CVE-2026-68237
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/userq: fix indefinite fence wait during GPU reset pre_reset only force-completes fences of MAPPED queues. A queue in any other state (e.g. mid-eviction) keeps its last_fence pending; after a GPU reset that fence never signals, so the eviction/suspend worker and process teardown (amdgpu_evf_mgr_flush_suspend) wait on it forever and wedge the machine: INFO: task kworker/6:28 blocked for more than 120 seconds. Workqueue: events amdgpu_eviction_fence_suspend_worker [amdgpu] Call Trace: dma_fence_wait_timeout+0x7e/0x130 amdgpu_userq_evict+0x67/0x140 [amdgpu] amdgpu_eviction_fence_suspend_worker+0xd8/0x160 [amdgpu] process_scheduled_works+0xa6/0x420 Force-complete every queue's fence regardless of state. The unmap and mark-hung step stays gated on MAPPED, since unmapping a queue that is not mapped is invalid. (cherry picked from commit 9102b39fa924dcc3dc75a3137bfa9633c40b88c0)
Affected versions
Linux kernel versions
6.19
and later are affected. Fixed in
7.1.6,
7.2
and their respective stable series.
References
2 totalFrequently asked questions
-
What is CVE-2026-68237?
CVE-2026-68237 is a unscored severity Linux kernel vulnerability . It affects Linux kernel versions from 6.19 onward and has been patched in 7.1.6 and 7.2. CVE-2026-68237 has not been confirmed as actively exploited and is not listed in the CISA KEV catalog.
-
Is there a patch available for CVE-2026-68237?
Yes. CVE-2026-68237 has been patched. Fixed versions include 7.1.6 and 7.2. If you are running Linux kernel 6.19 or later up to the fix versions, apply the relevant patch for your kernel branch.
-
Is CVE-2026-68237 actively exploited?
No. CVE-2026-68237 has not been confirmed as actively exploited. It is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog.