The ext2_error() function syncs the filesystem so it sleeps. The caller
is holding a spinlock so it's not allowed to sleep.
ext2_statfs() <- disables preempt
-> ext2_count_free_blocks()
-> ext2_get_group_desc()
Fix this by using WARN() to print an error message and a stack trace
instead of using ext2_error().
Link: https://lore.kernel.org/r/20210921203233.GA16529@kili
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jan Kara <jack@suse.cz>
The function uses the newly introduced rockchip_gpio_readl_bit()
which directly returns the actual value of the requeste bit.
So using the existing bit-wise check for the bit inside the value
will always return 0.
Fix this by dropping the bit manipulation on the result.
Fixes: 3bcbd1a85b ("gpio/rockchip: support next version gpio controller")
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
The gpio driver runs into issues on v1 gpio blocks, as the db_clk
and the whole extended debounce support is only ever defined on v2.
So checking for the IS_ERR on the db_clk is not enough, as it will
be NULL on v1.
Fix this by adding the needed condition for v2 first before checking
the existence of the db_clk.
This caused my rk3288-veyron-pinky to enter a reboot loop when it
tried to enable the power-key as adc-key device.
Fixes: 3bcbd1a85b ("gpio/rockchip: support next version gpio controller")
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
The current hwirq is calculated based on the old GPIO pin order(input
GPIO range is from 0 to ngpios - 1).
It should be calculated based on the current GPIO input pin order(input
GPIOs are 0, 2, 4, ..., (ngpios - 1) * 2).
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
The return type of irq_chip.irq_mask() and irq_chip.irq_unmask() should
be void.
Fixes: dbe776c2ca ("gpio: uniphier: add UniPhier GPIO controller driver")
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
Commit 8dcb7a15a5 ("gpiolib: acpi: Take into account debounce settings")
made the gpiolib-acpi code call gpio_set_debounce_timeout() when requesting
GPIOs.
This in itself is fine, but it also made gpio_set_debounce_timeout()
errors fatal, causing the requesting of the GPIO to fail. This is causing
regressions. E.g. on a HP ElitePad 1000 G2 various _AEI specified GPIO
ACPI event sources specify a debouncy timeout of 20 ms, but the
pinctrl-baytrail.c only supports certain fixed values, the closest
ones being 12 or 24 ms and pinctrl-baytrail.c responds with -EINVAL
when specified a value which is not one of the fixed values.
This is causing the acpi_request_own_gpiod() call to fail for 3
ACPI event sources on the HP ElitePad 1000 G2, which in turn is causing
e.g. the battery charging vs discharging status to never get updated,
even though a charger has been plugged-in or unplugged.
Make gpio_set_debounce_timeout() errors non fatal, warning about the
failure instead, to fix this regression.
Note we should probably also fix various pinctrl drivers to just
pick the first bigger discrete value rather then returning -EINVAL but
this will need to be done on a per driver basis, where as this fix
at least gets us back to where things were before and thus restores
functionality on devices where this was lost due to
gpio_set_debounce_timeout() errors.
Fixes: 8dcb7a15a5 ("gpiolib: acpi: Take into account debounce settings")
Depends-on: 2e2b496ceb ("gpiolib: acpi: Extract acpi_request_own_gpiod() helper")
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
Setting SCSI logging level with error=3, we saw some errors from enclosues:
[108017.360833] ses 0:0:9:0: tag#641 Done: NEEDS_RETRY Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK cmd_age=0s
[108017.360838] ses 0:0:9:0: tag#641 CDB: Receive Diagnostic 1c 01 01 00 20 00
[108017.427778] ses 0:0:9:0: Power-on or device reset occurred
[108017.427784] ses 0:0:9:0: tag#641 Done: SUCCESS Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
[108017.427788] ses 0:0:9:0: tag#641 CDB: Receive Diagnostic 1c 01 01 00 20 00
[108017.427791] ses 0:0:9:0: tag#641 Sense Key : Unit Attention [current]
[108017.427793] ses 0:0:9:0: tag#641 Add. Sense: Bus device reset function occurred
[108017.427801] ses 0:0:9:0: Failed to get diagnostic page 0x1
[108017.427804] ses 0:0:9:0: Failed to bind enclosure -19
[108017.427895] ses 0:0:10:0: Attached Enclosure device
[108017.427942] ses 0:0:10:0: Attached scsi generic sg18 type 13
Retry if the Send/Receive Diagnostic commands complete with a transient
error status (NOT_READY or UNIT_ATTENTION with ASC 0x29).
