CVE-2026-23034
BaseFortify
Publication date: 2026-01-31
Last updated on: 2026-02-03
Assigner: kernel.org
Description
Description
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| amdgpu | linux_kernel | to 2.11.0 (exc) |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-UNKNOWN |
Attack-Flow Graph
AI Powered Q&A
Can you explain this vulnerability to me?
This vulnerability is a reference leak in the Linux kernel's AMD GPU user mode queue (userq) driver. Specifically, the userq keeps a pointer to the most recent fence object (last_fence) which holds an extra reference. When the queue is destroyed, the driver frees related resources but fails to release this last_fence reference. As a result, the fence object remains allocated even after the driver unloads, causing memory objects to persist in the slab cache and triggering errors during module unload.
How can this vulnerability impact me? :
This vulnerability can cause memory leaks in the AMD GPU driver by leaving fence objects allocated after the driver unloads. This can lead to resource exhaustion or instability in the system, especially during driver unload or module removal operations. It may also trigger kernel errors related to slab cache objects remaining allocated.
How can this vulnerability be detected on my network or system? Can you suggest some commands?
This vulnerability can be detected by observing error messages during the unloading of the amdgpu driver module. Specifically, look for kernel log messages indicating a fence reference leak such as: BUG amdgpu_userq_fence: Objects remaining on __kmem_cache_shutdown() kmem_cache_destroy amdgpu_userq_fence: Slab cache still has objects You can check the kernel logs using the command: dmesg | grep amdgpu_userq_fence Additionally, monitoring for errors during module unload can be done with: sudo rmmod amdgpu and then checking dmesg for the above messages.
What immediate steps should I take to mitigate this vulnerability?
The immediate mitigation is to update the Linux kernel to a version that includes the fix for this vulnerability. The fix ensures that the last_fence reference is properly released during the amdgpu user mode queue teardown, preventing the fence reference leak. Until the update is applied, avoid unloading the amdgpu driver module to prevent the leak from occurring.