{"id":"CVE-2024-27411","state":"PUBLISHED","title":"drm/nouveau: keep DMA buffers required for suspend/resume","description":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/nouveau: keep DMA buffers required for suspend/resume\n\nNouveau deallocates a few buffers post GPU init which are required for GPU suspend/resume to function correctly.\nThis is likely not as big an issue on systems where the NVGPU is the only GPU, but on multi-GPU set ups it leads to a regression where the kernel module errors and results in a system-wide rendering freeze.\n\nThis commit addresses that regression by moving the two buffers required for suspend and resume to be deallocated at driver unload instead of post init.","publishedAt":"2024-05-17T11:50:46.770Z","updatedAt":"2026-08-05T11:29:42.130Z","severity":{"score":7.8,"label":"High","version":"3.1","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","source":"Linux"},"assigner":"Linux","products":["Linux Linux"],"affected":[{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"6190d4c08897d748dd25f0b78267a90aa1694e15","status":"affected","lessThan":"be00e15b240ed71fc30c0576af7ab670c8271661","versionType":"git"},{"version":"042b5f83841fbf7ce39474412db3b5e4765a7ea7","status":"affected","lessThan":"f6ecfdad359a01c7fd8a3bcfde3ef0acdf107e6e","versionType":"git"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c"]},{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"6.7.6","status":"affected","lessThan":"6.7.9","versionType":"semver"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c"]}],"references":["https://git.kernel.org/stable/c/be00e15b240ed71fc30c0576af7ab670c8271661","https://git.kernel.org/stable/c/f6ecfdad359a01c7fd8a3bcfde3ef0acdf107e6e"],"weaknesses":[],"ingestedAt":"2026-10-29T20:07:10.607Z","source":{"repository":"CVEProject/cvelistV5","path":"cves/2024/27xxx/CVE-2024-27411.json","commit":"556c9d3a8cd17e218bb8e452ba7cd619a384ff47"},"raw":{"dataType":"CVE_RECORD","dataVersion":"5.2","cveMetadata":{"cveId":"CVE-2024-27411","state":"PUBLISHED","datePublished":"2024-05-17T11:50:46.770Z","dateUpdated":"2026-08-05T11:29:42.130Z","assignerShortName":"Linux","assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","dateReserved":"2024-02-25T13:47:42.682Z"},"containers":{"cna":{"title":"drm/nouveau: keep DMA buffers required for suspend/resume","descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/nouveau: keep DMA buffers required for suspend/resume\n\nNouveau deallocates a few buffers post GPU init which are required for GPU suspend/resume to function correctly.\nThis is likely not as big an issue on systems where the NVGPU is the only GPU, but on multi-GPU set ups it leads to a regression where the kernel module errors and results in a system-wide rendering freeze.\n\nThis commit addresses that regression by moving the two buffers required for suspend and resume to be deallocated at driver unload instead of post init."}],"affected":[{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"6190d4c08897d748dd25f0b78267a90aa1694e15","status":"affected","lessThan":"be00e15b240ed71fc30c0576af7ab670c8271661","versionType":"git"},{"version":"042b5f83841fbf7ce39474412db3b5e4765a7ea7","status":"affected","lessThan":"f6ecfdad359a01c7fd8a3bcfde3ef0acdf107e6e","versionType":"git"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c"]},{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"6.7.6","status":"affected","lessThan":"6.7.9","versionType":"semver"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c"]}],"references":[{"url":"https://git.kernel.org/stable/c/be00e15b240ed71fc30c0576af7ab670c8271661"},{"url":"https://git.kernel.org/stable/c/f6ecfdad359a01c7fd8a3bcfde3ef0acdf107e6e"}],"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 vulnerable path is the nouveau GPU suspend/resume handler, reached through local DRM device-node usage and PCI runtime power management; no network or physical access is involved.\nAC:L - The condition is deterministic — `gsp->wpr_meta.data` and `gsp->rmargs.data` are unconditionally NULLed at postinit, so the very first suspend after driver load faults with no race and no memory-layout dependency.\nPR:L - An unprivileged local user with access to `/dev/dri/renderD*` (the default for a desktop session) can force a `pm_runtime_get_sync`/`put_autosuspend` cycle by opening and closing the render node, triggering runtime suspend within the 5s autosuspend delay; on hybrid-graphics systems it also fires from mere idling.\nUI:N - Nouveau enables runtime PM by default on Optimus/multi-GPU systems, so autosuspend fires automatically after idle and can be induced by the attacker's own process — no victim action such as closing a lid or invoking `systemctl suspend` is needed.\nS:U - The fault and its consequences remain within the kernel's own security authority; no VM, IOMMU, or sandbox boundary is crossed by the driver's own security scope.\nC:H - The freed RMARGS/WPR-meta DMA pages remain mapped into GSP-RM's libos address space and hold the physical address of GSP-RM's shared message queues, so the GPU continues reading memory the kernel has released and reallocated, exposing arbitrary reclaimed kernel data to the device.\nI:H - Premature `dma_free_coherent()` of buffers whose bus addresses are still programmed into the GPU is a device-side use-after-free; queue physical addresses sprayed into the reclaimed page are consumed by GSP-RM as DMA targets, yielding a write primitive into arbitrary kernel memory.\nA:H - A guaranteed NULL pointer dereference at NULL+0xa8 in `r535_gsp_fini()` oopses inside the PM path with device locks held, producing the reported system-wide rendering freeze and an unrecoverable suspend/resume failure."}]}],"providerMetadata":{"shortName":"Linux","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","dateUpdated":"2026-08-05T11:29:42.130Z"},"cpeApplicability":[{"nodes":[{"operator":"OR","negate":false,"cpeMatch":[{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"6.7.6","versionEndExcluding":"6.7.9"}]}]}],"x_generator":{"engine":"bippy-1.2.0"}},"adp":[{"title":"CVE Program Container","references":[{"url":"https://git.kernel.org/stable/c/be00e15b240ed71fc30c0576af7ab670c8271661","tags":["x_transferred"]},{"url":"https://git.kernel.org/stable/c/f6ecfdad359a01c7fd8a3bcfde3ef0acdf107e6e","tags":["x_transferred"]}],"providerMetadata":{"shortName":"CVE","orgId":"af854a3a-2127-422b-91ae-364da2661108","dateUpdated":"2024-08-02T00:34:52.193Z"}},{"title":"CISA ADP Vulnrichment","metrics":[{"other":{"type":"ssvc","content":{"id":"CVE-2024-27411","role":"CISA Coordinator","options":[{"Exploitation":"none"},{"Automatable":"no"},{"Technical Impact":"partial"}],"version":"2.0.3","timestamp":"2024-09-10T15:43:09.899488Z"}}}],"providerMetadata":{"shortName":"CISA-ADP","orgId":"134c704f-9b21-4f2e-91b3-4a467353bcc0","dateUpdated":"2024-09-11T17:33:25.106Z"}}]}}}