commit 9dc162e223 upstream.
The Focusrite Scarlett audio device does not behave correctly during
resumes. Below is what happens during every resume (captured with
Beagle 5000):
<Suspend>
<Resume>
<Reset>/<Chirp J>/<Tiny J>
<Reset/Target disconnected>
<High Speed>
The Scarlett disconnects and is enumerated again.
However from time to time it drops completely off the USB bus during
resume. Below is captured occurrence of such an event:
<Suspend>
<Resume>
<Reset>/<Chirp J>/<Tiny J>
<Reset>/<Chirp K>/<Tiny K>
<High Speed>
<Corrupted packet>
<Reset/Target disconnected>
To fix the condition a user has to unplug and plug the device again.
With USB_QUIRK_RESET_RESUME applied ("usbcore.quirks=1235:8211:b")
for the Scarlett audio device the issue still reproduces.
Applying USB_QUIRK_DISCONNECT_SUSPEND ("usbcore.quirks=1235:8211:m")
fixed the issue and the Scarlett audio device didn't drop off the USB
bus for ~5000 suspend/resume cycles where originally issue reproduced in
~100 or less suspend/resume cycles.
Signed-off-by: Łukasz Bartosik <lb@semihalf.com>
Cc: stable <stable@kernel.org>
Link: https://lore.kernel.org/r/20230724112911.1802577-1-lb@semihalf.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e835c0a4e2 upstream.
Commit c4a5153e87 ("usb: dwc3: core: Power-off core/PHYs on
system_suspend in host mode") replaces check for HOST only dr_mode with
current_dr_role. But during booting, the current_dr_role isn't
initialized, thus the device side reset is always issued even if dwc3
was configured as host-only. What's more, on some platforms with host
only dwc3, aways issuing device side reset by accessing device register
block can cause kernel panic.
Fixes: c4a5153e87 ("usb: dwc3: core: Power-off core/PHYs on system_suspend in host mode")
Cc: stable <stable@kernel.org>
Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/20230627162018.739-1-jszhang@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b32b8f2b95 upstream.
Hardware based on the Bay Trail / BYT SoCs require an external ULPI phy for
USB device-mode. The phy chip usually has its 'reset' and 'chip select'
lines connected to GPIOs described by ACPI fwnodes in the DSDT table.
Because of hardware with missing ACPI resources for the 'reset' and 'chip
select' GPIOs commit 5741022cbd ("usb: dwc3: pci: Add GPIO lookup table
on platforms without ACPI GPIO resources") introduced a fallback
gpiod_lookup_table with hard-coded mappings for Bay Trail devices.
However there are existing Bay Trail based devices, like the National
Instruments cRIO-903x series, where the phy chip has its 'reset' and
'chip-select' lines always asserted in hardware via resistor pull-ups. On
this hardware the phy chip is always enabled and the ACPI dsdt table is
missing information not only for the 'chip-select' and 'reset' lines but
also for the BYT GPIO controller itself "INT33FC".
With the introduction of the gpiod_lookup_table initializing the USB
device-mode on these hardware now errors out. The error comes from the
gpiod_get_optional() calls in dwc3_pci_quirks() which will now return an
-ENOENT error due to the missing ACPI entry for the INT33FC gpio controller
used in the aforementioned table.
This hardware used to work before because gpiod_get_optional() will return
NULL instead of -ENOENT if no GPIO has been assigned to the requested
function. The dwc3_pci_quirks() code for setting the 'cs' and 'reset' GPIOs
was then skipped (due to the NULL return). This is the correct behavior in
cases where the phy chip is hardwired and there are no GPIOs to control.
Since the gpiod_lookup_table relies on the presence of INT33FC fwnode
in ACPI tables only add the table if we know the entry for the INT33FC
gpio controller is present. This allows Bay Trail based devices with
hardwired dwc3 ULPI phys to continue working.
Fixes: 5741022cbd ("usb: dwc3: pci: Add GPIO lookup table on platforms without ACPI GPIO resources")
Cc: stable <stable@kernel.org>
Signed-off-by: Gratian Crisan <gratian.crisan@ni.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20230726184555.218091-2-gratian.crisan@ni.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 734ae15ab9 upstream.
This reverts commit b138e23d3d.
AutoRetry has been found to sometimes cause controller freezes when
communicating with buggy USB devices.
This controller feature allows the controller in host mode to send
non-terminating/burst retry ACKs instead of terminating retry ACKs
to devices when a transaction error (CRC error or overflow) occurs.
