eswitch_set_vport_state(struct mlx5_eswitch *esw, From e9d491a64755d11812b532fcf10b93b2f9535e45 Mon Sep 17 00:00:00 2001 From: Parav Pandit Date: Thu, 21 Oct 2021 18:21:30 +0300 Subject: [PATCH 0265/2048] net/mlx5: E-switch, move offloads mode callbacks to offloads file eswitch.c is mainly for common code between legacy and offloads mode. MAC address get and set via devlink is applicable only in offloads mode. Hence, move it to eswitch_offloads.c file. Signed-off-by: Parav Pandit Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/eswitch.c | 59 ------------------- .../mellanox/mlx5/core/eswitch_offloads.c | 59 +++++++++++++++++++ 2 files changed, 59 insertions(+), 59 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index c0526fc27ad6..ec5b1641d40c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1690,65 +1690,6 @@ bool mlx5_esw_is_sf_vport(struct mlx5_eswitch *esw, u16 vport_num) return mlx5_esw_check_port_type(esw, vport_num, MLX5_ESW_VPT_SF); } -static bool -is_port_function_supported(struct mlx5_eswitch *esw, u16 vport_num) -{ - return vport_num == MLX5_VPORT_PF || - mlx5_eswitch_is_vf_vport(esw, vport_num) || - mlx5_esw_is_sf_vport(esw, vport_num); -} - -int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port, - u8 *hw_addr, int *hw_addr_len, - struct netlink_ext_ack *extack) -{ - struct mlx5_eswitch *esw; - struct mlx5_vport *vport; - u16 vport_num; - - esw = mlx5_devlink_eswitch_get(port->devlink); - if (IS_ERR(esw)) - return PTR_ERR(esw); - - vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index); - if (!is_port_function_supported(esw, vport_num)) - return -EOPNOTSUPP; - - vport = mlx5_eswitch_get_vport(esw, vport_num); - if (IS_ERR(vport)) { - NL_SET_ERR_MSG_MOD(extack, "Invalid port"); - return PTR_ERR(vport); - } - - mutex_lock(&esw->state_lock); - ether_addr_copy(hw_addr, vport->info.mac); - *hw_addr_len = ETH_ALEN; - mutex_unlock(&esw->state_lock); - return 0; -} - -int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, - const u8 *hw_addr, int hw_addr_len, - struct netlink_ext_ack *extack) -{ - struct mlx5_eswitch *esw; - u16 vport_num; - - esw = mlx5_devlink_eswitch_get(port->devlink); - if (IS_ERR(esw)) { - NL_SET_ERR_MSG_MOD(extack, "Eswitch doesn't support set hw_addr"); - return PTR_ERR(esw); - } - - vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index); - if (!is_port_function_supported(esw, vport_num)) { - NL_SET_ERR_MSG_MOD(extack, "Port doesn't support set hw_addr"); - return -EINVAL; - } - - return mlx5_eswitch_set_vport_mac(esw, vport_num, hw_addr); -} - int mlx5_eswitch_set_vport_state(struct mlx5_eswitch *esw, u16 vport, int link_state) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index f4eaa5893886..4bd502ae82b6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -3862,3 +3862,62 @@ u32 mlx5_eswitch_get_vport_metadata_for_set(struct mlx5_eswitch *esw, return vport->metadata; } EXPORT_SYMBOL(mlx5_eswitch_get_vport_metadata_for_set); + +static bool +is_port_function_supported(struct mlx5_eswitch *esw, u16 vport_num) +{ + return vport_num == MLX5_VPORT_PF || + mlx5_eswitch_is_vf_vport(esw, vport_num) || + mlx5_esw_is_sf_vport(esw, vport_num); +} + +int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port, + u8 *hw_addr, int *hw_addr_len, + struct netlink_ext_ack *extack) +{ + struct mlx5_eswitch *esw; + struct mlx5_vport *vport; + u16 vport_num; + + esw = mlx5_devlink_eswitch_get(port->devlink); + if (IS_ERR(esw)) + return PTR_ERR(esw); + + vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index); + if (!is_port_function_supported(esw, vport_num)) + return -EOPNOTSUPP; + + vport = mlx5_eswitch_get_vport(esw, vport_num); + if (IS_ERR(vport)) { + NL_SET_ERR_MSG_MOD(extack, "Invalid port"); + return PTR_ERR(vport); + } + + mutex_lock(&esw->state_lock); + ether_addr_copy(hw_addr, vport->info.mac); + *hw_addr_len = ETH_ALEN; + mutex_unlock(&esw->state_lock); + return 0; +} + +int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port, + const u8 *hw_addr, int hw_addr_len, + struct netlink_ext_ack *extack) +{ + struct mlx5_eswitch *esw; + u16 vport_num; + + esw = mlx5_devlink_eswitch_get(port->devlink); + if (IS_ERR(esw)) { + NL_SET_ERR_MSG_MOD(extack, "Eswitch doesn't support set hw_addr"); + return PTR_ERR(esw); + } + + vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index); + if (!is_port_function_supported(esw, vport_num)) { + NL_SET_ERR_MSG_MOD(extack, "Port doesn't support set hw_addr"); + return -EINVAL; + } + + return mlx5_eswitch_set_vport_mac(esw, vport_num, hw_addr); +} From d7df09f5e7b46af0eb927c065113faa57411d100 Mon Sep 17 00:00:00 2001 From: Dmytro Linkin Date: Tue, 21 Sep 2021 18:45:42 +0300 Subject: [PATCH 0266/2048] net/mlx5: E-switch, Enable vport QoS on demand Vports' QoS is not commonly used but consume SW/HW resources, which becomes an issue on BlueField SoC systems. Don't enable QoS on vports by default on eswitch mode change and enable when it's going to be used by one of the top level users: - configuring TC matchall filter with police action; - setting rate with legacy NDO API; - calling devlink ops->rate_leaf_*() callbacks. Disable vport QoS on vport cleanup. Signed-off-by: Dmytro Linkin Reviewed-by: Parav Pandit Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../ethernet/mellanox/mlx5/core/esw/legacy.c | 4 +- .../net/ethernet/mellanox/mlx5/core/esw/qos.c | 80 +++++++++++++------ .../net/ethernet/mellanox/mlx5/core/esw/qos.h | 12 +-- .../net/ethernet/mellanox/mlx5/core/eswitch.c | 9 +-- 4 files changed, 63 insertions(+), 42 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c index df277a6cddc0..2b52f7c09152 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c @@ -522,9 +522,7 @@ int mlx5_eswitch_set_vport_rate(struct mlx5_eswitch *esw, u16 vport, return PTR_ERR(evport); mutex_lock(&esw->state_lock); - err = mlx5_esw_qos_set_vport_min_rate(esw, evport, min_rate, NULL); - if (!err) - err = mlx5_esw_qos_set_vport_max_rate(esw, evport, max_rate, NULL); + err = mlx5_esw_qos_set_vport_rate(esw, evport, max_rate, min_rate); mutex_unlock(&esw->state_lock); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index c6cc67cb4f6a..304abc293086 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -204,10 +204,8 @@ static int esw_qos_normalize_groups_min_rate(struct mlx5_eswitch *esw, u32 divid return 0; } -int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw, - struct mlx5_vport *evport, - u32 min_rate, - struct netlink_ext_ack *extack) +static int esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport, + u32 min_rate, struct netlink_ext_ack *extack) { u32 fw_max_bw_share, previous_min_rate; bool min_rate_supported; @@ -231,10 +229,8 @@ int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw, return err; } -int mlx5_esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw, - struct mlx5_vport *evport, - u32 max_rate, - struct netlink_ext_ack *extack) +static int esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport, + u32 max_rate, struct netlink_ext_ack *extack) { u32 act_max_rate = max_rate; bool max_rate_supported; @@ -605,8 +601,8 @@ unlock: mutex_unlock(&esw->state_lock); } -int mlx5_esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport, - u32 max_rate, u32 bw_share) +static int esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport, + u32 max_rate, u32 bw_share) { int err; @@ -615,7 +611,7 @@ int mlx5_esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport return 0; if (vport->qos.enabled) - return -EEXIST; + return 0; vport->qos.group = esw->qos.group0; @@ -645,31 +641,55 @@ void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vpo esw_warn(esw->dev, "E-Switch destroy TSAR vport element failed (vport=%d,err=%d)\n", vport->vport, err); - vport->qos.enabled = false; + memset(&vport->qos, 0, sizeof(vport->qos)); trace_mlx5_esw_vport_qos_destroy(vport); } +int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *vport, + u32 min_rate, u32 max_rate) +{ + int err; + + lockdep_assert_held(&esw->state_lock); + err = esw_qos_vport_enable(esw, vport, 0, 0); + if (err) + return err; + + err = esw_qos_set_vport_min_rate(esw, vport, min_rate, NULL); + if (!err) + err = esw_qos_set_vport_max_rate(esw, vport, max_rate, NULL); + + return err; +} + int mlx5_esw_qos_modify_vport_rate(struct mlx5_eswitch *esw, u16 vport_num, u32 rate_mbps) { u32 ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_vport *vport; u32 bitmask; + int err; vport = mlx5_eswitch_get_vport(esw, vport_num); if (IS_ERR(vport)) return PTR_ERR(vport); - if (!vport->qos.enabled) - return -EOPNOTSUPP; + mutex_lock(&esw->state_lock); + if (!vport->qos.enabled) { + /* Eswitch QoS wasn't enabled yet. Enable it and vport QoS. */ + err = esw_qos_vport_enable(esw, vport, rate_mbps, vport->qos.bw_share); + } else { + MLX5_SET(scheduling_context, ctx, max_average_bw, rate_mbps); - MLX5_SET(scheduling_context, ctx, max_average_bw, rate_mbps); - bitmask = MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_MAX_AVERAGE_BW; + bitmask = MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_MAX_AVERAGE_BW; + err = mlx5_modify_scheduling_element_cmd(esw->dev, + SCHEDULING_HIERARCHY_E_SWITCH, + ctx, + vport->qos.esw_tsar_ix, + bitmask); + } + mutex_unlock(&esw->state_lock); - return mlx5_modify_scheduling_element_cmd(esw->dev, - SCHEDULING_HIERARCHY_E_SWITCH, - ctx, - vport->qos.esw_tsar_ix, - bitmask); + return err; } #define MLX5_LINKSPEED_UNIT 125000 /* 1Mbps in Bps */ @@ -728,7 +748,12 @@ int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void return err; mutex_lock(&esw->state_lock); - err = mlx5_esw_qos_set_vport_min_rate(esw, vport, tx_share, extack); + err = esw_qos_vport_enable(esw, vport, 0, 0); + if (err) + goto unlock; + + err = esw_qos_set_vport_min_rate(esw, vport, tx_share, extack); +unlock: mutex_unlock(&esw->state_lock); return err; } @@ -749,7 +774,12 @@ int mlx5_esw_devlink_rate_leaf_tx_max_set(struct devlink_rate *rate_leaf, void * return err; mutex_lock(&esw->state_lock); - err = mlx5_esw_qos_set_vport_max_rate(esw, vport, tx_max, extack); + err = esw_qos_vport_enable(esw, vport, 0, 0); + if (err) + goto unlock; + + err = esw_qos_set_vport_max_rate(esw, vport, tx_max, extack); +unlock: mutex_unlock(&esw->state_lock); return err; } @@ -846,7 +876,9 @@ int mlx5_esw_qos_vport_update_group(struct mlx5_eswitch *esw, int err; mutex_lock(&esw->state_lock); - err = esw_qos_vport_update_group(esw, vport, group, extack); + err = esw_qos_vport_enable(esw, vport, 0, 0); + if (!err) + err = esw_qos_vport_update_group(esw, vport, group, extack); mutex_unlock(&esw->state_lock); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h index 28451abe2d2f..91b66c1b9881 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h @@ -6,18 +6,10 @@ #ifdef CONFIG_MLX5_ESWITCH -int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw, - struct mlx5_vport *evport, - u32 min_rate, - struct netlink_ext_ack *extack); -int mlx5_esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw, - struct mlx5_vport *evport, - u32 max_rate, - struct netlink_ext_ack *extack); +int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport, + u32 max_rate, u32 min_rate); void mlx5_esw_qos_create(struct mlx5_eswitch *esw); void mlx5_esw_qos_destroy(struct mlx5_eswitch *esw); -int mlx5_esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport, - u32 max_rate, u32 bw_share); void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vport); int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void *priv, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index ec5b1641d40c..2d188f462028 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -781,9 +781,6 @@ static int esw_vport_setup(struct mlx5_eswitch *esw, struct mlx5_vport *vport) if (err) return err; - /* Attach vport to the eswitch rate limiter */ - mlx5_esw_qos_vport_enable(esw, vport, vport->qos.max_rate, vport->qos.bw_share); - if (mlx5_esw_is_manager_vport(esw, vport_num)) return 0; @@ -1746,8 +1743,10 @@ int mlx5_eswitch_get_vport_config(struct mlx5_eswitch *esw, ivi->qos = evport->info.qos; ivi->spoofchk = evport->info.spoofchk; ivi->trusted = evport->info.trusted; - ivi->min_tx_rate = evport->qos.min_rate; - ivi->max_tx_rate = evport->qos.max_rate; + if (evport->qos.enabled) { + ivi->min_tx_rate = evport->qos.min_rate; + ivi->max_tx_rate = evport->qos.max_rate; + } mutex_unlock(&esw->state_lock); return 0; From 85c5f7c9200e5ce89f0c188d0c24ab4e731b6a51 Mon Sep 17 00:00:00 2001 From: Dmytro Linkin Date: Tue, 21 Sep 2021 19:08:38 +0300 Subject: [PATCH 0267/2048] net/mlx5: E-switch, Create QoS on demand Don't create eswitch QoS (root TSAR) on switch mode change. Create it on first child TSAR object creation - vport or rate group. Keep track root TSAR references and release root TSAR with last object deletion. No need to check for QoS is enabled when installing tc matchall filter. Remove related helper function due to no users of it. Signed-off-by: Dmytro Linkin Reviewed-by: Parav Pandit Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 6 - .../net/ethernet/mellanox/mlx5/core/esw/qos.c | 152 ++++++++++++------ .../net/ethernet/mellanox/mlx5/core/esw/qos.h | 2 - .../net/ethernet/mellanox/mlx5/core/eswitch.c | 9 +- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 11 +- 5 files changed, 111 insertions(+), 69 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 686bb2e08e9e..55e384abd364 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -4948,14 +4948,8 @@ static int scan_tc_matchall_fdb_actions(struct mlx5e_priv *priv, int mlx5e_tc_configure_matchall(struct mlx5e_priv *priv, struct tc_cls_matchall_offload *ma) { - struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; struct netlink_ext_ack *extack = ma->common.extack; - if (!mlx5_esw_qos_enabled(esw)) { - NL_SET_ERR_MSG_MOD(extack, "QoS is not supported on this device"); - return -EOPNOTSUPP; - } - if (ma->common.prio != 1) { NL_SET_ERR_MSG_MOD(extack, "only priority 1 is supported"); return -EINVAL; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index 304abc293086..ff0a07a91992 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -428,16 +428,13 @@ static int esw_qos_vport_update_group(struct mlx5_eswitch *esw, } static struct mlx5_esw_rate_group * -esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack) +__esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack) { u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_esw_rate_group *group; u32 divider; int err; - if (!MLX5_CAP_QOS(esw->dev, log_esw_max_sched_depth)) - return ERR_PTR(-EOPNOTSUPP); - group = kzalloc(sizeof(*group), GFP_KERNEL); if (!group) return ERR_PTR(-ENOMEM); @@ -478,9 +475,32 @@ err_sched_elem: return ERR_PTR(err); } -static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw, - struct mlx5_esw_rate_group *group, - struct netlink_ext_ack *extack) +static int esw_qos_get(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack); +static void esw_qos_put(struct mlx5_eswitch *esw); + +static struct mlx5_esw_rate_group * +esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack) +{ + struct mlx5_esw_rate_group *group; + int err; + + if (!MLX5_CAP_QOS(esw->dev, log_esw_max_sched_depth)) + return ERR_PTR(-EOPNOTSUPP); + + err = esw_qos_get(esw, extack); + if (err) + return ERR_PTR(err); + + group = __esw_qos_create_rate_group(esw, extack); + if (IS_ERR(group)) + esw_qos_put(esw); + + return group; +} + +static int __esw_qos_destroy_rate_group(struct mlx5_eswitch *esw, + struct mlx5_esw_rate_group *group, + struct netlink_ext_ack *extack) { u32 divider; int err; @@ -499,7 +519,21 @@ static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw, NL_SET_ERR_MSG_MOD(extack, "E-Switch destroy TSAR_ID failed"); trace_mlx5_esw_group_qos_destroy(esw->dev, group, group->tsar_ix); + kfree(group); + + return err; +} + +static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw, + struct mlx5_esw_rate_group *group, + struct netlink_ext_ack *extack) +{ + int err; + + err = __esw_qos_destroy_rate_group(esw, group, extack); + esw_qos_put(esw); + return err; } @@ -522,7 +556,7 @@ static bool esw_qos_element_type_supported(struct mlx5_core_dev *dev, int type) return false; } -void mlx5_esw_qos_create(struct mlx5_eswitch *esw) +static int esw_qos_create(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack) { u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_core_dev *dev = esw->dev; @@ -530,14 +564,10 @@ void mlx5_esw_qos_create(struct mlx5_eswitch *esw) int err; if (!MLX5_CAP_GEN(dev, qos) || !MLX5_CAP_QOS(dev, esw_scheduling)) - return; + return -EOPNOTSUPP; if (!esw_qos_element_type_supported(dev, SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR)) - return; - - mutex_lock(&esw->state_lock); - if (esw->qos.enabled) - goto unlock; + return -EOPNOTSUPP; MLX5_SET(scheduling_context, tsar_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR); @@ -551,75 +581,93 @@ void mlx5_esw_qos_create(struct mlx5_eswitch *esw) &esw->qos.root_tsar_ix); if (err) { esw_warn(dev, "E-Switch create root TSAR failed (%d)\n", err); - goto unlock; + return err; } INIT_LIST_HEAD(&esw->qos.groups); if (MLX5_CAP_QOS(dev, log_esw_max_sched_depth)) { - esw->qos.group0 = esw_qos_create_rate_group(esw, NULL); + esw->qos.group0 = __esw_qos_create_rate_group(esw, extack); if (IS_ERR(esw->qos.group0)) { esw_warn(dev, "E-Switch create rate group 0 failed (%ld)\n", PTR_ERR(esw->qos.group0)); goto err_group0; } } - esw->qos.enabled = true; -unlock: - mutex_unlock(&esw->state_lock); - return; + refcount_set(&esw->qos.refcnt, 1); + + return 0; err_group0: - err = mlx5_destroy_scheduling_element_cmd(esw->dev, - SCHEDULING_HIERARCHY_E_SWITCH, - esw->qos.root_tsar_ix); - if (err) - esw_warn(esw->dev, "E-Switch destroy root TSAR failed (%d)\n", err); - mutex_unlock(&esw->state_lock); + if (mlx5_destroy_scheduling_element_cmd(esw->dev, SCHEDULING_HIERARCHY_E_SWITCH, + esw->qos.root_tsar_ix)) + esw_warn(esw->dev, "E-Switch destroy root TSAR failed.\n"); + + return err; } -void mlx5_esw_qos_destroy(struct mlx5_eswitch *esw) +static void esw_qos_destroy(struct mlx5_eswitch *esw) { - struct devlink *devlink = priv_to_devlink(esw->dev); int err; - devlink_rate_nodes_destroy(devlink); - mutex_lock(&esw->state_lock); - if (!esw->qos.enabled) - goto unlock; - if (esw->qos.group0) - esw_qos_destroy_rate_group(esw, esw->qos.group0, NULL); + __esw_qos_destroy_rate_group(esw, esw->qos.group0, NULL); err = mlx5_destroy_scheduling_element_cmd(esw->dev, SCHEDULING_HIERARCHY_E_SWITCH, esw->qos.root_tsar_ix); if (err) esw_warn(esw->dev, "E-Switch destroy root TSAR failed (%d)\n", err); +} - esw->qos.enabled = false; -unlock: - mutex_unlock(&esw->state_lock); +static int esw_qos_get(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack) +{ + int err = 0; + + lockdep_assert_held(&esw->state_lock); + + if (!refcount_inc_not_zero(&esw->qos.refcnt)) { + /* esw_qos_create() set refcount to 1 only on success. + * No need to decrement on failure. + */ + err = esw_qos_create(esw, extack); + } + + return err; +} + +static void esw_qos_put(struct mlx5_eswitch *esw) +{ + lockdep_assert_held(&esw->state_lock); + if (refcount_dec_and_test(&esw->qos.refcnt)) + esw_qos_destroy(esw); } static int esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport, - u32 max_rate, u32 bw_share) + u32 max_rate, u32 bw_share, struct netlink_ext_ack *extack) { int err; lockdep_assert_held(&esw->state_lock); - if (!esw->qos.enabled) - return 0; - if (vport->qos.enabled) return 0; + err = esw_qos_get(esw, extack); + if (err) + return err; + vport->qos.group = esw->qos.group0; err = esw_qos_vport_create_sched_element(esw, vport, max_rate, bw_share); - if (!err) { - vport->qos.enabled = true; - trace_mlx5_esw_vport_qos_create(vport, bw_share, max_rate); - } + if (err) + goto err_out; + + vport->qos.enabled = true; + trace_mlx5_esw_vport_qos_create(vport, bw_share, max_rate); + + return 0; + +err_out: + esw_qos_put(esw); return err; } @@ -629,7 +677,7 @@ void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vpo int err; lockdep_assert_held(&esw->state_lock); - if (!esw->qos.enabled || !vport->qos.enabled) + if (!vport->qos.enabled) return; WARN(vport->qos.group && vport->qos.group != esw->qos.group0, "Disabling QoS on port before detaching it from group"); @@ -643,6 +691,8 @@ void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vpo memset(&vport->qos, 0, sizeof(vport->qos)); trace_mlx5_esw_vport_qos_destroy(vport); + + esw_qos_put(esw); } int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *vport, @@ -651,7 +701,7 @@ int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *vpo int err; lockdep_assert_held(&esw->state_lock); - err = esw_qos_vport_enable(esw, vport, 0, 0); + err = esw_qos_vport_enable(esw, vport, 0, 0, NULL); if (err) return err; @@ -676,7 +726,7 @@ int mlx5_esw_qos_modify_vport_rate(struct mlx5_eswitch *esw, u16 vport_num, u32 mutex_lock(&esw->state_lock); if (!vport->qos.enabled) { /* Eswitch QoS wasn't enabled yet. Enable it and vport QoS. */ - err = esw_qos_vport_enable(esw, vport, rate_mbps, vport->qos.bw_share); + err = esw_qos_vport_enable(esw, vport, rate_mbps, vport->qos.bw_share, NULL); } else { MLX5_SET(scheduling_context, ctx, max_average_bw, rate_mbps); @@ -748,7 +798,7 @@ int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void return err; mutex_lock(&esw->state_lock); - err = esw_qos_vport_enable(esw, vport, 0, 0); + err = esw_qos_vport_enable(esw, vport, 0, 0, extack); if (err) goto unlock; @@ -774,7 +824,7 @@ int mlx5_esw_devlink_rate_leaf_tx_max_set(struct devlink_rate *rate_leaf, void * return err; mutex_lock(&esw->state_lock); - err = esw_qos_vport_enable(esw, vport, 0, 0); + err = esw_qos_vport_enable(esw, vport, 0, 0, extack); if (err) goto unlock; @@ -876,7 +926,7 @@ int mlx5_esw_qos_vport_update_group(struct mlx5_eswitch *esw, int err; mutex_lock(&esw->state_lock); - err = esw_qos_vport_enable(esw, vport, 0, 0); + err = esw_qos_vport_enable(esw, vport, 0, 0, extack); if (!err) err = esw_qos_vport_update_group(esw, vport, group, extack); mutex_unlock(&esw->state_lock); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h index 91b66c1b9881..0141e9d52037 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h @@ -8,8 +8,6 @@ int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport, u32 max_rate, u32 min_rate); -void mlx5_esw_qos_create(struct mlx5_eswitch *esw); -void mlx5_esw_qos_destroy(struct mlx5_eswitch *esw); void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vport); int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void *priv, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 2d188f462028..46532dd42b43 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1257,8 +1257,6 @@ int mlx5_eswitch_enable_locked(struct mlx5_eswitch *esw, int mode, int num_vfs) mlx5_eswitch_update_num_of_vfs(esw, num_vfs); - mlx5_esw_qos_create(esw); - esw->mode = mode; if (mode == MLX5_ESWITCH_LEGACY) { @@ -1287,7 +1285,6 @@ abort: if (mode == MLX5_ESWITCH_OFFLOADS) mlx5_rescan_drivers(esw->dev); - mlx5_esw_qos_destroy(esw); mlx5_esw_acls_ns_cleanup(esw); return err; } @@ -1327,6 +1324,7 @@ int mlx5_eswitch_enable(struct mlx5_eswitch *esw, int num_vfs) void mlx5_eswitch_disable_locked(struct mlx5_eswitch *esw, bool clear_vf) { + struct devlink *devlink = priv_to_devlink(esw->dev); int old_mode; lockdep_assert_held_write(&esw->mode_lock); @@ -1356,7 +1354,8 @@ void mlx5_eswitch_disable_locked(struct mlx5_eswitch *esw, bool clear_vf) if (old_mode == MLX5_ESWITCH_OFFLOADS) mlx5_rescan_drivers(esw->dev); - mlx5_esw_qos_destroy(esw); + devlink_rate_nodes_destroy(devlink); + mlx5_esw_acls_ns_cleanup(esw); if (clear_vf) @@ -1565,6 +1564,7 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) lockdep_register_key(&esw->mode_lock_key); init_rwsem(&esw->mode_lock); lockdep_set_class(&esw->mode_lock, &esw->mode_lock_key); + refcount_set(&esw->qos.refcnt, 0); esw->enabled_vports = 0; esw->mode = MLX5_ESWITCH_NONE; @@ -1598,6 +1598,7 @@ void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw) esw->dev->priv.eswitch = NULL; destroy_workqueue(esw->work_queue); + WARN_ON(refcount_read(&esw->qos.refcnt)); lockdep_unregister_key(&esw->mode_lock_key); mutex_destroy(&esw->state_lock); WARN_ON(!xa_empty(&esw->offloads.vhca_map)); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 42f8ee2e5d9f..513f741d16c7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -308,10 +308,14 @@ struct mlx5_eswitch { atomic64_t user_count; struct { - bool enabled; u32 root_tsar_ix; struct mlx5_esw_rate_group *group0; struct list_head groups; /* Protected by esw->state_lock */ + + /* Protected by esw->state_lock. + * Initially 0, meaning no QoS users and QoS is disabled. + */ + refcount_t refcnt; } qos; struct mlx5_esw_bridge_offloads *br_offloads; @@ -516,11 +520,6 @@ int mlx5_eswitch_del_vlan_action(struct mlx5_eswitch *esw, int __mlx5_eswitch_set_vport_vlan(struct mlx5_eswitch *esw, u16 vport, u16 vlan, u8 qos, u8 set_flags); -static inline bool mlx5_esw_qos_enabled(struct mlx5_eswitch *esw) -{ - return esw->qos.enabled; -} - static inline bool mlx5_eswitch_vlan_actions_supported(struct mlx5_core_dev *dev, u8 vlan_depth) { From ea78548e0f98951fa7641037ad98a750137d6b6a Mon Sep 17 00:00:00 2001 From: Yucong Sun Date: Fri, 12 Nov 2021 11:25:32 -0800 Subject: [PATCH 0268/2048] selftests/bpf: Move summary line after the error logs Makes it easier to find the summary line when there is a lot of logs to scroll back. Signed-off-by: Yucong Sun Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112192535.898352-2-fallentree@fb.com --- tools/testing/selftests/bpf/test_progs.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c index c65986bd9d07..d129ea5c9a48 100644 --- a/tools/testing/selftests/bpf/test_progs.c +++ b/tools/testing/selftests/bpf/test_progs.c @@ -1198,11 +1198,11 @@ static int server_main(void) env.sub_succ_cnt += result->sub_succ_cnt; } + print_all_error_logs(); + fprintf(stdout, "Summary: %d/%d PASSED, %d SKIPPED, %d FAILED\n", env.succ_cnt, env.sub_succ_cnt, env.skip_cnt, env.fail_cnt); - print_all_error_logs(); - /* reap all workers */ for (i = 0; i < env.workers; i++) { int wstatus, pid; @@ -1484,11 +1484,11 @@ int main(int argc, char **argv) if (env.list_test_names) goto out; + print_all_error_logs(); + fprintf(stdout, "Summary: %d/%d PASSED, %d SKIPPED, %d FAILED\n", env.succ_cnt, env.sub_succ_cnt, env.skip_cnt, env.fail_cnt); - print_all_error_logs(); - close(env.saved_netns_fd); out: if (!env.list_test_names && env.has_testmod) From 67d61d30b8a8f33d2a4f269f3a548409827d7b01 Mon Sep 17 00:00:00 2001 From: Yucong Sun Date: Fri, 12 Nov 2021 11:25:33 -0800 Subject: [PATCH 0269/2048] selftests/bpf: Variable naming fix Change log_fd to log_fp to reflect its type correctly. Signed-off-by: Yucong Sun Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112192535.898352-3-fallentree@fb.com --- tools/testing/selftests/bpf/test_progs.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c index d129ea5c9a48..926475aa10bb 100644 --- a/tools/testing/selftests/bpf/test_progs.c +++ b/tools/testing/selftests/bpf/test_progs.c @@ -939,7 +939,7 @@ static void *dispatch_thread(void *ctx) { struct dispatch_data *data = ctx; int sock_fd; - FILE *log_fd = NULL; + FILE *log_fp = NULL; sock_fd = data->sock_fd; @@ -1002,8 +1002,8 @@ static void *dispatch_thread(void *ctx) /* collect all logs */ if (msg_test_done.test_done.have_log) { - log_fd = open_memstream(&result->log_buf, &result->log_cnt); - if (!log_fd) + log_fp = open_memstream(&result->log_buf, &result->log_cnt); + if (!log_fp) goto error; while (true) { @@ -1014,12 +1014,12 @@ static void *dispatch_thread(void *ctx) if (msg_log.type != MSG_TEST_LOG) goto error; - fprintf(log_fd, "%s", msg_log.test_log.log_buf); + fprintf(log_fp, "%s", msg_log.test_log.log_buf); if (msg_log.test_log.is_last) break; } - fclose(log_fd); - log_fd = NULL; + fclose(log_fp); + log_fp = NULL; } /* output log */ { @@ -1045,8 +1045,8 @@ error: if (env.debug) fprintf(stderr, "[%d]: Protocol/IO error: %s.\n", data->worker_id, strerror(errno)); - if (log_fd) - fclose(log_fd); + if (log_fp) + fclose(log_fp); done: { struct msg msg_exit; From db813d7bd919c521b869d657dc4a2a2335974cc4 Mon Sep 17 00:00:00 2001 From: Yucong Sun Date: Fri, 12 Nov 2021 11:25:34 -0800 Subject: [PATCH 0270/2048] selftests/bpf: Mark variable as static Fix warnings from checkstyle.pl Signed-off-by: Yucong Sun Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211112192535.898352-4-fallentree@fb.com --- tools/testing/selftests/bpf/test_progs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c index 926475aa10bb..296928948bb9 100644 --- a/tools/testing/selftests/bpf/test_progs.c +++ b/tools/testing/selftests/bpf/test_progs.c @@ -473,11 +473,11 @@ static struct prog_test_def prog_test_defs[] = { #include #undef DEFINE_TEST }; -const int prog_test_cnt = ARRAY_SIZE(prog_test_defs); +static const int prog_test_cnt = ARRAY_SIZE(prog_test_defs); const char *argp_program_version = "test_progs 0.1"; const char *argp_program_bug_address = ""; -const char argp_program_doc[] = "BPF selftests test runner"; +static const char argp_program_doc[] = "BPF selftests test runner"; enum ARG_KEYS { ARG_TEST_NUM = 'n', From 086c921a354089f209318501038d43c98d3f409f Mon Sep 17 00:00:00 2001 From: Sven Eckelmann Date: Mon, 15 Nov 2021 11:29:55 +0200 Subject: [PATCH 0271/2048] ath11k: Fix ETSI regd with weather radar overlap Some ETSI countries have a small overlap in the wireless-regdb with an ETSI channel (5590-5650). A good example is Australia: country AU: DFS-ETSI (2400 - 2483.5 @ 40), (36) (5150 - 5250 @ 80), (23), NO-OUTDOOR, AUTO-BW (5250 - 5350 @ 80), (20), NO-OUTDOOR, AUTO-BW, DFS (5470 - 5600 @ 80), (27), DFS (5650 - 5730 @ 80), (27), DFS (5730 - 5850 @ 80), (36) (57000 - 66000 @ 2160), (43), NO-OUTDOOR If the firmware (or the BDF) is shipped with these rules then there is only a 10 MHz overlap with the weather radar: * below: 5470 - 5590 * weather radar: 5590 - 5600 * above: (none for the rule "5470 - 5600 @ 80") There are several wrong assumption in the ath11k code: * there is always a valid range below the weather radar (actually: there could be no range below the weather radar range OR range could be smaller than 20 MHz) * intersected range in the weather radar range is valid (actually: the range could be smaller than 20 MHz) * range above weather radar is either empty or valid (actually: the range could be smaller than 20 MHz) These wrong assumption will lead in this example to a rule (5590 - 5600 @ 20), (N/A, 27), (600000 ms), DFS, AUTO-BW which is invalid according to is_valid_reg_rule() because the freq_diff is only 10 MHz but the max_bandwidth is set to 20 MHz. Which results in a rejection like: WARNING: at backports-20210222_001-4.4.60-b157d2276/net/wireless/reg.c:3984 [...] Call trace: [] reg_get_max_bandwidth+0x300/0x3a8 [cfg80211] [] regulatory_set_wiphy_regd_sync+0x3c/0x98 [cfg80211] [] ath11k_regd_update+0x1a8/0x210 [ath11k] [] ath11k_regd_update_work+0x18/0x20 [ath11k] [] process_one_work+0x1f8/0x340 [] worker_thread+0x25c/0x448 [] kthread+0xd0/0xd8 [] ret_from_fork+0x10/0x40 ath11k c000000.wifi: failed to perform regd update : -22 Invalid regulatory domain detected To avoid this, the algorithm has to be changed slightly. Instead of splitting a rule which overlaps with the weather radar range into 3 pieces and accepting the first two parts blindly, it must actually be checked for each piece whether it is a valid range. And only if it is valid, add it to the output array. When these checks are in place, the processed rules for AU would end up as country AU: DFS-ETSI (2400 - 2483 @ 40), (N/A, 36), (N/A) (5150 - 5250 @ 80), (6, 23), (N/A), NO-OUTDOOR, AUTO-BW (5250 - 5350 @ 80), (6, 20), (0 ms), NO-OUTDOOR, DFS, AUTO-BW (5470 - 5590 @ 80), (6, 27), (0 ms), DFS, AUTO-BW (5650 - 5730 @ 80), (6, 27), (0 ms), DFS, AUTO-BW (5730 - 5850 @ 80), (6, 36), (N/A), AUTO-BW and will be accepted by the wireless regulatory code. Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Signed-off-by: Sven Eckelmann Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211112153116.1214421-1-sven@narfation.org --- drivers/net/wireless/ath/ath11k/reg.c | 113 ++++++++++++++------------ 1 file changed, 61 insertions(+), 52 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/reg.c b/drivers/net/wireless/ath/ath11k/reg.c index a66b5bdd2167..8606170ba80d 100644 --- a/drivers/net/wireless/ath/ath11k/reg.c +++ b/drivers/net/wireless/ath/ath11k/reg.c @@ -456,6 +456,9 @@ ath11k_reg_adjust_bw(u16 start_freq, u16 end_freq, u16 max_bw) { u16 bw; + if (end_freq <= start_freq) + return 0; + bw = end_freq - start_freq; bw = min_t(u16, bw, max_bw); @@ -463,8 +466,10 @@ ath11k_reg_adjust_bw(u16 start_freq, u16 end_freq, u16 max_bw) bw = 80; else if (bw >= 40 && bw < 80) bw = 40; - else if (bw < 40) + else if (bw >= 20 && bw < 40) bw = 20; + else + bw = 0; return bw; } @@ -488,73 +493,77 @@ ath11k_reg_update_weather_radar_band(struct ath11k_base *ab, struct cur_reg_rule *reg_rule, u8 *rule_idx, u32 flags, u16 max_bw) { + u32 start_freq; u32 end_freq; u16 bw; u8 i; i = *rule_idx; + /* there might be situations when even the input rule must be dropped */ + i--; + + /* frequencies below weather radar */ bw = ath11k_reg_adjust_bw(reg_rule->start_freq, ETSI_WEATHER_RADAR_BAND_LOW, max_bw); + if (bw > 0) { + i++; - ath11k_reg_update_rule(regd->reg_rules + i, reg_rule->start_freq, - ETSI_WEATHER_RADAR_BAND_LOW, bw, - reg_rule->ant_gain, reg_rule->reg_power, - flags); + ath11k_reg_update_rule(regd->reg_rules + i, + reg_rule->start_freq, + ETSI_WEATHER_RADAR_BAND_LOW, bw, + reg_rule->ant_gain, reg_rule->reg_power, + flags); - ath11k_dbg(ab, ATH11K_DBG_REG, - "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", - i + 1, reg_rule->start_freq, ETSI_WEATHER_RADAR_BAND_LOW, - bw, reg_rule->ant_gain, reg_rule->reg_power, - regd->reg_rules[i].dfs_cac_ms, - flags); - - if (reg_rule->end_freq > ETSI_WEATHER_RADAR_BAND_HIGH) - end_freq = ETSI_WEATHER_RADAR_BAND_HIGH; - else - end_freq = reg_rule->end_freq; - - bw = ath11k_reg_adjust_bw(ETSI_WEATHER_RADAR_BAND_LOW, end_freq, - max_bw); - - i++; - - ath11k_reg_update_rule(regd->reg_rules + i, - ETSI_WEATHER_RADAR_BAND_LOW, end_freq, bw, - reg_rule->ant_gain, reg_rule->reg_power, - flags); - - regd->reg_rules[i].dfs_cac_ms = ETSI_WEATHER_RADAR_BAND_CAC_TIMEOUT; - - ath11k_dbg(ab, ATH11K_DBG_REG, - "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", - i + 1, ETSI_WEATHER_RADAR_BAND_LOW, end_freq, - bw, reg_rule->ant_gain, reg_rule->reg_power, - regd->reg_rules[i].dfs_cac_ms, - flags); - - if (end_freq == reg_rule->end_freq) { - regd->n_reg_rules--; - *rule_idx = i; - return; + ath11k_dbg(ab, ATH11K_DBG_REG, + "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", + i + 1, reg_rule->start_freq, + ETSI_WEATHER_RADAR_BAND_LOW, bw, reg_rule->ant_gain, + reg_rule->reg_power, regd->reg_rules[i].dfs_cac_ms, + flags); } + /* weather radar frequencies */ + start_freq = max_t(u32, reg_rule->start_freq, + ETSI_WEATHER_RADAR_BAND_LOW); + end_freq = min_t(u32, reg_rule->end_freq, ETSI_WEATHER_RADAR_BAND_HIGH); + + bw = ath11k_reg_adjust_bw(start_freq, end_freq, max_bw); + if (bw > 0) { + i++; + + ath11k_reg_update_rule(regd->reg_rules + i, start_freq, + end_freq, bw, reg_rule->ant_gain, + reg_rule->reg_power, flags); + + regd->reg_rules[i].dfs_cac_ms = ETSI_WEATHER_RADAR_BAND_CAC_TIMEOUT; + + ath11k_dbg(ab, ATH11K_DBG_REG, + "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", + i + 1, start_freq, end_freq, bw, + reg_rule->ant_gain, reg_rule->reg_power, + regd->reg_rules[i].dfs_cac_ms, flags); + } + + /* frequencies above weather radar */ bw = ath11k_reg_adjust_bw(ETSI_WEATHER_RADAR_BAND_HIGH, reg_rule->end_freq, max_bw); + if (bw > 0) { + i++; - i++; + ath11k_reg_update_rule(regd->reg_rules + i, + ETSI_WEATHER_RADAR_BAND_HIGH, + reg_rule->end_freq, bw, + reg_rule->ant_gain, reg_rule->reg_power, + flags); - ath11k_reg_update_rule(regd->reg_rules + i, ETSI_WEATHER_RADAR_BAND_HIGH, - reg_rule->end_freq, bw, - reg_rule->ant_gain, reg_rule->reg_power, - flags); - - ath11k_dbg(ab, ATH11K_DBG_REG, - "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", - i + 1, ETSI_WEATHER_RADAR_BAND_HIGH, reg_rule->end_freq, - bw, reg_rule->ant_gain, reg_rule->reg_power, - regd->reg_rules[i].dfs_cac_ms, - flags); + ath11k_dbg(ab, ATH11K_DBG_REG, + "\t%d. (%d - %d @ %d) (%d, %d) (%d ms) (FLAGS %d)\n", + i + 1, ETSI_WEATHER_RADAR_BAND_HIGH, + reg_rule->end_freq, bw, reg_rule->ant_gain, + reg_rule->reg_power, regd->reg_rules[i].dfs_cac_ms, + flags); + } *rule_idx = i; } From f951380a6022440335f668f85296096ba13071ba Mon Sep 17 00:00:00 2001 From: P Praneesh Date: Mon, 15 Nov 2021 11:50:52 +0200 Subject: [PATCH 0272/2048] ath11k: Disabling credit flow for WMI path Firmware credit flow control is enabled for WMI control services, which expects available tokens should be acquired before sending a command to the target. Also the token gets released when firmware receives the command. This credit-based flow limits driver to send WMI command only when the token available which is causing WMI commands to timeout and return -EAGAIN, whereas firmware has enough capability to process the WMI command. To fix this Tx starvation issue, introduce the ability to disable the credit flow for the WMI path. The driver sends WMI configuration for disabling credit flow to firmware by two ways. 1. By using a global flag (HTC_MSG_SETUP_COMPLETE_EX_ID msg type flags) 2. By using a local flag (ATH11K_HTC_CONN_FLAGS_DISABLE_CREDIT_FLOW_CTRL = 1 << 3) Ath11k uses both these configurations to disable credit flow for the WMI path completely. Also added a hw_param member for credit flow control by which we can enable or disable it based on per-target basis. Currently we are disabling credit flow for IPQ8074, IPQ6018, and QCN9074 as recommended by firmware. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01492-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ6018 hw1.0 AHB WLAN.HK.2.4.0.1-00330-QCAHKSWPL_SILICONZ-1 Co-developed-by: Pravas Kumar Panda Signed-off-by: Pravas Kumar Panda Signed-off-by: P Praneesh Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635156494-20059-1-git-send-email-quic_ppranees@quicinc.com --- drivers/net/wireless/ath/ath11k/ce.c | 37 +++++++++++--- drivers/net/wireless/ath/ath11k/ce.h | 3 +- drivers/net/wireless/ath/ath11k/core.c | 5 ++ drivers/net/wireless/ath/ath11k/htc.c | 71 +++++++++++++++++++------- drivers/net/wireless/ath/ath11k/htc.h | 9 ++-- drivers/net/wireless/ath/ath11k/hw.h | 1 + drivers/net/wireless/ath/ath11k/wmi.c | 54 +++++++++++++++++--- drivers/net/wireless/ath/ath11k/wmi.h | 1 + 8 files changed, 146 insertions(+), 35 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/ce.c b/drivers/net/wireless/ath/ath11k/ce.c index 3bc4242c1059..aaa7b05ff49d 100644 --- a/drivers/net/wireless/ath/ath11k/ce.c +++ b/drivers/net/wireless/ath/ath11k/ce.c @@ -14,6 +14,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = { .src_nentries = 16, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE1: target->host HTT + HTC control */ @@ -40,6 +41,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE4: host->target HTT */ @@ -73,6 +75,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE8: target autonomous hif_memcpy */ @@ -89,6 +92,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE10: target->host HTT */ @@ -142,6 +146,7 @@ const struct ce_attr ath11k_host_ce_config_qca6390[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE4: host->target HTT */ @@ -175,6 +180,7 @@ const struct ce_attr ath11k_host_ce_config_qca6390[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE8: target autonomous hif_memcpy */ @@ -220,6 +226,7 @@ const struct ce_attr ath11k_host_ce_config_qcn9074[] = { .src_nentries = 32, .src_sz_max = 2048, .dest_nentries = 0, + .send_cb = ath11k_htc_tx_completion_handler, }, /* CE4: host->target HTT */ @@ -489,18 +496,32 @@ err_unlock: return skb; } -static void ath11k_ce_send_done_cb(struct ath11k_ce_pipe *pipe) +static void ath11k_ce_tx_process_cb(struct ath11k_ce_pipe *pipe) { struct ath11k_base *ab = pipe->ab; struct sk_buff *skb; + struct sk_buff_head list; + __skb_queue_head_init(&list); while (!IS_ERR(skb = ath11k_ce_completed_send_next(pipe))) { if (!skb) continue; dma_unmap_single(ab->dev, ATH11K_SKB_CB(skb)->paddr, skb->len, DMA_TO_DEVICE); - dev_kfree_skb_any(skb); + + if ((!pipe->send_cb) || ab->hw_params.credit_flow) { + dev_kfree_skb_any(skb); + continue; + } + + __skb_queue_tail(&list, skb); + } + + while ((skb = __skb_dequeue(&list))) { + ath11k_dbg(ab, ATH11K_DBG_AHB, "tx ce pipe %d len %d\n", + pipe->pipe_num, skb->len); + pipe->send_cb(ab, skb); } } @@ -636,7 +657,7 @@ static int ath11k_ce_alloc_pipe(struct ath11k_base *ab, int ce_id) pipe->attr_flags = attr->flags; if (attr->src_nentries) { - pipe->send_cb = ath11k_ce_send_done_cb; + pipe->send_cb = attr->send_cb; nentries = roundup_pow_of_two(attr->src_nentries); desc_sz = ath11k_hal_ce_get_desc_size(HAL_CE_DESC_SRC); ring = ath11k_ce_alloc_ring(ab, nentries, desc_sz); @@ -667,9 +688,10 @@ static int ath11k_ce_alloc_pipe(struct ath11k_base *ab, int ce_id) void ath11k_ce_per_engine_service(struct ath11k_base *ab, u16 ce_id) { struct ath11k_ce_pipe *pipe = &ab->ce.ce_pipe[ce_id]; + const struct ce_attr *attr = &ab->hw_params.host_ce_config[ce_id]; - if (pipe->send_cb) - pipe->send_cb(pipe); + if (attr->src_nentries) + ath11k_ce_tx_process_cb(pipe); if (pipe->recv_cb) ath11k_ce_recv_process_cb(pipe); @@ -678,9 +700,10 @@ void ath11k_ce_per_engine_service(struct ath11k_base *ab, u16 ce_id) void ath11k_ce_poll_send_completed(struct ath11k_base *ab, u8 pipe_id) { struct ath11k_ce_pipe *pipe = &ab->ce.ce_pipe[pipe_id]; + const struct ce_attr *attr = &ab->hw_params.host_ce_config[pipe_id]; - if ((pipe->attr_flags & CE_ATTR_DIS_INTR) && pipe->send_cb) - pipe->send_cb(pipe); + if ((pipe->attr_flags & CE_ATTR_DIS_INTR) && attr->src_nentries) + ath11k_ce_tx_process_cb(pipe); } EXPORT_SYMBOL(ath11k_ce_per_engine_service); diff --git a/drivers/net/wireless/ath/ath11k/ce.h b/drivers/net/wireless/ath/ath11k/ce.h index 713f766cac22..8255b6cfab0c 100644 --- a/drivers/net/wireless/ath/ath11k/ce.h +++ b/drivers/net/wireless/ath/ath11k/ce.h @@ -101,6 +101,7 @@ struct ce_attr { unsigned int dest_nentries; void (*recv_cb)(struct ath11k_base *, struct sk_buff *); + void (*send_cb)(struct ath11k_base *, struct sk_buff *); }; #define CE_DESC_RING_ALIGN 8 @@ -154,7 +155,7 @@ struct ath11k_ce_pipe { unsigned int buf_sz; unsigned int rx_buf_needed; - void (*send_cb)(struct ath11k_ce_pipe *); + void (*send_cb)(struct ath11k_base *, struct sk_buff *); void (*recv_cb)(struct ath11k_base *, struct sk_buff *); struct tasklet_struct intr_tq; diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index 2ba3bf8d3fde..4e994db996db 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -81,6 +81,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), .fix_l1ss = true, + .credit_flow = false, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, @@ -133,6 +134,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), .fix_l1ss = true, + .credit_flow = false, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, @@ -184,6 +186,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_suspend = true, .hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074), .fix_l1ss = true, + .credit_flow = true, .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, @@ -235,6 +238,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_suspend = false, .hal_desc_sz = sizeof(struct hal_rx_desc_qcn9074), .fix_l1ss = true, + .credit_flow = false, .max_tx_ring = DP_TCL_NUM_RING_MAX, .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = true, @@ -286,6 +290,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .supports_suspend = true, .hal_desc_sz = sizeof(struct hal_rx_desc_wcn6855), .fix_l1ss = false, + .credit_flow = true, .max_tx_ring = DP_TCL_NUM_RING_MAX_QCA6390, .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, diff --git a/drivers/net/wireless/ath/ath11k/htc.c b/drivers/net/wireless/ath/ath11k/htc.c index 54b1d34724d7..6913b7494b9b 100644 --- a/drivers/net/wireless/ath/ath11k/htc.c +++ b/drivers/net/wireless/ath/ath11k/htc.c @@ -81,6 +81,8 @@ int ath11k_htc_send(struct ath11k_htc *htc, struct ath11k_base *ab = htc->ab; int credits = 0; int ret; + bool credit_flow_enabled = (ab->hw_params.credit_flow && + ep->tx_credit_flow_enabled); if (eid >= ATH11K_HTC_EP_COUNT) { ath11k_warn(ab, "Invalid endpoint id: %d\n", eid); @@ -89,7 +91,7 @@ int ath11k_htc_send(struct ath11k_htc *htc, skb_push(skb, sizeof(struct ath11k_htc_hdr)); - if (ep->tx_credit_flow_enabled) { + if (credit_flow_enabled) { credits = DIV_ROUND_UP(skb->len, htc->target_credit_size); spin_lock_bh(&htc->tx_lock); if (ep->tx_credits < credits) { @@ -126,7 +128,7 @@ int ath11k_htc_send(struct ath11k_htc *htc, err_unmap: dma_unmap_single(dev, skb_cb->paddr, skb->len, DMA_TO_DEVICE); err_credits: - if (ep->tx_credit_flow_enabled) { + if (credit_flow_enabled) { spin_lock_bh(&htc->tx_lock); ep->tx_credits += credits; ath11k_dbg(ab, ATH11K_DBG_HTC, @@ -203,23 +205,25 @@ static int ath11k_htc_process_trailer(struct ath11k_htc *htc, break; } - switch (record->hdr.id) { - case ATH11K_HTC_RECORD_CREDITS: - len = sizeof(struct ath11k_htc_credit_report); - if (record->hdr.len < len) { - ath11k_warn(ab, "Credit report too long\n"); - status = -EINVAL; + if (ab->hw_params.credit_flow) { + switch (record->hdr.id) { + case ATH11K_HTC_RECORD_CREDITS: + len = sizeof(struct ath11k_htc_credit_report); + if (record->hdr.len < len) { + ath11k_warn(ab, "Credit report too long\n"); + status = -EINVAL; + break; + } + ath11k_htc_process_credit_report(htc, + record->credit_report, + record->hdr.len, + src_eid); + break; + default: + ath11k_warn(ab, "Unhandled record: id:%d length:%d\n", + record->hdr.id, record->hdr.len); break; } - ath11k_htc_process_credit_report(htc, - record->credit_report, - record->hdr.len, - src_eid); - break; - default: - ath11k_warn(ab, "Unhandled record: id:%d length:%d\n", - record->hdr.id, record->hdr.len); - break; } if (status) @@ -245,6 +249,29 @@ static void ath11k_htc_suspend_complete(struct ath11k_base *ab, bool ack) complete(&ab->htc_suspend); } +void ath11k_htc_tx_completion_handler(struct ath11k_base *ab, + struct sk_buff *skb) +{ + struct ath11k_htc *htc = &ab->htc; + struct ath11k_htc_ep *ep; + void (*ep_tx_complete)(struct ath11k_base *, struct sk_buff *); + u8 eid; + + eid = ATH11K_SKB_CB(skb)->eid; + if (eid >= ATH11K_HTC_EP_COUNT) + return; + + ep = &htc->endpoint[eid]; + spin_lock_bh(&htc->tx_lock); + ep_tx_complete = ep->ep_ops.ep_tx_complete; + spin_unlock_bh(&htc->tx_lock); + if (!ep_tx_complete) { + dev_kfree_skb_any(skb); + return; + } + ep_tx_complete(htc->ab, skb); +} + void ath11k_htc_rx_completion_handler(struct ath11k_base *ab, struct sk_buff *skb) { @@ -607,6 +634,11 @@ int ath11k_htc_connect_service(struct ath11k_htc *htc, disable_credit_flow_ctrl = true; } + if (!ab->hw_params.credit_flow) { + flags |= ATH11K_HTC_CONN_FLAGS_DISABLE_CREDIT_FLOW_CTRL; + disable_credit_flow_ctrl = true; + } + req_msg->flags_len = FIELD_PREP(HTC_SVC_MSG_CONNECTIONFLAGS, flags); req_msg->msg_svc_id |= FIELD_PREP(HTC_SVC_MSG_SERVICE_ID, conn_req->service_id); @@ -732,7 +764,10 @@ int ath11k_htc_start(struct ath11k_htc *htc) msg->msg_id = FIELD_PREP(HTC_MSG_MESSAGEID, ATH11K_HTC_MSG_SETUP_COMPLETE_EX_ID); - ath11k_dbg(ab, ATH11K_DBG_HTC, "HTC is using TX credit flow control\n"); + if (ab->hw_params.credit_flow) + ath11k_dbg(ab, ATH11K_DBG_HTC, "HTC is using TX credit flow control\n"); + else + msg->flags |= ATH11K_GLOBAL_DISABLE_CREDIT_FLOW; status = ath11k_htc_send(htc, ATH11K_HTC_EP_0, skb); if (status) { diff --git a/drivers/net/wireless/ath/ath11k/htc.h b/drivers/net/wireless/ath/ath11k/htc.h index 6c8a469d7f9d..f429b37cfdf7 100644 --- a/drivers/net/wireless/ath/ath11k/htc.h +++ b/drivers/net/wireless/ath/ath11k/htc.h @@ -83,8 +83,8 @@ enum ath11k_htc_conn_flags { ATH11K_HTC_CONN_FLAGS_THRESHOLD_LEVEL_ONE_HALF = 0x1, ATH11K_HTC_CONN_FLAGS_THRESHOLD_LEVEL_THREE_FOURTHS = 0x2, ATH11K_HTC_CONN_FLAGS_THRESHOLD_LEVEL_UNITY = 0x3, - ATH11K_HTC_CONN_FLAGS_REDUCE_CREDIT_DRIBBLE = 1 << 2, - ATH11K_HTC_CONN_FLAGS_DISABLE_CREDIT_FLOW_CTRL = 1 << 3 + ATH11K_HTC_CONN_FLAGS_REDUCE_CREDIT_DRIBBLE = 0x4, + ATH11K_HTC_CONN_FLAGS_DISABLE_CREDIT_FLOW_CTRL = 0x8, }; enum ath11k_htc_conn_svc_status { @@ -116,6 +116,8 @@ struct ath11k_htc_conn_svc_resp { u32 svc_meta_pad; } __packed; +#define ATH11K_GLOBAL_DISABLE_CREDIT_FLOW BIT(1) + struct ath11k_htc_setup_complete_extended { u32 msg_id; u32 flags; @@ -305,5 +307,6 @@ int ath11k_htc_send(struct ath11k_htc *htc, enum ath11k_htc_ep_id eid, struct sk_buff *ath11k_htc_alloc_skb(struct ath11k_base *ar, int size); void ath11k_htc_rx_completion_handler(struct ath11k_base *ar, struct sk_buff *skb); - +void ath11k_htc_tx_completion_handler(struct ath11k_base *ab, + struct sk_buff *skb); #endif diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h index 3e64c9b94db0..dd2202f23b99 100644 --- a/drivers/net/wireless/ath/ath11k/hw.h +++ b/drivers/net/wireless/ath/ath11k/hw.h @@ -175,6 +175,7 @@ struct ath11k_hw_params { bool supports_suspend; u32 hal_desc_sz; bool fix_l1ss; + bool credit_flow; u8 max_tx_ring; const struct ath11k_hw_hal_params *hal_params; bool supports_dynamic_smps_6ghz; diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 6248e174bd8f..6bd355ce88d3 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -267,21 +267,39 @@ int ath11k_wmi_cmd_send(struct ath11k_pdev_wmi *wmi, struct sk_buff *skb, { struct ath11k_wmi_base *wmi_sc = wmi->wmi_ab; int ret = -EOPNOTSUPP; + struct ath11k_base *ab = wmi_sc->ab; might_sleep(); - wait_event_timeout(wmi_sc->tx_credits_wq, ({ - ret = ath11k_wmi_cmd_send_nowait(wmi, skb, cmd_id); + if (ab->hw_params.credit_flow) { + wait_event_timeout(wmi_sc->tx_credits_wq, ({ + ret = ath11k_wmi_cmd_send_nowait(wmi, skb, cmd_id); - if (ret && test_bit(ATH11K_FLAG_CRASH_FLUSH, &wmi_sc->ab->dev_flags)) - ret = -ESHUTDOWN; + if (ret && test_bit(ATH11K_FLAG_CRASH_FLUSH, + &wmi_sc->ab->dev_flags)) + ret = -ESHUTDOWN; - (ret != -EAGAIN); - }), WMI_SEND_TIMEOUT_HZ); + (ret != -EAGAIN); + }), WMI_SEND_TIMEOUT_HZ); + } else { + wait_event_timeout(wmi->tx_ce_desc_wq, ({ + ret = ath11k_wmi_cmd_send_nowait(wmi, skb, cmd_id); + + if (ret && test_bit(ATH11K_FLAG_CRASH_FLUSH, + &wmi_sc->ab->dev_flags)) + ret = -ESHUTDOWN; + + (ret != -ENOBUFS); + }), WMI_SEND_TIMEOUT_HZ); + } if (ret == -EAGAIN) ath11k_warn(wmi_sc->ab, "wmi command %d timeout\n", cmd_id); + if (ret == -ENOBUFS) + ath11k_warn(wmi_sc->ab, "ce desc not available for wmi command %d\n", + cmd_id); + return ret; } @@ -5816,7 +5834,30 @@ static void ath11k_wmi_op_ep_tx_credits(struct ath11k_base *ab) static void ath11k_wmi_htc_tx_complete(struct ath11k_base *ab, struct sk_buff *skb) { + struct ath11k_pdev_wmi *wmi = NULL; + u32 i; + u8 wmi_ep_count; + u8 eid; + + eid = ATH11K_SKB_CB(skb)->eid; dev_kfree_skb(skb); + + if (eid >= ATH11K_HTC_EP_COUNT) + return; + + wmi_ep_count = ab->htc.wmi_ep_count; + if (wmi_ep_count > ab->hw_params.max_radios) + return; + + for (i = 0; i < ab->htc.wmi_ep_count; i++) { + if (ab->wmi_ab.wmi[i].eid == eid) { + wmi = &ab->wmi_ab.wmi[i]; + break; + } + } + + if (wmi) + wake_up(&wmi->tx_ce_desc_wq); } static bool ath11k_reg_is_world_alpha(char *alpha) @@ -7207,6 +7248,7 @@ static int ath11k_connect_pdev_htc_service(struct ath11k_base *ab, ab->wmi_ab.wmi_endpoint_id[pdev_idx] = conn_resp.eid; ab->wmi_ab.wmi[pdev_idx].eid = conn_resp.eid; ab->wmi_ab.max_msg_len[pdev_idx] = conn_resp.max_msg_len; + init_waitqueue_head(&ab->wmi_ab.wmi[pdev_idx].tx_ce_desc_wq); return 0; } diff --git a/drivers/net/wireless/ath/ath11k/wmi.h b/drivers/net/wireless/ath/ath11k/wmi.h index 81d12405be54..b5b9f2c22e2e 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.h +++ b/drivers/net/wireless/ath/ath11k/wmi.h @@ -2522,6 +2522,7 @@ struct ath11k_pdev_wmi { enum ath11k_htc_ep_id eid; const struct wmi_peer_flags_map *peer_flags; u32 rx_decap_mode; + wait_queue_head_t tx_ce_desc_wq; }; struct vdev_create_params { From bd77f6b1d7104cf6451399a7c67d08afecb9a7c7 Mon Sep 17 00:00:00 2001 From: Rameshkumar Sundaram Date: Tue, 2 Nov 2021 11:11:33 +0530 Subject: [PATCH 0273/2048] ath11k: use cache line aligned buffers for dbring The DMA buffers of dbring which is used for spectral/cfr starts at certain offset from original kmalloc() returned buffer. This is not cache line aligned. And also driver tries to access the data that is immediately before this offset address (i.e. buff->paddr) after doing dma map. This will cause cache line sharing issues and data corruption, if CPU happen to write back cache after HW has dma'ed the data. Fix this by mapping a cache line aligned buffer to dma. Tested on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1 Signed-off-by: Rameshkumar Sundaram Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635831693-15962-1-git-send-email-quic_ramess@quicinc.com --- drivers/net/wireless/ath/ath11k/dbring.c | 16 ++++++++++++---- drivers/net/wireless/ath/ath11k/dbring.h | 2 +- 2 files changed, 13 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dbring.c b/drivers/net/wireless/ath/ath11k/dbring.c index fd98ba5b1130..de220a10bce3 100644 --- a/drivers/net/wireless/ath/ath11k/dbring.c +++ b/drivers/net/wireless/ath/ath11k/dbring.c @@ -87,17 +87,23 @@ static int ath11k_dbring_fill_bufs(struct ath11k *ar, req_entries = min(num_free, ring->bufs_max); num_remain = req_entries; align = ring->buf_align; - size = sizeof(*buff) + ring->buf_sz + align - 1; + size = ring->buf_sz + align - 1; while (num_remain > 0) { - buff = kzalloc(size, GFP_ATOMIC); + buff = kzalloc(sizeof(*buff), GFP_ATOMIC); if (!buff) break; + buff->payload = kzalloc(size, GFP_ATOMIC); + if (!buff->payload) { + kfree(buff); + break; + } ret = ath11k_dbring_bufs_replenish(ar, ring, buff); if (ret) { ath11k_warn(ar->ab, "failed to replenish db ring num_remain %d req_ent %d\n", num_remain, req_entries); + kfree(buff->payload); kfree(buff); break; } @@ -282,7 +288,7 @@ int ath11k_dbring_buffer_release_event(struct ath11k_base *ab, srng = &ab->hal.srng_list[ring->refill_srng.ring_id]; num_entry = ev->fixed.num_buf_release_entry; - size = sizeof(*buff) + ring->buf_sz + ring->buf_align - 1; + size = ring->buf_sz + ring->buf_align - 1; num_buff_reaped = 0; spin_lock_bh(&srng->lock); @@ -319,7 +325,8 @@ int ath11k_dbring_buffer_release_event(struct ath11k_base *ab, ring->handler(ar, &handler_data); } - memset(buff, 0, size); + buff->paddr = 0; + memset(buff->payload, 0, size); ath11k_dbring_bufs_replenish(ar, ring, buff); } @@ -346,6 +353,7 @@ void ath11k_dbring_buf_cleanup(struct ath11k *ar, struct ath11k_dbring *ring) idr_remove(&ring->bufs_idr, buf_id); dma_unmap_single(ar->ab->dev, buff->paddr, ring->buf_sz, DMA_FROM_DEVICE); + kfree(buff->payload); kfree(buff); } diff --git a/drivers/net/wireless/ath/ath11k/dbring.h b/drivers/net/wireless/ath/ath11k/dbring.h index f7fce9ef9c36..78a985faa0a1 100644 --- a/drivers/net/wireless/ath/ath11k/dbring.h +++ b/drivers/net/wireless/ath/ath11k/dbring.h @@ -13,7 +13,7 @@ struct ath11k_dbring_element { dma_addr_t paddr; - u8 payload[0]; + u8 *payload; }; struct ath11k_dbring_data { From 1ad6e4b00f29d017b196dda7ab96d1cfcbabd7d2 Mon Sep 17 00:00:00 2001 From: Anilkumar Kolli Date: Tue, 2 Nov 2021 18:22:38 +0530 Subject: [PATCH 0274/2048] ath11k: Add missing qmi_txn_cancel() Currently many functions do not follow this guidance when qmi_send_request() fails, therefore add missing qmi_txn_cancel() in the qmi_send_request() error path. Also remove initialization on 'struct qmi_txn' since qmi_tx_init() performs all necessary initialization. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1-01838-QCAHKSWPL_SILICONZ-1 Signed-off-by: Anilkumar Kolli Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635857558-21733-1-git-send-email-akolli@codeaurora.org --- drivers/net/wireless/ath/ath11k/qmi.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/qmi.c b/drivers/net/wireless/ath/ath11k/qmi.c index fa73118de6db..25eb22cbeaeb 100644 --- a/drivers/net/wireless/ath/ath11k/qmi.c +++ b/drivers/net/wireless/ath/ath11k/qmi.c @@ -1586,7 +1586,7 @@ static int ath11k_qmi_host_cap_send(struct ath11k_base *ab) { struct qmi_wlanfw_host_cap_req_msg_v01 req; struct qmi_wlanfw_host_cap_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0; memset(&req, 0, sizeof(req)); @@ -1640,6 +1640,7 @@ static int ath11k_qmi_host_cap_send(struct ath11k_base *ab) QMI_WLANFW_HOST_CAP_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_host_cap_req_msg_v01_ei, &req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send host capability request: %d\n", ret); goto out; } @@ -1705,6 +1706,7 @@ static int ath11k_qmi_fw_ind_register_send(struct ath11k_base *ab) QMI_WLANFW_IND_REGISTER_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_ind_register_req_msg_v01_ei, req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send indication register request: %d\n", ret); goto out; @@ -1734,7 +1736,7 @@ static int ath11k_qmi_respond_fw_mem_request(struct ath11k_base *ab) { struct qmi_wlanfw_respond_mem_req_msg_v01 *req; struct qmi_wlanfw_respond_mem_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0, i; bool delayed; @@ -1783,6 +1785,7 @@ static int ath11k_qmi_respond_fw_mem_request(struct ath11k_base *ab) QMI_WLANFW_RESPOND_MEM_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_respond_mem_req_msg_v01_ei, req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to respond qmi memory request: %d\n", ret); goto out; @@ -1911,7 +1914,7 @@ static int ath11k_qmi_request_target_cap(struct ath11k_base *ab) { struct qmi_wlanfw_cap_req_msg_v01 req; struct qmi_wlanfw_cap_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0; int r; @@ -1930,6 +1933,7 @@ static int ath11k_qmi_request_target_cap(struct ath11k_base *ab) QMI_WLANFW_CAP_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_cap_req_msg_v01_ei, &req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send qmi cap request: %d\n", ret); goto out; @@ -2000,7 +2004,7 @@ static int ath11k_qmi_load_file_target_mem(struct ath11k_base *ab, { struct qmi_wlanfw_bdf_download_req_msg_v01 *req; struct qmi_wlanfw_bdf_download_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; const u8 *temp = data; void __iomem *bdf_addr = NULL; int ret; @@ -2245,7 +2249,7 @@ static int ath11k_qmi_wlanfw_m3_info_send(struct ath11k_base *ab) struct m3_mem_region *m3_mem = &ab->qmi.m3_mem; struct qmi_wlanfw_m3_info_req_msg_v01 req; struct qmi_wlanfw_m3_info_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0; memset(&req, 0, sizeof(req)); @@ -2277,6 +2281,7 @@ static int ath11k_qmi_wlanfw_m3_info_send(struct ath11k_base *ab) QMI_WLANFW_M3_INFO_REQ_MSG_V01_MAX_MSG_LEN, qmi_wlanfw_m3_info_req_msg_v01_ei, &req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send m3 information request: %d\n", ret); goto out; @@ -2303,7 +2308,7 @@ static int ath11k_qmi_wlanfw_mode_send(struct ath11k_base *ab, { struct qmi_wlanfw_wlan_mode_req_msg_v01 req; struct qmi_wlanfw_wlan_mode_resp_msg_v01 resp; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0; memset(&req, 0, sizeof(req)); @@ -2325,6 +2330,7 @@ static int ath11k_qmi_wlanfw_mode_send(struct ath11k_base *ab, QMI_WLANFW_WLAN_MODE_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_wlan_mode_req_msg_v01_ei, &req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send wlan mode request (mode %d): %d\n", mode, ret); goto out; @@ -2358,7 +2364,7 @@ static int ath11k_qmi_wlanfw_wlan_cfg_send(struct ath11k_base *ab) struct qmi_wlanfw_wlan_cfg_resp_msg_v01 resp; struct ce_pipe_config *ce_cfg; struct service_to_pipe *svc_cfg; - struct qmi_txn txn = {}; + struct qmi_txn txn; int ret = 0, pipe_num; ce_cfg = (struct ce_pipe_config *)ab->qmi.ce_cfg.tgt_ce; @@ -2419,6 +2425,7 @@ static int ath11k_qmi_wlanfw_wlan_cfg_send(struct ath11k_base *ab) QMI_WLANFW_WLAN_CFG_REQ_MSG_V01_MAX_LEN, qmi_wlanfw_wlan_cfg_req_msg_v01_ei, req); if (ret < 0) { + qmi_txn_cancel(&txn); ath11k_warn(ab, "failed to send wlan config request: %d\n", ret); goto out; From fb12305aff12e735e599c79514dde5dac40f5a59 Mon Sep 17 00:00:00 2001 From: Venkateswara Naralasetty Date: Tue, 9 Nov 2021 12:05:55 +0530 Subject: [PATCH 0275/2048] ath11k: add trace log support This change is to add trace log support for, * WMI events * WMI commands * ath11k_dbg messages * ath11k_dbg_dump messages * ath11k_log_info messages * ath11k_log_warn messages * ath11k_log_err messages Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-00652-QCAHKSWPL_SILICONZ-1 Signed-off-by: Venkateswara Naralasetty Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636439755-30419-1-git-send-email-quic_vnaralas@quicinc.com --- drivers/net/wireless/ath/ath11k/debug.c | 12 +- drivers/net/wireless/ath/ath11k/debug.h | 3 +- drivers/net/wireless/ath/ath11k/trace.c | 1 + drivers/net/wireless/ath/ath11k/trace.h | 172 ++++++++++++++++++++++++ drivers/net/wireless/ath/ath11k/wmi.c | 4 + 5 files changed, 187 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/debug.c b/drivers/net/wireless/ath/ath11k/debug.c index c86de95fbdc5..958d87429062 100644 --- a/drivers/net/wireless/ath/ath11k/debug.c +++ b/drivers/net/wireless/ath/ath11k/debug.c @@ -17,7 +17,7 @@ void ath11k_info(struct ath11k_base *ab, const char *fmt, ...) va_start(args, fmt); vaf.va = &args; dev_info(ab->dev, "%pV", &vaf); - /* TODO: Trace the log */ + trace_ath11k_log_info(ab, &vaf); va_end(args); } EXPORT_SYMBOL(ath11k_info); @@ -32,7 +32,7 @@ void ath11k_err(struct ath11k_base *ab, const char *fmt, ...) va_start(args, fmt); vaf.va = &args; dev_err(ab->dev, "%pV", &vaf); - /* TODO: Trace the log */ + trace_ath11k_log_err(ab, &vaf); va_end(args); } EXPORT_SYMBOL(ath11k_err); @@ -47,7 +47,7 @@ void ath11k_warn(struct ath11k_base *ab, const char *fmt, ...) va_start(args, fmt); vaf.va = &args; dev_warn_ratelimited(ab->dev, "%pV", &vaf); - /* TODO: Trace the log */ + trace_ath11k_log_warn(ab, &vaf); va_end(args); } EXPORT_SYMBOL(ath11k_warn); @@ -68,7 +68,7 @@ void __ath11k_dbg(struct ath11k_base *ab, enum ath11k_debug_mask mask, if (ath11k_debug_mask & mask) dev_printk(KERN_DEBUG, ab->dev, "%pV", &vaf); - /* TODO: trace log */ + trace_ath11k_log_dbg(ab, mask, &vaf); va_end(args); } @@ -100,6 +100,10 @@ void ath11k_dbg_dump(struct ath11k_base *ab, dev_printk(KERN_DEBUG, ab->dev, "%s\n", linebuf); } } + + /* tracing code doesn't like null strings */ + trace_ath11k_log_dbg_dump(ab, msg ? msg : "", prefix ? prefix : "", + buf, len); } EXPORT_SYMBOL(ath11k_dbg_dump); diff --git a/drivers/net/wireless/ath/ath11k/debug.h b/drivers/net/wireless/ath/ath11k/debug.h index 659a275e2eb3..fbbd5fe02aa8 100644 --- a/drivers/net/wireless/ath/ath11k/debug.h +++ b/drivers/net/wireless/ath/ath11k/debug.h @@ -60,7 +60,8 @@ static inline void ath11k_dbg_dump(struct ath11k_base *ab, #define ath11k_dbg(ar, dbg_mask, fmt, ...) \ do { \ - if (ath11k_debug_mask & dbg_mask) \ + if ((ath11k_debug_mask & dbg_mask) || \ + trace_ath11k_log_dbg_enabled()) \ __ath11k_dbg(ar, dbg_mask, fmt, ##__VA_ARGS__); \ } while (0) diff --git a/drivers/net/wireless/ath/ath11k/trace.c b/drivers/net/wireless/ath/ath11k/trace.c index f0cc49ba0387..6620650d7845 100644 --- a/drivers/net/wireless/ath/ath11k/trace.c +++ b/drivers/net/wireless/ath/ath11k/trace.c @@ -7,3 +7,4 @@ #define CREATE_TRACE_POINTS #include "trace.h" +EXPORT_SYMBOL(__tracepoint_ath11k_log_dbg); diff --git a/drivers/net/wireless/ath/ath11k/trace.h b/drivers/net/wireless/ath/ath11k/trace.h index 25d18e9d5b0b..02003dc4207d 100644 --- a/drivers/net/wireless/ath/ath11k/trace.h +++ b/drivers/net/wireless/ath/ath11k/trace.h @@ -14,12 +14,24 @@ #if !defined(CONFIG_ATH11K_TRACING) #undef TRACE_EVENT #define TRACE_EVENT(name, proto, ...) \ +static inline void trace_ ## name(proto) {} \ +static inline bool trace_##name##_enabled(void) \ +{ \ + return false; \ +} + +#undef DECLARE_EVENT_CLASS +#define DECLARE_EVENT_CLASS(...) +#undef DEFINE_EVENT +#define DEFINE_EVENT(evt_class, name, proto, ...) \ static inline void trace_ ## name(proto) {} #endif /* !CONFIG_ATH11K_TRACING || __CHECKER__ */ #undef TRACE_SYSTEM #define TRACE_SYSTEM ath11k +#define ATH11K_MSG_MAX 400 + TRACE_EVENT(ath11k_htt_pktlog, TP_PROTO(struct ath11k *ar, const void *buf, u16 buf_len, u32 pktlog_checksum), @@ -108,6 +120,166 @@ TRACE_EVENT(ath11k_htt_rxdesc, ) ); +DECLARE_EVENT_CLASS(ath11k_log_event, + TP_PROTO(struct ath11k_base *ab, struct va_format *vaf), + TP_ARGS(ab, vaf), + TP_STRUCT__entry( + __string(device, dev_name(ab->dev)) + __string(driver, dev_driver_string(ab->dev)) + __dynamic_array(char, msg, ATH11K_MSG_MAX) + ), + TP_fast_assign( + __assign_str(device, dev_name(ab->dev)); + __assign_str(driver, dev_driver_string(ab->dev)); + WARN_ON_ONCE(vsnprintf(__get_dynamic_array(msg), + ATH11K_MSG_MAX, + vaf->fmt, + *vaf->va) >= ATH11K_MSG_MAX); + ), + TP_printk( + "%s %s %s", + __get_str(driver), + __get_str(device), + __get_str(msg) + ) +); + +DEFINE_EVENT(ath11k_log_event, ath11k_log_err, + TP_PROTO(struct ath11k_base *ab, struct va_format *vaf), + TP_ARGS(ab, vaf) +); + +DEFINE_EVENT(ath11k_log_event, ath11k_log_warn, + TP_PROTO(struct ath11k_base *ab, struct va_format *vaf), + TP_ARGS(ab, vaf) +); + +DEFINE_EVENT(ath11k_log_event, ath11k_log_info, + TP_PROTO(struct ath11k_base *ab, struct va_format *vaf), + TP_ARGS(ab, vaf) +); + +TRACE_EVENT(ath11k_wmi_cmd, + TP_PROTO(struct ath11k_base *ab, int id, const void *buf, size_t buf_len), + + TP_ARGS(ab, id, buf, buf_len), + + TP_STRUCT__entry( + __string(device, dev_name(ab->dev)) + __string(driver, dev_driver_string(ab->dev)) + __field(unsigned int, id) + __field(size_t, buf_len) + __dynamic_array(u8, buf, buf_len) + ), + + TP_fast_assign( + __assign_str(device, dev_name(ab->dev)); + __assign_str(driver, dev_driver_string(ab->dev)); + __entry->id = id; + __entry->buf_len = buf_len; + memcpy(__get_dynamic_array(buf), buf, buf_len); + ), + + TP_printk( + "%s %s id %d len %zu", + __get_str(driver), + __get_str(device), + __entry->id, + __entry->buf_len + ) +); + +TRACE_EVENT(ath11k_wmi_event, + TP_PROTO(struct ath11k_base *ab, int id, const void *buf, size_t buf_len), + + TP_ARGS(ab, id, buf, buf_len), + + TP_STRUCT__entry( + __string(device, dev_name(ab->dev)) + __string(driver, dev_driver_string(ab->dev)) + __field(unsigned int, id) + __field(size_t, buf_len) + __dynamic_array(u8, buf, buf_len) + ), + + TP_fast_assign( + __assign_str(device, dev_name(ab->dev)); + __assign_str(driver, dev_driver_string(ab->dev)); + __entry->id = id; + __entry->buf_len = buf_len; + memcpy(__get_dynamic_array(buf), buf, buf_len); + ), + + TP_printk( + "%s %s id %d len %zu", + __get_str(driver), + __get_str(device), + __entry->id, + __entry->buf_len + ) +); + +TRACE_EVENT(ath11k_log_dbg, + TP_PROTO(struct ath11k_base *ab, unsigned int level, struct va_format *vaf), + + TP_ARGS(ab, level, vaf), + + TP_STRUCT__entry( + __string(device, dev_name(ab->dev)) + __string(driver, dev_driver_string(ab->dev)) + __field(unsigned int, level) + __dynamic_array(char, msg, ATH11K_MSG_MAX) + ), + + TP_fast_assign( + __assign_str(device, dev_name(ab->dev)); + __assign_str(driver, dev_driver_string(ab->dev)); + __entry->level = level; + WARN_ON_ONCE(vsnprintf(__get_dynamic_array(msg), + ATH11K_MSG_MAX, vaf->fmt, + *vaf->va) >= ATH11K_MSG_MAX); + ), + + TP_printk( + "%s %s %s", + __get_str(driver), + __get_str(device), + __get_str(msg) + ) +); + +TRACE_EVENT(ath11k_log_dbg_dump, + TP_PROTO(struct ath11k_base *ab, const char *msg, const char *prefix, + const void *buf, size_t buf_len), + + TP_ARGS(ab, msg, prefix, buf, buf_len), + + TP_STRUCT__entry( + __string(device, dev_name(ab->dev)) + __string(driver, dev_driver_string(ab->dev)) + __string(msg, msg) + __string(prefix, prefix) + __field(size_t, buf_len) + __dynamic_array(u8, buf, buf_len) + ), + + TP_fast_assign( + __assign_str(device, dev_name(ab->dev)); + __assign_str(driver, dev_driver_string(ab->dev)); + __assign_str(msg, msg); + __assign_str(prefix, prefix); + __entry->buf_len = buf_len; + memcpy(__get_dynamic_array(buf), buf, buf_len); + ), + + TP_printk( + "%s %s %s/%s\n", + __get_str(driver), + __get_str(device), + __get_str(prefix), + __get_str(msg) + ) +); #endif /* _TRACE_H_ || TRACE_HEADER_MULTI_READ*/ /* we don't want to use include/trace/events */ diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 6bd355ce88d3..47e7fd127fd3 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -249,6 +249,8 @@ static int ath11k_wmi_cmd_send_nowait(struct ath11k_pdev_wmi *wmi, struct sk_buf cmd_hdr = (struct wmi_cmd_hdr *)skb->data; cmd_hdr->cmd_id = cmd; + trace_ath11k_wmi_cmd(ab, cmd_id, skb->data, skb->len); + memset(skb_cb, 0, sizeof(*skb_cb)); ret = ath11k_htc_send(&ab->htc, wmi->eid, skb); @@ -7103,6 +7105,8 @@ static void ath11k_wmi_tlv_op_rx(struct ath11k_base *ab, struct sk_buff *skb) cmd_hdr = (struct wmi_cmd_hdr *)skb->data; id = FIELD_GET(WMI_CMD_HDR_CMD_ID, (cmd_hdr->cmd_id)); + trace_ath11k_wmi_event(ab, id, skb->data, skb->len); + if (skb_pull(skb, sizeof(struct wmi_cmd_hdr)) == NULL) goto out; From 273703ebdb01b6c5f1aaf4b98fb57b177609055c Mon Sep 17 00:00:00 2001 From: Baochen Qiang Date: Tue, 26 Oct 2021 09:16:05 +0800 Subject: [PATCH 0276/2048] ath11k: Fix crash caused by uninitialized TX ring Commit 31582373a4a8 ("ath11k: Change number of TCL rings to one for QCA6390") avoids initializing the other entries of dp->tx_ring cause the corresponding TX rings on QCA6390/WCN6855 are not used, but leaves those ring masks in ath11k_hw_ring_mask_qca6390.tx unchanged. Normally this is OK because we will only get interrupts from the first TX ring on these chips and thus only the first entry of dp->tx_ring is involved. In case of one MSI vector, all DP rings share the same IRQ. For each interrupt, all rings have to be checked, which means the other entries of dp->tx_ring are involved. However since they are not initialized, system crashes. Fix this issue by simply removing those ring masks. crash stack: [ 102.907438] BUG: kernel NULL pointer dereference, address: 0000000000000028 [ 102.907447] #PF: supervisor read access in kernel mode [ 102.907451] #PF: error_code(0x0000) - not-present page [ 102.907453] PGD 1081f0067 P4D 1081f0067 PUD 1081f1067 PMD 0 [ 102.907460] Oops: 0000 [#1] SMP DEBUG_PAGEALLOC NOPTI [ 102.907465] CPU: 0 PID: 3511 Comm: apt-check Kdump: loaded Tainted: G E 5.15.0-rc4-wt-ath+ #20 [ 102.907470] Hardware name: AMD Celadon-RN/Celadon-RN, BIOS RCD1005E 10/08/2020 [ 102.907472] RIP: 0010:ath11k_dp_tx_completion_handler+0x201/0x830 [ath11k] [ 102.907497] Code: 3c 24 4e 8d ac 37 10 04 00 00 4a 8d bc 37 68 04 00 00 48 89 3c 24 48 63 c8 89 83 84 18 00 00 48 c1 e1 05 48 03 8b 78 18 00 00 <8b> 51 08 89 d6 83 e6 07 89 74 24 24 83 fe 03 74 04 85 f6 75 63 41 [ 102.907501] RSP: 0000:ffff9b7340003e08 EFLAGS: 00010202 [ 102.907505] RAX: 0000000000000001 RBX: ffff8e21530c0100 RCX: 0000000000000020 [ 102.907508] RDX: 0000000000000000 RSI: 00000000fffffe00 RDI: ffff8e21530c1938 [ 102.907511] RBP: ffff8e21530c0000 R08: 0000000000000001 R09: 0000000000000000 [ 102.907513] R10: ffff8e2145534c10 R11: 0000000000000001 R12: ffff8e21530c2938 [ 102.907515] R13: ffff8e21530c18e0 R14: 0000000000000100 R15: ffff8e21530c2978 [ 102.907518] FS: 00007f5d4297e740(0000) GS:ffff8e243d600000(0000) knlGS:0000000000000000 [ 102.907521] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 102.907524] CR2: 0000000000000028 CR3: 00000001034ea000 CR4: 0000000000350ef0 [ 102.907527] Call Trace: [ 102.907531] [ 102.907537] ath11k_dp_service_srng+0x5c/0x2f0 [ath11k] [ 102.907556] ath11k_pci_ext_grp_napi_poll+0x21/0x70 [ath11k_pci] [ 102.907562] __napi_poll+0x2c/0x160 [ 102.907570] net_rx_action+0x251/0x310 [ 102.907576] __do_softirq+0x107/0x2fc [ 102.907585] irq_exit_rcu+0x74/0x90 [ 102.907593] common_interrupt+0x83/0xa0 [ 102.907600] [ 102.907601] asm_common_interrupt+0x1e/0x40 Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1 Signed-off-by: Baochen Qiang Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211026011605.58615-1-quic_bqiang@quicinc.com --- drivers/net/wireless/ath/ath11k/hw.