From bb49c6fa8b845591b317b0d7afea4ae60ec7f3aa Mon Sep 17 00:00:00 2001
From: Stefano Garzarella <sgarzare@redhat.com>
Date: Fri, 11 Feb 2022 10:01:36 +0100
Subject: [PATCH 1/4] block: clear iocb->private in blkdev_bio_end_io_async()
iocb_bio_iopoll() expects iocb->private to be cleared before
releasing the bio.
We already do this in blkdev_bio_end_io(), but we forgot in the
recently added blkdev_bio_end_io_async().
Fixes: 54a88eb838d3 ("block: add single bio async direct IO helper")
Cc: asml.silence@gmail.com
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220211090136.44471-1-sgarzare@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
block/fops.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/block/fops.c b/block/fops.c
index 4f59e0f5bf30..a18e7fbd97b8 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -289,6 +289,8 @@ static void blkdev_bio_end_io_async(struct bio *bio)
struct kiocb *iocb = dio->iocb;
ssize_t ret;
+ WRITE_ONCE(iocb->private, NULL);
+
if (likely(!bio->bi_status)) {
ret = dio->size;
iocb->ki_pos += ret;
From 363f6368603743072e5f318c668c632bccb097a3 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Wed, 16 Feb 2022 15:07:15 +0100
Subject: [PATCH 2/4] nvme: don't return an error from nvme_configure_metadata
When a fabrics controller claims to support an invalidate metadata
configuration we already warn and disable metadata support. No need to
also return an error during revalidation.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Daniel Wagner <dwagner@suse.de>
Tested-by: Kanchan Joshi <joshi.k@samsung.com>
---
drivers/nvme/host/core.c | 13 +++++--------
1 file changed, 5 insertions(+), 8 deletions(-)
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 469f23186159..5fc040b279bf 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -1723,7 +1723,7 @@ static int nvme_setup_streams_ns(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
return 0;
}
-static int nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
+static void nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
{
struct nvme_ctrl *ctrl = ns->ctrl;
@@ -1739,7 +1739,8 @@ static int nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
ns->features &= ~(NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS);
if (!ns->ms || !(ctrl->ops->flags & NVME_F_METADATA_SUPPORTED))
- return 0;
+ return;
+
if (ctrl->ops->flags & NVME_F_FABRICS) {
/*
* The NVMe over Fabrics specification only supports metadata as
@@ -1747,7 +1748,7 @@ static int nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
* remap the separate metadata buffer from the block layer.
*/
if (WARN_ON_ONCE(!(id->flbas & NVME_NS_FLBAS_META_EXT)))
- return -EINVAL;
+ return;
ns->features |= NVME_NS_EXT_LBAS;
@@ -1774,8 +1775,6 @@ static int nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
else
ns->features |= NVME_NS_METADATA_SUPPORTED;
}
-
- return 0;
}
static void nvme_set_queue_limits(struct nvme_ctrl *ctrl,
@@ -1916,9 +1915,7 @@ static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
ns->lba_shift = id->lbaf[lbaf].ds;
nvme_set_queue_limits(ns->ctrl, ns->queue);
- ret = nvme_configure_metadata(ns, id);
- if (ret)
- goto out_unfreeze;
+ nvme_configure_metadata(ns, id);
nvme_set_chunk_sectors(ns, id);
nvme_update_disk_info(ns->disk, ns, id);
From 602e57c9799c19f27e440639deed3ec45cfe1651 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Wed, 16 Feb 2022 14:14:58 +0100
Subject: [PATCH 3/4] nvme: also mark passthrough-only namespaces ready in
nvme_update_ns_info
Commit e7d65803e2bb ("nvme-multipath: revalidate paths during rescan")
introduced the NVME_NS_READY flag, which nvme_path_is_disabled() uses
to check if a path can be used or not. We also need to set this flag
for devices that fail the ZNS feature validation and which are available
through passthrough devices only to that they can be used in multipathing
setups.
Fixes: e7d65803e2bb ("nvme-multipath: revalidate paths during rescan")
Reported-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Daniel Wagner <dwagner@suse.de>
Tested-by: Kanchan Joshi <joshi.k@samsung.com>
---
drivers/nvme/host/core.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 5fc040b279bf..fd4720d37cc0 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -1931,7 +1931,7 @@ static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
if (blk_queue_is_zoned(ns->queue)) {
ret = nvme_revalidate_zones(ns);
if (ret && !nvme_first_scan(ns->disk))
- goto out;
+ return ret;
}
if (nvme_ns_head_multipath(ns->head)) {
@@ -1946,16 +1946,16 @@ static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
return 0;
out_unfreeze:
- blk_mq_unfreeze_queue(ns->disk->queue);
-out:
/*
* If probing fails due an unsupported feature, hide the block device,
* but still allow other access.
