commit dfd6799304 upstream.
Resource sizes are a general concept for dealing with physical
addresses, and not specific to the Resource type, which is just one way
to access physical addresses. Thus, move the typedef to the io module.
Still keep a re-export under resource. This avoids this commit from
being a flag-day, but I also think it's a useful re-export in general so
that you can import
use kernel::io::resource::{Resource, ResourceSize};
instead of having to write
use kernel::io::{
resource::Resource,
ResourceSize,
};
in the specific cases where you need ResourceSize because you are using
the Resource type. Therefore I think it makes sense to keep this
re-export indefinitely and it is *not* intended as a temporary re-export
for migration purposes.
Cc: stable@vger.kernel.org # for v6.18 [1]
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/20251112-resource-phys-typedefs-v2-2-538307384f82@google.com
Link: https://lore.kernel.org/all/20251112-resource-phys-typedefs-v2-0-538307384f82@google.com/ [1]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e2f1081ca8 upstream.
MMIO backend of PCI Bar always assumes little-endian devices and
will convert to CPU endianness automatically. Remove the u32::from_le
conversion which would cause a bug on big-endian machines.
Cc: stable@vger.kernel.org
Reviewed-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Marko Turk <mt@markoturk.info>
Fixes: 685376d18e ("samples: rust: add Rust PCI sample driver")
Link: https://patch.msgid.link/20251210112503.62925-2-mt@markoturk.info
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d8932355f8 upstream.
Add dma_set_mask(), dma_set_coherent_mask(), dma_map_sgtable(), and
dma_max_mapping_size() helpers to fix a build error when
CONFIG_HAS_DMA is not enabled.
Note that when CONFIG_HAS_DMA is enabled, they are included in both
bindings_generated.rs and bindings_helpers_generated.rs. The former
takes precedence so behavior remains unchanged in that case.
This fixes the following build error on UML:
error[E0425]: cannot find function `dma_set_mask` in crate `bindings`
--> rust/kernel/dma.rs:46:38
|
46 | to_result(unsafe { bindings::dma_set_mask(self.as_ref().as_raw(), mask.value()) })
| ^^^^^^^^^^^^ help: a function with a similar name exists: `xa_set_mark`
|
::: rust/bindings/bindings_generated.rs:24690:5
|
24690 | pub fn xa_set_mark(arg1: *mut xarray, index: ffi::c_ulong, arg2: xa_mark_t);
| ---------------------------------------------------------------------------- similarly named function `xa_set_mark` defined here
error[E0425]: cannot find function `dma_set_coherent_mask` in crate `bindings`
--> rust/kernel/dma.rs:63:38
|
63 | to_result(unsafe { bindings::dma_set_coherent_mask(self.as_ref().as_raw(), mask.value()) })
| ^^^^^^^^^^^^^^^^^^^^^ help: a function with a similar name exists: `dma_coherent_ok`
|
::: rust/bindings/bindings_generated.rs:52745:5
|
52745 | pub fn dma_coherent_ok(dev: *mut device, phys: phys_addr_t, size: usize) -> bool_;
| ---------------------------------------------------------------------------------- similarly named function `dma_coherent_ok` defined here
error[E0425]: cannot find function `dma_map_sgtable` in crate `bindings`
--> rust/kernel/scatterlist.rs:212:23
|
212 | bindings::dma_map_sgtable(dev.as_raw(), sgt.as_ptr(), dir.into(), 0)
| ^^^^^^^^^^^^^^^ help: a function with a similar name exists: `dma_unmap_sgtable`
|
::: rust/bindings/bindings_helpers_generated.rs:1351:5
|
1351 | / pub fn dma_unmap_sgtable(
1352 | | dev: *mut device,
1353 | | sgt: *mut sg_table,
1354 | | dir: dma_data_direction,
1355 | | attrs: ffi::c_ulong,
1356 | | );
| |______- similarly named function `dma_unmap_sgtable` defined here
error[E0425]: cannot find function `dma_max_mapping_size` in crate `bindings`
--> rust/kernel/scatterlist.rs:356:52
|
356 | let max_segment = match unsafe { bindings::dma_max_mapping_size(dev.as_raw()) } {
| ^^^^^^^^^^^^^^^^^^^^ not found in `bindings`
error: aborting due to 4 previous errors
Cc: stable@vger.kernel.org # v6.17+
Fixes: 101d66828a ("rust: dma: add DMA addressing capabilities")
Signed-off-by: FUJITA Tomonori <fujita.tomonori@gmail.com>
Reviewed-by: David Gow <davidgow@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/20251204160639.364936-1-fujita.tomonori@gmail.com
[ Use relative paths in the error splat; add 'dma' prefix. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6c37bebd8c upstream.
Similar to the previous commit, List::remove is used on
delivered_deaths, so do not use mem::take on it as that may result in
violations of the List::remove safety requirements.
I don't think this particular case can be triggered because it requires
fd close to run in parallel with an ioctl on the same fd. But let's not
tempt fate.
Cc: stable@vger.kernel.org
Fixes: eafedbc7c0 ("rust_binder: add Rust Binder driver")
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://patch.msgid.link/20251111-binder-fix-list-remove-v1-2-8ed14a0da63d@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e54ad0cd36 upstream.