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/hw.c b/drivers/net/wireless/ath/ath11k/hw.c index da35fcf5bc56..2f0b526188e4 100644 --- a/drivers/net/wireless/ath/ath11k/hw.c +++ b/drivers/net/wireless/ath/ath11k/hw.c @@ -1061,8 +1061,6 @@ const struct ath11k_hw_ring_mask ath11k_hw_ring_mask_ipq8074 = { const struct ath11k_hw_ring_mask ath11k_hw_ring_mask_qca6390 = { .tx = { ATH11K_TX_RING_MASK_0, - ATH11K_TX_RING_MASK_1, - ATH11K_TX_RING_MASK_2, }, .rx_mon_status = { 0, 0, 0, 0, From fc95d10ac41d75c14a81afcc8722333d8b2cf80f Mon Sep 17 00:00:00 2001 From: Wen Gong Date: Mon, 15 Nov 2021 11:29:55 +0200 Subject: [PATCH 0277/2048] ath11k: add string type to search board data in board-2.bin for WCN6855 Currently ath11k only support string type with bus, chip id and board id such as "bus=ahb,qmi-chip-id=1,qmi-board-id=4" for ahb bus chip and "bus=pci,qmi-chip-id=0,qmi-board-id=255" for PCIe bus chip in board-2.bin. For WCN6855, it is not enough to distinguish all different chips. This is to add a new string type which include bus, chip id, board id, vendor, device, subsystem-vendor and subsystem-device for WCN6855. ath11k will first load board-2.bin and search in it for the board data with the above parameters, if matched one board data, then download it to firmware, if not matched any one, then ath11k will download the file board.bin to firmware. Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1 Signed-off-by: Wen Gong Signed-off-by: Jouni Malinen Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211111065340.20187-1-quic_wgong@quicinc.com --- drivers/net/wireless/ath/ath11k/core.c | 27 ++++++++++++++++++++------ drivers/net/wireless/ath/ath11k/core.h | 13 +++++++++++++ drivers/net/wireless/ath/ath11k/pci.c | 10 ++++++++++ 3 files changed, 44 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index 4e994db996db..a40bbca3e9af 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -412,11 +412,26 @@ static int ath11k_core_create_board_name(struct ath11k_base *ab, char *name, scnprintf(variant, sizeof(variant), ",variant=%s", ab->qmi.target.bdf_ext); - scnprintf(name, name_len, - "bus=%s,qmi-chip-id=%d,qmi-board-id=%d%s", - ath11k_bus_str(ab->hif.bus), - ab->qmi.target.chip_id, - ab->qmi.target.board_id, variant); + switch (ab->id.bdf_search) { + case ATH11K_BDF_SEARCH_BUS_AND_BOARD: + scnprintf(name, name_len, + "bus=%s,vendor=%04x,device=%04x,subsystem-vendor=%04x,subsystem-device=%04x,qmi-chip-id=%d,qmi-board-id=%d%s", + ath11k_bus_str(ab->hif.bus), + ab->id.vendor, ab->id.device, + ab->id.subsystem_vendor, + ab->id.subsystem_device, + ab->qmi.target.chip_id, + ab->qmi.target.board_id, + variant); + break; + default: + scnprintf(name, name_len, + "bus=%s,qmi-chip-id=%d,qmi-board-id=%d%s", + ath11k_bus_str(ab->hif.bus), + ab->qmi.target.chip_id, + ab->qmi.target.board_id, variant); + break; + } ath11k_dbg(ab, ATH11K_DBG_BOOT, "boot using board name '%s'\n", name); @@ -653,7 +668,7 @@ static int ath11k_core_fetch_board_data_api_1(struct ath11k_base *ab, return 0; } -#define BOARD_NAME_SIZE 100 +#define BOARD_NAME_SIZE 200 int ath11k_core_fetch_bdf(struct ath11k_base *ab, struct ath11k_board_data *bd) { char boardname[BOARD_NAME_SIZE]; diff --git a/drivers/net/wireless/ath/ath11k/core.h b/drivers/net/wireless/ath/ath11k/core.h index 2f1e10b7cc17..7ee5d28787e6 100644 --- a/drivers/net/wireless/ath/ath11k/core.h +++ b/drivers/net/wireless/ath/ath11k/core.h @@ -47,6 +47,11 @@ enum ath11k_supported_bw { ATH11K_BW_160 = 3, }; +enum ath11k_bdf_search { + ATH11K_BDF_SEARCH_DEFAULT, + ATH11K_BDF_SEARCH_BUS_AND_BOARD, +}; + enum wme_ac { WME_AC_BE, WME_AC_BK, @@ -760,6 +765,14 @@ struct ath11k_base { struct completion htc_suspend; + struct { + enum ath11k_bdf_search bdf_search; + u32 vendor; + u32 device; + u32 subsystem_vendor; + u32 subsystem_device; + } id; + /* must be last */ u8 drv_priv[0] __aligned(sizeof(void *)); }; diff --git a/drivers/net/wireless/ath/ath11k/pci.c b/drivers/net/wireless/ath/ath11k/pci.c index 3d353e7c9d5c..958c2b2c5b3e 100644 --- a/drivers/net/wireless/ath/ath11k/pci.c +++ b/drivers/net/wireless/ath/ath11k/pci.c @@ -1251,6 +1251,15 @@ static int ath11k_pci_probe(struct pci_dev *pdev, goto err_free_core; } + ath11k_dbg(ab, ATH11K_DBG_BOOT, "pci probe %04x:%04x %04x:%04x\n", + pdev->vendor, pdev->device, + pdev->subsystem_vendor, pdev->subsystem_device); + + ab->id.vendor = pdev->vendor; + ab->id.device = pdev->device; + ab->id.subsystem_vendor = pdev->subsystem_vendor; + ab->id.subsystem_device = pdev->subsystem_device; + switch (pci_dev->device) { case QCA6390_DEVICE_ID: ath11k_pci_read_hw_version(ab, &soc_hw_version_major, @@ -1273,6 +1282,7 @@ static int ath11k_pci_probe(struct pci_dev *pdev, ab->hw_rev = ATH11K_HW_QCN9074_HW10; break; case WCN6855_DEVICE_ID: + ab->id.bdf_search = ATH11K_BDF_SEARCH_BUS_AND_BOARD; ath11k_pci_read_hw_version(ab, &soc_hw_version_major, &soc_hw_version_minor); switch (soc_hw_version_major) { From 886433a984254c6d2c2074688dc8f48c40b1c070 Mon Sep 17 00:00:00 2001 From: John Crispin Date: Mon, 25 Oct 2021 21:40:54 +0530 Subject: [PATCH 0278/2048] ath11k: add support for BSS color change Whenever the MAC detects a color collision, or any of its associated stations detects one, the firmware will send out an event. Add the code to parse and handle this event and pass the data up to mac80211. The firmware does not provide an offload feature such as the one used for CSA. The color change process is hence triggered via the beacon offload tx completion events sent out by firmware. BSS color feature is enabled depending on service flag advertised by firmware, based on which color change functionality is invoked. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-00680-QCAHKSWPL_SILICONZ-1 Signed-off-by: John Crispin Co-developed-by: Lavanya Suresh Signed-off-by: Lavanya Suresh Signed-off-by: Rameshkumar Sundaram Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1635178254-17732-1-git-send-email-quic_ramess@quicinc.com --- drivers/net/wireless/ath/ath11k/core.h | 2 + drivers/net/wireless/ath/ath11k/mac.c | 57 +++++++++++++++++-- drivers/net/wireless/ath/ath11k/mac.h | 1 + drivers/net/wireless/ath/ath11k/wmi.c | 78 +++++++++++++++++++++++++- drivers/net/wireless/ath/ath11k/wmi.h | 15 +++++ 5 files changed, 147 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/core.h b/drivers/net/wireless/ath/ath11k/core.h index 7ee5d28787e6..bbfc10fd5c6d 100644 --- a/drivers/net/wireless/ath/ath11k/core.h +++ b/drivers/net/wireless/ath/ath11k/core.h @@ -255,6 +255,8 @@ struct ath11k_vif { int txpower; bool rsnie_present; bool wpaie_present; + bool bcca_zero_sent; + bool do_not_send_tmpl; struct ieee80211_chanctx_conf chanctx; }; diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 1320c8634d31..853f8f216913 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -1231,6 +1231,26 @@ static int ath11k_mac_setup_bcn_tmpl(struct ath11k_vif *arvif) return ret; } +void ath11k_mac_bcn_tx_event(struct ath11k_vif *arvif) +{ + struct ieee80211_vif *vif = arvif->vif; + + if (!vif->color_change_active && !arvif->bcca_zero_sent) + return; + + if (vif->color_change_active && ieee80211_beacon_cntdwn_is_complete(vif)) { + arvif->bcca_zero_sent = true; + ieee80211_color_change_finish(vif); + return; + } + + arvif->bcca_zero_sent = false; + + if (vif->color_change_active) + ieee80211_beacon_update_cntdwn(vif); + ath11k_mac_setup_bcn_tmpl(arvif); +} + static void ath11k_control_beaconing(struct ath11k_vif *arvif, struct ieee80211_bss_conf *info) { @@ -2894,10 +2914,17 @@ static void ath11k_mac_op_bss_info_changed(struct ieee80211_hw *hw, "Set staggered beacon mode for VDEV: %d\n", arvif->vdev_id); - ret = ath11k_mac_setup_bcn_tmpl(arvif); - if (ret) - ath11k_warn(ar->ab, "failed to update bcn template: %d\n", - ret); + if (!arvif->do_not_send_tmpl || !arvif->bcca_zero_sent) { + ret = ath11k_mac_setup_bcn_tmpl(arvif); + if (ret) + ath11k_warn(ar->ab, "failed to update bcn template: %d\n", + ret); + } + + if (arvif->bcca_zero_sent) + arvif->do_not_send_tmpl = true; + else + arvif->do_not_send_tmpl = false; } if (changed & (BSS_CHANGED_BEACON_INFO | BSS_CHANGED_BEACON)) { @@ -3108,6 +3135,25 @@ static void ath11k_mac_op_bss_info_changed(struct ieee80211_hw *hw, if (ret) ath11k_warn(ar->ab, "failed to set bss color collision on vdev %i: %d\n", arvif->vdev_id, ret); + + param_id = WMI_VDEV_PARAM_BSS_COLOR; + if (info->he_bss_color.enabled) + param_value = info->he_bss_color.color << + IEEE80211_HE_OPERATION_BSS_COLOR_OFFSET; + else + param_value = IEEE80211_HE_OPERATION_BSS_COLOR_DISABLED; + + ret = ath11k_wmi_vdev_set_param_cmd(ar, arvif->vdev_id, + param_id, + param_value); + if (ret) + ath11k_warn(ar->ab, + "failed to set bss color param on vdev %i: %d\n", + arvif->vdev_id, ret); + + ath11k_dbg(ar->ab, ATH11K_DBG_MAC, + "bss color param 0x%x set on vdev %i\n", + param_value, arvif->vdev_id); } else if (vif->type == NL80211_IFTYPE_STATION) { ret = ath11k_wmi_send_bss_color_change_enable_cmd(ar, arvif->vdev_id, @@ -7872,6 +7918,9 @@ static int __ath11k_mac_register(struct ath11k *ar) wiphy_ext_feature_set(ar->hw->wiphy, NL80211_EXT_FEATURE_CQM_RSSI_LIST); wiphy_ext_feature_set(ar->hw->wiphy, NL80211_EXT_FEATURE_STA_TX_PWR); + if (test_bit(WMI_TLV_SERVICE_BSS_COLOR_OFFLOAD, ar->ab->wmi_ab.svc_map)) + wiphy_ext_feature_set(ar->hw->wiphy, + NL80211_EXT_FEATURE_BSS_COLOR); ar->hw->wiphy->cipher_suites = cipher_suites; ar->hw->wiphy->n_cipher_suites = ARRAY_SIZE(cipher_suites); diff --git a/drivers/net/wireless/ath/ath11k/mac.h b/drivers/net/wireless/ath/ath11k/mac.h index 254ca4acc8e8..f6f37e8c8c6a 100644 --- a/drivers/net/wireless/ath/ath11k/mac.h +++ b/drivers/net/wireless/ath/ath11k/mac.h @@ -155,4 +155,5 @@ enum ath11k_supported_bw ath11k_mac_mac80211_bw_to_ath11k_bw(enum rate_info_bw b enum hal_encrypt_type ath11k_dp_tx_get_encrypt_type(u32 cipher); void ath11k_mac_handle_beacon(struct ath11k *ar, struct sk_buff *skb); void ath11k_mac_handle_beacon_miss(struct ath11k *ar, u32 vdev_id); +void ath11k_mac_bcn_tx_event(struct ath11k_vif *arvif); #endif diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 47e7fd127fd3..51ed0f2c8b73 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -128,6 +128,8 @@ static const struct wmi_tlv_policy wmi_tlv_policies[] = { .min_len = sizeof(struct wmi_probe_resp_tx_status_event) }, [WMI_TAG_VDEV_DELETE_RESP_EVENT] = { .min_len = sizeof(struct wmi_vdev_delete_resp_event) }, + [WMI_TAG_OBSS_COLOR_COLLISION_EVT] = { + .min_len = sizeof(struct wmi_obss_color_collision_event) }, }; #define PRIMAP(_hw_mode_) \ @@ -1633,6 +1635,15 @@ int ath11k_wmi_bcn_tmpl(struct ath11k *ar, u32 vdev_id, void *ptr; int ret, len; size_t aligned_len = roundup(bcn->len, 4); + struct ieee80211_vif *vif; + struct ath11k_vif *arvif = ath11k_mac_get_arvif(ar, vdev_id); + + if (!arvif) { + ath11k_warn(ar->ab, "failed to find arvif with vdev id %d\n", vdev_id); + return -EINVAL; + } + + vif = arvif->vif; len = sizeof(*cmd) + sizeof(*bcn_prb_info) + TLV_HDR_SIZE + aligned_len; @@ -1645,8 +1656,12 @@ int ath11k_wmi_bcn_tmpl(struct ath11k *ar, u32 vdev_id, FIELD_PREP(WMI_TLV_LEN, sizeof(*cmd) - TLV_HDR_SIZE); cmd->vdev_id = vdev_id; cmd->tim_ie_offset = offs->tim_offset; - cmd->csa_switch_count_offset = offs->cntdwn_counter_offs[0]; - cmd->ext_csa_switch_count_offset = offs->cntdwn_counter_offs[1]; + + if (vif->csa_active) { + cmd->csa_switch_count_offset = offs->cntdwn_counter_offs[0]; + cmd->ext_csa_switch_count_offset = offs->cntdwn_counter_offs[1]; + } + cmd->buf_len = bcn->len; ptr = skb->data + sizeof(*cmd); @@ -3450,6 +3465,53 @@ int ath11k_wmi_fils_discovery(struct ath11k *ar, u32 vdev_id, u32 interval, return ret; } +static void +ath11k_wmi_obss_color_collision_event(struct ath11k_base *ab, struct sk_buff *skb) +{ + const void **tb; + const struct wmi_obss_color_collision_event *ev; + struct ath11k_vif *arvif; + int ret; + + tb = ath11k_wmi_tlv_parse_alloc(ab, skb->data, skb->len, GFP_ATOMIC); + if (IS_ERR(tb)) { + ret = PTR_ERR(tb); + ath11k_warn(ab, "failed to parse tlv: %d\n", ret); + return; + } + + ev = tb[WMI_TAG_OBSS_COLOR_COLLISION_EVT]; + if (!ev) { + ath11k_warn(ab, "failed to fetch obss color collision ev"); + goto exit; + } + + arvif = ath11k_mac_get_arvif_by_vdev_id(ab, ev->vdev_id); + if (!arvif) { + ath11k_warn(ab, "failed to find arvif with vedv id %d in obss_color_collision_event\n", + ev->vdev_id); + goto exit; + } + + switch (ev->evt_type) { + case WMI_BSS_COLOR_COLLISION_DETECTION: + ieeee80211_obss_color_collision_notify(arvif->vif, ev->obss_color_bitmap); + ath11k_dbg(ab, ATH11K_DBG_WMI, + "OBSS color collision detected vdev:%d, event:%d, bitmap:%08llx\n", + ev->vdev_id, ev->evt_type, ev->obss_color_bitmap); + break; + case WMI_BSS_COLOR_COLLISION_DISABLE: + case WMI_BSS_COLOR_FREE_SLOT_TIMER_EXPIRY: + case WMI_BSS_COLOR_FREE_SLOT_AVAILABLE: + break; + default: + ath11k_warn(ab, "received unknown obss color collision detetction event\n"); + } + +exit: + kfree(tb); +} + static void ath11k_fill_band_to_mac_param(struct ath11k_base *soc, struct wmi_host_pdev_band_to_mac *band_to_mac) @@ -6157,6 +6219,7 @@ static void ath11k_vdev_start_resp_event(struct ath11k_base *ab, struct sk_buff static void ath11k_bcn_tx_status_event(struct ath11k_base *ab, struct sk_buff *skb) { + struct ath11k_vif *arvif; u32 vdev_id, tx_status; if (ath11k_pull_bcn_tx_status_ev(ab, skb->data, skb->len, @@ -6164,6 +6227,14 @@ static void ath11k_bcn_tx_status_event(struct ath11k_base *ab, struct sk_buff *s ath11k_warn(ab, "failed to extract bcn tx status"); return; } + + arvif = ath11k_mac_get_arvif_by_vdev_id(ab, vdev_id); + if (!arvif) { + ath11k_warn(ab, "invalid vdev id %d in bcn_tx_status", + vdev_id); + return; + } + ath11k_mac_bcn_tx_event(arvif); } static void ath11k_vdev_stopped_event(struct ath11k_base *ab, struct sk_buff *skb) @@ -7191,6 +7262,9 @@ static void ath11k_wmi_tlv_op_rx(struct ath11k_base *ab, struct sk_buff *skb) case WMI_OFFLOAD_PROB_RESP_TX_STATUS_EVENTID: ath11k_probe_resp_tx_status_event(ab, skb); break; + case WMI_OBSS_COLOR_COLLISION_DETECTION_EVENTID: + ath11k_wmi_obss_color_collision_event(ab, skb); + break; /* add Unsupported events here */ case WMI_TBTTOFFSET_EXT_UPDATE_EVENTID: case WMI_PEER_OPER_MODE_CHANGE_EVENTID: diff --git a/drivers/net/wireless/ath/ath11k/wmi.h b/drivers/net/wireless/ath/ath11k/wmi.h index b5b9f2c22e2e..4eb06cb7f883 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.h +++ b/drivers/net/wireless/ath/ath11k/wmi.h @@ -774,6 +774,8 @@ enum wmi_tlv_event_id { WMI_MDNS_STATS_EVENTID = WMI_TLV_CMD(WMI_GRP_MDNS_OFL), WMI_SAP_OFL_ADD_STA_EVENTID = WMI_TLV_CMD(WMI_GRP_SAP_OFL), WMI_SAP_OFL_DEL_STA_EVENTID, + WMI_OBSS_COLOR_COLLISION_DETECTION_EVENTID = + WMI_EVT_GRP_START_ID(WMI_GRP_OBSS_OFL), WMI_OCB_SET_CONFIG_RESP_EVENTID = WMI_TLV_CMD(WMI_GRP_OCB), WMI_OCB_GET_TSF_TIMER_RESP_EVENTID, WMI_DCC_GET_STATS_RESP_EVENTID, @@ -4916,6 +4918,13 @@ struct wmi_pdev_obss_pd_bitmap_cmd { #define ATH11K_BSS_COLOR_COLLISION_DETECTION_STA_PERIOD_MS 10000 #define ATH11K_BSS_COLOR_COLLISION_DETECTION_AP_PERIOD_MS 5000 +enum wmi_bss_color_collision { + WMI_BSS_COLOR_COLLISION_DISABLE = 0, + WMI_BSS_COLOR_COLLISION_DETECTION, + WMI_BSS_COLOR_FREE_SLOT_TIMER_EXPIRY, + WMI_BSS_COLOR_FREE_SLOT_AVAILABLE, +}; + struct wmi_obss_color_collision_cfg_params_cmd { u32 tlv_header; u32 vdev_id; @@ -4933,6 +4942,12 @@ struct wmi_bss_color_change_enable_params_cmd { u32 enable; } __packed; +struct wmi_obss_color_collision_event { + u32 vdev_id; + u32 evt_type; + u64 obss_color_bitmap; +} __packed; + #define ATH11K_IPV4_TH_SEED_SIZE 5 #define ATH11K_IPV6_TH_SEED_SIZE 11 From 436a4e88659842a7cf634d7cc088c8f2cc94ebf5 Mon Sep 17 00:00:00 2001 From: Karthikeyan Kathirvel Date: Mon, 15 Nov 2021 11:04:40 +0100 Subject: [PATCH 0279/2048] ath11k: clear the keys properly via DISABLE_KEY DISABLE_KEY sets the key_len to 0, firmware will not delete the keys if key_len is 0. Changing from security mode to open mode will cause mcast to be still encrypted without vdev restart. Set the proper key_len for DISABLE_KEY cmd to clear the keys in firmware. Tested-on: IPQ6018 hw1.0 AHB WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1 Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Reported-by: Sven Eckelmann Signed-off-by: Karthikeyan Kathirvel [sven@narfation.org: split into separate patches, clean up commit message] Signed-off-by: Sven Eckelmann Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211115100441.33771-1-sven@narfation.org --- drivers/net/wireless/ath/ath11k/mac.c | 4 +--- drivers/net/wireless/ath/ath11k/wmi.c | 3 ++- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index 853f8f216913..df6421fad160 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -3461,9 +3461,7 @@ static int ath11k_install_key(struct ath11k_vif *arvif, return 0; if (cmd == DISABLE_KEY) { - /* TODO: Check if FW expects value other than NONE for del */ - /* arg.key_cipher = WMI_CIPHER_NONE; */ - arg.key_len = 0; + arg.key_cipher = WMI_CIPHER_NONE; arg.key_data = NULL; goto install; } diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c index 51ed0f2c8b73..614b2f6bcc8e 100644 --- a/drivers/net/wireless/ath/ath11k/wmi.c +++ b/drivers/net/wireless/ath/ath11k/wmi.c @@ -1725,7 +1725,8 @@ int ath11k_wmi_vdev_install_key(struct ath11k *ar, tlv = (struct wmi_tlv *)(skb->data + sizeof(*cmd)); tlv->header = FIELD_PREP(WMI_TLV_TAG, WMI_TAG_ARRAY_BYTE) | FIELD_PREP(WMI_TLV_LEN, key_len_aligned); - memcpy(tlv->value, (u8 *)arg->key_data, key_len_aligned); + if (arg->key_data) + memcpy(tlv->value, (u8 *)arg->key_data, key_len_aligned); ret = ath11k_wmi_cmd_send(wmi, skb, WMI_VDEV_INSTALL_KEY_CMDID); if (ret) { From 64bc3aa02ae78b1fcb1b850e0eb1f0622002bfaa Mon Sep 17 00:00:00 2001 From: Karthikeyan Kathirvel Date: Mon, 15 Nov 2021 11:04:41 +0100 Subject: [PATCH 0280/2048] ath11k: reset RSN/WPA present state for open BSS The ath11k driver is caching the information about RSN/WPA IE in the configured beacon template. The cached information is used during associations to figure out whether 4-way PKT/2-way GTK peer flags need to be set or not. But the code never cleared the state when no such IE was found. This can for example happen when moving from an WPA/RSN to an open setup. The (seemingly connected) peer was then not able to communicate over the link because the firmware assumed a different (encryption enabled) state for the peer. Tested-on: IPQ6018 hw1.0 AHB WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1 Fixes: 01e34233c645 ("ath11k: fix wmi peer flags in peer assoc command") Cc: Venkateswara Naralasetty Reported-by: Sven Eckelmann Signed-off-by: Karthikeyan Kathirvel [sven@narfation.org: split into separate patches, clean up commit message] Signed-off-by: Sven Eckelmann Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211115100441.33771-2-sven@narfation.org --- drivers/net/wireless/ath/ath11k/mac.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index df6421fad160..292b2b7eab11 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -1214,11 +1214,15 @@ static int ath11k_mac_setup_bcn_tmpl(struct ath11k_vif *arvif) if (cfg80211_find_ie(WLAN_EID_RSN, ies, (skb_tail_pointer(bcn) - ies))) arvif->rsnie_present = true; + else + arvif->rsnie_present = false; if (cfg80211_find_vendor_ie(WLAN_OUI_MICROSOFT, WLAN_OUI_TYPE_MICROSOFT_WPA, ies, (skb_tail_pointer(bcn) - ies))) arvif->wpaie_present = true; + else + arvif->wpaie_present = false; ret = ath11k_wmi_bcn_tmpl(ar, arvif->vdev_id, &offs, bcn); From 136a3fa28a9f562e2d693e78b902bf8e5cfa1658 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:55:22 +0000 Subject: [PATCH 0281/2048] net: axienet: populate supported_interfaces member Populate the phy_interface_t bitmap for the Xilinx axienet driver with interfaces modes supported by the MAC. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c index 9b068b81ae09..8a0a43d71b51 100644 --- a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c +++ b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c @@ -2105,6 +2105,14 @@ static int axienet_probe(struct platform_device *pdev) lp->phylink_config.dev = &ndev->dev; lp->phylink_config.type = PHYLINK_NETDEV; + __set_bit(lp->phy_mode, lp->phylink_config.supported_interfaces); + if (lp->switch_x_sgmii) { + __set_bit(PHY_INTERFACE_MODE_1000BASEX, + lp->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_SGMII, + lp->phylink_config.supported_interfaces); + } + lp->phylink = phylink_create(&lp->phylink_config, pdev->dev.fwnode, lp->phy_mode, &axienet_phylink_ops); From 5703a4b664569e37d283dae067d97ea01fa80311 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:55:27 +0000 Subject: [PATCH 0282/2048] net: axienet: remove interface checks in axienet_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode in the validation function. Remove this to simplify it. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/xilinx/xilinx_axienet_main.c | 22 ------------------- 1 file changed, 22 deletions(-) diff --git a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c index 8a0a43d71b51..a058019ad9a1 100644 --- a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c +++ b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c @@ -1507,29 +1507,8 @@ static void axienet_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct net_device *ndev = to_net_dev(config->dev); - struct axienet_local *lp = netdev_priv(ndev); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - /* Only support the mode we are configured for */ - switch (state->interface) { - case PHY_INTERFACE_MODE_NA: - break; - case PHY_INTERFACE_MODE_1000BASEX: - case PHY_INTERFACE_MODE_SGMII: - if (lp->switch_x_sgmii) - break; - fallthrough; - default: - if (state->interface != lp->phy_mode) { - netdev_warn(ndev, "Cannot use PHY mode %s, supported: %s\n", - phy_modes(state->interface), - phy_modes(lp->phy_mode)); - linkmode_zero(supported); - return; - } - } - phylink_set(mask, Autoneg); phylink_set_port_modes(mask); @@ -1537,7 +1516,6 @@ static void axienet_validate(struct phylink_config *config, phylink_set(mask, Pause); switch (state->interface) { - case PHY_INTERFACE_MODE_NA: case PHY_INTERFACE_MODE_1000BASEX: case PHY_INTERFACE_MODE_SGMII: case PHY_INTERFACE_MODE_GMII: From 72a47e1aaf2e05e7b2be7feadc06396c87463ccb Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:55:32 +0000 Subject: [PATCH 0283/2048] net: axienet: use phylink_generic_validate() axienet has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/xilinx/xilinx_axienet_main.c | 41 ++----------------- 1 file changed, 3 insertions(+), 38 deletions(-) diff --git a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c index a058019ad9a1..3dabc1901416 100644 --- a/drivers/net/ethernet/xilinx/xilinx_axienet_main.c +++ b/drivers/net/ethernet/xilinx/xilinx_axienet_main.c @@ -1503,43 +1503,6 @@ static const struct ethtool_ops axienet_ethtool_ops = { .nway_reset = axienet_ethtools_nway_reset, }; -static void axienet_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set(mask, Autoneg); - phylink_set_port_modes(mask); - - phylink_set(mask, Asym_Pause); - phylink_set(mask, Pause); - - switch (state->interface) { - case PHY_INTERFACE_MODE_1000BASEX: - case PHY_INTERFACE_MODE_SGMII: - case PHY_INTERFACE_MODE_GMII: - case PHY_INTERFACE_MODE_RGMII: - case PHY_INTERFACE_MODE_RGMII_ID: - case PHY_INTERFACE_MODE_RGMII_RXID: - case PHY_INTERFACE_MODE_RGMII_TXID: - phylink_set(mask, 1000baseX_Full); - phylink_set(mask, 1000baseT_Full); - if (state->interface == PHY_INTERFACE_MODE_1000BASEX) - break; - fallthrough; - case PHY_INTERFACE_MODE_MII: - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 10baseT_Full); - fallthrough; - default: - break; - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static void axienet_mac_pcs_get_state(struct phylink_config *config, struct phylink_link_state *state) { @@ -1665,7 +1628,7 @@ static void axienet_mac_link_up(struct phylink_config *config, } static const struct phylink_mac_ops axienet_phylink_ops = { - .validate = axienet_validate, + .validate = phylink_generic_validate, .mac_pcs_get_state = axienet_mac_pcs_get_state, .mac_an_restart = axienet_mac_an_restart, .mac_prepare = axienet_mac_prepare, @@ -2082,6 +2045,8 @@ static int axienet_probe(struct platform_device *pdev) lp->phylink_config.dev = &ndev->dev; lp->phylink_config.type = PHYLINK_NETDEV; + lp->phylink_config.mac_capabilities = MAC_SYM_PAUSE | MAC_ASYM_PAUSE | + MAC_10FD | MAC_100FD | MAC_1000FD; __set_bit(lp->phy_mode, lp->phylink_config.supported_interfaces); if (lp->switch_x_sgmii) { From 4e5015df5211918242fd009014ddee10610b959d Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:58:58 +0000 Subject: [PATCH 0284/2048] net: enetc: populate supported_interfaces member Populate the phy_interface_t bitmap for the Freescale enetc driver with interfaces modes supported by the MAC. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/enetc/enetc_pf.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.c b/drivers/net/ethernet/freescale/enetc/enetc_pf.c index 0e87c7043b77..536454205590 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pf.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.c @@ -1112,6 +1112,16 @@ static int enetc_phylink_create(struct enetc_ndev_priv *priv, pf->phylink_config.dev = &priv->ndev->dev; pf->phylink_config.type = PHYLINK_NETDEV; + __set_bit(PHY_INTERFACE_MODE_INTERNAL, + pf->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_SGMII, + pf->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_2500BASEX, + pf->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_USXGMII, + pf->phylink_config.supported_interfaces); + phy_interface_set_rgmii(pf->phylink_config.supported_interfaces); + phylink = phylink_create(&pf->phylink_config, of_fwnode_handle(node), pf->if_mode, &enetc_mac_phylink_ops); if (IS_ERR(phylink)) { From 5a94c1ba8e33564c50ca9ce71add12a949fca82c Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:59:03 +0000 Subject: [PATCH 0285/2048] net: enetc: remove interface checks in enetc_pl_mac_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode in the validation function. Remove this to simplify it. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/enetc/enetc_pf.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.c b/drivers/net/ethernet/freescale/enetc/enetc_pf.c index 536454205590..61f05a021779 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pf.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.c @@ -936,16 +936,6 @@ static void enetc_pl_mac_validate(struct phylink_config *config, { __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - if (state->interface != PHY_INTERFACE_MODE_NA && - state->interface != PHY_INTERFACE_MODE_INTERNAL && - state->interface != PHY_INTERFACE_MODE_SGMII && - state->interface != PHY_INTERFACE_MODE_2500BASEX && - state->interface != PHY_INTERFACE_MODE_USXGMII && - !phy_interface_mode_is_rgmii(state->interface)) { - linkmode_zero(supported); - return; - } - phylink_set_port_modes(mask); phylink_set(mask, Autoneg); phylink_set(mask, Pause); From 75021cf02ff8ff999e8d3000b3473c7f7f25a873 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 09:59:08 +0000 Subject: [PATCH 0286/2048] net: enetc: use phylink_generic_validate() enetc has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/freescale/enetc/enetc_pf.c | 33 ++----------------- 1 file changed, 3 insertions(+), 30 deletions(-) diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.c b/drivers/net/ethernet/freescale/enetc/enetc_pf.c index 61f05a021779..fe6a544f37f0 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pf.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.c @@ -930,35 +930,6 @@ static void enetc_mdiobus_destroy(struct enetc_pf *pf) enetc_imdio_remove(pf); } -static void enetc_pl_mac_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set_port_modes(mask); - phylink_set(mask, Autoneg); - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 1000baseT_Half); - phylink_set(mask, 1000baseT_Full); - - if (state->interface == PHY_INTERFACE_MODE_INTERNAL || - state->interface == PHY_INTERFACE_MODE_2500BASEX || - state->interface == PHY_INTERFACE_MODE_USXGMII) { - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static void enetc_pl_mac_config(struct phylink_config *config, unsigned int mode, const struct phylink_link_state *state) @@ -1086,7 +1057,7 @@ static void enetc_pl_mac_link_down(struct phylink_config *config, } static const struct phylink_mac_ops enetc_mac_phylink_ops = { - .validate = enetc_pl_mac_validate, + .validate = phylink_generic_validate, .mac_config = enetc_pl_mac_config, .mac_link_up = enetc_pl_mac_link_up, .mac_link_down = enetc_pl_mac_link_down, @@ -1101,6 +1072,8 @@ static int enetc_phylink_create(struct enetc_ndev_priv *priv, pf->phylink_config.dev = &priv->ndev->dev; pf->phylink_config.type = PHYLINK_NETDEV; + pf->phylink_config.mac_capabilities = MAC_ASYM_PAUSE | MAC_SYM_PAUSE | + MAC_10 | MAC_100 | MAC_1000 | MAC_2500FD; __set_bit(PHY_INTERFACE_MODE_INTERNAL, pf->phylink_config.supported_interfaces); From ae089a8191764ba39d2867293eab5e7a9627f12c Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:02:01 +0000 Subject: [PATCH 0287/2048] net: sparx5: populate supported_interfaces member Populate the phy_interface_t bitmap for the Microchip Sparx5 driver with interfaces modes supported by the MAC. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../ethernet/microchip/sparx5/sparx5_main.c | 24 +++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c index 4625d4fb4cde..3cb6c1fe43ff 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c @@ -293,6 +293,30 @@ static int sparx5_create_port(struct sparx5 *sparx5, spx5_port->phylink_config.type = PHYLINK_NETDEV; spx5_port->phylink_config.pcs_poll = true; + __set_bit(PHY_INTERFACE_MODE_SGMII, + spx5_port->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_QSGMII, + spx5_port->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_1000BASEX, + spx5_port->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_2500BASEX, + spx5_port->phylink_config.supported_interfaces); + + if (spx5_port->conf.bandwidth == SPEED_5000 || + spx5_port->conf.bandwidth == SPEED_10000 || + spx5_port->conf.bandwidth == SPEED_25000) + __set_bit(PHY_INTERFACE_MODE_5GBASER, + spx5_port->phylink_config.supported_interfaces); + + if (spx5_port->conf.bandwidth == SPEED_10000 || + spx5_port->conf.bandwidth == SPEED_25000) + __set_bit(PHY_INTERFACE_MODE_10GBASER, + spx5_port->phylink_config.supported_interfaces); + + if (spx5_port->conf.bandwidth == SPEED_25000) + __set_bit(PHY_INTERFACE_MODE_25GBASER, + spx5_port->phylink_config.supported_interfaces); + phylink = phylink_create(&spx5_port->phylink_config, of_fwnode_handle(config->node), config->conf.phy_mode, From 9b5cc05fd91c1ea0f6eb4765464449a519d1eea4 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:02:06 +0000 Subject: [PATCH 0288/2048] net: sparx5: clean up sparx5_phylink_validate() sparx5_phylink_validate() no longer needs to check for PHY_INTERFACE_MODE_NA as phylink will walk the supported interface types to discover the link mode capabilities. Neither is it necessary to check the device capabilities as we will not be called for unsupported interface modes. Remove these checks. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../microchip/sparx5/sparx5_phylink.c | 67 +++++++------------ 1 file changed, 26 insertions(+), 41 deletions(-) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c b/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c index fb74752de0ca..e77ddded4811 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c @@ -30,7 +30,6 @@ static void sparx5_phylink_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct sparx5_port *port = netdev_priv(to_net_dev(config->dev)); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; phylink_set(mask, Autoneg); @@ -40,34 +39,23 @@ static void sparx5_phylink_validate(struct phylink_config *config, switch (state->interface) { case PHY_INTERFACE_MODE_5GBASER: + phylink_set(mask, 5000baseT_Full); + break; + case PHY_INTERFACE_MODE_10GBASER: + phylink_set(mask, 10000baseT_Full); + phylink_set(mask, 10000baseCR_Full); + phylink_set(mask, 10000baseSR_Full); + phylink_set(mask, 10000baseLR_Full); + phylink_set(mask, 10000baseLRM_Full); + phylink_set(mask, 10000baseER_Full); + break; + case PHY_INTERFACE_MODE_25GBASER: - case PHY_INTERFACE_MODE_NA: - if (port->conf.bandwidth == SPEED_5000) - phylink_set(mask, 5000baseT_Full); - if (port->conf.bandwidth == SPEED_10000) { - phylink_set(mask, 5000baseT_Full); - phylink_set(mask, 10000baseT_Full); - phylink_set(mask, 10000baseCR_Full); - phylink_set(mask, 10000baseSR_Full); - phylink_set(mask, 10000baseLR_Full); - phylink_set(mask, 10000baseLRM_Full); - phylink_set(mask, 10000baseER_Full); - } - if (port->conf.bandwidth == SPEED_25000) { - phylink_set(mask, 5000baseT_Full); - phylink_set(mask, 10000baseT_Full); - phylink_set(mask, 10000baseCR_Full); - phylink_set(mask, 10000baseSR_Full); - phylink_set(mask, 10000baseLR_Full); - phylink_set(mask, 10000baseLRM_Full); - phylink_set(mask, 10000baseER_Full); - phylink_set(mask, 25000baseCR_Full); - phylink_set(mask, 25000baseSR_Full); - } - if (state->interface != PHY_INTERFACE_MODE_NA) - break; - fallthrough; + phylink_set(mask, 25000baseCR_Full); + phylink_set(mask, 25000baseSR_Full); + break; + case PHY_INTERFACE_MODE_SGMII: case PHY_INTERFACE_MODE_QSGMII: phylink_set(mask, 10baseT_Half); @@ -76,21 +64,18 @@ static void sparx5_phylink_validate(struct phylink_config *config, phylink_set(mask, 100baseT_Full); phylink_set(mask, 1000baseT_Full); phylink_set(mask, 1000baseX_Full); - if (state->interface != PHY_INTERFACE_MODE_NA) - break; - fallthrough; - case PHY_INTERFACE_MODE_1000BASEX: - case PHY_INTERFACE_MODE_2500BASEX: - if (state->interface != PHY_INTERFACE_MODE_2500BASEX) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - } - if (state->interface == PHY_INTERFACE_MODE_2500BASEX || - state->interface == PHY_INTERFACE_MODE_NA) { - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - } break; + + case PHY_INTERFACE_MODE_1000BASEX: + phylink_set(mask, 1000baseT_Full); + phylink_set(mask, 1000baseX_Full); + break; + + case PHY_INTERFACE_MODE_2500BASEX: + phylink_set(mask, 2500baseT_Full); + phylink_set(mask, 2500baseX_Full); + break; + default: linkmode_zero(supported); return; From 319faa90b724ce68b3dccd24161c65feb4a77988 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:02:11 +0000 Subject: [PATCH 0289/2048] net: sparx5: use phylink_generic_validate() Sparx5 has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../ethernet/microchip/sparx5/sparx5_main.c | 3 + .../microchip/sparx5/sparx5_phylink.c | 60 +------------------ 2 files changed, 4 insertions(+), 59 deletions(-) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c index 3cb6c1fe43ff..16266275dd36 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c @@ -292,6 +292,9 @@ static int sparx5_create_port(struct sparx5 *sparx5, spx5_port->phylink_config.dev = &spx5_port->ndev->dev; spx5_port->phylink_config.type = PHYLINK_NETDEV; spx5_port->phylink_config.pcs_poll = true; + spx5_port->phylink_config.mac_capabilities = MAC_ASYM_PAUSE | + MAC_SYM_PAUSE | MAC_10 | MAC_100 | MAC_1000FD | + MAC_2500FD | MAC_5000FD | MAC_10000FD | MAC_25000FD; __set_bit(PHY_INTERFACE_MODE_SGMII, spx5_port->phylink_config.supported_interfaces); diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c b/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c index e77ddded4811..8ba33bc1a001 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_phylink.c @@ -26,64 +26,6 @@ static bool port_conf_has_changed(struct sparx5_port_config *a, struct sparx5_po return false; } -static void sparx5_phylink_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set(mask, Autoneg); - phylink_set_port_modes(mask); - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - - switch (state->interface) { - case PHY_INTERFACE_MODE_5GBASER: - phylink_set(mask, 5000baseT_Full); - break; - - case PHY_INTERFACE_MODE_10GBASER: - phylink_set(mask, 10000baseT_Full); - phylink_set(mask, 10000baseCR_Full); - phylink_set(mask, 10000baseSR_Full); - phylink_set(mask, 10000baseLR_Full); - phylink_set(mask, 10000baseLRM_Full); - phylink_set(mask, 10000baseER_Full); - break; - - case PHY_INTERFACE_MODE_25GBASER: - phylink_set(mask, 25000baseCR_Full); - phylink_set(mask, 25000baseSR_Full); - break; - - case PHY_INTERFACE_MODE_SGMII: - case PHY_INTERFACE_MODE_QSGMII: - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - break; - - case PHY_INTERFACE_MODE_1000BASEX: - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - break; - - case PHY_INTERFACE_MODE_2500BASEX: - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - break; - - default: - linkmode_zero(supported); - return; - } - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static void sparx5_phylink_mac_config(struct phylink_config *config, unsigned int mode, const struct phylink_link_state *state) @@ -187,7 +129,7 @@ const struct phylink_pcs_ops sparx5_phylink_pcs_ops = { }; const struct phylink_mac_ops sparx5_phylink_mac_ops = { - .validate = sparx5_phylink_validate, + .validate = phylink_generic_validate, .mac_config = sparx5_phylink_mac_config, .mac_link_down = sparx5_phylink_mac_link_down, .mac_link_up = sparx5_phylink_mac_link_up, From 83800d29f0c578e82554e7d4c6bfdbdf9b6cf428 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:06:43 +0000 Subject: [PATCH 0290/2048] net: mtk_eth_soc: populate supported_interfaces member Populate the phy interface mode bitmap for the Mediatek driver with interfaces modes supported by the MAC. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 75d67d1b5f6b..7f62298bc983 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -3009,6 +3009,26 @@ static int mtk_add_mac(struct mtk_eth *eth, struct device_node *np) mac->phylink_config.dev = ð->netdev[id]->dev; mac->phylink_config.type = PHYLINK_NETDEV; + __set_bit(PHY_INTERFACE_MODE_MII, + mac->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_GMII, + mac->phylink_config.supported_interfaces); + + if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_RGMII)) + phy_interface_set_rgmii(mac->phylink_config.supported_interfaces); + + if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_TRGMII) && !mac->id) + __set_bit(PHY_INTERFACE_MODE_TRGMII, + mac->phylink_config.supported_interfaces); + + if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_SGMII)) { + __set_bit(PHY_INTERFACE_MODE_SGMII, + mac->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_1000BASEX, + mac->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_2500BASEX, + mac->phylink_config.supported_interfaces); + } phylink = phylink_create(&mac->phylink_config, of_fwnode_handle(mac->of_node), From db81ca153814475d7e07365d46a4d1134bd122e2 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:06:48 +0000 Subject: [PATCH 0291/2048] net: mtk_eth_soc: remove interface checks in mtk_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode, nor handle PHY_INTERFACE_MODE_NA in the validation function. Remove these to simplify the implementation. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 34 --------------------- 1 file changed, 34 deletions(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 7f62298bc983..31872594c790 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -467,24 +467,8 @@ static void mtk_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct mtk_mac *mac = container_of(config, struct mtk_mac, - phylink_config); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - if (state->interface != PHY_INTERFACE_MODE_NA && - state->interface != PHY_INTERFACE_MODE_MII && - state->interface != PHY_INTERFACE_MODE_GMII && - !(MTK_HAS_CAPS(mac->hw->soc->caps, MTK_RGMII) && - phy_interface_mode_is_rgmii(state->interface)) && - !(MTK_HAS_CAPS(mac->hw->soc->caps, MTK_TRGMII) && - !mac->id && state->interface == PHY_INTERFACE_MODE_TRGMII) && - !(MTK_HAS_CAPS(mac->hw->soc->caps, MTK_SGMII) && - (state->interface == PHY_INTERFACE_MODE_SGMII || - phy_interface_mode_is_8023z(state->interface)))) { - linkmode_zero(supported); - return; - } - phylink_set_port_modes(mask); phylink_set(mask, Autoneg); @@ -511,7 +495,6 @@ static void mtk_validate(struct phylink_config *config, case PHY_INTERFACE_MODE_MII: case PHY_INTERFACE_MODE_RMII: case PHY_INTERFACE_MODE_REVMII: - case PHY_INTERFACE_MODE_NA: default: phylink_set(mask, 10baseT_Half); phylink_set(mask, 10baseT_Full); @@ -520,23 +503,6 @@ static void mtk_validate(struct phylink_config *config, break; } - if (state->interface == PHY_INTERFACE_MODE_NA) { - if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_SGMII)) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - phylink_set(mask, 2500baseX_Full); - } - if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_RGMII)) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseT_Half); - phylink_set(mask, 1000baseX_Full); - } - if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_GEPHY)) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseT_Half); - } - } - phylink_set(mask, Pause); phylink_set(mask, Asym_Pause); From 71d927494463c4f016d828e1134da26b7e961af5 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:06:53 +0000 Subject: [PATCH 0292/2048] net: mtk_eth_soc: drop use of phylink_helper_basex_speed() Now that we have a better method to select SFP interface modes, we no longer need to use phylink_helper_basex_speed() in a driver's validation function, and we can also get rid of our hack to indicate both 1000base-X and 2500base-X if the comphy is present to make that work. Remove this hack and use of phylink_helper_basex_speed(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 31872594c790..98f9a6ed9584 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -477,8 +477,9 @@ static void mtk_validate(struct phylink_config *config, phylink_set(mask, 1000baseT_Full); break; case PHY_INTERFACE_MODE_1000BASEX: - case PHY_INTERFACE_MODE_2500BASEX: phylink_set(mask, 1000baseX_Full); + break; + case PHY_INTERFACE_MODE_2500BASEX: phylink_set(mask, 2500baseX_Full); break; case PHY_INTERFACE_MODE_GMII: @@ -508,11 +509,6 @@ static void mtk_validate(struct phylink_config *config, linkmode_and(supported, supported, mask); linkmode_and(state->advertising, state->advertising, mask); - - /* We can only operate at 2500BaseX or 1000BaseX. If requested - * to advertise both, only report advertising at 2500BaseX. - */ - phylink_helper_basex_speed(state); } static const struct phylink_mac_ops mtk_phylink_ops = { From a4238f6ce151afa331375d74a5033b76da637644 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:06:58 +0000 Subject: [PATCH 0293/2048] net: mtk_eth_soc: use phylink_generic_validate() mtk_eth_soc has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 53 ++------------------- 1 file changed, 4 insertions(+), 49 deletions(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 98f9a6ed9584..de4152e2e3e4 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -463,56 +463,8 @@ static void mtk_mac_link_up(struct phylink_config *config, mtk_w32(mac->hw, mcr, MTK_MAC_MCR(mac->id)); } -static void mtk_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set_port_modes(mask); - phylink_set(mask, Autoneg); - - switch (state->interface) { - case PHY_INTERFACE_MODE_TRGMII: - phylink_set(mask, 1000baseT_Full); - break; - case PHY_INTERFACE_MODE_1000BASEX: - phylink_set(mask, 1000baseX_Full); - break; - case PHY_INTERFACE_MODE_2500BASEX: - phylink_set(mask, 2500baseX_Full); - break; - case PHY_INTERFACE_MODE_GMII: - case PHY_INTERFACE_MODE_RGMII: - case PHY_INTERFACE_MODE_RGMII_ID: - case PHY_INTERFACE_MODE_RGMII_RXID: - case PHY_INTERFACE_MODE_RGMII_TXID: - phylink_set(mask, 1000baseT_Half); - fallthrough; - case PHY_INTERFACE_MODE_SGMII: - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - fallthrough; - case PHY_INTERFACE_MODE_MII: - case PHY_INTERFACE_MODE_RMII: - case PHY_INTERFACE_MODE_REVMII: - default: - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - break; - } - - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static const struct phylink_mac_ops mtk_phylink_ops = { - .validate = mtk_validate, + .validate = phylink_generic_validate, .mac_pcs_get_state = mtk_mac_pcs_get_state, .mac_an_restart = mtk_mac_an_restart, .mac_config = mtk_mac_config, @@ -2971,6 +2923,9 @@ static int mtk_add_mac(struct mtk_eth *eth, struct device_node *np) mac->phylink_config.dev = ð->netdev[id]->dev; mac->phylink_config.type = PHYLINK_NETDEV; + mac->phylink_config.mac_capabilities = MAC_ASYM_PAUSE | MAC_SYM_PAUSE | + MAC_10 | MAC_100 | MAC_1000 | MAC_2500FD; + __set_bit(PHY_INTERFACE_MODE_MII, mac->phylink_config.supported_interfaces); __set_bit(PHY_INTERFACE_MODE_GMII, From 8ea8c5b492d4e593c64d71c986c320faade69e17 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:09:31 +0000 Subject: [PATCH 0294/2048] net: ocelot_net: populate supported_interfaces member Populate the phy interface mode bitmap for the MSCC Ocelot driver with the interface modes supported by the MAC. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot_net.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c index eaeba60b1bba..37c158df60ce 100644 --- a/drivers/net/ethernet/mscc/ocelot_net.c +++ b/drivers/net/ethernet/mscc/ocelot_net.c @@ -1655,6 +1655,9 @@ static int ocelot_port_phylink_create(struct ocelot *ocelot, int port, priv->phylink_config.dev = &priv->dev->dev; priv->phylink_config.type = PHYLINK_NETDEV; + __set_bit(ocelot_port->phy_mode, + priv->phylink_config.supported_interfaces); + phylink = phylink_create(&priv->phylink_config, of_fwnode_handle(portnp), phy_mode, &ocelot_phylink_ops); From a6f5248bc0a36ea1cbd0e04a8a744dc449022503 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:09:36 +0000 Subject: [PATCH 0295/2048] net: ocelot_net: remove interface checks in macb_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode in the validation function. Remove this to simplify it. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot_net.c | 9 --------- 1 file changed, 9 deletions(-) diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c index 37c158df60ce..21df548dcf64 100644 --- a/drivers/net/ethernet/mscc/ocelot_net.c +++ b/drivers/net/ethernet/mscc/ocelot_net.c @@ -1502,17 +1502,8 @@ static void vsc7514_phylink_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct net_device *ndev = to_net_dev(config->dev); - struct ocelot_port_private *priv = netdev_priv(ndev); - struct ocelot_port *ocelot_port = &priv->port; __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = {}; - if (state->interface != PHY_INTERFACE_MODE_NA && - state->interface != ocelot_port->phy_mode) { - linkmode_zero(supported); - return; - } - phylink_set_port_modes(mask); phylink_set(mask, Pause); From 7258aa5094dbc89919c13f9743571657cc99ed46 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 16 Nov 2021 10:09:41 +0000 Subject: [PATCH 0296/2048] net: ocelot_net: use phylink_generic_validate() ocelot_net has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot_net.c | 29 +++----------------------- 1 file changed, 3 insertions(+), 26 deletions(-) diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c index 21df548dcf64..0fcf359a6975 100644 --- a/drivers/net/ethernet/mscc/ocelot_net.c +++ b/drivers/net/ethernet/mscc/ocelot_net.c @@ -1498,31 +1498,6 @@ struct notifier_block ocelot_switchdev_blocking_nb __read_mostly = { .notifier_call = ocelot_switchdev_blocking_event, }; -static void vsc7514_phylink_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = {}; - - phylink_set_port_modes(mask); - - phylink_set(mask, Pause); - phylink_set(mask, Autoneg); - phylink_set(mask, Asym_Pause); - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 1000baseT_Half); - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static void vsc7514_phylink_mac_config(struct phylink_config *config, unsigned int link_an_mode, const struct phylink_link_state *state) @@ -1581,7 +1556,7 @@ static void vsc7514_phylink_mac_link_up(struct phylink_config *config, } static const struct phylink_mac_ops ocelot_phylink_ops = { - .validate = vsc7514_phylink_validate, + .validate = phylink_generic_validate, .mac_config = vsc7514_phylink_mac_config, .mac_link_down = vsc7514_phylink_mac_link_down, .mac_link_up = vsc7514_phylink_mac_link_up, @@ -1645,6 +1620,8 @@ static int ocelot_port_phylink_create(struct ocelot *ocelot, int port, priv->phylink_config.dev = &priv->dev->dev; priv->phylink_config.type = PHYLINK_NETDEV; + priv->phylink_config.mac_capabilities = MAC_ASYM_PAUSE | MAC_SYM_PAUSE | + MAC_10 | MAC_100 | MAC_1000FD | MAC_2500FD; __set_bit(ocelot_port->phy_mode, priv->phylink_config.supported_interfaces); From 738baea4970b36580cb1dd4f9b3fd5247aa1c7f5 Mon Sep 17 00:00:00 2001 From: Vasudev Kamath Date: Tue, 16 Nov 2021 12:51:48 +0530 Subject: [PATCH 0297/2048] Documentation: networking: net_failover: Fix documentation Update net_failover documentation with missing and incomplete details to get a proper working setup. Signed-off-by: Vasudev Kamath Reviewed-by: Krishna Kumar Signed-off-by: David S. Miller --- Documentation/networking/net_failover.rst | 111 +++++++++++++++++----- 1 file changed, 88 insertions(+), 23 deletions(-) diff --git a/Documentation/networking/net_failover.rst b/Documentation/networking/net_failover.rst index e143ab79a960..3a662f2b4d6e 100644 --- a/Documentation/networking/net_failover.rst +++ b/Documentation/networking/net_failover.rst @@ -35,7 +35,7 @@ To support this, the hypervisor needs to enable VIRTIO_NET_F_STANDBY feature on the virtio-net interface and assign the same MAC address to both virtio-net and VF interfaces. -Here is an example XML snippet that shows such configuration. +Here is an example libvirt XML snippet that shows such configuration: :: @@ -45,18 +45,32 @@ Here is an example XML snippet that shows such configuration. -
+ +
-
+ +In this configuration, the first device definition is for the virtio-net +interface and this acts as the 'persistent' device indicating that this +interface will always be plugged in. This is specified by the 'teaming' tag with +required attribute type having value 'persistent'. The link state for the +virtio-net device is set to 'down' to ensure that the 'failover' netdev prefers +the VF passthrough device for normal communication. The virtio-net device will +be brought UP during live migration to allow uninterrupted communication. + +The second device definition is for the VF passthrough interface. Here the +'teaming' tag is provided with type 'transient' indicating that this device may +periodically be unplugged. A second attribute - 'persistent' is provided and +points to the alias name declared for the virtio-net device. + Booting a VM with the above configuration will result in the following 3 -netdevs created in the VM. +interfaces created in the VM: :: 4: ens10: mtu 1500 qdisc noqueue state UP group default qlen 1000 @@ -65,13 +79,36 @@ netdevs created in the VM. valid_lft 42482sec preferred_lft 42482sec inet6 fe80::97d8:db2:8c10:b6d6/64 scope link valid_lft forever preferred_lft forever - 5: ens10nsby: mtu 1500 qdisc fq_codel master ens10 state UP group default qlen 1000 + 5: ens10nsby: mtu 1500 qdisc fq_codel master ens10 state DOWN group default qlen 1000 link/ether 52:54:00:00:12:53 brd ff:ff:ff:ff:ff:ff 7: ens11: mtu 1500 qdisc mq master ens10 state UP group default qlen 1000 link/ether 52:54:00:00:12:53 brd ff:ff:ff:ff:ff:ff -ens10 is the 'failover' master netdev, ens10nsby and ens11 are the slave -'standby' and 'primary' netdevs respectively. +Here, ens10 is the 'failover' master interface, ens10nsby is the slave 'standby' +virtio-net interface, and ens11 is the slave 'primary' VF passthrough interface. + +One point to note here is that some user space network configuration daemons +like systemd-networkd, ifupdown, etc, do not understand the 'net_failover' +device; and on the first boot, the VM might end up with both 'failover' device +and VF accquiring IP addresses (either same or different) from the DHCP server. +This will result in lack of connectivity to the VM. So some tweaks might be +needed to these network configuration daemons to make sure that an IP is +received only on the 'failover' device. + +Below is the patch snippet used with 'cloud-ifupdown-helper' script found on +Debian cloud images: + +:: + @@ -27,6 +27,8 @@ do_setup() { + local working="$cfgdir/.$INTERFACE" + local final="$cfgdir/$INTERFACE" + + + if [ -d "/sys/class/net/${INTERFACE}/master" ]; then exit 0; fi + + + if ifup --no-act "$INTERFACE" > /dev/null 2>&1; then + # interface is already known to ifupdown, no need to generate cfg + log "Skipping configuration generation for $INTERFACE" + Live Migration of a VM with SR-IOV VF & virtio-net in STANDBY mode ================================================================== @@ -80,40 +117,68 @@ net_failover also enables hypervisor controlled live migration to be supported with VMs that have direct attached SR-IOV VF devices by automatic failover to the paravirtual datapath when the VF is unplugged. -Here is a sample script that shows the steps to initiate live migration on -the source hypervisor. +Here is a sample script that shows the steps to initiate live migration from +the source hypervisor. Note: It is assumed that the VM is connected to a +software bridge 'br0' which has a single VF attached to it along with the vnet +device to the VM. This is not the VF that was passthrough'd to the VM (seen in +the vf.xml file). :: - # cat vf_xml + # cat vf.xml
-
