PT-2026-152501
CVE-2026-99621
·
Published
2026-11-20
·
Updated
2026-11-20
Description by
Score
N/A
Not availableNo severity ratings or metrics are available. When they are, we'll update the corresponding info on the page.
In the Linux kernel, the following vulnerability has been resolved:
drm/amdgpu: unlock the VM and BO when attaching a KFD BO fails
kfd mem attach() locks the page directory and the BO with a drm exec
transaction on its stack, but leaves through the unwind label without
drm exec fini() when locking fails, e.g. when MAP MEMORY TO GPU gets a
signal while the page directory or the BO is busy. Reservations already
acquired then remain locked with their ww acquire ctx pointing into the
expired stack frame.
With one thread doing GEM VA map/unmap on the KFD VM and
another one mapping KFD BOs while getting SIGUSR1 every 20 us:
BUG: KASAN: stack-out-of-bounds in ww mutex lock.constprop.0+0x34f5/0x3da0
ww mutex lock interruptible+0x2f/0x150
drm exec lock obj+0x107/0x2d0 [drm exec]
kfd mem attach+0x4dd/0x1b60 [amdgpu]
and the process ends up in D state.
Finish the transaction before unwinding.
(cherry picked from commit 7b885fa27e4b9fa5cbdc64e66fc35872aee296f5)
Found an issue in the description? Have something to add? Feel free to write us 👾