Commit 406f42fa0d ("net-next: When a bond have a massive amount
of VLANs...") introduced a rbtree for faster Ethernet address look
up. To maintain netdev->dev_addr in this tree we need to make all
the writes to it got through appropriate helpers.
Convert wireless from ether_addr_copy() to eth_hw_addr_set():
@@
expression dev, np;
@@
- ether_addr_copy(dev->dev_addr, np)
+ eth_hw_addr_set(dev, np)
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20211018235021.1279697-3-kuba@kernel.org
When powersaving (so either wifi powersaving or deep sleep, depending on
which state the firmware is in) is disabled, the way the firmware goes
into host sleep is different: Usually the firmware implicitely enters
host sleep on the next SLEEP event we get when we configured host sleep
via HSCFG before. When powersaving is disabled though, there are no
SLEEP events, the way we enter host sleep in that case is different: The
firmware will send us a HS_ACT_REQ event and after that we "manually"
make the firmware enter host sleep by sending it another HSCFG command
with the action HS_ACTIVATE.
Now waking up from host sleep appears to be different depending on
whether powersaving is enabled again: When powersaving is enabled, the
firmware implicitely leaves host sleep as soon as it wakes up and sends
us an AWAKE event. When powersaving is disabled though, it apparently
doesn't implicitely leave host sleep, but instead we need to send it a
HSCFG command with the HS_CONFIGURE action and the HS_CFG_CANCEL
condition. We didn't do that so far, which is why waking up from host
sleep was broken when powersaving is disabled.
So add some additional state to mwifiex_adapter where we keep track of
whether host sleep was activated manually via HS_ACTIVATE, and if that
was the case, deactivate it manually again via HS_CFG_CANCEL.
Signed-off-by: Jonas Dreßler <verdre@v0yd.nl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20211016153244.24353-6-verdre@v0yd.nl
'destroy_workqueue()' already drains the queue before destroying it, so
there is no need to flush it explicitly.
Remove the redundant 'flush_workqueue()' calls.
This was generated with coccinelle:
@@
expression E;
@@
- flush_workqueue(E);
destroy_workqueue(E);
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/0855d51423578ad019c0264dad3fe47a2e8af9c7.1633849511.git.christophe.jaillet@wanadoo.fr
We can deadlock when rmmod'ing the driver or going through firmware
reset, because the cfg80211_unregister_wdev() has to bring down the link
for us, ... which then grab the same wiphy lock.
nl80211_del_interface() already handles a very similar case, with a nice
description:
/*
* We hold RTNL, so this is safe, without RTNL opencount cannot
* reach 0, and thus the rdev cannot be deleted.
*
* We need to do it for the dev_close(), since that will call
* the netdev notifiers, and we need to acquire the mutex there
* but don't know if we get there from here or from some other
* place (e.g. "ip link set ... down").
*/
mutex_unlock(&rdev->wiphy.mtx);
...
Do similarly for mwifiex teardown, by ensuring we bring the link down
first.
Sample deadlock trace:
[ 247.103516] INFO: task rmmod:2119 blocked for more than 123 seconds.