+ - # Source Hypervisor + # Source Hypervisor migrate.sh #!/bin/bash - DOMAIN=fedora27-tap01 - PF=enp66s0f0 - VF_NUM=5 - TAP_IF=tap01 - VF_XML= + DOMAIN=vm-01 + PF=ens6np0 + VF=ens6v1 # VF attached to the bridge. + VF_NUM=1 + TAP_IF=vmtap01 # virtio-net interface in the VM. + VF_XML=vf.xml MAC=52:54:00:00:12:53 ZERO_MAC=00:00:00:00:00:00 + # Set the virtio-net interface up. virsh domif-setlink $DOMAIN $TAP_IF up - bridge fdb del $MAC dev $PF master - virsh detach-device $DOMAIN $VF_XML + + # Remove the VF that was passthrough'd to the VM. + virsh detach-device --live --config $DOMAIN $VF_XML + ip link set $PF vf $VF_NUM mac $ZERO_MAC - virsh migrate --live $DOMAIN qemu+ssh://$REMOTE_HOST/system + # Add FDB entry for traffic to continue going to the VM via + # the VF -> br0 -> vnet interface path. + bridge fdb add $MAC dev $VF + bridge fdb add $MAC dev $TAP_IF master - # Destination Hypervisor + # Migrate the VM + virsh migrate --live --persistent $DOMAIN qemu+ssh://$REMOTE_HOST/system + + # Clean up FDB entries after migration completes. + bridge fdb del $MAC dev $VF + bridge fdb del $MAC dev $TAP_IF master + +On the destination hypervisor, a shared bridge 'br0' is created before migration +starts, and a VF from the destination PF is added to the bridge. Similarly an +appropriate FDB entry is added. + +The following script is executed on the destination hypervisor once migration +completes, and it reattaches the VF to the VM and brings down the virtio-net +interface. + +:: + # reattach-vf.sh #!/bin/bash - virsh attach-device $DOMAIN $VF_XML - virsh domif-setlink $DOMAIN $TAP_IF down + bridge fdb del 52:54:00:00:12:53 dev ens36v0 + bridge fdb del 52:54:00:00:12:53 dev vmtap01 master + virsh attach-device --config --live vm01 vf.xml + virsh domif-setlink vm01 vmtap01 down From 65483559dc0aa527652de43d0634736c7fc7c72f Mon Sep 17 00:00:00 2001 From: Kurt Kanzenbach Date: Tue, 16 Nov 2021 09:03:25 +0100 Subject: [PATCH 0298/2048] net: ethernet: ti: cpsw: Enable PHY timestamping If the used PHYs also support hardware timestamping, all configuration requests should be forwared to the PHYs instead of being processed by the MAC driver itself. This enables PHY timestamping in combination with the cpsw driver. Tested with an am335x based board with two DP83640 PHYs connected to the cpsw switch. Signed-off-by: Kurt Kanzenbach Acked-by: Richard Cochran Signed-off-by: David S. Miller --- drivers/net/ethernet/ti/cpsw_priv.c | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/ti/cpsw_priv.c b/drivers/net/ethernet/ti/cpsw_priv.c index ecc2a6b7e28f..c99dd9735087 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.c +++ b/drivers/net/ethernet/ti/cpsw_priv.c @@ -710,20 +710,26 @@ int cpsw_ndo_ioctl(struct net_device *dev, struct ifreq *req, int cmd) struct cpsw_priv *priv = netdev_priv(dev); struct cpsw_common *cpsw = priv->cpsw; int slave_no = cpsw_slave_index(cpsw, priv); + struct phy_device *phy; if (!netif_running(dev)) return -EINVAL; - switch (cmd) { - case SIOCSHWTSTAMP: - return cpsw_hwtstamp_set(dev, req); - case SIOCGHWTSTAMP: - return cpsw_hwtstamp_get(dev, req); + phy = cpsw->slaves[slave_no].phy; + + if (!phy_has_hwtstamp(phy)) { + switch (cmd) { + case SIOCSHWTSTAMP: + return cpsw_hwtstamp_set(dev, req); + case SIOCGHWTSTAMP: + return cpsw_hwtstamp_get(dev, req); + } } - if (!cpsw->slaves[slave_no].phy) - return -EOPNOTSUPP; - return phy_mii_ioctl(cpsw->slaves[slave_no].phy, req, cmd); + if (phy) + return phy_mii_ioctl(phy, req, cmd); + + return -EOPNOTSUPP; } int cpsw_ndo_set_tx_maxrate(struct net_device *ndev, int queue, u32 rate) From 8160fb43d55d26d64607fd32fe69185a5f5fe41f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 16 Nov 2021 19:29:21 -0800 Subject: [PATCH 0299/2048] net: use an atomic_long_t for queue->trans_timeout tx_timeout_show() assumed dev_watchdog() would stop all the queues, to fetch queue->trans_timeout under protection of the queue->_xmit_lock. As we want to no longer disrupt transmits, we use an atomic_long_t instead. Signed-off-by: Eric Dumazet Cc: david decotigny Signed-off-by: David S. Miller --- include/linux/netdevice.h | 2 +- net/core/net-sysfs.c | 6 +----- net/sched/sch_generic.c | 2 +- 3 files changed, 3 insertions(+), 7 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index dd328364dfe9..1d22483cf78c 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -592,7 +592,7 @@ struct netdev_queue { * Number of TX timeouts for this queue * (/sys/class/net/DEV/Q/trans_timeout) */ - unsigned long trans_timeout; + atomic_long_t trans_timeout; /* Subordinate device that the queue has been assigned to */ struct net_device *sb_dev; diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 9c01c642cf9e..addbef5419fb 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -1201,11 +1201,7 @@ static const struct sysfs_ops netdev_queue_sysfs_ops = { static ssize_t tx_timeout_show(struct netdev_queue *queue, char *buf) { - unsigned long trans_timeout; - - spin_lock_irq(&queue->_xmit_lock); - trans_timeout = queue->trans_timeout; - spin_unlock_irq(&queue->_xmit_lock); + unsigned long trans_timeout = atomic_long_read(&queue->trans_timeout); return sprintf(buf, fmt_ulong, trans_timeout); } diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 3b0f62095803..1b4328bd495d 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -467,7 +467,7 @@ static void dev_watchdog(struct timer_list *t) time_after(jiffies, (trans_start + dev->watchdog_timeo))) { some_queue_timedout = 1; - txq->trans_timeout++; + atomic_long_inc(&txq->trans_timeout); break; } } From 5337824f4dc4bb26f38fbbba4ffb425a92803f15 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 16 Nov 2021 19:29:22 -0800 Subject: [PATCH 0300/2048] net: annotate accesses to queue->trans_start In following patches, dev_watchdog() will no longer stop all queues. It will read queue->trans_start locklessly. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 2 +- drivers/net/ethernet/atheros/ag71xx.c | 2 +- drivers/net/ethernet/freescale/dpaa/dpaa_eth.c | 4 ++-- drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 2 +- drivers/net/ethernet/ibm/ibmvnic.c | 2 +- drivers/net/ethernet/intel/igb/igb_main.c | 4 ++-- .../ethernet/mellanox/mlx5/core/en/reporter_tx.c | 2 +- .../net/ethernet/stmicro/stmmac/stmmac_main.c | 6 +++--- drivers/net/ethernet/ti/am65-cpsw-nuss.c | 2 +- drivers/net/virtio_net.c | 2 +- drivers/net/wireless/marvell/mwifiex/init.c | 2 +- drivers/staging/rtl8192e/rtllib_softmac.c | 2 +- include/linux/netdevice.h | 16 +++++++++++++--- net/sched/sch_generic.c | 8 ++++---- 14 files changed, 33 insertions(+), 23 deletions(-) diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c index 220dc42af31a..ff2d099aab21 100644 --- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c +++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c @@ -869,7 +869,7 @@ static void xgene_enet_timeout(struct net_device *ndev, unsigned int txqueue) for (i = 0; i < pdata->txq_cnt; i++) { txq = netdev_get_tx_queue(ndev, i); - txq->trans_start = jiffies; + txq_trans_cond_update(txq); netif_tx_start_queue(txq); } } diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index 88d2ab748399..e4f30bb7498f 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -766,7 +766,7 @@ static bool ag71xx_check_dma_stuck(struct ag71xx *ag) unsigned long timestamp; u32 rx_sm, tx_sm, rx_fd; - timestamp = netdev_get_tx_queue(ag->ndev, 0)->trans_start; + timestamp = READ_ONCE(netdev_get_tx_queue(ag->ndev, 0)->trans_start); if (likely(time_before(jiffies, timestamp + HZ / 10))) return false; diff --git a/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c b/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c index 6b2927d863e2..d6871437d951 100644 --- a/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c +++ b/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c @@ -2325,7 +2325,7 @@ dpaa_start_xmit(struct sk_buff *skb, struct net_device *net_dev) txq = netdev_get_tx_queue(net_dev, queue_mapping); /* LLTX requires to do our own update of trans_start */ - txq->trans_start = jiffies; + txq_trans_cond_update(txq); if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) { fd.cmd |= cpu_to_be32(FM_FD_CMD_UPD); @@ -2531,7 +2531,7 @@ static int dpaa_xdp_xmit_frame(struct net_device *net_dev, /* Bump the trans_start */ txq = netdev_get_tx_queue(net_dev, smp_processor_id()); - txq->trans_start = jiffies; + txq_trans_cond_update(txq); err = dpaa_xmit(priv, percpu_stats, smp_processor_id(), &fd); if (err) { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index 13835a37b3a2..d5100179f8d5 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -2679,7 +2679,7 @@ static bool hns3_get_tx_timeo_queue_info(struct net_device *ndev) unsigned long trans_start; q = netdev_get_tx_queue(ndev, i); - trans_start = q->trans_start; + trans_start = READ_ONCE(q->trans_start); if (netif_xmit_stopped(q) && time_after(jiffies, (trans_start + ndev->watchdog_timeo))) { diff --git a/drivers/net/ethernet/ibm/ibmvnic.c b/drivers/net/ethernet/ibm/ibmvnic.c index 3cca51735421..c327fc8860da 100644 --- a/drivers/net/ethernet/ibm/ibmvnic.c +++ b/drivers/net/ethernet/ibm/ibmvnic.c @@ -2058,7 +2058,7 @@ static netdev_tx_t ibmvnic_xmit(struct sk_buff *skb, struct net_device *netdev) tx_packets++; tx_bytes += skb->len; - txq->trans_start = jiffies; + txq_trans_cond_update(txq); ret = NETDEV_TX_OK; goto out; diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 836be0d3b291..18a019a47182 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -2927,7 +2927,7 @@ static int igb_xdp_xmit_back(struct igb_adapter *adapter, struct xdp_buff *xdp) nq = txring_txq(tx_ring); __netif_tx_lock(nq, cpu); /* Avoid transmit queue timeout since we share it with the slow path */ - nq->trans_start = jiffies; + txq_trans_cond_update(nq); ret = igb_xmit_xdp_ring(adapter, tx_ring, xdpf); __netif_tx_unlock(nq); @@ -2961,7 +2961,7 @@ static int igb_xdp_xmit(struct net_device *dev, int n, __netif_tx_lock(nq, cpu); /* Avoid transmit queue timeout since we share it with the slow path */ - nq->trans_start = jiffies; + txq_trans_cond_update(nq); for (i = 0; i < n; i++) { struct xdp_frame *xdpf = frames[i]; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c index 4f4bc8726ec4..860605133287 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c @@ -565,7 +565,7 @@ int mlx5e_reporter_tx_timeout(struct mlx5e_txqsq *sq) snprintf(err_str, sizeof(err_str), "TX timeout on queue: %d, SQ: 0x%x, CQ: 0x%x, SQ Cons: 0x%x SQ Prod: 0x%x, usecs since last trans: %u", sq->ch_ix, sq->sqn, sq->cq.mcq.cqn, sq->cc, sq->pc, - jiffies_to_usecs(jiffies - sq->txq->trans_start)); + jiffies_to_usecs(jiffies - READ_ONCE(sq->txq->trans_start))); mlx5e_health_report(priv, priv->tx_reporter, err_str, &err_ctx); return to_ctx.status; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 033c35c09a54..389d125310c1 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -2356,7 +2356,7 @@ static bool stmmac_xdp_xmit_zc(struct stmmac_priv *priv, u32 queue, u32 budget) bool work_done = true; /* Avoids TX time-out as we are sharing with slow path */ - nq->trans_start = jiffies; + txq_trans_cond_update(nq->trans_start); budget = min(budget, stmmac_tx_avail(priv, queue)); @@ -4657,7 +4657,7 @@ static int stmmac_xdp_xmit_back(struct stmmac_priv *priv, __netif_tx_lock(nq, cpu); /* Avoids TX time-out as we are sharing with slow path */ - nq->trans_start = jiffies; + txq_trans_cond_update(nq->trans_start); res = stmmac_xdp_xmit_xdpf(priv, queue, xdpf, false); if (res == STMMAC_XDP_TX) @@ -6293,7 +6293,7 @@ static int stmmac_xdp_xmit(struct net_device *dev, int num_frames, __netif_tx_lock(nq, cpu); /* Avoids TX time-out as we are sharing with slow path */ - nq->trans_start = jiffies; + txq_trans_cond_update(nq); for (i = 0; i < num_frames; i++) { int res; diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c index c092cb61416a..750cea23e9cd 100644 --- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c +++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c @@ -345,7 +345,7 @@ static void am65_cpsw_nuss_ndo_host_tx_timeout(struct net_device *ndev, netif_txq = netdev_get_tx_queue(ndev, txqueue); tx_chn = &common->tx_chns[txqueue]; - trans_start = netif_txq->trans_start; + trans_start = READ_ONCE(netif_txq->trans_start); netdev_err(ndev, "txq:%d DRV_XOFF:%d tmo:%u dql_avail:%d free_desc:%zu\n", txqueue, diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 1771d6e5224f..03e38e38ee4b 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -2694,7 +2694,7 @@ static void virtnet_tx_timeout(struct net_device *dev, unsigned int txqueue) netdev_err(dev, "TX timeout on queue: %u, sq: %s, vq: 0x%x, name: %s, %u usecs ago\n", txqueue, sq->name, sq->vq->index, sq->vq->name, - jiffies_to_usecs(jiffies - txq->trans_start)); + jiffies_to_usecs(jiffies - READ_ONCE(txq->trans_start))); } static const struct net_device_ops virtnet_netdev = { diff --git a/drivers/net/wireless/marvell/mwifiex/init.c b/drivers/net/wireless/marvell/mwifiex/init.c index f006a3d72b40..88c72d1827a0 100644 --- a/drivers/net/wireless/marvell/mwifiex/init.c +++ b/drivers/net/wireless/marvell/mwifiex/init.c @@ -332,7 +332,7 @@ void mwifiex_set_trans_start(struct net_device *dev) int i; for (i = 0; i < dev->num_tx_queues; i++) - netdev_get_tx_queue(dev, i)->trans_start = jiffies; + txq_trans_cond_update(netdev_get_tx_queue(dev, i)); netif_trans_update(dev); } diff --git a/drivers/staging/rtl8192e/rtllib_softmac.c b/drivers/staging/rtl8192e/rtllib_softmac.c index d2726d01c757..aabbea48223d 100644 --- a/drivers/staging/rtl8192e/rtllib_softmac.c +++ b/drivers/staging/rtl8192e/rtllib_softmac.c @@ -2515,7 +2515,7 @@ void rtllib_stop_all_queues(struct rtllib_device *ieee) unsigned int i; for (i = 0; i < ieee->dev->num_tx_queues; i++) - netdev_get_tx_queue(ieee->dev, i)->trans_start = jiffies; + txq_trans_cond_update(netdev_get_tx_queue(ieee->dev, i)); netif_tx_stop_all_queues(ieee->dev); } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 1d22483cf78c..279409ef2b18 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -4095,10 +4095,21 @@ static inline void __netif_tx_unlock_bh(struct netdev_queue *txq) spin_unlock_bh(&txq->_xmit_lock); } +/* + * txq->trans_start can be read locklessly from dev_watchdog() + */ static inline void txq_trans_update(struct netdev_queue *txq) { if (txq->xmit_lock_owner != -1) - txq->trans_start = jiffies; + WRITE_ONCE(txq->trans_start, jiffies); +} + +static inline void txq_trans_cond_update(struct netdev_queue *txq) +{ + unsigned long now = jiffies; + + if (READ_ONCE(txq->trans_start) != now) + WRITE_ONCE(txq->trans_start, now); } /* legacy drivers only, netdev_start_xmit() sets txq->trans_start */ @@ -4106,8 +4117,7 @@ static inline void netif_trans_update(struct net_device *dev) { struct netdev_queue *txq = netdev_get_tx_queue(dev, 0); - if (txq->trans_start != jiffies) - txq->trans_start = jiffies; + txq_trans_cond_update(txq); } /** diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 1b4328bd495d..02c46041f76e 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -434,9 +434,9 @@ unsigned long dev_trans_start(struct net_device *dev) dev = vlan_dev_real_dev(dev); else if (netif_is_macvlan(dev)) dev = macvlan_dev_real_dev(dev); - res = netdev_get_tx_queue(dev, 0)->trans_start; + res = READ_ONCE(netdev_get_tx_queue(dev, 0)->trans_start); for (i = 1; i < dev->num_tx_queues; i++) { - val = netdev_get_tx_queue(dev, i)->trans_start; + val = READ_ONCE(netdev_get_tx_queue(dev, i)->trans_start); if (val && time_after(val, res)) res = val; } @@ -462,7 +462,7 @@ static void dev_watchdog(struct timer_list *t) struct netdev_queue *txq; txq = netdev_get_tx_queue(dev, i); - trans_start = txq->trans_start; + trans_start = READ_ONCE(txq->trans_start); if (netif_xmit_stopped(txq) && time_after(jiffies, (trans_start + dev->watchdog_timeo))) { @@ -1148,7 +1148,7 @@ static void transition_one_qdisc(struct net_device *dev, rcu_assign_pointer(dev_queue->qdisc, new_qdisc); if (need_watchdog_p) { - dev_queue->trans_start = 0; + WRITE_ONCE(dev_queue->trans_start, 0); *need_watchdog_p = 1; } } From dab8fe320726b38a6b1dc6a7ca6e386c5f7779e8 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 16 Nov 2021 19:29:23 -0800 Subject: [PATCH 0301/2048] net: do not inline netif_tx_lock()/netif_tx_unlock() These are not fast path, there is no point in inlining them. Also provide netif_freeze_queues()/netif_unfreeze_queues() so that we can use them from dev_watchdog() in the following patch. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/netdevice.h | 39 ++---------------------------- net/sched/sch_generic.c | 51 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 53 insertions(+), 37 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 279409ef2b18..4f4a299e92de 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -4126,27 +4126,7 @@ static inline void netif_trans_update(struct net_device *dev) * * Get network device transmit lock */ -static inline void netif_tx_lock(struct net_device *dev) -{ - unsigned int i; - int cpu; - - spin_lock(&dev->tx_global_lock); - cpu = smp_processor_id(); - for (i = 0; i < dev->num_tx_queues; i++) { - struct netdev_queue *txq = netdev_get_tx_queue(dev, i); - - /* We are the only thread of execution doing a - * freeze, but we have to grab the _xmit_lock in - * order to synchronize with threads which are in - * the ->hard_start_xmit() handler and already - * checked the frozen bit. - */ - __netif_tx_lock(txq, cpu); - set_bit(__QUEUE_STATE_FROZEN, &txq->state); - __netif_tx_unlock(txq); - } -} +void netif_tx_lock(struct net_device *dev); static inline void netif_tx_lock_bh(struct net_device *dev) { @@ -4154,22 +4134,7 @@ static inline void netif_tx_lock_bh(struct net_device *dev) netif_tx_lock(dev); } -static inline void netif_tx_unlock(struct net_device *dev) -{ - unsigned int i; - - for (i = 0; i < dev->num_tx_queues; i++) { - struct netdev_queue *txq = netdev_get_tx_queue(dev, i); - - /* No need to grab the _xmit_lock here. If the - * queue is not stopped for another reason, we - * force a schedule. - */ - clear_bit(__QUEUE_STATE_FROZEN, &txq->state); - netif_schedule_queue(txq); - } - spin_unlock(&dev->tx_global_lock); -} +void netif_tx_unlock(struct net_device *dev); static inline void netif_tx_unlock_bh(struct net_device *dev) { diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 02c46041f76e..389e0d8fc68d 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -445,6 +445,57 @@ unsigned long dev_trans_start(struct net_device *dev) } EXPORT_SYMBOL(dev_trans_start); +static void netif_freeze_queues(struct net_device *dev) +{ + unsigned int i; + int cpu; + + cpu = smp_processor_id(); + for (i = 0; i < dev->num_tx_queues; i++) { + struct netdev_queue *txq = netdev_get_tx_queue(dev, i); + + /* We are the only thread of execution doing a + * freeze, but we have to grab the _xmit_lock in + * order to synchronize with threads which are in + * the ->hard_start_xmit() handler and already + * checked the frozen bit. + */ + __netif_tx_lock(txq, cpu); + set_bit(__QUEUE_STATE_FROZEN, &txq->state); + __netif_tx_unlock(txq); + } +} + +void netif_tx_lock(struct net_device *dev) +{ + spin_lock(&dev->tx_global_lock); + netif_freeze_queues(dev); +} +EXPORT_SYMBOL(netif_tx_lock); + +static void netif_unfreeze_queues(struct net_device *dev) +{ + unsigned int i; + + for (i = 0; i < dev->num_tx_queues; i++) { + struct netdev_queue *txq = netdev_get_tx_queue(dev, i); + + /* No need to grab the _xmit_lock here. If the + * queue is not stopped for another reason, we + * force a schedule. + */ + clear_bit(__QUEUE_STATE_FROZEN, &txq->state); + netif_schedule_queue(txq); + } +} + +void netif_tx_unlock(struct net_device *dev) +{ + netif_unfreeze_queues(dev); + spin_unlock(&dev->tx_global_lock); +} +EXPORT_SYMBOL(netif_tx_unlock); + static void dev_watchdog(struct timer_list *t) { struct net_device *dev = from_timer(dev, t, watchdog_timer); From bec251bc8b6ab83464f6fca6842ad4ee47307d2e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 16 Nov 2021 19:29:24 -0800 Subject: [PATCH 0302/2048] net: no longer stop all TX queues in dev_watchdog() There is no reason for stopping all TX queues from dev_watchdog() Not only this stops feeding the NIC, it also migrates all qdiscs to be serviced on the cpu calling netif_tx_unlock(), causing a potential latency artifact. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/sched/sch_generic.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 389e0d8fc68d..d33804d41c5c 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -500,7 +500,7 @@ static void dev_watchdog(struct timer_list *t) { struct net_device *dev = from_timer(dev, t, watchdog_timer); - netif_tx_lock(dev); + spin_lock(&dev->tx_global_lock); if (!qdisc_tx_is_noop(dev)) { if (netif_device_present(dev) && netif_running(dev) && @@ -523,11 +523,13 @@ static void dev_watchdog(struct timer_list *t) } } - if (some_queue_timedout) { + if (unlikely(some_queue_timedout)) { trace_net_dev_xmit_timeout(dev, i); WARN_ONCE(1, KERN_INFO "NETDEV WATCHDOG: %s (%s): transmit queue %u timed out\n", dev->name, netdev_drivername(dev), i); + netif_freeze_queues(dev); dev->netdev_ops->ndo_tx_timeout(dev, i); + netif_unfreeze_queues(dev); } if (!mod_timer(&dev->watchdog_timer, round_jiffies(jiffies + @@ -535,7 +537,7 @@ static void dev_watchdog(struct timer_list *t) dev_hold(dev); } } - netif_tx_unlock(dev); + spin_unlock(&dev->tx_global_lock); dev_put(dev); } From 0edbecd5705728e803a56d646194f4ccdef5845f Mon Sep 17 00:00:00 2001 From: Radoslaw Tyl Date: Wed, 30 Jun 2021 10:15:28 +0200 Subject: [PATCH 0303/2048] ixgbevf: Rename MSGTYPE to SUCCESS and FAILURE There is name similarity within IXGBE_VT_MSGTYPE_ACK and PFMAILBOX.ACK / VFMAILBOX.ACK. MSGTYPE macros are renamed to SUCCESS and FAILURE because they are not specified in datasheet and now will be easily distinguishable. Signed-off-by: Radoslaw Tyl Tested-by: Tony Brelinski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ixgbevf/ipsec.c | 2 +- drivers/net/ethernet/intel/ixgbevf/mbx.h | 17 +++++++----- drivers/net/ethernet/intel/ixgbevf/vf.c | 31 +++++++++++----------- 3 files changed, 27 insertions(+), 23 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ipsec.c b/drivers/net/ethernet/intel/ixgbevf/ipsec.c index e3e4676af9e4..9b84e122c7c3 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ipsec.c +++ b/drivers/net/ethernet/intel/ixgbevf/ipsec.c @@ -49,7 +49,7 @@ static int ixgbevf_ipsec_set_pf_sa(struct ixgbevf_adapter *adapter, goto out; ret = (int)msgbuf[1]; - if (msgbuf[0] & IXGBE_VT_MSGTYPE_NACK && ret >= 0) + if (msgbuf[0] & IXGBE_VT_MSGTYPE_FAILURE && ret >= 0) ret = -1; out: diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.h b/drivers/net/ethernet/intel/ixgbevf/mbx.h index 853796c8ef0e..a461b7d16206 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.h +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.h @@ -39,14 +39,17 @@ /* If it's a IXGBE_VF_* msg then it originates in the VF and is sent to the * PF. The reverse is true if it is IXGBE_PF_*. - * Message ACK's are the value or'd with 0xF0000000 + * Message results are the value or'd with 0xF0000000 */ -/* Messages below or'd with this are the ACK */ -#define IXGBE_VT_MSGTYPE_ACK 0x80000000 -/* Messages below or'd with this are the NACK */ -#define IXGBE_VT_MSGTYPE_NACK 0x40000000 -/* Indicates that VF is still clear to send requests */ -#define IXGBE_VT_MSGTYPE_CTS 0x20000000 +#define IXGBE_VT_MSGTYPE_SUCCESS 0x80000000 /* Messages or'd with this + * have succeeded + */ +#define IXGBE_VT_MSGTYPE_FAILURE 0x40000000 /* Messages or'd with this + * have failed + */ +#define IXGBE_VT_MSGTYPE_CTS 0x20000000 /* Indicates that VF is still + * clear to send requests + */ #define IXGBE_VT_MSGINFO_SHIFT 16 /* bits 23:16 are used for exra info for certain messages */ #define IXGBE_VT_MSGINFO_MASK (0xFF << IXGBE_VT_MSGINFO_SHIFT) diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c index d459f5c8e98f..f5d0d7969144 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.c +++ b/drivers/net/ethernet/intel/ixgbevf/vf.c @@ -108,11 +108,11 @@ static s32 ixgbevf_reset_hw_vf(struct ixgbe_hw *hw) * to indicate that no MAC address has yet been assigned for * the VF. */ - if (msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_ACK) && - msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_NACK)) + if (msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_SUCCESS) && + msgbuf[0] != (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_FAILURE)) return IXGBE_ERR_INVALID_MAC_ADDR; - if (msgbuf[0] == (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_ACK)) + if (msgbuf[0] == (IXGBE_VF_RESET | IXGBE_VT_MSGTYPE_SUCCESS)) ether_addr_copy(hw->mac.perm_addr, addr); hw->mac.mc_filter_type = msgbuf[IXGBE_VF_MC_TYPE_WORD]; @@ -269,7 +269,7 @@ static s32 ixgbevf_set_uc_addr_vf(struct ixgbe_hw *hw, u32 index, u8 *addr) if (!ret_val) { msgbuf[0] &= ~IXGBE_VT_MSGTYPE_CTS; - if (msgbuf[0] == (msgbuf_chk | IXGBE_VT_MSGTYPE_NACK)) + if (msgbuf[0] == (msgbuf_chk | IXGBE_VT_MSGTYPE_FAILURE)) return -ENOMEM; } @@ -336,14 +336,14 @@ int ixgbevf_get_reta_locked(struct ixgbe_hw *hw, u32 *reta, int num_rx_queues) msgbuf[0] &= ~IXGBE_VT_MSGTYPE_CTS; /* If the operation has been refused by a PF return -EPERM */ - if (msgbuf[0] == (IXGBE_VF_GET_RETA | IXGBE_VT_MSGTYPE_NACK)) + if (msgbuf[0] == (IXGBE_VF_GET_RETA | IXGBE_VT_MSGTYPE_FAILURE)) return -EPERM; /* If we didn't get an ACK there must have been * some sort of mailbox error so we should treat it * as such. */ - if (msgbuf[0] != (IXGBE_VF_GET_RETA | IXGBE_VT_MSGTYPE_ACK)) + if (msgbuf[0] != (IXGBE_VF_GET_RETA | IXGBE_VT_MSGTYPE_SUCCESS)) return IXGBE_ERR_MBX; /* ixgbevf doesn't support more than 2 queues at the moment */ @@ -403,14 +403,14 @@ int ixgbevf_get_rss_key_locked(struct ixgbe_hw *hw, u8 *rss_key) msgbuf[0] &= ~IXGBE_VT_MSGTYPE_CTS; /* If the operation has been refused by a PF return -EPERM */ - if (msgbuf[0] == (IXGBE_VF_GET_RSS_KEY | IXGBE_VT_MSGTYPE_NACK)) + if (msgbuf[0] == (IXGBE_VF_GET_RSS_KEY | IXGBE_VT_MSGTYPE_FAILURE)) return -EPERM; /* If we didn't get an ACK there must have been * some sort of mailbox error so we should treat it * as such. */ - if (msgbuf[0] != (IXGBE_VF_GET_RSS_KEY | IXGBE_VT_MSGTYPE_ACK)) + if (msgbuf[0] != (IXGBE_VF_GET_RSS_KEY | IXGBE_VT_MSGTYPE_SUCCESS)) return IXGBE_ERR_MBX; memcpy(rss_key, msgbuf + 1, IXGBEVF_RSS_HASH_KEY_SIZE); @@ -442,7 +442,7 @@ static s32 ixgbevf_set_rar_vf(struct ixgbe_hw *hw, u32 index, u8 *addr, /* if nacked the address was rejected, use "perm_addr" */ if (!ret_val && - (msgbuf[0] == (IXGBE_VF_SET_MAC_ADDR | IXGBE_VT_MSGTYPE_NACK))) { + (msgbuf[0] == (IXGBE_VF_SET_MAC_ADDR | IXGBE_VT_MSGTYPE_FAILURE))) { ixgbevf_get_mac_addr_vf(hw, hw->mac.addr); return IXGBE_ERR_MBX; } @@ -561,7 +561,7 @@ static s32 ixgbevf_update_xcast_mode(struct ixgbe_hw *hw, int xcast_mode) return err; msgbuf[0] &= ~IXGBE_VT_MSGTYPE_CTS; - if (msgbuf[0] == (IXGBE_VF_UPDATE_XCAST_MODE | IXGBE_VT_MSGTYPE_NACK)) + if (msgbuf[0] == (IXGBE_VF_UPDATE_XCAST_MODE | IXGBE_VT_MSGTYPE_FAILURE)) return -EPERM; return 0; @@ -606,7 +606,7 @@ static s32 ixgbevf_set_vfta_vf(struct ixgbe_hw *hw, u32 vlan, u32 vind, msgbuf[0] &= ~IXGBE_VT_MSGTYPE_CTS; msgbuf[0] &= ~(0xFF << IXGBE_VT_MSGINFO_SHIFT); - if (msgbuf[0] != (IXGBE_VF_SET_VLAN | IXGBE_VT_MSGTYPE_ACK)) + if (msgbuf[0] != (IXGBE_VF_SET_VLAN | IXGBE_VT_MSGTYPE_SUCCESS)) err = IXGBE_ERR_INVALID_ARGUMENT; mbx_err: @@ -710,7 +710,7 @@ static s32 ixgbevf_check_mac_link_vf(struct ixgbe_hw *hw, if (!(in_msg & IXGBE_VT_MSGTYPE_CTS)) { /* msg is not CTS and is NACK we must have lost CTS status */ - if (in_msg & IXGBE_VT_MSGTYPE_NACK) + if (in_msg & IXGBE_VT_MSGTYPE_FAILURE) ret_val = -1; goto out; } @@ -816,7 +816,7 @@ static s32 ixgbevf_set_rlpml_vf(struct ixgbe_hw *hw, u16 max_size) if (ret_val) return ret_val; if ((msgbuf[0] & IXGBE_VF_SET_LPE) && - (msgbuf[0] & IXGBE_VT_MSGTYPE_NACK)) + (msgbuf[0] & IXGBE_VT_MSGTYPE_FAILURE)) return IXGBE_ERR_MBX; return 0; @@ -863,7 +863,8 @@ static int ixgbevf_negotiate_api_version_vf(struct ixgbe_hw *hw, int api) msg[0] &= ~IXGBE_VT_MSGTYPE_CTS; /* Store value and return 0 on success */ - if (msg[0] == (IXGBE_VF_API_NEGOTIATE | IXGBE_VT_MSGTYPE_ACK)) { + if (msg[0] == (IXGBE_VF_API_NEGOTIATE | + IXGBE_VT_MSGTYPE_SUCCESS)) { hw->api_version = api; return 0; } @@ -918,7 +919,7 @@ int ixgbevf_get_queues(struct ixgbe_hw *hw, unsigned int *num_tcs, * some sort of mailbox error so we should treat it * as such */ - if (msg[0] != (IXGBE_VF_GET_QUEUE | IXGBE_VT_MSGTYPE_ACK)) + if (msg[0] != (IXGBE_VF_GET_QUEUE | IXGBE_VT_MSGTYPE_SUCCESS)) return IXGBE_ERR_MBX; /* record and validate values from message */ From 887a32031a8ae5d9ad805973f28744ebe685c4e5 Mon Sep 17 00:00:00 2001 From: Radoslaw Tyl Date: Wed, 30 Jun 2021 10:15:29 +0200 Subject: [PATCH 0304/2048] ixgbevf: Improve error handling in mailbox Add new handling for error codes: IXGBE_ERR_CONFIG - ixgbe_mbx_operations is not correctly set IXGBE_ERR_TIMEOUT - mailbox operation, e.g. poll for message, timeout Signed-off-by: Radoslaw Tyl Tested-by: Tony Brelinski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ixgbevf/defines.h | 3 +++ drivers/net/ethernet/intel/ixgbevf/mbx.c | 14 ++++++++++---- drivers/net/ethernet/intel/ixgbevf/mbx.h | 1 - 3 files changed, 13 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/defines.h b/drivers/net/ethernet/intel/ixgbevf/defines.h index 6bace746eaac..46fb1f9eab7f 100644 --- a/drivers/net/ethernet/intel/ixgbevf/defines.h +++ b/drivers/net/ethernet/intel/ixgbevf/defines.h @@ -281,6 +281,9 @@ struct ixgbe_adv_tx_context_desc { #define IXGBE_ERR_INVALID_MAC_ADDR -1 #define IXGBE_ERR_RESET_FAILED -2 #define IXGBE_ERR_INVALID_ARGUMENT -3 +#define IXGBE_ERR_CONFIG -4 +#define IXGBE_ERR_MBX -5 +#define IXGBE_ERR_TIMEOUT -6 /* Transmit Config masks */ #define IXGBE_TXDCTL_ENABLE 0x02000000 /* Ena specific Tx Queue */ diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.c b/drivers/net/ethernet/intel/ixgbevf/mbx.c index 6bc1953263b9..2c3762cb467d 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.c +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.c @@ -15,6 +15,9 @@ static s32 ixgbevf_poll_for_msg(struct ixgbe_hw *hw) struct ixgbe_mbx_info *mbx = &hw->mbx; int countdown = mbx->timeout; + if (!countdown || !mbx->ops.check_for_msg) + return IXGBE_ERR_CONFIG; + while (countdown && mbx->ops.check_for_msg(hw)) { countdown--; udelay(mbx->udelay); @@ -24,7 +27,7 @@ static s32 ixgbevf_poll_for_msg(struct ixgbe_hw *hw) if (!countdown) mbx->timeout = 0; - return countdown ? 0 : IXGBE_ERR_MBX; + return countdown ? 0 : IXGBE_ERR_TIMEOUT; } /** @@ -38,6 +41,9 @@ static s32 ixgbevf_poll_for_ack(struct ixgbe_hw *hw) struct ixgbe_mbx_info *mbx = &hw->mbx; int countdown = mbx->timeout; + if (!countdown || !mbx->ops.check_for_ack) + return IXGBE_ERR_CONFIG; + while (countdown && mbx->ops.check_for_ack(hw)) { countdown--; udelay(mbx->udelay); @@ -47,7 +53,7 @@ static s32 ixgbevf_poll_for_ack(struct ixgbe_hw *hw) if (!countdown) mbx->timeout = 0; - return countdown ? 0 : IXGBE_ERR_MBX; + return countdown ? 0 : IXGBE_ERR_TIMEOUT; } /** @@ -62,7 +68,7 @@ static s32 ixgbevf_poll_for_ack(struct ixgbe_hw *hw) static s32 ixgbevf_read_posted_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) { struct ixgbe_mbx_info *mbx = &hw->mbx; - s32 ret_val = IXGBE_ERR_MBX; + s32 ret_val = IXGBE_ERR_CONFIG; if (!mbx->ops.read) goto out; @@ -88,7 +94,7 @@ out: static s32 ixgbevf_write_posted_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) { struct ixgbe_mbx_info *mbx = &hw->mbx; - s32 ret_val = IXGBE_ERR_MBX; + s32 ret_val = IXGBE_ERR_CONFIG; /* exit if either we can't write or there isn't a defined timeout */ if (!mbx->ops.write || !mbx->timeout) diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.h b/drivers/net/ethernet/intel/ixgbevf/mbx.h index a461b7d16206..b3b83c95babf 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.h +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.h @@ -7,7 +7,6 @@ #include "vf.h" #define IXGBE_VFMAILBOX_SIZE 16 /* 16 32 bit words - 64 bytes */ -#define IXGBE_ERR_MBX -100 #define IXGBE_VFMAILBOX 0x002FC #define IXGBE_VFMBMEM 0x00200 From 9c9463c29d1b957f2e4cbe31ad88a4b6a3086199 Mon Sep 17 00:00:00 2001 From: Radoslaw Tyl Date: Wed, 30 Jun 2021 10:15:30 +0200 Subject: [PATCH 0305/2048] ixgbevf: Add legacy suffix to old API mailbox functions Add legacy suffix to mailbox functions which should be backwards compatible with older PF drivers. Communication during API negotiation always has to be done using the earlier implementation. Signed-off-by: Radoslaw Tyl Tested-by: Tony Brelinski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ixgbevf/ixgbevf.h | 2 +- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 +- drivers/net/ethernet/intel/ixgbevf/mbx.c | 14 +++++++------- 3 files changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h index a0e325774819..5d9284dd04dc 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h @@ -429,7 +429,7 @@ extern const struct ixgbevf_info ixgbevf_82599_vf_info; extern const struct ixgbevf_info ixgbevf_X540_vf_info; extern const struct ixgbevf_info ixgbevf_X550_vf_info; extern const struct ixgbevf_info ixgbevf_X550EM_x_vf_info; -extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops; +extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops_legacy; extern const struct ixgbevf_info ixgbevf_x550em_a_vf_info; extern const struct ixgbevf_info ixgbevf_82599_vf_hv_info; diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index d81811ab4ec4..cd93b47a61cd 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -4565,7 +4565,7 @@ static int ixgbevf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) memcpy(&hw->mac.ops, ii->mac_ops, sizeof(hw->mac.ops)); hw->mac.type = ii->mac; - memcpy(&hw->mbx.ops, &ixgbevf_mbx_ops, + memcpy(&hw->mbx.ops, &ixgbevf_mbx_ops_legacy, sizeof(struct ixgbe_mbx_operations)); /* setup the private structure */ diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.c b/drivers/net/ethernet/intel/ixgbevf/mbx.c index 2c3762cb467d..0edcfcbf5296 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.c +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.c @@ -224,14 +224,14 @@ static s32 ixgbevf_obtain_mbx_lock_vf(struct ixgbe_hw *hw) } /** - * ixgbevf_write_mbx_vf - Write a message to the mailbox + * ixgbevf_write_mbx_vf_legacy - Write a message to the mailbox * @hw: pointer to the HW structure * @msg: The message buffer * @size: Length of buffer * * returns 0 if it successfully copied message into the buffer **/ -static s32 ixgbevf_write_mbx_vf(struct ixgbe_hw *hw, u32 *msg, u16 size) +static s32 ixgbevf_write_mbx_vf_legacy(struct ixgbe_hw *hw, u32 *msg, u16 size) { s32 ret_val; u16 i; @@ -260,14 +260,14 @@ out_no_write: } /** - * ixgbevf_read_mbx_vf - Reads a message from the inbox intended for VF + * ixgbevf_read_mbx_vf_legacy - Reads a message from the inbox intended for VF * @hw: pointer to the HW structure * @msg: The message buffer * @size: Length of buffer * * returns 0 if it successfully read message from buffer **/ -static s32 ixgbevf_read_mbx_vf(struct ixgbe_hw *hw, u32 *msg, u16 size) +static s32 ixgbevf_read_mbx_vf_legacy(struct ixgbe_hw *hw, u32 *msg, u16 size) { s32 ret_val = 0; u16 i; @@ -318,10 +318,10 @@ static s32 ixgbevf_init_mbx_params_vf(struct ixgbe_hw *hw) return 0; } -const struct ixgbe_mbx_operations ixgbevf_mbx_ops = { +const struct ixgbe_mbx_operations ixgbevf_mbx_ops_legacy = { .init_params = ixgbevf_init_mbx_params_vf, - .read = ixgbevf_read_mbx_vf, - .write = ixgbevf_write_mbx_vf, + .read = ixgbevf_read_mbx_vf_legacy, + .write = ixgbevf_write_mbx_vf_legacy, .read_posted = ixgbevf_read_posted_mbx, .write_posted = ixgbevf_write_posted_mbx, .check_for_msg = ixgbevf_check_for_msg_vf, From c869259881a36fa86086efe661c74a6a9de5b169 Mon Sep 17 00:00:00 2001 From: Radoslaw Tyl Date: Wed, 30 Jun 2021 10:15:31 +0200 Subject: [PATCH 0306/2048] ixgbevf: Mailbox improvements Improve reliability of the mailbox communication and remove its potential flaws that may lead to the undefined or faulty behavior. Recently some users reported issues on ESX with 10G Intel NICs which were found to be caused by incorrect implementation of the PF-VF mailbox communication. Technical investigation highlighted areas to improve in the communication between PF or VF that wants to send the message (sender) and the other part which receives the message (receiver): - Locking the mailbox when the sender wants to send a message - Releasing the mailbox when the communication ends - Returning the result of the mailbox message execution Signed-off-by: Radoslaw Tyl Tested-by: Tony Brelinski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ixgbevf/defines.h | 1 + drivers/net/ethernet/intel/ixgbevf/ipsec.c | 8 +- drivers/net/ethernet/intel/ixgbevf/ixgbevf.h | 5 + drivers/net/ethernet/intel/ixgbevf/mbx.c | 315 +++++++++++++++---- drivers/net/ethernet/intel/ixgbevf/vf.c | 17 +- drivers/net/ethernet/intel/ixgbevf/vf.h | 5 +- 6 files changed, 266 insertions(+), 85 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/defines.h b/drivers/net/ethernet/intel/ixgbevf/defines.h index 46fb1f9eab7f..5f08779c0e4e 100644 --- a/drivers/net/ethernet/intel/ixgbevf/defines.h +++ b/drivers/net/ethernet/intel/ixgbevf/defines.h @@ -284,6 +284,7 @@ struct ixgbe_adv_tx_context_desc { #define IXGBE_ERR_CONFIG -4 #define IXGBE_ERR_MBX -5 #define IXGBE_ERR_TIMEOUT -6 +#define IXGBE_ERR_PARAM -7 /* Transmit Config masks */ #define IXGBE_TXDCTL_ENABLE 0x02000000 /* Ena specific Tx Queue */ diff --git a/drivers/net/ethernet/intel/ixgbevf/ipsec.c b/drivers/net/ethernet/intel/ixgbevf/ipsec.c index 9b84e122c7c3..d9935a42f0b7 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ipsec.c +++ b/drivers/net/ethernet/intel/ixgbevf/ipsec.c @@ -40,11 +40,11 @@ static int ixgbevf_ipsec_set_pf_sa(struct ixgbevf_adapter *adapter, spin_lock_bh(&adapter->mbx_lock); - ret = hw->mbx.ops.write_posted(hw, msgbuf, IXGBE_VFMAILBOX_SIZE); + ret = ixgbevf_write_mbx(hw, msgbuf, IXGBE_VFMAILBOX_SIZE); if (ret) goto out; - ret = hw->mbx.ops.read_posted(hw, msgbuf, 2); + ret = ixgbevf_poll_mbx(hw, msgbuf, 2); if (ret) goto out; @@ -77,11 +77,11 @@ static int ixgbevf_ipsec_del_pf_sa(struct ixgbevf_adapter *adapter, int pfsa) spin_lock_bh(&adapter->mbx_lock); - err = hw->mbx.ops.write_posted(hw, msgbuf, 2); + err = ixgbevf_write_mbx(hw, msgbuf, 2); if (err) goto out; - err = hw->mbx.ops.read_posted(hw, msgbuf, 2); + err = ixgbevf_poll_mbx(hw, msgbuf, 2); if (err) goto out; diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h index 5d9284dd04dc..e257390a4f6a 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h @@ -429,6 +429,7 @@ extern const struct ixgbevf_info ixgbevf_82599_vf_info; extern const struct ixgbevf_info ixgbevf_X540_vf_info; extern const struct ixgbevf_info ixgbevf_X550_vf_info; extern const struct ixgbevf_info ixgbevf_X550EM_x_vf_info; +extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops; extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops_legacy; extern const struct ixgbevf_info ixgbevf_x550em_a_vf_info; @@ -491,4 +492,8 @@ void ixgbe_napi_del_all(struct ixgbevf_adapter *adapter); #define hw_dbg(hw, format, arg...) \ netdev_dbg(ixgbevf_hw_to_netdev(hw), format, ## arg) + +s32 ixgbevf_poll_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size); +s32 ixgbevf_write_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size); + #endif /* _IXGBEVF_H_ */ diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.c b/drivers/net/ethernet/intel/ixgbevf/mbx.c index 0edcfcbf5296..a55dd978f7ca 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.c +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.c @@ -23,10 +23,6 @@ static s32 ixgbevf_poll_for_msg(struct ixgbe_hw *hw) udelay(mbx->udelay); } - /* if we failed, all future posted messages fail until reset */ - if (!countdown) - mbx->timeout = 0; - return countdown ? 0 : IXGBE_ERR_TIMEOUT; } @@ -49,82 +45,74 @@ static s32 ixgbevf_poll_for_ack(struct ixgbe_hw *hw) udelay(mbx->udelay); } - /* if we failed, all future posted messages fail until reset */ - if (!countdown) - mbx->timeout = 0; - return countdown ? 0 : IXGBE_ERR_TIMEOUT; } /** - * ixgbevf_read_posted_mbx - Wait for message notification and receive message - * @hw: pointer to the HW structure - * @msg: The message buffer - * @size: Length of buffer + * ixgbevf_read_mailbox_vf - read VF's mailbox register + * @hw: pointer to the HW structure * - * returns 0 if it successfully received a message notification and - * copied it into the receive buffer. + * This function is used to read the mailbox register dedicated for VF without + * losing the read to clear status bits. **/ -static s32 ixgbevf_read_posted_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) +static u32 ixgbevf_read_mailbox_vf(struct ixgbe_hw *hw) { - struct ixgbe_mbx_info *mbx = &hw->mbx; - s32 ret_val = IXGBE_ERR_CONFIG; + u32 vf_mailbox = IXGBE_READ_REG(hw, IXGBE_VFMAILBOX); - if (!mbx->ops.read) - goto out; + vf_mailbox |= hw->mbx.vf_mailbox; + hw->mbx.vf_mailbox |= vf_mailbox & IXGBE_VFMAILBOX_R2C_BITS; - ret_val = ixgbevf_poll_for_msg(hw); - - /* if ack received read message, otherwise we timed out */ - if (!ret_val) - ret_val = mbx->ops.read(hw, msg, size); -out: - return ret_val; + return vf_mailbox; } /** - * ixgbevf_write_posted_mbx - Write a message to the mailbox, wait for ack - * @hw: pointer to the HW structure - * @msg: The message buffer - * @size: Length of buffer + * ixgbevf_clear_msg_vf - clear PF status bit + * @hw: pointer to the HW structure * - * returns 0 if it successfully copied message into the buffer and - * received an ack to that message within delay * timeout period + * This function is used to clear PFSTS bit in the VFMAILBOX register **/ -static s32 ixgbevf_write_posted_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) +static void ixgbevf_clear_msg_vf(struct ixgbe_hw *hw) { - struct ixgbe_mbx_info *mbx = &hw->mbx; - s32 ret_val = IXGBE_ERR_CONFIG; + u32 vf_mailbox = ixgbevf_read_mailbox_vf(hw); - /* exit if either we can't write or there isn't a defined timeout */ - if (!mbx->ops.write || !mbx->timeout) - goto out; - - /* send msg */ - ret_val = mbx->ops.write(hw, msg, size); - - /* if msg sent wait until we receive an ack */ - if (!ret_val) - ret_val = ixgbevf_poll_for_ack(hw); -out: - return ret_val; + if (vf_mailbox & IXGBE_VFMAILBOX_PFSTS) { + hw->mbx.stats.reqs++; + hw->mbx.vf_mailbox &= ~IXGBE_VFMAILBOX_PFSTS; + } } /** - * ixgbevf_read_v2p_mailbox - read v2p mailbox - * @hw: pointer to the HW structure + * ixgbevf_clear_ack_vf - clear PF ACK bit + * @hw: pointer to the HW structure * - * This function is used to read the v2p mailbox without losing the read to - * clear status bits. + * This function is used to clear PFACK bit in the VFMAILBOX register **/ -static u32 ixgbevf_read_v2p_mailbox(struct ixgbe_hw *hw) +static void ixgbevf_clear_ack_vf(struct ixgbe_hw *hw) { - u32 v2p_mailbox = IXGBE_READ_REG(hw, IXGBE_VFMAILBOX); + u32 vf_mailbox = ixgbevf_read_mailbox_vf(hw); - v2p_mailbox |= hw->mbx.v2p_mailbox; - hw->mbx.v2p_mailbox |= v2p_mailbox & IXGBE_VFMAILBOX_R2C_BITS; + if (vf_mailbox & IXGBE_VFMAILBOX_PFACK) { + hw->mbx.stats.acks++; + hw->mbx.vf_mailbox &= ~IXGBE_VFMAILBOX_PFACK; + } +} - return v2p_mailbox; +/** + * ixgbevf_clear_rst_vf - clear PF reset bit + * @hw: pointer to the HW structure + * + * This function is used to clear reset indication and reset done bit in + * VFMAILBOX register after reset the shared resources and the reset sequence. + **/ +static void ixgbevf_clear_rst_vf(struct ixgbe_hw *hw) +{ + u32 vf_mailbox = ixgbevf_read_mailbox_vf(hw); + + if (vf_mailbox & (IXGBE_VFMAILBOX_RSTI | IXGBE_VFMAILBOX_RSTD)) { + hw->mbx.stats.rsts++; + hw->mbx.vf_mailbox &= ~(IXGBE_VFMAILBOX_RSTI | + IXGBE_VFMAILBOX_RSTD); + } } /** @@ -137,14 +125,12 @@ static u32 ixgbevf_read_v2p_mailbox(struct ixgbe_hw *hw) **/ static s32 ixgbevf_check_for_bit_vf(struct ixgbe_hw *hw, u32 mask) { - u32 v2p_mailbox = ixgbevf_read_v2p_mailbox(hw); + u32 vf_mailbox = ixgbevf_read_mailbox_vf(hw); s32 ret_val = IXGBE_ERR_MBX; - if (v2p_mailbox & mask) + if (vf_mailbox & mask) ret_val = 0; - hw->mbx.v2p_mailbox &= ~mask; - return ret_val; } @@ -178,6 +164,7 @@ static s32 ixgbevf_check_for_ack_vf(struct ixgbe_hw *hw) if (!ixgbevf_check_for_bit_vf(hw, IXGBE_VFMAILBOX_PFACK)) { ret_val = 0; + ixgbevf_clear_ack_vf(hw); hw->mbx.stats.acks++; } @@ -197,6 +184,7 @@ static s32 ixgbevf_check_for_rst_vf(struct ixgbe_hw *hw) if (!ixgbevf_check_for_bit_vf(hw, (IXGBE_VFMAILBOX_RSTD | IXGBE_VFMAILBOX_RSTI))) { ret_val = 0; + ixgbevf_clear_rst_vf(hw); hw->mbx.stats.rsts++; } @@ -211,14 +199,98 @@ static s32 ixgbevf_check_for_rst_vf(struct ixgbe_hw *hw) **/ static s32 ixgbevf_obtain_mbx_lock_vf(struct ixgbe_hw *hw) { - s32 ret_val = IXGBE_ERR_MBX; + struct ixgbe_mbx_info *mbx = &hw->mbx; + s32 ret_val = IXGBE_ERR_CONFIG; + int countdown = mbx->timeout; + u32 vf_mailbox; - /* Take ownership of the buffer */ - IXGBE_WRITE_REG(hw, IXGBE_VFMAILBOX, IXGBE_VFMAILBOX_VFU); + if (!mbx->timeout) + return ret_val; - /* reserve mailbox for VF use */ - if (ixgbevf_read_v2p_mailbox(hw) & IXGBE_VFMAILBOX_VFU) - ret_val = 0; + while (countdown--) { + /* Reserve mailbox for VF use */ + vf_mailbox = ixgbevf_read_mailbox_vf(hw); + vf_mailbox |= IXGBE_VFMAILBOX_VFU; + IXGBE_WRITE_REG(hw, IXGBE_VFMAILBOX, vf_mailbox); + + /* Verify that VF is the owner of the lock */ + if (ixgbevf_read_mailbox_vf(hw) & IXGBE_VFMAILBOX_VFU) { + ret_val = 0; + break; + } + + /* Wait a bit before trying again */ + udelay(mbx->udelay); + } + + if (ret_val) + ret_val = IXGBE_ERR_TIMEOUT; + + return ret_val; +} + +/** + * ixgbevf_release_mbx_lock_vf - release mailbox lock + * @hw: pointer to the HW structure + **/ +static void ixgbevf_release_mbx_lock_vf(struct ixgbe_hw *hw) +{ + u32 vf_mailbox; + + /* Return ownership of the buffer */ + vf_mailbox = ixgbevf_read_mailbox_vf(hw); + vf_mailbox &= ~IXGBE_VFMAILBOX_VFU; + IXGBE_WRITE_REG(hw, IXGBE_VFMAILBOX, vf_mailbox); +} + +/** + * ixgbevf_release_mbx_lock_vf_legacy - release mailbox lock + * @hw: pointer to the HW structure + **/ +static void ixgbevf_release_mbx_lock_vf_legacy(struct ixgbe_hw *__always_unused hw) +{ +} + +/** + * ixgbevf_write_mbx_vf - Write a message to the mailbox + * @hw: pointer to the HW structure + * @msg: The message buffer + * @size: Length of buffer + * + * returns 0 if it successfully copied message into the buffer + **/ +static s32 ixgbevf_write_mbx_vf(struct ixgbe_hw *hw, u32 *msg, u16 size) +{ + u32 vf_mailbox; + s32 ret_val; + u16 i; + + /* lock the mailbox to prevent PF/VF race condition */ + ret_val = ixgbevf_obtain_mbx_lock_vf(hw); + if (ret_val) + goto out_no_write; + + /* flush msg and acks as we are overwriting the message buffer */ + ixgbevf_clear_msg_vf(hw); + ixgbevf_clear_ack_vf(hw); + + /* copy the caller specified message to the mailbox memory buffer */ + for (i = 0; i < size; i++) + IXGBE_WRITE_REG_ARRAY(hw, IXGBE_VFMBMEM, i, msg[i]); + + /* update stats */ + hw->mbx.stats.msgs_tx++; + + /* interrupt the PF to tell it a message has been sent */ + vf_mailbox = ixgbevf_read_mailbox_vf(hw); + vf_mailbox |= IXGBE_VFMAILBOX_REQ; + IXGBE_WRITE_REG(hw, IXGBE_VFMAILBOX, vf_mailbox); + + /* if msg sent wait until we receive an ack */ + ret_val = ixgbevf_poll_for_ack(hw); + +out_no_write: + hw->mbx.ops.release(hw); return ret_val; } @@ -243,7 +315,9 @@ static s32 ixgbevf_write_mbx_vf_legacy(struct ixgbe_hw *hw, u32 *msg, u16 size) /* flush msg and acks as we are overwriting the message buffer */ ixgbevf_check_for_msg_vf(hw); + ixgbevf_clear_msg_vf(hw); ixgbevf_check_for_ack_vf(hw); + ixgbevf_clear_ack_vf(hw); /* copy the caller specified message to the mailbox memory buffer */ for (i = 0; i < size; i++) @@ -259,6 +333,42 @@ out_no_write: return ret_val; } +/** + * ixgbevf_read_mbx_vf - Reads a message from the inbox intended for VF + * @hw: pointer to the HW structure + * @msg: The message buffer + * @size: Length of buffer + * + * returns 0 if it successfully read message from buffer + **/ +static s32 ixgbevf_read_mbx_vf(struct ixgbe_hw *hw, u32 *msg, u16 size) +{ + u32 vf_mailbox; + s32 ret_val; + u16 i; + + /* check if there is a message from PF */ + ret_val = ixgbevf_check_for_msg_vf(hw); + if (ret_val) + return ret_val; + + ixgbevf_clear_msg_vf(hw); + + /* copy the message from the mailbox memory buffer */ + for (i = 0; i < size; i++) + msg[i] = IXGBE_READ_REG_ARRAY(hw, IXGBE_VFMBMEM, i); + + /* Acknowledge receipt */ + vf_mailbox = ixgbevf_read_mailbox_vf(hw); + vf_mailbox |= IXGBE_VFMAILBOX_ACK; + IXGBE_WRITE_REG(hw, IXGBE_VFMAILBOX, vf_mailbox); + + /* update stats */ + hw->mbx.stats.msgs_rx++; + + return ret_val; +} + /** * ixgbevf_read_mbx_vf_legacy - Reads a message from the inbox intended for VF * @hw: pointer to the HW structure @@ -304,7 +414,7 @@ static s32 ixgbevf_init_mbx_params_vf(struct ixgbe_hw *hw) /* start mailbox as timed out and let the reset_hw call set the timeout * value to begin communications */ - mbx->timeout = 0; + mbx->timeout = IXGBE_VF_MBX_INIT_TIMEOUT; mbx->udelay = IXGBE_VF_MBX_INIT_DELAY; mbx->size = IXGBE_VFMAILBOX_SIZE; @@ -318,12 +428,79 @@ static s32 ixgbevf_init_mbx_params_vf(struct ixgbe_hw *hw) return 0; } +/** + * ixgbevf_poll_mbx - Wait for message and read it from the mailbox + * @hw: pointer to the HW structure + * @msg: The message buffer + * @size: Length of buffer + * + * returns 0 if it successfully read message from buffer + **/ +s32 ixgbevf_poll_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) +{ + struct ixgbe_mbx_info *mbx = &hw->mbx; + s32 ret_val = IXGBE_ERR_CONFIG; + + if (!mbx->ops.read || !mbx->ops.check_for_msg || !mbx->timeout) + return ret_val; + + /* limit read to size of mailbox */ + if (size > mbx->size) + size = mbx->size; + + ret_val = ixgbevf_poll_for_msg(hw); + /* if ack received read message, otherwise we timed out */ + if (!ret_val) + ret_val = mbx->ops.read(hw, msg, size); + + return ret_val; +} + +/** + * ixgbevf_write_mbx - Write a message to the mailbox and wait for ACK + * @hw: pointer to the HW structure + * @msg: The message buffer + * @size: Length of buffer + * + * returns 0 if it successfully copied message into the buffer and + * received an ACK to that message within specified period + **/ +s32 ixgbevf_write_mbx(struct ixgbe_hw *hw, u32 *msg, u16 size) +{ + struct ixgbe_mbx_info *mbx = &hw->mbx; + s32 ret_val = IXGBE_ERR_CONFIG; + + /** + * exit if either we can't write, release + * or there is no timeout defined + */ + if (!mbx->ops.write || !mbx->ops.check_for_ack || !mbx->ops.release || + !mbx->timeout) + return ret_val; + + if (size > mbx->size) + ret_val = IXGBE_ERR_PARAM; + else + ret_val = mbx->ops.write(hw, msg, size); + + return ret_val; +} + +const struct ixgbe_mbx_operations ixgbevf_mbx_ops = { + .init_params = ixgbevf_init_mbx_params_vf, + .release = ixgbevf_release_mbx_lock_vf, + .read = ixgbevf_read_mbx_vf, + .write = ixgbevf_write_mbx_vf, + .check_for_msg = ixgbevf_check_for_msg_vf, + .check_for_ack = ixgbevf_check_for_ack_vf, + .check_for_rst = ixgbevf_check_for_rst_vf, +}; + const struct ixgbe_mbx_operations ixgbevf_mbx_ops_legacy = { .init_params = ixgbevf_init_mbx_params_vf, + .release = ixgbevf_release_mbx_lock_vf_legacy, .read = ixgbevf_read_mbx_vf_legacy, .write = ixgbevf_write_mbx_vf_legacy, - .read_posted = ixgbevf_read_posted_mbx, - .write_posted = ixgbevf_write_posted_mbx, .check_for_msg = ixgbevf_check_for_msg_vf, .check_for_ack = ixgbevf_check_for_ack_vf, .check_for_rst = ixgbevf_check_for_rst_vf, diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c index f5d0d7969144..a1672e518d97 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.c +++ b/drivers/net/ethernet/intel/ixgbevf/vf.c @@ -13,13 +13,12 @@ static inline s32 ixgbevf_write_msg_read_ack(struct ixgbe_hw *hw, u32 *msg, u32 *retmsg, u16 size) { - struct ixgbe_mbx_info *mbx = &hw->mbx; - s32 retval = mbx->ops.write_posted(hw, msg, size); + s32 retval = ixgbevf_write_mbx(hw, msg, size); if (retval) return retval; - return mbx->ops.read_posted(hw, retmsg, size); + return ixgbevf_poll_mbx(hw, retmsg, size); } /** @@ -92,7 +91,7 @@ static s32 ixgbevf_reset_hw_vf(struct ixgbe_hw *hw) mbx->timeout = IXGBE_VF_MBX_INIT_TIMEOUT; msgbuf[0] = IXGBE_VF_RESET; - mbx->ops.write_posted(hw, msgbuf, 1); + ixgbevf_write_mbx(hw, msgbuf, 1); mdelay(10); @@ -100,7 +99,7 @@ static s32 ixgbevf_reset_hw_vf(struct ixgbe_hw *hw) * also set up the mc_filter_type which is piggy backed * on the mac address in word 3 */ - ret_val = mbx->ops.read_posted(hw, msgbuf, IXGBE_VF_PERMADDR_MSG_LEN); + ret_val = ixgbevf_poll_mbx(hw, msgbuf, IXGBE_VF_PERMADDR_MSG_LEN); if (ret_val) return ret_val; @@ -323,12 +322,12 @@ int ixgbevf_get_reta_locked(struct ixgbe_hw *hw, u32 *reta, int num_rx_queues) msgbuf[0] = IXGBE_VF_GET_RETA; - err = hw->mbx.ops.write_posted(hw, msgbuf, 1); + err = ixgbevf_write_mbx(hw, msgbuf, 1); if (err) return err; - err = hw->mbx.ops.read_posted(hw, msgbuf, dwords + 1); + err = ixgbevf_poll_mbx(hw, msgbuf, dwords + 1); if (err) return err; @@ -390,12 +389,12 @@ int ixgbevf_get_rss_key_locked(struct ixgbe_hw *hw, u8 *rss_key) } msgbuf[0] = IXGBE_VF_GET_RSS_KEY; - err = hw->mbx.ops.write_posted(hw, msgbuf, 1); + err = ixgbevf_write_mbx(hw, msgbuf, 1); if (err) return err; - err = hw->mbx.ops.read_posted(hw, msgbuf, 11); + err = ixgbevf_poll_mbx(hw, msgbuf, 11); if (err) return err; diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.h b/drivers/net/ethernet/intel/ixgbevf/vf.h index 1d8209df4162..54158dac8707 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.h +++ b/drivers/net/ethernet/intel/ixgbevf/vf.h @@ -73,10 +73,9 @@ struct ixgbe_mac_info { struct ixgbe_mbx_operations { s32 (*init_params)(struct ixgbe_hw *hw); + void (*release)(struct ixgbe_hw *hw); s32 (*read)(struct ixgbe_hw *, u32 *, u16); s32 (*write)(struct ixgbe_hw *, u32 *, u16); - s32 (*read_posted)(struct ixgbe_hw *, u32 *, u16); - s32 (*write_posted)(struct ixgbe_hw *, u32 *, u16); s32 (*check_for_msg)(struct ixgbe_hw *); s32 (*check_for_ack)(struct ixgbe_hw *); s32 (*check_for_rst)(struct ixgbe_hw *); @@ -96,7 +95,7 @@ struct ixgbe_mbx_info { struct ixgbe_mbx_stats stats; u32 timeout; u32 udelay; - u32 v2p_mailbox; + u32 vf_mailbox; u16 size; }; From 339f28964147db9cb5c45d84a9b9d86a8ceb9260 Mon Sep 17 00:00:00 2001 From: Radoslaw Tyl Date: Wed, 30 Jun 2021 10:15:32 +0200 Subject: [PATCH 0307/2048] ixgbevf: Add support for new mailbox communication between PF and VF Provide improved mailbox communication, between PF and VF, which is defined as API version 1.5. Signed-off-by: Radoslaw Tyl Tested-by: Tony Brelinski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ixgbevf/ipsec.c | 1 + drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 9 +++++++++ drivers/net/ethernet/intel/ixgbevf/mbx.h | 1 + drivers/net/ethernet/intel/ixgbevf/vf.c | 14 ++++++++++++-- 4 files changed, 23 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ipsec.c b/drivers/net/ethernet/intel/ixgbevf/ipsec.c index d9935a42f0b7..e763cee0695e 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ipsec.c +++ b/drivers/net/ethernet/intel/ixgbevf/ipsec.c @@ -623,6 +623,7 @@ void ixgbevf_init_ipsec_offload(struct ixgbevf_adapter *adapter) switch (adapter->hw.api_version) { case ixgbe_mbox_api_14: + case ixgbe_mbox_api_15: break; default: return; diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index cd93b47a61cd..b1dfbaff8b31 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -2266,6 +2266,7 @@ static void ixgbevf_negotiate_api(struct ixgbevf_adapter *adapter) { struct ixgbe_hw *hw = &adapter->hw; static const int api[] = { + ixgbe_mbox_api_15, ixgbe_mbox_api_14, ixgbe_mbox_api_13, ixgbe_mbox_api_12, @@ -2284,6 +2285,12 @@ static void ixgbevf_negotiate_api(struct ixgbevf_adapter *adapter) idx++; } + if (hw->api_version >= ixgbe_mbox_api_15) { + hw->mbx.ops.init_params(hw); + memcpy(&hw->mbx.ops, &ixgbevf_mbx_ops, + sizeof(struct ixgbe_mbx_operations)); + } + spin_unlock_bh(&adapter->mbx_lock); } @@ -2627,6 +2634,7 @@ static void ixgbevf_set_num_queues(struct ixgbevf_adapter *adapter) case ixgbe_mbox_api_12: case ixgbe_mbox_api_13: case ixgbe_mbox_api_14: + case ixgbe_mbox_api_15: if (adapter->xdp_prog && hw->mac.max_tx_queues == rss) rss = rss > 3 ? 2 : 1; @@ -4625,6 +4633,7 @@ static int ixgbevf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) case ixgbe_mbox_api_12: case ixgbe_mbox_api_13: case ixgbe_mbox_api_14: + case ixgbe_mbox_api_15: netdev->max_mtu = IXGBE_MAX_JUMBO_FRAME_SIZE - (ETH_HLEN + ETH_FCS_LEN); break; diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.h b/drivers/net/ethernet/intel/ixgbevf/mbx.h index b3b83c95babf..7346ccf014a5 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.h +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.h @@ -65,6 +65,7 @@ enum ixgbe_pfvf_api_rev { ixgbe_mbox_api_12, /* API version 1.2, linux/freebsd VF driver */ ixgbe_mbox_api_13, /* API version 1.3, linux/freebsd VF driver */ ixgbe_mbox_api_14, /* API version 1.4, linux/freebsd VF driver */ + ixgbe_mbox_api_15, /* API version 1.5, linux/freebsd VF driver */ /* This value should always be last */ ixgbe_mbox_api_unknown, /* indicates that API version is not known */ }; diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c index a1672e518d97..61d8970c6d1d 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.c +++ b/drivers/net/ethernet/intel/ixgbevf/vf.c @@ -74,6 +74,9 @@ static s32 ixgbevf_reset_hw_vf(struct ixgbe_hw *hw) /* reset the api version */ hw->api_version = ixgbe_mbox_api_10; + hw->mbx.ops.init_params(hw); + memcpy(&hw->mbx.ops, &ixgbevf_mbx_ops_legacy, + sizeof(struct ixgbe_mbx_operations)); IXGBE_WRITE_REG(hw, IXGBE_VFCTRL, IXGBE_CTRL_RST); IXGBE_WRITE_FLUSH(hw); @@ -310,6 +313,7 @@ int ixgbevf_get_reta_locked(struct ixgbe_hw *hw, u32 *reta, int num_rx_queues) * is not supported for this device type. */ switch (hw->api_version) { + case ixgbe_mbox_api_15: case ixgbe_mbox_api_14: case ixgbe_mbox_api_13: case ixgbe_mbox_api_12: @@ -378,6 +382,7 @@ int ixgbevf_get_rss_key_locked(struct ixgbe_hw *hw, u8 *rss_key) * or if the operation is not supported for this device type. */ switch (hw->api_version) { + case ixgbe_mbox_api_15: case ixgbe_mbox_api_14: case ixgbe_mbox_api_13: case ixgbe_mbox_api_12: @@ -544,8 +549,9 @@ static s32 ixgbevf_update_xcast_mode(struct ixgbe_hw *hw, int xcast_mode) if (xcast_mode == IXGBEVF_XCAST_MODE_PROMISC) return -EOPNOTSUPP; fallthrough; - case ixgbe_mbox_api_14: case ixgbe_mbox_api_13: + case ixgbe_mbox_api_14: + case ixgbe_mbox_api_15: break; default: return -EOPNOTSUPP; @@ -704,8 +710,11 @@ static s32 ixgbevf_check_mac_link_vf(struct ixgbe_hw *hw, /* if the read failed it could just be a mailbox collision, best wait * until we are called again and don't report an error */ - if (mbx->ops.read(hw, &in_msg, 1)) + if (mbx->ops.read(hw, &in_msg, 1)) { + if (hw->api_version >= ixgbe_mbox_api_15) + mac->get_link_status = false; goto out; + } if (!(in_msg & IXGBE_VT_MSGTYPE_CTS)) { /* msg is not CTS and is NACK we must have lost CTS status */ @@ -901,6 +910,7 @@ int ixgbevf_get_queues(struct ixgbe_hw *hw, unsigned int *num_tcs, case ixgbe_mbox_api_12: case ixgbe_mbox_api_13: case ixgbe_mbox_api_14: + case ixgbe_mbox_api_15: break; default: return 0; From e92af33e472cf3aa25fd14ae7760c113f5ac8c48 Mon Sep 17 00:00:00 2001 From: Alexander Lobakin Date: Wed, 17 Nov 2021 16:29:17 +0100 Subject: [PATCH 0308/2048] stmmac: fix build due to brainos in trans_start changes txq_trans_cond_update() takes netdev_tx_queue *nq, not nq->trans_start. Fixes: 5337824f4dc4 ("net: annotate accesses to queue->trans_start") Signed-off-by: Alexander Lobakin Reviewed-by: Eric Dumazet Link: https://lore.kernel.org/r/20211117152917.3739-1-alexandr.lobakin@intel.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 389d125310c1..41e9d8838f05 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -2356,7 +2356,7 @@ static bool stmmac_xdp_xmit_zc(struct stmmac_priv *priv, u32 queue, u32 budget) bool work_done = true; /* Avoids TX time-out as we are sharing with slow path */ - txq_trans_cond_update(nq->trans_start); + txq_trans_cond_update(nq); budget = min(budget, stmmac_tx_avail(priv, queue)); @@ -4657,7 +4657,7 @@ static int stmmac_xdp_xmit_back(struct stmmac_priv *priv, __netif_tx_lock(nq, cpu); /* Avoids TX time-out as we are sharing with slow path */ - txq_trans_cond_update(nq->trans_start); + txq_trans_cond_update(nq); res = stmmac_xdp_xmit_xdpf(priv, queue, xdpf, false); if (res == STMMAC_XDP_TX) From 75082e7f46809432131749f4ecea66864d0f7438 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 17 Nov 2021 02:01:30 -0800 Subject: [PATCH 0309/2048] net: add missing include in include/net/gro.h This is needed for some arches, as reported by Geert Uytterhoeven, Randy Dunlap and Stephen Rothwell Fixes: 4721031c3559 ("net: move gro definitions to include/net/gro.h") Signed-off-by: Eric Dumazet Cc: Stephen Rothwell Cc: Randy Dunlap Tested-by: Geert Uytterhoeven Link: https://lore.kernel.org/r/20211117100130.2368319-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski --- include/net/gro.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/net/gro.h b/include/net/gro.h index d0e7df691a80..9c22a010369c 100644 --- a/include/net/gro.h +++ b/include/net/gro.h @@ -6,6 +6,7 @@ #include #include #include +#include #include #include From 3ff36bffaf3545d46e7dedcd8b89e62591de246d Mon Sep 17 00:00:00 2001 From: Dave Tucker Date: Fri, 12 Nov 2021 21:17:22 +0000 Subject: [PATCH 0310/2048] bpf, docs: Change underline in btf to match style guide This changes the type of underline used to follow the guidelines in Documentation/doc-guide/sphinx.rst which also ensures that the headings are rendered at the correct level in the HTML sidebar Signed-off-by: Dave Tucker Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/981b27485cc294206480df36fca46817e2553e39.1636749493.git.dave@dtucker.co.uk --- Documentation/bpf/btf.rst | 44 +++++++++++++++++++-------------------- 1 file changed, 22 insertions(+), 22 deletions(-) diff --git a/Documentation/bpf/btf.rst b/Documentation/bpf/btf.rst index d0ec40d00c28..1ebf4c5c7ddc 100644 --- a/Documentation/bpf/btf.rst +++ b/Documentation/bpf/btf.rst @@ -3,7 +3,7 @@ BPF Type Format (BTF) ===================== 1. Introduction -*************** +=============== BTF (BPF Type Format) is the metadata format which encodes the debug info related to BPF program/map. The name BTF was used initially to describe data @@ -30,7 +30,7 @@ sections are discussed in details in :ref:`BTF_Type_String`. .. _BTF_Type_String: 2. BTF Type and String Encoding -******************************* +=============================== The file ``include/uapi/linux/btf.h`` provides high-level definition of how types/strings are encoded. @@ -57,13 +57,13 @@ little-endian target. The ``btf_header`` is designed to be extensible with generated. 2.1 String Encoding -=================== +------------------- The first string in the string section must be a null string. The rest of string table is a concatenation of other null-terminated strings. 2.2 Type Encoding -================= +----------------- The type id ``0`` is reserved for ``void`` type. The type section is parsed sequentially and type id is assigned to each recognized type starting from id @@ -504,7 +504,7 @@ valid index (starting from 0) pointing to a member or an argument. * ``type``: the type with ``btf_type_tag`` attribute 3. BTF Kernel API -***************** +================= The following bpf syscall command involves BTF: * BPF_BTF_LOAD: load a blob of BTF data into kernel @@ -547,14 +547,14 @@ The workflow typically looks like: 3.1 BPF_BTF_LOAD -================ +---------------- Load a blob of BTF data into kernel. A blob of data, described in :ref:`BTF_Type_String`, can be directly loaded into the kernel. A ``btf_fd`` is returned to a userspace. 3.2 BPF_MAP_CREATE -================== +------------------ A map can be created with ``btf_fd`` and specified key/value type id.:: @@ -581,7 +581,7 @@ automatically. .. _BPF_Prog_Load: 3.3 BPF_PROG_LOAD -================= +----------------- During prog_load, func_info and line_info can be passed to kernel with proper values for the following attributes: @@ -631,7 +631,7 @@ For line_info, the line number and column number are defined as below: #define BPF_LINE_INFO_LINE_COL(line_col) ((line_col) & 0x3ff) 3.4 BPF_{PROG,MAP}_GET_NEXT_ID -============================== +------------------------------ In kernel, every loaded program, map or btf has a unique id. The id won't change during the lifetime of a program, map, or btf. @@ -641,13 +641,13 @@ each command, to user space, for bpf program or maps, respectively, so an inspection tool can inspect all programs and maps. 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID -=============================== +------------------------------- An introspection tool cannot use id to get details about program or maps. A file descriptor needs to be obtained first for reference-counting purpose. 3.6 BPF_OBJ_GET_INFO_BY_FD -========================== +-------------------------- Once a program/map fd is acquired, an introspection tool can get the detailed information from kernel about this fd, some of which are BTF-related. For @@ -656,7 +656,7 @@ example, ``bpf_map_info`` returns ``btf_id`` and key/value type ids. bpf byte codes, and jited_line_info. 3.7 BPF_BTF_GET_FD_BY_ID -======================== +------------------------ With ``btf_id`` obtained in ``bpf_map_info`` and ``bpf_prog_info``, bpf syscall command BPF_BTF_GET_FD_BY_ID can retrieve a btf fd. Then, with @@ -668,10 +668,10 @@ tool has full btf knowledge and is able to pretty print map key/values, dump func signatures and line info, along with byte/jit codes. 4. ELF File Format Interface -**************************** +============================ 4.1 .BTF section -================ +---------------- The .BTF section contains type and string data. The format of this section is same as the one describe in :ref:`BTF_Type_String`. @@ -679,7 +679,7 @@ same as the one describe in :ref:`BTF_Type_String`. .. _BTF_Ext_Section: 4.2 .BTF.ext section -==================== +-------------------- The .BTF.ext section encodes func_info and line_info which needs loader manipulation before loading into the kernel. @@ -743,7 +743,7 @@ bpf_insn``. For ELF API, the ``insn_off`` is the byte offset from the beginning of section (``btf_ext_info_sec->sec_name_off``). 4.2 .BTF_ids section -==================== +-------------------- The .BTF_ids section encodes BTF ID values that are used within the kernel. @@ -804,10 +804,10 @@ All the BTF ID lists and sets are compiled in the .BTF_ids section and resolved during the linking phase of kernel build by ``resolve_btfids`` tool. 5. Using BTF -************ +============ 5.1 bpftool map pretty print -============================ +---------------------------- With BTF, the map key/value can be printed based on fields rather than simply raw bytes. This is especially valuable for large structure or if your data @@ -849,7 +849,7 @@ bpftool is able to pretty print like below: ] 5.2 bpftool prog dump -===================== +--------------------- The following is an example showing how func_info and line_info can help prog dump with better kernel symbol names, function prototypes and line @@ -883,7 +883,7 @@ information.:: [...] 