Description
In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: Clear VRAM on allocation to prevent stale data exposure KFD VRAM allocations set AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE but not AMDGPU_GEM_CREATE_VRAM_CLEARED, leaving freshly allocated VRAM with stale data from prior use observable by compute kernels. The GEM ioctl path already sets VRAM_CLEARED for all userspace allocations via amdgpu_gem_create_ioctl() and amdgpu_mode_dumb_create(). The KFD path was missing this flag, allowing stale page table remnants to leak into user buffers. This causes crashes in RCCL P2P transport where non-zero data in ptrExchange/head/tail fields corrupts the protocol handshake.
Product status
6856e4b65f64eeb3f17148f79b36c1d60c627529 (git) before 1db431380879fd9d28b763a88a0c0431be5be8df
6856e4b65f64eeb3f17148f79b36c1d60c627529 (git) before 32b153658f017ad2f5bf8aab479e8d16ac95bc3a
6856e4b65f64eeb3f17148f79b36c1d60c627529 (git) before 77d0b5d11387071770246fd0185a69fa28e8e109
6856e4b65f64eeb3f17148f79b36c1d60c627529 (git) before 047d44d8d29a6a1a5757256837aa9dd78e3cd0b5
6856e4b65f64eeb3f17148f79b36c1d60c627529 (git) before ad52d61d82181dbdb7f05826de38352d5e550cc2
5.4
Any version before 5.4
6.6.140 (semver)
6.12.90 (semver)
6.18.32 (semver)
7.0.9 (semver)
7.1-rc1 (original_commit_for_fix)
References
git.kernel.org/...c/1db431380879fd9d28b763a88a0c0431be5be8df
git.kernel.org/...c/32b153658f017ad2f5bf8aab479e8d16ac95bc3a
git.kernel.org/...c/77d0b5d11387071770246fd0185a69fa28e8e109
git.kernel.org/...c/047d44d8d29a6a1a5757256837aa9dd78e3cd0b5
git.kernel.org/...c/ad52d61d82181dbdb7f05826de38352d5e550cc2