[ 247.110630] Not tainted 5.12.4 #5
[ 247.115796] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 247.124557] task:rmmod state:D stack: 0 pid: 2119 ppid: 2114 flags:0x00400208
[ 247.133905] Call trace:
[ 247.136644] __switch_to+0x130/0x170
[ 247.140643] __schedule+0x714/0xa0c
[ 247.144548] schedule_preempt_disabled+0x88/0xf4
[ 247.149714] __mutex_lock_common+0x43c/0x750
[ 247.154496] mutex_lock_nested+0x5c/0x68
[ 247.158884] cfg80211_netdev_notifier_call+0x280/0x4e0 [cfg80211]
[ 247.165769] raw_notifier_call_chain+0x4c/0x78
[ 247.170742] call_netdevice_notifiers_info+0x68/0xa4
[ 247.176305] __dev_close_many+0x7c/0x138
[ 247.180693] dev_close_many+0x7c/0x10c
[ 247.184893] unregister_netdevice_many+0xfc/0x654
[ 247.190158] unregister_netdevice_queue+0xb4/0xe0
[ 247.195424] _cfg80211_unregister_wdev+0xa4/0x204 [cfg80211]
[ 247.201816] cfg80211_unregister_wdev+0x20/0x2c [cfg80211]
[ 247.208016] mwifiex_del_virtual_intf+0xc8/0x188 [mwifiex]
[ 247.214174] mwifiex_uninit_sw+0x158/0x1b0 [mwifiex]
[ 247.219747] mwifiex_remove_card+0x38/0xa0 [mwifiex]
[ 247.225316] mwifiex_pcie_remove+0xd0/0xe0 [mwifiex_pcie]
[ 247.231451] pci_device_remove+0x50/0xe0
[ 247.235849] device_release_driver_internal+0x110/0x1b0
[ 247.241701] driver_detach+0x5c/0x9c
[ 247.245704] bus_remove_driver+0x84/0xb8
[ 247.250095] driver_unregister+0x3c/0x60
[ 247.254486] pci_unregister_driver+0x2c/0x90
[ 247.259267] cleanup_module+0x18/0xcdc [mwifiex_pcie]
Fixes: a05829a722 ("cfg80211: avoid holding the RTNL when calling the driver")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/linux-wireless/98392296-40ee-6300-369c-32e16cff3725@gmail.com/
Link: https://lore.kernel.org/linux-wireless/ab4d00ce52f32bd8e45ad0448a44737e@bewaar.me/
Reported-by: Maximilian Luz <luzmaximilian@gmail.com>
Reported-by: dave@bewaar.me
Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Tested-by: Maximilian Luz <luzmaximilian@gmail.com>
Tested-by: Dave Olsthoorn <dave@bewaar.me>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20210515024227.2159311-1-briannorris@chromium.org
Currently, _everything_ in cfg80211 holds the RTNL, and if you
have a slow USB device (or a few) you can get some bad lock
contention on that.
Fix that by re-adding a mutex to each wiphy/rdev as we had at
some point, so we have locking for the wireless_dev lists and
all the other things in there, and also so that drivers still
don't have to worry too much about it (they still won't get
parallel calls for a single device).
Then, we can restrict the RTNL to a few cases where we add or
remove interfaces and really need the added protection. Some
of the global list management still also uses the RTNL, since
we need to have it anyway for netdev management, but we only
hold the RTNL for very short periods of time here.
Link: https://lore.kernel.org/r/20210122161942.81df9f5e047a.I4a8e1a60b18863ea8c5e6d3a0faeafb2d45b2f40@changeid
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> [marvell driver issues]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The functions mwifiex_shutdown_sw() and mwifiex_reinit_sw() can be used
for more general purposes than the PCIe function level reset. Also, these
are even not PCIe-specific.
So, let's update the comments at the top of each function accordingly.
Signed-off-by: Tsuchiya Yuto <kitakar@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20201028142110.18144-3-kitakar@gmail.com
In file included from drivers/net/wireless/marvell/mwifiex//cmdevt.c:26:0:
drivers/net/wireless/marvell/mwifiex//wmm.h:41:17: warning: ‘tos_to_tid_inv’ defined but not used [-Wunused-const-variable=]
static const u8 tos_to_tid_inv[] = {
^~~~~~~~~~~~~~
drivers/net/wireless/marvell/mwifiex//wmm.h:34:18: warning: ‘mwifiex_1d_to_wmm_queue’ defined but not used [-Wunused-const-variable=]
static const u16 mwifiex_1d_to_wmm_queue[8] = { 1, 0, 0, 1, 2, 2, 3, 3 };
^~~~~~~~~~~~~~~~~~~~~~~
move the variables definition to .c file, and leave declarations
in the header file to fix these warnings.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200902140846.29024-1-yuehaibing@huawei.com
ENOTSUPP (double PP) is internal linux kernel code 524 available only in
kernel include file linux/errno.h and not exported to userspace.
EOPNOTSUPP (OP; double PP) is standard code 95 for reporting 'operation not
supported' available via kernel include file uapi/asm-generic/errno.h.
ENOTSUP (single P) is alias for EOPNOTSUPP defined only in userspace
include file bits/errno.h and not available in kernel.