5.3 Verifier Log -================ +---------------- The following is an example of how line_info can help debugging verification failure.:: @@ -909,7 +909,7 @@ failure.:: R2 offset is outside of the packet 6. BTF Generation -***************** +================= You need latest pahole @@ -1016,6 +1016,6 @@ format.:: .long 8206 # Line 8 Col 14 7. Testing -********** +========== Kernel bpf selftest `test_btf.c` provides extensive set of BTF-related tests. From f5b1c2ef43d79e054f471dc96996ac40bb262d8d Mon Sep 17 00:00:00 2001 From: Dave Tucker Date: Fri, 12 Nov 2021 21:17:23 +0000 Subject: [PATCH 0311/2048] bpf, docs: Rename bpf_lsm.rst to prog_lsm.rst This allows for documentation relating to BPF Program Types to be matched by the glob pattern prog_* for inclusion in a sphinx toctree Signed-off-by: Dave Tucker Signed-off-by: Daniel Borkmann Acked-by: KP Singh Link: https://lore.kernel.org/bpf/49fe0f370a2b28500c1b60f1fdb6fb7ec90de28a.1636749493.git.dave@dtucker.co.uk --- Documentation/bpf/{bpf_lsm.rst => prog_lsm.rst} | 0 MAINTAINERS | 2 +- 2 files changed, 1 insertion(+), 1 deletion(-) rename Documentation/bpf/{bpf_lsm.rst => prog_lsm.rst} (100%) diff --git a/Documentation/bpf/bpf_lsm.rst b/Documentation/bpf/prog_lsm.rst similarity index 100% rename from Documentation/bpf/bpf_lsm.rst rename to Documentation/bpf/prog_lsm.rst diff --git a/MAINTAINERS b/MAINTAINERS index 4c74516e4353..25a59950042d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3559,7 +3559,7 @@ R: Florent Revest R: Brendan Jackman L: bpf@vger.kernel.org S: Maintained -F: Documentation/bpf/bpf_lsm.rst +F: Documentation/bpf/prog_lsm.rst F: include/linux/bpf_lsm.h F: kernel/bpf/bpf_lsm.c F: security/bpf/ From 5931d9a3d0529dc803c792a10e52f0de1d0b9991 Mon Sep 17 00:00:00 2001 From: Dave Tucker Date: Fri, 12 Nov 2021 21:17:24 +0000 Subject: [PATCH 0312/2048] bpf, docs: Fix ordering of bpf documentation This commit fixes the display of the BPF documentation in the sidebar when rendered as HTML. Before this patch, the sidebar would render as follows for some sections: | BPF Documentation |- BPF Type Format (BTF) |- BPF Type Format (BTF) This was due to creating a heading in index.rst followed by a sphinx toctree, where the file referenced carries the same title as the section heading. To fix this I applied a pattern that has been established in other subfolders of Documentation: 1. Re-wrote index.rst to have a single toctree 2. Split the sections out in to their own files Additionally maps.rst and programs.rst make use of a glob pattern to include map_* or prog_* rst files in their toctree, meaning future map or program type documentation will be automatically included. Signed-off-by: Dave Tucker Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/1a1eed800e7b9dc13b458de113a489641519b0cc.1636749493.git.dave@dtucker.co.uk --- Documentation/bpf/faq.rst | 11 ++++ Documentation/bpf/helpers.rst | 7 +++ Documentation/bpf/index.rst | 97 ++++-------------------------- Documentation/bpf/libbpf/index.rst | 4 +- Documentation/bpf/maps.rst | 9 +++ Documentation/bpf/other.rst | 9 +++ Documentation/bpf/programs.rst | 9 +++ Documentation/bpf/syscall_api.rst | 11 ++++ Documentation/bpf/test_debug.rst | 9 +++ 9 files changed, 80 insertions(+), 86 deletions(-) create mode 100644 Documentation/bpf/faq.rst create mode 100644 Documentation/bpf/helpers.rst create mode 100644 Documentation/bpf/maps.rst create mode 100644 Documentation/bpf/other.rst create mode 100644 Documentation/bpf/programs.rst create mode 100644 Documentation/bpf/syscall_api.rst create mode 100644 Documentation/bpf/test_debug.rst diff --git a/Documentation/bpf/faq.rst b/Documentation/bpf/faq.rst new file mode 100644 index 000000000000..a622602ce9ad --- /dev/null +++ b/Documentation/bpf/faq.rst @@ -0,0 +1,11 @@ +================================ +Frequently asked questions (FAQ) +================================ + +Two sets of Questions and Answers (Q&A) are maintained. + +.. toctree:: + :maxdepth: 1 + + bpf_design_QA + bpf_devel_QA diff --git a/Documentation/bpf/helpers.rst b/Documentation/bpf/helpers.rst new file mode 100644 index 000000000000..c4ee0cc20dec --- /dev/null +++ b/Documentation/bpf/helpers.rst @@ -0,0 +1,7 @@ +Helper functions +================ + +* `bpf-helpers(7)`_ maintains a list of helpers available to eBPF programs. + +.. Links +.. _bpf-helpers(7): https://man7.org/linux/man-pages/man7/bpf-helpers.7.html \ No newline at end of file diff --git a/Documentation/bpf/index.rst b/Documentation/bpf/index.rst index 37f273a7e8b6..413f50101eca 100644 --- a/Documentation/bpf/index.rst +++ b/Documentation/bpf/index.rst @@ -12,97 +12,26 @@ BPF instruction-set. The Cilium project also maintains a `BPF and XDP Reference Guide`_ that goes into great technical depth about the BPF Architecture. -libbpf -====== - -Documentation/bpf/libbpf/libbpf.rst is a userspace library for loading and interacting with bpf programs. - -BPF Type Format (BTF) -===================== - .. toctree:: :maxdepth: 1 + libbpf/index btf - - -Frequently asked questions (FAQ) -================================ - -Two sets of Questions and Answers (Q&A) are maintained. - -.. toctree:: - :maxdepth: 1 - - bpf_design_QA - bpf_devel_QA - -Syscall API -=========== - -The primary info for the bpf syscall is available in the `man-pages`_ -for `bpf(2)`_. For more information about the userspace API, see -Documentation/userspace-api/ebpf/index.rst. - -Helper functions -================ - -* `bpf-helpers(7)`_ maintains a list of helpers available to eBPF programs. - - -Program types -============= - -.. toctree:: - :maxdepth: 1 - - prog_cgroup_sockopt - prog_cgroup_sysctl - prog_flow_dissector - bpf_lsm - prog_sk_lookup - - -Map types -========= - -.. toctree:: - :maxdepth: 1 - - map_cgroup_storage - - -Testing and debugging BPF -========================= - -.. toctree:: - :maxdepth: 1 - - drgn - s390 - - -Licensing -========= - -.. toctree:: - :maxdepth: 1 - + faq + syscall_api + helpers + programs + maps bpf_licensing + test_debug + other +.. only:: subproject and html -Other -===== + Indices + ======= -.. toctree:: - :maxdepth: 1 - - ringbuf - llvm_reloc + * :ref:`genindex` .. Links: -.. _networking-filter: ../networking/filter.rst -.. _man-pages: https://www.kernel.org/doc/man-pages/ -.. _bpf(2): https://man7.org/linux/man-pages/man2/bpf.2.html -.. _bpf-helpers(7): https://man7.org/linux/man-pages/man7/bpf-helpers.7.html -.. _BPF and XDP Reference Guide: https://docs.cilium.io/en/latest/bpf/ +.. _BPF and XDP Reference Guide: https://docs.cilium.io/en/latest/bpf/ \ No newline at end of file diff --git a/Documentation/bpf/libbpf/index.rst b/Documentation/bpf/libbpf/index.rst index 4f8adfc3ab83..4e8c656b539a 100644 --- a/Documentation/bpf/libbpf/index.rst +++ b/Documentation/bpf/libbpf/index.rst @@ -3,8 +3,6 @@ libbpf ====== -For API documentation see the `versioned API documentation site `_. - .. toctree:: :maxdepth: 1 @@ -14,6 +12,8 @@ For API documentation see the `versioned API documentation site `_. + All general BPF questions, including kernel functionality, libbpf APIs and their application, should be sent to bpf@vger.kernel.org mailing list. You can `subscribe `_ to the diff --git a/Documentation/bpf/maps.rst b/Documentation/bpf/maps.rst new file mode 100644 index 000000000000..2084b0e7cde8 --- /dev/null +++ b/Documentation/bpf/maps.rst @@ -0,0 +1,9 @@ +========= +Map Types +========= + +.. toctree:: + :maxdepth: 1 + :glob: + + map_* \ No newline at end of file diff --git a/Documentation/bpf/other.rst b/Documentation/bpf/other.rst new file mode 100644 index 000000000000..3d61963403b4 --- /dev/null +++ b/Documentation/bpf/other.rst @@ -0,0 +1,9 @@ +===== +Other +===== + +.. toctree:: + :maxdepth: 1 + + ringbuf + llvm_reloc \ No newline at end of file diff --git a/Documentation/bpf/programs.rst b/Documentation/bpf/programs.rst new file mode 100644 index 000000000000..620eb667ac7a --- /dev/null +++ b/Documentation/bpf/programs.rst @@ -0,0 +1,9 @@ +============= +Program Types +============= + +.. toctree:: + :maxdepth: 1 + :glob: + + prog_* diff --git a/Documentation/bpf/syscall_api.rst b/Documentation/bpf/syscall_api.rst new file mode 100644 index 000000000000..f0a1dff087ad --- /dev/null +++ b/Documentation/bpf/syscall_api.rst @@ -0,0 +1,11 @@ +=========== +Syscall API +=========== + +The primary info for the bpf syscall is available in the `man-pages`_ +for `bpf(2)`_. For more information about the userspace API, see +Documentation/userspace-api/ebpf/index.rst. + +.. Links: +.. _man-pages: https://www.kernel.org/doc/man-pages/ +.. _bpf(2): https://man7.org/linux/man-pages/man2/bpf.2.html \ No newline at end of file diff --git a/Documentation/bpf/test_debug.rst b/Documentation/bpf/test_debug.rst new file mode 100644 index 000000000000..ebf0caceb6a6 --- /dev/null +++ b/Documentation/bpf/test_debug.rst @@ -0,0 +1,9 @@ +========================= +Testing and debugging BPF +========================= + +.. toctree:: + :maxdepth: 1 + + drgn + s390 From dd7f091fd22b1dce6c20e8f7769aa068ed88ac6d Mon Sep 17 00:00:00 2001 From: Tirthendu Sarkar Date: Wed, 17 Nov 2021 18:06:13 +0530 Subject: [PATCH 0313/2048] selftests/bpf: Fix xdpxceiver failures for no hugepages xsk_configure_umem() needs hugepages to work in unaligned mode. So when hugepages are not configured, 'unaligned' tests should be skipped which is determined by the helper function hugepages_present(). This function erroneously returns true with MAP_NORESERVE flag even when no hugepages are configured. The removal of this flag fixes the issue. The test TEST_TYPE_UNALIGNED_INV_DESC also needs to be skipped when there are no hugepages. However, this was not skipped as there was no check for presence of hugepages and hence was failing. The check to skip the test has now been added. Fixes: a4ba98dd0c69 (selftests: xsk: Add test for unaligned mode) Signed-off-by: Tirthendu Sarkar Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211117123613.22288-1-tirthendu.sarkar@intel.com --- tools/testing/selftests/bpf/xdpxceiver.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/xdpxceiver.c b/tools/testing/selftests/bpf/xdpxceiver.c index fe7f423b8c3f..040164c7efc1 100644 --- a/tools/testing/selftests/bpf/xdpxceiver.c +++ b/tools/testing/selftests/bpf/xdpxceiver.c @@ -1217,7 +1217,7 @@ static bool hugepages_present(struct ifobject *ifobject) void *bufs; bufs = mmap(NULL, mmap_sz, PROT_READ | PROT_WRITE, - MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE | MAP_HUGETLB, -1, 0); + MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB, -1, 0); if (bufs == MAP_FAILED) return false; @@ -1364,6 +1364,10 @@ static void run_pkt_test(struct test_spec *test, enum test_mode mode, enum test_ testapp_invalid_desc(test); break; case TEST_TYPE_UNALIGNED_INV_DESC: + if (!hugepages_present(test->ifobj_tx)) { + ksft_test_result_skip("No 2M huge pages present.\n"); + return; + } test_spec_set_name(test, "UNALIGNED_INV_DESC"); test->ifobj_tx->umem->unaligned_mode = true; test->ifobj_rx->umem->unaligned_mode = true; From 8ff978b8b222bc9d51dd109a46b51026336c95d8 Mon Sep 17 00:00:00 2001 From: Riccardo Paolo Bestetti Date: Wed, 17 Nov 2021 10:00:11 +0100 Subject: [PATCH 0314/2048] ipv4/raw: support binding to nonlocal addresses Add support to inet v4 raw sockets for binding to nonlocal addresses through the IP_FREEBIND and IP_TRANSPARENT socket options, as well as the ipv4.ip_nonlocal_bind kernel parameter. Add helper function to inet_sock.h to check for bind address validity on the base of the address type and whether nonlocal address are enabled for the socket via any of the sockopts/sysctl, deduplicating checks in ipv4/ping.c, ipv4/af_inet.c, ipv6/af_inet6.c (for mapped v4->v6 addresses), and ipv4/raw.c. Add test cases with IP[V6]_FREEBIND verifying that both v4 and v6 raw sockets support binding to nonlocal addresses after the change. Add necessary support for the test cases to nettest. Signed-off-by: Riccardo Paolo Bestetti Reviewed-by: David Ahern Link: https://lore.kernel.org/r/20211117090010.125393-1-pbl@bestov.io Signed-off-by: Jakub Kicinski --- include/net/inet_sock.h | 12 +++++++ net/ipv4/af_inet.c | 7 ++-- net/ipv4/ping.c | 12 +++---- net/ipv4/raw.c | 13 ++++---- net/ipv6/af_inet6.c | 7 ++-- tools/testing/selftests/net/fcnal-test.sh | 40 +++++++++++++++++++++++ tools/testing/selftests/net/nettest.c | 33 ++++++++++++++++++- 7 files changed, 99 insertions(+), 25 deletions(-) diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h index 9e1111f5915b..234d70ae5f4c 100644 --- a/include/net/inet_sock.h +++ b/include/net/inet_sock.h @@ -372,4 +372,16 @@ static inline bool inet_can_nonlocal_bind(struct net *net, inet->freebind || inet->transparent; } +static inline bool inet_addr_valid_or_nonlocal(struct net *net, + struct inet_sock *inet, + __be32 addr, + int addr_type) +{ + return inet_can_nonlocal_bind(net, inet) || + addr == htonl(INADDR_ANY) || + addr_type == RTN_LOCAL || + addr_type == RTN_MULTICAST || + addr_type == RTN_BROADCAST; +} + #endif /* _INET_SOCK_H */ diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 6d52b6491255..c66b0563a267 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -489,11 +489,8 @@ int __inet_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, * is temporarily down) */ err = -EADDRNOTAVAIL; - if (!inet_can_nonlocal_bind(net, inet) && - addr->sin_addr.s_addr != htonl(INADDR_ANY) && - chk_addr_ret != RTN_LOCAL && - chk_addr_ret != RTN_MULTICAST && - chk_addr_ret != RTN_BROADCAST) + if (!inet_addr_valid_or_nonlocal(net, inet, addr->sin_addr.s_addr, + chk_addr_ret)) goto out; snum = ntohs(addr->sin_port); diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 1e44a43acfe2..e540b0dcf085 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -311,15 +311,11 @@ static int ping_check_bind_addr(struct sock *sk, struct inet_sock *isk, pr_debug("ping_check_bind_addr(sk=%p,addr=%pI4,port=%d)\n", sk, &addr->sin_addr.s_addr, ntohs(addr->sin_port)); - if (addr->sin_addr.s_addr == htonl(INADDR_ANY)) - chk_addr_ret = RTN_LOCAL; - else - chk_addr_ret = inet_addr_type(net, addr->sin_addr.s_addr); + chk_addr_ret = inet_addr_type(net, addr->sin_addr.s_addr); - if ((!inet_can_nonlocal_bind(net, isk) && - chk_addr_ret != RTN_LOCAL) || - chk_addr_ret == RTN_MULTICAST || - chk_addr_ret == RTN_BROADCAST) + if (!inet_addr_valid_or_nonlocal(net, inet_sk(sk), + addr->sin_addr.s_addr, + chk_addr_ret)) return -EADDRNOTAVAIL; #if IS_ENABLED(CONFIG_IPV6) diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 3910bb08a986..a53f256bf9d3 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -717,6 +717,7 @@ static int raw_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) { struct inet_sock *inet = inet_sk(sk); struct sockaddr_in *addr = (struct sockaddr_in *) uaddr; + struct net *net = sock_net(sk); u32 tb_id = RT_TABLE_LOCAL; int ret = -EINVAL; int chk_addr_ret; @@ -725,16 +726,16 @@ static int raw_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) goto out; if (sk->sk_bound_dev_if) - tb_id = l3mdev_fib_table_by_index(sock_net(sk), - sk->sk_bound_dev_if) ? : tb_id; + tb_id = l3mdev_fib_table_by_index(net, + sk->sk_bound_dev_if) ? : tb_id; - chk_addr_ret = inet_addr_type_table(sock_net(sk), addr->sin_addr.s_addr, - tb_id); + chk_addr_ret = inet_addr_type_table(net, addr->sin_addr.s_addr, tb_id); ret = -EADDRNOTAVAIL; - if (addr->sin_addr.s_addr && chk_addr_ret != RTN_LOCAL && - chk_addr_ret != RTN_MULTICAST && chk_addr_ret != RTN_BROADCAST) + if (!inet_addr_valid_or_nonlocal(net, inet, addr->sin_addr.s_addr, + chk_addr_ret)) goto out; + inet->inet_rcv_saddr = inet->inet_saddr = addr->sin_addr.s_addr; if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST) inet->inet_saddr = 0; /* Use device */ diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 0c4da163535a..1f28c9820c2e 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -337,11 +337,8 @@ static int __inet6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, chk_addr_ret = inet_addr_type_dev_table(net, dev, v4addr); rcu_read_unlock(); - if (!inet_can_nonlocal_bind(net, inet) && - v4addr != htonl(INADDR_ANY) && - chk_addr_ret != RTN_LOCAL && - chk_addr_ret != RTN_MULTICAST && - chk_addr_ret != RTN_BROADCAST) { + if (!inet_addr_valid_or_nonlocal(net, inet, v4addr, + chk_addr_ret)) { err = -EADDRNOTAVAIL; goto out; } diff --git a/tools/testing/selftests/net/fcnal-test.sh b/tools/testing/selftests/net/fcnal-test.sh index 3313566ce906..7caa4f0e067d 100755 --- a/tools/testing/selftests/net/fcnal-test.sh +++ b/tools/testing/selftests/net/fcnal-test.sh @@ -66,6 +66,10 @@ NSB_LO_IP=172.16.2.2 NSA_LO_IP6=2001:db8:2::1 NSB_LO_IP6=2001:db8:2::2 +# non-local addresses for freebind tests +NL_IP=172.17.1.1 +NL_IP6=2001:db8:4::1 + MD5_PW=abc123 MD5_WRONG_PW=abc1234 @@ -316,6 +320,9 @@ addr2str() ${NSB_LO_IP6}) echo "ns-B loopback IPv6";; ${NSB_LINKIP6}|${NSB_LINKIP6}%*) echo "ns-B IPv6 LLA";; + ${NL_IP}) echo "nonlocal IP";; + ${NL_IP6}) echo "nonlocal IPv6";; + ${VRF_IP}) echo "VRF IP";; ${VRF_IP6}) echo "VRF IPv6";; @@ -1767,6 +1774,14 @@ ipv4_addr_bind_novrf() log_test_addr ${a} $? 0 "Raw socket bind to local address after device bind" done + # + # raw socket with nonlocal bind + # + a=${NL_IP} + log_start + run_cmd nettest -s -R -P icmp -f -l ${a} -I ${NSA_DEV} -b + log_test_addr ${a} $? 0 "Raw socket bind to nonlocal address after device bind" + # # tcp sockets # @@ -1815,6 +1830,14 @@ ipv4_addr_bind_vrf() run_cmd nettest -s -R -P icmp -l ${a} -I ${VRF} -b log_test_addr ${a} $? 1 "Raw socket bind to out of scope address after VRF bind" + # + # raw socket with nonlocal bind + # + a=${NL_IP} + log_start + run_cmd nettest -s -R -P icmp -f -l ${a} -I ${VRF} -b + log_test_addr ${a} $? 0 "Raw socket bind to nonlocal address after VRF bind" + # # tcp sockets # @@ -1965,6 +1988,7 @@ ipv4_rt() a=${NSA_IP} log_start + run_cmd nettest ${varg} -s & sleep 1 run_cmd nettest ${varg} -d ${NSA_DEV} -r ${a} & @@ -3402,6 +3426,14 @@ ipv6_addr_bind_novrf() log_test_addr ${a} $? 0 "Raw socket bind to local address after device bind" done + # + # raw socket with nonlocal bind + # + a=${NL_IP6} + log_start + run_cmd nettest -6 -s -R -P icmp -f -l ${a} -I ${NSA_DEV} -b + log_test_addr ${a} $? 0 "Raw socket bind to nonlocal address" + # # tcp sockets # @@ -3443,6 +3475,14 @@ ipv6_addr_bind_vrf() run_cmd nettest -6 -s -R -P ipv6-icmp -l ${a} -I ${VRF} -b log_test_addr ${a} $? 1 "Raw socket bind to invalid local address after vrf bind" + # + # raw socket with nonlocal bind + # + a=${NL_IP6} + log_start + run_cmd nettest -6 -s -R -P icmp -f -l ${a} -I ${VRF} -b + log_test_addr ${a} $? 0 "Raw socket bind to nonlocal address after VRF bind" + # # tcp sockets # diff --git a/tools/testing/selftests/net/nettest.c b/tools/testing/selftests/net/nettest.c index b599003eb5ba..d9a6fd2cd9d3 100644 --- a/tools/testing/selftests/net/nettest.c +++ b/tools/testing/selftests/net/nettest.c @@ -85,6 +85,7 @@ struct sock_args { int version; /* AF_INET/AF_INET6 */ int use_setsockopt; + int use_freebind; int use_cmsg; const char *dev; const char *server_dev; @@ -514,6 +515,29 @@ static int set_membership(int sd, uint32_t grp, uint32_t addr, int ifindex) return 0; } +static int set_freebind(int sd, int version) +{ + unsigned int one = 1; + int rc = 0; + + switch (version) { + case AF_INET: + if (setsockopt(sd, SOL_IP, IP_FREEBIND, &one, sizeof(one))) { + log_err_errno("setsockopt(IP_FREEBIND)"); + rc = -1; + } + break; + case AF_INET6: + if (setsockopt(sd, SOL_IPV6, IPV6_FREEBIND, &one, sizeof(one))) { + log_err_errno("setsockopt(IPV6_FREEBIND"); + rc = -1; + } + break; + } + + return rc; +} + static int set_broadcast(int sd) { unsigned int one = 1; @@ -1419,6 +1443,9 @@ static int lsock_init(struct sock_args *args) set_unicast_if(sd, args->ifindex, args->version)) goto err; + if (args->use_freebind && set_freebind(sd, args->version)) + goto err; + if (bind_socket(sd, args)) goto err; @@ -1827,7 +1854,7 @@ static int ipc_parent(int cpid, int fd, struct sock_args *args) return client_status; } -#define GETOPT_STR "sr:l:c:p:t:g:P:DRn:M:X:m:d:I:BN:O:SCi6xL:0:1:2:3:Fbq" +#define GETOPT_STR "sr:l:c:p:t:g:P:DRn:M:X:m:d:I:BN:O:SCi6xL:0:1:2:3:Fbqf" #define OPT_FORCE_BIND_KEY_IFINDEX 1001 #define OPT_NO_BIND_KEY_IFINDEX 1002 @@ -1864,6 +1891,7 @@ static void print_usage(char *prog) " -I dev bind socket to given device name - server mode\n" " -S use setsockopt (IP_UNICAST_IF or IP_MULTICAST_IF)\n" " to set device binding\n" + " -f bind socket with the IP[V6]_FREEBIND option\n" " -C use cmsg and IP_PKTINFO to specify device binding\n" "\n" " -L len send random message of given length\n" @@ -1999,6 +2027,9 @@ int main(int argc, char *argv[]) case 'S': args.use_setsockopt = 1; break; + case 'f': + args.use_freebind = 1; + break; case 'C': args.use_cmsg = 1; break; From 6c950ca7c11c17e025326713b7668fd1fb2b996c Mon Sep 17 00:00:00 2001 From: Bhupesh Sharma Date: Wed, 17 Nov 2021 16:35:38 +0530 Subject: [PATCH 0315/2048] net: stmmac: dwmac-qcom-ethqos: add platform level clocks management Split clocks settings from init callback into clks_config callback, which could support platform level clock management. Cc: David S. Miller Signed-off-by: Bhupesh Sharma Signed-off-by: David S. Miller --- .../stmicro/stmmac/dwmac-qcom-ethqos.c | 26 ++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-qcom-ethqos.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-qcom-ethqos.c index 5c74b6279d69..8fea48e477e6 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-qcom-ethqos.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-qcom-ethqos.c @@ -447,6 +447,24 @@ static void ethqos_fix_mac_speed(void *priv, unsigned int speed) ethqos_configure(ethqos); } +static int ethqos_clks_config(void *priv, bool enabled) +{ + struct qcom_ethqos *ethqos = priv; + int ret = 0; + + if (enabled) { + ret = clk_prepare_enable(ethqos->rgmii_clk); + if (ret) { + dev_err(ðqos->pdev->dev, "rgmii_clk enable failed\n"); + return ret; + } + } else { + clk_disable_unprepare(ethqos->rgmii_clk); + } + + return ret; +} + static int qcom_ethqos_probe(struct platform_device *pdev) { struct device_node *np = pdev->dev.of_node; @@ -466,6 +484,8 @@ static int qcom_ethqos_probe(struct platform_device *pdev) return PTR_ERR(plat_dat); } + plat_dat->clks_config = ethqos_clks_config; + ethqos = devm_kzalloc(&pdev->dev, sizeof(*ethqos), GFP_KERNEL); if (!ethqos) { ret = -ENOMEM; @@ -489,7 +509,7 @@ static int qcom_ethqos_probe(struct platform_device *pdev) goto err_mem; } - ret = clk_prepare_enable(ethqos->rgmii_clk); + ret = ethqos_clks_config(ethqos, true); if (ret) goto err_mem; @@ -512,7 +532,7 @@ static int qcom_ethqos_probe(struct platform_device *pdev) return ret; err_clk: - clk_disable_unprepare(ethqos->rgmii_clk); + ethqos_clks_config(ethqos, false); err_mem: stmmac_remove_config_dt(pdev, plat_dat); @@ -530,7 +550,7 @@ static int qcom_ethqos_remove(struct platform_device *pdev) return -ENODEV; ret = stmmac_pltfr_remove(pdev); - clk_disable_unprepare(ethqos->rgmii_clk); + ethqos_clks_config(ethqos, false); return ret; } From 680e9d2cd4bf0d2a46e578aed5b330f946bd7620 Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 17 Nov 2021 16:46:20 +0000 Subject: [PATCH 0316/2048] net: ag71xx: populate supported_interfaces member Populate the phy_interface_t bitmap for the Atheros ag71xx driver with interfaces modes supported by the MAC. Signed-off-by: Russell King Signed-off-by: David S. Miller --- drivers/net/ethernet/atheros/ag71xx.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index e4f30bb7498f..d21293ecdb9a 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -1178,6 +1178,32 @@ static int ag71xx_phylink_setup(struct ag71xx *ag) ag->phylink_config.dev = &ag->ndev->dev; ag->phylink_config.type = PHYLINK_NETDEV; + if ((ag71xx_is(ag, AR9330) && ag->mac_idx == 0) || + ag71xx_is(ag, AR9340) || + ag71xx_is(ag, QCA9530) || + (ag71xx_is(ag, QCA9550) && ag->mac_idx == 1)) + __set_bit(PHY_INTERFACE_MODE_MII, + ag->phylink_config.supported_interfaces); + + if ((ag71xx_is(ag, AR9330) && ag->mac_idx == 1) || + (ag71xx_is(ag, AR9340) && ag->mac_idx == 1) || + (ag71xx_is(ag, QCA9530) && ag->mac_idx == 1)) + __set_bit(PHY_INTERFACE_MODE_GMII, + ag->phylink_config.supported_interfaces); + + if (ag71xx_is(ag, QCA9550) && ag->mac_idx == 0) + __set_bit(PHY_INTERFACE_MODE_SGMII, + ag->phylink_config.supported_interfaces); + + if (ag71xx_is(ag, AR9340) && ag->mac_idx == 0) + __set_bit(PHY_INTERFACE_MODE_RMII, + ag->phylink_config.supported_interfaces); + + if ((ag71xx_is(ag, AR9340) && ag->mac_idx == 0) || + (ag71xx_is(ag, QCA9550) && ag->mac_idx == 1)) + __set_bit(PHY_INTERFACE_MODE_RGMII, + ag->phylink_config.supported_interfaces); + phylink = phylink_create(&ag->phylink_config, ag->pdev->dev.fwnode, ag->phy_if_mode, &ag71xx_phylink_mac_ops); if (IS_ERR(phylink)) From 5e20a8aa48a0c32dad6165b26d5030a6131a25ed Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Wed, 17 Nov 2021 16:46:25 +0000 Subject: [PATCH 0317/2048] net: ag71xx: remove interface checks in ag71xx_mac_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode, nor handle PHY_INTERFACE_MODE_NA in the validation function. Remove these to simplify the implementation. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/atheros/ag71xx.c | 41 +-------------------------- 1 file changed, 1 insertion(+), 40 deletions(-) diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index d21293ecdb9a..d3e7af2db3b9 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -1028,42 +1028,8 @@ static void ag71xx_mac_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct ag71xx *ag = netdev_priv(to_net_dev(config->dev)); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - switch (state->interface) { - case PHY_INTERFACE_MODE_NA: - break; - case PHY_INTERFACE_MODE_MII: - if ((ag71xx_is(ag, AR9330) && ag->mac_idx == 0) || - ag71xx_is(ag, AR9340) || - ag71xx_is(ag, QCA9530) || - (ag71xx_is(ag, QCA9550) && ag->mac_idx == 1)) - break; - goto unsupported; - case PHY_INTERFACE_MODE_GMII: - if ((ag71xx_is(ag, AR9330) && ag->mac_idx == 1) || - (ag71xx_is(ag, AR9340) && ag->mac_idx == 1) || - (ag71xx_is(ag, QCA9530) && ag->mac_idx == 1)) - break; - goto unsupported; - case PHY_INTERFACE_MODE_SGMII: - if (ag71xx_is(ag, QCA9550) && ag->mac_idx == 0) - break; - goto unsupported; - case PHY_INTERFACE_MODE_RMII: - if (ag71xx_is(ag, AR9340) && ag->mac_idx == 0) - break; - goto unsupported; - case PHY_INTERFACE_MODE_RGMII: - if ((ag71xx_is(ag, AR9340) && ag->mac_idx == 0) || - (ag71xx_is(ag, QCA9550) && ag->mac_idx == 1)) - break; - goto unsupported; - default: - goto unsupported; - } - phylink_set(mask, MII); phylink_set(mask, Pause); @@ -1074,8 +1040,7 @@ static void ag71xx_mac_validate(struct phylink_config *config, phylink_set(mask, 100baseT_Half); phylink_set(mask, 100baseT_Full); - if (state->interface == PHY_INTERFACE_MODE_NA || - state->interface == PHY_INTERFACE_MODE_SGMII || + if (state->interface == PHY_INTERFACE_MODE_SGMII || state->interface == PHY_INTERFACE_MODE_RGMII || state->interface == PHY_INTERFACE_MODE_GMII) { phylink_set(mask, 1000baseT_Full); @@ -1084,10 +1049,6 @@ static void ag71xx_mac_validate(struct phylink_config *config, linkmode_and(supported, supported, mask); linkmode_and(state->advertising, state->advertising, mask); - - return; -unsupported: - linkmode_zero(supported); } static void ag71xx_mac_pcs_get_state(struct phylink_config *config, From c8fa4bac30e1ca4b0213f85fd868d0f14545d4c7 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Wed, 17 Nov 2021 16:46:31 +0000 Subject: [PATCH 0318/2048] net: ag71xx: use phylink_generic_validate() ag71xx apparently only supports MII port type, which makes it different from other implementations. However, Oleksij says there is no special reason for this. Convert the driver to use phylink_generic_validate(), which will allow all ethtool port linkmodes instead of only MII, giving the driver consistent behaviour with other drivers. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/atheros/ag71xx.c | 31 +++------------------------ 1 file changed, 3 insertions(+), 28 deletions(-) diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index d3e7af2db3b9..ff924f06581e 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -1024,33 +1024,6 @@ static void ag71xx_mac_config(struct phylink_config *config, unsigned int mode, ag71xx_wr(ag, AG71XX_REG_FIFO_CFG3, ag->fifodata[2]); } -static void ag71xx_mac_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set(mask, MII); - - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - phylink_set(mask, Autoneg); - phylink_set(mask, 10baseT_Half); - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Half); - phylink_set(mask, 100baseT_Full); - - if (state->interface == PHY_INTERFACE_MODE_SGMII || - state->interface == PHY_INTERFACE_MODE_RGMII || - state->interface == PHY_INTERFACE_MODE_GMII) { - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); -} - static void ag71xx_mac_pcs_get_state(struct phylink_config *config, struct phylink_link_state *state) { @@ -1124,7 +1097,7 @@ static void ag71xx_mac_link_up(struct phylink_config *config, } static const struct phylink_mac_ops ag71xx_phylink_mac_ops = { - .validate = ag71xx_mac_validate, + .validate = phylink_generic_validate, .mac_pcs_get_state = ag71xx_mac_pcs_get_state, .mac_an_restart = ag71xx_mac_an_restart, .mac_config = ag71xx_mac_config, @@ -1138,6 +1111,8 @@ static int ag71xx_phylink_setup(struct ag71xx *ag) ag->phylink_config.dev = &ag->ndev->dev; ag->phylink_config.type = PHYLINK_NETDEV; + ag->phylink_config.mac_capabilities = MAC_SYM_PAUSE | MAC_ASYM_PAUSE | + MAC_10 | MAC_100 | MAC_1000FD; if ((ag71xx_is(ag, AR9330) && ag->mac_idx == 0) || ag71xx_is(ag, AR9340) || From 15d0b14cec1c3ee3739d15fdcec3ebe1b847dd7c Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 17 Nov 2021 17:24:02 +0000 Subject: [PATCH 0319/2048] net: dpaa2-mac: populate supported_interfaces member Populate the phy interface mode bitmap for the Freescale DPAA2 driver with interfaces modes supported by the MAC. Signed-off-by: Russell King Signed-off-by: David S. Miller --- .../net/ethernet/freescale/dpaa2/dpaa2-mac.c | 21 +++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c index ef8f0a055024..176ce0a03716 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c @@ -336,9 +336,30 @@ int dpaa2_mac_connect(struct dpaa2_mac *mac) return err; } + memset(&mac->phylink_config, 0, sizeof(mac->phylink_config)); mac->phylink_config.dev = &net_dev->dev; mac->phylink_config.type = PHYLINK_NETDEV; + /* We support the current interface mode, and if we have a PCS + * similar interface modes that do not require the PLLs to be + * reconfigured. + */ + __set_bit(mac->if_mode, mac->phylink_config.supported_interfaces); + if (mac->pcs) { + switch (mac->if_mode) { + case PHY_INTERFACE_MODE_1000BASEX: + case PHY_INTERFACE_MODE_SGMII: + __set_bit(PHY_INTERFACE_MODE_1000BASEX, + mac->phylink_config.supported_interfaces); + __set_bit(PHY_INTERFACE_MODE_SGMII, + mac->phylink_config.supported_interfaces); + break; + + default: + break; + } + } + phylink = phylink_create(&mac->phylink_config, dpmac_node, mac->if_mode, &dpaa2_mac_phylink_ops); From 22de481d23c90727510c9c4a37b89deb34e6ba3a Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Wed, 17 Nov 2021 17:24:07 +0000 Subject: [PATCH 0320/2048] net: dpaa2-mac: remove interface checks in dpaa2_mac_validate() As phylink checks the interface mode against the supported_interfaces bitmap, we no longer need to validate the interface mode, nor handle PHY_INTERFACE_MODE_NA in the validation function. Remove these to simplify the implementation. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/freescale/dpaa2/dpaa2-mac.c | 35 ------------------- 1 file changed, 35 deletions(-) diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c index 176ce0a03716..bcc7fe127d91 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c @@ -90,53 +90,18 @@ static int dpaa2_mac_get_if_mode(struct fwnode_handle *dpmac_node, return err; } -static bool dpaa2_mac_phy_mode_mismatch(struct dpaa2_mac *mac, - phy_interface_t interface) -{ - switch (interface) { - /* We can switch between SGMII and 1000BASE-X at runtime with - * pcs-lynx - */ - case PHY_INTERFACE_MODE_SGMII: - case PHY_INTERFACE_MODE_1000BASEX: - if (mac->pcs && - (mac->if_mode == PHY_INTERFACE_MODE_SGMII || - mac->if_mode == PHY_INTERFACE_MODE_1000BASEX)) - return false; - return interface != mac->if_mode; - - case PHY_INTERFACE_MODE_10GBASER: - case PHY_INTERFACE_MODE_USXGMII: - case PHY_INTERFACE_MODE_QSGMII: - case PHY_INTERFACE_MODE_RGMII: - case PHY_INTERFACE_MODE_RGMII_ID: - case PHY_INTERFACE_MODE_RGMII_RXID: - case PHY_INTERFACE_MODE_RGMII_TXID: - return (interface != mac->if_mode); - default: - return true; - } -} - static void dpaa2_mac_validate(struct phylink_config *config, unsigned long *supported, struct phylink_link_state *state) { - struct dpaa2_mac *mac = phylink_to_dpaa2_mac(config); __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - if (state->interface != PHY_INTERFACE_MODE_NA && - dpaa2_mac_phy_mode_mismatch(mac, state->interface)) { - goto empty_set; - } - phylink_set_port_modes(mask); phylink_set(mask, Autoneg); phylink_set(mask, Pause); phylink_set(mask, Asym_Pause); switch (state->interface) { - case PHY_INTERFACE_MODE_NA: case PHY_INTERFACE_MODE_10GBASER: case PHY_INTERFACE_MODE_USXGMII: phylink_set_10g_modes(mask); From 6d386f6613269f53b9d21664e582d22e8cda7505 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Wed, 17 Nov 2021 17:24:13 +0000 Subject: [PATCH 0321/2048] net: dpaa2-mac: use phylink_generic_validate() DPAA2 has no special behaviour in its validation implementation, so can be switched to phylink_generic_validate(). Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- .../net/ethernet/freescale/dpaa2/dpaa2-mac.c | 53 ++----------------- 1 file changed, 5 insertions(+), 48 deletions(-) diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c index bcc7fe127d91..34b2a73c347f 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c @@ -90,53 +90,6 @@ static int dpaa2_mac_get_if_mode(struct fwnode_handle *dpmac_node, return err; } -static void dpaa2_mac_validate(struct phylink_config *config, - unsigned long *supported, - struct phylink_link_state *state) -{ - __ETHTOOL_DECLARE_LINK_MODE_MASK(mask) = { 0, }; - - phylink_set_port_modes(mask); - phylink_set(mask, Autoneg); - phylink_set(mask, Pause); - phylink_set(mask, Asym_Pause); - - switch (state->interface) { - case PHY_INTERFACE_MODE_10GBASER: - case PHY_INTERFACE_MODE_USXGMII: - phylink_set_10g_modes(mask); - if (state->interface == PHY_INTERFACE_MODE_10GBASER) - break; - phylink_set(mask, 5000baseT_Full); - phylink_set(mask, 2500baseT_Full); - fallthrough; - case PHY_INTERFACE_MODE_SGMII: - case PHY_INTERFACE_MODE_QSGMII: - case PHY_INTERFACE_MODE_1000BASEX: - case PHY_INTERFACE_MODE_RGMII: - case PHY_INTERFACE_MODE_RGMII_ID: - case PHY_INTERFACE_MODE_RGMII_RXID: - case PHY_INTERFACE_MODE_RGMII_TXID: - phylink_set(mask, 1000baseX_Full); - phylink_set(mask, 1000baseT_Full); - if (state->interface == PHY_INTERFACE_MODE_1000BASEX) - break; - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 10baseT_Full); - break; - default: - goto empty_set; - } - - linkmode_and(supported, supported, mask); - linkmode_and(state->advertising, state->advertising, mask); - - return; - -empty_set: - linkmode_zero(supported); -} - static void dpaa2_mac_config(struct phylink_config *config, unsigned int mode, const struct phylink_link_state *state) { @@ -208,7 +161,7 @@ static void dpaa2_mac_link_down(struct phylink_config *config, } static const struct phylink_mac_ops dpaa2_mac_phylink_ops = { - .validate = dpaa2_mac_validate, + .validate = phylink_generic_validate, .mac_config = dpaa2_mac_config, .mac_link_up = dpaa2_mac_link_up, .mac_link_down = dpaa2_mac_link_down, @@ -305,6 +258,10 @@ int dpaa2_mac_connect(struct dpaa2_mac *mac) mac->phylink_config.dev = &net_dev->dev; mac->phylink_config.type = PHYLINK_NETDEV; + mac->phylink_config.mac_capabilities = MAC_SYM_PAUSE | MAC_ASYM_PAUSE | + MAC_10FD | MAC_100FD | MAC_1000FD | MAC_2500FD | MAC_5000FD | + MAC_10000FD; + /* We support the current interface mode, and if we have a PCS * similar interface modes that do not require the PLLs to be * reconfigured. From 867ae8a7993b26fc742703f748663cc2ac421db7 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Wed, 17 Nov 2021 09:36:29 -0800 Subject: [PATCH 0322/2048] net: mdio: Replaced BUG_ON() with WARN() Killing the kernel because a certain MDIO bus object is not in the desired state at various points in the registration or unregistration paths is excessive and is not helping in troubleshooting or fixing issues. Replace the BUG_ON() with WARN() and print out the MDIO bus name to facilitate debugging. Signed-off-by: Florian Fainelli Reviewed-by: Andrew Lunn Signed-off-by: David S. Miller --- drivers/net/phy/mdio_bus.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c index c204067f1890..9b6f2df07211 100644 --- a/drivers/net/phy/mdio_bus.c +++ b/drivers/net/phy/mdio_bus.c @@ -176,9 +176,11 @@ static void mdiobus_release(struct device *d) { struct mii_bus *bus = to_mii_bus(d); - BUG_ON(bus->state != MDIOBUS_RELEASED && - /* for compatibility with error handling in drivers */ - bus->state != MDIOBUS_ALLOCATED); + WARN(bus->state != MDIOBUS_RELEASED && + /* for compatibility with error handling in drivers */ + bus->state != MDIOBUS_ALLOCATED, + "%s: not in RELEASED or ALLOCATED state\n", + bus->id); kfree(bus); } @@ -529,8 +531,9 @@ int __mdiobus_register(struct mii_bus *bus, struct module *owner) bus->parent->of_node->fwnode.flags |= FWNODE_FLAG_NEEDS_CHILD_BOUND_ON_ADD; - BUG_ON(bus->state != MDIOBUS_ALLOCATED && - bus->state != MDIOBUS_UNREGISTERED); + WARN(bus->state != MDIOBUS_ALLOCATED && + bus->state != MDIOBUS_UNREGISTERED, + "%s: not in ALLOCATED or UNREGISTERED state\n", bus->id); bus->owner = owner; bus->dev.parent = bus->parent; @@ -658,7 +661,8 @@ void mdiobus_free(struct mii_bus *bus) return; } - BUG_ON(bus->state != MDIOBUS_UNREGISTERED); + WARN(bus->state != MDIOBUS_UNREGISTERED, + "%s: not in UNREGISTERED state\n", bus->id); bus->state = MDIOBUS_RELEASED; put_device(&bus->dev); From df6160deb3debe6f964c16349f9431157ff67dda Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 17 Nov 2021 17:57:29 -0800 Subject: [PATCH 0323/2048] tcp: add missing htmldocs for skb->ll_node and sk->defer_list Add missing entries to fix these "make htmldocs" warnings. ./include/linux/skbuff.h:953: warning: Function parameter or member 'll_node' not described in 'sk_buff' ./include/net/sock.h:540: warning: Function parameter or member 'defer_list' not described in 'sock' Fixes: f35f821935d8 ("tcp: defer skb freeing after socket lock is released") Signed-off-by: Eric Dumazet Reported-by: Stephen Rothwell Signed-off-by: David S. Miller --- include/linux/skbuff.h | 1 + include/net/sock.h | 1 + 2 files changed, 2 insertions(+) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index b8b806512e16..100fd604fbc9 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -627,6 +627,7 @@ typedef unsigned char *sk_buff_data_t; * for retransmit timer * @rbnode: RB tree node, alternative to next/prev for netem/tcp * @list: queue head + * @ll_node: anchor in an llist (eg socket defer_list) * @sk: Socket we are owned by * @ip_defrag_offset: (aka @sk) alternate use of @sk, used in * fragmentation management diff --git a/include/net/sock.h b/include/net/sock.h index f09c0c4736c4..a79fc772324e 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -292,6 +292,7 @@ struct bpf_local_storage; * @sk_pacing_shift: scaling factor for TCP Small Queues * @sk_lingertime: %SO_LINGER l_linger setting * @sk_backlog: always used with the per-socket spinlock held + * @defer_list: head of llist storing skbs to be freed * @sk_callback_lock: used with the callbacks in the end of this struct * @sk_error_queue: rarely used * @sk_prot_creator: sk_prot of original sock creator (see ipv6_setsockopt, From 4cdf85ef2371e851d2a93602b7965fd5758e3ac4 Mon Sep 17 00:00:00 2001 From: Yao Jing Date: Thu, 18 Nov 2021 06:10:18 +0000 Subject: [PATCH 0324/2048] ipv6: ah6: use swap() to make code cleaner Use the macro 'swap()' defined in 'include/linux/minmax.h' to avoid opencoding it. Reported-by: Zeal Robot Signed-off-by: Yao Jing Signed-off-by: David S. Miller --- net/ipv6/ah6.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c index 828e62514260..b5995c1f4d7a 100644 --- a/net/ipv6/ah6.c +++ b/net/ipv6/ah6.c @@ -175,7 +175,6 @@ static void ipv6_rearrange_destopt(struct ipv6hdr *iph, struct ipv6_opt_hdr *des * See 11.3.2 of RFC 3775 for details. */ if (opt[off] == IPV6_TLV_HAO) { - struct in6_addr final_addr; struct ipv6_destopt_hao *hao; hao = (struct ipv6_destopt_hao *)&opt[off]; @@ -184,9 +183,7 @@ static void ipv6_rearrange_destopt(struct ipv6hdr *iph, struct ipv6_opt_hdr *des hao->length); goto bad; } - final_addr = hao->addr; - hao->addr = iph->saddr; - iph->saddr = final_addr; + swap(hao->addr, iph->saddr); } break; } From f6ef47e5bdc6f652176e433b02317fc83049f8d7 Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Thu, 18 Nov 2021 14:57:23 +0800 Subject: [PATCH 0325/2048] mctp/test: Update refcount checking in route fragment tests In 99ce45d5e, we moved a route refcount decrement from mctp_do_fragment_route into the caller. This invalidates the assumption that the route test makes about refcount behaviour, so the route tests fail. This change fixes the test case to suit the new refcount behaviour. Fixes: 99ce45d5e7db ("mctp: Implement extended addressing") Signed-off-by: Jeremy Kerr Signed-off-by: David S. Miller --- net/mctp/test/route-test.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c index 36fac3daf86a..86ad15abf897 100644 --- a/net/mctp/test/route-test.c +++ b/net/mctp/test/route-test.c @@ -150,11 +150,6 @@ static void mctp_test_fragment(struct kunit *test) rt = mctp_test_create_route(&init_net, NULL, 10, mtu); KUNIT_ASSERT_TRUE(test, rt); - /* The refcount would usually be incremented as part of a route lookup, - * but we're setting the route directly here. - */ - refcount_inc(&rt->rt.refs); - rc = mctp_do_fragment_route(&rt->rt, skb, mtu, MCTP_TAG_OWNER); KUNIT_EXPECT_FALSE(test, rc); From 0568c3bf3f34ad2f86e6b2dfaa0855aad9c1c562 Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:11:57 +0800 Subject: [PATCH 0326/2048] net: mscc: ocelot: add MAC table stream learn and lookup operations ocelot_mact_learn_streamdata() can be used in VSC9959 to overwrite an FDB entry with stream data. The stream data includes SFID and SSID which can be used for PSFP and FRER set. ocelot_mact_lookup() can be used to check if the given {DMAC, VID} FDB entry is exist, and also can retrieve the DEST_IDX and entry type for the FDB entry. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot.c | 81 +++++++++++++++++++++++++++--- drivers/net/ethernet/mscc/ocelot.h | 13 ----- include/soc/mscc/ocelot.h | 22 ++++++++ 3 files changed, 97 insertions(+), 19 deletions(-) diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c index e6c18b598d5c..9e981913d6ba 100644 --- a/drivers/net/ethernet/mscc/ocelot.c +++ b/drivers/net/ethernet/mscc/ocelot.c @@ -61,9 +61,9 @@ static void ocelot_mact_select(struct ocelot *ocelot, } -int ocelot_mact_learn(struct ocelot *ocelot, int port, - const unsigned char mac[ETH_ALEN], - unsigned int vid, enum macaccess_entry_type type) +static int __ocelot_mact_learn(struct ocelot *ocelot, int port, + const unsigned char mac[ETH_ALEN], + unsigned int vid, enum macaccess_entry_type type) { u32 cmd = ANA_TABLES_MACACCESS_VALID | ANA_TABLES_MACACCESS_DEST_IDX(port) | @@ -83,8 +83,6 @@ int ocelot_mact_learn(struct ocelot *ocelot, int port, if (mc_ports & BIT(ocelot->num_phys_ports)) cmd |= ANA_TABLES_MACACCESS_MAC_CPU_COPY; - mutex_lock(&ocelot->mact_lock); - ocelot_mact_select(ocelot, mac, vid); /* Issue a write command */ @@ -92,9 +90,20 @@ int ocelot_mact_learn(struct ocelot *ocelot, int port, err = ocelot_mact_wait_for_completion(ocelot); + return err; +} + +int ocelot_mact_learn(struct ocelot *ocelot, int port, + const unsigned char mac[ETH_ALEN], + unsigned int vid, enum macaccess_entry_type type) +{ + int ret; + + mutex_lock(&ocelot->mact_lock); + ret = __ocelot_mact_learn(ocelot, port, mac, vid, type); mutex_unlock(&ocelot->mact_lock); - return err; + return ret; } EXPORT_SYMBOL(ocelot_mact_learn); @@ -120,6 +129,66 @@ int ocelot_mact_forget(struct ocelot *ocelot, } EXPORT_SYMBOL(ocelot_mact_forget); +int ocelot_mact_lookup(struct ocelot *ocelot, int *dst_idx, + const unsigned char mac[ETH_ALEN], + unsigned int vid, enum macaccess_entry_type *type) +{ + int val; + + mutex_lock(&ocelot->mact_lock); + + ocelot_mact_select(ocelot, mac, vid); + + /* Issue a read command with MACACCESS_VALID=1. */ + ocelot_write(ocelot, ANA_TABLES_MACACCESS_VALID | + ANA_TABLES_MACACCESS_MAC_TABLE_CMD(MACACCESS_CMD_READ), + ANA_TABLES_MACACCESS); + + if (ocelot_mact_wait_for_completion(ocelot)) { + mutex_unlock(&ocelot->mact_lock); + return -ETIMEDOUT; + } + + /* Read back the entry flags */ + val = ocelot_read(ocelot, ANA_TABLES_MACACCESS); + + mutex_unlock(&ocelot->mact_lock); + + if (!(val & ANA_TABLES_MACACCESS_VALID)) + return -ENOENT; + + *dst_idx = ANA_TABLES_MACACCESS_DEST_IDX_X(val); + *type = ANA_TABLES_MACACCESS_ENTRYTYPE_X(val); + + return 0; +} +EXPORT_SYMBOL(ocelot_mact_lookup); + +int ocelot_mact_learn_streamdata(struct ocelot *ocelot, int dst_idx, + const unsigned char mac[ETH_ALEN], + unsigned int vid, + enum macaccess_entry_type type, + int sfid, int ssid) +{ + int ret; + + mutex_lock(&ocelot->mact_lock); + + ocelot_write(ocelot, + (sfid < 0 ? 0 : ANA_TABLES_STREAMDATA_SFID_VALID) | + ANA_TABLES_STREAMDATA_SFID(sfid) | + (ssid < 0 ? 0 : ANA_TABLES_STREAMDATA_SSID_VALID) | + ANA_TABLES_STREAMDATA_SSID(ssid), + ANA_TABLES_STREAMDATA); + + ret = __ocelot_mact_learn(ocelot, dst_idx, mac, vid, type); + + mutex_unlock(&ocelot->mact_lock); + + return ret; +} +EXPORT_SYMBOL(ocelot_mact_learn_streamdata); + static void ocelot_mact_init(struct ocelot *ocelot) { /* Configure the learning mode entries attributes: diff --git a/drivers/net/ethernet/mscc/ocelot.h b/drivers/net/ethernet/mscc/ocelot.h index e43da09b8f91..1eb0b5ad51e9 100644 --- a/drivers/net/ethernet/mscc/ocelot.h +++ b/drivers/net/ethernet/mscc/ocelot.h @@ -55,19 +55,6 @@ struct ocelot_dump_ctx { int idx; }; -/* MAC table entry types. - * ENTRYTYPE_NORMAL is subject to aging. - * ENTRYTYPE_LOCKED is not subject to aging. - * ENTRYTYPE_MACv4 is not subject to aging. For IPv4 multicast. - * ENTRYTYPE_MACv6 is not subject to aging. For IPv6 multicast. - */ -enum macaccess_entry_type { - ENTRYTYPE_NORMAL = 0, - ENTRYTYPE_LOCKED, - ENTRYTYPE_MACv4, - ENTRYTYPE_MACv6, -}; - /* A (PGID) port mask structure, encoding the 2^ocelot->num_phys_ports * possibilities of egress port masks for L2 multicast traffic. * For a switch with 9 user ports, there are 512 possible port masks, but the diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h index fef3a36b0210..1d5ff11e4100 100644 --- a/include/soc/mscc/ocelot.h +++ b/include/soc/mscc/ocelot.h @@ -593,6 +593,19 @@ enum ocelot_sb_pool { OCELOT_SB_POOL_NUM, }; +/* MAC table entry types. + * ENTRYTYPE_NORMAL is subject to aging. + * ENTRYTYPE_LOCKED is not subject to aging. + * ENTRYTYPE_MACv4 is not subject to aging. For IPv4 multicast. + * ENTRYTYPE_MACv6 is not subject to aging. For IPv6 multicast. + */ +enum macaccess_entry_type { + ENTRYTYPE_NORMAL = 0, + ENTRYTYPE_LOCKED, + ENTRYTYPE_MACv4, + ENTRYTYPE_MACv6, +}; + #define OCELOT_QUIRK_PCS_PERFORMS_RATE_ADAPTATION BIT(0) #define OCELOT_QUIRK_QSGMII_PORTS_MUST_BE_UP BIT(1) @@ -870,6 +883,15 @@ void ocelot_phylink_mac_link_up(struct ocelot *ocelot, int port, bool tx_pause, bool rx_pause, unsigned long quirks); +int ocelot_mact_lookup(struct ocelot *ocelot, int *dst_idx, + const unsigned char mac[ETH_ALEN], + unsigned int vid, enum macaccess_entry_type *type); +int ocelot_mact_learn_streamdata(struct ocelot *ocelot, int dst_idx, + const unsigned char mac[ETH_ALEN], + unsigned int vid, + enum macaccess_entry_type type, + int sfid, int ssid); + #if IS_ENABLED(CONFIG_BRIDGE_MRP) int ocelot_mrp_add(struct ocelot *ocelot, int port, const struct switchdev_obj_mrp *mrp); From 5b1918a54a9180fdb2276dd4a90976e066c410da Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:11:58 +0800 Subject: [PATCH 0327/2048] net: mscc: ocelot: set vcap IS2 chain to goto PSFP chain Some chips in the ocelot series such as VSC9959 support Per-Stream Filtering and Policing(PSFP), which is processing after VCAP blocks. We set this block on chain 30000 and set vcap IS2 chain to goto PSFP chain if hardware support. Signed-off-by: Xiaoliang Yang Reviewed-by: Vladimir Oltean Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot_flower.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mscc/ocelot_flower.c b/drivers/net/ethernet/mscc/ocelot_flower.c index 769a8159373e..ed609bc4398e 100644 --- a/drivers/net/ethernet/mscc/ocelot_flower.c +++ b/drivers/net/ethernet/mscc/ocelot_flower.c @@ -20,6 +20,9 @@ (1 * VCAP_BLOCK + (lookup) * VCAP_LOOKUP) #define VCAP_IS2_CHAIN(lookup, pag) \ (2 * VCAP_BLOCK + (lookup) * VCAP_LOOKUP + (pag)) +/* PSFP chain and block ID */ +#define PSFP_BLOCK_ID OCELOT_NUM_VCAP_BLOCKS +#define OCELOT_PSFP_CHAIN (3 * VCAP_BLOCK) static int ocelot_chain_to_block(int chain, bool ingress) { @@ -46,6 +49,9 @@ static int ocelot_chain_to_block(int chain, bool ingress) if (chain == VCAP_IS2_CHAIN(lookup, pag)) return VCAP_IS2; + if (chain == OCELOT_PSFP_CHAIN) + return PSFP_BLOCK_ID; + return -EOPNOTSUPP; } @@ -84,7 +90,8 @@ static bool ocelot_is_goto_target_valid(int goto_target, int chain, goto_target == VCAP_IS1_CHAIN(1) || goto_target == VCAP_IS1_CHAIN(2) || goto_target == VCAP_IS2_CHAIN(0, 0) || - goto_target == VCAP_IS2_CHAIN(1, 0)); + goto_target == VCAP_IS2_CHAIN(1, 0) || + goto_target == OCELOT_PSFP_CHAIN); if (chain == VCAP_IS1_CHAIN(0)) return (goto_target == VCAP_IS1_CHAIN(1)); @@ -111,7 +118,11 @@ static bool ocelot_is_goto_target_valid(int goto_target, int chain, if (chain == VCAP_IS2_CHAIN(0, pag)) return (goto_target == VCAP_IS2_CHAIN(1, pag)); - /* VCAP IS2 lookup 1 cannot jump anywhere */ + /* VCAP IS2 lookup 1 can goto to PSFP block if hardware support */ + for (pag = 0; pag < VCAP_IS2_NUM_PAG; pag++) + if (chain == VCAP_IS2_CHAIN(1, pag)) + return (goto_target == OCELOT_PSFP_CHAIN); + return false; } @@ -407,7 +418,7 @@ static int ocelot_flower_parse_action(struct ocelot *ocelot, int port, if (filter->goto_target == -1) { if ((filter->block_id == VCAP_IS2 && filter->lookup == 1) || - chain == 0) { + chain == 0 || filter->block_id == PSFP_BLOCK_ID) { allow_missing_goto_target = true; } else { NL_SET_ERR_MSG_MOD(extack, "Missing GOTO action"); From 23e2c506ad6c588b469e3d06cc20299434440d02 Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:11:59 +0800 Subject: [PATCH 0328/2048] net: mscc: ocelot: add gate and police action offload to PSFP PSFP support gate and police action. This patch add the gate and police action to flower parse action, check chain ID to determine which block to offload. Adding psfp callback functions to add, delete and update gate and police in PSFP table if hardware supports it. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/ethernet/mscc/ocelot.c | 3 ++ drivers/net/ethernet/mscc/ocelot_flower.c | 52 ++++++++++++++++++++++- include/soc/mscc/ocelot.h | 5 +++ include/soc/mscc/ocelot_vcap.h | 1 + 4 files changed, 59 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c index 9e981913d6ba..95920668feb0 100644 --- a/drivers/net/ethernet/mscc/ocelot.c +++ b/drivers/net/ethernet/mscc/ocelot.c @@ -2352,6 +2352,9 @@ int ocelot_init(struct ocelot *ocelot) ocelot_vcap_init(ocelot); ocelot_cpu_port_init(ocelot); + if (ocelot->ops->psfp_init) + ocelot->ops->psfp_init(ocelot); + for (port = 0; port < ocelot->num_phys_ports; port++) { /* Clear all counters (5 groups) */ ocelot_write(ocelot, SYS_STAT_CFG_STAT_VIEW(port) | diff --git a/drivers/net/ethernet/mscc/ocelot_flower.c b/drivers/net/ethernet/mscc/ocelot_flower.c index ed609bc4398e..b22966e15acf 100644 --- a/drivers/net/ethernet/mscc/ocelot_flower.c +++ b/drivers/net/ethernet/mscc/ocelot_flower.c @@ -280,10 +280,14 @@ static int ocelot_flower_parse_action(struct ocelot *ocelot, int port, filter->type = OCELOT_VCAP_FILTER_OFFLOAD; break; case FLOW_ACTION_POLICE: + if (filter->block_id == PSFP_BLOCK_ID) { + filter->type = OCELOT_PSFP_FILTER_OFFLOAD; + break; + } if (filter->block_id != VCAP_IS2 || filter->lookup != 0) { NL_SET_ERR_MSG_MOD(extack, - "Police action can only be offloaded to VCAP IS2 lookup 0"); + "Police action can only be offloaded to VCAP IS2 lookup 0 or PSFP"); return -EOPNOTSUPP; } if (filter->goto_target != -1) { @@ -410,6 +414,14 @@ static int ocelot_flower_parse_action(struct ocelot *ocelot, int port, filter->action.pcp_a_val = a->vlan.prio; filter->type = OCELOT_VCAP_FILTER_OFFLOAD; break; + case FLOW_ACTION_GATE: + if (filter->block_id != PSFP_BLOCK_ID) { + NL_SET_ERR_MSG_MOD(extack, + "Gate action can only be offloaded to PSFP chain"); + return -EOPNOTSUPP; + } + filter->type = OCELOT_PSFP_FILTER_OFFLOAD; + break; default: NL_SET_ERR_MSG_MOD(extack, "Cannot offload action"); return -EOPNOTSUPP; @@ -700,6 +712,10 @@ static int ocelot_flower_parse(struct ocelot *ocelot, int port, bool ingress, if (ret) return ret; + /* PSFP filter need to parse key by stream identification function. */ + if (filter->type == OCELOT_PSFP_FILTER_OFFLOAD) + return 0; + return ocelot_flower_parse_key(ocelot, port, ingress, f, filter); } @@ -803,6 +819,15 @@ int ocelot_cls_flower_replace(struct ocelot *ocelot, int port, if (filter->type == OCELOT_VCAP_FILTER_DUMMY) return ocelot_vcap_dummy_filter_add(ocelot, filter); + if (filter->type == OCELOT_PSFP_FILTER_OFFLOAD) { + kfree(filter); + if (ocelot->ops->psfp_filter_add) + return ocelot->ops->psfp_filter_add(ocelot, f); + + NL_SET_ERR_MSG_MOD(extack, "PSFP chain is not supported in HW"); + return -EOPNOTSUPP; + } + return ocelot_vcap_filter_add(ocelot, filter, f->common.extack); } EXPORT_SYMBOL_GPL(ocelot_cls_flower_replace); @@ -818,6 +843,13 @@ int ocelot_cls_flower_destroy(struct ocelot *ocelot, int port, if (block_id < 0) return 0; + if (block_id == PSFP_BLOCK_ID) { + if (ocelot->ops->psfp_filter_del) + return ocelot->ops->psfp_filter_del(ocelot, f); + + return -EOPNOTSUPP; + } + block = &ocelot->block[block_id]; filter = ocelot_vcap_block_find_filter_by_id(block, f->cookie, true); @@ -836,12 +868,25 @@ int ocelot_cls_flower_stats(struct ocelot *ocelot, int port, { struct ocelot_vcap_filter *filter; struct ocelot_vcap_block *block; + struct flow_stats stats = {0}; int block_id, ret; block_id = ocelot_chain_to_block(f->common.chain_index, ingress); if (block_id < 0) return 0; + if (block_id == PSFP_BLOCK_ID) { + if (ocelot->ops->psfp_stats_get) { + ret = ocelot->ops->psfp_stats_get(ocelot, f, &stats); + if (ret) + return ret; + + goto stats_update; + } + + return -EOPNOTSUPP; + } + block = &ocelot->block[block_id]; filter = ocelot_vcap_block_find_filter_by_id(block, f->cookie, true); @@ -852,7 +897,10 @@ int ocelot_cls_flower_stats(struct ocelot *ocelot, int port, if (ret) return ret; - flow_stats_update(&f->stats, 0x0, filter->stats.pkts, 0, 0x0, + stats.pkts = filter->stats.pkts; + +stats_update: + flow_stats_update(&f->stats, 0x0, stats.pkts, stats.drops, 0x0, FLOW_ACTION_HW_STATS_IMMEDIATE); return 0; } diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h index 1d5ff11e4100..e9985ace59c0 100644 --- a/include/soc/mscc/ocelot.h +++ b/include/soc/mscc/ocelot.h @@ -555,6 +555,11 @@ struct ocelot_ops { u16 (*wm_enc)(u16 value); u16 (*wm_dec)(u16 value); void (*wm_stat)(u32 val, u32 *inuse, u32 *maxuse); + void (*psfp_init)(struct ocelot *ocelot); + int (*psfp_filter_add)(struct ocelot *ocelot, struct flow_cls_offload *f); + int (*psfp_filter_del)(struct ocelot *ocelot, struct flow_cls_offload *f); + int (*psfp_stats_get)(struct ocelot *ocelot, struct flow_cls_offload *f, + struct flow_stats *stats); }; struct ocelot_vcap_block { diff --git a/include/soc/mscc/ocelot_vcap.h b/include/soc/mscc/ocelot_vcap.h index eeb1142aa1b1..9cca2f8e61a2 100644 --- a/include/soc/mscc/ocelot_vcap.h +++ b/include/soc/mscc/ocelot_vcap.h @@ -656,6 +656,7 @@ enum ocelot_vcap_filter_type { OCELOT_VCAP_FILTER_DUMMY, OCELOT_VCAP_FILTER_PAG, OCELOT_VCAP_FILTER_OFFLOAD, + OCELOT_PSFP_FILTER_OFFLOAD, }; struct ocelot_vcap_id { From 7d4b564d6adde3167dd015f7dbb7aee1d7a4294e Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:12:00 +0800 Subject: [PATCH 0329/2048] net: dsa: felix: support psfp filter on vsc9959 VSC9959 supports Per-Stream Filtering and Policing(PSFP) that complies with the IEEE 802.1Qci standard. The stream is identified by Null stream identification(DMAC and VLAN ID) defined in IEEE802.1CB. For PSFP, four tables need to be set up: stream table, stream filter table, stream gate table, and flow meter table. Identify the stream by parsing the tc flower keys and add it to the stream table. The stream filter table is automatically maintained, and its index is determined by SGID(flow gate index) and FMID(flow meter index). Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/dsa/ocelot/felix_vsc9959.c | 443 ++++++++++++++++++++++++- include/soc/mscc/ocelot.h | 8 + include/soc/mscc/ocelot_ana.h | 10 + 3 files changed, 451 insertions(+), 10 deletions(-) diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index 45c5ec7a83ea..d643e926d1bd 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -5,6 +5,7 @@ #include #include #include +#include #include #include #include @@ -292,7 +293,7 @@ static const u32 vsc9959_sys_regmap[] = { REG_RESERVED(SYS_MMGT_FAST), REG_RESERVED(SYS_EVENTS_DIF), REG_RESERVED(SYS_EVENTS_CORE), - REG_RESERVED(SYS_CNT), + REG(SYS_CNT, 0x000000), REG(SYS_PTP_STATUS, 0x000f14), REG(SYS_PTP_TXSTAMP, 0x000f18), REG(SYS_PTP_NXT, 0x000f1c), @@ -1020,15 +1021,6 @@ static void vsc9959_wm_stat(u32 val, u32 *inuse, u32 *maxuse) *maxuse = val & GENMASK(11, 0); } -static const struct ocelot_ops vsc9959_ops = { - .reset = vsc9959_reset, - .wm_enc = vsc9959_wm_enc, - .wm_dec = vsc9959_wm_dec, - .wm_stat = vsc9959_wm_stat, - .port_to_netdev = felix_port_to_netdev, - .netdev_to_port = felix_netdev_to_port, -}; - static int vsc9959_mdio_bus_alloc(struct ocelot *ocelot) { struct felix *felix = ocelot_to_felix(ocelot); @@ -1344,6 +1336,437 @@ static int vsc9959_port_setup_tc(struct dsa_switch *ds, int port, } } +#define VSC9959_PSFP_SFID_MAX 175 +#define VSC9959_PSFP_GATE_ID_MAX 183 +#define VSC9959_PSFP_POLICER_MAX 383 + +struct felix_stream { + struct list_head list; + unsigned long id; + u8 dmac[ETH_ALEN]; + u16 vid; + s8 prio; + u8 sfid_valid; + u8 ssid_valid; + u32 sfid; + u32 ssid; +}; + +struct felix_stream_filter { + struct list_head list; + refcount_t refcount; + u32 index; + u8 enable; + u8 sg_valid; + u32 sgid; + u8 fm_valid; + u32 fmid; + u8 prio_valid; + u8 prio; + u32 maxsdu; +}; + +struct felix_stream_filter_counters { + u32 match; + u32 not_pass_gate; + u32 not_pass_sdu; + u32 red; +}; + +static int vsc9959_stream_identify(struct flow_cls_offload *f, + struct felix_stream *stream) +{ + struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct flow_dissector *dissector = rule->match.dissector; + + if (dissector->used_keys & + ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) | + BIT(FLOW_DISSECTOR_KEY_BASIC) | + BIT(FLOW_DISSECTOR_KEY_VLAN) | + BIT(FLOW_DISSECTOR_KEY_ETH_ADDRS))) + return -EOPNOTSUPP; + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ETH_ADDRS)) { + struct flow_match_eth_addrs match; + + flow_rule_match_eth_addrs(rule, &match); + ether_addr_copy(stream->dmac, match.key->dst); + if (!is_zero_ether_addr(match.mask->src)) + return -EOPNOTSUPP; + } else { + return -EOPNOTSUPP; + } + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) { + struct flow_match_vlan match; + + flow_rule_match_vlan(rule, &match); + if (match.mask->vlan_priority) + stream->prio = match.key->vlan_priority; + else + stream->prio = -1; + + if (!match.mask->vlan_id) + return -EOPNOTSUPP; + stream->vid = match.key->vlan_id; + } else { + return -EOPNOTSUPP; + } + + stream->id = f->cookie; + + return 0; +} + +static int vsc9959_mact_stream_set(struct ocelot *ocelot, + struct felix_stream *stream, + struct netlink_ext_ack *extack) +{ + enum macaccess_entry_type type; + int ret, sfid, ssid; + u32 vid, dst_idx; + u8 mac[ETH_ALEN]; + + ether_addr_copy(mac, stream->dmac); + vid = stream->vid; + + /* Stream identification desn't support to add a stream with non + * existent MAC (The MAC entry has not been learned in MAC table). + */ + ret = ocelot_mact_lookup(ocelot, &dst_idx, mac, vid, &type); + if (ret) { + if (extack) + NL_SET_ERR_MSG_MOD(extack, "Stream is not learned in MAC table"); + return -EOPNOTSUPP; + } + + if ((stream->sfid_valid || stream->ssid_valid) && + type == ENTRYTYPE_NORMAL) + type = ENTRYTYPE_LOCKED; + + sfid = stream->sfid_valid ? stream->sfid : -1; + ssid = stream->ssid_valid ? stream->ssid : -1; + + ret = ocelot_mact_learn_streamdata(ocelot, dst_idx, mac, vid, type, + sfid, ssid); + + return ret; +} + +static struct felix_stream * +vsc9959_stream_table_lookup(struct list_head *stream_list, + struct felix_stream *stream) +{ + struct felix_stream *tmp; + + list_for_each_entry(tmp, stream_list, list) + if (ether_addr_equal(tmp->dmac, stream->dmac) && + tmp->vid == stream->vid) + return tmp; + + return NULL; +} + +static int vsc9959_stream_table_add(struct ocelot *ocelot, + struct list_head *stream_list, + struct felix_stream *stream, + struct netlink_ext_ack *extack) +{ + struct felix_stream *stream_entry; + int ret; + + stream_entry = kzalloc(sizeof(*stream_entry), GFP_KERNEL); + if (!stream_entry) + return -ENOMEM; + + memcpy(stream_entry, stream, sizeof(*stream_entry)); + + ret = vsc9959_mact_stream_set(ocelot, stream_entry, extack); + if (ret) { + kfree(stream_entry); + return ret; + } + + list_add_tail(&stream_entry->list, stream_list); + + return 0; +} + +static struct felix_stream * +vsc9959_stream_table_get(struct list_head *stream_list, unsigned long id) +{ + struct felix_stream *tmp; + + list_for_each_entry(tmp, stream_list, list) + if (tmp->id == id) + return tmp; + + return NULL; +} + +static void vsc9959_stream_table_del(struct ocelot *ocelot, + struct felix_stream *stream) +{ + vsc9959_mact_stream_set(ocelot, stream, NULL); + + list_del(&stream->list); + kfree(stream); +} + +static u32 vsc9959_sfi_access_status(struct ocelot *ocelot) +{ + return ocelot_read(ocelot, ANA_TABLES_SFIDACCESS); +} + +static int vsc9959_psfp_sfi_set(struct ocelot *ocelot, + struct felix_stream_filter *sfi) +{ + u32 val; + + if (sfi->index > VSC9959_PSFP_SFID_MAX) + return -EINVAL; + + if (!sfi->enable) { + ocelot_write(ocelot, ANA_TABLES_SFIDTIDX_SFID_INDEX(sfi->index), + ANA_TABLES_SFIDTIDX); + + val = ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(SFIDACCESS_CMD_WRITE); + ocelot_write(ocelot, val, ANA_TABLES_SFIDACCESS); + + return readx_poll_timeout(vsc9959_sfi_access_status, ocelot, val, + (!ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(val)), + 10, 100000); + } + + if (sfi->sgid > VSC9959_PSFP_GATE_ID_MAX || + sfi->fmid > VSC9959_PSFP_POLICER_MAX) + return -EINVAL; + + ocelot_write(ocelot, + (sfi->sg_valid ? ANA_TABLES_SFIDTIDX_SGID_VALID : 0) | + ANA_TABLES_SFIDTIDX_SGID(sfi->sgid) | + (sfi->fm_valid ? ANA_TABLES_SFIDTIDX_POL_ENA : 0) | + ANA_TABLES_SFIDTIDX_POL_IDX(sfi->fmid) | + ANA_TABLES_SFIDTIDX_SFID_INDEX(sfi->index), + ANA_TABLES_SFIDTIDX); + + ocelot_write(ocelot, + (sfi->prio_valid ? ANA_TABLES_SFIDACCESS_IGR_PRIO_MATCH_ENA : 0) | + ANA_TABLES_SFIDACCESS_IGR_PRIO(sfi->prio) | + ANA_TABLES_SFIDACCESS_MAX_SDU_LEN(sfi->maxsdu) | + ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(SFIDACCESS_CMD_WRITE), + ANA_TABLES_SFIDACCESS); + + return readx_poll_timeout(vsc9959_sfi_access_status, ocelot, val, + (!ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(val)), + 10, 100000); +} + +static int vsc9959_psfp_sfi_table_add(struct ocelot *ocelot, + struct felix_stream_filter *sfi) +{ + struct felix_stream_filter *sfi_entry, *tmp; + struct list_head *pos, *q, *last; + struct ocelot_psfp_list *psfp; + u32 insert = 0; + int ret; + + psfp = &ocelot->psfp; + last = &psfp->sfi_list; + + list_for_each_safe(pos, q, &psfp->sfi_list) { + tmp = list_entry(pos, struct felix_stream_filter, list); + if (sfi->sg_valid == tmp->sg_valid && + sfi->fm_valid == tmp->fm_valid && + tmp->sgid == sfi->sgid && + tmp->fmid == sfi->fmid) { + sfi->index = tmp->index; + refcount_inc(&tmp->refcount); + return 0; + } + /* Make sure that the index is increasing in order. */ + if (tmp->index == insert) { + last = pos; + insert++; + } + } + sfi->index = insert; + + sfi_entry = kzalloc(sizeof(*sfi_entry), GFP_KERNEL); + if (!sfi_entry) + return -ENOMEM; + + memcpy(sfi_entry, sfi, sizeof(*sfi_entry)); + refcount_set(&sfi_entry->refcount, 1); + + ret = vsc9959_psfp_sfi_set(ocelot, sfi_entry); + if (ret) { + kfree(sfi_entry); + return ret; + } + + list_add(&sfi_entry->list, last); + + return 0; +} + +static void vsc9959_psfp_sfi_table_del(struct ocelot *ocelot, u32 index) +{ + struct felix_stream_filter *tmp, *n; + struct ocelot_psfp_list *psfp; + u8 z; + + psfp = &ocelot->psfp; + + list_for_each_entry_safe(tmp, n, &psfp->sfi_list, list) + if (tmp->index == index) { + z = refcount_dec_and_test(&tmp->refcount); + if (z) { + tmp->enable = 0; + vsc9959_psfp_sfi_set(ocelot, tmp); + list_del(&tmp->list); + kfree(tmp); + } + break; + } +} + +static void vsc9959_psfp_counters_get(struct ocelot *ocelot, u32 index, + struct felix_stream_filter_counters *counters) +{ + ocelot_rmw(ocelot, SYS_STAT_CFG_STAT_VIEW(index), + SYS_STAT_CFG_STAT_VIEW_M, + SYS_STAT_CFG); + + counters->match = ocelot_read_gix(ocelot, SYS_CNT, 0x200); + counters->not_pass_gate = ocelot_read_gix(ocelot, SYS_CNT, 0x201); + counters->not_pass_sdu = ocelot_read_gix(ocelot, SYS_CNT, 0x202); + counters->red = ocelot_read_gix(ocelot, SYS_CNT, 0x203); + + /* Clear the PSFP counter. */ + ocelot_write(ocelot, + SYS_STAT_CFG_STAT_VIEW(index) | + SYS_STAT_CFG_STAT_CLEAR_SHOT(0x10), + SYS_STAT_CFG); +} + +static int vsc9959_psfp_filter_add(struct ocelot *ocelot, + struct flow_cls_offload *f) +{ + struct netlink_ext_ack *extack = f->common.extack; + struct felix_stream_filter sfi = {0}; + const struct flow_action_entry *a; + struct felix_stream *stream_entry; + struct felix_stream stream = {0}; + struct ocelot_psfp_list *psfp; + int ret, i; + + psfp = &ocelot->psfp; + + ret = vsc9959_stream_identify(f, &stream); + if (ret) { + NL_SET_ERR_MSG_MOD(extack, "Only can match on VID, PCP, and dest MAC"); + return ret; + } + + flow_action_for_each(i, a, &f->rule->action) { + switch (a->id) { + case FLOW_ACTION_GATE: + case FLOW_ACTION_POLICE: + default: + return -EOPNOTSUPP; + } + } + + /* Check if stream is set. */ + stream_entry = vsc9959_stream_table_lookup(&psfp->stream_list, &stream); + if (stream_entry) { + NL_SET_ERR_MSG_MOD(extack, "This stream is already added"); + return -EEXIST; + } + + sfi.prio_valid = (stream.prio < 0 ? 