Unfortunately, if the USB device continues to respond with a CRC error,
the controller will not complete endpoint-related commands while it
keeps trying to auto-retry. [3] The xHCI driver will notice this once
it tries to abort the transfer using a Stop Endpoint command and
does not receive a completion in time. [1]
This situation is reported to dmesg:
[sda] tag#29 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD IN
[sda] tag#29 CDB: opcode=0x28 28 00 00 69 42 80 00 00 48 00
xhci-hcd: xHCI host not responding to stop endpoint command
xhci-hcd: xHCI host controller not responding, assume dead
xhci-hcd: HC died; cleaning up
Some users observed this problem on an Odroid HC2 with the JMS578
USB3-to-SATA bridge. The issue can be triggered by starting
a read-heavy workload on an attached SSD. After a while, the host
controller would die and the SSD would disappear from the system. [1]
Further analysis by Synopsys determined that controller revisions
other than the one in Odroid HC2 are also affected by this.
The recommended solution was to disable AutoRetry altogether.
This change does not have a noticeable performance impact. [2]
Revert the enablement commit. This will keep the AutoRetry bit in
the default state configured during SoC design [2].
Fixes: b138e23d3d ("usb: dwc3: core: Enable AutoRetry feature in the controller")
Link: https://lore.kernel.org/r/a21f34c04632d250cd0a78c7c6f4a1c9c7a43142.camel@gmail.com/ [1]
Link: https://lore.kernel.org/r/20230711214834.kyr6ulync32d4ktk@synopsys.com/ [2]
Link: https://lore.kernel.org/r/20230712225518.2smu7wse6djc7l5o@synopsys.com/ [3]
Cc: stable@vger.kernel.org
Cc: Mauro Ribeiro <mauro.ribeiro@hardkernel.com>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Suggested-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Jakub Vanek <linuxtardis@gmail.com>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/20230714122419.27741-1-linuxtardis@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f8a2da6ec2 upstream.
After an initial link up the CAN device is in ERROR-ACTIVE mode. Due
to a missing CAN_STATE_STOPPED in gs_can_close() it doesn't change to
STOPPED after a link down:
| ip link set dev can0 up
| ip link set dev can0 down
| ip --details link show can0
| 13: can0: <NOARP,ECHO> mtu 16 qdisc pfifo_fast state DOWN mode DEFAULT group default qlen 10
| link/can promiscuity 0 allmulti 0 minmtu 0 maxmtu 0
| can state ERROR-ACTIVE restart-ms 1000
Add missing assignment of CAN_STATE_STOPPED in gs_can_close().
Cc: stable@vger.kernel.org
Fixes: d08e973a77 ("can: gs_usb: Added support for the GS_USB CAN devices")
Link: https://lore.kernel.org/all/20230718-gs_usb-fix-can-state-v1-1-f19738ae2c23@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit dd92c8a1f9 upstream.
Add the device and product ID for this CAN bus interface / license
dongle. The device is usable either directly from user space or can be
attached to a kernel CAN interface with slcan_attach.
Reported-by: Kaufmann Automotive GmbH <info@kaufmann-automotive.ch>
Tested-by: Kaufmann Automotive GmbH <info@kaufmann-automotive.ch>
Signed-off-by: Oliver Neukum <oneukum@suse.com>
[ johan: amend commit message and move entries in sort order ]
Cc: stable@vger.kernel.org
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4dd8752a14 upstream.
The runtime PM state should not be changed by drivers that do not
implement runtime PM even if it happens to work around a bug in PM core.
With the wake irq arming now fixed, drop the bogus runtime PM state
update which left the device in active state (and could potentially
prevent a parent device from suspending).
Fixes: f3974413cf ("tty: serial: qcom_geni_serial: Wakeup IRQ cleanup")
Cc: 5.6+ <stable@vger.kernel.org> # 5.6+
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 26a0652cb4 upstream.
Reject KVM_SET_SREGS{2} with -EINVAL if the incoming CR0 is invalid,
e.g. due to setting bits 63:32, illegal combinations, or to a value that
isn't allowed in VMX (non-)root mode. The VMX checks in particular are
"fun" as failure to disallow Real Mode for an L2 that is configured with
unrestricted guest disabled, when KVM itself has unrestricted guest
enabled, will result in KVM forcing VM86 mode to virtual Real Mode for
L2, but then fail to unwind the related metadata when synthesizing a
nested VM-Exit back to L1 (which has unrestricted guest enabled).