Because Linux kernel does not support ENOTSUP (single P) and because
userspace does not support ENOTSUPP (double PP), report error code for
'operation not supported' via EOPNOTSUPP macro.
This patch fixes problem that mwifiex kernel driver sends to userspace
unsupported error codes like: "failed: -524 (No error information)".
After applying this patch userspace see: "failed: -95 (Not supported)".
Signed-off-by: Pali Rohár <pali@kernel.org>
Acked-by: Ganapathi Bhat <ganapathi.bhat@nxp.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20200703112151.18917-1-pali@kernel.org
As of 6-DEC-2019, NXP has acquired Marvell’s Wireless business
unit. This change is to update the license text accordingly.
Signed-off-by: James Cao <zheng.cao@nxp.com>
Signed-off-by: Cathy Luo <xiaohua.luo@nxp.com>
Signed-off-by: Ganapathi Bhat <ganapathi.bhat@nxp.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The SDIO HW reset procedure in mwifiex_sdio_card_reset_work() is broken,
when the SDIO card is shared with another SDIO func driver. This is the
case when the Bluetooth btmrvl driver is being used in combination with
mwifiex. More precisely, when mwifiex_sdio_card_reset_work() runs to resets
the SDIO card, the btmrvl driver doesn't get notified about it. Beyond that
point, the btmrvl driver will fail to communicate with the SDIO card.
This is a generic problem for SDIO func drivers sharing an SDIO card, which
are about to be addressed in subsequent changes to the mmc core and the
mmc_hw_reset() interface. In principle, these changes means the
mmc_hw_reset() interface starts to return 1 if the are multiple drivers for
the SDIO card, as to indicate to the caller that the reset needed to be
scheduled asynchronously through a hotplug mechanism of the SDIO card.
Let's prepare the mwifiex driver to support the upcoming new behaviour of
mmc_hw_reset(), which means extending the mwifiex_sdio_card_reset_work() to
support the asynchronous SDIO HW reset path. This also means, we need to
allow the ->remove() callback to run, without waiting for the FW to be
loaded. Additionally, during system suspend, mwifiex_sdio_suspend() may be
called when a reset has been scheduled, but waiting to be executed. In this
scenario let's simply return -EBUSY to abort the suspend process, as to
allow the reset to be completed first.
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Douglas Anderson <dianders@chromium.org>
Cc: stable@vger.kernel.org # v5.4+
Acked-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
main_proc_lock and int_lock (in mwifiex_adapter) are the only spinlocks
used in hardirq contexts. The rest are only in task or softirq contexts.
Convert every other lock from *_irq{save,restore}() variants to _bh()
variants.
This is a mechanical transformation of all spinlock usage in mwifiex
using the following:
Step 1:
I ran this nasty sed script:
sed -i -E '/spin_lock_irqsave|spin_unlock_irqrestore/ {
/main_proc_lock|int_lock/! {
s:(spin_(un|)lock)_irq(save|restore):\1_bh: ;
# Join broken lines.
:a /;$/! {
N;
s/\s*\n\s*//;
ba
}
/,.*\);$/ s:,.*\):\):
}
}' drivers/net/wireless/marvell/mwifiex/*
Step 2:
Manually delete the flags / ra_list_flags args from:
mwifiex_send_single_packet()
mwifiex_11n_aggregate_pkt()
mwifiex_send_processed_packet()
which are now unused.
Step 3:
Apply this semantic patch (coccinelle) to remove the unused 'flags'
variables:
// <smpl>
@@
type T;
identifier i;
@@
(
extern T i;
|
- T i;
... when != i
)
// </smpl>
(Usage is something like this:
make coccicheck COCCI=./patch.cocci MODE=patch M=drivers/net/wireless/marvell/mwifiex/
although this skips *.h files for some reasons, so I had to massage
stuff.)