0 : 1); + sfi.prio = (sfi.prio_valid ? stream.prio : 0); + sfi.enable = 1; + + ret = vsc9959_psfp_sfi_table_add(ocelot, &sfi); + if (ret) + return ret; + + stream.sfid = sfi.index; + stream.sfid_valid = 1; + ret = vsc9959_stream_table_add(ocelot, &psfp->stream_list, + &stream, extack); + if (ret) + vsc9959_psfp_sfi_table_del(ocelot, stream.sfid); + + return ret; +} + +static int vsc9959_psfp_filter_del(struct ocelot *ocelot, + struct flow_cls_offload *f) +{ + struct ocelot_psfp_list *psfp; + struct felix_stream *stream; + + psfp = &ocelot->psfp; + + stream = vsc9959_stream_table_get(&psfp->stream_list, f->cookie); + if (!stream) + return -ENOMEM; + + vsc9959_psfp_sfi_table_del(ocelot, stream->sfid); + + stream->sfid_valid = 0; + vsc9959_stream_table_del(ocelot, stream); + + return 0; +} + +static int vsc9959_psfp_stats_get(struct ocelot *ocelot, + struct flow_cls_offload *f, + struct flow_stats *stats) +{ + struct felix_stream_filter_counters counters; + struct ocelot_psfp_list *psfp; + struct felix_stream *stream; + + psfp = &ocelot->psfp; + stream = vsc9959_stream_table_get(&psfp->stream_list, f->cookie); + if (!stream) + return -ENOMEM; + + vsc9959_psfp_counters_get(ocelot, stream->sfid, &counters); + + stats->pkts = counters.match; + stats->drops = counters.not_pass_gate + counters.not_pass_sdu + + counters.red; + + return 0; +} + +static void vsc9959_psfp_init(struct ocelot *ocelot) +{ + struct ocelot_psfp_list *psfp = &ocelot->psfp; + + INIT_LIST_HEAD(&psfp->stream_list); + INIT_LIST_HEAD(&psfp->sfi_list); + INIT_LIST_HEAD(&psfp->sgi_list); +} + +static const struct ocelot_ops vsc9959_ops = { + .reset = vsc9959_reset, + .wm_enc = vsc9959_wm_enc, + .wm_dec = vsc9959_wm_dec, + .wm_stat = vsc9959_wm_stat, + .port_to_netdev = felix_port_to_netdev, + .netdev_to_port = felix_netdev_to_port, + .psfp_init = vsc9959_psfp_init, + .psfp_filter_add = vsc9959_psfp_filter_add, + .psfp_filter_del = vsc9959_psfp_filter_del, + .psfp_stats_get = vsc9959_psfp_stats_get, +}; + static const struct felix_info felix_info_vsc9959 = { .target_io_res = vsc9959_target_io_res, .port_io_res = vsc9959_port_io_res, diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h index e9985ace59c0..5ea72d274d7f 100644 --- a/include/soc/mscc/ocelot.h +++ b/include/soc/mscc/ocelot.h @@ -586,6 +586,12 @@ enum ocelot_port_tag_config { OCELOT_PORT_TAG_TRUNK = 3, }; +struct ocelot_psfp_list { + struct list_head stream_list; + struct list_head sfi_list; + struct list_head sgi_list; +}; + enum ocelot_sb { OCELOT_SB_BUF, OCELOT_SB_REF, @@ -687,6 +693,8 @@ struct ocelot { struct ocelot_vcap_block block[3]; struct vcap_props *vcap; + struct ocelot_psfp_list psfp; + /* Workqueue to check statistics for overflow with its lock */ struct mutex stats_lock; u64 *stats; diff --git a/include/soc/mscc/ocelot_ana.h b/include/soc/mscc/ocelot_ana.h index 1669481d9779..67e0ae05a5ab 100644 --- a/include/soc/mscc/ocelot_ana.h +++ b/include/soc/mscc/ocelot_ana.h @@ -227,6 +227,11 @@ #define ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(x) ((x) & GENMASK(1, 0)) #define ANA_TABLES_SFIDACCESS_SFID_TBL_CMD_M GENMASK(1, 0) +#define SFIDACCESS_CMD_IDLE 0 +#define SFIDACCESS_CMD_READ 1 +#define SFIDACCESS_CMD_WRITE 2 +#define SFIDACCESS_CMD_INIT 3 + #define ANA_TABLES_SFIDTIDX_SGID_VALID BIT(26) #define ANA_TABLES_SFIDTIDX_SGID(x) (((x) << 18) & GENMASK(25, 18)) #define ANA_TABLES_SFIDTIDX_SGID_M GENMASK(25, 18) @@ -255,6 +260,11 @@ #define ANA_SG_CONFIG_REG_3_INIT_IPS(x) (((x) << 21) & GENMASK(24, 21)) #define ANA_SG_CONFIG_REG_3_INIT_IPS_M GENMASK(24, 21) #define ANA_SG_CONFIG_REG_3_INIT_IPS_X(x) (((x) & GENMASK(24, 21)) >> 21) +#define ANA_SG_CONFIG_REG_3_IPV_VALID BIT(24) +#define ANA_SG_CONFIG_REG_3_IPV_INVALID(x) (((x) << 24) & GENMASK(24, 24)) +#define ANA_SG_CONFIG_REG_3_INIT_IPV(x) (((x) << 21) & GENMASK(23, 21)) +#define ANA_SG_CONFIG_REG_3_INIT_IPV_M GENMASK(23, 21) +#define ANA_SG_CONFIG_REG_3_INIT_IPV_X(x) (((x) & GENMASK(23, 21)) >> 21) #define ANA_SG_CONFIG_REG_3_INIT_GATE_STATE BIT(25) #define ANA_SG_GCL_GS_CONFIG_RSZ 0x4 From 23ae3a7877718931474684ef4fbbaf1d1511ee84 Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:12:01 +0800 Subject: [PATCH 0330/2048] net: dsa: felix: add stream gate settings for psfp This patch adds stream gate settings for PSFP. Use SGI table to store stream gate entries. Disable the gate entry when it is not used by any stream. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/dsa/ocelot/felix_vsc9959.c | 217 ++++++++++++++++++++++++- 1 file changed, 213 insertions(+), 4 deletions(-) diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index d643e926d1bd..18a2e538f573 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -8,6 +8,7 @@ #include #include #include +#include #include #include #include @@ -1339,6 +1340,8 @@ static int vsc9959_port_setup_tc(struct dsa_switch *ds, int port, #define VSC9959_PSFP_SFID_MAX 175 #define VSC9959_PSFP_GATE_ID_MAX 183 #define VSC9959_PSFP_POLICER_MAX 383 +#define VSC9959_PSFP_GATE_LIST_NUM 4 +#define VSC9959_PSFP_GATE_CYCLETIME_MIN 5000 struct felix_stream { struct list_head list; @@ -1373,6 +1376,24 @@ struct felix_stream_filter_counters { u32 red; }; +struct felix_stream_gate { + u32 index; + u8 enable; + u8 ipv_valid; + u8 init_ipv; + u64 basetime; + u64 cycletime; + u64 cycletime_ext; + u32 num_entries; + struct action_gate_entry entries[0]; +}; + +struct felix_stream_gate_entry { + struct list_head list; + refcount_t refcount; + u32 index; +}; + static int vsc9959_stream_identify(struct flow_cls_offload *f, struct felix_stream *stream) { @@ -1610,6 +1631,18 @@ static int vsc9959_psfp_sfi_table_add(struct ocelot *ocelot, return 0; } +static struct felix_stream_filter * +vsc9959_psfp_sfi_table_get(struct list_head *sfi_list, u32 index) +{ + struct felix_stream_filter *tmp; + + list_for_each_entry(tmp, sfi_list, list) + if (tmp->index == index) + return tmp; + + return NULL; +} + static void vsc9959_psfp_sfi_table_del(struct ocelot *ocelot, u32 index) { struct felix_stream_filter *tmp, *n; @@ -1631,6 +1664,152 @@ static void vsc9959_psfp_sfi_table_del(struct ocelot *ocelot, u32 index) } } +static void vsc9959_psfp_parse_gate(const struct flow_action_entry *entry, + struct felix_stream_gate *sgi) +{ + sgi->index = entry->gate.index; + sgi->ipv_valid = (entry->gate.prio < 0) ? 0 : 1; + sgi->init_ipv = (sgi->ipv_valid) ? entry->gate.prio : 0; + sgi->basetime = entry->gate.basetime; + sgi->cycletime = entry->gate.cycletime; + sgi->num_entries = entry->gate.num_entries; + sgi->enable = 1; + + memcpy(sgi->entries, entry->gate.entries, + entry->gate.num_entries * sizeof(struct action_gate_entry)); +} + +static u32 vsc9959_sgi_cfg_status(struct ocelot *ocelot) +{ + return ocelot_read(ocelot, ANA_SG_ACCESS_CTRL); +} + +static int vsc9959_psfp_sgi_set(struct ocelot *ocelot, + struct felix_stream_gate *sgi) +{ + struct action_gate_entry *e; + struct timespec64 base_ts; + u32 interval_sum = 0; + u32 val; + int i; + + if (sgi->index > VSC9959_PSFP_GATE_ID_MAX) + return -EINVAL; + + ocelot_write(ocelot, ANA_SG_ACCESS_CTRL_SGID(sgi->index), + ANA_SG_ACCESS_CTRL); + + if (!sgi->enable) { + ocelot_rmw(ocelot, ANA_SG_CONFIG_REG_3_INIT_GATE_STATE, + ANA_SG_CONFIG_REG_3_INIT_GATE_STATE | + ANA_SG_CONFIG_REG_3_GATE_ENABLE, + ANA_SG_CONFIG_REG_3); + + return 0; + } + + if (sgi->cycletime < VSC9959_PSFP_GATE_CYCLETIME_MIN || + sgi->cycletime > NSEC_PER_SEC) + return -EINVAL; + + if (sgi->num_entries > VSC9959_PSFP_GATE_LIST_NUM) + return -EINVAL; + + vsc9959_new_base_time(ocelot, sgi->basetime, sgi->cycletime, &base_ts); + ocelot_write(ocelot, base_ts.tv_nsec, ANA_SG_CONFIG_REG_1); + val = lower_32_bits(base_ts.tv_sec); + ocelot_write(ocelot, val, ANA_SG_CONFIG_REG_2); + + val = upper_32_bits(base_ts.tv_sec); + ocelot_write(ocelot, + (sgi->ipv_valid ? ANA_SG_CONFIG_REG_3_IPV_VALID : 0) | + ANA_SG_CONFIG_REG_3_INIT_IPV(sgi->init_ipv) | + ANA_SG_CONFIG_REG_3_GATE_ENABLE | + ANA_SG_CONFIG_REG_3_LIST_LENGTH(sgi->num_entries) | + ANA_SG_CONFIG_REG_3_INIT_GATE_STATE | + ANA_SG_CONFIG_REG_3_BASE_TIME_SEC_MSB(val), + ANA_SG_CONFIG_REG_3); + + ocelot_write(ocelot, sgi->cycletime, ANA_SG_CONFIG_REG_4); + + e = sgi->entries; + for (i = 0; i < sgi->num_entries; i++) { + u32 ips = (e[i].ipv < 0) ? 0 : (e[i].ipv + 8); + + ocelot_write_rix(ocelot, ANA_SG_GCL_GS_CONFIG_IPS(ips) | + (e[i].gate_state ? + ANA_SG_GCL_GS_CONFIG_GATE_STATE : 0), + ANA_SG_GCL_GS_CONFIG, i); + + interval_sum += e[i].interval; + ocelot_write_rix(ocelot, interval_sum, ANA_SG_GCL_TI_CONFIG, i); + } + + ocelot_rmw(ocelot, ANA_SG_ACCESS_CTRL_CONFIG_CHANGE, + ANA_SG_ACCESS_CTRL_CONFIG_CHANGE, + ANA_SG_ACCESS_CTRL); + + return readx_poll_timeout(vsc9959_sgi_cfg_status, ocelot, val, + (!(ANA_SG_ACCESS_CTRL_CONFIG_CHANGE & val)), + 10, 100000); +} + +static int vsc9959_psfp_sgi_table_add(struct ocelot *ocelot, + struct felix_stream_gate *sgi) +{ + struct felix_stream_gate_entry *tmp; + struct ocelot_psfp_list *psfp; + int ret; + + psfp = &ocelot->psfp; + + list_for_each_entry(tmp, &psfp->sgi_list, list) + if (tmp->index == sgi->index) { + refcount_inc(&tmp->refcount); + return 0; + } + + tmp = kzalloc(sizeof(*tmp), GFP_KERNEL); + if (!tmp) + return -ENOMEM; + + ret = vsc9959_psfp_sgi_set(ocelot, sgi); + if (ret) { + kfree(tmp); + return ret; + } + + tmp->index = sgi->index; + refcount_set(&tmp->refcount, 1); + list_add_tail(&tmp->list, &psfp->sgi_list); + + return 0; +} + +static void vsc9959_psfp_sgi_table_del(struct ocelot *ocelot, + u32 index) +{ + struct felix_stream_gate_entry *tmp, *n; + struct felix_stream_gate sgi = {0}; + struct ocelot_psfp_list *psfp; + u8 z; + + psfp = &ocelot->psfp; + + list_for_each_entry_safe(tmp, n, &psfp->sgi_list, list) + if (tmp->index == index) { + z = refcount_dec_and_test(&tmp->refcount); + if (z) { + sgi.index = index; + sgi.enable = 0; + vsc9959_psfp_sgi_set(ocelot, &sgi); + list_del(&tmp->list); + kfree(tmp); + } + break; + } +} + static void vsc9959_psfp_counters_get(struct ocelot *ocelot, u32 index, struct felix_stream_filter_counters *counters) { @@ -1658,8 +1837,9 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, const struct flow_action_entry *a; struct felix_stream *stream_entry; struct felix_stream stream = {0}; + struct felix_stream_gate *sgi; struct ocelot_psfp_list *psfp; - int ret, i; + int ret, i, size; psfp = &ocelot->psfp; @@ -1672,6 +1852,18 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, flow_action_for_each(i, a, &f->rule->action) { switch (a->id) { case FLOW_ACTION_GATE: + size = struct_size(sgi, entries, a->gate.num_entries); + sgi = kzalloc(size, GFP_KERNEL); + vsc9959_psfp_parse_gate(a, sgi); + ret = vsc9959_psfp_sgi_table_add(ocelot, sgi); + if (ret) { + kfree(sgi); + return ret; + } + sfi.sg_valid = 1; + sfi.sgid = sgi->index; + kfree(sgi); + break; case FLOW_ACTION_POLICE: default: return -EOPNOTSUPP; @@ -1682,7 +1874,8 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, stream_entry = vsc9959_stream_table_lookup(&psfp->stream_list, &stream); if (stream_entry) { NL_SET_ERR_MSG_MOD(extack, "This stream is already added"); - return -EEXIST; + ret = -EEXIST; + goto err; } sfi.prio_valid = (stream.prio < 0 ? 0 : 1); @@ -1691,14 +1884,22 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, ret = vsc9959_psfp_sfi_table_add(ocelot, &sfi); if (ret) - return ret; + goto err; stream.sfid = sfi.index; stream.sfid_valid = 1; ret = vsc9959_stream_table_add(ocelot, &psfp->stream_list, &stream, extack); - if (ret) + if (ret) { vsc9959_psfp_sfi_table_del(ocelot, stream.sfid); + goto err; + } + + return 0; + +err: + if (sfi.sg_valid) + vsc9959_psfp_sgi_table_del(ocelot, sfi.sgid); return ret; } @@ -1706,6 +1907,7 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, static int vsc9959_psfp_filter_del(struct ocelot *ocelot, struct flow_cls_offload *f) { + static struct felix_stream_filter *sfi; struct ocelot_psfp_list *psfp; struct felix_stream *stream; @@ -1715,6 +1917,13 @@ static int vsc9959_psfp_filter_del(struct ocelot *ocelot, if (!stream) return -ENOMEM; + sfi = vsc9959_psfp_sfi_table_get(&psfp->sfi_list, stream->sfid); + if (!sfi) + return -ENOMEM; + + if (sfi->sg_valid) + vsc9959_psfp_sgi_table_del(ocelot, sfi->sgid); + vsc9959_psfp_sfi_table_del(ocelot, stream->sfid); stream->sfid_valid = 0; From 77043c37096d4753b9f40e51445f31eb9dc40295 Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:12:02 +0800 Subject: [PATCH 0331/2048] net: mscc: ocelot: use index to set vcap policer Policer was previously automatically assigned from the highest index to the lowest index from policer pool. But police action of tc flower now uses index to set an police entry. This patch uses the police index to set vcap policers, so that one policer can be shared by multiple rules. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/dsa/ocelot/felix.c | 4 + drivers/net/dsa/ocelot/felix.h | 4 + drivers/net/dsa/ocelot/felix_vsc9959.c | 6 ++ drivers/net/dsa/ocelot/seville_vsc9953.c | 8 ++ drivers/net/ethernet/mscc/ocelot_flower.c | 15 +++ drivers/net/ethernet/mscc/ocelot_vcap.c | 109 +++++++++++++-------- drivers/net/ethernet/mscc/ocelot_vsc7514.c | 7 ++ include/soc/mscc/ocelot.h | 14 ++- 8 files changed, 126 insertions(+), 41 deletions(-) diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c index 327cc4654806..e487143709da 100644 --- a/drivers/net/dsa/ocelot/felix.c +++ b/drivers/net/dsa/ocelot/felix.c @@ -989,6 +989,10 @@ static int felix_init_structs(struct felix *felix, int num_phys_ports) ocelot->num_stats = felix->info->num_stats; ocelot->num_mact_rows = felix->info->num_mact_rows; ocelot->vcap = felix->info->vcap; + ocelot->vcap_pol.base = felix->info->vcap_pol_base; + ocelot->vcap_pol.max = felix->info->vcap_pol_max; + ocelot->vcap_pol.base2 = felix->info->vcap_pol_base2; + ocelot->vcap_pol.max2 = felix->info->vcap_pol_max2; ocelot->ops = felix->info->ops; ocelot->npi_inj_prefix = OCELOT_TAG_PREFIX_SHORT; ocelot->npi_xtr_prefix = OCELOT_TAG_PREFIX_SHORT; diff --git a/drivers/net/dsa/ocelot/felix.h b/drivers/net/dsa/ocelot/felix.h index be3e42e135c0..dfe08dddd262 100644 --- a/drivers/net/dsa/ocelot/felix.h +++ b/drivers/net/dsa/ocelot/felix.h @@ -21,6 +21,10 @@ struct felix_info { int num_ports; int num_tx_queues; struct vcap_props *vcap; + u16 vcap_pol_base; + u16 vcap_pol_max; + u16 vcap_pol_base2; + u16 vcap_pol_max2; int switch_pci_bar; int imdio_pci_bar; const struct ptp_clock_info *ptp_caps; diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index 18a2e538f573..f8d770384344 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -19,6 +19,8 @@ #include "felix.h" #define VSC9959_TAS_GCL_ENTRY_MAX 63 +#define VSC9959_VCAP_POLICER_BASE 63 +#define VSC9959_VCAP_POLICER_MAX 383 static const u32 vsc9959_ana_regmap[] = { REG(ANA_ADVLEARN, 0x0089a0), @@ -1986,6 +1988,10 @@ static const struct felix_info felix_info_vsc9959 = { .stats_layout = vsc9959_stats_layout, .num_stats = ARRAY_SIZE(vsc9959_stats_layout), .vcap = vsc9959_vcap_props, + .vcap_pol_base = VSC9959_VCAP_POLICER_BASE, + .vcap_pol_max = VSC9959_VCAP_POLICER_MAX, + .vcap_pol_base2 = 0, + .vcap_pol_max2 = 0, .num_mact_rows = 2048, .num_ports = 6, .num_tx_queues = OCELOT_NUM_TC, diff --git a/drivers/net/dsa/ocelot/seville_vsc9953.c b/drivers/net/dsa/ocelot/seville_vsc9953.c index 92eae63150ea..899b98193b4a 100644 --- a/drivers/net/dsa/ocelot/seville_vsc9953.c +++ b/drivers/net/dsa/ocelot/seville_vsc9953.c @@ -18,6 +18,10 @@ #define MSCC_MIIM_CMD_REGAD_SHIFT 20 #define MSCC_MIIM_CMD_PHYAD_SHIFT 25 #define MSCC_MIIM_CMD_VLD BIT(31) +#define VSC9953_VCAP_POLICER_BASE 11 +#define VSC9953_VCAP_POLICER_MAX 31 +#define VSC9953_VCAP_POLICER_BASE2 120 +#define VSC9953_VCAP_POLICER_MAX2 161 static const u32 vsc9953_ana_regmap[] = { REG(ANA_ADVLEARN, 0x00b500), @@ -1172,6 +1176,10 @@ static const struct felix_info seville_info_vsc9953 = { .stats_layout = vsc9953_stats_layout, .num_stats = ARRAY_SIZE(vsc9953_stats_layout), .vcap = vsc9953_vcap_props, + .vcap_pol_base = VSC9953_VCAP_POLICER_BASE, + .vcap_pol_max = VSC9953_VCAP_POLICER_MAX, + .vcap_pol_base2 = VSC9953_VCAP_POLICER_BASE2, + .vcap_pol_max2 = VSC9953_VCAP_POLICER_MAX2, .num_mact_rows = 2048, .num_ports = 10, .num_tx_queues = OCELOT_NUM_TC, diff --git a/drivers/net/ethernet/mscc/ocelot_flower.c b/drivers/net/ethernet/mscc/ocelot_flower.c index b22966e15acf..b54b52fd9e1b 100644 --- a/drivers/net/ethernet/mscc/ocelot_flower.c +++ b/drivers/net/ethernet/mscc/ocelot_flower.c @@ -222,6 +222,7 @@ static int ocelot_flower_parse_action(struct ocelot *ocelot, int port, const struct flow_action_entry *a; enum ocelot_tag_tpid_sel tpid; int i, chain, egress_port; + u32 pol_ix, pol_max; u64 rate; int err; @@ -301,6 +302,20 @@ static int ocelot_flower_parse_action(struct ocelot *ocelot, int port, return -EOPNOTSUPP; } filter->action.police_ena = true; + + pol_ix = a->police.index + ocelot->vcap_pol.base; + pol_max = ocelot->vcap_pol.max; + + if (ocelot->vcap_pol.max2 && pol_ix > pol_max) { + pol_ix += ocelot->vcap_pol.base2 - pol_max - 1; + pol_max = ocelot->vcap_pol.max2; + } + + if (pol_ix >= pol_max) + return -EINVAL; + + filter->action.pol_ix = pol_ix; + rate = a->police.rate_bytes_ps; filter->action.pol.rate = div_u64(rate, 1000) * 8; filter->action.pol.burst = a->police.burst; diff --git a/drivers/net/ethernet/mscc/ocelot_vcap.c b/drivers/net/ethernet/mscc/ocelot_vcap.c index 99d7376a70a7..18ab0fd303c8 100644 --- a/drivers/net/ethernet/mscc/ocelot_vcap.c +++ b/drivers/net/ethernet/mscc/ocelot_vcap.c @@ -887,10 +887,18 @@ static void vcap_entry_set(struct ocelot *ocelot, int ix, return es0_entry_set(ocelot, ix, filter); } -static int ocelot_vcap_policer_add(struct ocelot *ocelot, u32 pol_ix, - struct ocelot_policer *pol) +struct vcap_policer_entry { + struct list_head list; + refcount_t refcount; + u32 pol_ix; +}; + +int ocelot_vcap_policer_add(struct ocelot *ocelot, u32 pol_ix, + struct ocelot_policer *pol) { struct qos_policer_conf pp = { 0 }; + struct vcap_policer_entry *tmp; + int ret; if (!pol) return -EINVAL; @@ -899,57 +907,74 @@ static int ocelot_vcap_policer_add(struct ocelot *ocelot, u32 pol_ix, pp.pir = pol->rate; pp.pbs = pol->burst; - return qos_policer_conf_set(ocelot, 0, pol_ix, &pp); -} - -static void ocelot_vcap_policer_del(struct ocelot *ocelot, - struct ocelot_vcap_block *block, - u32 pol_ix) -{ - struct ocelot_vcap_filter *filter; - struct qos_policer_conf pp = {0}; - int index = -1; - - if (pol_ix < block->pol_lpr) - return; - - list_for_each_entry(filter, &block->rules, list) { - index++; - if (filter->block_id == VCAP_IS2 && - filter->action.police_ena && - filter->action.pol_ix < pol_ix) { - filter->action.pol_ix += 1; - ocelot_vcap_policer_add(ocelot, filter->action.pol_ix, - &filter->action.pol); - is2_entry_set(ocelot, index, filter); + list_for_each_entry(tmp, &ocelot->vcap_pol.pol_list, list) + if (tmp->pol_ix == pol_ix) { + refcount_inc(&tmp->refcount); + return 0; } + + tmp = kzalloc(sizeof(*tmp), GFP_KERNEL); + if (!tmp) + return -ENOMEM; + + ret = qos_policer_conf_set(ocelot, 0, pol_ix, &pp); + if (ret) { + kfree(tmp); + return ret; } - pp.mode = MSCC_QOS_RATE_MODE_DISABLED; - qos_policer_conf_set(ocelot, 0, pol_ix, &pp); + tmp->pol_ix = pol_ix; + refcount_set(&tmp->refcount, 1); + list_add_tail(&tmp->list, &ocelot->vcap_pol.pol_list); - block->pol_lpr++; + return 0; } +EXPORT_SYMBOL(ocelot_vcap_policer_add); -static void ocelot_vcap_filter_add_to_block(struct ocelot *ocelot, - struct ocelot_vcap_block *block, - struct ocelot_vcap_filter *filter) +int ocelot_vcap_policer_del(struct ocelot *ocelot, u32 pol_ix) +{ + struct qos_policer_conf pp = {0}; + struct vcap_policer_entry *tmp, *n; + u8 z = 0; + + list_for_each_entry_safe(tmp, n, &ocelot->vcap_pol.pol_list, list) + if (tmp->pol_ix == pol_ix) { + z = refcount_dec_and_test(&tmp->refcount); + if (z) { + list_del(&tmp->list); + kfree(tmp); + } + } + + if (z) { + pp.mode = MSCC_QOS_RATE_MODE_DISABLED; + return qos_policer_conf_set(ocelot, 0, pol_ix, &pp); + } + + return 0; +} +EXPORT_SYMBOL(ocelot_vcap_policer_del); + +static int ocelot_vcap_filter_add_to_block(struct ocelot *ocelot, + struct ocelot_vcap_block *block, + struct ocelot_vcap_filter *filter) { struct ocelot_vcap_filter *tmp; struct list_head *pos, *n; + int ret; if (filter->block_id == VCAP_IS2 && filter->action.police_ena) { - block->pol_lpr--; - filter->action.pol_ix = block->pol_lpr; - ocelot_vcap_policer_add(ocelot, filter->action.pol_ix, - &filter->action.pol); + ret = ocelot_vcap_policer_add(ocelot, filter->action.pol_ix, + &filter->action.pol); + if (ret) + return ret; } block->count++; if (list_empty(&block->rules)) { list_add(&filter->list, &block->rules); - return; + return 0; } list_for_each_safe(pos, n, &block->rules) { @@ -958,6 +983,8 @@ static void ocelot_vcap_filter_add_to_block(struct ocelot *ocelot, break; } list_add(&filter->list, pos->prev); + + return 0; } static bool ocelot_vcap_filter_equal(const struct ocelot_vcap_filter *a, @@ -1132,7 +1159,7 @@ int ocelot_vcap_filter_add(struct ocelot *ocelot, struct netlink_ext_ack *extack) { struct ocelot_vcap_block *block = &ocelot->block[filter->block_id]; - int i, index; + int i, index, ret; if (!ocelot_exclusive_mac_etype_filter_rules(ocelot, filter)) { NL_SET_ERR_MSG_MOD(extack, @@ -1141,7 +1168,9 @@ int ocelot_vcap_filter_add(struct ocelot *ocelot, } /* Add filter to the linked list */ - ocelot_vcap_filter_add_to_block(ocelot, block, filter); + ret = ocelot_vcap_filter_add_to_block(ocelot, block, filter); + if (ret) + return ret; /* Get the index of the inserted filter */ index = ocelot_vcap_block_get_filter_index(block, filter); @@ -1174,7 +1203,7 @@ static void ocelot_vcap_block_remove_filter(struct ocelot *ocelot, if (ocelot_vcap_filter_equal(filter, tmp)) { if (tmp->block_id == VCAP_IS2 && tmp->action.police_ena) - ocelot_vcap_policer_del(ocelot, block, + ocelot_vcap_policer_del(ocelot, tmp->action.pol_ix); list_del(pos); @@ -1350,13 +1379,13 @@ int ocelot_vcap_init(struct ocelot *ocelot) struct vcap_props *vcap = &ocelot->vcap[i]; INIT_LIST_HEAD(&block->rules); - block->pol_lpr = OCELOT_POLICER_DISCARD - 1; ocelot_vcap_detect_constants(ocelot, vcap); ocelot_vcap_init_one(ocelot, vcap); } INIT_LIST_HEAD(&ocelot->dummy_rules); + INIT_LIST_HEAD(&ocelot->vcap_pol.pol_list); return 0; } diff --git a/drivers/net/ethernet/mscc/ocelot_vsc7514.c b/drivers/net/ethernet/mscc/ocelot_vsc7514.c index 38103b0255b0..cd3eb101f159 100644 --- a/drivers/net/ethernet/mscc/ocelot_vsc7514.c +++ b/drivers/net/ethernet/mscc/ocelot_vsc7514.c @@ -20,6 +20,9 @@ #include #include "ocelot.h" +#define VSC7514_VCAP_POLICER_BASE 128 +#define VSC7514_VCAP_POLICER_MAX 191 + static const u32 ocelot_ana_regmap[] = { REG(ANA_ADVLEARN, 0x009000), REG(ANA_VLANMASK, 0x009004), @@ -1129,6 +1132,10 @@ static int mscc_ocelot_probe(struct platform_device *pdev) ocelot->num_flooding_pgids = 1; ocelot->vcap = vsc7514_vcap_props; + + ocelot->vcap_pol.base = VSC7514_VCAP_POLICER_BASE; + ocelot->vcap_pol.max = VSC7514_VCAP_POLICER_MAX; + ocelot->npi = -1; err = ocelot_init(ocelot); diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h index 5ea72d274d7f..2a41685b5c7d 100644 --- a/include/soc/mscc/ocelot.h +++ b/include/soc/mscc/ocelot.h @@ -562,10 +562,17 @@ struct ocelot_ops { struct flow_stats *stats); }; +struct ocelot_vcap_policer { + struct list_head pol_list; + u16 base; + u16 max; + u16 base2; + u16 max2; +}; + struct ocelot_vcap_block { struct list_head rules; int count; - int pol_lpr; }; struct ocelot_bridge_vlan { @@ -691,6 +698,7 @@ struct ocelot { struct list_head dummy_rules; struct ocelot_vcap_block block[3]; + struct ocelot_vcap_policer vcap_pol; struct vcap_props *vcap; struct ocelot_psfp_list psfp; @@ -905,6 +913,10 @@ int ocelot_mact_learn_streamdata(struct ocelot *ocelot, int dst_idx, enum macaccess_entry_type type, int sfid, int ssid); +int ocelot_vcap_policer_add(struct ocelot *ocelot, u32 pol_ix, + struct ocelot_policer *pol); +int ocelot_vcap_policer_del(struct ocelot *ocelot, u32 pol_ix); + #if IS_ENABLED(CONFIG_BRIDGE_MRP) int ocelot_mrp_add(struct ocelot *ocelot, int port, const struct switchdev_obj_mrp *mrp); From 76c13ede7120e55b5a31580d2231e30b26246212 Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:12:03 +0800 Subject: [PATCH 0332/2048] net: dsa: felix: use vcap policer to set flow meter for psfp This patch add police action to set flow meter table which is defined in IEEE802.1Qci. Flow metering is two rates two buckets and three color marker to policing the frames, we only enable one rate one bucket in this patch. Flow metering shares a same policer pool with VCAP policers, so the PSFP policer calls ocelot_vcap_policer_add() and ocelot_vcap_policer_del() to set flow meter police. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/dsa/ocelot/felix_vsc9959.c | 32 +++++++++++++++++++++++++- 1 file changed, 31 insertions(+), 1 deletion(-) diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index f8d770384344..eb6c05f29883 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -1341,6 +1341,7 @@ static int vsc9959_port_setup_tc(struct dsa_switch *ds, int port, #define VSC9959_PSFP_SFID_MAX 175 #define VSC9959_PSFP_GATE_ID_MAX 183 +#define VSC9959_PSFP_POLICER_BASE 63 #define VSC9959_PSFP_POLICER_MAX 383 #define VSC9959_PSFP_GATE_LIST_NUM 4 #define VSC9959_PSFP_GATE_CYCLETIME_MIN 5000 @@ -1841,7 +1842,10 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, struct felix_stream stream = {0}; struct felix_stream_gate *sgi; struct ocelot_psfp_list *psfp; + struct ocelot_policer pol; int ret, i, size; + u64 rate, burst; + u32 index; psfp = &ocelot->psfp; @@ -1860,13 +1864,33 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, ret = vsc9959_psfp_sgi_table_add(ocelot, sgi); if (ret) { kfree(sgi); - return ret; + goto err; } sfi.sg_valid = 1; sfi.sgid = sgi->index; kfree(sgi); break; case FLOW_ACTION_POLICE: + index = a->police.index + VSC9959_PSFP_POLICER_BASE; + if (index > VSC9959_PSFP_POLICER_MAX) { + ret = -EINVAL; + goto err; + } + + rate = a->police.rate_bytes_ps; + burst = rate * PSCHED_NS2TICKS(a->police.burst); + pol = (struct ocelot_policer) { + .burst = div_u64(burst, PSCHED_TICKS_PER_SEC), + .rate = div_u64(rate, 1000) * 8, + }; + ret = ocelot_vcap_policer_add(ocelot, index, &pol); + if (ret) + goto err; + + sfi.fm_valid = 1; + sfi.fmid = index; + sfi.maxsdu = a->police.mtu; + break; default: return -EOPNOTSUPP; } @@ -1903,6 +1927,9 @@ err: if (sfi.sg_valid) vsc9959_psfp_sgi_table_del(ocelot, sfi.sgid); + if (sfi.fm_valid) + ocelot_vcap_policer_del(ocelot, sfi.fmid); + return ret; } @@ -1926,6 +1953,9 @@ static int vsc9959_psfp_filter_del(struct ocelot *ocelot, if (sfi->sg_valid) vsc9959_psfp_sgi_table_del(ocelot, sfi->sgid); + if (sfi->fm_valid) + ocelot_vcap_policer_del(ocelot, sfi->fmid); + vsc9959_psfp_sfi_table_del(ocelot, stream->sfid); stream->sfid_valid = 0; From a7e13edf37beee65f2c2ec60c42e5fb89a2958ce Mon Sep 17 00:00:00 2001 From: Xiaoliang Yang Date: Thu, 18 Nov 2021 18:12:04 +0800 Subject: [PATCH 0333/2048] net: dsa: felix: restrict psfp rules on ingress port PSFP rules take effect on the streams from any port of VSC9959 switch. This patch use ingress port to limit the rule only active on this port. Each stream can only match two ingress source ports in VSC9959. Streams from lowest port gets the configuration of SFID pointed by MAC Table lookup and streams from highest port gets the configuration of (SFID+1) pointed by MAC Table lookup. This patch defines the PSFP rule on highest port as dummy rule, which means that it does not modify the MAC table. Signed-off-by: Xiaoliang Yang Signed-off-by: David S. Miller --- drivers/net/dsa/ocelot/felix_vsc9959.c | 220 +++++++++++++++++----- drivers/net/ethernet/mscc/ocelot_flower.c | 2 +- include/soc/mscc/ocelot.h | 3 +- 3 files changed, 178 insertions(+), 47 deletions(-) diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index eb6c05f29883..42ac1952b39a 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -1349,6 +1349,9 @@ static int vsc9959_port_setup_tc(struct dsa_switch *ds, int port, struct felix_stream { struct list_head list; unsigned long id; + bool dummy; + int ports; + int port; u8 dmac[ETH_ALEN]; u16 vid; s8 prio; @@ -1363,6 +1366,7 @@ struct felix_stream_filter { refcount_t refcount; u32 index; u8 enable; + int portmask; u8 sg_valid; u32 sgid; u8 fm_valid; @@ -1505,10 +1509,12 @@ static int vsc9959_stream_table_add(struct ocelot *ocelot, memcpy(stream_entry, stream, sizeof(*stream_entry)); - ret = vsc9959_mact_stream_set(ocelot, stream_entry, extack); - if (ret) { - kfree(stream_entry); - return ret; + if (!stream->dummy) { + ret = vsc9959_mact_stream_set(ocelot, stream_entry, extack); + if (ret) { + kfree(stream_entry); + return ret; + } } list_add_tail(&stream_entry->list, stream_list); @@ -1531,7 +1537,8 @@ vsc9959_stream_table_get(struct list_head *stream_list, unsigned long id) static void vsc9959_stream_table_del(struct ocelot *ocelot, struct felix_stream *stream) { - vsc9959_mact_stream_set(ocelot, stream, NULL); + if (!stream->dummy) + vsc9959_mact_stream_set(ocelot, stream, NULL); list_del(&stream->list); kfree(stream); @@ -1586,36 +1593,37 @@ static int vsc9959_psfp_sfi_set(struct ocelot *ocelot, 10, 100000); } -static int vsc9959_psfp_sfi_table_add(struct ocelot *ocelot, - struct felix_stream_filter *sfi) +static int vsc9959_psfp_sfidmask_set(struct ocelot *ocelot, u32 sfid, int ports) { - struct felix_stream_filter *sfi_entry, *tmp; - struct list_head *pos, *q, *last; - struct ocelot_psfp_list *psfp; - u32 insert = 0; + u32 val; + + ocelot_rmw(ocelot, + ANA_TABLES_SFIDTIDX_SFID_INDEX(sfid), + ANA_TABLES_SFIDTIDX_SFID_INDEX_M, + ANA_TABLES_SFIDTIDX); + + ocelot_write(ocelot, + ANA_TABLES_SFID_MASK_IGR_PORT_MASK(ports) | + ANA_TABLES_SFID_MASK_IGR_SRCPORT_MATCH_ENA, + ANA_TABLES_SFID_MASK); + + ocelot_rmw(ocelot, + ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(SFIDACCESS_CMD_WRITE), + ANA_TABLES_SFIDACCESS_SFID_TBL_CMD_M, + ANA_TABLES_SFIDACCESS); + + return readx_poll_timeout(vsc9959_sfi_access_status, ocelot, val, + (!ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(val)), + 10, 100000); +} + +static int vsc9959_psfp_sfi_list_add(struct ocelot *ocelot, + struct felix_stream_filter *sfi, + struct list_head *pos) +{ + struct felix_stream_filter *sfi_entry; int ret; - psfp = &ocelot->psfp; - last = &psfp->sfi_list; - - list_for_each_safe(pos, q, &psfp->sfi_list) { - tmp = list_entry(pos, struct felix_stream_filter, list); - if (sfi->sg_valid == tmp->sg_valid && - sfi->fm_valid == tmp->fm_valid && - tmp->sgid == sfi->sgid && - tmp->fmid == sfi->fmid) { - sfi->index = tmp->index; - refcount_inc(&tmp->refcount); - return 0; - } - /* Make sure that the index is increasing in order. */ - if (tmp->index == insert) { - last = pos; - insert++; - } - } - sfi->index = insert; - sfi_entry = kzalloc(sizeof(*sfi_entry), GFP_KERNEL); if (!sfi_entry) return -ENOMEM; @@ -1629,11 +1637,79 @@ static int vsc9959_psfp_sfi_table_add(struct ocelot *ocelot, return ret; } - list_add(&sfi_entry->list, last); + vsc9959_psfp_sfidmask_set(ocelot, sfi->index, sfi->portmask); + + list_add(&sfi_entry->list, pos); return 0; } +static int vsc9959_psfp_sfi_table_add(struct ocelot *ocelot, + struct felix_stream_filter *sfi) +{ + struct list_head *pos, *q, *last; + struct felix_stream_filter *tmp; + struct ocelot_psfp_list *psfp; + u32 insert = 0; + + psfp = &ocelot->psfp; + last = &psfp->sfi_list; + + list_for_each_safe(pos, q, &psfp->sfi_list) { + tmp = list_entry(pos, struct felix_stream_filter, list); + if (sfi->sg_valid == tmp->sg_valid && + sfi->fm_valid == tmp->fm_valid && + sfi->portmask == tmp->portmask && + tmp->sgid == sfi->sgid && + tmp->fmid == sfi->fmid) { + sfi->index = tmp->index; + refcount_inc(&tmp->refcount); + return 0; + } + /* Make sure that the index is increasing in order. */ + if (tmp->index == insert) { + last = pos; + insert++; + } + } + sfi->index = insert; + + return vsc9959_psfp_sfi_list_add(ocelot, sfi, last); +} + +static int vsc9959_psfp_sfi_table_add2(struct ocelot *ocelot, + struct felix_stream_filter *sfi, + struct felix_stream_filter *sfi2) +{ + struct felix_stream_filter *tmp; + struct list_head *pos, *q, *last; + struct ocelot_psfp_list *psfp; + u32 insert = 0; + int ret; + + psfp = &ocelot->psfp; + last = &psfp->sfi_list; + + list_for_each_safe(pos, q, &psfp->sfi_list) { + tmp = list_entry(pos, struct felix_stream_filter, list); + /* Make sure that the index is increasing in order. */ + if (tmp->index >= insert + 2) + break; + + insert = tmp->index + 1; + last = pos; + } + sfi->index = insert; + + ret = vsc9959_psfp_sfi_list_add(ocelot, sfi, last); + if (ret) + return ret; + + sfi2->index = insert + 1; + + return vsc9959_psfp_sfi_list_add(ocelot, sfi2, last->next); +} + static struct felix_stream_filter * vsc9959_psfp_sfi_table_get(struct list_head *sfi_list, u32 index) { @@ -1832,10 +1908,11 @@ static void vsc9959_psfp_counters_get(struct ocelot *ocelot, u32 index, SYS_STAT_CFG); } -static int vsc9959_psfp_filter_add(struct ocelot *ocelot, +static int vsc9959_psfp_filter_add(struct ocelot *ocelot, int port, struct flow_cls_offload *f) { struct netlink_ext_ack *extack = f->common.extack; + struct felix_stream_filter old_sfi, *sfi_entry; struct felix_stream_filter sfi = {0}; const struct flow_action_entry *a; struct felix_stream *stream_entry; @@ -1896,21 +1973,61 @@ static int vsc9959_psfp_filter_add(struct ocelot *ocelot, } } - /* Check if stream is set. */ - stream_entry = vsc9959_stream_table_lookup(&psfp->stream_list, &stream); - if (stream_entry) { - NL_SET_ERR_MSG_MOD(extack, "This stream is already added"); - ret = -EEXIST; - goto err; - } + stream.ports = BIT(port); + stream.port = port; + sfi.portmask = stream.ports; sfi.prio_valid = (stream.prio < 0 ? 0 : 1); sfi.prio = (sfi.prio_valid ? stream.prio : 0); sfi.enable = 1; - ret = vsc9959_psfp_sfi_table_add(ocelot, &sfi); - if (ret) - goto err; + /* Check if stream is set. */ + stream_entry = vsc9959_stream_table_lookup(&psfp->stream_list, &stream); + if (stream_entry) { + if (stream_entry->ports & BIT(port)) { + NL_SET_ERR_MSG_MOD(extack, + "The stream is added on this port"); + ret = -EEXIST; + goto err; + } + + if (stream_entry->ports != BIT(stream_entry->port)) { + NL_SET_ERR_MSG_MOD(extack, + "The stream is added on two ports"); + ret = -EEXIST; + goto err; + } + + stream_entry->ports |= BIT(port); + stream.ports = stream_entry->ports; + + sfi_entry = vsc9959_psfp_sfi_table_get(&psfp->sfi_list, + stream_entry->sfid); + memcpy(&old_sfi, sfi_entry, sizeof(old_sfi)); + + vsc9959_psfp_sfi_table_del(ocelot, stream_entry->sfid); + + old_sfi.portmask = stream_entry->ports; + sfi.portmask = stream.ports; + + if (stream_entry->port > port) { + ret = vsc9959_psfp_sfi_table_add2(ocelot, &sfi, + &old_sfi); + stream_entry->dummy = true; + } else { + ret = vsc9959_psfp_sfi_table_add2(ocelot, &old_sfi, + &sfi); + stream.dummy = true; + } + if (ret) + goto err; + + stream_entry->sfid = old_sfi.index; + } else { + ret = vsc9959_psfp_sfi_table_add(ocelot, &sfi); + if (ret) + goto err; + } stream.sfid = sfi.index; stream.sfid_valid = 1; @@ -1936,9 +2053,9 @@ err: static int vsc9959_psfp_filter_del(struct ocelot *ocelot, struct flow_cls_offload *f) { + struct felix_stream *stream, tmp, *stream_entry; static struct felix_stream_filter *sfi; struct ocelot_psfp_list *psfp; - struct felix_stream *stream; psfp = &ocelot->psfp; @@ -1958,9 +2075,22 @@ static int vsc9959_psfp_filter_del(struct ocelot *ocelot, vsc9959_psfp_sfi_table_del(ocelot, stream->sfid); + memcpy(&tmp, stream, sizeof(tmp)); + stream->sfid_valid = 0; vsc9959_stream_table_del(ocelot, stream); + stream_entry = vsc9959_stream_table_lookup(&psfp->stream_list, &tmp); + if (stream_entry) { + stream_entry->ports = BIT(stream_entry->port); + if (stream_entry->dummy) { + stream_entry->dummy = false; + vsc9959_mact_stream_set(ocelot, stream_entry, NULL); + } + vsc9959_psfp_sfidmask_set(ocelot, stream_entry->sfid, + stream_entry->ports); + } + return 0; } diff --git a/drivers/net/ethernet/mscc/ocelot_flower.c b/drivers/net/ethernet/mscc/ocelot_flower.c index b54b52fd9e1b..58fce173f95b 100644 --- a/drivers/net/ethernet/mscc/ocelot_flower.c +++ b/drivers/net/ethernet/mscc/ocelot_flower.c @@ -837,7 +837,7 @@ int ocelot_cls_flower_replace(struct ocelot *ocelot, int port, if (filter->type == OCELOT_PSFP_FILTER_OFFLOAD) { kfree(filter); if (ocelot->ops->psfp_filter_add) - return ocelot->ops->psfp_filter_add(ocelot, f); + return ocelot->ops->psfp_filter_add(ocelot, port, f); NL_SET_ERR_MSG_MOD(extack, "PSFP chain is not supported in HW"); return -EOPNOTSUPP; diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h index 2a41685b5c7d..89d17629efe5 100644 --- a/include/soc/mscc/ocelot.h +++ b/include/soc/mscc/ocelot.h @@ -556,7 +556,8 @@ struct ocelot_ops { u16 (*wm_dec)(u16 value); void (*wm_stat)(u32 val, u32 *inuse, u32 *maxuse); void (*psfp_init)(struct ocelot *ocelot); - int (*psfp_filter_add)(struct ocelot *ocelot, struct flow_cls_offload *f); + int (*psfp_filter_add)(struct ocelot *ocelot, int port, + struct flow_cls_offload *f); int (*psfp_filter_del)(struct ocelot *ocelot, struct flow_cls_offload *f); int (*psfp_stats_get)(struct ocelot *ocelot, struct flow_cls_offload *f, struct flow_stats *stats); From a6df95cae40bee555e01a37b4023ce8e97ffa249 Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:34 +0000 Subject: [PATCH 0334/2048] lan78xx: Fix memory allocation bug Fix memory allocation that fails to check for NULL return. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 38 +++++++++++++++++++++----------------- 1 file changed, 21 insertions(+), 17 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index f20376c1ef3f..3ddacc6239a3 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -4106,18 +4106,20 @@ static int lan78xx_probe(struct usb_interface *intf, period = ep_intr->desc.bInterval; maxp = usb_maxpacket(dev->udev, dev->pipe_intr, 0); buf = kmalloc(maxp, GFP_KERNEL); - if (buf) { - dev->urb_intr = usb_alloc_urb(0, GFP_KERNEL); - if (!dev->urb_intr) { - ret = -ENOMEM; - kfree(buf); - goto out3; - } else { - usb_fill_int_urb(dev->urb_intr, dev->udev, - dev->pipe_intr, buf, maxp, - intr_complete, dev, period); - dev->urb_intr->transfer_flags |= URB_FREE_BUFFER; - } + if (!buf) { + ret = -ENOMEM; + goto out3; + } + + dev->urb_intr = usb_alloc_urb(0, GFP_KERNEL); + if (!dev->urb_intr) { + ret = -ENOMEM; + goto out4; + } else { + usb_fill_int_urb(dev->urb_intr, dev->udev, + dev->pipe_intr, buf, maxp, + intr_complete, dev, period); + dev->urb_intr->transfer_flags |= URB_FREE_BUFFER; } dev->maxpacket = usb_maxpacket(dev->udev, dev->pipe_out, 1); @@ -4125,7 +4127,7 @@ static int lan78xx_probe(struct usb_interface *intf, /* Reject broken descriptors. */ if (dev->maxpacket == 0) { ret = -ENODEV; - goto out4; + goto out5; } /* driver requires remote-wakeup capability during autosuspend. */ @@ -4133,12 +4135,12 @@ static int lan78xx_probe(struct usb_interface *intf, ret = lan78xx_phy_init(dev); if (ret < 0) - goto out4; + goto out5; ret = register_netdev(netdev); if (ret != 0) { netif_err(dev, probe, netdev, "couldn't register the device\n"); - goto out5; + goto out6; } usb_set_intfdata(intf, dev); @@ -4153,10 +4155,12 @@ static int lan78xx_probe(struct usb_interface *intf, return 0; -out5: +out6: phy_disconnect(netdev->phydev); -out4: +out5: usb_free_urb(dev->urb_intr); +out4: + kfree(buf); out3: lan78xx_unbind(dev, intf); out2: From d383216a7efec3b38b8bd4925a208383b2492d84 Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:35 +0000 Subject: [PATCH 0335/2048] lan78xx: Introduce Tx URB processing improvements This patch introduces a new approach to allocating and managing Tx URBs that contributes to improving driver throughput and reducing CPU load. A pool of Tx URBs is created during driver instantiation. A URB is allocated from the pool when there is data to transmit. The URB is released back to the pool when the data has been transmitted by the device. The default URB buffer size is different for each USB bus speed. The chosen sizes provide good USB utilisation with little impact on overall packet latency. SKBs to be transmitted are added to a pending queue for processing. The driver tracks the available Tx URB buffer space and copies as much pending data as possible into each free URB. Each full URB is then submitted to the USB host controller for transmission. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 678 ++++++++++++++++++++++++++------------ 1 file changed, 463 insertions(+), 215 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index 3ddacc6239a3..7187aac01e7e 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -68,6 +68,7 @@ #define DEFAULT_VLAN_FILTER_ENABLE (true) #define DEFAULT_VLAN_RX_OFFLOAD (true) #define TX_OVERHEAD (8) +#define TX_ALIGNMENT (4) #define RXW_PADDING 2 #define LAN78XX_USB_VENDOR_ID (0x0424) @@ -90,6 +91,21 @@ WAKE_MCAST | WAKE_BCAST | \ WAKE_ARP | WAKE_MAGIC) +#define TX_URB_NUM 10 +#define TX_SS_URB_NUM TX_URB_NUM +#define TX_HS_URB_NUM TX_URB_NUM +#define TX_FS_URB_NUM TX_URB_NUM + +/* A single URB buffer must be large enough to hold a complete jumbo packet + */ +#define TX_SS_URB_SIZE (32 * 1024) +#define TX_HS_URB_SIZE (16 * 1024) +#define TX_FS_URB_SIZE (10 * 1024) + +#define TX_CMD_LEN 8 +#define TX_SKB_MIN_LEN (TX_CMD_LEN + ETH_HLEN) +#define LAN78XX_TSO_SIZE(dev) ((dev)->tx_urb_size - TX_SKB_MIN_LEN) + /* USB related defines */ #define BULK_IN_PIPE 1 #define BULK_OUT_PIPE 2 @@ -385,11 +401,15 @@ struct lan78xx_net { struct usb_interface *intf; void *driver_priv; + unsigned int tx_pend_data_len; + size_t n_tx_urbs; + size_t tx_urb_size; + int rx_qlen; - int tx_qlen; struct sk_buff_head rxq; - struct sk_buff_head txq; struct sk_buff_head done; + struct sk_buff_head txq_free; + struct sk_buff_head txq; struct sk_buff_head txq_pend; struct tasklet_struct bh; @@ -443,6 +463,107 @@ static int msg_level = -1; module_param(msg_level, int, 0); MODULE_PARM_DESC(msg_level, "Override default message level"); +static struct sk_buff *lan78xx_get_buf(struct sk_buff_head *buf_pool) +{ + if (skb_queue_empty(buf_pool)) + return NULL; + + return skb_dequeue(buf_pool); +} + +static void lan78xx_release_buf(struct sk_buff_head *buf_pool, + struct sk_buff *buf) +{ + buf->data = buf->head; + skb_reset_tail_pointer(buf); + + buf->len = 0; + buf->data_len = 0; + + skb_queue_tail(buf_pool, buf); +} + +static void lan78xx_free_buf_pool(struct sk_buff_head *buf_pool) +{ + struct skb_data *entry; + struct sk_buff *buf; + + while (!skb_queue_empty(buf_pool)) { + buf = skb_dequeue(buf_pool); + if (buf) { + entry = (struct skb_data *)buf->cb; + usb_free_urb(entry->urb); + dev_kfree_skb_any(buf); + } + } +} + +static int lan78xx_alloc_buf_pool(struct sk_buff_head *buf_pool, + size_t n_urbs, size_t urb_size, + struct lan78xx_net *dev) +{ + struct skb_data *entry; + struct sk_buff *buf; + struct urb *urb; + int i; + + skb_queue_head_init(buf_pool); + + for (i = 0; i < n_urbs; i++) { + buf = alloc_skb(urb_size, GFP_ATOMIC); + if (!buf) + goto error; + + if (skb_linearize(buf) != 0) { + dev_kfree_skb_any(buf); + goto error; + } + + urb = usb_alloc_urb(0, GFP_ATOMIC); + if (!urb) { + dev_kfree_skb_any(buf); + goto error; + } + + entry = (struct skb_data *)buf->cb; + entry->urb = urb; + entry->dev = dev; + entry->length = 0; + entry->num_of_packet = 0; + + skb_queue_tail(buf_pool, buf); + } + + return 0; + +error: + lan78xx_free_buf_pool(buf_pool); + + return -ENOMEM; +} + +static struct sk_buff *lan78xx_get_tx_buf(struct lan78xx_net *dev) +{ + return lan78xx_get_buf(&dev->txq_free); +} + +static void lan78xx_release_tx_buf(struct lan78xx_net *dev, + struct sk_buff *tx_buf) +{ + lan78xx_release_buf(&dev->txq_free, tx_buf); +} + +static void lan78xx_free_tx_resources(struct lan78xx_net *dev) +{ + lan78xx_free_buf_pool(&dev->txq_free); +} + +static int lan78xx_alloc_tx_resources(struct lan78xx_net *dev) +{ + return lan78xx_alloc_buf_pool(&dev->txq_free, + dev->n_tx_urbs, dev->tx_urb_size, dev); +} + static int lan78xx_read_reg(struct lan78xx_net *dev, u32 index, u32 *data) { u32 *buf; @@ -2557,6 +2678,32 @@ static void lan78xx_init_ltm(struct lan78xx_net *dev) lan78xx_write_reg(dev, LTM_INACTIVE1, regs[5]); } +static int lan78xx_urb_config_init(struct lan78xx_net *dev) +{ + int result = 0; + + switch (dev->udev->speed) { + case USB_SPEED_SUPER: + dev->tx_urb_size = TX_SS_URB_SIZE; + dev->n_tx_urbs = TX_SS_URB_NUM; + break; + case USB_SPEED_HIGH: + dev->tx_urb_size = TX_HS_URB_SIZE; + dev->n_tx_urbs = TX_HS_URB_NUM; + break; + case USB_SPEED_FULL: + dev->tx_urb_size = TX_FS_URB_SIZE; + dev->n_tx_urbs = TX_FS_URB_NUM; + break; + default: + netdev_warn(dev->net, "USB bus speed not supported\n"); + result = -EIO; + break; + } + + return result; +} + static int lan78xx_start_hw(struct lan78xx_net *dev, u32 reg, u32 hw_enable) { return lan78xx_update_reg(dev, reg, hw_enable, hw_enable); @@ -2768,17 +2915,14 @@ static int lan78xx_reset(struct lan78xx_net *dev) buf = DEFAULT_BURST_CAP_SIZE / SS_USB_PKT_SIZE; dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; dev->rx_qlen = 4; - dev->tx_qlen = 4; } else if (dev->udev->speed == USB_SPEED_HIGH) { buf = DEFAULT_BURST_CAP_SIZE / HS_USB_PKT_SIZE; dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; dev->rx_qlen = RX_MAX_QUEUE_MEMORY / dev->rx_urb_size; - dev->tx_qlen = RX_MAX_QUEUE_MEMORY / dev->hard_mtu; } else { buf = DEFAULT_BURST_CAP_SIZE / FS_USB_PKT_SIZE; dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; dev->rx_qlen = 4; - dev->tx_qlen = 4; } ret = lan78xx_write_reg(dev, BURST_CAP, buf); @@ -3020,6 +3164,8 @@ static void lan78xx_terminate_urbs(struct lan78xx_net *dev) usb_free_urb(entry->urb); dev_kfree_skb(skb); } + + skb_queue_purge(&dev->txq_pend); } static int lan78xx_stop(struct net_device *net) @@ -3071,48 +3217,6 @@ static int lan78xx_stop(struct net_device *net) return 0; } -static struct sk_buff *lan78xx_tx_prep(struct lan78xx_net *dev, - struct sk_buff *skb, gfp_t flags) -{ - u32 tx_cmd_a, tx_cmd_b; - void *ptr; - - if (skb_cow_head(skb, TX_OVERHEAD)) { - dev_kfree_skb_any(skb); - return NULL; - } - - if (skb_linearize(skb)) { - dev_kfree_skb_any(skb); - return NULL; - } - - tx_cmd_a = (u32)(skb->len & TX_CMD_A_LEN_MASK_) | TX_CMD_A_FCS_; - - if (skb->ip_summed == CHECKSUM_PARTIAL) - tx_cmd_a |= TX_CMD_A_IPE_ | TX_CMD_A_TPE_; - - tx_cmd_b = 0; - if (skb_is_gso(skb)) { - u16 mss = max(skb_shinfo(skb)->gso_size, TX_CMD_B_MSS_MIN_); - - tx_cmd_b = (mss << TX_CMD_B_MSS_SHIFT_) & TX_CMD_B_MSS_MASK_; - - tx_cmd_a |= TX_CMD_A_LSO_; - } - - if (skb_vlan_tag_present(skb)) { - tx_cmd_a |= TX_CMD_A_IVTG_; - tx_cmd_b |= skb_vlan_tag_get(skb) & TX_CMD_B_VTAG_MASK_; - } - - ptr = skb_push(skb, 8); - put_unaligned_le32(tx_cmd_a, ptr); - put_unaligned_le32(tx_cmd_b, ptr + 4); - - return skb; -} - static enum skb_state defer_bh(struct lan78xx_net *dev, struct sk_buff *skb, struct sk_buff_head *list, enum skb_state state) { @@ -3146,7 +3250,7 @@ static void tx_complete(struct urb *urb) dev->net->stats.tx_packets += entry->num_of_packet; dev->net->stats.tx_bytes += entry->length; } else { - dev->net->stats.tx_errors++; + dev->net->stats.tx_errors += entry->num_of_packet; switch (urb->status) { case -EPIPE: @@ -3179,7 +3283,15 @@ static void tx_complete(struct urb *urb) usb_autopm_put_interface_async(dev->intf); - defer_bh(dev, skb, &dev->txq, tx_done); + skb_unlink(skb, &dev->txq); + + lan78xx_release_tx_buf(dev, skb); + + /* Re-schedule tasklet if Tx data pending but no URBs in progress. + */ + if (skb_queue_empty(&dev->txq) && + !skb_queue_empty(&dev->txq_pend)) + tasklet_schedule(&dev->bh); } static void lan78xx_queue_skb(struct sk_buff_head *list, @@ -3191,36 +3303,97 @@ static void lan78xx_queue_skb(struct sk_buff_head *list, entry->state = state; } +static unsigned int lan78xx_tx_urb_space(struct lan78xx_net *dev) +{ + return skb_queue_len(&dev->txq_free) * dev->tx_urb_size; +} + +static unsigned int lan78xx_tx_pend_data_len(struct lan78xx_net *dev) +{ + return dev->tx_pend_data_len; +} + +static void lan78xx_tx_pend_skb_add(struct lan78xx_net *dev, + struct sk_buff *skb, + unsigned int *tx_pend_data_len) +{ + unsigned long flags; + + spin_lock_irqsave(&dev->txq_pend.lock, flags); + + __skb_queue_tail(&dev->txq_pend, skb); + + dev->tx_pend_data_len += skb->len; + *tx_pend_data_len = dev->tx_pend_data_len; + + spin_unlock_irqrestore(&dev->txq_pend.lock, flags); +} + +static void lan78xx_tx_pend_skb_head_add(struct lan78xx_net *dev, + struct sk_buff *skb, + unsigned int *tx_pend_data_len) +{ + unsigned long flags; + + spin_lock_irqsave(&dev->txq_pend.lock, flags); + + __skb_queue_head(&dev->txq_pend, skb); + + dev->tx_pend_data_len += skb->len; + *tx_pend_data_len = dev->tx_pend_data_len; + + spin_unlock_irqrestore(&dev->txq_pend.lock, flags); +} + +static void lan78xx_tx_pend_skb_get(struct lan78xx_net *dev, + struct sk_buff **skb, + unsigned int *tx_pend_data_len) +{ + unsigned long flags; + + spin_lock_irqsave(&dev->txq_pend.lock, flags); + + *skb = __skb_dequeue(&dev->txq_pend); + if (*skb) + dev->tx_pend_data_len -= (*skb)->len; + *tx_pend_data_len = dev->tx_pend_data_len; + + spin_unlock_irqrestore(&dev->txq_pend.lock, flags); +} + static netdev_tx_t lan78xx_start_xmit(struct sk_buff *skb, struct net_device *net) { struct lan78xx_net *dev = netdev_priv(net); - struct sk_buff *skb2 = NULL; + unsigned int tx_pend_data_len; if (test_bit(EVENT_DEV_ASLEEP, &dev->flags)) schedule_delayed_work(&dev->wq, 0); - if (skb) { - skb_tx_timestamp(skb); - skb2 = lan78xx_tx_prep(dev, skb, GFP_ATOMIC); + skb_tx_timestamp(skb); + + lan78xx_tx_pend_skb_add(dev, skb, &tx_pend_data_len); + + /* Set up a Tx URB if none is in progress */ + + if (skb_queue_empty(&dev->txq)) + tasklet_schedule(&dev->bh); + + /* Stop stack Tx queue if we have enough data to fill + * all the free Tx URBs. + */ + if (tx_pend_data_len > lan78xx_tx_urb_space(dev)) { + netif_stop_queue(net); + + netif_dbg(dev, hw, dev->net, "tx data len: %u, urb space %u", + tx_pend_data_len, lan78xx_tx_urb_space(dev)); + + /* Kick off transmission of pending data */ + + if (!skb_queue_empty(&dev->txq_free)) + tasklet_schedule(&dev->bh); } - if (skb2) { - skb_queue_tail(&dev->txq_pend, skb2); - - /* throttle TX patch at slower than SUPER SPEED USB */ - if ((dev->udev->speed < USB_SPEED_SUPER) && - (skb_queue_len(&dev->txq_pend) > 10)) - netif_stop_queue(net); - } else { - netif_dbg(dev, tx_err, dev->net, - "lan78xx_tx_prep return NULL\n"); - dev->net->stats.tx_errors++; - dev->net->stats.tx_dropped++; - } - - tasklet_schedule(&dev->bh); - return NETDEV_TX_OK; } @@ -3600,139 +3773,191 @@ static void rx_complete(struct urb *urb) netif_dbg(dev, rx_err, dev->net, "no read resubmitted\n"); } +static void lan78xx_fill_tx_cmd_words(struct sk_buff *skb, u8 *buffer) +{ + u32 tx_cmd_a; + u32 tx_cmd_b; + + tx_cmd_a = (u32)(skb->len & TX_CMD_A_LEN_MASK_) | TX_CMD_A_FCS_; + + if (skb->ip_summed == CHECKSUM_PARTIAL) + tx_cmd_a |= TX_CMD_A_IPE_ | TX_CMD_A_TPE_; + + tx_cmd_b = 0; + if (skb_is_gso(skb)) { + u16 mss = max(skb_shinfo(skb)->gso_size, TX_CMD_B_MSS_MIN_); + + tx_cmd_b = (mss << TX_CMD_B_MSS_SHIFT_) & TX_CMD_B_MSS_MASK_; + + tx_cmd_a |= TX_CMD_A_LSO_; + } + + if (skb_vlan_tag_present(skb)) { + tx_cmd_a |= TX_CMD_A_IVTG_; + tx_cmd_b |= skb_vlan_tag_get(skb) & TX_CMD_B_VTAG_MASK_; + } + + put_unaligned_le32(tx_cmd_a, buffer); + put_unaligned_le32(tx_cmd_b, buffer + 4); +} + +static struct skb_data *lan78xx_tx_buf_fill(struct lan78xx_net *dev, + struct sk_buff *tx_buf) +{ + struct skb_data *entry = (struct skb_data *)tx_buf->cb; + int remain = dev->tx_urb_size; + u8 *tx_data = tx_buf->data; + u32 urb_len = 0; + + entry->num_of_packet = 0; + entry->length = 0; + + /* Work through the pending SKBs and copy the data of each SKB into + * the URB buffer if there room for all the SKB data. + * + * There must be at least DST+SRC+TYPE in the SKB (with padding enabled) + */ + while (remain >= TX_SKB_MIN_LEN) { + unsigned int pending_bytes; + unsigned int align_bytes; + struct sk_buff *skb; + unsigned int len; + + lan78xx_tx_pend_skb_get(dev, &skb, &pending_bytes); + + if (!skb) + break; + + align_bytes = (TX_ALIGNMENT - (urb_len % TX_ALIGNMENT)) % + TX_ALIGNMENT; + len = align_bytes + TX_CMD_LEN + skb->len; + if (len > remain) { + lan78xx_tx_pend_skb_head_add(dev, skb, &pending_bytes); + break; + } + + tx_data += align_bytes; + + lan78xx_fill_tx_cmd_words(skb, tx_data); + tx_data += TX_CMD_LEN; + + len = skb->len; + if (skb_copy_bits(skb, 0, tx_data, len) < 0) { + struct net_device_stats *stats = &dev->net->stats; + + stats->tx_dropped++; + dev_kfree_skb_any(skb); + tx_data -= TX_CMD_LEN; + continue; + } + + tx_data += len; + entry->length += len; + entry->num_of_packet += skb_shinfo(skb)->gso_segs ?: 1; + + dev_kfree_skb_any(skb); + + urb_len = (u32)(tx_data - (u8 *)tx_buf->data); + + remain = dev->tx_urb_size - urb_len; + } + + skb_put(tx_buf, urb_len); + + return entry; +} + static void lan78xx_tx_bh(struct lan78xx_net *dev) { - int length; - struct urb *urb = NULL; - struct skb_data *entry; - unsigned long flags; - struct sk_buff_head *tqp = &dev->txq_pend; - struct sk_buff *skb, *skb2; int ret; - int count, pos; - int skb_totallen, pkt_cnt; - skb_totallen = 0; - pkt_cnt = 0; - count = 0; - length = 0; - spin_lock_irqsave(&tqp->lock, flags); - skb_queue_walk(tqp, skb) { - if (skb_is_gso(skb)) { - if (!skb_queue_is_first(tqp, skb)) { - /* handle previous packets first */ - break; - } - count = 1; - length = skb->len - TX_OVERHEAD; - __skb_unlink(skb, tqp); - spin_unlock_irqrestore(&tqp->lock, flags); - goto gso_skb; - } + /* Start the stack Tx queue if it was stopped + */ + netif_tx_lock(dev->net); + if (netif_queue_stopped(dev->net)) { + if (lan78xx_tx_pend_data_len(dev) < lan78xx_tx_urb_space(dev)) + netif_wake_queue(dev->net); + } + netif_tx_unlock(dev->net); - if ((skb_totallen + skb->len) > MAX_SINGLE_PACKET_SIZE) + /* Go through the Tx pending queue and set up URBs to transfer + * the data to the device. Stop if no more pending data or URBs, + * or if an error occurs when a URB is submitted. + */ + do { + struct skb_data *entry; + struct sk_buff *tx_buf; + unsigned long flags; + + if (skb_queue_empty(&dev->txq_pend)) break; - skb_totallen = skb->len + roundup(skb_totallen, sizeof(u32)); - pkt_cnt++; - } - spin_unlock_irqrestore(&tqp->lock, flags); - /* copy to a single skb */ - skb = alloc_skb(skb_totallen, GFP_ATOMIC); - if (!skb) - goto drop; + tx_buf = lan78xx_get_tx_buf(dev); + if (!tx_buf) + break; - skb_put(skb, skb_totallen); + entry = lan78xx_tx_buf_fill(dev, tx_buf); - for (count = pos = 0; count < pkt_cnt; count++) { - skb2 = skb_dequeue(tqp); - if (skb2) { - length += (skb2->len - TX_OVERHEAD); - memcpy(skb->data + pos, skb2->data, skb2->len); - pos += roundup(skb2->len, sizeof(u32)); - dev_kfree_skb(skb2); + spin_lock_irqsave(&dev->txq.lock, flags); + ret = usb_autopm_get_interface_async(dev->intf); + if (ret < 0) { + spin_unlock_irqrestore(&dev->txq.lock, flags); + goto out; } - } -gso_skb: - urb = usb_alloc_urb(0, GFP_ATOMIC); - if (!urb) - goto drop; + usb_fill_bulk_urb(entry->urb, dev->udev, dev->pipe_out, + tx_buf->data, tx_buf->len, tx_complete, + tx_buf); - entry = (struct skb_data *)skb->cb; - entry->urb = urb; - entry->dev = dev; - entry->length = length; - entry->num_of_packet = count; - - spin_lock_irqsave(&dev->txq.lock, flags); - ret = usb_autopm_get_interface_async(dev->intf); - if (ret < 0) { - spin_unlock_irqrestore(&dev->txq.lock, flags); - goto drop; - } - - usb_fill_bulk_urb(urb, dev->udev, dev->pipe_out, - skb->data, skb->len, tx_complete, skb); - - if (length % dev->maxpacket == 0) { - /* send USB_ZERO_PACKET */ - urb->transfer_flags |= URB_ZERO_PACKET; - } + if (tx_buf->len % dev->maxpacket == 0) { + /* send USB_ZERO_PACKET */ + entry->urb->transfer_flags |= URB_ZERO_PACKET; + } #ifdef CONFIG_PM - /* if this triggers the device is still a sleep */ - if (test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { - /* transmission will be done in resume */ - usb_anchor_urb(urb, &dev->deferred); - /* no use to process more packets */ - netif_stop_queue(dev->net); - usb_put_urb(urb); - spin_unlock_irqrestore(&dev->txq.lock, flags); - netdev_dbg(dev->net, "Delaying transmission for resumption\n"); - return; - } -#endif - - ret = usb_submit_urb(urb, GFP_ATOMIC); - switch (ret) { - case 0: - netif_trans_update(dev->net); - lan78xx_queue_skb(&dev->txq, skb, tx_start); - if (skb_queue_len(&dev->txq) >= dev->tx_qlen) + /* if device is asleep stop outgoing packet processing */ + if (test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { + usb_anchor_urb(entry->urb, &dev->deferred); netif_stop_queue(dev->net); - break; - case -EPIPE: - netif_stop_queue(dev->net); - lan78xx_defer_kevent(dev, EVENT_TX_HALT); - usb_autopm_put_interface_async(dev->intf); - break; - case -ENODEV: - case -ENOENT: - netif_dbg(dev, tx_err, dev->net, - "tx: submit urb err %d (disconnected?)", ret); - netif_device_detach(dev->net); - break; - default: - usb_autopm_put_interface_async(dev->intf); - netif_dbg(dev, tx_err, dev->net, - "tx: submit urb err %d\n", ret); - break; - } + spin_unlock_irqrestore(&dev->txq.lock, flags); + netdev_dbg(dev->net, + "Delaying transmission for resumption\n"); + return; + } +#endif + ret = usb_submit_urb(entry->urb, GFP_ATOMIC); + switch (ret) { + case 0: + netif_trans_update(dev->net); + lan78xx_queue_skb(&dev->txq, tx_buf, tx_start); + break; + case -EPIPE: + netif_stop_queue(dev->net); + lan78xx_defer_kevent(dev, EVENT_TX_HALT); + usb_autopm_put_interface_async(dev->intf); + break; + case -ENODEV: + case -ENOENT: + netif_dbg(dev, tx_err, dev->net, + "tx submit urb err %d (disconnected?)", ret); + netif_device_detach(dev->net); + break; + default: + usb_autopm_put_interface_async(dev->intf); + netif_dbg(dev, tx_err, dev->net, + "tx submit urb err %d\n", ret); + break; + } - spin_unlock_irqrestore(&dev->txq.lock, flags); + spin_unlock_irqrestore(&dev->txq.lock, flags); - if (ret) { - netif_dbg(dev, tx_err, dev->net, "drop, code %d\n", ret); -drop: - dev->net->stats.tx_dropped++; - if (skb) - dev_kfree_skb_any(skb); - usb_free_urb(urb); - } else { - netif_dbg(dev, tx_queued, dev->net, - "> tx, len %d, type 0x%x\n", length, skb->protocol); - } + if (ret) { + netdev_warn(dev->net, "failed to tx urb %d\n", ret); +out: + dev->net->stats.tx_dropped += entry->num_of_packet; + lan78xx_release_tx_buf(dev, tx_buf); + } + } while (ret == 0); } static void lan78xx_rx_bh(struct lan78xx_net *dev) @@ -3753,8 +3978,6 @@ static void lan78xx_rx_bh(struct lan78xx_net *dev) if (skb_queue_len(&dev->rxq) < dev->rx_qlen) tasklet_schedule(&dev->bh); } - if (skb_queue_len(&dev->txq) < dev->tx_qlen) - netif_wake_queue(dev->net); } static void lan78xx_bh(struct tasklet_struct *t) @@ -3770,10 +3993,6 @@ static void lan78xx_bh(struct tasklet_struct *t) entry->state = rx_cleanup; rx_process(dev, skb); continue; - case tx_done: - usb_free_urb(entry->urb); - dev_kfree_skb(skb); - continue; case rx_cleanup: usb_free_urb(entry->urb); dev_kfree_skb(skb); @@ -3792,11 +4011,26 @@ static void lan78xx_bh(struct tasklet_struct *t) jiffies + STAT_UPDATE_TIMER); } - if (!skb_queue_empty(&dev->txq_pend)) - lan78xx_tx_bh(dev); - if (!test_bit(EVENT_RX_HALT, &dev->flags)) lan78xx_rx_bh(dev); + + lan78xx_tx_bh(dev); + + if (!skb_queue_empty(&dev->done)) { + tasklet_schedule(&dev->bh); + } else if (netif_carrier_ok(dev->net)) { + if (skb_queue_empty(&dev->txq) && + !skb_queue_empty(&dev->txq_pend)) { + tasklet_schedule(&dev->bh); + } else { + netif_tx_lock(dev->net); + if (netif_queue_stopped(dev->net)) { + netif_wake_queue(dev->net); + tasklet_schedule(&dev->bh); + } + netif_tx_unlock(dev->net); + } + } } } @@ -3961,6 +4195,8 @@ static void lan78xx_disconnect(struct usb_interface *intf) lan78xx_unbind(dev, intf); + lan78xx_free_tx_resources(dev); + usb_kill_urb(dev->urb_intr); usb_free_urb(dev->urb_intr); @@ -3980,7 +4216,9 @@ static netdev_features_t lan78xx_features_check(struct sk_buff *skb, struct net_device *netdev, netdev_features_t features) { - if (skb->len + TX_OVERHEAD > MAX_SINGLE_PACKET_SIZE) + struct lan78xx_net *dev = netdev_priv(netdev); + + if (skb->len > LAN78XX_TSO_SIZE(dev)) features &= ~NETIF_F_GSO_MASK; features = vlan_features_check(skb, features); @@ -4051,6 +4289,16 @@ static int lan78xx_probe(struct usb_interface *intf, mutex_init(&dev->phy_mutex); mutex_init(&dev->dev_mutex); + ret = lan78xx_urb_config_init(dev); + if (ret < 0) + goto out2; + + ret = lan78xx_alloc_tx_resources(dev); + if (ret < 0) + goto out2; + + netif_set_gso_max_size(netdev, LAN78XX_TSO_SIZE(dev)); + tasklet_setup(&dev->bh, lan78xx_bh); INIT_DELAYED_WORK(&dev->wq, lan78xx_delayedwork); init_usb_anchor(&dev->deferred); @@ -4066,27 +4314,27 @@ static int lan78xx_probe(struct usb_interface *intf, if (intf->cur_altsetting->desc.bNumEndpoints < 3) { ret = -ENODEV; - goto out2; + goto out3; } dev->pipe_in = usb_rcvbulkpipe(udev, BULK_IN_PIPE); ep_blkin = usb_pipe_endpoint(udev, dev->pipe_in); if (!ep_blkin || !usb_endpoint_is_bulk_in(&ep_blkin->desc)) { ret = -ENODEV; - goto out2; + goto out3; } dev->pipe_out = usb_sndbulkpipe(udev, BULK_OUT_PIPE); ep_blkout = usb_pipe_endpoint(udev, dev->pipe_out); if (!ep_blkout || !usb_endpoint_is_bulk_out(&ep_blkout->desc)) { ret = -ENODEV; - goto out2; + goto out3; } ep_intr = &intf->cur_altsetting->endpoint[2]; if (!usb_endpoint_is_int_in(&ep_intr->desc)) { ret = -ENODEV; - goto out2; + goto out3; } dev->pipe_intr = usb_rcvintpipe(dev->udev, @@ -4094,7 +4342,7 @@ static int lan78xx_probe(struct usb_interface *intf, ret = lan78xx_bind(dev, intf); if (ret < 0) - goto out2; + goto out3; if (netdev->mtu > (dev->hard_mtu - netdev->hard_header_len)) netdev->mtu = dev->hard_mtu - netdev->hard_header_len; @@ -4108,13 +4356,13 @@ static int lan78xx_probe(struct usb_interface *intf, buf = kmalloc(maxp, GFP_KERNEL); if (!buf) { ret = -ENOMEM; - goto out3; + goto out4; } dev->urb_intr = usb_alloc_urb(0, GFP_KERNEL); if (!dev->urb_intr) { ret = -ENOMEM; - goto out4; + goto out5; } else { usb_fill_int_urb(dev->urb_intr, dev->udev, dev->pipe_intr, buf, maxp, @@ -4127,7 +4375,7 @@ static int lan78xx_probe(struct usb_interface *intf, /* Reject broken descriptors. */ if (dev->maxpacket == 0) { ret = -ENODEV; - goto out5; + goto out6; } /* driver requires remote-wakeup capability during autosuspend. */ @@ -4135,12 +4383,12 @@ static int lan78xx_probe(struct usb_interface *intf, ret = lan78xx_phy_init(dev); if (ret < 0) - goto out5; + goto out6; ret = register_netdev(netdev); if (ret != 0) { netif_err(dev, probe, netdev, "couldn't register the device\n"); - goto out6; + goto out7; } usb_set_intfdata(intf, dev); @@ -4155,14 +4403,16 @@ static int lan78xx_probe(struct usb_interface *intf, return 0; -out6: +out7: phy_disconnect(netdev->phydev); -out5: +out6: usb_free_urb(dev->urb_intr); -out4: +out5: kfree(buf); -out3: +out4: lan78xx_unbind(dev, intf); +out3: + lan78xx_free_tx_resources(dev); out2: free_netdev(netdev); out1: @@ -4583,8 +4833,7 @@ static bool lan78xx_submit_deferred_urbs(struct lan78xx_net *dev) if (!netif_device_present(dev->net) || !netif_carrier_ok(dev->net) || pipe_halted) { - usb_free_urb(urb); - dev_kfree_skb(skb); + lan78xx_release_tx_buf(dev, skb); continue; } @@ -4594,15 +4843,14 @@ static bool lan78xx_submit_deferred_urbs(struct lan78xx_net *dev) netif_trans_update(dev->net); lan78xx_queue_skb(&dev->txq, skb, tx_start); } else { - usb_free_urb(urb); - dev_kfree_skb(skb); - if (ret == -EPIPE) { netif_stop_queue(dev->net); pipe_halted = true; } else if (ret == -ENODEV) { netif_device_detach(dev->net); } + + lan78xx_release_tx_buf(dev, skb); } } @@ -4654,7 +4902,7 @@ static int lan78xx_resume(struct usb_interface *intf) if (!pipe_halted && netif_device_present(dev->net) && - (skb_queue_len(&dev->txq) < dev->tx_qlen)) + (lan78xx_tx_pend_data_len(dev) < lan78xx_tx_urb_space(dev))) netif_start_queue(dev->net); ret = lan78xx_start_tx_path(dev); From c450a8eb187a4d7ebe92e0afec54ac31ce403e79 Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:36 +0000 Subject: [PATCH 0336/2048] lan78xx: Introduce Rx URB processing improvements This patch introduces a new approach to allocating and managing Rx URBs that contributes to improving driver throughput and reducing CPU load. A pool of Rx URBs is created during driver instantiation. All the URBs are initially submitted to the USB host controller for processing. The default URB buffer size is different for each USB bus speed. The chosen sizes provide good USB utilisation with little impact on overall packet latency. Completed URBs are processed in the driver bottom half. The URB buffer contents are copied to a dynamically allocated SKB, which is then passed to the network stack. The URB is then re-submitted to the USB host controller. NOTE: the call to skb_copy() in rx_process() that copies the URB contents to a new SKB is a temporary change to make this patch work in its own right. This call will be removed when the NAPI processing is introduced by patch 6 in this patch set. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 282 ++++++++++++++++++++++---------------- 1 file changed, 166 insertions(+), 116 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index 7187aac01e7e..3dfd46c91093 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -102,6 +102,20 @@ #define TX_HS_URB_SIZE (16 * 1024) #define TX_FS_URB_SIZE (10 * 1024) +#define RX_SS_URB_NUM 30 +#define RX_HS_URB_NUM 10 +#define RX_FS_URB_NUM 10 +#define RX_SS_URB_SIZE TX_SS_URB_SIZE +#define RX_HS_URB_SIZE TX_HS_URB_SIZE +#define RX_FS_URB_SIZE TX_FS_URB_SIZE + +#define SS_BURST_CAP_SIZE RX_SS_URB_SIZE +#define SS_BULK_IN_DELAY 0x2000 +#define HS_BURST_CAP_SIZE RX_HS_URB_SIZE +#define HS_BULK_IN_DELAY 0x2000 +#define FS_BURST_CAP_SIZE RX_FS_URB_SIZE +#define FS_BULK_IN_DELAY 0x2000 + #define TX_CMD_LEN 8 #define TX_SKB_MIN_LEN (TX_CMD_LEN + ETH_HLEN) #define LAN78XX_TSO_SIZE(dev) ((dev)->tx_urb_size - TX_SKB_MIN_LEN) @@ -403,11 +417,13 @@ struct lan78xx_net { unsigned int tx_pend_data_len; size_t n_tx_urbs; + size_t n_rx_urbs; size_t tx_urb_size; + size_t rx_urb_size; - int rx_qlen; + struct sk_buff_head rxq_free; struct sk_buff_head rxq; - struct sk_buff_head done; + struct sk_buff_head rxq_done; struct sk_buff_head txq_free; struct sk_buff_head txq; struct sk_buff_head txq_pend; @@ -425,7 +441,9 @@ struct lan78xx_net { unsigned int pipe_in, pipe_out, pipe_intr; u32 hard_mtu; /* count any extra framing */ - size_t rx_urb_size; /* size for rx urbs */ + + unsigned int bulk_in_delay; + unsigned int burst_cap; unsigned long flags; @@ -542,6 +560,28 @@ error: return -ENOMEM; } +static struct sk_buff *lan78xx_get_rx_buf(struct lan78xx_net *dev) +{ + return lan78xx_get_buf(&dev->rxq_free); +} + +static void lan78xx_release_rx_buf(struct lan78xx_net *dev, + struct sk_buff *rx_buf) +{ + lan78xx_release_buf(&dev->rxq_free, rx_buf); +} + +static void lan78xx_free_rx_resources(struct lan78xx_net *dev) +{ + lan78xx_free_buf_pool(&dev->rxq_free); +} + +static int lan78xx_alloc_rx_resources(struct lan78xx_net *dev) +{ + return lan78xx_alloc_buf_pool(&dev->rxq_free, + dev->n_rx_urbs, dev->rx_urb_size, dev); +} + static struct sk_buff *lan78xx_get_tx_buf(struct lan78xx_net *dev) { return lan78xx_get_buf(&dev->txq_free); @@ -1321,6 +1361,8 @@ static int lan78xx_update_flowcontrol(struct lan78xx_net *dev, u8 duplex, return 0; } +static void lan78xx_rx_urb_submit_all(struct lan78xx_net *dev); + static int lan78xx_mac_reset(struct lan78xx_net *dev) { unsigned long start_time = jiffies; @@ -1452,6 +1494,8 @@ static int lan78xx_link_reset(struct lan78xx_net *dev) jiffies + STAT_UPDATE_TIMER); } + lan78xx_rx_urb_submit_all(dev); + tasklet_schedule(&dev->bh); } @@ -2684,16 +2728,28 @@ static int lan78xx_urb_config_init(struct lan78xx_net *dev) switch (dev->udev->speed) { case USB_SPEED_SUPER: + dev->rx_urb_size = RX_SS_URB_SIZE; dev->tx_urb_size = TX_SS_URB_SIZE; + dev->n_rx_urbs = RX_SS_URB_NUM; dev->n_tx_urbs = TX_SS_URB_NUM; + dev->bulk_in_delay = SS_BULK_IN_DELAY; + dev->burst_cap = SS_BURST_CAP_SIZE / SS_USB_PKT_SIZE; break; case USB_SPEED_HIGH: + dev->rx_urb_size = RX_HS_URB_SIZE; dev->tx_urb_size = TX_HS_URB_SIZE; + dev->n_rx_urbs = RX_HS_URB_NUM; dev->n_tx_urbs = TX_HS_URB_NUM; + dev->bulk_in_delay = HS_BULK_IN_DELAY; + dev->burst_cap = HS_BURST_CAP_SIZE / HS_USB_PKT_SIZE; break; case USB_SPEED_FULL: + dev->rx_urb_size = RX_FS_URB_SIZE; dev->tx_urb_size = TX_FS_URB_SIZE; + dev->n_rx_urbs = RX_FS_URB_NUM; dev->n_tx_urbs = TX_FS_URB_NUM; + dev->bulk_in_delay = FS_BULK_IN_DELAY; + dev->burst_cap = FS_BURST_CAP_SIZE / FS_USB_PKT_SIZE; break; default: netdev_warn(dev->net, "USB bus speed not supported\n"); @@ -2911,25 +2967,11 @@ static int lan78xx_reset(struct lan78xx_net *dev) /* Init LTM */ lan78xx_init_ltm(dev); - if (dev->udev->speed == USB_SPEED_SUPER) { - buf = DEFAULT_BURST_CAP_SIZE / SS_USB_PKT_SIZE; - dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; - dev->rx_qlen = 4; - } else if (dev->udev->speed == USB_SPEED_HIGH) { - buf = DEFAULT_BURST_CAP_SIZE / HS_USB_PKT_SIZE; - dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; - dev->rx_qlen = RX_MAX_QUEUE_MEMORY / dev->rx_urb_size; - } else { - buf = DEFAULT_BURST_CAP_SIZE / FS_USB_PKT_SIZE; - dev->rx_urb_size = DEFAULT_BURST_CAP_SIZE; - dev->rx_qlen = 4; - } - - ret = lan78xx_write_reg(dev, BURST_CAP, buf); + ret = lan78xx_write_reg(dev, BURST_CAP, dev->burst_cap); if (ret < 0) return ret; - ret = lan78xx_write_reg(dev, BULK_IN_DLY, DEFAULT_BULK_IN_DELAY); + ret = lan78xx_write_reg(dev, BULK_IN_DLY, dev->bulk_in_delay); if (ret < 0) return ret; @@ -3155,14 +3197,12 @@ static void lan78xx_terminate_urbs(struct lan78xx_net *dev) dev->wait = NULL; remove_wait_queue(&unlink_wakeup, &wait); - while (!skb_queue_empty(&dev->done)) { - struct skb_data *entry; - struct sk_buff *skb; + /* empty Rx done and Tx pend queues + */ + while (!skb_queue_empty(&dev->rxq_done)) { + struct sk_buff *skb = skb_dequeue(&dev->rxq_done); - skb = skb_dequeue(&dev->done); - entry = (struct skb_data *)(skb->cb); - usb_free_urb(entry->urb); - dev_kfree_skb(skb); + lan78xx_release_rx_buf(dev, skb); } skb_queue_purge(&dev->txq_pend); @@ -3230,12 +3270,12 @@ static enum skb_state defer_bh(struct lan78xx_net *dev, struct sk_buff *skb, __skb_unlink(skb, list); spin_unlock(&list->lock); - spin_lock(&dev->done.lock); + spin_lock(&dev->rxq_done.lock); - __skb_queue_tail(&dev->done, skb); - if (skb_queue_len(&dev->done) == 1) + __skb_queue_tail(&dev->rxq_done, skb); + if (skb_queue_len(&dev->rxq_done) == 1) tasklet_schedule(&dev->bh); - spin_unlock_irqrestore(&dev->done.lock, flags); + spin_unlock_irqrestore(&dev->rxq_done.lock, flags); return old_state; } @@ -3624,43 +3664,32 @@ static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb) static inline void rx_process(struct lan78xx_net *dev, struct sk_buff *skb) { - if (!lan78xx_rx(dev, skb)) { + struct sk_buff *rx_buf = skb_copy(skb, GFP_ATOMIC); + + if (!lan78xx_rx(dev, rx_buf)) { dev->net->stats.rx_errors++; - goto done; + return; } - if (skb->len) { - lan78xx_skb_return(dev, skb); + if (rx_buf->len) { + lan78xx_skb_return(dev, rx_buf); return; } netif_dbg(dev, rx_err, dev->net, "drop\n"); dev->net->stats.rx_errors++; -done: - skb_queue_tail(&dev->done, skb); } static void rx_complete(struct urb *urb); -static int rx_submit(struct lan78xx_net *dev, struct urb *urb, gfp_t flags) +static int rx_submit(struct lan78xx_net *dev, struct sk_buff *skb, gfp_t flags) { - struct sk_buff *skb; - struct skb_data *entry; - unsigned long lockflags; + struct skb_data *entry = (struct skb_data *)skb->cb; size_t size = dev->rx_urb_size; + struct urb *urb = entry->urb; + unsigned long lockflags; int ret = 0; - skb = netdev_alloc_skb_ip_align(dev->net, size); - if (!skb) { - usb_free_urb(urb); - return -ENOMEM; - } - - entry = (struct skb_data *)skb->cb; - entry->urb = urb; - entry->dev = dev; - entry->length = 0; - usb_fill_bulk_urb(urb, dev->udev, dev->pipe_in, skb->data, size, rx_complete, skb); @@ -3670,7 +3699,7 @@ static int rx_submit(struct lan78xx_net *dev, struct urb *urb, gfp_t flags) netif_running(dev->net) && !test_bit(EVENT_RX_HALT, &dev->flags) && !test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { - ret = usb_submit_urb(urb, GFP_ATOMIC); + ret = usb_submit_urb(urb, flags); switch (ret) { case 0: lan78xx_queue_skb(&dev->rxq, skb, rx_start); @@ -3685,21 +3714,23 @@ static int rx_submit(struct lan78xx_net *dev, struct urb *urb, gfp_t flags) break; case -EHOSTUNREACH: ret = -ENOLINK; + tasklet_schedule(&dev->bh); break; default: netif_dbg(dev, rx_err, dev->net, "rx submit, %d\n", ret); tasklet_schedule(&dev->bh); + break; } } else { netif_dbg(dev, ifdown, dev->net, "rx: stopped\n"); ret = -ENOLINK; } spin_unlock_irqrestore(&dev->rxq.lock, lockflags); - if (ret) { - dev_kfree_skb_any(skb); - usb_free_urb(urb); - } + + if (ret) + lan78xx_release_rx_buf(dev, skb); + return ret; } @@ -3711,9 +3742,14 @@ static void rx_complete(struct urb *urb) int urb_status = urb->status; enum skb_state state; + netif_dbg(dev, rx_status, dev->net, + "rx done: status %d", urb->status); + skb_put(skb, urb->actual_length); state = rx_done; - entry->urb = NULL; + + if (urb != entry->urb) + netif_warn(dev, rx_err, dev->net, "URB pointer mismatch"); switch (urb_status) { case 0: @@ -3735,16 +3771,12 @@ static void rx_complete(struct urb *urb) netif_dbg(dev, ifdown, dev->net, "rx shutdown, code %d\n", urb_status); state = rx_cleanup; - entry->urb = urb; - urb = NULL; break; case -EPROTO: case -ETIME: case -EILSEQ: dev->net->stats.rx_errors++; state = rx_cleanup; - entry->urb = urb; - urb = NULL; break; /* data overrun ... flush fifo? */ @@ -3760,17 +3792,31 @@ static void rx_complete(struct urb *urb) } state = defer_bh(dev, skb, &dev->rxq, state); +} - if (urb) { - if (netif_running(dev->net) && - !test_bit(EVENT_RX_HALT, &dev->flags) && - state != unlink_start) { - rx_submit(dev, urb, GFP_ATOMIC); - return; - } - usb_free_urb(urb); +static void lan78xx_rx_urb_submit_all(struct lan78xx_net *dev) +{ + struct sk_buff *rx_buf; + + /* Ensure the maximum number of Rx URBs is submitted + */ + while ((rx_buf = lan78xx_get_rx_buf(dev)) != NULL) { + if (rx_submit(dev, rx_buf, GFP_ATOMIC) != 0) + break; } - netif_dbg(dev, rx_err, dev->net, "no read resubmitted\n"); +} + +static void lan78xx_rx_urb_resubmit(struct lan78xx_net *dev, + struct sk_buff *rx_buf) +{ + /* reset SKB data pointers */ + + rx_buf->data = rx_buf->head; + skb_reset_tail_pointer(rx_buf); + rx_buf->len = 0; + rx_buf->data_len = 0; + + rx_submit(dev, rx_buf, GFP_ATOMIC); } static void lan78xx_fill_tx_cmd_words(struct sk_buff *skb, u8 *buffer) @@ -3960,47 +4006,41 @@ out: } while (ret == 0); } -static void lan78xx_rx_bh(struct lan78xx_net *dev) -{ - struct urb *urb; - int i; - - if (skb_queue_len(&dev->rxq) < dev->rx_qlen) { - for (i = 0; i < 10; i++) { - if (skb_queue_len(&dev->rxq) >= dev->rx_qlen) - break; - urb = usb_alloc_urb(0, GFP_ATOMIC); - if (urb) - if (rx_submit(dev, urb, GFP_ATOMIC) == -ENOLINK) - return; - } - - if (skb_queue_len(&dev->rxq) < dev->rx_qlen) - tasklet_schedule(&dev->bh); - } -} - static void lan78xx_bh(struct tasklet_struct *t) { struct lan78xx_net *dev = from_tasklet(dev, t, bh); - struct sk_buff *skb; + struct sk_buff_head done; + struct sk_buff *rx_buf; struct skb_data *entry; + unsigned long flags; - while ((skb = skb_dequeue(&dev->done))) { - entry = (struct skb_data *)(skb->cb); + /* Take a snapshot of the done queue and move items to a + * temporary queue. Rx URB completions will continue to add + * to the done queue. + */ + __skb_queue_head_init(&done); + + spin_lock_irqsave(&dev->rxq_done.lock, flags); + skb_queue_splice_init(&dev->rxq_done, &done); + spin_unlock_irqrestore(&dev->rxq_done.lock, flags); + + /* Extract receive frames from completed URBs and + * pass them to the stack. Re-submit each completed URB. + */ + while ((rx_buf = __skb_dequeue(&done))) { + entry = (struct skb_data *)(rx_buf->cb); switch (entry->state) { case rx_done: - entry->state = rx_cleanup; - rx_process(dev, skb); - continue; + rx_process(dev, rx_buf); + break; case rx_cleanup: - usb_free_urb(entry->urb); - dev_kfree_skb(skb); - continue; + break; default: netdev_dbg(dev->net, "skb state %d\n", entry->state); - return; + break; } + + lan78xx_rx_urb_resubmit(dev, rx_buf); } if (netif_device_present(dev->net) && netif_running(dev->net)) { @@ -4012,11 +4052,14 @@ static void lan78xx_bh(struct tasklet_struct *t) } if (!test_bit(EVENT_RX_HALT, &dev->flags)) - lan78xx_rx_bh(dev); + lan78xx_rx_urb_submit_all(dev); lan78xx_tx_bh(dev); - if (!skb_queue_empty(&dev->done)) { + /* Start a new polling cycle if data was received or + * data is waiting to be transmitted. + */ + if (!skb_queue_empty(&dev->rxq_done)) { tasklet_schedule(&dev->bh); } else if (netif_carrier_ok(dev->net)) { if (skb_queue_empty(&dev->txq) && @@ -4196,6 +4239,7 @@ static void lan78xx_disconnect(struct usb_interface *intf) lan78xx_unbind(dev, intf); lan78xx_free_tx_resources(dev); + lan78xx_free_rx_resources(dev); usb_kill_urb(dev->urb_intr); usb_free_urb(dev->urb_intr); @@ -4284,7 +4328,7 @@ static int lan78xx_probe(struct usb_interface *intf, skb_queue_head_init(&dev->rxq); skb_queue_head_init(&dev->txq); - skb_queue_head_init(&dev->done); + skb_queue_head_init(&dev->rxq_done); skb_queue_head_init(&dev->txq_pend); mutex_init(&dev->phy_mutex); mutex_init(&dev->dev_mutex); @@ -4297,6 +4341,10 @@ static int lan78xx_probe(struct usb_interface *intf, if (ret < 0) goto out2; + ret = lan78xx_alloc_rx_resources(dev); + if (ret < 0) + goto out3; + netif_set_gso_max_size(netdev, LAN78XX_TSO_SIZE(dev)); tasklet_setup(&dev->bh, lan78xx_bh); @@ -4314,27 +4362,27 @@ static int lan78xx_probe(struct usb_interface *intf, if (intf->cur_altsetting->desc.bNumEndpoints < 3) { ret = -ENODEV; - goto out3; + goto out4; } dev->pipe_in = usb_rcvbulkpipe(udev, BULK_IN_PIPE); ep_blkin = usb_pipe_endpoint(udev, dev->pipe_in); if (!ep_blkin || !usb_endpoint_is_bulk_in(&ep_blkin->desc)) { ret = -ENODEV; - goto out3; + goto out4; } dev->pipe_out = usb_sndbulkpipe(udev, BULK_OUT_PIPE); ep_blkout = usb_pipe_endpoint(udev, dev->pipe_out); if (!ep_blkout || !usb_endpoint_is_bulk_out(&ep_blkout->desc)) { ret = -ENODEV; - goto out3; + goto out4; } ep_intr = &intf->cur_altsetting->endpoint[2]; if (!usb_endpoint_is_int_in(&ep_intr->desc)) { ret = -ENODEV; - goto out3; + goto out4; } dev->pipe_intr = usb_rcvintpipe(dev->udev, @@ -4342,7 +4390,7 @@ static int lan78xx_probe(struct usb_interface *intf, ret = lan78xx_bind(dev, intf); if (ret < 0) - goto out3; + goto out4; if (netdev->mtu > (dev->hard_mtu - netdev->hard_header_len)) netdev->mtu = dev->hard_mtu - netdev->hard_header_len; @@ -4356,13 +4404,13 @@ static int lan78xx_probe(struct usb_interface *intf, buf = kmalloc(maxp, GFP_KERNEL); if (!buf) { ret = -ENOMEM; - goto out4; + goto out5; } dev->urb_intr = usb_alloc_urb(0, GFP_KERNEL); if (!dev->urb_intr) { ret = -ENOMEM; - goto out5; + goto out6; } else { usb_fill_int_urb(dev->urb_intr, dev->udev, dev->pipe_intr, buf, maxp, @@ -4383,12 +4431,12 @@ static int lan78xx_probe(struct usb_interface *intf, ret = lan78xx_phy_init(dev); if (ret < 0) - goto out6; + goto out7; ret = register_netdev(netdev); if (ret != 0) { netif_err(dev, probe, netdev, "couldn't register the device\n"); - goto out7; + goto out8; } usb_set_intfdata(intf, dev); @@ -4403,14 +4451,16 @@ static int lan78xx_probe(struct usb_interface *intf, return 0; -out7: +out8: phy_disconnect(netdev->phydev); -out6: +out7: usb_free_urb(dev->urb_intr); -out5: +out6: kfree(buf); -out4: +out5: lan78xx_unbind(dev, intf); +out4: + lan78xx_free_rx_resources(dev); out3: lan78xx_free_tx_resources(dev); out2: From 9d2da72189a8d233cdcf6ebe5ff0689730e93e1f Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:37 +0000 Subject: [PATCH 0337/2048] lan78xx: Re-order rx_submit() to remove forward declaration Move position of rx_submit() to remove forward declaration of rx_complete() which is now no longer required. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 106 +++++++++++++++++++------------------- 1 file changed, 52 insertions(+), 54 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index 3dfd46c91093..ebd3d9fc5c41 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -3680,60 +3680,6 @@ static inline void rx_process(struct lan78xx_net *dev, struct sk_buff *skb) dev->net->stats.rx_errors++; } -static void rx_complete(struct urb *urb); - -static int rx_submit(struct lan78xx_net *dev, struct sk_buff *skb, gfp_t flags) -{ - struct skb_data *entry = (struct skb_data *)skb->cb; - size_t size = dev->rx_urb_size; - struct urb *urb = entry->urb; - unsigned long lockflags; - int ret = 0; - - usb_fill_bulk_urb(urb, dev->udev, dev->pipe_in, - skb->data, size, rx_complete, skb); - - spin_lock_irqsave(&dev->rxq.lock, lockflags); - - if (netif_device_present(dev->net) && - netif_running(dev->net) && - !test_bit(EVENT_RX_HALT, &dev->flags) && - !test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { - ret = usb_submit_urb(urb, flags); - switch (ret) { - case 0: - lan78xx_queue_skb(&dev->rxq, skb, rx_start); - break; - case -EPIPE: - lan78xx_defer_kevent(dev, EVENT_RX_HALT); - break; - case -ENODEV: - case -ENOENT: - netif_dbg(dev, ifdown, dev->net, "device gone\n"); - netif_device_detach(dev->net); - break; - case -EHOSTUNREACH: - ret = -ENOLINK; - tasklet_schedule(&dev->bh); - break; - default: - netif_dbg(dev, rx_err, dev->net, - "rx submit, %d\n", ret); - tasklet_schedule(&dev->bh); - break; - } - } else { - netif_dbg(dev, ifdown, dev->net, "rx: stopped\n"); - ret = -ENOLINK; - } - spin_unlock_irqrestore(&dev->rxq.lock, lockflags); - - if (ret) - lan78xx_release_rx_buf(dev, skb); - - return ret; -} - static void rx_complete(struct urb *urb) { struct sk_buff *skb = (struct sk_buff *)urb->context; @@ -3794,6 +3740,58 @@ static void rx_complete(struct urb *urb) state = defer_bh(dev, skb, &dev->rxq, state); } +static int rx_submit(struct lan78xx_net *dev, struct sk_buff *skb, gfp_t flags) +{ + struct skb_data *entry = (struct skb_data *)skb->cb; + size_t size = dev->rx_urb_size; + struct urb *urb = entry->urb; + unsigned long lockflags; + int ret = 0; + + usb_fill_bulk_urb(urb, dev->udev, dev->pipe_in, + skb->data, size, rx_complete, skb); + + spin_lock_irqsave(&dev->rxq.lock, lockflags); + + if (netif_device_present(dev->net) && + netif_running(dev->net) && + !test_bit(EVENT_RX_HALT, &dev->flags) && + !test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { + ret = usb_submit_urb(urb, flags); + switch (ret) { + case 0: + lan78xx_queue_skb(&dev->rxq, skb, rx_start); + break; + case -EPIPE: + lan78xx_defer_kevent(dev, EVENT_RX_HALT); + break; + case -ENODEV: + case -ENOENT: + netif_dbg(dev, ifdown, dev->net, "device gone\n"); + netif_device_detach(dev->net); + break; + case -EHOSTUNREACH: + ret = -ENOLINK; + tasklet_schedule(&dev->bh); + break; + default: + netif_dbg(dev, rx_err, dev->net, + "rx submit, %d\n", ret); + tasklet_schedule(&dev->bh); + break; + } + } else { + netif_dbg(dev, ifdown, dev->net, "rx: stopped\n"); + ret = -ENOLINK; + } + spin_unlock_irqrestore(&dev->rxq.lock, lockflags); + + if (ret) + lan78xx_release_rx_buf(dev, skb); + + return ret; +} + static void lan78xx_rx_urb_submit_all(struct lan78xx_net *dev) { struct sk_buff *rx_buf; From 0dd87266c1337db64cd6d3bbd1503553b922aa88 Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:38 +0000 Subject: [PATCH 0338/2048] lan78xx: Remove hardware-specific header update Remove hardware-specific header length adjustment as it is no longer required. It also breaks generic receive offload (GRO) processing of received TCP frames that results in a TCP ACK being sent for each received frame. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 51 ++++++++++++--------------------------- 1 file changed, 16 insertions(+), 35 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index ebd3d9fc5c41..64f60cf6c911 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -67,7 +67,6 @@ #define DEFAULT_TSO_CSUM_ENABLE (true) #define DEFAULT_VLAN_FILTER_ENABLE (true) #define DEFAULT_VLAN_RX_OFFLOAD (true) -#define TX_OVERHEAD (8) #define TX_ALIGNMENT (4) #define RXW_PADDING 2 @@ -120,6 +119,10 @@ #define TX_SKB_MIN_LEN (TX_CMD_LEN + ETH_HLEN) #define LAN78XX_TSO_SIZE(dev) ((dev)->tx_urb_size - TX_SKB_MIN_LEN) +#define RX_CMD_LEN 10 +#define RX_SKB_MIN_LEN (RX_CMD_LEN + ETH_HLEN) +#define RX_MAX_FRAME_LEN(mtu) ((mtu) + ETH_HLEN + VLAN_HLEN) + /* USB related defines */ #define BULK_IN_PIPE 1 #define BULK_OUT_PIPE 2 @@ -440,8 +443,6 @@ struct lan78xx_net { struct mutex phy_mutex; /* for phy access */ unsigned int pipe_in, pipe_out, pipe_intr; - u32 hard_mtu; /* count any extra framing */ - unsigned int bulk_in_delay; unsigned int burst_cap; @@ -2536,37 +2537,24 @@ found: static int lan78xx_change_mtu(struct net_device *netdev, int new_mtu) { struct lan78xx_net *dev = netdev_priv(netdev); - int ll_mtu = new_mtu + netdev->hard_header_len; - int old_hard_mtu = dev->hard_mtu; - int old_rx_urb_size = dev->rx_urb_size; + int max_frame_len = RX_MAX_FRAME_LEN(new_mtu); int ret; /* no second zero-length packet read wanted after mtu-sized packets */ - if ((ll_mtu % dev->maxpacket) == 0) + if ((max_frame_len % dev->maxpacket) == 0) return -EDOM; ret = usb_autopm_get_interface(dev->intf); if (ret < 0) return ret; - lan78xx_set_rx_max_frame_length(dev, new_mtu + VLAN_ETH_HLEN); - - netdev->mtu = new_mtu; - - dev->hard_mtu = netdev->mtu + netdev->hard_header_len; - if (dev->rx_urb_size == old_hard_mtu) { - dev->rx_urb_size = dev->hard_mtu; - if (dev->rx_urb_size > old_rx_urb_size) { - if (netif_running(dev->net)) { - unlink_urbs(dev, &dev->rxq); - tasklet_schedule(&dev->bh); - } - } - } + ret = lan78xx_set_rx_max_frame_length(dev, max_frame_len); + if (!ret) + netdev->mtu = new_mtu; usb_autopm_put_interface(dev->intf); - return 0; + return ret; } static int lan78xx_set_mac_addr(struct net_device *netdev, void *p) @@ -3084,7 +3072,7 @@ static int lan78xx_reset(struct lan78xx_net *dev) return ret; ret = lan78xx_set_rx_max_frame_length(dev, - dev->net->mtu + VLAN_ETH_HLEN); + RX_MAX_FRAME_LEN(dev->net->mtu)); return ret; } @@ -3489,9 +3477,6 @@ static int lan78xx_bind(struct lan78xx_net *dev, struct usb_interface *intf) goto out1; } - dev->net->hard_header_len += TX_OVERHEAD; - dev->hard_mtu = dev->net->mtu + dev->net->hard_header_len; - /* Init all registers */ ret = lan78xx_reset(dev); if (ret) { @@ -3592,7 +3577,7 @@ static void lan78xx_skb_return(struct lan78xx_net *dev, struct sk_buff *skb) static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb) { - if (skb->len < dev->net->hard_header_len) + if (skb->len < RX_SKB_MIN_LEN) return 0; while (skb->len > 0) { @@ -3699,7 +3684,7 @@ static void rx_complete(struct urb *urb) switch (urb_status) { case 0: - if (skb->len < dev->net->hard_header_len) { + if (skb->len < RX_SKB_MIN_LEN) { state = rx_cleanup; dev->net->stats.rx_errors++; dev->net->stats.rx_length_errors++; @@ -4343,6 +4328,9 @@ static int lan78xx_probe(struct usb_interface *intf, if (ret < 0) goto out3; + /* MTU range: 68 - 9000 */ + netdev->max_mtu = MAX_SINGLE_PACKET_SIZE; + netif_set_gso_max_size(netdev, LAN78XX_TSO_SIZE(dev)); tasklet_setup(&dev->bh, lan78xx_bh); @@ -4390,13 +4378,6 @@ static int lan78xx_probe(struct usb_interface *intf, if (ret < 0) goto out4; - if (netdev->mtu > (dev->hard_mtu - netdev->hard_header_len)) - netdev->mtu = dev->hard_mtu - netdev->hard_header_len; - - /* MTU range: 68 - 9000 */ - netdev->max_mtu = MAX_SINGLE_PACKET_SIZE; - netif_set_gso_max_size(netdev, MAX_SINGLE_PACKET_SIZE - MAX_HEADER); - period = ep_intr->desc.bInterval; maxp = usb_maxpacket(dev->udev, dev->pipe_intr, 0); buf = kmalloc(maxp, GFP_KERNEL); From ec4c7e12396b1a30fbacfa68425118f5b46ea878 Mon Sep 17 00:00:00 2001 From: John Efstathiades Date: Thu, 18 Nov 2021 11:01:39 +0000 Subject: [PATCH 0339/2048] lan78xx: Introduce NAPI polling support This patch introduces a NAPI-style approach for processing completed Rx URBs that contributes to improving driver throughput and reducing CPU load. Packets in completed URBs are copied to NAPI SKBs and passed to the network stack for processing. Each frame passed to the stack is one work item in the NAPI budget. If the NAPI budget is consumed and frames remain, they are added to an overflow queue that is processed at the start of the next NAPI polling cycle. The NAPI handler is also responsible for copying pending Tx data to Tx URBs and submitting them to the USB host controller for transmission. Signed-off-by: John Efstathiades Signed-off-by: David S. Miller --- drivers/net/usb/lan78xx.c | 182 ++++++++++++++++++++++++-------------- 1 file changed, 114 insertions(+), 68 deletions(-) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index 64f60cf6c911..a9e7cbe15f20 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -90,6 +90,8 @@ WAKE_MCAST | WAKE_BCAST | \ WAKE_ARP | WAKE_MAGIC) +#define LAN78XX_NAPI_WEIGHT 64 + #define TX_URB_NUM 10 #define TX_SS_URB_NUM TX_URB_NUM #define TX_HS_URB_NUM TX_URB_NUM @@ -427,11 +429,13 @@ struct lan78xx_net { struct sk_buff_head rxq_free; struct sk_buff_head rxq; struct sk_buff_head rxq_done; + struct sk_buff_head rxq_overflow; struct sk_buff_head txq_free; struct sk_buff_head txq; struct sk_buff_head txq_pend; - struct tasklet_struct bh; + struct napi_struct napi; + struct delayed_work wq; int msg_enable; @@ -1497,7 +1501,7 @@ static int lan78xx_link_reset(struct lan78xx_net *dev) lan78xx_rx_urb_submit_all(dev); - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); } return 0; @@ -3152,6 +3156,8 @@ static int lan78xx_open(struct net_device *net) dev->link_on = false; + napi_enable(&dev->napi); + lan78xx_defer_kevent(dev, EVENT_LINK_RESET); done: mutex_unlock(&dev->dev_mutex); @@ -3185,7 +3191,7 @@ static void lan78xx_terminate_urbs(struct lan78xx_net *dev) dev->wait = NULL; remove_wait_queue(&unlink_wakeup, &wait); - /* empty Rx done and Tx pend queues + /* empty Rx done, Rx overflow and Tx pend queues */ while (!skb_queue_empty(&dev->rxq_done)) { struct sk_buff *skb = skb_dequeue(&dev->rxq_done); @@ -3193,6 +3199,7 @@ static void lan78xx_terminate_urbs(struct lan78xx_net *dev) lan78xx_release_rx_buf(dev, skb); } + skb_queue_purge(&dev->rxq_overflow); skb_queue_purge(&dev->txq_pend); } @@ -3209,7 +3216,7 @@ static int lan78xx_stop(struct net_device *net) clear_bit(EVENT_DEV_OPEN, &dev->flags); netif_stop_queue(net); - tasklet_kill(&dev->bh); + napi_disable(&dev->napi); lan78xx_terminate_urbs(dev); @@ -3262,7 +3269,8 @@ static enum skb_state defer_bh(struct lan78xx_net *dev, struct sk_buff *skb, __skb_queue_tail(&dev->rxq_done, skb); if (skb_queue_len(&dev->rxq_done) == 1) - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); + spin_unlock_irqrestore(&dev->rxq_done.lock, flags); return old_state; @@ -3315,11 +3323,11 @@ static void tx_complete(struct urb *urb) lan78xx_release_tx_buf(dev, skb); - /* Re-schedule tasklet if Tx data pending but no URBs in progress. + /* Re-schedule NAPI if Tx data pending but no URBs in progress. */ if (skb_queue_empty(&dev->txq) && !skb_queue_empty(&dev->txq_pend)) - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); } static void lan78xx_queue_skb(struct sk_buff_head *list, @@ -3405,7 +3413,7 @@ lan78xx_start_xmit(struct sk_buff *skb, struct net_device *net) /* Set up a Tx URB if none is in progress */ if (skb_queue_empty(&dev->txq)) - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); /* Stop stack Tx queue if we have enough data to fill * all the free Tx URBs. @@ -3419,7 +3427,7 @@ lan78xx_start_xmit(struct sk_buff *skb, struct net_device *net) /* Kick off transmission of pending data */ if (!skb_queue_empty(&dev->txq_free)) - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); } return NETDEV_TX_OK; @@ -3555,8 +3563,6 @@ static void lan78xx_rx_vlan_offload(struct lan78xx_net *dev, static void lan78xx_skb_return(struct lan78xx_net *dev, struct sk_buff *skb) { - int status; - dev->net->stats.rx_packets++; dev->net->stats.rx_bytes += skb->len; @@ -3569,21 +3575,21 @@ static void lan78xx_skb_return(struct lan78xx_net *dev, struct sk_buff *skb) if (skb_defer_rx_timestamp(skb)) return; - status = netif_rx(skb); - if (status != NET_RX_SUCCESS) - netif_dbg(dev, rx_err, dev->net, - "netif_rx status %d\n", status); + napi_gro_receive(&dev->napi, skb); } -static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb) +static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb, + int budget, int *work_done) { if (skb->len < RX_SKB_MIN_LEN) return 0; + /* Extract frames from the URB buffer and pass each one to + * the stack in a new NAPI SKB. + */ while (skb->len > 0) { u32 rx_cmd_a, rx_cmd_b, align_count, size; u16 rx_cmd_c; - struct sk_buff *skb2; unsigned char *packet; rx_cmd_a = get_unaligned_le32(skb->data); @@ -3605,41 +3611,36 @@ static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb) netif_dbg(dev, rx_err, dev->net, "Error rx_cmd_a=0x%08x", rx_cmd_a); } else { - /* last frame in this batch */ - if (skb->len == size) { - lan78xx_rx_csum_offload(dev, skb, - rx_cmd_a, rx_cmd_b); - lan78xx_rx_vlan_offload(dev, skb, - rx_cmd_a, rx_cmd_b); + u32 frame_len = size - ETH_FCS_LEN; + struct sk_buff *skb2; - skb_trim(skb, skb->len - 4); /* remove fcs */ - skb->truesize = size + sizeof(struct sk_buff); - - return 1; - } - - skb2 = skb_clone(skb, GFP_ATOMIC); - if (unlikely(!skb2)) { - netdev_warn(dev->net, "Error allocating skb"); + skb2 = napi_alloc_skb(&dev->napi, frame_len); + if (!skb2) return 0; - } - skb2->len = size; - skb2->data = packet; - skb_set_tail_pointer(skb2, size); + memcpy(skb2->data, packet, frame_len); + + skb_put(skb2, frame_len); lan78xx_rx_csum_offload(dev, skb2, rx_cmd_a, rx_cmd_b); lan78xx_rx_vlan_offload(dev, skb2, rx_cmd_a, rx_cmd_b); - skb_trim(skb2, skb2->len - 4); /* remove fcs */ - skb2->truesize = size + sizeof(struct sk_buff); - - lan78xx_skb_return(dev, skb2); + /* Processing of the URB buffer must complete once + * it has started. If the NAPI work budget is exhausted + * while frames remain they are added to the overflow + * queue for delivery in the next NAPI polling cycle. + */ + if (*work_done < budget) { + lan78xx_skb_return(dev, skb2); + ++(*work_done); + } else { + skb_queue_tail(&dev->rxq_overflow, skb2); + } } skb_pull(skb, size); - /* padding bytes before the next frame starts */ + /* skip padding bytes before the next frame starts */ if (skb->len) skb_pull(skb, align_count); } @@ -3647,22 +3648,13 @@ static int lan78xx_rx(struct lan78xx_net *dev, struct sk_buff *skb) return 1; } -static inline void rx_process(struct lan78xx_net *dev, struct sk_buff *skb) +static inline void rx_process(struct lan78xx_net *dev, struct sk_buff *skb, + int budget, int *work_done) { - struct sk_buff *rx_buf = skb_copy(skb, GFP_ATOMIC); - - if (!lan78xx_rx(dev, rx_buf)) { + if (!lan78xx_rx(dev, skb, budget, work_done)) { + netif_dbg(dev, rx_err, dev->net, "drop\n"); dev->net->stats.rx_errors++; - return; } - - if (rx_buf->len) { - lan78xx_skb_return(dev, rx_buf); - return; - } - - netif_dbg(dev, rx_err, dev->net, "drop\n"); - dev->net->stats.rx_errors++; } static void rx_complete(struct urb *urb) @@ -3757,12 +3749,12 @@ static int rx_submit(struct lan78xx_net *dev, struct sk_buff *skb, gfp_t flags) break; case -EHOSTUNREACH: ret = -ENOLINK; - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); break; default: netif_dbg(dev, rx_err, dev->net, "rx submit, %d\n", ret); - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); break; } } else { @@ -3989,13 +3981,21 @@ out: } while (ret == 0); } -static void lan78xx_bh(struct tasklet_struct *t) +static int lan78xx_bh(struct lan78xx_net *dev, int budget) { - struct lan78xx_net *dev = from_tasklet(dev, t, bh); struct sk_buff_head done; struct sk_buff *rx_buf; struct skb_data *entry; unsigned long flags; + int work_done = 0; + + /* Pass frames received in the last NAPI cycle before + * working on newly completed URBs. + */ + while (!skb_queue_empty(&dev->rxq_overflow)) { + lan78xx_skb_return(dev, skb_dequeue(&dev->rxq_overflow)); + ++work_done; + } /* Take a snapshot of the done queue and move items to a * temporary queue. Rx URB completions will continue to add @@ -4010,22 +4010,32 @@ static void lan78xx_bh(struct tasklet_struct *t) /* Extract receive frames from completed URBs and * pass them to the stack. Re-submit each completed URB. */ - while ((rx_buf = __skb_dequeue(&done))) { + while ((work_done < budget) && + (rx_buf = __skb_dequeue(&done))) { entry = (struct skb_data *)(rx_buf->cb); switch (entry->state) { case rx_done: - rx_process(dev, rx_buf); + rx_process(dev, rx_buf, budget, &work_done); break; case rx_cleanup: break; default: - netdev_dbg(dev->net, "skb state %d\n", entry->state); + netdev_dbg(dev->net, "rx buf state %d\n", + entry->state); break; } lan78xx_rx_urb_resubmit(dev, rx_buf); } + /* If budget was consumed before processing all the URBs put them + * back on the front of the done queue. They will be first to be + * processed in the next NAPI cycle. + */ + spin_lock_irqsave(&dev->rxq_done.lock, flags); + skb_queue_splice(&done, &dev->rxq_done); + spin_unlock_irqrestore(&dev->rxq_done.lock, flags); + if (netif_device_present(dev->net) && netif_running(dev->net)) { /* reset update timer delta */ if (timer_pending(&dev->stat_monitor) && (dev->delta != 1)) { @@ -4034,30 +4044,61 @@ static void lan78xx_bh(struct tasklet_struct *t) jiffies + STAT_UPDATE_TIMER); } + /* Submit all free Rx URBs */ + if (!test_bit(EVENT_RX_HALT, &dev->flags)) lan78xx_rx_urb_submit_all(dev); + /* Submit new Tx URBs */ + lan78xx_tx_bh(dev); + } + + return work_done; +} + +static int lan78xx_poll(struct napi_struct *napi, int budget) +{ + struct lan78xx_net *dev = container_of(napi, struct lan78xx_net, napi); + int result = budget; + int work_done; + + /* Don't do any work if the device is suspended */ + + if (test_bit(EVENT_DEV_ASLEEP, &dev->flags)) { + napi_complete_done(napi, 0); + return 0; + } + + /* Process completed URBs and submit new URBs */ + + work_done = lan78xx_bh(dev, budget); + + if (work_done < budget) { + napi_complete_done(napi, work_done); /* Start a new polling cycle if data was received or * data is waiting to be transmitted. */ if (!skb_queue_empty(&dev->rxq_done)) { - tasklet_schedule(&dev->bh); + napi_schedule(napi); } else if (netif_carrier_ok(dev->net)) { if (skb_queue_empty(&dev->txq) && !skb_queue_empty(&dev->txq_pend)) { - tasklet_schedule(&dev->bh); + napi_schedule(napi); } else { netif_tx_lock(dev->net); if (netif_queue_stopped(dev->net)) { netif_wake_queue(dev->net); - tasklet_schedule(&dev->bh); + napi_schedule(napi); } netif_tx_unlock(dev->net); } } + result = work_done; } + + return result; } static void lan78xx_delayedwork(struct work_struct *work) @@ -4103,7 +4144,7 @@ static void lan78xx_delayedwork(struct work_struct *work) status); } else { clear_bit(EVENT_RX_HALT, &dev->flags); - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); } } @@ -4197,6 +4238,8 @@ static void lan78xx_disconnect(struct usb_interface *intf) set_bit(EVENT_DEV_DISCONNECT, &dev->flags); + netif_napi_del(&dev->napi); + udev = interface_to_usbdev(intf); net = dev->net; @@ -4236,7 +4279,7 @@ static void lan78xx_tx_timeout(struct net_device *net, unsigned int txqueue) struct lan78xx_net *dev = netdev_priv(net); unlink_urbs(dev, &dev->txq); - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); } static netdev_features_t lan78xx_features_check(struct sk_buff *skb, @@ -4313,6 +4356,7 @@ static int lan78xx_probe(struct usb_interface *intf, skb_queue_head_init(&dev->txq); skb_queue_head_init(&dev->rxq_done); skb_queue_head_init(&dev->txq_pend); + skb_queue_head_init(&dev->rxq_overflow); mutex_init(&dev->phy_mutex); mutex_init(&dev->dev_mutex); @@ -4333,7 +4377,8 @@ static int lan78xx_probe(struct usb_interface *intf, netif_set_gso_max_size(netdev, LAN78XX_TSO_SIZE(dev)); - tasklet_setup(&dev->bh, lan78xx_bh); + netif_napi_add(netdev, &dev->napi, lan78xx_poll, LAN78XX_NAPI_WEIGHT); + INIT_DELAYED_WORK(&dev->wq, lan78xx_delayedwork); init_usb_anchor(&dev->deferred); @@ -4439,6 +4484,7 @@ out6: out5: lan78xx_unbind(dev, intf); out4: + netif_napi_del(&dev->napi); lan78xx_free_rx_resources(dev); out3: lan78xx_free_tx_resources(dev); @@ -4938,7 +4984,7 @@ static int lan78xx_resume(struct usb_interface *intf) if (ret < 0) goto out; - tasklet_schedule(&dev->bh); + napi_schedule(&dev->napi); if (!timer_pending(&dev->stat_monitor)) { dev->delta = 1; From 29ad850a5cae84757bcd4c60e0d74232ef8c5157 Mon Sep 17 00:00:00 2001 From: Ilya Leoshkevich Date: Thu, 18 Nov 2021 12:52:25 +0100 Subject: [PATCH 0340/2048] selfetests/bpf: Adapt vmtest.sh to s390 libbpf CI changes [1] added s390 support to libbpf CI and added an ${ARCH} prefix to a number of paths and identifiers in libbpf GitHub repo, which vmtest.sh relies upon. Update these and make use of the new s390 support. [1] https://github.com/libbpf/libbpf/pull/204 Co-developed-by: Andrii Nakryiko Signed-off-by: Andrii Nakryiko Signed-off-by: Ilya Leoshkevich Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211118115225.1349726-1-iii@linux.ibm.com --- tools/testing/selftests/bpf/vmtest.sh | 46 ++++++++++++++++++--------- 1 file changed, 31 insertions(+), 15 deletions(-) diff --git a/tools/testing/selftests/bpf/vmtest.sh b/tools/testing/selftests/bpf/vmtest.sh index 027198768fad..5e43c79ddc6e 100755 --- a/tools/testing/selftests/bpf/vmtest.sh +++ b/tools/testing/selftests/bpf/vmtest.sh @@ -4,17 +4,34 @@ set -u set -e -# This script currently only works for x86_64, as -# it is based on the VM image used by the BPF CI which is -# x86_64. -QEMU_BINARY="${QEMU_BINARY:="qemu-system-x86_64"}" -X86_BZIMAGE="arch/x86/boot/bzImage" +# This script currently only works for x86_64 and s390x, as +# it is based on the VM image used by the BPF CI, which is +# available only for these architectures. +ARCH="$(uname -m)" +case "${ARCH}" in +s390x) + QEMU_BINARY=qemu-system-s390x + QEMU_CONSOLE="ttyS1" + QEMU_FLAGS=(-smp 2) + BZIMAGE="arch/s390/boot/compressed/vmlinux" + ;; +x86_64) + QEMU_BINARY=qemu-system-x86_64 + QEMU_CONSOLE="ttyS0,115200" + QEMU_FLAGS=(-cpu host -smp 8) + BZIMAGE="arch/x86/boot/bzImage" + ;; +*) + echo "Unsupported architecture" + exit 1 + ;; +esac DEFAULT_COMMAND="./test_progs" MOUNT_DIR="mnt" ROOTFS_IMAGE="root.img" OUTPUT_DIR="$HOME/.bpf_selftests" -KCONFIG_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/latest.config" -KCONFIG_API_URL="https://api.github.com/repos/libbpf/libbpf/contents/travis-ci/vmtest/configs/latest.config" +KCONFIG_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/config-latest.${ARCH}" +KCONFIG_API_URL="https://api.github.com/repos/libbpf/libbpf/contents/travis-ci/vmtest/configs/config-latest.${ARCH}" INDEX_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/INDEX" NUM_COMPILE_JOBS="$(nproc)" LOG_FILE_BASE="$(date +"bpf_selftests.%Y-%m-%d_%H-%M-%S")" @@ -85,7 +102,7 @@ newest_rootfs_version() { { for file in "${!URLS[@]}"; do - if [[ $file =~ ^libbpf-vmtest-rootfs-(.*)\.tar\.zst$ ]]; then + if [[ $file =~ ^"${ARCH}"/libbpf-vmtest-rootfs-(.*)\.tar\.zst$ ]]; then echo "${BASH_REMATCH[1]}" fi done @@ -102,7 +119,7 @@ download_rootfs() exit 1 fi - download "libbpf-vmtest-rootfs-$rootfsversion.tar.zst" | + download "${ARCH}/libbpf-vmtest-rootfs-$rootfsversion.tar.zst" | zstd -d | sudo tar -C "$dir" -x } @@ -224,13 +241,12 @@ EOF -nodefaults \ -display none \ -serial mon:stdio \ - -cpu host \ + "${qemu_flags[@]}" \ -enable-kvm \ - -smp 8 \ -m 4G \ -drive file="${rootfs_img}",format=raw,index=1,media=disk,if=virtio,cache=none \ -kernel "${kernel_bzimage}" \ - -append "root=/dev/vda rw console=ttyS0,115200" + -append "root=/dev/vda rw console=${QEMU_CONSOLE}" } copy_logs() @@ -282,7 +298,7 @@ main() local kernel_checkout=$(realpath "${script_dir}"/../../../../) # By default the script searches for the kernel in the checkout directory but # it also obeys environment variables O= and KBUILD_OUTPUT= - local kernel_bzimage="${kernel_checkout}/${X86_BZIMAGE}" + local kernel_bzimage="${kernel_checkout}/${BZIMAGE}" local command="${DEFAULT_COMMAND}" local update_image="no" local exit_command="poweroff -f" @@ -337,13 +353,13 @@ main() if is_rel_path "${O}"; then O="$(realpath "${PWD}/${O}")" fi - kernel_bzimage="${O}/${X86_BZIMAGE}" + kernel_bzimage="${O}/${BZIMAGE}" make_command="${make_command} O=${O}" elif [[ "${KBUILD_OUTPUT:=""}" != "" ]]; then if is_rel_path "${KBUILD_OUTPUT}"; then KBUILD_OUTPUT="$(realpath "${PWD}/${KBUILD_OUTPUT}")" fi - kernel_bzimage="${KBUILD_OUTPUT}/${X86_BZIMAGE}" + kernel_bzimage="${KBUILD_OUTPUT}/${BZIMAGE}" make_command="${make_command} KBUILD_OUTPUT=${KBUILD_OUTPUT}" fi From c6e7871894a3b975ae45995d69cf761566904595 Mon Sep 17 00:00:00 2001 From: luo penghao Date: Thu, 4 Nov 2021 03:19:31 +0000 Subject: [PATCH 0341/2048] ipv6/esp6: Remove structure variables and alignment statements The definition of this variable is just to find the length of the structure after aligning the structure. The PTR alignment function is to optimize the size of the structure. In fact, it doesn't seem to be of much use, because both members of the structure are of type u32. So I think that the definition of the variable and the corresponding alignment can be deleted, the value of extralen can be directly passed in the size of the structure. The clang_analyzer complains as follows: net/ipv6/esp6.c:117:27 warning: Value stored to 'extra' during its initialization is never read Reported-by: Zeal Robot Signed-off-by: luo penghao Signed-off-by: Steffen Klassert --- net/ipv6/esp6.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c index ed2f061b8768..c35c211c9cb7 100644 --- a/net/ipv6/esp6.c +++ b/net/ipv6/esp6.c @@ -114,7 +114,6 @@ static inline struct scatterlist *esp_req_sg(struct crypto_aead *aead, static void esp_ssg_unref(struct xfrm_state *x, void *tmp) { - struct esp_output_extra *extra = esp_tmp_extra(tmp); struct crypto_aead *aead = x->data; int extralen = 0; u8 *iv; @@ -122,7 +121,7 @@ static void esp_ssg_unref(struct xfrm_state *x, void *tmp) struct scatterlist *sg; if (x->props.flags & XFRM_STATE_ESN) - extralen += sizeof(*extra); + extralen += sizeof(struct esp_output_extra); iv = esp_tmp_iv(aead, tmp, extralen); req = esp_tmp_req(aead, iv); From 2e1809208a4a52e883527c288b4fca61a90100b2 Mon Sep 17 00:00:00 2001 From: luo penghao Date: Thu, 4 Nov 2021 06:26:21 +0000 Subject: [PATCH 0342/2048] xfrm: Remove duplicate assignment The statement in the switch is repeated with the statement at the beginning of the while loop, so this statement is meaningless. The clang_analyzer complains as follows: net/xfrm/xfrm_policy.c:3392:2 warning: Value stored to 'exthdr' is never read Reported-by: Zeal Robot Signed-off-by: luo penghao Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_policy.c | 1 - 1 file changed, 1 deletion(-) diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 1a06585022ab..edc673e78114 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -3392,7 +3392,6 @@ decode_session6(struct sk_buff *skb, struct flowi *fl, bool reverse) case NEXTHDR_DEST: offset += ipv6_optlen(exthdr); nexthdr = exthdr->nexthdr; - exthdr = (struct ipv6_opt_hdr *)(nh + offset); break; case IPPROTO_UDP: case IPPROTO_UDPLITE: From 081e2d6476e30399433b509684d5da4d1844e430 Mon Sep 17 00:00:00 2001 From: Seevalamuthu Mariappan Date: Wed, 17 Nov 2021 09:39:41 +0200 Subject: [PATCH 0343/2048] ath11k: add hw_param for wakeup_mhi Wakeup mhi is needed before pci_read/write only for QCA6390 and WCN6855. Since wakeup & release mhi is enabled for all hardwares, below mhi assert is seen in QCN9074 when doing 'rmmod ath11k_pci': Kernel panic - not syncing: dev_wake != 0 CPU: 2 PID: 13535 Comm: procd Not tainted 4.4.60 #1 Hardware name: Generic DT based system [<80316dac>] (unwind_backtrace) from [<80313700>] (show_stack+0x10/0x14) [<80313700>] (show_stack) from [<805135dc>] (dump_stack+0x7c/0x9c) [<805135dc>] (dump_stack) from [<8032136c>] (panic+0x84/0x1f8) [<8032136c>] (panic) from [<80549b24>] (mhi_pm_disable_transition+0x3b8/0x5b8) [<80549b24>] (mhi_pm_disable_transition) from [<80549ddc>] (mhi_power_down+0xb8/0x100) [<80549ddc>] (mhi_power_down) from [<7f5242b0>] (ath11k_mhi_op_status_cb+0x284/0x3ac [ath11k_pci]) [E][__mhi_device_get_sync] Did not enter M0 state, cur_state:RESET pm_state:SHUTDOWN Process [E][__mhi_device_get_sync] Did not enter M0 state, cur_state:RESET pm_state:SHUTDOWN Process [E][__mhi_device_get_sync] Did not enter M0 state, cur_state:RESET pm_state:SHUTDOWN Process [<7f5242b0>] (ath11k_mhi_op_status_cb [ath11k_pci]) from [<7f524878>] (ath11k_mhi_stop+0x10/0x20 [ath11k_pci]) [<7f524878>] (ath11k_mhi_stop [ath11k_pci]) from [<7f525b94>] (ath11k_pci_power_down+0x54/0x90 [ath11k_pci]) [<7f525b94>] (ath11k_pci_power_down [ath11k_pci]) from [<8056b2a8>] (pci_device_shutdown+0x30/0x44) [<8056b2a8>] (pci_device_shutdown) from [<805cfa0c>] (device_shutdown+0x124/0x174) [<805cfa0c>] (device_shutdown) from [<8033aaa4>] (kernel_restart+0xc/0x50) [<8033aaa4>] (kernel_restart) from [<8033ada8>] (SyS_reboot+0x178/0x1ec) [<8033ada8>] (SyS_reboot) from [<80301b80>] (ret_fast_syscall+0x0/0x34) Hence, disable wakeup/release mhi using hw_param for other hardwares. Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.5.0.1-01060-QCAHKSWPL_SILICONZ-1 Fixes: a05bd8513335 ("ath11k: read and write registers below unwindowed address") Signed-off-by: Seevalamuthu Mariappan Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/1636702019-26142-1-git-send-email-quic_seevalam@quicinc.com --- drivers/net/wireless/ath/ath11k/core.c | 5 +++++ drivers/net/wireless/ath/ath11k/hw.h | 1 + drivers/net/wireless/ath/ath11k/pci.c | 12 ++++++++---- 3 files changed, 14 insertions(+), 4 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c index a40bbca3e9af..dd1a1bb078c3 100644 --- a/drivers/net/wireless/ath/ath11k/core.c +++ b/drivers/net/wireless/ath/ath11k/core.c @@ -86,6 +86,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, .alloc_cacheable_memory = true, + .wakeup_mhi = false, }, { .hw_rev = ATH11K_HW_IPQ6018_HW10, @@ -139,6 +140,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = false, .alloc_cacheable_memory = true, + .wakeup_mhi = false, }, { .name = "qca6390 hw2.0", @@ -191,6 +193,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, .alloc_cacheable_memory = false, + .wakeup_mhi = true, }, { .name = "qcn9074 hw1.0", @@ -243,6 +246,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .hal_params = &ath11k_hw_hal_params_ipq8074, .supports_dynamic_smps_6ghz = true, .alloc_cacheable_memory = true, + .wakeup_mhi = false, }, { .name = "wcn6855 hw2.0", @@ -295,6 +299,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = { .hal_params = &ath11k_hw_hal_params_qca6390, .supports_dynamic_smps_6ghz = false, .alloc_cacheable_memory = false, + .wakeup_mhi = true, }, }; diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h index dd2202f23b99..2c9d232ebfed 100644 --- a/drivers/net/wireless/ath/ath11k/hw.h +++ b/drivers/net/wireless/ath/ath11k/hw.h @@ -180,6 +180,7 @@ struct ath11k_hw_params { const struct ath11k_hw_hal_params *hal_params; bool supports_dynamic_smps_6ghz; bool alloc_cacheable_memory; + bool wakeup_mhi; }; struct ath11k_hw_ops { diff --git a/drivers/net/wireless/ath/ath11k/pci.c b/drivers/net/wireless/ath/ath11k/pci.c index 958c2b2c5b3e..c2af8184e4a2 100644 --- a/drivers/net/wireless/ath/ath11k/pci.c +++ b/drivers/net/wireless/ath/ath11k/pci.c @@ -182,7 +182,8 @@ void ath11k_pci_write32(struct ath11k_base *ab, u32 offset, u32 value) /* for offset beyond BAR + 4K - 32, may * need to wakeup MHI to access. */ - if (test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && + if (ab->hw_params.wakeup_mhi && + test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && offset >= ACCESS_ALWAYS_OFF) mhi_device_get_sync(ab_pci->mhi_ctrl->mhi_dev); @@ -206,7 +207,8 @@ void ath11k_pci_write32(struct ath11k_base *ab, u32 offset, u32 value) } } - if (test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && + if (ab->hw_params.wakeup_mhi && + test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && offset >= ACCESS_ALWAYS_OFF) mhi_device_put(ab_pci->mhi_ctrl->mhi_dev); } @@ -219,7 +221,8 @@ u32 ath11k_pci_read32(struct ath11k_base *ab, u32 offset) /* for offset beyond BAR + 4K - 32, may * need to wakeup MHI to access. */ - if (test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && + if (ab->hw_params.wakeup_mhi && + test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && offset >= ACCESS_ALWAYS_OFF) mhi_device_get_sync(ab_pci->mhi_ctrl->mhi_dev); @@ -243,7 +246,8 @@ u32 ath11k_pci_read32(struct ath11k_base *ab, u32 offset) } } - if (test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && + if (ab->hw_params.wakeup_mhi && + test_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags) && offset >= ACCESS_ALWAYS_OFF) mhi_device_put(ab_pci->mhi_ctrl->mhi_dev); From 5125b9a9c420ee91e2c258ea96a8261a49e7449d Mon Sep 17 00:00:00 2001 From: Peter Seiderer Date: Tue, 16 Nov 2021 23:07:20 +0100 Subject: [PATCH 0344/2048] ath9k: fix intr_txqs setting The struct ath_hw member intr_txqs is never reset/assigned outside of ath9k_hw_init_queues() and with the used bitwise-or in the interrupt handling ar9002_hw_get_isr() accumulates all ever set interrupt flags. Fix this by using a pure assign instead of bitwise-or for the first line (note: intr_txqs is only evaluated in case ATH9K_INT_TX bit is set). Signed-off-by: Peter Seiderer Signed-off-by: Kalle Valo Link: https://lore.kernel.org/r/20211116220720.30145-1-ps.report@gmx.net --- drivers/net/wireless/ath/ath9k/ar9002_mac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath9k/ar9002_mac.c b/drivers/net/wireless/ath/ath9k/ar9002_mac.c index ce9a0a53771e..fba5a847c3bb 100644 --- a/drivers/net/wireless/ath/ath9k/ar9002_mac.c +++ b/drivers/net/wireless/ath/ath9k/ar9002_mac.c @@ -120,7 +120,7 @@ static bool ar9002_hw_get_isr(struct ath_hw *ah, enum ath9k_int *masked, AR_ISR_TXEOL); } - ah->intr_txqs |= MS(s0_s, AR_ISR_S0_QCU_TXOK); + ah->intr_txqs = MS(s0_s, AR_ISR_S0_QCU_TXOK); ah->intr_txqs |= MS(s0_s, AR_ISR_S0_QCU_TXDESC); ah->intr_txqs |= MS(s1_s, AR_ISR_S1_QCU_TXERR); ah->intr_txqs |= MS(s1_s, AR_ISR_S1_QCU_TXEOL); From bc2dfc02836b1133d1bf4d22aa13d48ac98eabef Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Sat, 23 Oct 2021 11:10:50 +0200 Subject: [PATCH 0345/2048] cfg80211: implement APIs for dedicated radar detection HW If a dedicated (off-channel) radar detection hardware (chain) is available in the hardware/driver, allow this to be used by calling the NL80211_CMD_RADAR_DETECT command with a new flag attribute requesting off-channel radar detection is used. Offchannel CAC (channel availability check) avoids the CAC downtime when switching to a radar channel or when turning on the AP. Drivers advertise support for this using the new feature flag NL80211_EXT_FEATURE_RADAR_OFFCHAN. Tested-by: Evelyn Tsai Signed-off-by: Lorenzo Bianconi Link: https://lore.kernel.org/r/7468e291ef5d05d692c1738d25b8f778d8ea5c3f.1634979655.git.lorenzo@kernel.org Link: https://lore.kernel.org/r/1e60e60fef00e14401adae81c3d49f3e5f307537.1634979655.git.lorenzo@kernel.org Link: https://lore.kernel.org/r/85fa50f57fc3adb2934c8d9ca0be30394de6b7e8.1634979655.git.lorenzo@kernel.org Link: https://lore.kernel.org/r/4b6c08671ad59aae0ac46fc94c02f31b1610eb72.1634979655.git.lorenzo@kernel.org Link: https://lore.kernel.org/r/241849ccaf2c228873c6f8495bf87b19159ba458.1634979655.git.lorenzo@kernel.org [remove offchan_mutex, fix cfg80211_stop_offchan_radar_detection(), remove gfp_t argument, fix documentation, fix tracing] Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 25 ++++++++ include/uapi/linux/nl80211.h | 13 ++++ net/wireless/core.c | 3 + net/wireless/core.h | 13 ++++ net/wireless/mlme.c | 113 +++++++++++++++++++++++++++++++++++ net/wireless/nl80211.c | 17 ++++-- net/wireless/rdev-ops.h | 17 ++++++ net/wireless/trace.h | 19 ++++++ 8 files changed, 214 insertions(+), 6 deletions(-) diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 423f97b982ff..db8866d42a4b 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -4072,6 +4072,15 @@ struct mgmt_frame_regs { * @set_fils_aad: Set FILS AAD data to the AP driver so that the driver can use * those to decrypt (Re)Association Request and encrypt (Re)Association * Response frame. + * + * @set_radar_offchan: Configure dedicated offchannel chain available for + * radar/CAC detection on some hw. This chain can't be used to transmit + * or receive frames and it is bounded to a running wdev. + * Offchannel radar/CAC detection allows to avoid the CAC downtime + * switching to a different channel during CAC detection on the selected + * radar channel. + * The caller is expected to set chandef pointer to NULL in order to + * disable offchannel CAC/radar detection. */ struct cfg80211_ops { int (*suspend)(struct wiphy *wiphy, struct cfg80211_wowlan *wow); @@ -4404,6 +4413,8 @@ struct cfg80211_ops { struct cfg80211_color_change_settings *params); int (*set_fils_aad)(struct wiphy *wiphy, struct net_device *dev, struct cfg80211_fils_aad *fils_aad); + int (*set_radar_offchan)(struct wiphy *wiphy, + struct cfg80211_chan_def *chandef); }; /* @@ -7633,6 +7644,20 @@ void cfg80211_cac_event(struct net_device *netdev, const struct cfg80211_chan_def *chandef, enum nl80211_radar_event event, gfp_t gfp); +/** + * cfg80211_offchan_cac_event - Channel Availability Check (CAC) offchan event + * @wiphy: the wiphy + * @chandef: chandef for the current channel + * @event: type of event + * + * This function is called when a Channel Availability Check (CAC) is finished, + * started or aborted by a offchannel dedicated chain. + * + * Note that this acquires the wiphy lock. + */ +void cfg80211_offchan_cac_event(struct wiphy *wiphy, + const struct cfg80211_chan_def *chandef, + enum nl80211_radar_event event); /** * cfg80211_gtk_rekey_notify - notify userspace about driver rekeying diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 61cab81e920d..3e734826792f 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -2639,6 +2639,13 @@ enum nl80211_commands { * Mandatory parameter for the transmitting interface to enable MBSSID. * Optional for the non-transmitting interfaces. * + * @NL80211_ATTR_RADAR_OFFCHAN: Configure dedicated offchannel chain available for + * radar/CAC detection on some hw. This chain can't be used to transmit + * or receive frames and it is bounded to a running wdev. + * Offchannel radar/CAC detection allows to avoid the CAC downtime + * switching on a different channel during CAC detection on the selected + * radar channel. + * * @NUM_NL80211_ATTR: total number of nl80211_attrs available * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use @@ -3145,6 +3152,8 @@ enum nl80211_attrs { NL80211_ATTR_MBSSID_CONFIG, NL80211_ATTR_MBSSID_ELEMS, + NL80211_ATTR_RADAR_OFFCHAN, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -6051,6 +6060,9 @@ enum nl80211_feature_flags { * frames. Userspace has to share FILS AAD details to the driver by using * @NL80211_CMD_SET_FILS_AAD. * + * @NL80211_EXT_FEATURE_RADAR_OFFCHAN: Device supports offchannel radar/CAC + * detection. + * * @NUM_NL80211_EXT_FEATURES: number of extended features. * @MAX_NL80211_EXT_FEATURES: highest extended feature index. */ @@ -6117,6 +6129,7 @@ enum nl80211_ext_feature_index { NL80211_EXT_FEATURE_PROT_RANGE_NEGO_AND_MEASURE, NL80211_EXT_FEATURE_BSS_COLOR, NL80211_EXT_FEATURE_FILS_CRYPTO_OFFLOAD, + NL80211_EXT_FEATURE_RADAR_OFFCHAN, /* add new features before the definition below */ NUM_NL80211_EXT_FEATURES, diff --git a/net/wireless/core.c b/net/wireless/core.c index eb297e1015e0..39b2d4ae581d 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -545,6 +545,7 @@ use_default_name: INIT_WORK(&rdev->rfkill_block, cfg80211_rfkill_block_work); INIT_WORK(&rdev->conn_work, cfg80211_conn_work); INIT_WORK(&rdev->event_work, cfg80211_event_work); + INIT_DELAYED_WORK(&rdev->offchan_cac_work, cfg80211_offchan_cac_work); init_waitqueue_head(&rdev->dev_wait); @@ -1207,6 +1208,8 @@ void __cfg80211_leave(struct cfg80211_registered_device *rdev, cfg80211_pmsr_wdev_down(wdev); + cfg80211_stop_offchan_radar_detection(wdev); + switch (wdev->iftype) { case NL80211_IFTYPE_ADHOC: __cfg80211_leave_ibss(rdev, dev, true); diff --git a/net/wireless/core.h b/net/wireless/core.h index 1720abf36f92..612d460dcde0 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -84,6 +84,10 @@ struct cfg80211_registered_device { struct delayed_work dfs_update_channels_wk; + struct wireless_dev *offchan_radar_wdev; + struct cfg80211_chan_def offchan_radar_chandef; + struct delayed_work offchan_cac_work; + /* netlink port which started critical protocol (0 means not started) */ u32 crit_proto_nlportid; @@ -491,6 +495,15 @@ cfg80211_chandef_dfs_cac_time(struct wiphy *wiphy, void cfg80211_sched_dfs_chan_update(struct cfg80211_registered_device *rdev); +int +cfg80211_start_offchan_radar_detection(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev, + struct cfg80211_chan_def *chandef); + +void cfg80211_stop_offchan_radar_detection(struct wireless_dev *wdev); + +void cfg80211_offchan_cac_work(struct work_struct *work); + bool cfg80211_any_wiphy_oper_chan(struct wiphy *wiphy, struct ieee80211_channel *chan); diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 783acd2c4211..46f2ec4d50d7 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -970,3 +970,116 @@ void cfg80211_cac_event(struct net_device *netdev, nl80211_radar_notify(rdev, chandef, event, netdev, gfp); } EXPORT_SYMBOL(cfg80211_cac_event); + +void cfg80211_offchan_cac_work(struct work_struct *work) +{ + struct delayed_work *delayed_work = to_delayed_work(work); + struct cfg80211_registered_device *rdev; + + rdev = container_of(delayed_work, struct cfg80211_registered_device, + offchan_cac_work); + cfg80211_offchan_cac_event(&rdev->wiphy, &rdev->offchan_radar_chandef, + NL80211_RADAR_CAC_FINISHED); +} + +static void +__cfg80211_offchan_cac_event(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev, + const struct cfg80211_chan_def *chandef, + enum nl80211_radar_event event) +{ + struct wiphy *wiphy = &rdev->wiphy; + struct net_device *netdev; + + lockdep_assert_wiphy(&rdev->wiphy); + + if (event != NL80211_RADAR_CAC_STARTED && !rdev->offchan_radar_wdev) + return; + + switch (event) { + case NL80211_RADAR_CAC_FINISHED: + cfg80211_set_dfs_state(wiphy, chandef, NL80211_DFS_AVAILABLE); + memcpy(&rdev->cac_done_chandef, chandef, sizeof(*chandef)); + queue_work(cfg80211_wq, &rdev->propagate_cac_done_wk); + cfg80211_sched_dfs_chan_update(rdev); + wdev = rdev->offchan_radar_wdev; + rdev->offchan_radar_wdev = NULL; + break; + case NL80211_RADAR_CAC_ABORTED: + cancel_delayed_work(&rdev->offchan_cac_work); + wdev = rdev->offchan_radar_wdev; + rdev->offchan_radar_wdev = NULL; + break; + case NL80211_RADAR_CAC_STARTED: + WARN_ON(!wdev); + rdev->offchan_radar_wdev = wdev; + break; + default: + return; + } + + netdev = wdev ? wdev->netdev : NULL; + nl80211_radar_notify(rdev, chandef, event, netdev, GFP_KERNEL); +} + +void cfg80211_offchan_cac_event(struct wiphy *wiphy, + const struct cfg80211_chan_def *chandef, + enum nl80211_radar_event event) +{ + struct cfg80211_registered_device *rdev = wiphy_to_rdev(wiphy); + + wiphy_lock(wiphy); + __cfg80211_offchan_cac_event(rdev, NULL, chandef, event); + wiphy_unlock(wiphy); +} +EXPORT_SYMBOL(cfg80211_offchan_cac_event); + +int +cfg80211_start_offchan_radar_detection(struct cfg80211_registered_device *rdev, + struct wireless_dev *wdev, + struct cfg80211_chan_def *chandef) +{ + unsigned int cac_time_ms; + int err; + + lockdep_assert_wiphy(&rdev->wiphy); + + if (!wiphy_ext_feature_isset(&rdev->wiphy, + NL80211_EXT_FEATURE_RADAR_OFFCHAN)) + return -EOPNOTSUPP; + + if (rdev->offchan_radar_wdev) + return -EBUSY; + + err = rdev_set_radar_offchan(rdev, chandef); + if (err) + return err; + + cac_time_ms = cfg80211_chandef_dfs_cac_time(&rdev->wiphy, chandef); + if (!cac_time_ms) + cac_time_ms = IEEE80211_DFS_MIN_CAC_TIME_MS; + + rdev->offchan_radar_chandef = *chandef; + __cfg80211_offchan_cac_event(rdev, wdev, chandef, + NL80211_RADAR_CAC_STARTED); + queue_delayed_work(cfg80211_wq, &rdev->offchan_cac_work, + msecs_to_jiffies(cac_time_ms)); + + return 0; +} + +void cfg80211_stop_offchan_radar_detection(struct wireless_dev *wdev) +{ + struct wiphy *wiphy = wdev->wiphy; + struct cfg80211_registered_device *rdev = wiphy_to_rdev(wiphy); + + lockdep_assert_wiphy(wiphy); + + if (wdev != rdev->offchan_radar_wdev) + return; + + rdev_set_radar_offchan(rdev, NULL); + + __cfg80211_offchan_cac_event(rdev, NULL, NULL, + NL80211_RADAR_CAC_ABORTED); +} diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index a27b3b5fa210..83a1ba96e172 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -776,6 +776,7 @@ static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [NL80211_ATTR_MBSSID_CONFIG] = NLA_POLICY_NESTED(nl80211_mbssid_config_policy), [NL80211_ATTR_MBSSID_ELEMS] = { .type = NLA_NESTED }, + [NL80211_ATTR_RADAR_OFFCHAN] = { .type = NLA_FLAG }, }; /* policy for the key attributes */ @@ -9284,12 +9285,6 @@ static int nl80211_start_radar_detection(struct sk_buff *skb, if (err) return err; - if (netif_carrier_ok(dev)) - return -EBUSY; - - if (wdev->cac_started) - return -EBUSY; - err = cfg80211_chandef_dfs_required(wiphy, &chandef, wdev->iftype); if (err < 0) return err; @@ -9300,6 +9295,16 @@ static int nl80211_start_radar_detection(struct sk_buff *skb, if (!cfg80211_chandef_dfs_usable(wiphy, &chandef)) return -EINVAL; + if (nla_get_flag(info->attrs[NL80211_ATTR_RADAR_OFFCHAN])) + return cfg80211_start_offchan_radar_detection(rdev, wdev, + &chandef); + + if (netif_carrier_ok(dev)) + return -EBUSY; + + if (wdev->cac_started) + return -EBUSY; + /* CAC start is offloaded to HW and can't be started manually */ if (wiphy_ext_feature_isset(wiphy, NL80211_EXT_FEATURE_DFS_OFFLOAD)) return -EOPNOTSUPP; diff --git a/net/wireless/rdev-ops.h b/net/wireless/rdev-ops.h index cc1efec4b27b..8672b3ef99e4 100644 --- a/net/wireless/rdev-ops.h +++ b/net/wireless/rdev-ops.h @@ -1395,4 +1395,21 @@ rdev_set_fils_aad(struct cfg80211_registered_device *rdev, return ret; } +static inline int +rdev_set_radar_offchan(struct cfg80211_registered_device *rdev, + struct cfg80211_chan_def *chandef) +{ + struct wiphy *wiphy = &rdev->wiphy; + int ret; + + if (!rdev->ops->set_radar_offchan) + return -EOPNOTSUPP; + + trace_rdev_set_radar_offchan(wiphy, chandef); + ret = rdev->ops->set_radar_offchan(wiphy, chandef); + trace_rdev_return_int(wiphy, ret); + + return ret; +} + #endif /* __CFG80211_RDEV_OPS */ diff --git a/net/wireless/trace.h b/net/wireless/trace.h index ad6c16a06bcb..0b27eaa14a18 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -3674,6 +3674,25 @@ TRACE_EVENT(cfg80211_bss_color_notify, __entry->color_bitmap) ); +TRACE_EVENT(rdev_set_radar_offchan, + TP_PROTO(struct wiphy *wiphy, struct cfg80211_chan_def *chandef), + + TP_ARGS(wiphy, chandef), + + TP_STRUCT__entry( + WIPHY_ENTRY + CHAN_DEF_ENTRY + ), + + TP_fast_assign( + WIPHY_ASSIGN; + CHAN_DEF_ASSIGN(chandef) + ), + + TP_printk(WIPHY_PR_FMT ", " CHAN_DEF_PR_FMT, + WIPHY_PR_ARG, CHAN_DEF_PR_ARG) +); + #endif /* !__RDEV_OPS_TRACE || TRACE_HEADER_MULTI_READ */ #undef TRACE_INCLUDE_PATH From 237337c230b94e78a5a0f88d1705259ab543fc40 Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Sat, 23 Oct 2021 11:10:51 +0200 Subject: [PATCH 0346/2048] mac80211: introduce set_radar_offchan callback Similar to cfg80211, introduce set_radar_offchan callback in mac80211_ops in order to configure a dedicated offchannel chain available on some hw (e.g. mt7915) to perform offchannel CAC detection and avoid tx/rx downtime. Tested-by: Evelyn Tsai Signed-off-by: Lorenzo Bianconi Link: https://lore.kernel.org/r/201110606d4f3a7dfdf31440e351f2e2c375d4f0.1634979655.git.lorenzo@kernel.org Signed-off-by: Johannes Berg --- include/net/mac80211.h | 10 ++++++++++ net/mac80211/cfg.c | 13 +++++++++++++ 2 files changed, 23 insertions(+) diff --git a/include/net/mac80211.h b/include/net/mac80211.h index dd757f0987b0..775dbb982654 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -3944,6 +3944,14 @@ struct ieee80211_prep_tx_info { * twt structure. * @twt_teardown_request: Update the hw with TWT teardown request received * from the peer. + * @set_radar_offchan: Configure dedicated offchannel chain available for + * radar/CAC detection on some hw. This chain can't be used to transmit + * or receive frames and it is bounded to a running wdev. + * Offchannel radar/CAC detection allows to avoid the CAC downtime + * switching to a different channel during CAC detection on the selected + * radar channel. + * The caller is expected to set chandef pointer to NULL in order to + * disable offchannel CAC/radar detection. */ struct ieee80211_ops { void (*tx)(struct ieee80211_hw *hw, @@ -4272,6 +4280,8 @@ struct ieee80211_ops { struct ieee80211_twt_setup *twt); void (*twt_teardown_request)(struct ieee80211_hw *hw, struct ieee80211_sta *sta, u8 flowid); + int (*set_radar_offchan)(struct ieee80211_hw *hw, + struct cfg80211_chan_def *chandef); }; /** diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index bd3d3195097f..45334d59fe06 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -4383,6 +4383,18 @@ out: return err; } +static int +ieee80211_set_radar_offchan(struct wiphy *wiphy, + struct cfg80211_chan_def *chandef) +{ + struct ieee80211_local *local = wiphy_priv(wiphy); + + if (!local->ops->set_radar_offchan) + return -EOPNOTSUPP; + + return local->ops->set_radar_offchan(&local->hw, chandef); +} + const struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -4487,4 +4499,5 @@ const struct cfg80211_ops mac80211_config_ops = { .reset_tid_config = ieee80211_reset_tid_config, .set_sar_specs = ieee80211_set_sar_specs, .color_change = ieee80211_color_change, + .set_radar_offchan = ieee80211_set_radar_offchan, }; From f5d32a7b10713427655a14d4777af7f598d3c1fa Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Tue, 26 Oct 2021 10:40:00 +0100 Subject: [PATCH 0347/2048] mac80211_hwsim: Fix spelling mistake "Droping" -> "Dropping" There is a spelling mistake in a comment, fix it. Signed-off-by: Colin Ian King Link: https://lore.kernel.org/r/20211026094000.209463-1-colin.i.king@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/mac80211_hwsim.