*/
if (ret == -ENODEV) {
ns->disk->flags |= GENHD_FL_HIDDEN;
+ set_bit(NVME_NS_READY, &ns->flags);
ret = 0;
}
+ blk_mq_unfreeze_queue(ns->disk->queue);
return ret;
}
From c2700d2886a87f83f31e0a301de1d2350b52c79b Mon Sep 17 00:00:00 2001
From: Varun Prakash <varun@chelsio.com>
Date: Sat, 22 Jan 2022 22:27:44 +0530
Subject: [PATCH 4/4] nvme-tcp: send H2CData PDUs based on MAXH2CDATA
As per NVMe/TCP specification (revision 1.0a, section 3.6.2.3)
Maximum Host to Controller Data length (MAXH2CDATA): Specifies the
maximum number of PDU-Data bytes per H2CData PDU in bytes. This value
is a multiple of dwords and should be no less than 4,096.
Current code sets H2CData PDU data_length to r2t_length,
it does not check MAXH2CDATA value. Fix this by setting H2CData PDU
data_length to min(req->h2cdata_left, queue->maxh2cdata).
Also validate MAXH2CDATA value returned by target in ICResp PDU,
if it is not a multiple of dword or if it is less than 4096 return
-EINVAL from nvme_tcp_init_connection().
Signed-off-by: Varun Prakash <varun@chelsio.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/nvme/host/tcp.c | 63 +++++++++++++++++++++++++++++++---------
include/linux/nvme-tcp.h | 1 +
2 files changed, 50 insertions(+), 14 deletions(-)
diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index 891a36d02e7c..65e00c64a588 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -44,6 +44,8 @@ struct nvme_tcp_request {
u32 data_len;
u32 pdu_len;
u32 pdu_sent;
+ u32 h2cdata_left;
+ u32 h2cdata_offset;
u16 ttag;
__le16 status;
struct list_head entry;
@@ -95,6 +97,7 @@ struct nvme_tcp_queue {
struct nvme_tcp_request *request;
int queue_size;
+ u32 maxh2cdata;
size_t cmnd_capsule_len;
struct nvme_tcp_ctrl *ctrl;
unsigned long flags;
@@ -572,23 +575,26 @@ static int nvme_tcp_handle_comp(struct nvme_tcp_queue *queue,
return ret;
}
-static void nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
- struct nvme_tcp_r2t_pdu *pdu)
+static void nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req)
{
struct nvme_tcp_data_pdu *data = req->pdu;
struct nvme_tcp_queue *queue = req->queue;
struct request *rq = blk_mq_rq_from_pdu(req);
+ u32 h2cdata_sent = req->pdu_len;
u8 hdgst = nvme_tcp_hdgst_len(queue);
u8 ddgst = nvme_tcp_ddgst_len(queue);
req->state = NVME_TCP_SEND_H2C_PDU;
req->offset = 0;
- req->pdu_len = le32_to_cpu(pdu->r2t_length);
+ req->pdu_len = min(req->h2cdata_left, queue->maxh2cdata);
req->pdu_sent = 0;
+ req->h2cdata_left -= req->pdu_len;
+ req->h2cdata_offset += h2cdata_sent;
memset(data, 0, sizeof(*data));
data->hdr.type = nvme_tcp_h2c_data;
- data->hdr.flags = NVME_TCP_F_DATA_LAST;
+ if (!req->h2cdata_left)
+ data->hdr.flags = NVME_TCP_F_DATA_LAST;
if (queue->hdr_digest)
data->hdr.flags |= NVME_TCP_F_HDGST;
if (queue->data_digest)
@@ -597,9 +603,9 @@ static void nvme_tcp_setup_h2c_data_pdu(struct nvme_tcp_request *req,
data->hdr.pdo = data->hdr.hlen + hdgst;
data->hdr.plen =
cpu_to_le32(data->hdr.hlen + hdgst + req->pdu_len + ddgst);
- data->ttag = pdu->ttag;
+ data->ttag = req->ttag;
data->command_id = nvme_cid(rq);
- data->data_offset = pdu->r2t_offset;
+ data->data_offset = cpu_to_le32(req->h2cdata_offset);
data->data_length = cpu_to_le32(req->pdu_len);
}
@@ -609,6 +615,7 @@ static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
struct nvme_tcp_request *req;
struct request *rq;
u32 r2t_length = le32_to_cpu(pdu->r2t_length);
+ u32 r2t_offset = le32_to_cpu(pdu->r2t_offset);
rq = nvme_find_rq(nvme_tcp_tagset(queue), pdu->command_id);
if (!rq) {
@@ -633,14 +640,19 @@ static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
return -EPROTO;
}
