GHSA-P526-PM6J-FR73
Vulnerability from github – Published: 2026-08-10 15:33 – Updated: 2026-08-10 15:33In 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": [],
"aliases": [
"CVE-2026-68237"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-08-10T13:20:12Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/amdgpu/userq: fix indefinite fence wait during GPU reset\n\npre_reset only force-completes fences of MAPPED queues. A queue in any\nother state (e.g. mid-eviction) keeps its last_fence pending; after a\nGPU reset that fence never signals, so the eviction/suspend worker and\nprocess teardown (amdgpu_evf_mgr_flush_suspend) wait on it forever and\nwedge the machine:\n\n INFO: task kworker/6:28 blocked for more than 120 seconds.\n Workqueue: events amdgpu_eviction_fence_suspend_worker [amdgpu]\n Call Trace:\n dma_fence_wait_timeout+0x7e/0x130\n amdgpu_userq_evict+0x67/0x140 [amdgpu]\n amdgpu_eviction_fence_suspend_worker+0xd8/0x160 [amdgpu]\n process_scheduled_works+0xa6/0x420\n\nForce-complete every queue\u0027s fence regardless of state. The unmap and\nmark-hung step stays gated on MAPPED, since unmapping a queue that is\nnot mapped is invalid.\n\n(cherry picked from commit 9102b39fa924dcc3dc75a3137bfa9633c40b88c0)",
"id": "GHSA-p526-pm6j-fr73",
"modified": "2026-08-10T15:33:42Z",
"published": "2026-08-10T15:33:42Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-68237"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/3085ae8695e025b39d208f288c6265edc75abbe8"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/5d75ec2e5f1736c2f10c7d6f4565bf1bf29f29a7"
}
],
"schema_version": "1.4.0",
"severity": []
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.