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c index 23219f3747f8..0307a6677907 100644 --- a/drivers/net/wireless/mac80211_hwsim.c +++ b/drivers/net/wireless/mac80211_hwsim.c @@ -1276,7 +1276,7 @@ static void mac80211_hwsim_tx_frame_nl(struct ieee80211_hw *hw, hdr->frame_control |= cpu_to_le16(IEEE80211_FCTL_PM); /* If the queue contains MAX_QUEUE skb's drop some */ if (skb_queue_len(&data->pending) >= MAX_QUEUE) { - /* Droping until WARN_QUEUE level */ + /* Dropping until WARN_QUEUE level */ while (skb_queue_len(&data->pending) >= WARN_QUEUE) { ieee80211_free_txskb(hw, skb_dequeue(&data->pending)); data->tx_dropped++; From 1507b153198137dfa9cb4bec7c5dee07089ec3af Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Wed, 27 Oct 2021 11:03:42 +0200 Subject: [PATCH 0348/2048] cfg80211: move offchan_cac_event to a dedicated work In order to make cfg80211_offchan_cac_abort() (renamed from cfg80211_offchan_cac_event) callable in other contexts and without so much locking restrictions, make it trigger a new work instead of operating directly. Do some other renames while at it to clarify. Signed-off-by: Lorenzo Bianconi Link: https://lore.kernel.org/r/6145c3d0f30400a568023f67981981d24c7c6133.1635325205.git.lorenzo@kernel.org [rewrite commit log] Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 14 +++-------- net/wireless/core.c | 6 ++++- net/wireless/core.h | 7 ++++-- net/wireless/mlme.c | 56 +++++++++++++++++++++++++++--------------- 4 files changed, 50 insertions(+), 33 deletions(-) diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index db8866d42a4b..362da9f6bf39 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -7645,19 +7645,13 @@ void cfg80211_cac_event(struct net_device *netdev, enum nl80211_radar_event event, gfp_t gfp); /** - * cfg80211_offchan_cac_event - Channel Availability Check (CAC) offchan event + * cfg80211_offchan_cac_abort - Channel Availability Check offchan abort event * @wiphy: the wiphy - * @chandef: chandef for the current channel - * @event: type of event * - * This function is called when a Channel Availability Check (CAC) is finished, - * started or aborted by a offchannel dedicated chain. - * - * Note that this acquires the wiphy lock. + * This function is called by the driver when a Channel Availability Check + * (CAC) is aborted by a offchannel dedicated chain. */ -void cfg80211_offchan_cac_event(struct wiphy *wiphy, - const struct cfg80211_chan_def *chandef, - enum nl80211_radar_event event); +void cfg80211_offchan_cac_abort(struct wiphy *wiphy); /** * cfg80211_gtk_rekey_notify - notify userspace about driver rekeying diff --git a/net/wireless/core.c b/net/wireless/core.c index 39b2d4ae581d..c4ea903f8184 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -545,7 +545,9 @@ use_default_name: INIT_WORK(&rdev->rfkill_block, cfg80211_rfkill_block_work); INIT_WORK(&rdev->conn_work, cfg80211_conn_work); INIT_WORK(&rdev->event_work, cfg80211_event_work); - INIT_DELAYED_WORK(&rdev->offchan_cac_work, cfg80211_offchan_cac_work); + INIT_WORK(&rdev->offchan_cac_abort_wk, cfg80211_offchan_cac_abort_wk); + INIT_DELAYED_WORK(&rdev->offchan_cac_done_wk, + cfg80211_offchan_cac_done_wk); init_waitqueue_head(&rdev->dev_wait); @@ -1055,11 +1057,13 @@ void wiphy_unregister(struct wiphy *wiphy) cancel_work_sync(&rdev->conn_work); flush_work(&rdev->event_work); cancel_delayed_work_sync(&rdev->dfs_update_channels_wk); + cancel_delayed_work_sync(&rdev->offchan_cac_done_wk); flush_work(&rdev->destroy_work); flush_work(&rdev->sched_scan_stop_wk); flush_work(&rdev->propagate_radar_detect_wk); flush_work(&rdev->propagate_cac_done_wk); flush_work(&rdev->mgmt_registrations_update_wk); + flush_work(&rdev->offchan_cac_abort_wk); #ifdef CONFIG_PM if (rdev->wiphy.wowlan_config && rdev->ops->set_wakeup) diff --git a/net/wireless/core.h b/net/wireless/core.h index 612d460dcde0..fb8d9006d838 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -86,7 +86,8 @@ struct cfg80211_registered_device { struct wireless_dev *offchan_radar_wdev; struct cfg80211_chan_def offchan_radar_chandef; - struct delayed_work offchan_cac_work; + struct delayed_work offchan_cac_done_wk; + struct work_struct offchan_cac_abort_wk; /* netlink port which started critical protocol (0 means not started) */ u32 crit_proto_nlportid; @@ -502,7 +503,9 @@ cfg80211_start_offchan_radar_detection(struct cfg80211_registered_device *rdev, void cfg80211_stop_offchan_radar_detection(struct wireless_dev *wdev); -void cfg80211_offchan_cac_work(struct work_struct *work); +void cfg80211_offchan_cac_done_wk(struct work_struct *work); + +void cfg80211_offchan_cac_abort_wk(struct work_struct *work); bool cfg80211_any_wiphy_oper_chan(struct wiphy *wiphy, struct ieee80211_channel *chan); diff --git a/net/wireless/mlme.c b/net/wireless/mlme.c index 46f2ec4d50d7..840795828d4f 100644 --- a/net/wireless/mlme.c +++ b/net/wireless/mlme.c @@ -971,17 +971,6 @@ void cfg80211_cac_event(struct net_device *netdev, } EXPORT_SYMBOL(cfg80211_cac_event); -void cfg80211_offchan_cac_work(struct work_struct *work) -{ - struct delayed_work *delayed_work = to_delayed_work(work); - struct cfg80211_registered_device *rdev; - - rdev = container_of(delayed_work, struct cfg80211_registered_device, - offchan_cac_work); - cfg80211_offchan_cac_event(&rdev->wiphy, &rdev->offchan_radar_chandef, - NL80211_RADAR_CAC_FINISHED); -} - static void __cfg80211_offchan_cac_event(struct cfg80211_registered_device *rdev, struct wireless_dev *wdev, @@ -1006,7 +995,7 @@ __cfg80211_offchan_cac_event(struct cfg80211_registered_device *rdev, rdev->offchan_radar_wdev = NULL; break; case NL80211_RADAR_CAC_ABORTED: - cancel_delayed_work(&rdev->offchan_cac_work); + cancel_delayed_work(&rdev->offchan_cac_done_wk); wdev = rdev->offchan_radar_wdev; rdev->offchan_radar_wdev = NULL; break; @@ -1022,17 +1011,44 @@ __cfg80211_offchan_cac_event(struct cfg80211_registered_device *rdev, nl80211_radar_notify(rdev, chandef, event, netdev, GFP_KERNEL); } -void cfg80211_offchan_cac_event(struct wiphy *wiphy, - const struct cfg80211_chan_def *chandef, - enum nl80211_radar_event event) +static void +cfg80211_offchan_cac_event(struct cfg80211_registered_device *rdev, + const struct cfg80211_chan_def *chandef, + enum nl80211_radar_event event) +{ + wiphy_lock(&rdev->wiphy); + __cfg80211_offchan_cac_event(rdev, NULL, chandef, event); + wiphy_unlock(&rdev->wiphy); +} + +void cfg80211_offchan_cac_done_wk(struct work_struct *work) +{ + struct delayed_work *delayed_work = to_delayed_work(work); + struct cfg80211_registered_device *rdev; + + rdev = container_of(delayed_work, struct cfg80211_registered_device, + offchan_cac_done_wk); + cfg80211_offchan_cac_event(rdev, &rdev->offchan_radar_chandef, + NL80211_RADAR_CAC_FINISHED); +} + +void cfg80211_offchan_cac_abort_wk(struct work_struct *work) +{ + struct cfg80211_registered_device *rdev; + + rdev = container_of(work, struct cfg80211_registered_device, + offchan_cac_abort_wk); + cfg80211_offchan_cac_event(rdev, &rdev->offchan_radar_chandef, + NL80211_RADAR_CAC_ABORTED); +} + +void cfg80211_offchan_cac_abort(struct wiphy *wiphy) { struct cfg80211_registered_device *rdev = wiphy_to_rdev(wiphy); - wiphy_lock(wiphy); - __cfg80211_offchan_cac_event(rdev, NULL, chandef, event); - wiphy_unlock(wiphy); + queue_work(cfg80211_wq, &rdev->offchan_cac_abort_wk); } -EXPORT_SYMBOL(cfg80211_offchan_cac_event); +EXPORT_SYMBOL(cfg80211_offchan_cac_abort); int cfg80211_start_offchan_radar_detection(struct cfg80211_registered_device *rdev, @@ -1062,7 +1078,7 @@ cfg80211_start_offchan_radar_detection(struct cfg80211_registered_device *rdev, rdev->offchan_radar_chandef = *chandef; __cfg80211_offchan_cac_event(rdev, wdev, chandef, NL80211_RADAR_CAC_STARTED); - queue_delayed_work(cfg80211_wq, &rdev->offchan_cac_work, + queue_delayed_work(cfg80211_wq, &rdev->offchan_cac_done_wk, msecs_to_jiffies(cac_time_ms)); return 0; From e291422c8f008c47a04788e98d557dc961d72777 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 06:27:17 -0800 Subject: [PATCH 0349/2048] net: ax88796c: don't write to netdev->dev_addr directly The future is here, convert the new driver as we are about to make netdev->dev_addr const. Acked-by: Lukasz Stelmach Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/asix/ax88796c_main.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/asix/ax88796c_main.c b/drivers/net/ethernet/asix/ax88796c_main.c index e230d8d0ff73..e7a9f9863258 100644 --- a/drivers/net/ethernet/asix/ax88796c_main.c +++ b/drivers/net/ethernet/asix/ax88796c_main.c @@ -144,12 +144,13 @@ static void ax88796c_set_mac_addr(struct net_device *ndev) static void ax88796c_load_mac_addr(struct net_device *ndev) { struct ax88796c_device *ax_local = to_ax88796c_device(ndev); + u8 addr[ETH_ALEN]; u16 temp; lockdep_assert_held(&ax_local->spi_lock); /* Try the device tree first */ - if (!eth_platform_get_mac_address(&ax_local->spi->dev, ndev->dev_addr) && + if (!platform_get_ethdev_address(&ax_local->spi->dev, ndev) && is_valid_ether_addr(ndev->dev_addr)) { if (netif_msg_probe(ax_local)) dev_info(&ax_local->spi->dev, @@ -159,18 +160,19 @@ static void ax88796c_load_mac_addr(struct net_device *ndev) /* Read the MAC address from AX88796C */ temp = AX_READ(&ax_local->ax_spi, P3_MACASR0); - ndev->dev_addr[5] = (u8)temp; - ndev->dev_addr[4] = (u8)(temp >> 8); + addr[5] = (u8)temp; + addr[4] = (u8)(temp >> 8); temp = AX_READ(&ax_local->ax_spi, P3_MACASR1); - ndev->dev_addr[3] = (u8)temp; - ndev->dev_addr[2] = (u8)(temp >> 8); + addr[3] = (u8)temp; + addr[2] = (u8)(temp >> 8); temp = AX_READ(&ax_local->ax_spi, P3_MACASR2); - ndev->dev_addr[1] = (u8)temp; - ndev->dev_addr[0] = (u8)(temp >> 8); + addr[1] = (u8)temp; + addr[0] = (u8)(temp >> 8); - if (is_valid_ether_addr(ndev->dev_addr)) { + if (is_valid_ether_addr(addr)) { + eth_hw_addr_set(ndev, addr); if (netif_msg_probe(ax_local)) dev_info(&ax_local->spi->dev, "MAC address read from ASIX chip\n"); From 54612b4a8bc78f4fc62f6c5639c89e4ea5a2ac79 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 06:27:18 -0800 Subject: [PATCH 0350/2048] mlxsw: constify address in mlxsw_sp_port_dev_addr_set Argument comes from netdev->dev_addr directly, it needs a const. Reviewed-by: Ido Schimmel Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 5925db386b1b..4ce07f9905f6 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -303,7 +303,7 @@ int mlxsw_sp_port_admin_status_set(struct mlxsw_sp_port *mlxsw_sp_port, } static int mlxsw_sp_port_dev_addr_set(struct mlxsw_sp_port *mlxsw_sp_port, - unsigned char *addr) + const unsigned char *addr) { struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; char ppad_pl[MLXSW_REG_PPAD_LEN]; From b09d58025e3c6c78a874c234769cc95b9276e94e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 06:27:19 -0800 Subject: [PATCH 0351/2048] wilc1000: copy address before calling wilc_set_mac_address wilc_set_mac_address() calls IO routines which don't guarantee the pointer won't be written to. Make a copy. Acked-by: Kalle Valo Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/wireless/microchip/wilc1000/netdev.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/wireless/microchip/wilc1000/netdev.c b/drivers/net/wireless/microchip/wilc1000/netdev.c index 690572e01a2a..4712cd7dff9f 100644 --- a/drivers/net/wireless/microchip/wilc1000/netdev.c +++ b/drivers/net/wireless/microchip/wilc1000/netdev.c @@ -574,6 +574,7 @@ static int wilc_mac_open(struct net_device *ndev) struct wilc *wl = vif->wilc; int ret = 0; struct mgmt_frame_regs mgmt_regs = {}; + u8 addr[ETH_ALEN] __aligned(2); if (!wl || !wl->dev) { netdev_err(ndev, "device not ready\n"); @@ -596,10 +597,9 @@ static int wilc_mac_open(struct net_device *ndev) vif->idx); if (is_valid_ether_addr(ndev->dev_addr)) { - wilc_set_mac_address(vif, ndev->dev_addr); + ether_addr_copy(addr, ndev->dev_addr); + wilc_set_mac_address(vif, addr); } else { - u8 addr[ETH_ALEN]; - wilc_get_mac_address(vif, addr); eth_hw_addr_set(ndev, addr); } From a608e6794b080594590c52d2e373068d5148954c Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 06:27:20 -0800 Subject: [PATCH 0352/2048] ipw2200: constify address in ipw_send_adapter_address Add const to the address param of ipw_send_adapter_address() all the functions down the chain have already been changed. Not sure how I lost this in the rebase. Acked-by: Kalle Valo Signed-off-by: Jakub Kicinski Acked-by: Stanislav Yakovlev Signed-off-by: David S. Miller --- drivers/net/wireless/intel/ipw2x00/ipw2200.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/intel/ipw2x00/ipw2200.c b/drivers/net/wireless/intel/ipw2x00/ipw2200.c index 23037bfc9e4c..5727c7c00a28 100644 --- a/drivers/net/wireless/intel/ipw2x00/ipw2200.c +++ b/drivers/net/wireless/intel/ipw2x00/ipw2200.c @@ -2303,7 +2303,7 @@ static int ipw_send_ssid(struct ipw_priv *priv, u8 * ssid, int len) ssid); } -static int ipw_send_adapter_address(struct ipw_priv *priv, u8 * mac) +static int ipw_send_adapter_address(struct ipw_priv *priv, const u8 * mac) { if (!priv || !mac) { IPW_ERROR("Invalid args\n"); From 0222ee53c483df7572eea7ba8585dda59328d46e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:19 -0800 Subject: [PATCH 0353/2048] amd: lance: use eth_hw_addr_set() IO reads, so save to an array then eth_hw_addr_set(). Fixes build on x86 (32bit). Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/lance.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/amd/lance.c b/drivers/net/ethernet/amd/lance.c index 945bf1d87507..462016666752 100644 --- a/drivers/net/ethernet/amd/lance.c +++ b/drivers/net/ethernet/amd/lance.c @@ -480,6 +480,7 @@ static int __init lance_probe1(struct net_device *dev, int ioaddr, int irq, int unsigned long flags; int err = -ENOMEM; void __iomem *bios; + u8 addr[ETH_ALEN]; /* First we look for special cases. Check for HP's on-board ethernet by looking for 'HP' in the BIOS. @@ -541,7 +542,8 @@ static int __init lance_probe1(struct net_device *dev, int ioaddr, int irq, int /* There is a 16 byte station address PROM at the base address. The first six bytes are the station address. */ for (i = 0; i < 6; i++) - dev->dev_addr[i] = inb(ioaddr + i); + addr[i] = inb(ioaddr + i); + eth_hw_addr_set(dev, addr); printk("%pM", dev->dev_addr); dev->base_addr = ioaddr; From 69ede3097b871dbc793dd6e21f38fde56d273963 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:20 -0800 Subject: [PATCH 0354/2048] amd: ni65: use eth_hw_addr_set() IO reads, so save to an array then eth_hw_addr_set(). Fixes build on x86 (32bit). Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/ni65.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/amd/ni65.c b/drivers/net/ethernet/amd/ni65.c index 032e8922b482..8ba579b89b75 100644 --- a/drivers/net/ethernet/amd/ni65.c +++ b/drivers/net/ethernet/amd/ni65.c @@ -251,7 +251,7 @@ static void ni65_recv_intr(struct net_device *dev,int); static void ni65_xmit_intr(struct net_device *dev,int); static int ni65_open(struct net_device *dev); static int ni65_lance_reinit(struct net_device *dev); -static void ni65_init_lance(struct priv *p,unsigned char*,int,int); +static void ni65_init_lance(struct priv *p,const unsigned char*,int,int); static netdev_tx_t ni65_send_packet(struct sk_buff *skb, struct net_device *dev); static void ni65_timeout(struct net_device *dev, unsigned int txqueue); @@ -418,6 +418,7 @@ static int __init ni65_probe1(struct net_device *dev,int ioaddr) { int i,j; struct priv *p; + u8 addr[ETH_ALEN]; unsigned long flags; dev->irq = irq; @@ -444,7 +445,8 @@ static int __init ni65_probe1(struct net_device *dev,int ioaddr) return -ENODEV; for(j=0;j<6;j++) - dev->dev_addr[j] = inb(ioaddr+cards[i].addr_offset+j); + addr[j] = inb(ioaddr+cards[i].addr_offset+j); + eth_hw_addr_set(dev, addr); if( (j=ni65_alloc_buffer(dev)) < 0) { release_region(ioaddr, cards[i].total_size); @@ -566,7 +568,7 @@ static int __init ni65_probe1(struct net_device *dev,int ioaddr) /* * set lance register and trigger init */ -static void ni65_init_lance(struct priv *p,unsigned char *daddr,int filter,int mode) +static void ni65_init_lance(struct priv *p,const unsigned char *daddr,int filter,int mode) { int i; u32 pib; From 285e4c664d6461b175b4613fc77126b5006a1912 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:21 -0800 Subject: [PATCH 0355/2048] amd: a2065/ariadne: use eth_hw_addr_set() dev_addr is initialized byte by byte from series. Fixes build on x86 (32bit). Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/a2065.c | 18 ++++++++++-------- drivers/net/ethernet/amd/ariadne.c | 20 +++++++++++--------- 2 files changed, 21 insertions(+), 17 deletions(-) diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c index 2f808dbc8b0e..3a351d3396bf 100644 --- a/drivers/net/ethernet/amd/a2065.c +++ b/drivers/net/ethernet/amd/a2065.c @@ -680,6 +680,7 @@ static int a2065_init_one(struct zorro_dev *z, unsigned long base_addr = board + A2065_LANCE; unsigned long mem_start = board + A2065_RAM; struct resource *r1, *r2; + u8 addr[ETH_ALEN]; u32 serial; int err; @@ -706,17 +707,18 @@ static int a2065_init_one(struct zorro_dev *z, r2->name = dev->name; serial = be32_to_cpu(z->rom.er_SerialNumber); - dev->dev_addr[0] = 0x00; + addr[0] = 0x00; if (z->id != ZORRO_PROD_AMERISTAR_A2065) { /* Commodore */ - dev->dev_addr[1] = 0x80; - dev->dev_addr[2] = 0x10; + addr[1] = 0x80; + addr[2] = 0x10; } else { /* Ameristar */ - dev->dev_addr[1] = 0x00; - dev->dev_addr[2] = 0x9f; + addr[1] = 0x00; + addr[2] = 0x9f; } - dev->dev_addr[3] = (serial >> 16) & 0xff; - dev->dev_addr[4] = (serial >> 8) & 0xff; - dev->dev_addr[5] = serial & 0xff; + addr[3] = (serial >> 16) & 0xff; + addr[4] = (serial >> 8) & 0xff; + addr[5] = serial & 0xff; + eth_hw_addr_set(dev, addr); dev->base_addr = (unsigned long)ZTWO_VADDR(base_addr); dev->mem_start = (unsigned long)ZTWO_VADDR(mem_start); dev->mem_end = dev->mem_start + A2065_RAM_SIZE; diff --git a/drivers/net/ethernet/amd/ariadne.c b/drivers/net/ethernet/amd/ariadne.c index 5e0f645f5bde..4ea7b9f3c424 100644 --- a/drivers/net/ethernet/amd/ariadne.c +++ b/drivers/net/ethernet/amd/ariadne.c @@ -441,11 +441,11 @@ static int ariadne_open(struct net_device *dev) /* Set the Ethernet Hardware Address */ lance->RAP = CSR12; /* Physical Address Register, PADR[15:0] */ - lance->RDP = ((u_short *)&dev->dev_addr[0])[0]; + lance->RDP = ((const u_short *)&dev->dev_addr[0])[0]; lance->RAP = CSR13; /* Physical Address Register, PADR[31:16] */ - lance->RDP = ((u_short *)&dev->dev_addr[0])[1]; + lance->RDP = ((const u_short *)&dev->dev_addr[0])[1]; lance->RAP = CSR14; /* Physical Address Register, PADR[47:32] */ - lance->RDP = ((u_short *)&dev->dev_addr[0])[2]; + lance->RDP = ((const u_short *)&dev->dev_addr[0])[2]; /* Set the Init Block Mode */ lance->RAP = CSR15; /* Mode Register */ @@ -717,6 +717,7 @@ static int ariadne_init_one(struct zorro_dev *z, unsigned long mem_start = board + ARIADNE_RAM; struct resource *r1, *r2; struct net_device *dev; + u8 addr[ETH_ALEN]; u32 serial; int err; @@ -740,12 +741,13 @@ static int ariadne_init_one(struct zorro_dev *z, r2->name = dev->name; serial = be32_to_cpu(z->rom.er_SerialNumber); - dev->dev_addr[0] = 0x00; - dev->dev_addr[1] = 0x60; - dev->dev_addr[2] = 0x30; - dev->dev_addr[3] = (serial >> 16) & 0xff; - dev->dev_addr[4] = (serial >> 8) & 0xff; - dev->dev_addr[5] = serial & 0xff; + addr[0] = 0x00; + addr[1] = 0x60; + addr[2] = 0x30; + addr[3] = (serial >> 16) & 0xff; + addr[4] = (serial >> 8) & 0xff; + addr[5] = serial & 0xff; + eth_hw_addr_set(dev, addr); dev->base_addr = (unsigned long)ZTWO_VADDR(base_addr); dev->mem_start = (unsigned long)ZTWO_VADDR(mem_start); dev->mem_end = dev->mem_start + ARIADNE_RAM_SIZE; From 21942eef062781429b356974589d7965952940fb Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:22 -0800 Subject: [PATCH 0356/2048] amd: hplance: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/hplance.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/amd/hplance.c b/drivers/net/ethernet/amd/hplance.c index 6784f8748638..055fda11c572 100644 --- a/drivers/net/ethernet/amd/hplance.c +++ b/drivers/net/ethernet/amd/hplance.c @@ -129,6 +129,7 @@ static void hplance_init(struct net_device *dev, struct dio_dev *d) { unsigned long va = (d->resource.start + DIO_VIRADDRBASE); struct hplance_private *lp; + u8 addr[ETH_ALEN]; int i; /* reset the board */ @@ -144,9 +145,10 @@ static void hplance_init(struct net_device *dev, struct dio_dev *d) /* The NVRAM holds our ethernet address, one nibble per byte, * at bytes NVRAMOFF+1,3,5,7,9... */ - dev->dev_addr[i] = ((in_8(va + HPLANCE_NVRAMOFF + i*4 + 1) & 0xF) << 4) + addr[i] = ((in_8(va + HPLANCE_NVRAMOFF + i*4 + 1) & 0xF) << 4) | (in_8(va + HPLANCE_NVRAMOFF + i*4 + 3) & 0xF); } + eth_hw_addr_set(dev, addr); lp = netdev_priv(dev); lp->lance.name = d->name; From c3dc2f7196ca0f59d9baeb5d3b927e703944dc6c Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:23 -0800 Subject: [PATCH 0357/2048] amd: atarilance: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/atarilance.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/amd/atarilance.c b/drivers/net/ethernet/amd/atarilance.c index 9c7d9690d00c..27869164c6e6 100644 --- a/drivers/net/ethernet/amd/atarilance.c +++ b/drivers/net/ethernet/amd/atarilance.c @@ -471,6 +471,7 @@ static unsigned long __init lance_probe1( struct net_device *dev, int i; static int did_version; unsigned short save1, save2; + u8 addr[ETH_ALEN]; PROBE_PRINT(( "Probing for Lance card at mem %#lx io %#lx\n", (long)memaddr, (long)ioaddr )); @@ -585,14 +586,16 @@ static unsigned long __init lance_probe1( struct net_device *dev, eth_hw_addr_set(dev, OldRieblDefHwaddr); break; case NEW_RIEBL: - lp->memcpy_f(dev->dev_addr, RIEBL_HWADDR_ADDR, ETH_ALEN); + lp->memcpy_f(addr, RIEBL_HWADDR_ADDR, ETH_ALEN); + eth_hw_addr_set(dev, addr); break; case PAM_CARD: i = IO->eeprom; for( i = 0; i < 6; ++i ) - dev->dev_addr[i] = + addr[i] = ((((unsigned short *)MEM)[i*2] & 0x0f) << 4) | ((((unsigned short *)MEM)[i*2+1] & 0x0f)); + eth_hw_addr_set(dev, addr); i = IO->mem; break; } From cc71b8b9376ff5072d23b191654408c144dac6aa Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:24 -0800 Subject: [PATCH 0358/2048] amd: mvme147: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/mvme147.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/amd/mvme147.c b/drivers/net/ethernet/amd/mvme147.c index da97fccea9ea..410c7b67eba4 100644 --- a/drivers/net/ethernet/amd/mvme147.c +++ b/drivers/net/ethernet/amd/mvme147.c @@ -74,6 +74,7 @@ static struct net_device * __init mvme147lance_probe(void) static int called; static const char name[] = "MVME147 LANCE"; struct m147lance_private *lp; + u8 macaddr[ETH_ALEN]; u_long *addr; u_long address; int err; @@ -93,15 +94,16 @@ static struct net_device * __init mvme147lance_probe(void) addr = (u_long *)ETHERNET_ADDRESS; address = *addr; - dev->dev_addr[0] = 0x08; - dev->dev_addr[1] = 0x00; - dev->dev_addr[2] = 0x3e; + macaddr[0] = 0x08; + macaddr[1] = 0x00; + macaddr[2] = 0x3e; address = address >> 8; - dev->dev_addr[5] = address&0xff; + macaddr[5] = address&0xff; address = address >> 8; - dev->dev_addr[4] = address&0xff; + macaddr[4] = address&0xff; address = address >> 8; - dev->dev_addr[3] = address&0xff; + macaddr[3] = address&0xff; + eth_hw_addr_set(dev, macaddr); printk("%s: MVME147 at 0x%08lx, irq %d, Hardware Address %pM\n", dev->name, dev->base_addr, MVME147_LANCE_IRQ, From 5114ddf8dd881f9059147d3f130e9415ce94125e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:25 -0800 Subject: [PATCH 0359/2048] 8390: smc-ultra: use eth_hw_addr_set() IO reads, so save to an array then eth_hw_addr_set(). Fixes build on Alpha. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/8390/smc-ultra.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/8390/smc-ultra.c b/drivers/net/ethernet/8390/smc-ultra.c index 0890fa493f70..6e62c37c9400 100644 --- a/drivers/net/ethernet/8390/smc-ultra.c +++ b/drivers/net/ethernet/8390/smc-ultra.c @@ -204,6 +204,7 @@ static int __init ultra_probe1(struct net_device *dev, int ioaddr) { int i, retval; int checksum = 0; + u8 macaddr[ETH_ALEN]; const char *model_name; unsigned char eeprom_irq = 0; static unsigned version_printed; @@ -239,7 +240,8 @@ static int __init ultra_probe1(struct net_device *dev, int ioaddr) model_name = (idreg & 0xF0) == 0x20 ? "SMC Ultra" : "SMC EtherEZ"; for (i = 0; i < 6; i++) - dev->dev_addr[i] = inb(ioaddr + 8 + i); + macaddr[i] = inb(ioaddr + 8 + i); + eth_hw_addr_set(dev, macaddr); netdev_info(dev, "%s at %#3x, %pM", model_name, ioaddr, dev->dev_addr); From d7d28e90e229a8af0472421015c5828f5cd1ad2e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:26 -0800 Subject: [PATCH 0360/2048] 8390: hydra: use eth_hw_addr_set() Loop with offsetting to every second byte, so use a temp buffer. Fixes m68k build. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/8390/hydra.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/8390/hydra.c b/drivers/net/ethernet/8390/hydra.c index 941754ea78ec..1df7601af86a 100644 --- a/drivers/net/ethernet/8390/hydra.c +++ b/drivers/net/ethernet/8390/hydra.c @@ -116,6 +116,7 @@ static int hydra_init(struct zorro_dev *z) unsigned long ioaddr = board+HYDRA_NIC_BASE; const char name[] = "NE2000"; int start_page, stop_page; + u8 macaddr[ETH_ALEN]; int j; int err; @@ -129,7 +130,8 @@ static int hydra_init(struct zorro_dev *z) return -ENOMEM; for (j = 0; j < ETH_ALEN; j++) - dev->dev_addr[j] = *((u8 *)(board + HYDRA_ADDRPROM + 2*j)); + macaddr[j] = *((u8 *)(board + HYDRA_ADDRPROM + 2*j)); + eth_hw_addr_set(dev, macaddr); /* We must set the 8390 for word mode. */ z_writeb(0x4b, ioaddr + NE_EN0_DCFG); From 973a34c087f43b61b26570110a284faf48d08d5d Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:27 -0800 Subject: [PATCH 0361/2048] 8390: mac8390: use eth_hw_addr_set() Use temp to pass to the reading function, the function is generic so can't fix there. Fixes m68k build. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/8390/mac8390.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/8390/mac8390.c b/drivers/net/ethernet/8390/mac8390.c index 91b04abfd687..7fb819b9b89a 100644 --- a/drivers/net/ethernet/8390/mac8390.c +++ b/drivers/net/ethernet/8390/mac8390.c @@ -292,6 +292,7 @@ static bool mac8390_rsrc_init(struct net_device *dev, struct nubus_dirent ent; int offset; volatile unsigned short *i; + u8 addr[ETH_ALEN]; dev->irq = SLOT2IRQ(board->slot); /* This is getting to be a habit */ @@ -314,7 +315,8 @@ static bool mac8390_rsrc_init(struct net_device *dev, return false; } - nubus_get_rsrc_mem(dev->dev_addr, &ent, 6); + nubus_get_rsrc_mem(addr, &ent, 6); + eth_hw_addr_set(dev, addr); if (useresources[cardtype] == 1) { nubus_rewinddir(&dir); From f95f8e890a2aa576425402fea44bfa657e8ccaa6 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:28 -0800 Subject: [PATCH 0362/2048] 8390: wd: use eth_hw_addr_set() IO reads, so save to an array then eth_hw_addr_set(). Fixes build on x86 (32bit). Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/8390/wd.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/8390/wd.c b/drivers/net/ethernet/8390/wd.c index 263a942d81fa..5b00c452bede 100644 --- a/drivers/net/ethernet/8390/wd.c +++ b/drivers/net/ethernet/8390/wd.c @@ -168,6 +168,7 @@ static int __init wd_probe1(struct net_device *dev, int ioaddr) int checksum = 0; int ancient = 0; /* An old card without config registers. */ int word16 = 0; /* 0 = 8 bit, 1 = 16 bit */ + u8 addr[ETH_ALEN]; const char *model_name; static unsigned version_printed; struct ei_device *ei_local = netdev_priv(dev); @@ -191,7 +192,8 @@ static int __init wd_probe1(struct net_device *dev, int ioaddr) netdev_info(dev, version); for (i = 0; i < 6; i++) - dev->dev_addr[i] = inb(ioaddr + 8 + i); + addr[i] = inb(ioaddr + 8 + i); + eth_hw_addr_set(dev, addr); netdev_info(dev, "WD80x3 at %#3x, %pM", ioaddr, dev->dev_addr); From 80db345e7df0c507a83bd12ac7766fb054443804 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:29 -0800 Subject: [PATCH 0363/2048] smc9194: use eth_hw_addr_set() dev_addr is set from IO reads, and broken from a u16 value. Fixes build on Alpha. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/smsc/smc9194.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/smsc/smc9194.c b/drivers/net/ethernet/smsc/smc9194.c index 0ce403fa5f1a..af661c65ffe2 100644 --- a/drivers/net/ethernet/smsc/smc9194.c +++ b/drivers/net/ethernet/smsc/smc9194.c @@ -856,6 +856,7 @@ static int __init smc_probe(struct net_device *dev, int ioaddr) word configuration_register; word memory_info_register; word memory_cfg_register; + u8 addr[ETH_ALEN]; /* Grab the region so that no one else tries to probe our ioports. */ if (!request_region(ioaddr, SMC_IO_EXTENT, DRV_NAME)) @@ -924,9 +925,10 @@ static int __init smc_probe(struct net_device *dev, int ioaddr) word address; address = inw( ioaddr + ADDR0 + i ); - dev->dev_addr[ i + 1] = address >> 8; - dev->dev_addr[ i ] = address & 0xFF; + addr[i + 1] = address >> 8; + addr[i] = address & 0xFF; } + eth_hw_addr_set(dev, addr); /* get the memory information */ From 5b6d5affd27486ec9152a66df1741cf67782712a Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:30 -0800 Subject: [PATCH 0364/2048] lasi_82594: use eth_hw_addr_set() dev_addr is set from IO reads, passed to an arch-specific helper. Note that the helper never reads it so uninitialized temp is fine. Fixes build on parisc. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/i825xx/lasi_82596.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/i825xx/lasi_82596.c b/drivers/net/ethernet/i825xx/lasi_82596.c index 48e001881c75..0af70094aba3 100644 --- a/drivers/net/ethernet/i825xx/lasi_82596.c +++ b/drivers/net/ethernet/i825xx/lasi_82596.c @@ -147,6 +147,7 @@ lan_init_chip(struct parisc_device *dev) struct net_device *netdevice; struct i596_private *lp; int retval = -ENOMEM; + u8 addr[ETH_ALEN]; int i; if (!dev->irq) { @@ -167,13 +168,14 @@ lan_init_chip(struct parisc_device *dev) netdevice->base_addr = dev->hpa.start; netdevice->irq = dev->irq; - if (pdc_lan_station_id(netdevice->dev_addr, netdevice->base_addr)) { + if (pdc_lan_station_id(addr, netdevice->base_addr)) { for (i = 0; i < 6; i++) { - netdevice->dev_addr[i] = gsc_readb(LAN_PROM_ADDR + i); + addr[i] = gsc_readb(LAN_PROM_ADDR + i); } printk(KERN_INFO "%s: MAC of HP700 LAN read from EEPROM\n", __FILE__); } + eth_hw_addr_set(netdevice, addr); lp = netdev_priv(netdevice); lp->options = dev->id.sversion == 0x72 ? OPT_SWAP_PORT : 0; From e217fc4affc8c7392e4db48488b36d2a5d446e9d Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:31 -0800 Subject: [PATCH 0365/2048] apple: macmace: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/apple/macmace.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/apple/macmace.c b/drivers/net/ethernet/apple/macmace.c index 95d3061c61be..8fcaf1639920 100644 --- a/drivers/net/ethernet/apple/macmace.c +++ b/drivers/net/ethernet/apple/macmace.c @@ -92,7 +92,7 @@ static void mace_reset(struct net_device *dev); static irqreturn_t mace_interrupt(int irq, void *dev_id); static irqreturn_t mace_dma_intr(int irq, void *dev_id); static void mace_tx_timeout(struct net_device *dev, unsigned int txqueue); -static void __mace_set_address(struct net_device *dev, void *addr); +static void __mace_set_address(struct net_device *dev, const void *addr); /* * Load a receive DMA channel with a base address and ring length @@ -197,6 +197,7 @@ static int mace_probe(struct platform_device *pdev) unsigned char *addr; struct net_device *dev; unsigned char checksum = 0; + u8 macaddr[ETH_ALEN]; int err; dev = alloc_etherdev(PRIV_BYTES); @@ -229,8 +230,9 @@ static int mace_probe(struct platform_device *pdev) for (j = 0; j < 6; ++j) { u8 v = bitrev8(addr[j<<4]); checksum ^= v; - dev->dev_addr[j] = v; + macaddr[j] = v; } + eth_hw_addr_set(dev, macaddr); for (; j < 8; ++j) { checksum ^= bitrev8(addr[j<<4]); } @@ -315,11 +317,12 @@ static void mace_reset(struct net_device *dev) * Load the address on a mace controller. */ -static void __mace_set_address(struct net_device *dev, void *addr) +static void __mace_set_address(struct net_device *dev, const void *addr) { struct mace_data *mp = netdev_priv(dev); volatile struct mace *mb = mp->mace; - unsigned char *p = addr; + const unsigned char *p = addr; + u8 macaddr[ETH_ALEN]; int i; /* load up the hardware address */ @@ -331,7 +334,8 @@ static void __mace_set_address(struct net_device *dev, void *addr) ; } for (i = 0; i < 6; ++i) - mb->padr = dev->dev_addr[i] = p[i]; + mb->padr = macaddr[i] = p[i]; + eth_hw_addr_set(dev, macaddr); if (mp->chipid != BROKEN_ADDRCHG_REV) mb->iac = 0; } From 9a962aedd30f7fceb828d3161a80e0526e358eb5 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:32 -0800 Subject: [PATCH 0366/2048] cirrus: mac89x0: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/cirrus/mac89x0.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cirrus/mac89x0.c b/drivers/net/ethernet/cirrus/mac89x0.c index 84251b85fc93..21a70b1f0ac5 100644 --- a/drivers/net/ethernet/cirrus/mac89x0.c +++ b/drivers/net/ethernet/cirrus/mac89x0.c @@ -242,12 +242,15 @@ static int mac89x0_device_probe(struct platform_device *pdev) pr_info("No EEPROM, giving up now.\n"); goto out1; } else { + u8 addr[ETH_ALEN]; + for (i = 0; i < ETH_ALEN; i += 2) { /* Big-endian (why??!) */ unsigned short s = readreg(dev, PP_IA + i); - dev->dev_addr[i] = s >> 8; - dev->dev_addr[i+1] = s & 0xff; + addr[i] = s >> 8; + addr[i+1] = s & 0xff; } + eth_hw_addr_set(dev, addr); } dev->irq = SLOT2IRQ(slot); From bb52aff3e32144ce7e59541302cbce58464a450e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 18 Nov 2021 23:10:33 -0800 Subject: [PATCH 0367/2048] natsemi: macsonic: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/natsemi/macsonic.c | 27 ++++++++++++++++--------- 1 file changed, 17 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/natsemi/macsonic.c b/drivers/net/ethernet/natsemi/macsonic.c index 8709d700e15a..b16f7c830f9b 100644 --- a/drivers/net/ethernet/natsemi/macsonic.c +++ b/drivers/net/ethernet/natsemi/macsonic.c @@ -203,6 +203,7 @@ static void mac_onboard_sonic_ethernet_addr(struct net_device *dev) struct sonic_local *lp = netdev_priv(dev); const int prom_addr = ONBOARD_SONIC_PROM_BASE; unsigned short val; + u8 addr[ETH_ALEN]; /* * On NuBus boards we can sometimes look in the ROM resources. @@ -213,7 +214,8 @@ static void mac_onboard_sonic_ethernet_addr(struct net_device *dev) int i; for (i = 0; i < 6; i++) - dev->dev_addr[i] = SONIC_READ_PROM(i); + addr[i] = SONIC_READ_PROM(i); + eth_hw_addr_set(dev, addr); if (!INVALID_MAC(dev->dev_addr)) return; @@ -222,7 +224,8 @@ static void mac_onboard_sonic_ethernet_addr(struct net_device *dev) * source has a rather long and detailed historical account of * why this is so. */ - bit_reverse_addr(dev->dev_addr); + bit_reverse_addr(addr); + eth_hw_addr_set(dev, addr); if (!INVALID_MAC(dev->dev_addr)) return; @@ -243,14 +246,15 @@ static void mac_onboard_sonic_ethernet_addr(struct net_device *dev) SONIC_WRITE(SONIC_CEP, 15); val = SONIC_READ(SONIC_CAP2); - dev->dev_addr[5] = val >> 8; - dev->dev_addr[4] = val & 0xff; + addr[5] = val >> 8; + addr[4] = val & 0xff; val = SONIC_READ(SONIC_CAP1); - dev->dev_addr[3] = val >> 8; - dev->dev_addr[2] = val & 0xff; + addr[3] = val >> 8; + addr[2] = val & 0xff; val = SONIC_READ(SONIC_CAP0); - dev->dev_addr[1] = val >> 8; - dev->dev_addr[0] = val & 0xff; + addr[1] = val >> 8; + addr[0] = val & 0xff; + eth_hw_addr_set(dev, addr); if (!INVALID_MAC(dev->dev_addr)) return; @@ -355,13 +359,16 @@ static int mac_onboard_sonic_probe(struct net_device *dev) static int mac_sonic_nubus_ethernet_addr(struct net_device *dev, unsigned long prom_addr, int id) { + u8 addr[ETH_ALEN]; int i; + for(i = 0; i < 6; i++) - dev->dev_addr[i] = SONIC_READ_PROM(i); + addr[i] = SONIC_READ_PROM(i); /* Some of the addresses are bit-reversed */ if (id != MACSONIC_DAYNA) - bit_reverse_addr(dev->dev_addr); + bit_reverse_addr(addr); + eth_hw_addr_set(dev, addr); return 0; } From 832585d2172fdaa29ee6497a94c0b9da0ab97eec Mon Sep 17 00:00:00 2001 From: Julian Wiedmann Date: Thu, 18 Nov 2021 17:06:02 +0100 Subject: [PATCH 0368/2048] s390/qeth: allocate RX queue at probe time We always need an RX queue, and there's no reconfig situation either where we would need to free & rebuild the queue. So allocate the RX queue right from the start, and avoid freeing it during unrelated qeth_free_qdio_queues() calls. Signed-off-by: Julian Wiedmann Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- drivers/s390/net/qeth_core_main.c | 35 +++++++++++++++---------------- 1 file changed, 17 insertions(+), 18 deletions(-) diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c index 26c55f67289f..d32aa8b705db 100644 --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -194,9 +194,6 @@ static void qeth_clear_working_pool_list(struct qeth_card *card) &card->qdio.in_buf_pool.entry_list, list) list_del(&pool_entry->list); - if (!queue) - return; - for (i = 0; i < ARRAY_SIZE(queue->bufs); i++) queue->bufs[i].pool_entry = NULL; } @@ -275,8 +272,8 @@ int qeth_resize_buffer_pool(struct qeth_card *card, unsigned int count) QETH_CARD_TEXT(card, 2, "realcbp"); - /* Defer until queue is allocated: */ - if (!card->qdio.in_q) + /* Defer until pool is allocated: */ + if (list_empty(&pool->entry_list)) goto out; /* Remove entries from the pool: */ @@ -2557,14 +2554,9 @@ static int qeth_alloc_qdio_queues(struct qeth_card *card) QETH_QDIO_ALLOCATED) != QETH_QDIO_UNINITIALIZED) return 0; - QETH_CARD_TEXT(card, 2, "inq"); - card->qdio.in_q = qeth_alloc_qdio_queue(); - if (!card->qdio.in_q) - goto out_nomem; - /* inbound buffer pool */ if (qeth_alloc_buffer_pool(card)) - goto out_freeinq; + goto out_buffer_pool; /* outbound */ for (i = 0; i < card->qdio.no_out_queues; ++i) { @@ -2605,10 +2597,7 @@ out_freeoutq: card->qdio.out_qs[i] = NULL; } qeth_free_buffer_pool(card); -out_freeinq: - qeth_free_qdio_queue(card->qdio.in_q); - card->qdio.in_q = NULL; -out_nomem: +out_buffer_pool: atomic_set(&card->qdio.state, QETH_QDIO_UNINITIALIZED); return -ENOMEM; } @@ -2623,11 +2612,12 @@ static void qeth_free_qdio_queues(struct qeth_card *card) qeth_free_cq(card); for (j = 0; j < QDIO_MAX_BUFFERS_PER_Q; ++j) { - if (card->qdio.in_q->bufs[j].rx_skb) + if (card->qdio.in_q->bufs[j].rx_skb) { consume_skb(card->qdio.in_q->bufs[j].rx_skb); + card->qdio.in_q->bufs[j].rx_skb = NULL; + } } - qeth_free_qdio_queue(card->qdio.in_q); - card->qdio.in_q = NULL; + /* inbound buffer pool */ qeth_free_buffer_pool(card); /* free outbound qdio_qs */ @@ -6447,6 +6437,12 @@ static int qeth_core_probe_device(struct ccwgroup_device *gdev) qeth_determine_capabilities(card); qeth_set_blkt_defaults(card); + card->qdio.in_q = qeth_alloc_qdio_queue(); + if (!card->qdio.in_q) { + rc = -ENOMEM; + goto err_rx_queue; + } + card->qdio.no_out_queues = card->dev->num_tx_queues; rc = qeth_update_from_chp_desc(card); if (rc) @@ -6473,6 +6469,8 @@ static int qeth_core_probe_device(struct ccwgroup_device *gdev) err_setup_disc: err_chp_desc: + qeth_free_qdio_queue(card->qdio.in_q); +err_rx_queue: free_netdev(card->dev); err_card: qeth_core_free_card(card); @@ -6494,6 +6492,7 @@ static void qeth_core_remove_device(struct ccwgroup_device *gdev) qeth_free_qdio_queues(card); + qeth_free_qdio_queue(card->qdio.in_q); free_netdev(card->dev); qeth_core_free_card(card); put_device(&gdev->dev); From 682026a5e9344fc6d0e779c1d49edad91cc2fbdb Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 18 Nov 2021 17:06:03 +0100 Subject: [PATCH 0369/2048] net/iucv: fix kernel doc comments Fix kernel doc comments where appropriate or remove incorrect kernel doc indicators. Also move kernel doc comments directly before functions. Signed-off-by: Heiko Carstens Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- net/iucv/iucv.c | 124 ++++++++++++++++++++++++------------------------ 1 file changed, 62 insertions(+), 62 deletions(-) diff --git a/net/iucv/iucv.c b/net/iucv/iucv.c index f3343a8541a5..8f4d49a7d3e8 100644 --- a/net/iucv/iucv.c +++ b/net/iucv/iucv.c @@ -276,8 +276,8 @@ static union iucv_param *iucv_param[NR_CPUS]; static union iucv_param *iucv_param_irq[NR_CPUS]; /** - * iucv_call_b2f0 - * @code: identifier of IUCV call to CP. + * __iucv_call_b2f0 + * @command: identifier of IUCV call to CP. * @parm: pointer to a struct iucv_parm block * * Calls CP to execute IUCV commands. @@ -309,7 +309,7 @@ static inline int iucv_call_b2f0(int command, union iucv_param *parm) return ccode == 1 ? parm->ctrl.iprcode : ccode; } -/** +/* * iucv_query_maxconn * * Determines the maximum number of connections that may be established. @@ -493,8 +493,8 @@ static void iucv_retrieve_cpu(void *data) cpumask_clear_cpu(cpu, &iucv_buffer_cpumask); } -/** - * iucv_setmask_smp +/* + * iucv_setmask_mp * * Allow iucv interrupts on all cpus. */ @@ -512,7 +512,7 @@ static void iucv_setmask_mp(void) cpus_read_unlock(); } -/** +/* * iucv_setmask_up * * Allow iucv interrupts on a single cpu. @@ -529,7 +529,7 @@ static void iucv_setmask_up(void) smp_call_function_single(cpu, iucv_block_cpu, NULL, 1); } -/** +/* * iucv_enable * * This function makes iucv ready for use. It allocates the pathid @@ -564,7 +564,7 @@ out: return rc; } -/** +/* * iucv_disable * * This function shuts down iucv. It disables iucv interrupts, retrieves @@ -1347,8 +1347,9 @@ EXPORT_SYMBOL(iucv_message_send); * @srccls: source class of message * @buffer: address of send buffer or address of struct iucv_array * @size: length of send buffer - * @ansbuf: address of answer buffer or address of struct iucv_array + * @answer: address of answer buffer or address of struct iucv_array * @asize: size of reply buffer + * @residual: ignored * * This function transmits data to another application. Data to be * transmitted is in a buffer. The receiver of the send is expected to @@ -1400,13 +1401,6 @@ out: } EXPORT_SYMBOL(iucv_message_send2way); -/** - * iucv_path_pending - * @data: Pointer to external interrupt buffer - * - * Process connection pending work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_path_pending { u16 ippathid; u8 ipflags1; @@ -1420,6 +1414,13 @@ struct iucv_path_pending { u8 res4[3]; } __packed; +/** + * iucv_path_pending + * @data: Pointer to external interrupt buffer + * + * Process connection pending work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_path_pending(struct iucv_irq_data *data) { struct iucv_path_pending *ipp = (void *) data; @@ -1461,13 +1462,6 @@ out_sever: iucv_sever_pathid(ipp->ippathid, error); } -/** - * iucv_path_complete - * @data: Pointer to external interrupt buffer - * - * Process connection complete work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_path_complete { u16 ippathid; u8 ipflags1; @@ -1481,6 +1475,13 @@ struct iucv_path_complete { u8 res4[3]; } __packed; +/** + * iucv_path_complete + * @data: Pointer to external interrupt buffer + * + * Process connection complete work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_path_complete(struct iucv_irq_data *data) { struct iucv_path_complete *ipc = (void *) data; @@ -1492,13 +1493,6 @@ static void iucv_path_complete(struct iucv_irq_data *data) path->handler->path_complete(path, ipc->ipuser); } -/** - * iucv_path_severed - * @data: Pointer to external interrupt buffer - * - * Process connection severed work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_path_severed { u16 ippathid; u8 res1; @@ -1511,6 +1505,13 @@ struct iucv_path_severed { u8 res5[3]; } __packed; +/** + * iucv_path_severed + * @data: Pointer to external interrupt buffer + * + * Process connection severed work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_path_severed(struct iucv_irq_data *data) { struct iucv_path_severed *ips = (void *) data; @@ -1528,13 +1529,6 @@ static void iucv_path_severed(struct iucv_irq_data *data) } } -/** - * iucv_path_quiesced - * @data: Pointer to external interrupt buffer - * - * Process connection quiesced work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_path_quiesced { u16 ippathid; u8 res1; @@ -1547,6 +1541,13 @@ struct iucv_path_quiesced { u8 res5[3]; } __packed; +/** + * iucv_path_quiesced + * @data: Pointer to external interrupt buffer + * + * Process connection quiesced work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_path_quiesced(struct iucv_irq_data *data) { struct iucv_path_quiesced *ipq = (void *) data; @@ -1556,13 +1557,6 @@ static void iucv_path_quiesced(struct iucv_irq_data *data) path->handler->path_quiesced(path, ipq->ipuser); } -/** - * iucv_path_resumed - * @data: Pointer to external interrupt buffer - * - * Process connection resumed work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_path_resumed { u16 ippathid; u8 res1; @@ -1575,6 +1569,13 @@ struct iucv_path_resumed { u8 res5[3]; } __packed; +/** + * iucv_path_resumed + * @data: Pointer to external interrupt buffer + * + * Process connection resumed work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_path_resumed(struct iucv_irq_data *data) { struct iucv_path_resumed *ipr = (void *) data; @@ -1584,13 +1585,6 @@ static void iucv_path_resumed(struct iucv_irq_data *data) path->handler->path_resumed(path, ipr->ipuser); } -/** - * iucv_message_complete - * @data: Pointer to external interrupt buffer - * - * Process message complete work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_message_complete { u16 ippathid; u8 ipflags1; @@ -1606,6 +1600,13 @@ struct iucv_message_complete { u8 res2[3]; } __packed; +/** + * iucv_message_complete + * @data: Pointer to external interrupt buffer + * + * Process message complete work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_message_complete(struct iucv_irq_data *data) { struct iucv_message_complete *imc = (void *) data; @@ -1624,13 +1625,6 @@ static void iucv_message_complete(struct iucv_irq_data *data) } } -/** - * iucv_message_pending - * @data: Pointer to external interrupt buffer - * - * Process message pending work item. Called from tasklet while holding - * iucv_table_lock. - */ struct iucv_message_pending { u16 ippathid; u8 ipflags1; @@ -1653,6 +1647,13 @@ struct iucv_message_pending { u8 res2[3]; } __packed; +/** + * iucv_message_pending + * @data: Pointer to external interrupt buffer + * + * Process message pending work item. Called from tasklet while holding + * iucv_table_lock. + */ static void iucv_message_pending(struct iucv_irq_data *data) { struct iucv_message_pending *imp = (void *) data; @@ -1673,7 +1674,7 @@ static void iucv_message_pending(struct iucv_irq_data *data) } } -/** +/* * iucv_tasklet_fn: * * This tasklet loops over the queue of irq buffers created by @@ -1717,7 +1718,7 @@ static void iucv_tasklet_fn(unsigned long ignored) spin_unlock(&iucv_table_lock); } -/** +/* * iucv_work_fn: * * This work function loops over the queue of path pending irq blocks @@ -1748,9 +1749,8 @@ static void iucv_work_fn(struct work_struct *work) spin_unlock_bh(&iucv_table_lock); } -/** +/* * iucv_external_interrupt - * @code: irq code * * Handles external interrupts coming in from CP. * Places the interrupt buffer on a queue and schedules iucv_tasklet_fn(). From 7c8e1a9155ef7db11672107ee99ad55a0815e3a6 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 18 Nov 2021 17:06:04 +0100 Subject: [PATCH 0370/2048] net/af_iucv: fix kernel doc comments Fix kernel doc comments where appropriate, or remove incorrect kernel doc indicators. Acked-by: Julian Wiedmann Signed-off-by: Heiko Carstens Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- net/iucv/af_iucv.c | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index 18316ee3c692..996ccf3665e3 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -142,7 +142,7 @@ static inline size_t iucv_msg_length(struct iucv_message *msg) * iucv_sock_in_state() - check for specific states * @sk: sock structure * @state: first iucv sk state - * @state: second iucv sk state + * @state2: second iucv sk state * * Returns true if the socket in either in the first or second state. */ @@ -172,7 +172,7 @@ static inline int iucv_below_msglim(struct sock *sk) (atomic_read(&iucv->pendings) <= 0)); } -/** +/* * iucv_sock_wake_msglim() - Wake up thread waiting on msg limit */ static void iucv_sock_wake_msglim(struct sock *sk) @@ -187,7 +187,7 @@ static void iucv_sock_wake_msglim(struct sock *sk) rcu_read_unlock(); } -/** +/* * afiucv_hs_send() - send a message through HiperSockets transport */ static int afiucv_hs_send(struct iucv_message *imsg, struct sock *sock, @@ -1831,9 +1831,9 @@ static void afiucv_swap_src_dest(struct sk_buff *skb) memset(skb->data, 0, ETH_HLEN); } -/** +/* * afiucv_hs_callback_syn - react on received SYN - **/ + */ static int afiucv_hs_callback_syn(struct sock *sk, struct sk_buff *skb) { struct af_iucv_trans_hdr *trans_hdr = iucv_trans_hdr(skb); @@ -1896,9 +1896,9 @@ out: return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_callback_synack() - react on received SYN-ACK - **/ + */ static int afiucv_hs_callback_synack(struct sock *sk, struct sk_buff *skb) { struct iucv_sock *iucv = iucv_sk(sk); @@ -1917,9 +1917,9 @@ static int afiucv_hs_callback_synack(struct sock *sk, struct sk_buff *skb) return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_callback_synfin() - react on received SYN_FIN - **/ + */ static int afiucv_hs_callback_synfin(struct sock *sk, struct sk_buff *skb) { struct iucv_sock *iucv = iucv_sk(sk); @@ -1937,9 +1937,9 @@ static int afiucv_hs_callback_synfin(struct sock *sk, struct sk_buff *skb) return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_callback_fin() - react on received FIN - **/ + */ static int afiucv_hs_callback_fin(struct sock *sk, struct sk_buff *skb) { struct iucv_sock *iucv = iucv_sk(sk); @@ -1960,9 +1960,9 @@ static int afiucv_hs_callback_fin(struct sock *sk, struct sk_buff *skb) return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_callback_win() - react on received WIN - **/ + */ static int afiucv_hs_callback_win(struct sock *sk, struct sk_buff *skb) { struct iucv_sock *iucv = iucv_sk(sk); @@ -1978,9 +1978,9 @@ static int afiucv_hs_callback_win(struct sock *sk, struct sk_buff *skb) return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_callback_rx() - react on received data - **/ + */ static int afiucv_hs_callback_rx(struct sock *sk, struct sk_buff *skb) { struct iucv_sock *iucv = iucv_sk(sk); @@ -2022,11 +2022,11 @@ static int afiucv_hs_callback_rx(struct sock *sk, struct sk_buff *skb) return NET_RX_SUCCESS; } -/** +/* * afiucv_hs_rcv() - base function for arriving data through HiperSockets * transport * called from netif RX softirq - **/ + */ static int afiucv_hs_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) { @@ -2128,10 +2128,10 @@ static int afiucv_hs_rcv(struct sk_buff *skb, struct net_device *dev, return err; } -/** +/* * afiucv_hs_callback_txnotify() - handle send notifications from HiperSockets * transport - **/ + */ static void afiucv_hs_callback_txnotify(struct sock *sk, enum iucv_tx_notify n) { struct iucv_sock *iucv = iucv_sk(sk); From 9961d6d50b7f53ab75049ebf3bea1ff560014bed Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 18 Nov 2021 17:06:05 +0100 Subject: [PATCH 0371/2048] s390/ctcm: fix format string The second parameter as specified by the format string is actually a string not an integer. Acked-by: Julian Wiedmann Signed-off-by: Heiko Carstens Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- drivers/s390/net/ctcm_fsms.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/s390/net/ctcm_fsms.c b/drivers/s390/net/ctcm_fsms.c index de2423c72b02..5db591cf7215 100644 --- a/drivers/s390/net/ctcm_fsms.c +++ b/drivers/s390/net/ctcm_fsms.c @@ -1406,7 +1406,7 @@ static void ctcmpc_chx_rx(fsm_instance *fi, int event, void *arg) if (new_skb == NULL) { CTCM_DBF_TEXT_(MPC_ERROR, CTC_DBF_ERROR, - "%s(%d): skb allocation failed", + "%s(%s): skb allocation failed", CTCM_FUNTAIL, dev->name); fsm_event(priv->mpcg->fsm, MPCG_EVENT_INOP, dev); goto again; From dddbf91387a04a03a8e071cb4857403df219382e Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 18 Nov 2021 17:06:06 +0100 Subject: [PATCH 0372/2048] s390/ctcm: add __printf format attribute to ctcm_dbf_longtext MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Allow the compiler to recognize and check format strings and parameters. As reported with allmodconfig and W=1: drivers/s390/net/ctcm_dbug.c: In function ‘ctcm_dbf_longtext’: drivers/s390/net/ctcm_dbug.c:73:9: error: function ‘ctcm_dbf_longtext’ might be a candidate for ‘gnu_printf’ format attribute [-Werror=suggest-attribute=format] 73 | vsnprintf(dbf_txt_buf, sizeof(dbf_txt_buf), fmt, args); | ^~~~~~~~~ Acked-by: Julian Wiedmann Signed-off-by: Heiko Carstens Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- drivers/s390/net/ctcm_dbug.h | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/s390/net/ctcm_dbug.h b/drivers/s390/net/ctcm_dbug.h index 675575ef162e..cce11daf3245 100644 --- a/drivers/s390/net/ctcm_dbug.h +++ b/drivers/s390/net/ctcm_dbug.h @@ -65,6 +65,7 @@ extern struct ctcm_dbf_info ctcm_dbf[CTCM_DBF_INFOS]; int ctcm_register_dbf_views(void); void ctcm_unregister_dbf_views(void); +__printf(3, 4) void ctcm_dbf_longtext(enum ctcm_dbf_names dbf_nix, int level, char *text, ...); static inline const char *strtail(const char *s, int n) From 09ae598271f8cfe46148cece8e03509b9da16334 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 18 Nov 2021 17:06:07 +0100 Subject: [PATCH 0373/2048] s390/lcs: add braces around empty function body MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix allmodconfig + W=1 compile breakage: drivers/s390/net/lcs.c: In function ‘lcs_get_frames_cb’: drivers/s390/net/lcs.c:1823:25: error: suggest braces around empty body in an ‘else’ statement [-Werror=empty-body] 1823 | ; // FIXME: error message ? | ^ Acked-by: Julian Wiedmann Signed-off-by: Heiko Carstens Signed-off-by: Karsten Graul Signed-off-by: David S. Miller --- drivers/s390/net/lcs.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/s390/net/lcs.c b/drivers/s390/net/lcs.c index 2a6479740600..a61d38a1b4ed 100644 --- a/drivers/s390/net/lcs.c +++ b/drivers/s390/net/lcs.c @@ -1808,19 +1808,20 @@ lcs_get_frames_cb(struct lcs_channel *channel, struct lcs_buffer *buffer) return; } /* What kind of frame is it? */ - if (lcs_hdr->type == LCS_FRAME_TYPE_CONTROL) + if (lcs_hdr->type == LCS_FRAME_TYPE_CONTROL) { /* Control frame. */ lcs_get_control(card, (struct lcs_cmd *) lcs_hdr); - else if (lcs_hdr->type == LCS_FRAME_TYPE_ENET || - lcs_hdr->type == LCS_FRAME_TYPE_TR || - lcs_hdr->type == LCS_FRAME_TYPE_FDDI) + } else if (lcs_hdr->type == LCS_FRAME_TYPE_ENET || + lcs_hdr->type == LCS_FRAME_TYPE_TR || + lcs_hdr->type == LCS_FRAME_TYPE_FDDI) { /* Normal network packet. */ lcs_get_skb(card, (char *)(lcs_hdr + 1), lcs_hdr->offset - offset - sizeof(struct lcs_header)); - else + } else { /* Unknown frame type. */ ; // FIXME: error message ? + } /* Proceed to next frame. */ offset = lcs_hdr->offset; lcs_hdr->offset = LCS_ILLEGAL_OFFSET; From ec574d9ee5d2e74c913cb3e927e7706be7bfb64d Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Thu, 18 Nov 2021 18:07:06 +0000 Subject: [PATCH 0374/2048] net: phylink: add 1000base-KX to phylink_caps_to_linkmodes() 1000base-KX was missed in phylink_caps_to_linkmodes(), add it. This will be necessary to convert stmmac with xpcs to ensure we don't drop any supported linkmodes. Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/phy/phylink.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index f7156b6868e7..da17b874a5e7 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -197,6 +197,7 @@ static void phylink_caps_to_linkmodes(unsigned long *linkmodes, if (caps & MAC_1000FD) { __set_bit(ETHTOOL_LINK_MODE_1000baseT_Full_BIT, linkmodes); + __set_bit(ETHTOOL_LINK_MODE_1000baseKX_Full_BIT, linkmodes); __set_bit(ETHTOOL_LINK_MODE_1000baseX_Full_BIT, linkmodes); __set_bit(ETHTOOL_LINK_MODE_1000baseT1_Full_BIT, linkmodes); } From 88181f1d3474f59fb897ed01e4e4d8f77802fa4a Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 10:41:42 -0800 Subject: [PATCH 0375/2048] cxgb3: Use struct_group() for memcpy() region In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy(), memmove(), and memset(), avoid intentionally writing across neighboring fields. Use struct_group() in struct rss_hdr around members imm_data and intr_gen, so they can be referenced together. This will allow memcpy() and sizeof() to more easily reason about sizes, improve readability, and avoid future warnings about writing beyond the end of imm_data. "pahole" shows no size nor member offset changes to struct rss_hdr. "objdump -d" shows no object code changes. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- drivers/net/ethernet/chelsio/cxgb3/sge.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/chelsio/cxgb3/sge.c b/drivers/net/ethernet/chelsio/cxgb3/sge.c index c3afec1041f8..70f528a9c727 100644 --- a/drivers/net/ethernet/chelsio/cxgb3/sge.c +++ b/drivers/net/ethernet/chelsio/cxgb3/sge.c @@ -126,8 +126,10 @@ struct rsp_desc { /* response queue descriptor */ struct rss_header rss_hdr; __be32 flags; __be32 len_cq; - u8 imm_data[47]; - u8 intr_gen; + struct_group(immediate, + u8 imm_data[47]; + u8 intr_gen; + ); }; /* @@ -925,7 +927,8 @@ static inline struct sk_buff *get_imm_packet(const struct rsp_desc *resp) if (skb) { __skb_put(skb, IMMED_PKT_SIZE); - skb_copy_to_linear_data(skb, resp->imm_data, IMMED_PKT_SIZE); + BUILD_BUG_ON(IMMED_PKT_SIZE != sizeof(resp->immediate)); + skb_copy_to_linear_data(skb, &resp->immediate, IMMED_PKT_SIZE); } return skb; } From 641d3ef00ce3982aadceeeba6c3e24bee43ce62e Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 10:42:35 -0800 Subject: [PATCH 0376/2048] cxgb4: Use struct_group() for memcpy() region In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy(), memmove(), and memset(), avoid intentionally writing across neighboring fields. Use struct_group() in struct fw_eth_tx_pkt_vm_wr around members ethmacdst, ethmacsrc, ethtype, and vlantci, so they can be referenced together. This will allow memcpy() and sizeof() to more easily reason about sizes, improve readability, and avoid future warnings about writing beyond the end of ethmacdst. "pahole" shows no size nor member offset changes to struct fw_eth_tx_pkt_vm_wr. "objdump -d" shows no object code changes. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- drivers/net/ethernet/chelsio/cxgb4/sge.c | 8 +++++--- drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h | 10 ++++++---- drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 7 ++----- 3 files changed, 13 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/chelsio/cxgb4/sge.c b/drivers/net/ethernet/chelsio/cxgb4/sge.c index fa5b596ff23a..f889f404305c 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/sge.c +++ b/drivers/net/ethernet/chelsio/cxgb4/sge.c @@ -1842,8 +1842,10 @@ static netdev_tx_t cxgb4_vf_eth_xmit(struct sk_buff *skb, * (including the VLAN tag) into the header so we reject anything * smaller than that ... */ - fw_hdr_copy_len = sizeof(wr->ethmacdst) + sizeof(wr->ethmacsrc) + - sizeof(wr->ethtype) + sizeof(wr->vlantci); + BUILD_BUG_ON(sizeof(wr->firmware) != + (sizeof(wr->ethmacdst) + sizeof(wr->ethmacsrc) + + sizeof(wr->ethtype) + sizeof(wr->vlantci))); + fw_hdr_copy_len = sizeof(wr->firmware); ret = cxgb4_validate_skb(skb, dev, fw_hdr_copy_len); if (ret) goto out_free; @@ -1924,7 +1926,7 @@ static netdev_tx_t cxgb4_vf_eth_xmit(struct sk_buff *skb, wr->equiq_to_len16 = cpu_to_be32(wr_mid); wr->r3[0] = cpu_to_be32(0); wr->r3[1] = cpu_to_be32(0); - skb_copy_from_linear_data(skb, (void *)wr->ethmacdst, fw_hdr_copy_len); + skb_copy_from_linear_data(skb, &wr->firmware, fw_hdr_copy_len); end = (u64 *)wr + flits; /* If this is a Large Send Offload packet we'll put in an LSO CPL diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h index 0a326c054707..2419459a0b85 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h +++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h @@ -794,10 +794,12 @@ struct fw_eth_tx_pkt_vm_wr { __be32 op_immdlen; __be32 equiq_to_len16; __be32 r3[2]; - u8 ethmacdst[6]; - u8 ethmacsrc[6]; - __be16 ethtype; - __be16 vlantci; + struct_group(firmware, + u8 ethmacdst[ETH_ALEN]; + u8 ethmacsrc[ETH_ALEN]; + __be16 ethtype; + __be16 vlantci; + ); }; #define FW_CMD_MAX_TIMEOUT 10000 diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c index 0295b2406646..43b2ceb6aa32 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c @@ -1167,10 +1167,7 @@ netdev_tx_t t4vf_eth_xmit(struct sk_buff *skb, struct net_device *dev) struct cpl_tx_pkt_core *cpl; const struct skb_shared_info *ssi; dma_addr_t addr[MAX_SKB_FRAGS + 1]; - const size_t fw_hdr_copy_len = (sizeof(wr->ethmacdst) + - sizeof(wr->ethmacsrc) + - sizeof(wr->ethtype) + - sizeof(wr->vlantci)); + const size_t fw_hdr_copy_len = sizeof(wr->firmware); /* * The chip minimum packet length is 10 octets but the firmware @@ -1267,7 +1264,7 @@ netdev_tx_t t4vf_eth_xmit(struct sk_buff *skb, struct net_device *dev) wr->equiq_to_len16 = cpu_to_be32(wr_mid); wr->r3[0] = cpu_to_be32(0); wr->r3[1] = cpu_to_be32(0); - skb_copy_from_linear_data(skb, (void *)wr->ethmacdst, fw_hdr_copy_len); + skb_copy_from_linear_data(skb, &wr->firmware, fw_hdr_copy_len); end = (u64 *)wr + flits; /* From 29fd0ec65e912613945c2d9fe7d551a842a26b0d Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 10:42:53 -0800 Subject: [PATCH 0377/2048] bnx2x: Use struct_group() for memcpy() region In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy(), memmove(), and memset(), avoid intentionally writing across neighboring fields. Use struct_group() in struct nig_stats around members egress_mac_pkt0_lo, egress_mac_pkt0_hi, egress_mac_pkt1_lo, and egress_mac_pkt1_hi (and the respective members in struct bnx2x_eth_stats), so they can be referenced together. This will allow memcpy() and sizeof() to more easily reason about sizes, improve readability, and avoid future warnings about writing beyond the end of struct bnx2x_eth_stats's rx_stat_ifhcinbadoctets_hi. "pahole" shows no size nor member offset changes to either struct. "objdump -d" shows no meaningful object code changes (i.e. only source line number induced differences and optimizations). Additionally adds BUILD_BUG_ON() to compare the separate struct group sizes. Reviewed-by: Prabhakar Kushwaha Link: https://lore.kernel.org/lkml/DM5PR18MB2229B0413C372CC6E49D59A3B2C59@DM5PR18MB2229.namprd18.prod.outlook.com Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.c | 7 ++++--- drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.h | 14 ++++++++++---- 2 files changed, 14 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.c index 0b193edb73b8..2bb133ae61c3 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.c @@ -849,7 +849,8 @@ static int bnx2x_hw_stats_update(struct bnx2x *bp) memcpy(old, new, sizeof(struct nig_stats)); - memcpy(&(estats->rx_stat_ifhcinbadoctets_hi), &(pstats->mac_stx[1]), + BUILD_BUG_ON(sizeof(estats->shared) != sizeof(pstats->mac_stx[1])); + memcpy(&(estats->shared), &(pstats->mac_stx[1]), sizeof(struct mac_stx)); estats->brb_drop_hi = pstats->brb_drop_hi; estats->brb_drop_lo = pstats->brb_drop_lo; @@ -1634,9 +1635,9 @@ void bnx2x_stats_init(struct bnx2x *bp) REG_RD(bp, NIG_REG_STAT0_BRB_TRUNCATE + port*0x38); if (!CHIP_IS_E3(bp)) { REG_RD_DMAE(bp, NIG_REG_STAT0_EGRESS_MAC_PKT0 + port*0x50, - &(bp->port.old_nig_stats.egress_mac_pkt0_lo), 2); + &(bp->port.old_nig_stats.egress_mac_pkt0), 2); REG_RD_DMAE(bp, NIG_REG_STAT0_EGRESS_MAC_PKT1 + port*0x50, - &(bp->port.old_nig_stats.egress_mac_pkt1_lo), 2); + &(bp->port.old_nig_stats.egress_mac_pkt1), 2); } /* Prepare statistics ramrod data */ diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.h index d55e63692cf3..ae93c078707b 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_stats.h @@ -36,10 +36,14 @@ struct nig_stats { u32 pbf_octets; u32 pbf_packet; u32 safc_inp; - u32 egress_mac_pkt0_lo; - u32 egress_mac_pkt0_hi; - u32 egress_mac_pkt1_lo; - u32 egress_mac_pkt1_hi; + struct_group(egress_mac_pkt0, + u32 egress_mac_pkt0_lo; + u32 egress_mac_pkt0_hi; + ); + struct_group(egress_mac_pkt1, + u32 egress_mac_pkt1_lo; + u32 egress_mac_pkt1_hi; + ); }; enum bnx2x_stats_event { @@ -83,6 +87,7 @@ struct bnx2x_eth_stats { u32 no_buff_discard_hi; u32 no_buff_discard_lo; + struct_group(shared, u32 rx_stat_ifhcinbadoctets_hi; u32 rx_stat_ifhcinbadoctets_lo; u32 tx_stat_ifhcoutbadoctets_hi; @@ -159,6 +164,7 @@ struct bnx2x_eth_stats { u32 tx_stat_dot3statsinternalmactransmiterrors_lo; u32 tx_stat_bmac_ufl_hi; u32 tx_stat_bmac_ufl_lo; + ); u32 pause_frames_received_hi; u32 pause_frames_received_lo; From a6366b13c165e5d8d8b06d8d732a4f98bfb8ef66 Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Thu, 18 Nov 2021 20:37:15 +0100 Subject: [PATCH 0378/2048] net: ipa: Use 'for_each_clear_bit' when possible Use 'for_each_clear_bit()' instead of hand writing it. It is much less version. Signed-off-by: Christophe JAILLET Signed-off-by: David S. Miller --- drivers/net/ipa/ipa_mem.