Testing: I've played with a variety of stress tests, including download
stress tests on the same APs which caught regressions with commit
5188d5453b ("mwifiex: restructure rx_reorder_tbl_lock usage"). I've
primarily tested on Marvell 8997 / PCIe, although I've given 8897 / SDIO
a quick spin as well.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
In set_mac_address, driver check for interfaces with same bss_type
For first STA entry, this would return 3 interfaces since all priv's have
bss_type as 0 due to kzalloc. Thus mac address gets changed for STA
unexpected. This patch adds check for first STA and avoids mac address
change. This patch also adds mac_address change for p2p based on bss_num
type.
Signed-off-by: Sharvari Harisangam <sharvari@marvell.com>
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
After the previous patch, all the callers of ndo_select_queue()
provide as a 'fallback' argument netdev_pick_tx.
The only exceptions are nested calls to ndo_select_queue(),
which pass down the 'fallback' available in the current scope
- still netdev_pick_tx.
We can drop such argument and replace fallback() invocation with
netdev_pick_tx(). This avoids an indirect call per xmit packet
in some scenarios (TCP syn, UDP unconnected, XDP generic, pktgen)
with device drivers implementing such ndo. It also clean the code
a bit.
Tested with ixgbe and CONFIG_FCOE=m
With pktgen using queue xmit:
threads vanilla patched
(kpps) (kpps)
1 2334 2428
2 4166 4278
4 7895 8100
v1 -> v2:
- rebased after helper's name change
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Driver is using boolean variables to maintain vairous status
information of adapter. These status variables are accessed by
multiple threads and there is a possibility of a race. To avoid
this, convert these variables to a set of bitops flags, to be
operated atomically.
Below variables of mwifiex_adapter are converted to bitop flags:
surprise_removed
is_cmd_timedout
is_suspended
is_hs_configured
hs_enabling
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This patch makes it so that instead of passing a void pointer as the
accel_priv we instead pass a net_device pointer as sb_dev. Making this
change allows us to pass the subordinate device through to the fallback
function eventually so that we can keep the actual code in the
ndo_select_queue call as focused on possible on the exception cases.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
During changing virtual interface, keep using previous net device
mac address.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
When interface type changed, firmware using a new connction pointer.
We need Re-configure the mac address.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
For user configurated mac address, directly set to firmware with no change.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.
Fix this by returning 'netdev_tx_t' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Multiple interfaces with same bss type could affect each other if
they are sharing the same mac address. In this patch, different
mac address is assigned to new interface which have same bss type
with exist interfaces.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This patch refactor current device dump code to make it generic
for subsequent implementation on usb interface.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Cathy Luo <cluo@marvell.com>
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly.
Cc: Kalle Valo <kvalo@codeaurora.org>
Cc: Amitkumar Karwar <amitkarwar@gmail.com>
Cc: Nishant Sarmukadam <nishants@marvell.com>
Cc: Ganapathi Bhat <gbhat@marvell.com>
Cc: Xinming Hu <huxm@marvell.com>
Cc: Arvind Yadav <arvind.yadav.cs@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Johannes Berg <johannes.berg@intel.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Andrew Zaborowski <andrew.zaborowski@intel.com>
Cc: libertas-dev@lists.infradead.org
Cc: linux-wireless@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Per below statement about p2p device address in WFA P2P
spec $2.4.3:
The P2P Device Address of a P2P Device shall be its globally
administered MAC address, or its globally administered MAC
address with the locally administered bit set.
This patch follow above statement, using a separate device
address for p2p interface
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Cathy Luo <cluo@marvell.com>
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Trivial fix to spelling mistake in aggr_ctrl module parameter
message text.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
We inited wakeup info at the beginning of mwifiex_add_card, so we need
to uninit it in the error handling.
It's much the same as what we did in:
36908c4 mwifiex: uninit wakeup info when removing device
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
This print isn't very useful. It's also different between
mwifiex_add_card() and mwifiex_reinit_sw(), and I'd like to consolidate
them eventually.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
In reading through _mwifiex_fw_dpc(), I noticed that after we've
registered our wiphy, we still have error paths that don't free it back
up. Let's do that.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
It's possible for some control interfaces (e.g., scans, set freq) to be
active after we've stopped our main work queue and the netif TX queues.
These don't get completely shut out until we've unregistered the wdevs
and wiphy.
So let's only free command buffers and poison our lists after
wiphy_unregister().
