CVE-2024-36949: amd/amdkfd: sync all devices to wait all processes being evicted
In the Linux kernel, the following vulnerability has been resolved:
amd/amdkfd: sync all devices to wait all processes being evicted
If there are more than one device doing reset in parallel, the first
device will call kfd_suspend_all_processes() to evict all processes
on all devices, this call takes time to finish. other device will
start reset and recover without waiting. if the process has not been
evicted before doing recover, it will be restored, then caused page
fault.
Security readout for executives and security teams
Plain-English summary
CVE-2024-36949 is a Linux kernel issue in AMD GPU compute handling. During simultaneous GPU device resets, processes may be restored before eviction completes, causing page faults. The main business risk appears to be system or workload instability on affected Linux systems using AMD GPU/KFD functionality.
Executive priority
Treat as a targeted infrastructure maintenance item. Prioritize GPU compute platforms and systems where instability affects production workloads. There is no source-supported evidence of active exploitation, but kernel fixes should be included in normal patch cycles.
Technical view
The flaw is in amd/amdkfd reset recovery synchronization. One device may start recovery while another is still evicting processes across devices via kfd_suspend_all_processes(). If recovery occurs before eviction finishes, a process can be restored prematurely and trigger a page fault. Kernel stable commits are referenced as the resolution.
Likely exposure
Exposure is most relevant to Linux systems with AMD GPU/KFD support, especially multi-GPU or GPU-compute environments where parallel device resets can occur. Distribution backports may change exposure, so version-only checks are insufficient.
Exploitation context
No active exploitation is indicated in the provided sources, and the CVE is not listed as KEV. The source material describes a race-condition stability issue, not a remote attack path. Required conditions and security impact are not fully characterized.
Researcher notes
The public record provides limited severity detail: no CVSS, CWE, or complete exploitability analysis. Assessment should focus on reproducing reset concurrency behavior safely, confirming patch presence, and mapping distro-specific affected builds. Avoid assuming all Linux systems are meaningfully exposed.
Mitigation direction
Apply Linux kernel updates containing the referenced stable fixes.
Check your Linux distribution advisory for backported fixes and package names.
Prioritize AMD GPU compute, multi-GPU, and reset-prone systems.
Schedule maintenance if kernel replacement requires reboot.
Validation and detection
Inventory Linux hosts using AMDGPU/KFD functionality.
Confirm installed kernels include the referenced stable commits or distro backports.
Review vendor changelogs for CVE-2024-36949 or amdkfd reset synchronization fixes.
Monitor affected systems for GPU reset, KFD eviction, and page fault events.
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Potential ATT&CK relevance
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
cve · low confidence lookup
CVE-2024-36949 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.