Link: https://lore.kernel.org/r/1631849061-10210-2-git-send-email-wenxiong@linux.ibm.com
Reviewed-by: Brian King <brking@linux.ibm.com>
Reviewed-by: James Bottomley <jejb@linux.ibm.com>
Signed-off-by: Wen Xiong <wenxiong@linux.ibm.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
I fixed a stringop-overread warning earlier this year, now a second copy of
the original code was added and the warning came back:
drivers/scsi/lpfc/lpfc_attr.c: In function 'lpfc_cmf_info_show':
drivers/scsi/lpfc/lpfc_attr.c:289:25: error: 'strnlen' specified bound 4095 exceeds source size 24 [-Werror=stringop-overread]
289 | strnlen(LPFC_INFO_MORE_STR, PAGE_SIZE - 1),
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix it the same way as the other copy.
Link: https://lore.kernel.org/r/20210920095628.1191676-1-arnd@kernel.org
Fixes: ada48ba70f ("scsi: lpfc: Fix gcc -Wstringop-overread warning")
Fixes: 74a7baa2a3 ("scsi: lpfc: Add cmf_info sysfs entry")
Reviewed-by: James Smart <jsmart2021@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The acornscsi driver has a config option to enable tagged queuing, but this
option gets disabled in the driver itself with the comment 'needs to be
debugged'. As this is a _really_ old driver I doubt anyone will be wanting
to invest time here, so remove the tagged queue vestiges and make our lives
easier.
[jpg: Use scsi_cmd_to_rq()]
Link: https://lore.kernel.org/r/1631696835-136198-3-git-send-email-john.garry@huawei.com
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This reverts commit a113eaaf86.
There are a couple of issues with the commit:
1. It causes deadlocks.
2. It causes the shost->eh_cmd_q list of failed requests not to be
processed, ever.
So revert it.
1. Deadlocks
The SCSI error handler runs with requests blocked beginning when
scsi_schedule_eh() sets SHOST_RECOVERY state, continuing through
scsi_error_handler() callback ->eh_strategy_handler() until
scsi_restart_operations() is called. By setting eh_strategy_handler to
ufshcd_err_handler, the patch changed the UFS error handler to run with
requests blocked, including PM requests, for the entire run of the error
handler.
That conflicts with UFS error handler existing synchronization with UFS
device PM operations. The UFS error handler synchronizes with runtime PM
by doing pm_runtime_get_sync() prior to blocking requests itself. It
synchronizes with system PM by use of hba->host_sem, again before blocking
requests itself. However, if requests are already blocked, then PM
operations will block. So:
the UFS error handler blocks waiting on PM
+ PM blocks waiting on SCSI PM requests to process or fail
+ PM requests are blocked waiting on error handling to finish
= deadlock
This happens both for runtime PM and system PM.
Prior to the patch, these deadlocks could not happen even if SCSI error
handling was running, because the presence of requests in shost->eh_cmd_q
would mean the queues could not be suspended, which would mean that, should
the UFS error handler run at the same time, it would not need to wait for
PM or vice versa.
Please note these scenarios are not just theoretical, they were found
during testing on a Samsung Galaxy Book S.
2. ->eh_strategy_handler() must process shost->eh_cmd_q list of failed
requests, as all other eh_strategy_handler's do except UFS error handler.