This resolves various use-after-free issues seen when resetting the
device.
Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
In general, it's helpful to use the same code for device removal as for
device reset, as this tends to have fewer bugs. Let's move the wiphy
unregistration code into the common reset and removal code.
In particular, it's very hard to properly handle the reset sequence when
something fails. Currently, if mwifiex_reinit_sw() fails, we've failed
to unregister the associated wiphy, and so running something as simple
as "iw phy" can trigger an OOPS, as the wiphy still has hooks back into
freed mwifiex data structures. For example, KASAN complained:
[... see reset fail for other reasons ...]
[ 1184.821158] mwifiex_pcie 0000:01:00.0: info: dnld wifi firmware from 174948 bytes
[ 1186.870914] mwifiex_pcie 0000:01:00.0: info: FW download over, size 608396 bytes
[ 1187.685990] mwifiex_pcie 0000:01:00.0: WLAN FW is active
[ 1187.692673] mwifiex_pcie 0000:01:00.0: cmd_wait_q terminated: -512
[ 1187.699075] mwifiex_pcie 0000:01:00.0: info: _mwifiex_fw_dpc: unregister device
[ 1187.713476] mwifiex: Failed to bring up adapter: -5
[ 1187.718644] mwifiex_pcie 0000:01:00.0: reinit failed: -5
[... run `iw phy` ...]
[ 1212.902419] ==================================================================
[ 1212.909806] BUG: KASAN: use-after-free in mwifiex_cfg80211_get_antenna+0x54/0xfc [mwifiex] at addr ffffffc0ad1a8028
[ 1212.920246] Read of size 1 by task iw/3127
[...]
[ 1212.934946] page dumped because: kasan: bad access detected
[...]
[ 1212.950665] Call trace:
[ 1212.953148] [<ffffffc00020a69c>] dump_backtrace+0x0/0x190
[ 1212.958572] [<ffffffc00020a96c>] show_stack+0x20/0x28
[ 1212.963648] [<ffffffc0005ce18c>] dump_stack+0xa4/0xcc
[ 1212.968723] [<ffffffc0003c4430>] kasan_report+0x378/0x500
[ 1212.974140] [<ffffffc0003c3358>] __asan_load1+0x44/0x4c
[ 1212.979462] [<ffffffbffc2e8360>] mwifiex_cfg80211_get_antenna+0x54/0xfc [mwifiex]
[ 1212.987131] [<ffffffbffc084fc4>] nl80211_send_wiphy+0x75c/0x2de0 [cfg80211]
[ 1212.994246] [<ffffffbffc094f60>] nl80211_dump_wiphy+0x32c/0x438 [cfg80211]
[ 1213.001149] [<ffffffc000ab6404>] genl_lock_dumpit+0x48/0x64
[ 1213.006746] [<ffffffc000ab3474>] netlink_dump+0x178/0x398
[ 1213.012171] [<ffffffc000ab3d18>] __netlink_dump_start+0x1bc/0x260
[...]
This all goes away if we just tear down the wiphy on the way down, and
set it back up if/when we bring the device back up.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
When resetting the device, we might have queued up interrupts that
didn't get a chance to finish processing. We really don't need to handle
them at this point; we just want to make sure they don't cause us to try
to process old commands from before the device was reset.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
When PCIe FLR code was added, it explicitly copy-and-pasted much of
mwifiex_remove_card() into mwifiex_shutdown_sw(). This is unnecessary,
as almost all of the code should be reused.
Let's reunite what we can for now.
The only functional changes for now:
* call netif_device_detach() in the remove() code path -- this wasn't
done before, but it really should be a no-op, when the device is
getting totally unregistered soon anyway
* call the ->down_dev() driver callback only after we've finished all
SW teardown -- this should have no significant effect, since the only
user (pcie.c) does very minimal work there, and it doesn't matter
that we reorder this
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Network devices can allocate reasources and private memory using
netdev_ops->ndo_init(). However, the release of these resources
can occur in one of two different places.
Either netdev_ops->ndo_uninit() or netdev->destructor().