Invariants should be prefixed with a # to turn it into a header.
There are no functional changes in this patch.
Cc: stable@vger.kernel.org
Fixes: c284d3e423 ("rust: drm: gem: Add GEM object abstraction")
Signed-off-by: Lyude Paul <lyude@redhat.com>
Link: https://patch.msgid.link/20251107202603.465932-1-lyude@redhat.com
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit cb6d5aa9c0 ]
In smb3_reconfigure(), if smb3_sync_session_ctx_passwords() fails, the
function returns immediately without freeing and erasing the newly
allocated new_password and new_password2. This causes both a memory leak
and a potential information leak.
Fix this by calling kfree_sensitive() on both password buffers before
returning in this error case.
Fixes: 0f0e357902 ("cifs: during remount, make sure passwords are in sync")
Signed-off-by: Zilin Guan <zilin@seu.edu.cn>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d8ee3cfdc8 ]
vhost_vsock_get() uses hash_for_each_possible_rcu() to find the
`vhost_vsock` associated with the `guest_cid`. hash_for_each_possible_rcu()
should only be called within an RCU read section, as mentioned in the
following comment in include/linux/rculist.h:
/**
* hlist_for_each_entry_rcu - iterate over rcu list of given type
* @pos: the type * to use as a loop cursor.
* @head: the head for your list.
* @member: the name of the hlist_node within the struct.
* @cond: optional lockdep expression if called from non-RCU protection.
*
* This list-traversal primitive may safely run concurrently with
* the _rcu list-mutation primitives such as hlist_add_head_rcu()
* as long as the traversal is guarded by rcu_read_lock().
*/
Currently, all calls to vhost_vsock_get() are between rcu_read_lock()
and rcu_read_unlock() except for calls in vhost_vsock_set_cid() and
vhost_vsock_reset_orphans(). In both cases, the current code is safe,
but we can make improvements to make it more robust.
About vhost_vsock_set_cid(), when building the kernel with
CONFIG_PROVE_RCU_LIST enabled, we get the following RCU warning when the
user space issues `ioctl(dev, VHOST_VSOCK_SET_GUEST_CID, ...)` :
WARNING: suspicious RCU usage
6.18.0-rc7 #62 Not tainted
-----------------------------
drivers/vhost/vsock.c:74 RCU-list traversed in non-reader section!!
other info that might help us debug this:
rcu_scheduler_active = 2, debug_locks = 1
1 lock held by rpc-libvirtd/3443:
#0: ffffffffc05032a8 (vhost_vsock_mutex){+.+.}-{4:4}, at: vhost_vsock_dev_ioctl+0x2ff/0x530 [vhost_vsock]
stack backtrace:
CPU: 2 UID: 0 PID: 3443 Comm: rpc-libvirtd Not tainted 6.18.0-rc7 #62 PREEMPT(none)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.17.0-7.fc42 06/10/2025
Call Trace:
<TASK>
dump_stack_lvl+0x75/0xb0
dump_stack+0x14/0x1a
lockdep_rcu_suspicious.cold+0x4e/0x97
vhost_vsock_get+0x8f/0xa0 [vhost_vsock]
vhost_vsock_dev_ioctl+0x307/0x530 [vhost_vsock]
__x64_sys_ioctl+0x4f2/0xa00
x64_sys_call+0xed0/0x1da0
do_syscall_64+0x73/0xfa0
entry_SYSCALL_64_after_hwframe+0x76/0x7e
...
</TASK>
This is not a real problem, because the vhost_vsock_get() caller, i.e.
vhost_vsock_set_cid(), holds the `vhost_vsock_mutex` used by the hash
table writers. Anyway, to prevent that warning, add lockdep_is_held()
condition to hash_for_each_possible_rcu() to verify that either the
caller is in an RCU read section or `vhost_vsock_mutex` is held when
CONFIG_PROVE_RCU_LIST is enabled; and also clarify the comment for
vhost_vsock_get() to better describe the locking requirements and the
scope of the returned pointer validity.
About vhost_vsock_reset_orphans(), currently this function is only
called via vsock_for_each_connected_socket(), which holds the
`vsock_table_lock` spinlock (which is also an RCU read-side critical
section). However, add an explicit RCU read lock there to make the code
more robust and explicit about the RCU requirements, and to prevent
issues if the calling context changes in the future or if
vhost_vsock_reset_orphans() is called from other contexts.
Fixes: 834e772c8d ("vhost/vsock: fix use-after-free in network stack callers")
Cc: stefanha@redhat.com
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20251126133826.142496-1-sgarzare@redhat.com>
Message-ID: <20251126210313.GA499503@fedora>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1ddb815fdf ]
The "dev->clt_device_id" variable is set using ida_alloc_max() which
returns an int and in particular it returns negative error codes.
Change the type from u32 to int to fix the error checking.
Fixes: c9b5645fd8 ("block: rnbd-clt: Fix leaked ID in init_dev()")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit daa24603d9 ]
If a ublk server process releases a ublk char device file, any requests
dispatched to the ublk server but not yet completed will retain a ref
value of UBLK_REFCOUNT_INIT. Before commit e63d2228ef ("ublk: simplify
aborting ublk request"), __ublk_fail_req() would decrement the reference
count before completing the failed request. However, that commit
optimized __ublk_fail_req() to call __ublk_complete_rq() directly
without decrementing the request reference count.