Refer for example: scsi_unjam_host(), ata_scsi_error() and
sas_scsi_recover_host().
Link: https://lore.kernel.org/r/20210917144349.14058-1-adrian.hunter@intel.com
Fixes: a113eaaf86 ("scsi: ufs: Synchronize SCSI and UFS error handling")
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Julian Wiedmann says:
====================
s390/qeth: fixes 2021-09-21
This brings two fixes for deadlocks when a device is removed while it
has certain types of async work pending. And one additional fix for a
missing NULL check in an error case.
====================
Link: https://lore.kernel.org/r/20210921145217.1584654-1-jwi@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 0b9902c1fc ("s390/qeth: fix deadlock during recovery") removed
taking discipline_mutex inside qeth_do_reset(), fixing potential
deadlocks. An error path was missed though, that still takes
discipline_mutex and thus has the original deadlock potential.
Intermittent deadlocks were seen when a qeth channel path is configured
offline, causing a race between qeth_do_reset and ccwgroup_remove.
Call qeth_set_offline() directly in the qeth_do_reset() error case and
then a new variant of ccwgroup_set_offline(), without taking
discipline_mutex.
Fixes: b41b554c1e ("s390/qeth: fix locking for discipline setup / removal")
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Problem: qeth_close_dev_handler is a worker that tries to acquire
card->discipline_mutex via drv->set_offline() in ccwgroup_set_offline().
Since commit b41b554c1e
("s390/qeth: fix locking for discipline setup / removal")
qeth_remove_discipline() is called under card->discipline_mutex and
cancels the work and waits for it to finish.
STOPLAN reception with reason code IPA_RC_VEPA_TO_VEB_TRANSITION is the
only situation that schedules close_dev_work. In that situation scheduling
qeth recovery will also result in an offline interface, when resetting the
isolation mode fails, if the external switch is still set to VEB.
And since commit 0b9902c1fc ("s390/qeth: fix deadlock during recovery")
qeth recovery does not aquire card->discipline_mutex anymore.
So we accept the longer pathlength of qeth_schedule_recovery in this
error situation and re-use the existing function.
As a side-benefit this changes the hwtrap to behave like during recovery
instead of like during a user-triggered set_offline.
Fixes: b41b554c1e ("s390/qeth: fix locking for discipline setup / removal")
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Acked-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When qeth_set_online() calls qeth_clear_working_pool_list() to roll
back after an error exit from qeth_hardsetup_card(), we are at risk of
accessing card->qdio.in_q before it was allocated by
qeth_alloc_qdio_queues() via qeth_mpc_initialize().
qeth_clear_working_pool_list() then dereferences NULL, and by writing to
queue->bufs[i].pool_entry scribbles all over the CPU's lowcore.
Resulting in a crash when those lowcore areas are used next (eg. on
the next machine-check interrupt).
Such a scenario would typically happen when the device is first set
online and its queues aren't allocated yet. An early IO error or certain
misconfigs (eg. mismatched transport mode, bad portno) then cause us to
error out from qeth_hardsetup_card() with card->qdio.in_q still being
NULL.
Fix it by checking the pointer for NULL before accessing it.
Note that we also have (rare) paths inside qeth_mpc_initialize() where
a configuration change can cause us to free the existing queues,
expecting that subsequent code will allocate them again. If we then
error out before that re-allocation happens, the same bug occurs.
Fixes: eff73e16ee ("s390/qeth: tolerate pre-filled RX buffer")
Reported-by: Stefan Raspl <raspl@linux.ibm.com>
Root-caused-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The problem is the mismatched types between "ctx->total_len" which is
an unsigned int, "rc" which is an int, and "ctx->rc" which is a
ssize_t. The code does:
ctx->rc = (rc == 0) ? ctx->total_len : rc;
We want "ctx->rc" to store the negative "rc" error code. But what
happens is that "rc" is type promoted to a high unsigned int and
'ctx->rc" will store the high positive value instead of a negative
value.
