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.
History

Fri, 09 May 2025 07:00:00 +0000

Type Values Removed Values Added
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.
Title drm/amdkfd: Fix mode1 reset crash issue
References

cve-icon MITRE

Status: PUBLISHED

Assigner: Linux

Published: 2025-05-09T06:42:00.663Z

Updated: 2025-05-09T06:42:00.663Z

Reserved: 2025-04-16T04:51:23.956Z

Link: CVE-2025-37854

cve-icon Vulnrichment

No data.

cve-icon NVD

Status : Received

Published: 2025-05-09T07:16:06.367

Modified: 2025-05-09T07:16:06.367

Link: CVE-2025-37854

cve-icon Redhat

No data.