Opportunistically fix a benign typo in the prototype for is_valid_cr4().
Cc: stable@vger.kernel.org
Reported-by: syzbot+5feef0b9ee9c8e9e5689@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/000000000000f316b705fdf6e2b4@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20230613203037.1968489-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c4abd73520 upstream.
Stuff CR0 and/or CR4 to be compliant with a restricted guest if and only
if KVM itself is not configured to utilize unrestricted guests, i.e. don't
stuff CR0/CR4 for a restricted L2 that is running as the guest of an
unrestricted L1. Any attempt to VM-Enter a restricted guest with invalid
CR0/CR4 values should fail, i.e. in a nested scenario, KVM (as L0) should
never observe a restricted L2 with incompatible CR0/CR4, since nested
VM-Enter from L1 should have failed.
And if KVM does observe an active, restricted L2 with incompatible state,
e.g. due to a KVM bug, fudging CR0/CR4 instead of letting VM-Enter fail
does more harm than good, as KVM will often neglect to undo the side
effects, e.g. won't clear rmode.vm86_active on nested VM-Exit, and thus
the damage can easily spill over to L1. On the other hand, letting
VM-Enter fail due to bad guest state is more likely to contain the damage
to L2 as KVM relies on hardware to perform most guest state consistency
checks, i.e. KVM needs to be able to reflect a failed nested VM-Enter into
L1 irrespective of (un)restricted guest behavior.
Cc: Jim Mattson <jmattson@google.com>
Cc: stable@vger.kernel.org
Fixes: bddd82d19e ("KVM: nVMX: KVM needs to unset "unrestricted guest" VM-execution control in vmcs02 if vmcs12 doesn't set it")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20230613203037.1968489-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f4fc01af5b upstream.
The legacy gadget driver omitted calling usb_gadget_check_config()
to ensure that the USB device controller (UDC) has adequate resources,
including sufficient endpoint numbers and types, to support the given
configuration.
Previously, usb_add_config() was solely invoked by the legacy gadget
driver. Adds the necessary usb_gadget_check_config() after the bind()
operation to fix the issue.
Fixes: dce49449e0 ("usb: cdns3: allocate TX FIFO size according to composite EP number")
Cc: stable <stable@kernel.org>
Reported-by: Ravi Gunasekaran <r-gunasekaran@ti.com>
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20230707230015.494999-1-Frank.Li@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a8291be6b5 upstream.
This reverts commit f08aa7c80d.
The reverted commit was based on static analysis and a misunderstanding
of how PTR_ERR() and NULLs are supposed to work. When a function
returns both pointer errors and NULL then normally the NULL means
"continue operating without a feature because it was deliberately
turned off". The NULL should not be treated as a failure. If a driver
cannot work when that feature is disabled then the KConfig should
enforce that the function cannot return NULL. We should not need to
test for it.
In this driver, the bug means that probe cannot succeed when CONFIG_PM
is disabled.
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Fixes: f08aa7c80d ("usb: gadget: tegra-xudc: Fix error check in tegra_xudc_powerdomain_init()")
Cc: stable <stable@kernel.org>
Link: https://lore.kernel.org/r/ZKQoBa84U/ykEh3C@moroto
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit dea499781a ]
Warning happened in trace_buffered_event_disable() at
WARN_ON_ONCE(!trace_buffered_event_ref)
Call Trace:
? __warn+0xa5/0x1b0
? trace_buffered_event_disable+0x189/0x1b0
__ftrace_event_enable_disable+0x19e/0x3e0
free_probe_data+0x3b/0xa0
unregister_ftrace_function_probe_func+0x6b8/0x800
event_enable_func+0x2f0/0x3d0
ftrace_process_regex.isra.0+0x12d/0x1b0
ftrace_filter_write+0xe6/0x140
vfs_write+0x1c9/0x6f0
[...]
The cause of the warning is in __ftrace_event_enable_disable(),
trace_buffered_event_enable() was called once while
trace_buffered_event_disable() was called twice.