The fix is to change "rc" from an int to a ssize_t.
Fixes: c610c4b619 ("CIFS: Add asynchronous write support through kernel AIO")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
During the v5.13 cycle we updated the SPI subsystem to generate OF style
modaliases for SPI devices, replacing the old Linux style modalises we
used to generate based on spi_device_id which are the DT style name with
the vendor removed. Unfortunately this means that we start only
reporting OF style modalises and not the old ones and there is nothing
that ensures that drivers list every possible OF compatible string in
their OF ID table. The result is that there are systems which have been
relying on loading modules based on the old style that are now broken,
as found by Russell King with spi-nor on Macchiatobin.
spi-nor is a particularly problematic case for this, it only lists a
single generic DT compatible jedec,spi-nor in the driver but supports a
huge raft of device specific compatibles, with a large set of part
numbers many of which are offered by multiple vendors. Russell's
searches of upstream device trees has turned up examples with vendor
names written in non-standard ways too. To make matters worse up until
8ff16cf77c ("Documentation: devicetree: m25p80: add "nor-jedec"
binding") the generic compatible was not part of the binding so there
are device trees out there written to that binding version which don't
list it all. The sheer number of parts supported together with our
previous approach of ignoring the vendor ID makes robustly fixing this
by adding compatibles to the spi-nor driver seem problematic, the
current DT binding document does not list all the parts supported by the
driver at the minute (further patches will fix this).
I've also investigated supporting both formats of modalias
simultaneously but that doesn't seem possible, especially without
breaking our userspace ABI which is obviously not viable.
Instead revert the relevant changes for now:
e09f2ab8ee ("spi: update modalias_show after of_device_uevent_modalias support")
3ce6c9e261 ("spi: add of_device_uevent_modalias support")
This will unfortunately mean that any system which had started having
modules autoload based on the OF compatibles for drivers that list
things there but not in the spi_device_ids will now not have those
modules load which is itself a regression. Since it affects a narrower
time window and the particularly problematic spi-nor driver may be
critical to system boot on smaller systems this seems the best of a
series of bad options. I will start an audit of SPI drivers to identify
and fix cases where things won't autoload using spi_device_id, this is
not great but seems to be the best way forward that anyone has been able
to identify.
Thanks to Russell for both his report and the additional diagnostic and
analysis work he has done here, the detailed research above was his
work.
Fixes: e09f2ab8ee ("spi: update modalias_show after of_device_uevent_modalias support")
Fixes: 3ce6c9e261 ("spi: add of_device_uevent_modalias support")
Reported-by: Russell King (Oracle) <linux@armlinux.org.uk>
Suggested-by: Russell King (Oracle) <linux@armlinux.org.uk>
Signed-off-by: Mark Brown <broonie@kernel.org>
Tested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Cc: Andreas Schwab <schwab@suse.de>
Cc: Marco Felsch <m.felsch@pengutronix.de>
Windows client expect to get default stream name(::DATA) in
FILE_STREAM_INFORMATION response even if there is no stream data in file.
This patch fix update failure when writing ppt or doc files.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Reviewed-By: Tom Talpey <tom@talpey.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
While we are working through detailed security reviews
of ksmbd server code we should remind users that it is an
experimental module by adding a warning when the module
loads. Currently the module shows as experimental
in Kconfig and is disabled by default, but we don't want
to confuse users.
Although ksmbd passes a wide variety of the
important functional tests (since initial focus had
been largely on functional testing such as smbtorture,
xfstests etc.), and ksmbd has added key security
features (e.g. GCM256 encryption, Kerberos support),
there are ongoing detailed reviews of the code base
for path processing and network buffer decoding, and
this patch reminds users that the module should be
considered "experimental."
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Reviewed-by: Paulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
During initialization of a signal testcase, features declared as required
are properly checked against the running system but no action is then taken
to effectively skip such a testcase.
Fix core signals test logic to abort initialization and report such a
testcase as skipped to the KSelfTest framework.
