KernelScan.io

HIGH Introduced in 4.2

amdgpu UvdDecode OOB

CVE-2026-80540

CVSS 7.8 / 10.0 NVD

CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

KernelScan AI7.7HIGH

01

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu: Fix UVD decode image min size calculation This needs to use pitch instead of width. Also reject pitch over 4096 to avoid overflow. (cherry picked from commit b41c8cb12e202b220353332ab87dc01a11f69304)

02

Engine v0.6.0

Risk summary

A local user with access to the AMD GPU UVD video decode interface can trigger an out-of-bounds write by submitting a crafted decode command with an oversized pitch, corrupting kernel heap memory and potentially escalating privileges or crashing the system.

Affecteddrivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c (amdgpu UVD)

Vulnerability analysis

The UVD decode command parser computes the minimum image buffer size using the image width instead of the pitch, and does not reject pitch values exceeding 4096. A local attacker with access to the AMD GPU device can submit a video decode command stream with a large pitch, causing the computed buffer size to overflow and the kernel to write beyond the allocated decode buffer. The fix changes the size calculation to use pitch times height and rejects pitch values over 4096, preventing the overflow. The vulnerable path is reachable from a local process with access to the amdgpu DRM device.

03

BranchIntroducedFixed inPatch commit
5.104.25.10.266bc7397a033ac
6.14.26.1.184b7549e3f96c7
6.64.26.6.15360539d517e84
6.124.26.12.105271a7da84a62
6.184.26.18.4625ee120f3803
7.14.27.1.105cbd8af02b0b
mainline4.27.2b8bb9ba3f101
5.154.25.15.217d058f7a67094
cip/4.194.24.19.325-cip13627ed0ac2bcc8
cip/4.19-rt4.24.19.325-cip136-rt5027ed0ac2bcc8
cip/4.44.24.4.302-cip1157090efa2b92c
cip/4.4-rt4.24.4.302-cip115-rt637090efa2b92c
cip/5.104.25.10.268-cip77bc7397a033ac
cip/5.10-rt4.25.10.268-cip77-rt34bc7397a033ac
cip/6.14.26.1.187-cip60b7549e3f96c7
cip/6.124.26.12.108-cip31271a7da84a62