The leaked reference count incorrectly allows user copy and zero copy
operations on the completed ublk request. It also triggers the
WARN_ON_ONCE(refcount_read(&io->ref)) warnings in ublk_queue_reinit()
and ublk_deinit_queue().
Commit c5c5eb24ed ("ublk: avoid ublk_io_release() called after ublk
char dev is closed") already fixed the issue for ublk devices using
UBLK_F_SUPPORT_ZERO_COPY or UBLK_F_AUTO_BUF_REG. However, the reference
count leak also affects UBLK_F_USER_COPY, the other reference-counted
data copy mode. Fix the condition in ublk_check_and_reset_active_ref()
to include all reference-counted data copy modes. This ensures that any
ublk requests still owned by the ublk server when it exits have their
reference counts reset to 0.
Signed-off-by: Caleb Sander Mateos <csander@purestorage.com>
Fixes: e63d2228ef ("ublk: simplify aborting ublk request")
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ef3b04091f ]
Move the call to preempt_prepare_postamble() after verifying that
preempt_postamble_ptr is valid. If preempt_postamble_ptr is NULL,
dereferencing it in preempt_prepare_postamble() would lead to a crash.
This change avoids calling the preparation function when the
postamble allocation has failed, preventing potential NULL pointer
dereference and ensuring proper error handling.
Fixes: 50117cad0c ("drm/msm/a6xx: Use posamble to reset counters on preemption")
Signed-off-by: Alok Tiwari <alok.a.tiwari@oracle.com>
Patchwork: https://patchwork.freedesktop.org/patch/687659/
Message-ID: <20251113082839.3821867-1-alok.a.tiwari@oracle.com>
Signed-off-by: Rob Clark <robin.clark@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 129049d4fe ]
On plaforms with an a7xx GPU not supporting IFPC, the ifpc_reglist
if still deferenced in a7xx_patch_pwrup_reglist() which causes
a kernel crash:
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
...
pc : a6xx_hw_init+0x155c/0x1e4c [msm]
lr : a6xx_hw_init+0x9a8/0x1e4c [msm]
...
Call trace:
a6xx_hw_init+0x155c/0x1e4c [msm] (P)
msm_gpu_hw_init+0x58/0x88 [msm]
adreno_load_gpu+0x94/0x1fc [msm]
msm_open+0xe4/0xf4 [msm]
drm_file_alloc+0x1a0/0x2e4 [drm]
drm_client_init+0x7c/0x104 [drm]
drm_fbdev_client_setup+0x94/0xcf0 [drm_client_lib]
drm_client_setup+0xb4/0xd8 [drm_client_lib]
msm_drm_kms_post_init+0x2c/0x3c [msm]
msm_drm_init+0x1a4/0x228 [msm]
msm_drm_bind+0x30/0x3c [msm]
...
Check the validity of ifpc_reglist before deferencing the table
to setup the register values.
Fixes: a6a0157cc6 ("drm/msm/a6xx: Enable IFPC on Adreno X1-85")
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Akhil P Oommen <akhilpo@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/688944/
Message-ID: <20251117-topic-sm8x50-fix-a6xx-non-ifpc-v1-1-e4473cbf5903@linaro.org>
Signed-off-by: Rob Clark <robin.clark@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1f7d6e2efe ]
The atomic write enable module param is "atomic_wr", and not
"atomic_write", so fix the module param description.
Fixes: 84f3a3c01d ("scsi: scsi_debug: Atomic write support")
Signed-off-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://patch.msgid.link/20251211100651.9056-1-john.g.garry@oracle.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 36dac9a3dd ]
Since commit e424054000 ("MIPS: Tracing: Reduce the overhead of
dynamic Function Tracer"), the macro UASM_i_LA_mostly has been used,
and this macro can generate more than 2 instructions. At the same
time, the code in ftrace assumes that no more than 2 instructions can
be generated, which is why it stores them in an int[2] array. However,
as previously noted, the macro UASM_i_LA_mostly (and now UASM_i_LA)
causes a buffer overflow when _mcount is beyond 32 bits. This leads to
corruption of the variables located in the __read_mostly section.
This corruption was observed because the variable
__cpu_primary_thread_mask was corrupted, causing a hang very early
during boot.
This fix prevents the corruption by avoiding the generation of
instructions if they could exceed 2 instructions in
length. Fortunately, insn_la_mcount is only used if the instrumented
code is located outside the kernel code section, so dynamic ftrace can
still be used, albeit in a more limited scope. This is still
preferable to corrupting memory and/or crashing the kernel.
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b169e1733c ]
Dell Pro Rugged 10/12 tablets has a reliable VGBS method.
If VGBS is not called on boot, the on-screen keyboard won't appear if the
device is booted without a keyboard.
Call VGBS on boot on thess devices to get the initial state of
SW_TABLET_MODE in a reliable way.
Signed-off-by: Chia-Lin Kao (AceLan) <acelan.kao@canonical.com>
Reviewed-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Link: https://patch.msgid.link/20251127070407.656463-1-acelan.kao@canonical.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 541dfb49dc ]
./drivers/hwmon/emc2305.c:597:4-15: ERROR: probable double put
Device node iterators put the previous value of the index variable, so an
explicit put causes a double put.