Reproduction script show as below, for analysis, see the comments:
```
#!/bin/bash
cd /sys/kernel/tracing/
# 1. Register a 'disable_event' command, then:
# 1) SOFT_DISABLED_BIT was set;
# 2) trace_buffered_event_enable() was called first time;
echo 'cmdline_proc_show:disable_event:initcall:initcall_finish' > \
set_ftrace_filter
# 2. Enable the event registered, then:
# 1) SOFT_DISABLED_BIT was cleared;
# 2) trace_buffered_event_disable() was called first time;
echo 1 > events/initcall/initcall_finish/enable
# 3. Try to call into cmdline_proc_show(), then SOFT_DISABLED_BIT was
# set again!!!
cat /proc/cmdline
# 4. Unregister the 'disable_event' command, then:
# 1) SOFT_DISABLED_BIT was cleared again;
# 2) trace_buffered_event_disable() was called second time!!!
echo '!cmdline_proc_show:disable_event:initcall:initcall_finish' > \
set_ftrace_filter
```
To fix it, IIUC, we can change to call trace_buffered_event_enable() at
fist time soft-mode enabled, and call trace_buffered_event_disable() at
last time soft-mode disabled.
Link: https://lore.kernel.org/linux-trace-kernel/20230726095804.920457-1-zhengyejian1@huawei.com
Cc: <mhiramat@kernel.org>
Fixes: 0fc1b09ff1 ("tracing: Use temp buffer when filtering events")
Signed-off-by: Zheng Yejian <zhengyejian1@huawei.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2d093282b0 ]
When pages are removed in rb_remove_pages(), 'cpu_buffer->read' is set
to 0 in order to make sure any read iterators reset themselves. However,
this will mess 'entries' stating, see following steps:
# cd /sys/kernel/tracing/
# 1. Enlarge ring buffer prepare for later reducing:
# echo 20 > per_cpu/cpu0/buffer_size_kb
# 2. Write a log into ring buffer of cpu0:
# taskset -c 0 echo "hello1" > trace_marker
# 3. Read the log:
# cat per_cpu/cpu0/trace_pipe
<...>-332 [000] ..... 62.406844: tracing_mark_write: hello1
# 4. Stop reading and see the stats, now 0 entries, and 1 event readed:
# cat per_cpu/cpu0/stats
entries: 0
[...]
read events: 1
# 5. Reduce the ring buffer
# echo 7 > per_cpu/cpu0/buffer_size_kb
# 6. Now entries became unexpected 1 because actually no entries!!!
# cat per_cpu/cpu0/stats
entries: 1
[...]
read events: 0
To fix it, introduce 'page_removed' field to count total removed pages
since last reset, then use it to let read iterators reset themselves
instead of changing the 'read' pointer.
Link: https://lore.kernel.org/linux-trace-kernel/20230724054040.3489499-1-zhengyejian1@huawei.com
Cc: <mhiramat@kernel.org>
Cc: <vnagarnaik@google.com>
Fixes: 83f40318da ("ring-buffer: Make removal of ring buffer pages atomic")
Signed-off-by: Zheng Yejian <zhengyejian1@huawei.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3fc2febb0f ]
The global function triggers a warning because of the missing prototype
drivers/ata/pata_ns87415.c:263:6: warning: no previous prototype for 'ns87560_tf_read' [-Wmissing-prototypes]
263 | void ns87560_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
There are no other references to this, so just make it static.
Fixes: c4b5b7b6c4 ("pata_ns87415: Initial cut at 87415/87560 IDE support")
Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Reviewed-by: Serge Semin <fancer.lancer@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 53e7d08f6d ]
In ublk_ctrl_start_dev(), if wait_for_completion_interruptible() is
interrupted by signal, queues aren't setup successfully yet, so we
have to fail UBLK_CMD_START_DEV, otherwise kernel oops can be triggered.
Reported by German when working on qemu-storage-deamon which requires
single thread ublk daemon.
Fixes: 71f28f3136 ("ublk_drv: add io_uring based userspace block driver")
Reported-by: German Maglione <gmaglione@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230726144502.566785-2-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1b5d0ddcb3 ]
A fence id of zero is expected to be invalid, and is not removed from
the fence_idr table. If userspace is requesting to specify the fence
id with the FENCE_SN_IN flag, we need to reject a zero fence id value.
Fixes: 17154addc5 ("drm/msm: Add MSM_SUBMIT_FENCE_SN_IN")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/549180/
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c01aebeef3 ]
If the second call to amdgpu_bo_create_kernel() fails, the memory
allocated from the first call should be cleared. If the third call
fails, the memory from the second call should be cleared.
