]> git.ipfire.org Git - thirdparty/kernel/linux.git/commit
pds_core: fix deadlock between reset thread and remove
authorNikhil P. Rao <nikhil.rao@amd.com>
Tue, 14 Jul 2026 18:02:22 +0000 (18:02 +0000)
committerJakub Kicinski <kuba@kernel.org>
Tue, 21 Jul 2026 19:42:26 +0000 (12:42 -0700)
commitab0eec0ff0a421737a37f510ceab5c6ea59cd05a
tree8b61d95f4ba9c94576af8e219501458dd4bb3da5
parentdefd52c1eab2926d874fa95408187dac6f890ff1
pds_core: fix deadlock between reset thread and remove

pci_reset_function() acquires device_lock before performing the reset.
pdsc_remove() is called by the PCI core with device_lock already held.
If pdsc_pci_reset_thread() is running when pdsc_remove() is called,
destroy_workqueue() will block waiting for the work to complete, while
the work is blocked waiting for device_lock - deadlock.

Use pci_try_reset_function() which uses pci_dev_trylock() internally.
This acquires both the device lock and the PCI config access lock
without blocking - if either lock is contended, it returns -EAGAIN
immediately. This avoids the deadlock while also ensuring proper
config space access serialization during the reset.

The pci_dev_get/put calls are also removed as they were unnecessary -
the driver-owned workqueue is destroyed in pdsc_remove(), guaranteeing
the work completes before remove returns. The PCI core holds its
reference to pci_dev throughout the entire unbind sequence.

Fixes: 81665adf25d2 ("pds_core: Fix pdsc_check_pci_health function to use work thread")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Closes: https://patchwork.kernel.org/comment/27002369/
Signed-off-by: Nikhil P. Rao <nikhil.rao@amd.com>
Reviewed-by: Harshitha Ramamurthy <hramamurthy@google.com>
Link: https://patch.msgid.link/20260714180223.1642792-2-nikhil.rao@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
drivers/net/ethernet/amd/pds_core/core.c