Signed-off-by: Pei Xiao <xiaopei01@kylinos.cn>
Link: https://lore.kernel.org/r/tencent_CD373F952BE48697C949E39CB5EB77841D06@qq.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 13989207ee ]
With authentication, in addition to EKEYREJECTED there is also no point in
retrying reconnects when status is ENOKEY. Thus, add -ENOKEY as another
criteria to determine when to stop retries.
Cc: Daniel Wagner <wagi@kernel.org>
Cc: Hannes Reinecke <hare@suse.de>
Closes: https://lore.kernel.org/linux-nvme/20250829-nvme-fc-sync-v3-0-d69c87e63aee@kernel.org/
Signed-off-by: Justin Tee <justintee8345@gmail.com>
Tested-by: Daniel Wagner <wagi@kernel.org>
Reviewed-by: Daniel Wagner <wagi@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4910da6b36 ]
The for_each_child_of_node() macro automatically manages device node
reference counts during normal iteration. However, when breaking out
of the loop early with return, the current iteration's node is not
automatically released, leading to a reference count leak.
Fix this by adding of_node_put(child) before returning from the loop
when emc2305_set_single_tz() fails.
This issue could lead to memory leaks over multiple probe cycles.
Signed-off-by: Pei Xiao <xiaopei01@kylinos.cn>
Link: https://lore.kernel.org/r/tencent_5CDC08544C901D5ECA270573D5AEE3117108@qq.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b71cbcf7d1 ]
nvme_fc_ctrl_put can acquire the rport lock when freeing the
ctrl object:
nvme_fc_ctrl_put
nvme_fc_ctrl_free
spin_lock_irqsave(rport->lock)
Thus we can't hold the rport lock when calling nvme_fc_ctrl_put.
Justin suggested use the safe list iterator variant because
nvme_fc_ctrl_put will also modify the rport->list.
Cc: Justin Tee <justin.tee@broadcom.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Daniel Wagner <wagi@kernel.org>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 55715d7ad5 ]
Add Legion Go 2 SKU's to the Extreme Mode quirks table.
Signed-off-by: Derek J. Clark <derekjohn.clark@gmail.com>
Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Link: https://patch.msgid.link/20251127151605.1018026-4-derekjohn.clark@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d3429178ee ]
When probing the I2C master, disable SMBus interrupts to prevent
storms caused by broken firmware mis-configuring IC_SMBUS=1; the
handler never services them and a mis-configured SMBUS Master
extend-clock timeout or SMBUS Slave extend-clock timeout can
flood the CPU.
Signed-off-by: Jinhui Guo <guojinhui.liam@bytedance.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://lore.kernel.org/r/20251021075714.3712-2-guojinhui.liam@bytedance.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 29e8a0c587 ]
In sprd_clk_init(), when devm_clk_get() returns -EPROBE_DEFER
for either uart or source clock, we should propagate the
error instead of just warning and continuing with NULL clocks.
Currently the driver only emits a warning when clock acquisition
fails and proceeds with NULL clock pointers. This can lead to
issues later when the clocks are actually needed. More importantly,
when the clock provider is not ready yet and returns -EPROBE_DEFER,
we should return this error to allow deferred probing.
This change adds explicit checks for -EPROBE_DEFER after both:
1. devm_clk_get(uport->dev, uart)
2. devm_clk_get(uport->dev, source)
When -EPROBE_DEFER is encountered, the function now returns
-EPROBE_DEFER to let the driver framework retry probing
later when the clock dependencies are resolved.
Signed-off-by: Wenhua Lin <Wenhua.Lin@unisoc.com>
Link: https://patch.msgid.link/20251022030840.956589-1-Wenhua.Lin@unisoc.com
Reviewed-by: Cixi Geng <cixi.geng@linux.dev>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e6aec6d9f5 ]
Some old HCs ignore transfer ring link TRBs whose chain bit is unset.
This breaks endpoint operation and sometimes makes it execute other
ring's TDs, which may corrupt their buffers or cause unwanted device
action. We avoid this by chaining all link TRBs on affected rings.
Fix an omission which allows them to be unchained by cancelling TDs.
The patch was tested by reproducing this condition on an isochronous
endpoint (non-power-of-two TDs are sometimes split not to cross 64K)
and printing link TRBs in trb_to_noop() on good and buggy HCs.
Actual hardware malfunction is rare since it requires Missed Service
Error shortly before the unchained link TRB, at least on NEC and AMD.
I have never seen it after commit bb0ba4cb10 ("usb: xhci: Apply the
link chain quirk on NEC isoc endpoints"), but it's Russian roulette
and I can't test all affected hosts and workloads. Fairly often MSEs
happen after cancellation because the endpoint was stopped.