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/net/ipa/ipa_mem.c b/drivers/net/ipa/ipa_mem.c index 4337b0920d3d..1e9eae208e44 100644 --- a/drivers/net/ipa/ipa_mem.c +++ b/drivers/net/ipa/ipa_mem.c @@ -266,9 +266,7 @@ static bool ipa_mem_valid(struct ipa *ipa, const struct ipa_mem_data *mem_data) } /* Now see if any required regions are not defined */ - for (mem_id = find_first_zero_bit(regions, IPA_MEM_COUNT); - mem_id < IPA_MEM_COUNT; - mem_id = find_next_zero_bit(regions, IPA_MEM_COUNT, mem_id + 1)) { + for_each_clear_bit(mem_id, regions, IPA_MEM_COUNT) { if (ipa_mem_id_required(ipa, mem_id)) dev_err(dev, "required memory region %u missing\n", mem_id); From 92e888bc6f1b84cb4dee23d768777362221be1d4 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Thu, 18 Nov 2021 21:04:23 +0100 Subject: [PATCH 0379/2048] sky2: use PCI VPD API in eeprom ethtool ops Recently pci_read/write_vpd_any() have been added to the PCI VPD API. These functions allow to access VPD address space outside the auto-detected VPD, and they can be used to significantly simplify the eeprom ethtool ops. Tested with a 88E8070 card with 1KB EEPROM. Signed-off-by: Heiner Kallweit Reviewed-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/sky2.c | 84 +++++------------------------ 1 file changed, 12 insertions(+), 72 deletions(-) diff --git a/drivers/net/ethernet/marvell/sky2.c b/drivers/net/ethernet/marvell/sky2.c index 28b5b9341145..d53ce826befd 100644 --- a/drivers/net/ethernet/marvell/sky2.c +++ b/drivers/net/ethernet/marvell/sky2.c @@ -4266,96 +4266,36 @@ static int sky2_get_eeprom_len(struct net_device *dev) return 1 << ( ((reg2 & PCI_VPD_ROM_SZ) >> 14) + 8); } -static int sky2_vpd_wait(const struct sky2_hw *hw, int cap, u16 busy) -{ - unsigned long start = jiffies; - - while ( (sky2_pci_read16(hw, cap + PCI_VPD_ADDR) & PCI_VPD_ADDR_F) == busy) { - /* Can take up to 10.6 ms for write */ - if (time_after(jiffies, start + HZ/4)) { - dev_err(&hw->pdev->dev, "VPD cycle timed out\n"); - return -ETIMEDOUT; - } - msleep(1); - } - - return 0; -} - -static int sky2_vpd_read(struct sky2_hw *hw, int cap, void *data, - u16 offset, size_t length) -{ - int rc = 0; - - while (length > 0) { - u32 val; - - sky2_pci_write16(hw, cap + PCI_VPD_ADDR, offset); - rc = sky2_vpd_wait(hw, cap, 0); - if (rc) - break; - - val = sky2_pci_read32(hw, cap + PCI_VPD_DATA); - - memcpy(data, &val, min(sizeof(val), length)); - offset += sizeof(u32); - data += sizeof(u32); - length -= sizeof(u32); - } - - return rc; -} - -static int sky2_vpd_write(struct sky2_hw *hw, int cap, const void *data, - u16 offset, unsigned int length) -{ - unsigned int i; - int rc = 0; - - for (i = 0; i < length; i += sizeof(u32)) { - u32 val = *(u32 *)(data + i); - - sky2_pci_write32(hw, cap + PCI_VPD_DATA, val); - sky2_pci_write32(hw, cap + PCI_VPD_ADDR, offset | PCI_VPD_ADDR_F); - - rc = sky2_vpd_wait(hw, cap, PCI_VPD_ADDR_F); - if (rc) - break; - } - return rc; -} - static int sky2_get_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, u8 *data) { struct sky2_port *sky2 = netdev_priv(dev); - int cap = pci_find_capability(sky2->hw->pdev, PCI_CAP_ID_VPD); - - if (!cap) - return -EINVAL; + int rc; eeprom->magic = SKY2_EEPROM_MAGIC; + rc = pci_read_vpd_any(sky2->hw->pdev, eeprom->offset, eeprom->len, + data); + if (rc < 0) + return rc; - return sky2_vpd_read(sky2->hw, cap, data, eeprom->offset, eeprom->len); + eeprom->len = rc; + + return 0; } static int sky2_set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, u8 *data) { struct sky2_port *sky2 = netdev_priv(dev); - int cap = pci_find_capability(sky2->hw->pdev, PCI_CAP_ID_VPD); - - if (!cap) - return -EINVAL; + int rc; if (eeprom->magic != SKY2_EEPROM_MAGIC) return -EINVAL; - /* Partial writes not supported */ - if ((eeprom->offset & 3) || (eeprom->len & 3)) - return -EINVAL; + rc = pci_write_vpd_any(sky2->hw->pdev, eeprom->offset, eeprom->len, + data); - return sky2_vpd_write(sky2->hw, cap, data, eeprom->offset, eeprom->len); + return rc < 0 ? rc : 0; } static netdev_features_t sky2_fix_features(struct net_device *dev, From f5455a1d9d499e4028b8ae8370758efad00b00e9 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 12:30:19 -0800 Subject: [PATCH 0380/2048] net: dccp: Use memset_startat() for TP zeroing In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memset(), avoid intentionally writing across neighboring fields. Use memset_startat() so memset() doesn't get confused about writing beyond the destination member that is intended to be the starting point of zeroing through the end of the struct. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- net/dccp/trace.h | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/net/dccp/trace.h b/net/dccp/trace.h index 5062421beee9..5a43b3508c7f 100644 --- a/net/dccp/trace.h +++ b/net/dccp/trace.h @@ -60,9 +60,7 @@ TRACE_EVENT(dccp_probe, __entry->tx_t_ipi = hc->tx_t_ipi; } else { __entry->tx_s = 0; - memset(&__entry->tx_rtt, 0, (void *)&__entry->tx_t_ipi - - (void *)&__entry->tx_rtt + - sizeof(__entry->tx_t_ipi)); + memset_startat(__entry, 0, tx_rtt); } ), From e3617433c3da3d0859a4bc67f3f975e87f650ebf Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 12:30:45 -0800 Subject: [PATCH 0381/2048] net: 802: Use memset_startat() to clear struct fields In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memset(), avoid intentionally writing across neighboring fields. Use memset_startat() so memset() doesn't get confused about writing beyond the destination member that is intended to be the starting point of zeroing through the end of the struct. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- net/802/hippi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/802/hippi.c b/net/802/hippi.c index 887e73d520e4..1997b7dd265e 100644 --- a/net/802/hippi.c +++ b/net/802/hippi.c @@ -65,7 +65,7 @@ static int hippi_header(struct sk_buff *skb, struct net_device *dev, hip->le.src_addr_type = 2; /* 12 bit SC address */ memcpy(hip->le.src_switch_addr, dev->dev_addr + 3, 3); - memset(&hip->le.reserved, 0, 16); + memset_startat(&hip->le, 0, reserved); hip->snap.dsap = HIPPI_EXTENDED_SAP; hip->snap.ssap = HIPPI_EXTENDED_SAP; From 8f2a83b454c9b17989843756313e0180827a7ace Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 12:32:41 -0800 Subject: [PATCH 0382/2048] ipv6: Use memset_after() to zero rt6_info In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memset(), avoid intentionally writing across neighboring fields. Use memset_after() to clear everything after the dst_entry member of struct rt6_info. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- net/ipv6/route.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 5e8f2f15607d..ecc6df6592a8 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -328,9 +328,7 @@ static const struct rt6_info ip6_blk_hole_entry_template = { static void rt6_info_init(struct rt6_info *rt) { - struct dst_entry *dst = &rt->dst; - - memset(dst + 1, 0, sizeof(*rt) - sizeof(*dst)); + memset_after(rt, 0, dst); INIT_LIST_HEAD(&rt->rt6i_uncached); } From b5d8cf0af167f3ab9f4cfe44918cde01e20a1222 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 12:34:07 -0800 Subject: [PATCH 0383/2048] net/af_iucv: Use struct_group() to zero struct iucv_sock region In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memset(), avoid intentionally writing across neighboring fields. Add struct_group() to mark the region of struct iucv_sock that gets initialized to zero. Avoid the future warning: In function 'fortify_memset_chk', inlined from 'iucv_sock_alloc' at net/iucv/af_iucv.c:476:2: ./include/linux/fortify-string.h:199:4: warning: call to '__write_overflow_field' declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Wattribute-warning] 199 | __write_overflow_field(p_size_field, size); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Acked-by: Karsten Graul Link: https://lore.kernel.org/lkml/19ff61a0-0cda-6000-ce56-dc6b367c00d6@linux.ibm.com/ Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- include/net/iucv/af_iucv.h | 10 ++++++---- net/iucv/af_iucv.c | 2 +- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/include/net/iucv/af_iucv.h b/include/net/iucv/af_iucv.h index ff06246dbbb9..df85d19fbf84 100644 --- a/include/net/iucv/af_iucv.h +++ b/include/net/iucv/af_iucv.h @@ -112,10 +112,12 @@ enum iucv_tx_notify { struct iucv_sock { struct sock sk; - char src_user_id[8]; - char src_name[8]; - char dst_user_id[8]; - char dst_name[8]; + struct_group(init, + char src_user_id[8]; + char src_name[8]; + char dst_user_id[8]; + char dst_name[8]; + ); struct list_head accept_q; spinlock_t accept_q_lock; struct sock *parent; diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index 996ccf3665e3..49ecbe8d176a 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -473,7 +473,7 @@ static struct sock *iucv_sock_alloc(struct socket *sock, int proto, gfp_t prio, atomic_set(&iucv->msg_recv, 0); iucv->path = NULL; iucv->sk_txnotify = afiucv_hs_callback_txnotify; - memset(&iucv->src_user_id , 0, 32); + memset(&iucv->init, 0, sizeof(iucv->init)); if (pr_iucv) iucv->transport = AF_IUCV_TRANS_IUCV; else From 812ad3d270cb900d156a1542360d81f749dbe040 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 18 Nov 2021 12:34:56 -0800 Subject: [PATCH 0384/2048] ethtool: stats: Use struct_group() to clear all stats at once In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memset(), avoid intentionally writing across neighboring fields. Add struct_group() to mark region of struct stats_reply_data that should be initialized, which can now be done in a single memset() call. Signed-off-by: Kees Cook Signed-off-by: David S. Miller --- net/ethtool/stats.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/net/ethtool/stats.c b/net/ethtool/stats.c index ec07f5765e03..a20e0a24ff61 100644 --- a/net/ethtool/stats.c +++ b/net/ethtool/stats.c @@ -14,10 +14,12 @@ struct stats_req_info { struct stats_reply_data { struct ethnl_reply_data base; - struct ethtool_eth_phy_stats phy_stats; - struct ethtool_eth_mac_stats mac_stats; - struct ethtool_eth_ctrl_stats ctrl_stats; - struct ethtool_rmon_stats rmon_stats; + struct_group(stats, + struct ethtool_eth_phy_stats phy_stats; + struct ethtool_eth_mac_stats mac_stats; + struct ethtool_eth_ctrl_stats ctrl_stats; + struct ethtool_rmon_stats rmon_stats; + ); const struct ethtool_rmon_hist_range *rmon_ranges; }; @@ -117,10 +119,7 @@ static int stats_prepare_data(const struct ethnl_req_info *req_base, /* Mark all stats as unset (see ETHTOOL_STAT_NOT_SET) to prevent them * from being reported to user space in case driver did not set them. */ - memset(&data->phy_stats, 0xff, sizeof(data->phy_stats)); - memset(&data->mac_stats, 0xff, sizeof(data->mac_stats)); - memset(&data->ctrl_stats, 0xff, sizeof(data->ctrl_stats)); - memset(&data->rmon_stats, 0xff, sizeof(data->rmon_stats)); + memset(&data->stats, 0xff, sizeof(data->stats)); if (test_bit(ETHTOOL_STATS_ETH_PHY, req_info->stat_mask) && dev->ethtool_ops->get_eth_phy_stats) From eaa54d66145e4314eb785a507341a7c8b13f6df8 Mon Sep 17 00:00:00 2001 From: Yu Xiao Date: Fri, 19 Nov 2021 09:58:01 +0100 Subject: [PATCH 0385/2048] nfp: flower: correction of error handling Removing reduplicated error handling when running into error path of `nfp_compile_flow_metadata`. Signed-off-by: Yu Xiao Signed-off-by: Yinjun Zhang Signed-off-by: Simon Horman Signed-off-by: David S. Miller --- .../ethernet/netronome/nfp/flower/metadata.c | 20 ------------------- 1 file changed, 20 deletions(-) diff --git a/drivers/net/ethernet/netronome/nfp/flower/metadata.c b/drivers/net/ethernet/netronome/nfp/flower/metadata.c index 2af9faee96c5..0c60a436a8f2 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/metadata.c +++ b/drivers/net/ethernet/netronome/nfp/flower/metadata.c @@ -338,11 +338,6 @@ int nfp_compile_flow_metadata(struct nfp_app *app, u32 cookie, nfp_flow->meta.mask_len, &nfp_flow->meta.flags, &new_mask_id)) { NL_SET_ERR_MSG_MOD(extack, "invalid entry: cannot allocate a new mask id"); - if (nfp_release_stats_entry(app, stats_cxt)) { - NL_SET_ERR_MSG_MOD(extack, "invalid entry: cannot release stats context"); - err = -EINVAL; - goto err_remove_rhash; - } err = -ENOENT; goto err_remove_rhash; } @@ -359,21 +354,6 @@ int nfp_compile_flow_metadata(struct nfp_app *app, u32 cookie, check_entry = nfp_flower_search_fl_table(app, cookie, netdev); if (check_entry) { NL_SET_ERR_MSG_MOD(extack, "invalid entry: cannot offload duplicate flow entry"); - if (nfp_release_stats_entry(app, stats_cxt)) { - NL_SET_ERR_MSG_MOD(extack, "invalid entry: cannot release stats context"); - err = -EINVAL; - goto err_remove_mask; - } - - if (!nfp_flow->pre_tun_rule.dev && - !nfp_check_mask_remove(app, nfp_flow->mask_data, - nfp_flow->meta.mask_len, - NULL, &new_mask_id)) { - NL_SET_ERR_MSG_MOD(extack, "invalid entry: cannot release mask id"); - err = -EINVAL; - goto err_remove_mask; - } - err = -EEXIST; goto err_remove_mask; } From 520fbdf7fb19b7744e370d36d9244a446299ceb7 Mon Sep 17 00:00:00 2001 From: Bernard Zhao Date: Thu, 18 Nov 2021 18:06:42 -0800 Subject: [PATCH 0386/2048] net/bridge: replace simple_strtoul to kstrtol simple_strtoull is obsolete, use kstrtol instead. Signed-off-by: Bernard Zhao Signed-off-by: David S. Miller --- net/bridge/br_sysfs_br.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/net/bridge/br_sysfs_br.c b/net/bridge/br_sysfs_br.c index d9a89ddd0331..11c490694296 100644 --- a/net/bridge/br_sysfs_br.c +++ b/net/bridge/br_sysfs_br.c @@ -36,15 +36,14 @@ static ssize_t store_bridge_parm(struct device *d, struct net_bridge *br = to_bridge(d); struct netlink_ext_ack extack = {0}; unsigned long val; - char *endp; int err; if (!ns_capable(dev_net(br->dev)->user_ns, CAP_NET_ADMIN)) return -EPERM; - val = simple_strtoul(buf, &endp, 0); - if (endp == buf) - return -EINVAL; + err = kstrtoul(buf, 10, &val); + if (err != 0) + return err; if (!rtnl_trylock()) return restart_syscall(); From 7615209f42a1976894cd0df97a380a034911656a Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Thu, 18 Nov 2021 09:40:54 -0800 Subject: [PATCH 0387/2048] libbpf: Add runtime APIs to query libbpf version MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Libbpf provided LIBBPF_MAJOR_VERSION and LIBBPF_MINOR_VERSION macros to check libbpf version at compilation time. This doesn't cover all the needs, though, because version of libbpf that application is compiled against doesn't necessarily match the version of libbpf at runtime, especially if libbpf is used as a shared library. Add libbpf_major_version() and libbpf_minor_version() returning major and minor versions, respectively, as integers. Also add a convenience libbpf_version_string() for various tooling using libbpf to print out libbpf version in a human-readable form. Currently it will return "v0.6", but in the future it can contains some extra information, so the format itself is not part of a stable API and shouldn't be relied upon. Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann Acked-by: John Fastabend Acked-by: Toke Høiland-Jørgensen Link: https://lore.kernel.org/bpf/20211118174054.2699477-1-andrii@kernel.org --- tools/lib/bpf/libbpf.c | 19 +++++++++++++++++++ tools/lib/bpf/libbpf.h | 4 ++++ tools/lib/bpf/libbpf.map | 3 +++ 3 files changed, 26 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index de7e09a6b5ec..78de238f975a 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -168,6 +168,25 @@ int libbpf_set_strict_mode(enum libbpf_strict_mode mode) return 0; } +__u32 libbpf_major_version(void) +{ + return LIBBPF_MAJOR_VERSION; +} + +__u32 libbpf_minor_version(void) +{ + return LIBBPF_MINOR_VERSION; +} + +const char *libbpf_version_string(void) +{ +#define __S(X) #X +#define _S(X) __S(X) + return "v" _S(LIBBPF_MAJOR_VERSION) "." _S(LIBBPF_MINOR_VERSION); +#undef _S +#undef __S +} + enum kern_feature_id { /* v4.14: kernel support for program & map names. */ FEAT_PROG_NAME, diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 4ec69f224342..003fdc5cf3a8 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -24,6 +24,10 @@ extern "C" { #endif +LIBBPF_API __u32 libbpf_major_version(void); +LIBBPF_API __u32 libbpf_minor_version(void); +LIBBPF_API const char *libbpf_version_string(void); + enum libbpf_errno { __LIBBPF_ERRNO__START = 4000, diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index 6a59514a48cf..bea6791272e5 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -410,6 +410,9 @@ LIBBPF_0.6.0 { btf__type_cnt; btf_dump__new; btf_dump__new_deprecated; + libbpf_major_version; + libbpf_minor_version; + libbpf_version_string; perf_buffer__new; perf_buffer__new_deprecated; perf_buffer__new_raw; From efdd3eb8015e7447095f02a26eaabd164cd18004 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Wed, 17 Nov 2021 11:41:13 -0800 Subject: [PATCH 0388/2048] libbpf: Accommodate DWARF/compiler bug with duplicated structs According to [0], compilers sometimes might produce duplicate DWARF definitions for exactly the same struct/union within the same compilation unit (CU). We've had similar issues with identical arrays and handled them with a similar workaround in 6b6e6b1d09aa ("libbpf: Accomodate DWARF/compiler bug with duplicated identical arrays"). Do the same for struct/union by ensuring that two structs/unions are exactly the same, down to the integer values of field referenced type IDs. Solving this more generically (allowing referenced types to be equivalent, but using different type IDs, all within a single CU) requires a huge complexity increase to handle many-to-many mappings between canonidal and candidate type graphs. Before we invest in that, let's see if this approach handles all the instances of this issue in practice. Thankfully it's pretty rare, it seems. [0] https://lore.kernel.org/bpf/YXr2NFlJTAhHdZqq@krava/ Reported-by: Jiri Olsa Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211117194114.347675-1-andrii@kernel.org --- tools/lib/bpf/btf.c | 45 +++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 41 insertions(+), 4 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index b6be579e0dc6..e97217a77196 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -3477,8 +3477,8 @@ static long btf_hash_struct(struct btf_type *t) } /* - * Check structural compatibility of two FUNC_PROTOs, ignoring referenced type - * IDs. This check is performed during type graph equivalence check and + * Check structural compatibility of two STRUCTs/UNIONs, ignoring referenced + * type IDs. This check is performed during type graph equivalence check and * referenced types equivalence is checked separately. */ static bool btf_shallow_equal_struct(struct btf_type *t1, struct btf_type *t2) @@ -3851,6 +3851,31 @@ static int btf_dedup_identical_arrays(struct btf_dedup *d, __u32 id1, __u32 id2) return btf_equal_array(t1, t2); } +/* Check if given two types are identical STRUCT/UNION definitions */ +static bool btf_dedup_identical_structs(struct btf_dedup *d, __u32 id1, __u32 id2) +{ + const struct btf_member *m1, *m2; + struct btf_type *t1, *t2; + int n, i; + + t1 = btf_type_by_id(d->btf, id1); + t2 = btf_type_by_id(d->btf, id2); + + if (!btf_is_composite(t1) || btf_kind(t1) != btf_kind(t2)) + return false; + + if (!btf_shallow_equal_struct(t1, t2)) + return false; + + m1 = btf_members(t1); + m2 = btf_members(t2); + for (i = 0, n = btf_vlen(t1); i < n; i++, m1++, m2++) { + if (m1->type != m2->type) + return false; + } + return true; +} + /* * Check equivalence of BTF type graph formed by candidate struct/union (we'll * call it "candidate graph" in this description for brevity) to a type graph @@ -3962,6 +3987,8 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id, hypot_type_id = d->hypot_map[canon_id]; if (hypot_type_id <= BTF_MAX_NR_TYPES) { + if (hypot_type_id == cand_id) + return 1; /* In some cases compiler will generate different DWARF types * for *identical* array type definitions and use them for * different fields within the *same* struct. This breaks type @@ -3970,8 +3997,18 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id, * types within a single CU. So work around that by explicitly * allowing identical array types here. */ - return hypot_type_id == cand_id || - btf_dedup_identical_arrays(d, hypot_type_id, cand_id); + if (btf_dedup_identical_arrays(d, hypot_type_id, cand_id)) + return 1; + /* It turns out that similar situation can happen with + * struct/union sometimes, sigh... Handle the case where + * structs/unions are exactly the same, down to the referenced + * type IDs. Anything more complicated (e.g., if referenced + * types are different, but equivalent) is *way more* + * complicated and requires a many-to-many equivalence mapping. + */ + if (btf_dedup_identical_structs(d, hypot_type_id, cand_id)) + return 1; + return 0; } if (btf_dedup_hypot_map_add(d, canon_id, cand_id)) From 9a49afe6f5a516eb33bec24be0f81cb35ca79445 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 17 Nov 2021 11:41:14 -0800 Subject: [PATCH 0389/2048] selftests/bpf: Add btf_dedup case with duplicated structs within CU Add an artificial minimal example simulating compilers producing two different types within a single CU that correspond to identical struct definitions. Signed-off-by: Jiri Olsa Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211117194114.347675-2-andrii@kernel.org --- .../bpf/prog_tests/btf_dedup_split.c | 113 ++++++++++++++++++ 1 file changed, 113 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c b/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c index 9d3b8d7a1537..94ff9757557a 100644 --- a/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c +++ b/tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c @@ -314,6 +314,117 @@ cleanup: btf__free(btf1); } +static void btf_add_dup_struct_in_cu(struct btf *btf, int start_id) +{ +#define ID(n) (start_id + n) + btf__set_pointer_size(btf, 8); /* enforce 64-bit arch */ + + btf__add_int(btf, "int", 4, BTF_INT_SIGNED); /* [1] int */ + + btf__add_struct(btf, "s", 8); /* [2] struct s { */ + btf__add_field(btf, "a", ID(3), 0, 0); /* struct anon a; */ + btf__add_field(btf, "b", ID(4), 0, 0); /* struct anon b; */ + /* } */ + + btf__add_struct(btf, "(anon)", 8); /* [3] struct anon { */ + btf__add_field(btf, "f1", ID(1), 0, 0); /* int f1; */ + btf__add_field(btf, "f2", ID(1), 32, 0); /* int f2; */ + /* } */ + + btf__add_struct(btf, "(anon)", 8); /* [4] struct anon { */ + btf__add_field(btf, "f1", ID(1), 0, 0); /* int f1; */ + btf__add_field(btf, "f2", ID(1), 32, 0); /* int f2; */ + /* } */ +#undef ID +} + +static void test_split_dup_struct_in_cu() +{ + struct btf *btf1, *btf2; + int err; + + /* generate the base data.. */ + btf1 = btf__new_empty(); + if (!ASSERT_OK_PTR(btf1, "empty_main_btf")) + return; + + btf_add_dup_struct_in_cu(btf1, 0); + + VALIDATE_RAW_BTF( + btf1, + "[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[2] STRUCT 's' size=8 vlen=2\n" + "\t'a' type_id=3 bits_offset=0\n" + "\t'b' type_id=4 bits_offset=0", + "[3] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=1 bits_offset=0\n" + "\t'f2' type_id=1 bits_offset=32", + "[4] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=1 bits_offset=0\n" + "\t'f2' type_id=1 bits_offset=32"); + + /* ..dedup them... */ + err = btf__dedup(btf1, NULL, NULL); + if (!ASSERT_OK(err, "btf_dedup")) + goto cleanup; + + VALIDATE_RAW_BTF( + btf1, + "[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[2] STRUCT 's' size=8 vlen=2\n" + "\t'a' type_id=3 bits_offset=0\n" + "\t'b' type_id=3 bits_offset=0", + "[3] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=1 bits_offset=0\n" + "\t'f2' type_id=1 bits_offset=32"); + + /* and add the same data on top of it */ + btf2 = btf__new_empty_split(btf1); + if (!ASSERT_OK_PTR(btf2, "empty_split_btf")) + goto cleanup; + + btf_add_dup_struct_in_cu(btf2, 3); + + VALIDATE_RAW_BTF( + btf2, + "[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[2] STRUCT 's' size=8 vlen=2\n" + "\t'a' type_id=3 bits_offset=0\n" + "\t'b' type_id=3 bits_offset=0", + "[3] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=1 bits_offset=0\n" + "\t'f2' type_id=1 bits_offset=32", + "[4] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[5] STRUCT 's' size=8 vlen=2\n" + "\t'a' type_id=6 bits_offset=0\n" + "\t'b' type_id=7 bits_offset=0", + "[6] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=4 bits_offset=0\n" + "\t'f2' type_id=4 bits_offset=32", + "[7] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=4 bits_offset=0\n" + "\t'f2' type_id=4 bits_offset=32"); + + err = btf__dedup(btf2, NULL, NULL); + if (!ASSERT_OK(err, "btf_dedup")) + goto cleanup; + + /* after dedup it should match the original data */ + VALIDATE_RAW_BTF( + btf2, + "[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED", + "[2] STRUCT 's' size=8 vlen=2\n" + "\t'a' type_id=3 bits_offset=0\n" + "\t'b' type_id=3 bits_offset=0", + "[3] STRUCT '(anon)' size=8 vlen=2\n" + "\t'f1' type_id=1 bits_offset=0\n" + "\t'f2' type_id=1 bits_offset=32"); + +cleanup: + btf__free(btf2); + btf__free(btf1); +} + void test_btf_dedup_split() { if (test__start_subtest("split_simple")) @@ -322,4 +433,6 @@ void test_btf_dedup_split() test_split_struct_duped(); if (test__start_subtest("split_fwd_resolve")) test_split_fwd_resolve(); + if (test__start_subtest("split_dup_struct_in_cu")) + test_split_dup_struct_in_cu(); } From 8cccee9e91e19207671b94af40bacf7c1d2e74ef Mon Sep 17 00:00:00 2001 From: Florent Revest Date: Fri, 19 Nov 2021 19:00:35 +0100 Subject: [PATCH 0390/2048] libbpf: Change bpf_program__set_extra_flags to bpf_program__set_flags bpf_program__set_extra_flags has just been introduced so we can still change it without breaking users. This new interface is a bit more flexible (for example if someone wants to clear a flag). Signed-off-by: Florent Revest Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20211119180035.1396139-1-revest@chromium.org --- tools/lib/bpf/libbpf.c | 4 ++-- tools/lib/bpf/libbpf.h | 2 +- tools/lib/bpf/libbpf.map | 2 +- tools/testing/selftests/bpf/testing_helpers.c | 4 +++- 4 files changed, 7 insertions(+), 5 deletions(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 78de238f975a..af405c38aadc 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -8315,12 +8315,12 @@ __u32 bpf_program__flags(const struct bpf_program *prog) return prog->prog_flags; } -int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags) +int bpf_program__set_flags(struct bpf_program *prog, __u32 flags) { if (prog->obj->loaded) return libbpf_err(-EBUSY); - prog->prog_flags |= extra_flags; + prog->prog_flags = flags; return 0; } diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 003fdc5cf3a8..d02139fec4ac 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -498,7 +498,7 @@ bpf_program__set_expected_attach_type(struct bpf_program *prog, enum bpf_attach_type type); LIBBPF_API __u32 bpf_program__flags(const struct bpf_program *prog); -LIBBPF_API int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags); +LIBBPF_API int bpf_program__set_flags(struct bpf_program *prog, __u32 flags); LIBBPF_API int bpf_program__set_attach_target(struct bpf_program *prog, int attach_prog_fd, diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index bea6791272e5..69bc069f0a68 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -400,7 +400,7 @@ LIBBPF_0.6.0 { bpf_program__flags; bpf_program__insn_cnt; bpf_program__insns; - bpf_program__set_extra_flags; + bpf_program__set_flags; btf__add_btf; btf__add_decl_tag; btf__add_type_tag; diff --git a/tools/testing/selftests/bpf/testing_helpers.c b/tools/testing/selftests/bpf/testing_helpers.c index 52c2f24e0898..0f1c37ac6f2c 100644 --- a/tools/testing/selftests/bpf/testing_helpers.c +++ b/tools/testing/selftests/bpf/testing_helpers.c @@ -91,6 +91,7 @@ int bpf_prog_test_load(const char *file, enum bpf_prog_type type, struct bpf_object_load_attr attr = {}; struct bpf_object *obj; struct bpf_program *prog; + __u32 flags; int err; obj = bpf_object__open(file); @@ -106,7 +107,8 @@ int bpf_prog_test_load(const char *file, enum bpf_prog_type type, if (type != BPF_PROG_TYPE_UNSPEC) bpf_program__set_type(prog, type); - bpf_program__set_extra_flags(prog, BPF_F_TEST_RND_HI32); + flags = bpf_program__flags(prog) | BPF_F_TEST_RND_HI32; + bpf_program__set_flags(prog, flags); attr.obj = obj; attr.log_level = extra_prog_load_log_flags; From 8d22679dc89a6d9e1d41b2514902e3f7ef51547a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 18 Nov 2021 18:23:55 -0800 Subject: [PATCH 0391/2048] ipv6: ip6_skb_dst_mtu() cleanups Use const attribute where we can, and cache skb_dst() Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/r/20211119022355.2985984-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski --- include/net/ip6_route.h | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h index 5efd0b71dc67..ca2d6b60e1ec 100644 --- a/include/net/ip6_route.h +++ b/include/net/ip6_route.h @@ -263,19 +263,19 @@ static inline bool ipv6_anycast_destination(const struct dst_entry *dst, int ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, int (*output)(struct net *, struct sock *, struct sk_buff *)); -static inline unsigned int ip6_skb_dst_mtu(struct sk_buff *skb) +static inline unsigned int ip6_skb_dst_mtu(const struct sk_buff *skb) { + const struct ipv6_pinfo *np = skb->sk && !dev_recursion_level() ? + inet6_sk(skb->sk) : NULL; + const struct dst_entry *dst = skb_dst(skb); unsigned int mtu; - struct ipv6_pinfo *np = skb->sk && !dev_recursion_level() ? - inet6_sk(skb->sk) : NULL; - if (np && np->pmtudisc >= IPV6_PMTUDISC_PROBE) { - mtu = READ_ONCE(skb_dst(skb)->dev->mtu); - mtu -= lwtunnel_headroom(skb_dst(skb)->lwtstate, mtu); - } else - mtu = dst_mtu(skb_dst(skb)); - + mtu = READ_ONCE(dst->dev->mtu); + mtu -= lwtunnel_headroom(dst->lwtstate, mtu); + } else { + mtu = dst_mtu(dst); + } return mtu; } From d9f31aeaa1e5aefa68130878af3c3513d41c1e2d Mon Sep 17 00:00:00 2001 From: Yang Li Date: Thu, 18 Nov 2021 17:48:03 +0800 Subject: [PATCH 0392/2048] ethernet: renesas: Use div64_ul instead of do_div do_div() does a 64-by-32 division. Here the divisor is an unsigned long which on some platforms is 64 bit wide. So use div64_ul instead of do_div to avoid a possible truncation. Eliminate the following coccicheck warning: ./drivers/net/ethernet/renesas/ravb_main.c:2492:1-7: WARNING: do_div() does a 64-by-32 division, please consider using div64_ul instead. Reported-by: Abaci Robot Signed-off-by: Yang Li Reviewed-by: Geert Uytterhoeven Reviewed-by: Sergey Shtylyov Link: https://lore.kernel.org/r/1637228883-100100-1-git-send-email-yang.lee@linux.alibaba.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/renesas/ravb_main.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index b4c597f4040c..151cce2fe36d 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -30,8 +30,7 @@ #include #include #include - -#include +#include #include "ravb.h" @@ -2488,8 +2487,7 @@ static int ravb_set_gti(struct net_device *ndev) if (!rate) return -EINVAL; - inc = 1000000000ULL << 20; - do_div(inc, rate); + inc = div64_ul(1000000000ULL << 20, rate); if (inc < GTI_TIV_MIN || inc > GTI_TIV_MAX) { dev_err(dev, "gti.tiv increment 0x%llx is outside the range 0x%x - 0x%x\n", From 1388d4ad9d828f7d5ad8327d12fa2afb9abdf08a Mon Sep 17 00:00:00 2001 From: Hari Nagalla Date: Thu, 18 Nov 2021 04:25:32 -0600 Subject: [PATCH 0393/2048] net: phy: add support for TI DP83561-SP phy Add support for the TI DP83561-SP Gigabit ethernet phy device. The dp83561-sp is a radiation hardened space grade gigabit ethernet PHY. It has been tested for single event latch upto 121 MeV, the critical reliability parameter for space designs. It interfaces directly to twisted pair media through an external transformer. And the device also interfaces directly to the MAC layer through Reduced GMII (RGMII) and MII. DP83867, DP83869 and DP83561-SP, all these parts support 1000Base-T/ 100Base-TX/ and 10Base-Te standards and have similar register map for the core functionality. The data sheet for this part is at https://www.ti.com/product/DP83561-SP Signed-off-by: Hari Nagalla Signed-off-by: Geet Modi Reviewed-by: Andrew Lunn Link: https://lore.kernel.org/r/20211118102532.9835-1-hnagalla@ti.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/dp83869.c | 44 ++++++++++++++++++++------------------- 1 file changed, 23 insertions(+), 21 deletions(-) diff --git a/drivers/net/phy/dp83869.c b/drivers/net/phy/dp83869.c index 7113925606f7..b4ff9c5073a3 100644 --- a/drivers/net/phy/dp83869.c +++ b/drivers/net/phy/dp83869.c @@ -16,6 +16,7 @@ #include #define DP83869_PHY_ID 0x2000a0f1 +#define DP83561_PHY_ID 0x2000a1a4 #define DP83869_DEVADDR 0x1f #define MII_DP83869_PHYCTRL 0x10 @@ -878,34 +879,35 @@ static int dp83869_phy_reset(struct phy_device *phydev) return dp83869_config_init(phydev); } + +#define DP83869_PHY_DRIVER(_id, _name) \ +{ \ + PHY_ID_MATCH_MODEL(_id), \ + .name = (_name), \ + .probe = dp83869_probe, \ + .config_init = dp83869_config_init, \ + .soft_reset = dp83869_phy_reset, \ + .config_intr = dp83869_config_intr, \ + .handle_interrupt = dp83869_handle_interrupt, \ + .read_status = dp83869_read_status, \ + .get_tunable = dp83869_get_tunable, \ + .set_tunable = dp83869_set_tunable, \ + .get_wol = dp83869_get_wol, \ + .set_wol = dp83869_set_wol, \ + .suspend = genphy_suspend, \ + .resume = genphy_resume, \ +} + static struct phy_driver dp83869_driver[] = { - { - PHY_ID_MATCH_MODEL(DP83869_PHY_ID), - .name = "TI DP83869", + DP83869_PHY_DRIVER(DP83869_PHY_ID, "TI DP83869"), + DP83869_PHY_DRIVER(DP83561_PHY_ID, "TI DP83561-SP"), - .probe = dp83869_probe, - .config_init = dp83869_config_init, - .soft_reset = dp83869_phy_reset, - - /* IRQ related */ - .config_intr = dp83869_config_intr, - .handle_interrupt = dp83869_handle_interrupt, - .read_status = dp83869_read_status, - - .get_tunable = dp83869_get_tunable, - .set_tunable = dp83869_set_tunable, - - .get_wol = dp83869_get_wol, - .set_wol = dp83869_set_wol, - - .suspend = genphy_suspend, - .resume = genphy_resume, - }, }; module_phy_driver(dp83869_driver); static struct mdio_device_id __maybe_unused dp83869_tbl[] = { { PHY_ID_MATCH_MODEL(DP83869_PHY_ID) }, + { PHY_ID_MATCH_MODEL(DP83561_PHY_ID) }, { } }; MODULE_DEVICE_TABLE(mdio, dp83869_tbl); From 0f98d7e478430b64d9520a23585e02be5f8b1b2a Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:49 -0800 Subject: [PATCH 0394/2048] 82596: use eth_hw_addr_set() Byte by byte assignments. Fixes build on m68k. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/i825xx/82596.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/i825xx/82596.c b/drivers/net/ethernet/i825xx/82596.c index b482f6f633bd..3ee89ae496d0 100644 --- a/drivers/net/ethernet/i825xx/82596.c +++ b/drivers/net/ethernet/i825xx/82596.c @@ -1178,7 +1178,8 @@ found: DEB(DEB_PROBE,printk(KERN_INFO "%s: 82596 at %#3lx,", dev->name, dev->base_addr)); for (i = 0; i < 6; i++) - DEB(DEB_PROBE,printk(" %2.2X", dev->dev_addr[i] = eth_addr[i])); + DEB(DEB_PROBE,printk(" %2.2X", eth_addr[i])); + eth_hw_addr_set(dev, eth_addr); DEB(DEB_PROBE,printk(" IRQ %d.\n", dev->irq)); From c9646a18033ea2c3a6d27c951dedf4f60a290bf9 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:50 -0800 Subject: [PATCH 0395/2048] bnx2x: constify static inline stub for dev_addr bnx2x_vfpf_config_mac() was constified by not its stub. Reported-by: kernel test robot Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h index 8c2cf5519787..2dac704dc346 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sriov.h @@ -586,7 +586,7 @@ static inline int bnx2x_vfpf_release(struct bnx2x *bp) {return 0; } static inline int bnx2x_vfpf_init(struct bnx2x *bp) {return 0; } static inline void bnx2x_vfpf_close_vf(struct bnx2x *bp) {} static inline int bnx2x_vfpf_setup_q(struct bnx2x *bp, struct bnx2x_fastpath *fp, bool is_leading) {return 0; } -static inline int bnx2x_vfpf_config_mac(struct bnx2x *bp, u8 *addr, +static inline int bnx2x_vfpf_config_mac(struct bnx2x *bp, const u8 *addr, u8 vf_qid, bool set) {return 0; } static inline int bnx2x_vfpf_config_rss(struct bnx2x *bp, struct bnx2x_config_rss_params *params) {return 0; } From adeef3e32146a8d2a73c399dc6f5d76a449131b1 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:51 -0800 Subject: [PATCH 0396/2048] net: constify netdev->dev_addr Commit 406f42fa0d3c ("net-next: When a bond have a massive amount of VLANs...") introduced a rbtree for faster Ethernet address look up. We converted all users to make modifications via appropriate helpers, make netdev->dev_addr const. The update helpers need to upcast from the buffer to struct netdev_hw_addr. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- include/linux/netdevice.h | 14 +++++--------- net/core/dev_addr_lists.c | 10 ++++++++++ 2 files changed, 15 insertions(+), 9 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 4f4a299e92de..2462195784a9 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2117,7 +2117,7 @@ struct net_device { * Cache lines mostly used on receive path (including eth_type_trans()) */ /* Interface address info used in eth_type_trans() */ - unsigned char *dev_addr; + const unsigned char *dev_addr; struct netdev_rx_queue *_rx; unsigned int num_rx_queues; @@ -4268,10 +4268,13 @@ void __hw_addr_unsync_dev(struct netdev_hw_addr_list *list, void __hw_addr_init(struct netdev_hw_addr_list *list); /* Functions used for device addresses handling */ +void dev_addr_mod(struct net_device *dev, unsigned int offset, + const void *addr, size_t len); + static inline void __dev_addr_set(struct net_device *dev, const void *addr, size_t len) { - memcpy(dev->dev_addr, addr, len); + dev_addr_mod(dev, 0, addr, len); } static inline void dev_addr_set(struct net_device *dev, const u8 *addr) @@ -4279,13 +4282,6 @@ static inline void dev_addr_set(struct net_device *dev, const u8 *addr) __dev_addr_set(dev, addr, dev->addr_len); } -static inline void -dev_addr_mod(struct net_device *dev, unsigned int offset, - const void *addr, size_t len) -{ - memcpy(&dev->dev_addr[offset], addr, len); -} - int dev_addr_add(struct net_device *dev, const unsigned char *addr, unsigned char addr_type); int dev_addr_del(struct net_device *dev, const unsigned char *addr, diff --git a/net/core/dev_addr_lists.c b/net/core/dev_addr_lists.c index f0cb38344126..ae8b1ef00fec 100644 --- a/net/core/dev_addr_lists.c +++ b/net/core/dev_addr_lists.c @@ -549,6 +549,16 @@ int dev_addr_init(struct net_device *dev) } EXPORT_SYMBOL(dev_addr_init); +void dev_addr_mod(struct net_device *dev, unsigned int offset, + const void *addr, size_t len) +{ + struct netdev_hw_addr *ha; + + ha = container_of(dev->dev_addr, struct netdev_hw_addr, addr[0]); + memcpy(&ha->addr[offset], addr, len); +} +EXPORT_SYMBOL(dev_addr_mod); + /** * dev_addr_add - Add a device address * @dev: device From 5f0b69238427f19a20063e90486754b1ea3ccd3b Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:52 -0800 Subject: [PATCH 0397/2048] net: unexport dev_addr_init() & dev_addr_flush() There are no module callers in-tree and it's hard to justify why anyone would init or flush addresses of a netdev (note the flush is more of a destructor, it frees netdev->dev_addr). Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- net/core/dev_addr_lists.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/net/core/dev_addr_lists.c b/net/core/dev_addr_lists.c index ae8b1ef00fec..a23a83ac18e5 100644 --- a/net/core/dev_addr_lists.c +++ b/net/core/dev_addr_lists.c @@ -513,7 +513,6 @@ void dev_addr_flush(struct net_device *dev) __hw_addr_flush(&dev->dev_addrs); dev->dev_addr = NULL; } -EXPORT_SYMBOL(dev_addr_flush); /** * dev_addr_init - Init device address list @@ -547,7 +546,6 @@ int dev_addr_init(struct net_device *dev) } return err; } -EXPORT_SYMBOL(dev_addr_init); void dev_addr_mod(struct net_device *dev, unsigned int offset, const void *addr, size_t len) From d07b26f5bbea9ade34dfd6abea7b3ca056c03cd1 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:53 -0800 Subject: [PATCH 0398/2048] dev_addr: add a modification check netdev->dev_addr should only be modified via helpers, but someone may be casting off the const. Add a runtime check to catch abuses. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- include/linux/netdevice.h | 5 +++++ net/core/dev.c | 1 + net/core/dev_addr_lists.c | 19 +++++++++++++++++++ 3 files changed, 25 insertions(+) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 2462195784a9..cb7f2661d187 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1942,6 +1942,8 @@ enum netdev_ml_priv_type { * @unlink_list: As netif_addr_lock() can be called recursively, * keep a list of interfaces to be deleted. * + * @dev_addr_shadow: Copy of @dev_addr to catch direct writes. + * * FIXME: cleanup struct net_device such that network protocol info * moves out. */ @@ -2268,6 +2270,8 @@ struct net_device { /* protected by rtnl_lock */ struct bpf_xdp_entity xdp_state[__MAX_XDP_MODE]; + + u8 dev_addr_shadow[MAX_ADDR_LEN]; }; #define to_net_dev(d) container_of(d, struct net_device, dev) @@ -4288,6 +4292,7 @@ int dev_addr_del(struct net_device *dev, const unsigned char *addr, unsigned char addr_type); void dev_addr_flush(struct net_device *dev); int dev_addr_init(struct net_device *dev); +void dev_addr_check(struct net_device *dev); /* Functions used for unicast addresses handling */ int dev_uc_add(struct net_device *dev, const unsigned char *addr); diff --git a/net/core/dev.c b/net/core/dev.c index 92c9258cbf28..9219e319e901 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1377,6 +1377,7 @@ static int __dev_open(struct net_device *dev, struct netlink_ext_ack *extack) int ret; ASSERT_RTNL(); + dev_addr_check(dev); if (!netif_device_present(dev)) { /* may be detached because parent is runtime-suspended */ diff --git a/net/core/dev_addr_lists.c b/net/core/dev_addr_lists.c index a23a83ac18e5..969942734951 100644 --- a/net/core/dev_addr_lists.c +++ b/net/core/dev_addr_lists.c @@ -498,6 +498,21 @@ EXPORT_SYMBOL(__hw_addr_init); * Device addresses handling functions */ +/* Check that netdev->dev_addr is not written to directly as this would + * break the rbtree layout. All changes should go thru dev_addr_set() and co. + * Remove this check in mid-2024. + */ +void dev_addr_check(struct net_device *dev) +{ + if (!memcmp(dev->dev_addr, dev->dev_addr_shadow, MAX_ADDR_LEN)) + return; + + netdev_warn(dev, "Current addr: %*ph\n", MAX_ADDR_LEN, dev->dev_addr); + netdev_warn(dev, "Expected addr: %*ph\n", + MAX_ADDR_LEN, dev->dev_addr_shadow); + netdev_WARN(dev, "Incorrect netdev->dev_addr\n"); +} + /** * dev_addr_flush - Flush device address list * @dev: device @@ -509,6 +524,7 @@ EXPORT_SYMBOL(__hw_addr_init); void dev_addr_flush(struct net_device *dev) { /* rtnl_mutex must be held here */ + dev_addr_check(dev); __hw_addr_flush(&dev->dev_addrs); dev->dev_addr = NULL; @@ -552,8 +568,11 @@ void dev_addr_mod(struct net_device *dev, unsigned int offset, { struct netdev_hw_addr *ha; + dev_addr_check(dev); + ha = container_of(dev->dev_addr, struct netdev_hw_addr, addr[0]); memcpy(&ha->addr[offset], addr, len); + memcpy(&dev->dev_addr_shadow[offset], addr, len); } EXPORT_SYMBOL(dev_addr_mod); From a387ff8e5ddad4d9e4c8fed54bc99be09d501b78 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:54 -0800 Subject: [PATCH 0399/2048] dev_addr_list: put the first addr on the tree Since all netdev->dev_addr modifications go via dev_addr_mod() we can put it on the list. When address is change remove it and add it back. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- net/core/dev_addr_lists.c | 62 +++++++++++++++++++++------------------ 1 file changed, 34 insertions(+), 28 deletions(-) diff --git a/net/core/dev_addr_lists.c b/net/core/dev_addr_lists.c index 969942734951..bead38ca50bd 100644 --- a/net/core/dev_addr_lists.c +++ b/net/core/dev_addr_lists.c @@ -16,6 +16,35 @@ * General list handling functions */ +static int __hw_addr_insert(struct netdev_hw_addr_list *list, + struct netdev_hw_addr *new, int addr_len) +{ + struct rb_node **ins_point = &list->tree.rb_node, *parent = NULL; + struct netdev_hw_addr *ha; + + while (*ins_point) { + int diff; + + ha = rb_entry(*ins_point, struct netdev_hw_addr, node); + diff = memcmp(new->addr, ha->addr, addr_len); + if (diff == 0) + diff = memcmp(&new->type, &ha->type, sizeof(new->type)); + + parent = *ins_point; + if (diff < 0) + ins_point = &parent->rb_left; + else if (diff > 0) + ins_point = &parent->rb_right; + else + return -EEXIST; + } + + rb_link_node_rcu(&new->node, parent, ins_point); + rb_insert_color(&new->node, &list->tree); + + return 0; +} + static struct netdev_hw_addr* __hw_addr_create(const unsigned char *addr, int addr_len, unsigned char addr_type, bool global, bool sync) @@ -50,11 +79,6 @@ static int __hw_addr_add_ex(struct netdev_hw_addr_list *list, if (addr_len > MAX_ADDR_LEN) return -EINVAL; - ha = list_first_entry(&list->list, struct netdev_hw_addr, list); - if (ha && !memcmp(addr, ha->addr, addr_len) && - (!addr_type || addr_type == ha->type)) - goto found_it; - while (*ins_point) { int diff; @@ -69,7 +93,6 @@ static int __hw_addr_add_ex(struct netdev_hw_addr_list *list, } else if (diff > 0) { ins_point = &parent->rb_right; } else { -found_it: if (exclusive) return -EEXIST; if (global) { @@ -94,16 +117,8 @@ found_it: if (!ha) return -ENOMEM; - /* The first address in dev->dev_addrs is pointed to by dev->dev_addr - * and mutated freely by device drivers and netdev ops, so if we insert - * it into the tree we'll end up with an invalid rbtree. - */ - if (list->count > 0) { - rb_link_node(&ha->node, parent, ins_point); - rb_insert_color(&ha->node, &list->tree); - } else { - RB_CLEAR_NODE(&ha->node); - } + rb_link_node(&ha->node, parent, ins_point); + rb_insert_color(&ha->node, &list->tree); list_add_tail_rcu(&ha->list, &list->list); list->count++; @@ -138,8 +153,7 @@ static int __hw_addr_del_entry(struct netdev_hw_addr_list *list, if (--ha->refcount) return 0; - if (!RB_EMPTY_NODE(&ha->node)) - rb_erase(&ha->node, &list->tree); + rb_erase(&ha->node, &list->tree); list_del_rcu(&ha->list); kfree_rcu(ha, rcu_head); @@ -151,18 +165,8 @@ static struct netdev_hw_addr *__hw_addr_lookup(struct netdev_hw_addr_list *list, const unsigned char *addr, int addr_len, unsigned char addr_type) { - struct netdev_hw_addr *ha; struct rb_node *node; - /* The first address isn't inserted into the tree because in the dev->dev_addrs - * list it's the address pointed to by dev->dev_addr which is freely mutated - * in place, so we need to check it separately. - */ - ha = list_first_entry(&list->list, struct netdev_hw_addr, list); - if (ha && !memcmp(addr, ha->addr, addr_len) && - (!addr_type || addr_type == ha->type)) - return ha; - node = list->tree.rb_node; while (node) { @@ -571,8 +575,10 @@ void dev_addr_mod(struct net_device *dev, unsigned int offset, dev_addr_check(dev); ha = container_of(dev->dev_addr, struct netdev_hw_addr, addr[0]); + rb_erase(&ha->node, &dev->dev_addrs.tree); memcpy(&ha->addr[offset], addr, len); memcpy(&dev->dev_addr_shadow[offset], addr, len); + WARN_ON(__hw_addr_insert(&dev->dev_addrs, ha, dev->addr_len)); } EXPORT_SYMBOL(dev_addr_mod); From 2c193f2cb11074bb7859605db3562ad653ab5ec7 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Nov 2021 06:21:55 -0800 Subject: [PATCH 0400/2048] net: kunit: add a test for dev_addr_lists Add a KUnit test for the dev_addr API. Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller --- net/Kconfig | 5 + net/core/Makefile | 2 + net/core/dev_addr_lists_test.c | 236 +++++++++++++++++++++++++++++++++ 3 files changed, 243 insertions(+) create mode 100644 net/core/dev_addr_lists_test.c diff --git a/net/Kconfig b/net/Kconfig index 074472dfa94a..8a1f9d0287de 100644 --- a/net/Kconfig +++ b/net/Kconfig @@ -455,4 +455,9 @@ config ETHTOOL_NETLINK netlink. It provides better extensibility and some new features, e.g. notification messages. +config NETDEV_ADDR_LIST_TEST + tristate "Unit tests for device address list" + default KUNIT_ALL_TESTS + depends on KUNIT + endif # if NET diff --git a/net/core/Makefile b/net/core/Makefile index 6bdcb2cafed8..a8e4f737692b 100644 --- a/net/core/Makefile +++ b/net/core/Makefile @@ -13,6 +13,8 @@ obj-y += dev.o dev_addr_lists.o dst.o netevent.o \ sock_diag.o dev_ioctl.o tso.o sock_reuseport.o \ fib_notifier.o xdp.o flow_offload.o gro.o +obj-$(CONFIG_NETDEV_ADDR_LIST_TEST) += dev_addr_lists_test.o + obj-y += net-sysfs.o obj-$(CONFIG_PAGE_POOL) += page_pool.o obj-$(CONFIG_PROC_FS) += net-procfs.o diff --git a/net/core/dev_addr_lists_test.c b/net/core/dev_addr_lists_test.c new file mode 100644 index 000000000000..049cfbc58aa9 --- /dev/null +++ b/net/core/dev_addr_lists_test.c @@ -0,0 +1,236 @@ +// SPDX-License-Identifier: GPL-2.0-or-later + +#include +#include +#include +#include + +static const struct net_device_ops dummy_netdev_ops = { +}; + +struct dev_addr_test_priv { + u32 addr_seen; +}; + +static int dev_addr_test_sync(struct net_device *netdev, const unsigned char *a) +{ + struct dev_addr_test_priv *datp = netdev_priv(netdev); + + if (a[0] < 31 && !memchr_inv(a, a[0], ETH_ALEN)) + datp->addr_seen |= 1 << a[0]; + return 0; +} + +static int dev_addr_test_unsync(struct net_device *netdev, + const unsigned char *a) +{ + struct dev_addr_test_priv *datp = netdev_priv(netdev); + + if (a[0] < 31 && !memchr_inv(a, a[0], ETH_ALEN)) + datp->addr_seen &= ~(1 << a[0]); + return 0; +} + +static int dev_addr_test_init(struct kunit *test) +{ + struct dev_addr_test_priv *datp; + struct net_device *netdev; + int err; + + netdev = alloc_etherdev(sizeof(*datp)); + KUNIT_ASSERT_TRUE(test, !!netdev); + + test->priv = netdev; + netdev->netdev_ops = &dummy_netdev_ops; + + err = register_netdev(netdev); + if (err) { + free_netdev(netdev); + KUNIT_FAIL(test, "Can't register netdev %d", err); + } + + rtnl_lock(); + return 0; +} + +static void dev_addr_test_exit(struct kunit *test) +{ + struct net_device *netdev = test->priv; + + rtnl_unlock(); + unregister_netdev(netdev); + free_netdev(netdev); +} + +static void dev_addr_test_basic(struct kunit *test) +{ + struct net_device *netdev = test->priv; + u8 addr[ETH_ALEN]; + + KUNIT_EXPECT_TRUE(test, !!netdev->dev_addr); + + memset(addr, 2, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + KUNIT_EXPECT_EQ(test, 0, memcmp(netdev->dev_addr, addr, sizeof(addr))); + + memset(addr, 3, sizeof(addr)); + dev_addr_set(netdev, addr); + KUNIT_EXPECT_EQ(test, 0, memcmp(netdev->dev_addr, addr, sizeof(addr))); +} + +static void dev_addr_test_sync_one(struct kunit *test) +{ + struct net_device *netdev = test->priv; + struct dev_addr_test_priv *datp; + u8 addr[ETH_ALEN]; + + datp = netdev_priv(netdev); + + memset(addr, 1, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + KUNIT_EXPECT_EQ(test, 2, datp->addr_seen); + + memset(addr, 2, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + + datp->addr_seen = 0; + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + /* It's not going to sync anything because the main address is + * considered synced and we overwrite in place. + */ + KUNIT_EXPECT_EQ(test, 0, datp->addr_seen); +} + +static void dev_addr_test_add_del(struct kunit *test) +{ + struct net_device *netdev = test->priv; + struct dev_addr_test_priv *datp; + u8 addr[ETH_ALEN]; + int i; + + datp = netdev_priv(netdev); + + for (i = 1; i < 4; i++) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, 0, dev_addr_add(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + } + /* Add 3 again */ + KUNIT_EXPECT_EQ(test, 0, dev_addr_add(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + KUNIT_EXPECT_EQ(test, 0xf, datp->addr_seen); + + KUNIT_EXPECT_EQ(test, 0, dev_addr_del(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + KUNIT_EXPECT_EQ(test, 0xf, datp->addr_seen); + + for (i = 1; i < 4; i++) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, 0, dev_addr_del(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + } + + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + KUNIT_EXPECT_EQ(test, 1, datp->addr_seen); +} + +static void dev_addr_test_del_main(struct kunit *test) +{ + struct net_device *netdev = test->priv; + u8 addr[ETH_ALEN]; + + memset(addr, 1, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + + KUNIT_EXPECT_EQ(test, -ENOENT, dev_addr_del(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + KUNIT_EXPECT_EQ(test, 0, dev_addr_add(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + KUNIT_EXPECT_EQ(test, 0, dev_addr_del(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + KUNIT_EXPECT_EQ(test, -ENOENT, dev_addr_del(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); +} + +static void dev_addr_test_add_set(struct kunit *test) +{ + struct net_device *netdev = test->priv; + struct dev_addr_test_priv *datp; + u8 addr[ETH_ALEN]; + int i; + + datp = netdev_priv(netdev); + + /* There is no external API like dev_addr_add_excl(), + * so shuffle the tree a little bit and exploit aliasing. + */ + for (i = 1; i < 16; i++) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, 0, dev_addr_add(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + } + + memset(addr, i, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + KUNIT_EXPECT_EQ(test, 0, dev_addr_add(netdev, addr, + NETDEV_HW_ADDR_T_LAN)); + memset(addr, 0, sizeof(addr)); + eth_hw_addr_set(netdev, addr); + + __hw_addr_sync_dev(&netdev->dev_addrs, netdev, dev_addr_test_sync, + dev_addr_test_unsync); + KUNIT_EXPECT_EQ(test, 0xffff, datp->addr_seen); +} + +static void dev_addr_test_add_excl(struct kunit *test) +{ + struct net_device *netdev = test->priv; + u8 addr[ETH_ALEN]; + int i; + + for (i = 0; i < 10; i++) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, 0, dev_uc_add_excl(netdev, addr)); + } + KUNIT_EXPECT_EQ(test, -EEXIST, dev_uc_add_excl(netdev, addr)); + + for (i = 0; i < 10; i += 2) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, 0, dev_uc_del(netdev, addr)); + } + for (i = 1; i < 10; i += 2) { + memset(addr, i, sizeof(addr)); + KUNIT_EXPECT_EQ(test, -EEXIST, dev_uc_add_excl(netdev, addr)); + } +} + +static struct kunit_case dev_addr_test_cases[] = { + KUNIT_CASE(dev_addr_test_basic), + KUNIT_CASE(dev_addr_test_sync_one), + KUNIT_CASE(dev_addr_test_add_del), + KUNIT_CASE(dev_addr_test_del_main), + KUNIT_CASE(dev_addr_test_add_set), + KUNIT_CASE(dev_addr_test_add_excl), + {} +}; + +static struct kunit_suite dev_addr_test_suite = { + .name = "dev-addr-list-test", + .test_cases = dev_addr_test_cases, + .init = dev_addr_test_init, + .exit = dev_addr_test_exit, +}; +kunit_test_suite(dev_addr_test_suite); + +MODULE_LICENSE("GPL"); From 4f47d5d507d6f211ebceac76a5f0b83c2eae154b Mon Sep 17 00:00:00 2001 From: Poorva Sonparote Date: Fri, 19 Nov 2021 12:41:34 -0800 Subject: [PATCH 0401/2048] ipv4: Exposing __ip_sock_set_tos() in ip.h Making the static function __ip_sock_set_tos() from net/ipv4/ip_sockglue.c accessible by declaring it in include/net/ip.h The reason for doing this is to use this function to set IP_TOS value in mptcp_setsockopt() without the lock. Signed-off-by: Poorva Sonparote Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- include/net/ip.h | 1 + net/ipv4/ip_sockglue.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/include/net/ip.h b/include/net/ip.h index 7d1088888c10..81e23a102a0d 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -783,5 +783,6 @@ int ip_sock_set_mtu_discover(struct sock *sk, int val); void ip_sock_set_pktinfo(struct sock *sk); void ip_sock_set_recverr(struct sock *sk); void ip_sock_set_tos(struct sock *sk, int val); +void __ip_sock_set_tos(struct sock *sk, int val); #endif /* _IP_H */ diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index 38d29b175ca6..445a9ecaefa1 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -576,7 +576,7 @@ out: return err; } -static void __ip_sock_set_tos(struct sock *sk, int val) +void __ip_sock_set_tos(struct sock *sk, int val) { if (sk->sk_type == SOCK_STREAM) { val &= ~INET_ECN_MASK; From ffcacff87cd65dc09b97982acf3a771713983a28 Mon Sep 17 00:00:00 2001 From: Poorva Sonparote Date: Fri, 19 Nov 2021 12:41:35 -0800 Subject: [PATCH 0402/2048] mptcp: Support for IP_TOS for MPTCP setsockopt() SOL_IP provides a way to configure network layer attributes in a socket. This patch adds support for IP_TOS for setsockopt(.. ,SOL_IP, ..) Support for SOL_IP is added in mptcp_setsockopt() and IP_TOS is handled in a private function. The idea here is to take in the value passed for IP_TOS and set it to the current subflow, open subflows as well new subflows that might be created after the initial call to setsockopt(). This sync is done using sync_socket_options(.., ssk) and setting the value of tos using __ip_sock_set_tos(ssk,..). The patch has been tested using the packetdrill script here - https://github.com/multipath-tcp/mptcp_net-next/issues/220#issuecomment-947863717 Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/220 Signed-off-by: Poorva Sonparote Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- net/mptcp/sockopt.c | 40 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 40 insertions(+) diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c index 0f1e661c2032..b452571e8d9b 100644 --- a/net/mptcp/sockopt.c +++ b/net/mptcp/sockopt.c @@ -598,6 +598,42 @@ static int mptcp_setsockopt_sol_tcp_congestion(struct mptcp_sock *msk, sockptr_t return ret; } +static int mptcp_setsockopt_v4_set_tos(struct mptcp_sock *msk, int optname, + sockptr_t optval, unsigned int optlen) +{ + struct mptcp_subflow_context *subflow; + struct sock *sk = (struct sock *)msk; + int err, val; + + err = ip_setsockopt(sk, SOL_IP, optname, optval, optlen); + + if (err != 0) + return err; + + lock_sock(sk); + sockopt_seq_inc(msk); + val = inet_sk(sk)->tos; + mptcp_for_each_subflow(msk, subflow) { + struct sock *ssk = mptcp_subflow_tcp_sock(subflow); + + __ip_sock_set_tos(ssk, val); + } + release_sock(sk); + + return err; +} + +static int mptcp_setsockopt_v4(struct mptcp_sock *msk, int optname, + sockptr_t optval, unsigned int optlen) +{ + switch (optname) { + case IP_TOS: + return mptcp_setsockopt_v4_set_tos(msk, optname, optval, optlen); + } + + return -EOPNOTSUPP; +} + static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname, sockptr_t optval, unsigned int optlen) { @@ -637,6 +673,9 @@ int mptcp_setsockopt(struct sock *sk, int level, int optname, if (ssk) return tcp_setsockopt(ssk, level, optname, optval, optlen); + if (level == SOL_IP) + return mptcp_setsockopt_v4(msk, optname, optval, optlen); + if (level == SOL_IPV6) return mptcp_setsockopt_v6(msk, optname, optval, optlen); @@ -1003,6 +1042,7 @@ static void sync_socket_options(struct mptcp_sock *msk, struct sock *ssk) ssk->sk_priority = sk->sk_priority; ssk->sk_bound_dev_if = sk->sk_bound_dev_if; ssk->sk_incoming_cpu = sk->sk_incoming_cpu; + __ip_sock_set_tos(ssk, inet_sk(sk)->tos); if (sk->sk_userlocks & tx_rx_locks) { ssk->sk_userlocks |= sk->sk_userlocks & tx_rx_locks; From c9406a23c1161c3a9a672d055ec840b00d730a6f Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Fri, 19 Nov 2021 12:41:36 -0800 Subject: [PATCH 0403/2048] mptcp: sockopt: add SOL_IP freebind & transparent options These options also need to be set before bind, so do the sync of msk to new ssk socket a bit earlier. Signed-off-by: Florian Westphal Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- net/mptcp/sockopt.c | 66 ++++++++++++++++++++++++++++++++++++++++++++- net/mptcp/subflow.c | 3 ++- 2 files changed, 67 insertions(+), 2 deletions(-) diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c index b452571e8d9b..fb43e145cb57 100644 --- a/net/mptcp/sockopt.c +++ b/net/mptcp/sockopt.c @@ -390,6 +390,8 @@ static int mptcp_setsockopt_v6(struct mptcp_sock *msk, int optname, switch (optname) { case IPV6_V6ONLY: + case IPV6_TRANSPARENT: + case IPV6_FREEBIND: lock_sock(sk); ssock = __mptcp_nmpc_socket(msk); if (!ssock) { @@ -398,8 +400,24 @@ static int mptcp_setsockopt_v6(struct mptcp_sock *msk, int optname, } ret = tcp_setsockopt(ssock->sk, SOL_IPV6, optname, optval, optlen); - if (ret == 0) + if (ret != 0) { + release_sock(sk); + return ret; + } + + sockopt_seq_inc(msk); + + switch (optname) { + case IPV6_V6ONLY: sk->sk_ipv6only = ssock->sk->sk_ipv6only; + break; + case IPV6_TRANSPARENT: + inet_sk(sk)->transparent = inet_sk(ssock->sk)->transparent; + break; + case IPV6_FREEBIND: + inet_sk(sk)->freebind = inet_sk(ssock->sk)->freebind; + break; + } release_sock(sk); break; @@ -598,6 +616,46 @@ static int mptcp_setsockopt_sol_tcp_congestion(struct mptcp_sock *msk, sockptr_t return ret; } +static int mptcp_setsockopt_sol_ip_set_transparent(struct mptcp_sock *msk, int optname, + sockptr_t optval, unsigned int optlen) +{ + struct sock *sk = (struct sock *)msk; + struct inet_sock *issk; + struct socket *ssock; + int err; + + err = ip_setsockopt(sk, SOL_IP, optname, optval, optlen); + if (err != 0) + return err; + + lock_sock(sk); + + ssock = __mptcp_nmpc_socket(msk); + if (!ssock) { + release_sock(sk); + return -EINVAL; + } + + issk = inet_sk(ssock->sk); + + switch (optname) { + case IP_FREEBIND: + issk->freebind = inet_sk(sk)->freebind; + break; + case IP_TRANSPARENT: + issk->transparent = inet_sk(sk)->transparent; + break; + default: + release_sock(sk); + WARN_ON_ONCE(1); + return -EOPNOTSUPP; + } + + sockopt_seq_inc(msk); + release_sock(sk); + return 0; +} + static int mptcp_setsockopt_v4_set_tos(struct mptcp_sock *msk, int optname, sockptr_t optval, unsigned int optlen) { @@ -627,6 +685,9 @@ static int mptcp_setsockopt_v4(struct mptcp_sock *msk, int optname, sockptr_t optval, unsigned int optlen) { switch (optname) { + case IP_FREEBIND: + case IP_TRANSPARENT: + return mptcp_setsockopt_sol_ip_set_transparent(msk, optname, optval, optlen); case IP_TOS: return mptcp_setsockopt_v4_set_tos(msk, optname, optval, optlen); } @@ -1068,6 +1129,9 @@ static void sync_socket_options(struct mptcp_sock *msk, struct sock *ssk) if (inet_csk(sk)->icsk_ca_ops != inet_csk(ssk)->icsk_ca_ops) tcp_set_congestion_control(ssk, msk->ca_name, false, true); + + inet_sk(ssk)->transparent = inet_sk(sk)->transparent; + inet_sk(ssk)->freebind = inet_sk(sk)->freebind; } static void __mptcp_sockopt_sync(struct mptcp_sock *msk, struct sock *ssk) diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 49787a1d7b34..b8dd3441f7d0 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1425,6 +1425,8 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc, if (addr.ss_family == AF_INET6) addrlen = sizeof(struct sockaddr_in6); #endif + mptcp_sockopt_sync(msk, ssk); + ssk->sk_bound_dev_if = ifindex; err = kernel_bind(sf, (struct sockaddr *)&addr, addrlen); if (err) @@ -1441,7 +1443,6 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc, mptcp_info2sockaddr(remote, &addr, ssk->sk_family); mptcp_add_pending_subflow(msk, subflow); - mptcp_sockopt_sync(msk, ssk); err = kernel_connect(sf, (struct sockaddr *)&addr, addrlen, O_NONBLOCK); if (err && err != -EINPROGRESS) goto failed_unlink; From 5fb62e9cd3adadd95303447ee8e3f62ee98b0e73 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Fri, 19 Nov 2021 12:41:37 -0800 Subject: [PATCH 0404/2048] selftests: mptcp: add tproxy test case No hard dependencies here, just skip if test environ lacks nft binary or the needed kernel config options. The test case spawns listener in ns2 but ns1 will connect to the ip address of ns4. policy routing + tproxy rule will redirect packets to ns2 instead of forward. v3: - update mptcp/config (Mat Martineau) - more verbose SKIP messages in mptcp_connect.sh Signed-off-by: Florian Westphal Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- tools/testing/selftests/net/mptcp/config | 8 +- .../selftests/net/mptcp/mptcp_connect.c | 51 +++++++++++- .../selftests/net/mptcp/mptcp_connect.sh | 80 +++++++++++++++++++ 3 files changed, 136 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/net/mptcp/config b/tools/testing/selftests/net/mptcp/config index 0faaccd21447..419e71560fd1 100644 --- a/tools/testing/selftests/net/mptcp/config +++ b/tools/testing/selftests/net/mptcp/config @@ -13,5 +13,9 @@ CONFIG_NFT_COUNTER=m CONFIG_NFT_COMPAT=m CONFIG_NETFILTER_XTABLES=m CONFIG_NETFILTER_XT_MATCH_BPF=m -CONFIG_NF_TABLES_IPV4=y -CONFIG_NF_TABLES_IPV6=y +CONFIG_NF_TABLES_INET=y +CONFIG_NFT_TPROXY=m +CONFIG_NFT_SOCKET=m +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_IP_MULTIPLE_TABLES=y +CONFIG_IPV6_MULTIPLE_TABLES=y diff --git a/tools/testing/selftests/net/mptcp/mptcp_connect.c b/tools/testing/selftests/net/mptcp/mptcp_connect.c index 95e81d557b08..ada9b80774d4 100644 --- a/tools/testing/selftests/net/mptcp/mptcp_connect.c +++ b/tools/testing/selftests/net/mptcp/mptcp_connect.c @@ -75,7 +75,12 @@ struct cfg_cmsg_types { unsigned int timestampns:1; }; +struct cfg_sockopt_types { + unsigned int transparent:1; +}; + static struct cfg_cmsg_types cfg_cmsg_types; +static struct cfg_sockopt_types cfg_sockopt_types; static void die_usage(void) { @@ -93,6 +98,7 @@ static void die_usage(void) fprintf(stderr, "\t-u -- check mptcp ulp\n"); fprintf(stderr, "\t-w num -- wait num sec before closing the socket\n"); fprintf(stderr, "\t-c cmsg -- test cmsg type \n"); + fprintf(stderr, "\t-o option -- test sockopt