{"dataType":"CVE_RECORD","dataVersion":"5.2","cveMetadata":{"cveId":"CVE-2023-53707","assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","state":"PUBLISHED","assignerShortName":"Linux","dateReserved":"2025-10-22T13:21:37.346Z","datePublished":"2025-10-22T13:23:43.822Z","dateUpdated":"2026-08-05T09:15:38.286Z"},"containers":{"cna":{"providerMetadata":{"orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux","dateUpdated":"2026-08-05T09:15:38.286Z"},"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/amdgpu: Fix integer overflow in amdgpu_cs_pass1\n\nThe type of size is unsigned int, if size is 0x40000000, there will\nbe an integer overflow, size will be zero after size *= sizeof(uint32_t),\nwill cause uninitialized memory to be referenced later."}],"metrics":[{"cvssV3_1":{"version":"3.1","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","baseScore":7.8,"baseSeverity":"HIGH"},"scenarios":[{"lang":"en","value":"AV:L - The vulnerability is reached only through the DRM_IOCTL_AMDGPU_CS ioctl on a local amdgpu device node (/dev/dri/renderD* or card*), requiring local access to the machine. There is no network or remote-peer path into amdgpu_cs_pass1().\nAC:L - The attacker directly supplies the overflowing value (length_dw >= 0x40000000) in a single ioctl call with no race, timing, or memory-layout condition to win, and can free memory and retry to make the large allocation succeed. On the multi-GiB systems where AMD GPUs are deployed (workstations, Steam Deck, cloud GPU nodes) the allocation reliably succeeds.\nPR:L - AMDGPU_CS is registered with DRM_RENDER_ALLOW, so no DRM master, authentication, or capability is required — any unprivileged local user or sandboxed app holding an fd on the render node can invoke it, which is standard for desktop sessions, Android apps, and GPU-enabled containers.\nUI:N - Exploitation is a self-contained sequence of ioctls issued by the attacker's own process; no victim action, file open, or mount is needed.\nS:U - The corrupted state and the uninitialized data consumed all live within the kernel's own security authority on the same host; no VM, IOMMU, or sandbox boundary is crossed.\nC:H - A multi-gigabyte kvmalloc/vmalloc buffer is left completely uninitialized and then consumed as GEM handles, BO-list descriptors, user pointers and fence offsets, exposing stale kernel/page contents; the resulting lookup success/failure and the observable fence write offset form oracles that let an attacker read back that uninitialized kernel memory.\nI:H - Uninitialized kernel memory — attacker-groomable via page spraying — is used as GPU virtual addresses, IB sizes, flags, buffer-object handles and __user pointers that drive command submission, BO list construction and the user-fence write target, giving control-relevant fields that can be steered toward memory corruption.\nA:H - An unprivileged caller can force repeated 4 GiB kernel allocations causing reclaim storms and OOM kills, trips a WARN_ON_ONCE() in kvmalloc_node() (a panic under panic_on_warn), and drives the CS parser through uninitialized state, producing kernel crashes and denial of service."}]}],"affected":[{"product":"Linux","vendor":"Linux","defaultStatus":"unaffected","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c"],"versions":[{"version":"d38ceaf99ed015f2a0b9af3499791bd3a3daae21","lessThan":"9f55d300541cb5b435984d269087810581580b00","status":"affected","versionType":"git"},{"version":"d38ceaf99ed015f2a0b9af3499791bd3a3daae21","lessThan":"c3deb091398e9e469d08dd1599b6d76fd6b29df8","status":"affected","versionType":"git"},{"version":"d38ceaf99ed015f2a0b9af3499791bd3a3daae21","lessThan":"87c2213e85bd81e4a9a4d0880c256568794ae388","status":"affected","versionType":"git"}]},{"product":"Linux","vendor":"Linux","defaultStatus":"affected","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c"],"versions":[{"version":"4.2","status":"affected"},{"version":"0","lessThan":"4.2","status":"unaffected","versionType":"semver"},{"version":"6.1.47","lessThanOrEqual":"6.1.*","status":"unaffected","versionType":"semver"},{"version":"6.4.12","lessThanOrEqual":"6.4.*","status":"unaffected","versionType":"semver"},{"version":"6.5","lessThanOrEqual":"*","status":"unaffected","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"operator":"OR","negate":false,"cpeMatch":[{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"4.2","versionEndExcluding":"6.1.47"},{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"4.2","versionEndExcluding":"6.4.12"},{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"4.2","versionEndExcluding":"6.5"}]}]}],"references":[{"url":"https://git.kernel.org/stable/c/9f55d300541cb5b435984d269087810581580b00"},{"url":"https://git.kernel.org/stable/c/c3deb091398e9e469d08dd1599b6d76fd6b29df8"},{"url":"https://git.kernel.org/stable/c/87c2213e85bd81e4a9a4d0880c256568794ae388"}],"title":"drm/amdgpu: Fix integer overflow in amdgpu_cs_pass1","x_generator":{"engine":"bippy-1.2.0"}}}}