Signed-off-by: Michal Pecio <michal.pecio@gmail.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://patch.msgid.link/20251119142417.2820519-11-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 955a48a535 ]
The optical drive of EL-R12 has the same vid and pid as INIC-3069,
as follows:
T: Bus=02 Lev=02 Prnt=02 Port=01 Cnt=01 Dev#= 3 Spd=5000 MxCh= 0
D: Ver= 3.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1
P: Vendor=13fd ProdID=3940 Rev= 3.10
S: Manufacturer=HL-DT-ST
S: Product= DVD+-RW GT80N
S: SerialNumber=423349524E4E38303338323439202020
C:* #Ifs= 1 Cfg#= 1 Atr=80 MxPwr=144mA
I:* If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=02 Prot=50 Driver=usb-storage
E: Ad=83(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms
E: Ad=0a(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms
This will result in the optical drive device also adding
the quirks of US_FL_NO_ATA_1X. When performing an erase operation,
it will fail, and the reason for the failure is as follows:
[ 388.967742] sr 5:0:0:0: [sr0] tag#0 Send: scmd 0x00000000d20c33a7
[ 388.967742] sr 5:0:0:0: [sr0] tag#0 CDB: ATA command pass through(12)/Blank a1 11 00 00 00 00 00 00 00 00 00 00
[ 388.967773] sr 5:0:0:0: [sr0] tag#0 Done: SUCCESS Result: hostbyte=DID_TARGET_FAILURE driverbyte=DRIVER_OK cmd_age=0s
[ 388.967773] sr 5:0:0:0: [sr0] tag#0 CDB: ATA command pass through(12)/Blank a1 11 00 00 00 00 00 00 00 00 00 00
[ 388.967803] sr 5:0:0:0: [sr0] tag#0 Sense Key : Illegal Request [current]
[ 388.967803] sr 5:0:0:0: [sr0] tag#0 Add. Sense: Invalid field in cdb
[ 388.967803] sr 5:0:0:0: [sr0] tag#0 scsi host busy 1 failed 0
[ 388.967803] sr 5:0:0:0: Notifying upper driver of completion (result 8100002)
[ 388.967834] sr 5:0:0:0: [sr0] tag#0 0 sectors total, 0 bytes done.
For the EL-R12 standard optical drive, all operational commands
and usage scenarios were tested without adding the IGNORE_RESIDUE quirks,
and no issues were encountered. It can be reasonably concluded
that removing the IGNORE_RESIDUE quirks has no impact.
Signed-off-by: Chen Changcheng <chenchangcheng@kylinos.cn>
Link: https://patch.msgid.link/20251121064020.29332-1-chenchangcheng@kylinos.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c9fb952360 ]
FIELD_PREP() checks that a value fits into the available bitfield, add a
check for step_avg to fix gcc complains.
which gcc complains about:
drivers/iio/adc/ti_am335x_adc.c: In function 'tiadc_step_config':
include/linux/compiler_types.h:572:38: error: call to
'__compiletime_assert_491' declared with attribute error: FIELD_PREP: value
too large for the field include/linux/mfd/ti_am335x_tscadc.h:58:29: note:
in expansion of macro 'FIELD_PREP'
#define STEPCONFIG_AVG(val) FIELD_PREP(GENMASK(4, 2), (val))
^~~~~~~~~~
drivers/iio/adc/ti_am335x_adc.c:127:17: note: in expansion of macro 'STEPCONFIG_AVG'
stepconfig = STEPCONFIG_AVG(ffs(adc_dev->step_avg[i]) - 1)
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202510102117.Jqxrw1vF-lkp@intel.com/
Signed-off-by: Pei Xiao <xiaopei01@kylinos.cn>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 30cd2cb1ab ]
The UCSI spec states that the num_connectors field is 7 bits, and the
8th bit is reserved and should be set to zero.
Some buggy FW has been known to set this bit, and it can lead to a
system not booting.
Flag that the FW is not behaving correctly, and auto-fix the value
so that the system boots correctly.
Found on Lenovo P1 G8 during Linux enablement program. The FW will
be fixed, but seemed worth addressing in case it hit platforms that
aren't officially Linux supported.
Signed-off-by: Mark Pearson <mpearson-lenovo@squebb.ca>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20250821185319.2585023-1-mpearson-lenovo@squebb.ca
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 09bf21bf52 ]
Interrupts are disabled before entering usb_hcd_giveback_urb().
A spinlock_t becomes a sleeping lock on PREEMPT_RT, so it cannot be
acquired with disabled interrupts.
Save the interrupt status and restore it after usb_hcd_giveback_urb().
syz reported:
BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
Call Trace:
dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
rt_spin_lock+0xc7/0x2c0 kernel/locking/spinlock_rt.c:57
spin_lock include/linux/spinlock_rt.h:44 [inline]
mon_bus_complete drivers/usb/mon/mon_main.c:134 [inline]
mon_complete+0x5c/0x200 drivers/usb/mon/mon_main.c:147
usbmon_urb_complete include/linux/usb/hcd.h:738 [inline]
__usb_hcd_giveback_urb+0x254/0x5e0 drivers/usb/core/hcd.c:1647
vhci_urb_enqueue+0xb4f/0xe70 drivers/usb/usbip/vhci_hcd.c:818
Reported-by: syzbot+205ef33a3b636b4181fb@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=205ef33a3b636b4181fb
Signed-off-by: Lizhi Xu <lizhi.xu@windriver.com>
Acked-by: Shuah Khan <skhan@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250916014143.1439759-1-lizhi.xu@windriver.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4e163c39dd ]
xfstests generic/363 was failing due to unzeroed post-EOF page
cache that allowed mmap writes beyond EOF to become visible
after file extension.