Fixes: f96bf43403 ("kselftest: arm64: mangle_pstate_invalid_compat_toggle and common utils")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20210920121228.35368-1-cristian.marussi@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Pull s390 eBPF fixes from Vasily Gorbik:
"Johan Almbladh has implemented a number of new testcases for eBPF [1],
which uncovered three miscompilation issues in the s390 eBPF JIT"
Link: https://lore.kernel.org/bpf/20210902185229.1840281-1-johan.almbladh@anyfinetworks.com/ [1]
* tag 's390-5.15-ebpf-jit-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/bpf: Fix optimizing out zero-extensions
s390/bpf: Fix 64-bit subtraction of the -0x80000000 constant
s390/bpf: Fix branch shortening during codegen pass
The "> max" tests should be ">= max" to prevent an out of bounds access
on the next lines.
Fixes: e1a4541ec0 ("ceph: flush the mdlog before waiting on unsafe reqs")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
In commit b7213ffa0e ("qnx4: avoid stringop-overread errors") I tried
to teach gcc about how the directory entry structure can be two
different things depending on a status flag. It made the code clearer,
and it seemed to make gcc happy.
However, Arnd points to a gcc bug, where despite using two different
members of a union, gcc then gets confused, and uses the size of one of
the members to decide if a string overrun happens. And not necessarily
the rigth one.
End result: with some configurations, gcc-11 will still complain about
the source buffer size being overread:
fs/qnx4/dir.c: In function 'qnx4_readdir':
fs/qnx4/dir.c:76:32: error: 'strnlen' specified bound [16, 48] exceeds source size 1 [-Werror=stringop-overread]
76 | size = strnlen(name, size);
| ^~~~~~~~~~~~~~~~~~~
fs/qnx4/dir.c:26:22: note: source object declared here
26 | char de_name;
| ^~~~~~~
because gcc will get confused about which union member entry is actually
getting accessed, even when the source code is very clear about it. Gcc
internally will have combined two "redundant" pointers (pointing to
different union elements that are at the same offset), and takes the
size checking from one or the other - not necessarily the right one.
This is clearly a gcc bug, but we can work around it fairly easily. The
biggest thing here is the big honking comment about why we do what we
do.
Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99578#c6
Reported-and-tested-by: Arnd Bergmann <arnd@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Defining PCI_IOBASE results in pci resource handling working but the
addresses generated for IO accesses are wrong since the ioremap in the pci core
function 'pci_parse_request_of_pci_ranges' tries to remap to a fixed virtual
address (PC_IOBASE) which can't work for KSEG1 addresses. To get it working this
way, we would need to put PCI_IOBASE somewhere into KSEG2 which will result in
creating TLB entries for IO addresses, which most of the time isn't needed on
MIPS because of access via KSEG1. So avoid to define PCI_IOBASE and increase
IO_SPACE_LIMIT resource for ralink MIPS platform instead, to get valid IO
addresses for resources from pci core 'pci_address_to_pio' function.
Fixes: 222b27713d ("MIPS: ralink: Define PCI_IOBASE")
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Link: https://lore.kernel.org/r/20210822161005.22467-2-sergio.paracuellos@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Lockdep complains about rtw_free_assoc_resources() taking the sta_hash_lock
followed by it calling rtw_free_stainfo() which takes xmitpriv->lock.
While the rtl8723bs_xmit_thread takes the sta_hash_lock while already
holding the xmitpriv->lock:
[ 103.849756] ======================================================
[ 103.849761] WARNING: possible circular locking dependency detected
[ 103.849767] 5.15.0-rc1+ #470 Tainted: G C E
[ 103.849773] ------------------------------------------------------
[ 103.849776] wpa_supplicant/695 is trying to acquire lock:
[ 103.849781] ffffa5d0c0562b00 (&pxmitpriv->lock){+.-.}-{2:2}, at: rtw_free_stainfo+0x8a/0x510 [r8723bs]
[ 103.849840]
but task is already holding lock:
[ 103.849843] ffffa5d0c05636a8 (&pstapriv->sta_hash_lock){+.-.}-{2:2}, at: rtw_free_assoc_resources+0x48/0x110 [r8723bs]
[ 103.849881]
which lock already depends on the new lock.
