CVE-2026-99631: drm/amdgpu: check amdgpu_vm_ready() in dma-buf move notify
In the Linux kernel, the following vulnerability has been resolved:
drm/amdgpu: check amdgpuvmready() in dma-buf move notify
amdgpudmabufmovenotify() touches the page tables without first calling amdgpuvmready(). When a process is SIGKILLed, its VM entities are stopped. Teardown of an imported dma-buf then tries to clear PTEs by submitting to one of those stopped entities. That eventually hangs the task.
Commit aa5fc4362fac ("drm/amdgpu: fix task hang from failed job submission during process kill") added the stopped entity check in amdgpuvmready() and used it in amdgpucsvmhandling(). That did not cover a call in amdgpudmabufmovenotify(), what was already mentioned in: https://lists.freedesktop.org/archives/amd-gfx/2025-October/133076.html
Tested by SIGKILLing a GPU workload holding imported dma-bufs. Without the patch "Trying to push to a killed entity" errors are logged, then a task hangs. With the patch there are no hung tasks. This was tested on a 6.18 kernel.
(cherry picked from commit bee5c9d7a34cb1b510ad3b4c18c30f9d38b0ab7d)
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.17 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.18.64 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5 - Upgrade
Upgrade
debian/linuxto a version that resolves this vulnerability.Fixed in 6.1.176-1Fixed in 6.1.192-2Fixed in 6.12.107-1Fixed in 6.12.117-1Fixed in 7.3.5-1 - Upgrade
Upgrade
Linux kernelto a version that resolves this vulnerability.Fixed in 6.18
Event History
Frequently Asked Questions
What conditions are needed to trigger the hang?
A process must be using a GPU workload that holds imported dma-bufs and then be terminated with SIGKILL. During teardown, the driver may attempt to clear page-table entries through VM entities that have already been stopped.
What symptoms indicate a system may be affected?
Kernel logs may contain "Trying to push to a killed entity" errors, followed by a hung task. The issue occurs during teardown of an imported dma-buf after the associated process is killed.
What does the fix change?
The fix checks amdgpu_vm_ready() before amdgpu_dma_buf_move_notify() accesses page tables. This prevents job submission through stopped VM entities during process teardown.