For example, in following xfs_io sequence, 0x22 should not be
written to the file but would become visible after the extension:
xfs_io -f -t -c "pwrite -S 0x11 0 8" \
-c "mmap 0 4096" \
-c "mwrite -S 0x22 32 32" \
-c "munmap" \
-c "pwrite -S 0x33 512 32" \
$testfile
This violates the expected behavior where writes beyond EOF via
mmap should not persist after the file is extended. Instead, the
extended region should contain zeros.
Fix this by using truncate_pagecache() to truncate the page cache
after the current EOF when extending the file.
Signed-off-by: Yuezhang Mo <Yuezhang.Mo@sony.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 51fc7b4ce1 ]
The kernel test robot reported that the exFAT remount operation
failed. The reason for the failure was that the process's umask
is different between mount and remount, causing fs_fmask and
fs_dmask are changed.
Potentially, both gid and uid may also be changed. Therefore, when
initializing fs_context for remount, inherit these mount options
from the options used during mount.
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202511251637.81670f5c-lkp@intel.com
Signed-off-by: Yuezhang Mo <Yuezhang.Mo@sony.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f3d8b64ee4 ]
RESET_CONTROL_FLAGS_BIT_* macros use BIT(), but reset.h does not
include bits.h. This causes compilation errors when including
reset.h standalone.
Include bits.h to make reset.h self-contained.
Suggested-by: Troy Mitchell <troy.mitchell@linux.dev>
Reviewed-by: Troy Mitchell <troy.mitchell@linux.dev>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Encrow Thorne <jyc0019@gmail.com>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e5bf5ee266 ]
ffs_epfile_open() can race with removal, ending up with file->private_data
pointing to freed object.
There is a total count of opened files on functionfs (both ep0 and
dynamic ones) and when it hits zero, dynamic files get removed.
Unfortunately, that removal can happen while another thread is
in ffs_epfile_open(), but has not incremented the count yet.
In that case open will succeed, leaving us with UAF on any subsequent
read() or write().
The root cause is that ffs->opened is misused; atomic_dec_and_test() vs.
atomic_add_return() is not a good idea, when object remains visible all
along.
To untangle that
* serialize openers on ffs->mutex (both for ep0 and for dynamic files)
* have dynamic ones use atomic_inc_not_zero() and fail if we had
zero ->opened; in that case the file we are opening is doomed.
* have the inodes of dynamic files marked on removal (from the
callback of simple_recursive_removal()) - clear ->i_private there.
* have open of dynamic ones verify they hadn't been already removed,
along with checking that state is FFS_ACTIVE.
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7aa31ee9ec ]
The VIA watchdog driver uses allocate_resource() to reserve a MMIO
region for the watchdog control register. However, the allocated
resource was not given a name, which causes the kernel resource tree
to contain an entry marked as "<BAD>" under /proc/iomem on x86
platforms.
During boot, this unnamed resource can lead to a critical hang because
subsequent resource lookups and conflict checks fail to handle the
invalid entry properly.
Signed-off-by: Li Qiang <liqiang01@kylinos.cn>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b359af8275 ]
generic_file_direct_write() also does this and has a large
comment about.
Reproducer here is xfstest's generic/209, which is exactly to
have competing DIO write and cached IO read.
Signed-off-by: Bernd Schubert <bschubert@ddn.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1ce120dcef ]
This was done as condition on direct_io_allow_mmap, but I believe
this is not right, as a file might be open two times - once with
write-back enabled another time with FOPEN_DIRECT_IO.
Signed-off-by: Bernd Schubert <bschubert@ddn.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 957aa59749 ]
If a mailbox command completes immediately after
wait_for_completion_timeout() times out, ha->mbx_intr_comp could be left
in an inconsistent state, causing the next mailbox command not to wait
for the hardware. Fix by reinitializing the completion before use.
Signed-off-by: Tony Battersby <tonyb@cybernetics.com>
Link: https://patch.msgid.link/11b6485e-0bfd-4784-8f99-c06a196dad94@cybernetics.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8f58fc64d5 ]
When given the module parameter qlini_mode=exclusive, qla2xxx in
initiator mode is initially unable to successfully send SCSI commands to
devices it finds while scanning, resulting in an escalating series of
resets until an adapter reset clears the issue. Fix by checking the
active mode instead of the module parameter.
Signed-off-by: Tony Battersby <tonyb@cybernetics.com>
Link: https://patch.msgid.link/1715ec14-ba9a-45dc-9cf2-d41aa6b81b5e@cybernetics.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4f6aaade2a ]
When qla2xxx is loaded with qlini_mode=disabled,
ha->flags.disable_msix_handshake is used before it is set, resulting in
the wrong interrupt handler being used on certain HBAs
(qla2xxx_msix_rsp_q_hs() is used when qla2xxx_msix_rsp_q() should be
used). The only difference between these two interrupt handlers is that
the _hs() version writes to a register to clear the "RISC" interrupt,
whereas the other version does not. So this bug results in the RISC
interrupt being cleared when it should not be. This occasionally causes
a different interrupt handler qla24xx_msix_default() for a different
vector to see ((stat & HSRX_RISC_INT) == 0) and ignore its interrupt,
which then causes problems like:
qla2xxx [0000:02:00.0]-d04c:6: MBX Command timeout for cmd 20,
iocontrol=8 jiffies=1090c0300 mb[0-3]=[0x4000 0x0 0x40 0xda] mb7 0x500
host_status 0x40000010 hccr 0x3f00
qla2xxx [0000:02:00.0]-101e:6: Mailbox cmd timeout occurred, cmd=0x20,
mb[0]=0x20. Scheduling ISP abort
(the cmd varies; sometimes it is 0x20, 0x22, 0x54, 0x5a, 0x5d, or 0x6a)
This problem can be reproduced with a 16 or 32 Gbps HBA by loading
qla2xxx with qlini_mode=disabled and running a high IOPS test while
triggering frequent RSCN database change events.