The decision of which operation frees the resources depends upon
whether it is necessary for all netdev refs to be released before it
is safe to perform the freeing.
netdev_ops->ndo_uninit() presumably can occur right after the
NETDEV_UNREGISTER notifier completes and the unicast and multicast
address lists are flushed.
netdev->destructor(), on the other hand, does not run until the
netdev references all go away.
Further complicating the situation is that netdev->destructor()
almost universally does also a free_netdev().
This creates a problem for the logic in register_netdevice().
Because all callers of register_netdevice() manage the freeing
of the netdev, and invoke free_netdev(dev) if register_netdevice()
fails.
If netdev_ops->ndo_init() succeeds, but something else fails inside
of register_netdevice(), it does call ndo_ops->ndo_uninit(). But
it is not able to invoke netdev->destructor().
This is because netdev->destructor() will do a free_netdev() and
then the caller of register_netdevice() will do the same.
However, this means that the resources that would normally be released
by netdev->destructor() will not be.
Over the years drivers have added local hacks to deal with this, by
invoking their destructor parts by hand when register_netdevice()
fails.
Many drivers do not try to deal with this, and instead we have leaks.
Let's close this hole by formalizing the distinction between what
private things need to be freed up by netdev->destructor() and whether
the driver needs unregister_netdevice() to perform the free_netdev().
netdev->priv_destructor() performs all actions to free up the private
resources that used to be freed by netdev->destructor(), except for
free_netdev().
netdev->needs_free_netdev is a boolean that indicates whether
free_netdev() should be done at the end of unregister_netdevice().
Now, register_netdevice() can sanely release all resources after
ndo_ops->ndo_init() succeeds, by invoking both ndo_ops->ndo_uninit()
and netdev->priv_destructor().
And at the end of unregister_netdevice(), we invoke
netdev->priv_destructor() and optionally call free_netdev().
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of using 4KB packet buffer for data transfer, new chipset have
more device memory. This patch try to aggregation packets in an 16KB
buffer. In this way, totally usb transaction cost will be reduced.
Thoughput test on usb 2.0 show both TCP TX and UPD TX promote ~40M,
from ~240M to ~280M.
This feature is default disabled, and can be enabled by module
parameter, like:
insmod mwifiex.ko aggr_ctrl=1
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Cathy Luo <cluo@marvell.com>
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Have proper request id filled in the SCHED_SCAN_RESULTS and
SCHED_SCAN_STOPPED notifications toward user-space by having the
driver provide it through the api.
Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Quite a lot of patches for rtlwifi and iwlwifi this time, but changes
also for other active wireless drivers.
Major changes:
ath9k
* add support for Dell Wireless 1601 PCI device
* add debugfs file to manually override noise floor
ath10k
* bump up FW API to 6 for a new QCA6174 firmware branch
wil6210
* support 8 kB RX buffers
iwlwifi
* work to support A000 devices continues
* add support for FW API 30
* add Geographical and Dynamic Specific Absorption Rate (SAR) support
* support a few new PCI device IDs
rtlwifi
* work on adding Bluetooth coexistance support, not finished yet
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJY+cipAAoJEG4XJFUm622b+04H/2u7zhIqv3R5NsmGRlwjKmvd
6us1hcYzUiBW1kBFFtcvokwgqv3Rw/EUi+NYA5UCHgo30HTVJWTqW0q+BUEI7U0X
iwuFdHbs/qjj8hXKgZwaJMVyRhXAbq+iHYaMY5OEDNTRSbxqXA4V69SHGK9vLX0l
KQLJBZX3etnwsyRWrmd4E0roGGfD5tMPpGMUDSlvTPWt0I+1QFUxRI9ohn1fRBpD
p+UHHC+SYF6NzxvfqyPwm6ndb2XOwdAKYljlf4KrBwHtSXvAH8p5EASm4PjGbavY
0mrkv0cb8QOKQNVe9q4B4AmM4iAITZ9i3sQBPJ+EBN2P+kS/lFU5L/+5z3VQjms=
=KTEG
-----END PGP SIGNATURE-----
Merge tag 'wireless-drivers-next-for-davem-2017-04-21' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next
Kalle Valo says:
====================
wireless-drivers-next patches for 4.12
Quite a lot of patches for rtlwifi and iwlwifi this time, but changes
also for other active wireless drivers.