- if (unlikely(le32_to_cpu(pdu->r2t_offset) < req->data_sent)) {
+ if (unlikely(r2t_offset < req->data_sent)) {
dev_err(queue->ctrl->ctrl.device,
"req %d unexpected r2t offset %u (expected %zu)\n",
- rq->tag, le32_to_cpu(pdu->r2t_offset), req->data_sent);
+ rq->tag, r2t_offset, req->data_sent);
return -EPROTO;
}
- nvme_tcp_setup_h2c_data_pdu(req, pdu);
+ req->pdu_len = 0;
+ req->h2cdata_left = r2t_length;
+ req->h2cdata_offset = r2t_offset;
+ req->ttag = pdu->ttag;
+
+ nvme_tcp_setup_h2c_data_pdu(req);
nvme_tcp_queue_request(req, false, true);
return 0;
@@ -928,6 +940,7 @@ static int nvme_tcp_try_send_data(struct nvme_tcp_request *req)
{
struct nvme_tcp_queue *queue = req->queue;
int req_data_len = req->data_len;
+ u32 h2cdata_left = req->h2cdata_left;
while (true) {
struct page *page = nvme_tcp_req_cur_page(req);
@@ -972,7 +985,10 @@ static int nvme_tcp_try_send_data(struct nvme_tcp_request *req)
req->state = NVME_TCP_SEND_DDGST;
req->offset = 0;
} else {
- nvme_tcp_done_send_req(queue);
+ if (h2cdata_left)
+ nvme_tcp_setup_h2c_data_pdu(req);
+ else
+ nvme_tcp_done_send_req(queue);
}
return 1;
}
@@ -1030,9 +1046,14 @@ static int nvme_tcp_try_send_data_pdu(struct nvme_tcp_request *req)
if (queue->hdr_digest && !req->offset)
nvme_tcp_hdgst(queue->snd_hash, pdu, sizeof(*pdu));
- ret = kernel_sendpage(queue->sock, virt_to_page(pdu),
- offset_in_page(pdu) + req->offset, len,
- MSG_DONTWAIT | MSG_MORE | MSG_SENDPAGE_NOTLAST);
+ if (!req->h2cdata_left)
+ ret = kernel_sendpage(queue->sock, virt_to_page(pdu),
+ offset_in_page(pdu) + req->offset, len,
+ MSG_DONTWAIT | MSG_MORE | MSG_SENDPAGE_NOTLAST);
+ else
+ ret = sock_no_sendpage(queue->sock, virt_to_page(pdu),
+ offset_in_page(pdu) + req->offset, len,
+ MSG_DONTWAIT | MSG_MORE);
if (unlikely(ret <= 0))
return ret;
@@ -1052,6 +1073,7 @@ static int nvme_tcp_try_send_ddgst(struct nvme_tcp_request *req)
{
struct nvme_tcp_queue *queue = req->queue;
size_t offset = req->offset;
+ u32 h2cdata_left = req->h2cdata_left;
int ret;
struct msghdr msg = { .msg_flags = MSG_DONTWAIT };
struct kvec iov = {
@@ -1069,7 +1091,10 @@ static int nvme_tcp_try_send_ddgst(struct nvme_tcp_request *req)
return ret;
if (offset + ret == NVME_TCP_DIGEST_LENGTH) {
- nvme_tcp_done_send_req(queue);
+ if (h2cdata_left)
+ nvme_tcp_setup_h2c_data_pdu(req);
+ else
+ nvme_tcp_done_send_req(queue);
return 1;
}
@@ -1261,6 +1286,7 @@ static int nvme_tcp_init_connection(struct nvme_tcp_queue *queue)
struct msghdr msg = {};
struct kvec iov;
bool ctrl_hdgst, ctrl_ddgst;
+ u32 maxh2cdata;
int ret;
icreq = kzalloc(sizeof(*icreq), GFP_KERNEL);
@@ -1344,6 +1370,14 @@ static int nvme_tcp_init_connection(struct nvme_tcp_queue *queue)
goto free_icresp;
}
+ maxh2cdata = le32_to_cpu(icresp->maxdata);
+ if ((maxh2cdata % 4) || (maxh2cdata < NVME_TCP_MIN_MAXH2CDATA)) {
+ pr_err("queue %d: invalid maxh2cdata returned %u\n",
+ nvme_tcp_queue_id(queue), maxh2cdata);
+ goto free_icresp;
+ }
+ queue->maxh2cdata = maxh2cdata;
+
ret = 0;
free_icresp:
kfree(icresp);
@@ -2329,6 +2363,7 @@ static blk_status_t nvme_tcp_setup_cmd_pdu(struct nvme_ns *ns,
req->data_sent = 0;
req->pdu_len = 0;
req->pdu_sent = 0;
+ req->h2cdata_left = 0;
req->data_len = blk_rq_nr_phys_segments(rq) ?
blk_rq_payload_bytes(rq) : 0;
req->curr_bio = rq->bio;
diff --git a/include/linux/nvme-tcp.h b/include/linux/nvme-tcp.h
index 959e0bd9a913..75470159a194 100644
--- a/include/linux/nvme-tcp.h
+++ b/include/linux/nvme-tcp.h
@@ -12,6 +12,7 @@
#define NVME_TCP_DISC_PORT 8009
#define NVME_TCP_ADMIN_CCSZ SZ_8K
#define NVME_TCP_DIGEST_LENGTH 4
+#define NVME_TCP_MIN_MAXH2CDATA 4096
enum nvme_tcp_pfv {
NVME_TCP_PFV_1_0 = 0x0,