[ 103.849884]
the existing dependency chain (in reverse order) is:
[ 103.849887]
-> #1 (&pstapriv->sta_hash_lock){+.-.}-{2:2}:
[ 103.849898] _raw_spin_lock_bh+0x34/0x40
[ 103.849913] rtw_get_stainfo+0x93/0x110 [r8723bs]
[ 103.849948] rtw_make_wlanhdr+0x14a/0x270 [r8723bs]
[ 103.849983] rtw_xmitframe_coalesce+0x5c/0x6c0 [r8723bs]
[ 103.850019] rtl8723bs_xmit_thread+0x4ac/0x620 [r8723bs]
[ 103.850050] kthread+0x143/0x160
[ 103.850058] ret_from_fork+0x22/0x30
[ 103.850067]
-> #0 (&pxmitpriv->lock){+.-.}-{2:2}:
[ 103.850077] __lock_acquire+0x1158/0x1de0
[ 103.850084] lock_acquire+0xb5/0x2b0
[ 103.850090] _raw_spin_lock_bh+0x34/0x40
[ 103.850095] rtw_free_stainfo+0x8a/0x510 [r8723bs]
[ 103.850130] rtw_free_assoc_resources+0x53/0x110 [r8723bs]
[ 103.850159] PHY_IQCalibrate_8723B+0x122b/0x36a0 [r8723bs]
[ 103.850189] cfg80211_disconnect+0x173/0x320 [cfg80211]
[ 103.850331] nl80211_disconnect+0x6e/0xb0 [cfg80211]
[ 103.850422] genl_family_rcv_msg_doit+0xcd/0x110
[ 103.850430] genl_rcv_msg+0xce/0x1c0
[ 103.850435] netlink_rcv_skb+0x50/0xf0
[ 103.850441] genl_rcv+0x24/0x40
[ 103.850446] netlink_unicast+0x16d/0x230
[ 103.850452] netlink_sendmsg+0x22b/0x450
[ 103.850457] sock_sendmsg+0x5e/0x60
[ 103.850465] ____sys_sendmsg+0x22f/0x270
[ 103.850472] ___sys_sendmsg+0x81/0xc0
[ 103.850479] __sys_sendmsg+0x49/0x80
[ 103.850485] do_syscall_64+0x3b/0x90
[ 103.850493] entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 103.850500]
other info that might help us debug this:
[ 103.850504] Possible unsafe locking scenario:
[ 103.850507] CPU0 CPU1
[ 103.850510] ---- ----
[ 103.850512] lock(&pstapriv->sta_hash_lock);
[ 103.850518] lock(&pxmitpriv->lock);
[ 103.850524] lock(&pstapriv->sta_hash_lock);
[ 103.850530] lock(&pxmitpriv->lock);
[ 103.850535]
*** DEADLOCK ***
Push the taking of sta_hash_lock down into rtw_free_stainfo(),
where the critical section is, this allows taking the lock after
rtw_free_stainfo() has released pxmitpriv->lock.
This requires changing rtw_free_all_stainfo() so that it does its freeing
in 2 steps, first moving all stainfo-s to free to a local list while
holding the sta_hash_lock and then walking that list to call
rtw_free_stainfo() on them without holding the sta_hash_lock.
Pushing the taking of sta_hash_lock down into rtw_free_stainfo(),
also fixes a whole bunch of callers of rtw_free_stainfo() which
were not holding that lock even though they should.
Note that this also fixes the deadlock from the "remove possible
deadlock when disconnect" patch in a different way. But the
changes from that patch offer a nice locking cleanup regardless.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20210920145502.155454-2-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