Major changes:
ath9k
* add support for Dell Wireless 1601 PCI device
* add debugfs file to manually override noise floor
ath10k
* bump up FW API to 6 for a new QCA6174 firmware branch
wil6210
* support 8 kB RX buffers
iwlwifi
* work to support A000 devices continues
* add support for FW API 30
* add Geographical and Dynamic Specific Absorption Rate (SAR) support
* support a few new PCI device IDs
rtlwifi
* work on adding Bluetooth coexistance support, not finished yet
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
* connection quality monitoring with multiple thresholds
* support for FILS shared key authentication offload
* pre-CAC regulatory compliance - only ETSI allows this
* sanity check for some rate confusion that hit ChromeOS
(but nobody else uses it, evidently)
* some documentation updates
* lots of cleanups
-----BEGIN PGP SIGNATURE-----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=f4VT
-----END PGP SIGNATURE-----
Merge tag 'mac80211-next-for-davem-2017-04-18' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next
Johannes Berg says:
====================
My last pull request has been a while, we now have:
* connection quality monitoring with multiple thresholds
* support for FILS shared key authentication offload
* pre-CAC regulatory compliance - only ETSI allows this
* sanity check for some rate confusion that hit ChromeOS
(but nobody else uses it, evidently)
* some documentation updates
* lots of cleanups
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
'chan_stats' is (re)allocated in _mwifiex_fw_dpc() ->
mwifiex_init_channel_scan_gap(), which is called whenever the device is
initialized -- at probe or at reset.
But we only free it in we completely unregister the adapter, meaning we
leak a copy of it during every reset.
Let's free it in the shutdown / removal paths instead (and in the
error-handling path), to avoid the leak.
Ideally, we can eventually unify much of mwifiex_shutdown_sw() and
mwifiex_remove_card() (way too much copy-and-paste) to reduce the burden
on bugfixes like this. But that's work for tomorrow.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
If we reset because of a command timeout, we should reset this flag.
Otherwise, we might erroneously think the next command after reset is
timing out, and trigger another reset.
The above behavior effectively neuters the automatic card_reset()
behavior, as it means we will never recover from a command timeout
properly (and in fact, we might enter an infinite loop:
timeout -> reset -> (fake) timeout -> reset -> ...
This fixes a bug introduced with introduction of PCIe function level
reset support, but it was carried into the SDIO driver when it was
converted to use the same codepaths. And this is currently mostly a
problem only in the SDIO driver, because it's the only one with
automatic card_reset() support (e.g., on command timeout). But it will
be a problem for PCIe too, as I'm working on supporting automatic
card_reset() for PCIe.
Fixes: c742e623e9 ("mwifiex: sdio card reset enhancement")
Fixes: 4c5dae59d2 ("mwifiex: add PCIe function level reset support")
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
mwifiex_dbg will do nothing before adapter->dev get assigned. several logs
lost in this case. it can be avoided by fall back to pr_info.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Reviewed-by: Dmitry Torokhov <dtor@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
If wakeup interrupt handler is called, we know that the wakeup
interrupt number is valid, there is no need to check it.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Cathy Luo <cluo@marvell.com>
Reviewed-by: Dmitry Torokhov <dtor@chromium.org>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Instead passing both flags, which can be NULL, and vif_params,
which are never NULL, move the flags into the vif_params and
use BIT(0), which is invalid from userspace, to indicate that
the flags were changed.
While updating all drivers, fix a small bug in wil6210 where
it was setting the flags to 0 instead of leaving them unchanged.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Lots of bugfixes as usual but also some new features.