While analyzing the problem I discovered that even with
disable_msix_handshake forced to 0, it is not necessary to clear the
RISC interrupt from qla2xxx_msix_rsp_q_hs() (more below). So just
completely remove qla2xxx_msix_rsp_q_hs() and the logic for selecting
it, which also fixes the bug with qlini_mode=disabled.
The test below describes the justification for not needing
qla2xxx_msix_rsp_q_hs():
Force disable_msix_handshake to 0:
qla24xx_config_rings():
if (0 && (ha->fw_attributes & BIT_6) && (IS_MSIX_NACK_CAPABLE(ha)) &&
(ha->flags.msix_enabled)) {
In qla24xx_msix_rsp_q() and qla2xxx_msix_rsp_q_hs(), check:
(rd_reg_dword(®->host_status) & HSRX_RISC_INT)
Count the number of calls to each function with HSRX_RISC_INT set and
the number with HSRX_RISC_INT not set while performing some I/O.
If qla2xxx_msix_rsp_q_hs() clears the RISC interrupt (original code):
qla24xx_msix_rsp_q: 50% of calls have HSRX_RISC_INT set
qla2xxx_msix_rsp_q_hs: 5% of calls have HSRX_RISC_INT set
(# of qla2xxx_msix_rsp_q_hs interrupts) =
(# of qla24xx_msix_rsp_q interrupts) * 3
If qla2xxx_msix_rsp_q_hs() does not clear the RISC interrupt (patched
code):
qla24xx_msix_rsp_q: 100% of calls have HSRX_RISC_INT set
qla2xxx_msix_rsp_q_hs: 9% of calls have HSRX_RISC_INT set
(# of qla2xxx_msix_rsp_q_hs interrupts) =
(# of qla24xx_msix_rsp_q interrupts) * 3
In the case of the original code, qla24xx_msix_rsp_q() was seeing
HSRX_RISC_INT set only 50% of the time because qla2xxx_msix_rsp_q_hs()
was clearing it when it shouldn't have been. In the patched code,
qla24xx_msix_rsp_q() sees HSRX_RISC_INT set 100% of the time, which
makes sense if that interrupt handler needs to clear the RISC interrupt
(which it does). qla2xxx_msix_rsp_q_hs() sees HSRX_RISC_INT only 9% of
the time, which is just overlap from the other interrupt during the
high IOPS test.
Tested with SCST on:
QLE2742 FW:v9.08.02 (32 Gbps 2-port)
QLE2694L FW:v9.10.11 (16 Gbps 4-port)
QLE2694L FW:v9.08.02 (16 Gbps 4-port)
QLE2672 FW:v8.07.12 (16 Gbps 2-port)
both initiator and target mode
Signed-off-by: Tony Battersby <tonyb@cybernetics.com>
Link: https://patch.msgid.link/56d378eb-14ad-49c7-bae9-c649b6c7691e@cybernetics.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 825ce89a3e ]
The PUT_64[LB]E() macros need to cast the value to unsigned long long
like the GET_64[LB]E() macros. Caused lots of warnings when compiled
on 32-bit, and clobbered addresses (36-bit P4080).
Signed-off-by: Ben Collins <bcollins@kernel.org>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Madhavan Srinivasan <maddy@linux.ibm.com>
Link: https://patch.msgid.link/2025042122-mustard-wrasse-694572@boujee-and-buff
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f0e6bc0c3e ]
CP110 based platforms rely on the bootloader for pci port
initialization.
TF-A actively prevents non-uboot re-configuration of pci lanes, and many
boards do not have software control over the pci card reset.
If a pci port had link at boot-time and the clock is stopped at a later
point, the link fails and can not be recovered.
PCI controller driver probe - and by extension ownership of a driver for
the pci clocks - may be delayed especially on large modular kernels,
causing the clock core to start disabling unused clocks.
Add the CLK_IGNORE_UNUSED flag to the three pci port's clocks to ensure
they are not stopped before the pci controller driver has taken
ownership and tested for an existing link.
This fixes failed pci link detection when controller driver probes late,
e.g. with arm64 defconfig and CONFIG_PHY_MVEBU_CP110_COMPHY=m.
Closes: https://lore.kernel.org/r/b71596c7-461b-44b6-89ab-3cfbd492639f@solid-run.com
Signed-off-by: Josua Mayer <josua@solid-run.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 07caedc6a3 ]
It's possible for an unstable link to repeatedly bounce allowing a FLOGI
retry, but then bounce again forcing an abort of the FLOGI. Ensure that
the initial reference count on the FLOGI ndlp is restored in this faulty
link scenario.
