Skip to content
CVE PROGRAM SOURCE RECORD7.8 High

CVE-2025-37854

drm/amdkfd: Fix mode1 reset crash issue

Read JSON ↗CVE.org ↗
Status
Published
Published
Updated
Assigning organization
Linux

Description

In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: Fix mode1 reset crash issue If HW scheduler hangs and mode1 reset is used to recover GPU, KFD signal user space to abort the processes. After process abort exit, user queues still use the GPU to access system memory before h/w is reset while KFD cleanup worker free system memory and free VRAM. There is use-after-free race bug that KFD allocate and reuse the freed system memory, and user queue write to the same system memory to corrupt the data structure and cause driver crash. To fix this race, KFD cleanup worker terminate user queues, then flush reset_domain wq to wait for any GPU ongoing reset complete, and then free outstanding BOs.

Affected products

Version statuses are supplied by the assigning organization. Consult the original record for detailed ranges and platform conditions.

Linux Linux

Default version status: unaffected

  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < 57c9dabda80ac167de8cd71231baae37cc2f442daffected
  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < 89af6b39f028c130d4362f57042927f005423e6aaffected
  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < ffd37d7d44d7e0b6e769d4fe6590e327f8cc3951affected
  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < 6f30a847432cae84c7428e9b684b3e3fa49b2391affected
  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < 9c4bcdf4068aae3e17e31c144300be405cfa03ffaffected
  • a4c63cafa58b4bd9e15511bab77a4752b93d3aa0 to < f0b4440cdc1807bb6ec3dce0d6de81170803569baffected

Linux Linux

Default version status: affected

  • 5.18affected
  • 0 to < 5.18unaffected
  • 6.1.135 to ≤ 6.1.*unaffected
  • 6.6.88 to ≤ 6.6.*unaffected
  • 6.12.24 to ≤ 6.12.*unaffected
  • 6.13.12 to ≤ 6.13.*unaffected
  • 6.14.3 to ≤ 6.14.*unaffected
  • 6.15 to ≤ *unaffected

References