Major changes:
ath10k
* improve firmware download time for QCA6174 and QCA9377, especially
helps resume time
ath9k_htc
* add support AirTies 1eda:2315 AR9271 device
rt2x00
* add support MT7620
mwifiex
* enable auto deep sleep mode for USB chipsets
brcmfmac
* add support for network namespaces (WIPHY_FLAG_NETNS_OK)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJY55+nAAoJEG4XJFUm622bRUAIAJfCM5YFyh9Y/XV147JshdGi
xScDAgwIA1/o+iHfvTsjSH3/uoH3JhsiqfcXN7R80kyvG7HrUeXaebmUDgbBvOOj
FTMuytiD+xkgEKVjWIwtXUYqctyzQ8ofxIQJ2W5E9CbEYAZ43uEcXzdnhBKVaVuY
XPw59MF5vRGDDXTnTf4af4OC+L1QqwUqsyi4j7oMIfexieMGQxQL0JYOyHweCnCV
gMz/kTxAIcmC1yOiunu8VyU4kK8borW36wC7XEE3MOXhqSKnQjyhI/efA76AKX0j
O7sfEKCFlZU1xXQbkB9ecbKc2jyMefiE0gwLWWI5dwKhnUS2qQ6GS8ML+NMfV0g=
=B2X+
-----END PGP SIGNATURE-----
Merge tag 'wireless-drivers-next-for-davem-2017-04-07' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next
Kalle Valo says:
====================
wireless-drivers-next patches for 4.12
Lots of bugfixes as usual but also some new features.
Major changes:
ath10k
* improve firmware download time for QCA6174 and QCA9377, especially
helps resume time
ath9k_htc
* add support AirTies 1eda:2315 AR9271 device
rt2x00
* add support MT7620
mwifiex
* enable auto deep sleep mode for USB chipsets
brcmfmac
* add support for network namespaces (WIPHY_FLAG_NETNS_OK)
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When resetting the device, we take a synchronous firmware-loading code
path, which borrows a lot from the asynchronous path used at probe time.
We don't catch errors correctly though, which means that in the PCIe
driver, we may try to dereference the 'adapter' struct after
mwifiex_fw_dpc() has freed it. See this (erronous) print in
mwifiex_pcie_reset_notify():
mwifiex_dbg(adapter, INFO, "%s, successful\n", __func__);
Let's instead refactor the synchronous (or "!req_fw_nowait") path so
that we propagate errors and handle them properly.
This fixes a use-after-free issue in the PCIe driver, as well as a
misleading debug message ("successful"). It looks like the SDIO driver
doesn't have these problems, since it doesn't do anything after
mwifiex_reinit_sw().
Fixes: 4c5dae59d2 ("mwifiex: add PCIe function level reset support")
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
If we fail to reinit the FW when resetting the device (in the
synchronous version of mwifiex_init_hw_fw() -> mwifiex_fw_dpc()),
mwifiex_fw_dpc() will tear down the interface and free up the adapter.
But we don't actually check for all failure cases of mwifiex_fw_dpc(),
so some of them fall through and dereference adapter->fw_done with a
freed adapter, causing a use-after-free bug.
In any case, mwifiex_fw_dpc() will always signal FW completion -- in the
error OR success case -- so at best, this was repeat work. Let's not do
it.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Currrently we are disabling this wake irq after receiving it. If this
happens before we finish suspend and the pm event check is disabled,
the system will continue suspending, and this irq would not work again.
We may need to abort system suspend to avoid that.
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
We manually init wakeup info, but we don't detach it on device removal.
This means that if we (for example) rmmod + modprobe the driver, the
device framework might return -EEXIST the second time, and we'll
complain in the logs:
[ 839.311881] mwifiex_pcie 0000:01:00.0: fail to init wakeup for mwifiex
AFAICT, there's no other negative effect.
But we can fix this by disabling wakeup on remove, similar to what a few
other drivers do (e.g., the power supply framework).
This code (and bug) has existed on SDIO for a while, but it got moved
around and enabled for PCIe with commit 853402a008 ("mwifiex: Enable
WoWLAN for both sdio and pcie").
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The mwifiex_dbg() log handler utilizes the struct device in
adapter->dev. Without it, it decides not to print anything.
As of commit 2e02b58142 ("mwifiex: Allow mwifiex early access to device
structure"), we started assigning that pointer only after we finished
mwifiex_register() -- this effectively neuters any mwifiex_dbg() logging
done before this point.
Let's move the device assignment into mwifiex_register().
Fixes: 2e02b58142 ("mwifiex: Allow mwifiex early access to device structure")
Cc: Rajat Jain <rajatja@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