Fixes: b95b539168 ("drm/amdgpu/psp: move PSP memory alloc from hw_init to sw_init")
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7d5fff8982 ]
__md_stop_writes() and __md_stop() will modify many fields that are
protected by 'reconfig_mutex', and all the callers will grab
'reconfig_mutex' except for md_stop().
Also, update md_stop() to make certain 'reconfig_mutex' is held using
lockdep_assert_held().
Fixes: 9d09e663d5 ("dm: raid456 basic support")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e74c874eab ]
There are four equivalent goto tags in raid_ctr(), clean them up to
use just one.
There is no functional change and this is preparation to fix
raid_ctr()'s unprotected md_stop().
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
Stable-dep-of: 7d5fff8982 ("dm raid: protect md_stop() with 'reconfig_mutex'")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bae3028799 ]
In the error paths 'bad_stripe_cache' and 'bad_check_reshape',
'reconfig_mutex' is still held after raid_ctr() returns.
Fixes: 9dbd1aa3a8 ("dm raid: add reshaping support to the target")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1982655821 ]
The NTLMSSP_NEGOTIATE_VERSION flag only needs to be sent during
the NTLMSSP NEGOTIATE (not the AUTH) request, so filter it out for
NTLMSSP AUTH requests. See MS-NLMP 2.2.1.3
This fixes a problem found by the gssntlmssp server.
Link: https://github.com/gssapi/gss-ntlmssp/issues/95
Fixes: 52d005337b ("smb3: send NTLMSSP version information")
Acked-by: Roy Shterman <roy.shterman@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4cf67d3cc9 ]
KASAN and KFENCE detected an user-after-free in the CXL driver. This
happens in the cxl_decoder_add() fail path. KASAN prints the following
error:
BUG: KASAN: slab-use-after-free in cxl_parse_cfmws (drivers/cxl/acpi.c:299)
This happens in cxl_parse_cfmws(), where put_device() is called,
releasing cxld, which is accessed later.
Use the local variables in the dev_err() instead of pointing to the
released memory. Since the dev_err() is printing a resource, change the open
coded print format to use the %pr format specifier.
Fixes: e50fe01e1f ("cxl/core: Drop ->platform_res attribute for root decoders")
Signed-off-by: Breno Leitao <leitao@debian.org>
Link: https://lore.kernel.org/r/20230714093146.2253438-1-leitao@debian.org
Reviewed-by: Alison Schofield <alison.schofield@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b5bbc65512 ]
HW may generate completions that indicates QP is destroyed.
Driver should not be scheduling any more completion handlers
for this QP, after the QP is destroyed. Since CQs are active
during the QP destroy, driver may still schedule completion
handlers. This can cause a race where the destroy_cq and poll_cq
running simultaneously.
Snippet of kernel panic while doing bnxt_re driver load unload in loop.
This indicates a poll after the CQ is freed.
[77786.481636] Call Trace:
[77786.481640] <TASK>
[77786.481644] bnxt_re_poll_cq+0x14a/0x620 [bnxt_re]
[77786.481658] ? kvm_clock_read+0x14/0x30
[77786.481693] __ib_process_cq+0x57/0x190 [ib_core]
[77786.481728] ib_cq_poll_work+0x26/0x80 [ib_core]
[77786.481761] process_one_work+0x1e5/0x3f0
[77786.481768] worker_thread+0x50/0x3a0
[77786.481785] ? __pfx_worker_thread+0x10/0x10
[77786.481790] kthread+0xe2/0x110
[77786.481794] ? __pfx_kthread+0x10/0x10
[77786.481797] ret_from_fork+0x2c/0x50
To avoid this, complete all completion handlers before returning the
destroy QP. If free_cq is called soon after destroy_qp, IB stack
will cancel the CQ work before invoking the destroy_cq verb and
this will prevent any race mentioned.
Fixes: 1ac5a40479 ("RDMA/bnxt_re: Add bnxt_re RoCE driver")
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1689322969-25402-2-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dc52aadbc1 ]
Commit 21c2fe94ab ("RDMA/mthca: Combine special QP struct with mthca QP")
introduced a new struct mthca_sqp which doesn't contain struct mthca_qp
any longer. Placing a pointer of this new struct into qptable leads
to crashes, because mthca_poll_one() expects a qp pointer. Fix this
by putting the correct pointer into qptable.
Fixes: 21c2fe94ab ("RDMA/mthca: Combine special QP struct with mthca QP")
Signed-off-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Link: https://lore.kernel.org/r/20230713141658.9426-1-tbogendoerfer@suse.de
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