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://patch.msgid.link/20251106224639.139176-8-justintee8345@gmail.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 48e6b7e708 ]
Add support for new Hurray Data controller.
All entries are in HEX.
Add PCI IDs for Hurray Data controllers:
VID / DID / SVID / SDID
---- ---- ---- ----
9005 028f 207d 4840
Reviewed-by: Scott Benesh <scott.benesh@microchip.com>
Reviewed-by: Scott Teel <scott.teel@microchip.com>
Reviewed-by: Mike McGowen <mike.mcgowen@microchip.com>
Signed-off-by: David Strahan <David.Strahan@microchip.com>
Signed-off-by: Don Brace <don.brace@microchip.com>
Link: https://patch.msgid.link/20251106163823.786828-4-don.brace@microchip.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3f61783920 ]
am33xx.dtsi has the same clock setup as am35xx.dtsi, setting
ti,no-reset-on-init and ti,no-idle on timer1_target and timer2_target,
so AM33 needs the same workaround as AM35 to avoid ti-sysc probe
failing on certain target modules.
Signed-off-by: Matthias Schiffer <matthias.schiffer@tq-group.com>
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
Link: https://lore.kernel.org/r/20250825131114.2206804-1-alexander.stein@ew.tq-group.com
Signed-off-by: Kevin Hilman <khilman@baylibre.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7b5d441696 ]
This is currently done in uml_finishsetup(), but e.g. with
KCOV enabled we'll crash because some init code can call
into e.g. memparse(), which has coverage annotations, and
then the checks in check_kcov_mode() crash because current
is NULL.
Simply initialize the cpu_tasks[] array statically, which
fixes the crash. For the later SMP work, it seems to have
not really caused any problems yet, but initialize all of
the entries anyway.
Link: https://patch.msgid.link/20250924113214.c76cd74d0583.I974f691ebb1a2b47915bd2b04cc38e5263b9447f@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 81fb53feb6 ]
With mailbox channel requested, there is possibility that interrupts may
come in, so need to make sure the workqueue is initialized before
the queue is scheduled by mailbox rx callback.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 014de20bb3 ]
Address a race condition between shutdown and suspend operations in the
UFS Mediatek driver. Before entering suspend, check if a shutdown is in
progress to prevent conflicts and ensure system stability.
Signed-off-by: Peter Wang <peter.wang@mediatek.com>
Acked-by: Chun-Hung Wu <chun-hung.wu@mediatek.com>
Link: https://patch.msgid.link/20250924094527.2992256-6-peter.wang@mediatek.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6bd30d8fc5 ]
channel_handler() sets intf->channels_ready to true but never
clears it, so __scan_channels() skips any rescan. When the BMC
firmware changes a rescan is required. Allow it by clearing
the flag before starting a new scan.
Signed-off-by: Jinhui Guo <guojinhui.liam@bytedance.com>
Message-ID: <20250930074239.2353-3-guojinhui.liam@bytedance.com>
Signed-off-by: Corey Minyard <corey@minyard.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 936750fdba ]
The race window between __scan_channels() and deliver_response() causes
the parameters of some channels to be set to 0.
1.[CPUA] __scan_channels() issues an IPMI request and waits with
wait_event() until all channels have been scanned.
wait_event() internally calls might_sleep(), which might
yield the CPU. (Moreover, an interrupt can preempt
wait_event() and force the task to yield the CPU.)
2.[CPUB] deliver_response() is invoked when the CPU receives the
IPMI response. After processing a IPMI response,
deliver_response() directly assigns intf->wchannels to
intf->channel_list and sets intf->channels_ready to true.
However, not all channels are actually ready for use.
3.[CPUA] Since intf->channels_ready is already true, wait_event()
never enters __wait_event(). __scan_channels() immediately
clears intf->null_user_handler and exits.
4.[CPUB] Once intf->null_user_handler is set to NULL, deliver_response()
ignores further IPMI responses, leaving the remaining
channels zero-initialized and unusable.
CPUA CPUB
------------------------------- -----------------------------
__scan_channels()
intf->null_user_handler
= channel_handler;
send_channel_info_cmd(intf,
0);
wait_event(intf->waitq,
intf->channels_ready);
do {
might_sleep();
deliver_response()
channel_handler()
intf->channel_list =
intf->wchannels + set;
intf->channels_ready = true;
send_channel_info_cmd(intf,
intf->curr_channel);
if (condition)
break;
__wait_event(wq_head,
condition);
} while(0)
intf->null_user_handler
= NULL;
deliver_response()
if (!msg->user)
if (intf->null_user_handler)
rv = -EINVAL;
return rv;
------------------------------- -----------------------------
Fix the race between __scan_channels() and deliver_response() by
deferring both the assignment intf->channel_list = intf->wchannels
and the flag intf->channels_ready = true until all channels have
been successfully scanned or until the IPMI request has failed.
Signed-off-by: Jinhui Guo <guojinhui.liam@bytedance.com>
Message-ID: <20250930074239.2353-2-guojinhui.liam@bytedance.com>
Signed-off-by: Corey Minyard <corey@minyard.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
