{"id":"CVE-2026-31484","state":"PUBLISHED","title":"io_uring/fdinfo: fix OOB read in SQE_MIXED wrap check","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nio_uring/fdinfo: fix OOB read in SQE_MIXED wrap check\n\n__io_uring_show_fdinfo() iterates over pending SQEs and, for 128-byte\nSQEs on an IORING_SETUP_SQE_MIXED ring, needs to detect when the second\nhalf of the SQE would be past the end of the sq_sqes array. The current\ncheck tests (++sq_head & sq_mask) == 0, but sq_head is only incremented\nwhen a 128-byte SQE is encountered, not on every iteration. The actual\narray index is sq_idx = (i + sq_head) & sq_mask, which can be sq_mask\n(the last slot) while the wrap check passes.\n\nFix by checking sq_idx directly. Keep the sq_head increment so the loop\nstill skips the second half of the 128-byte SQE on the next iteration.","publishedAt":"2026-04-22T13:54:10.223Z","updatedAt":"2026-08-05T12:22:59.333Z","severity":{"score":7.1,"label":"High","version":"3.1","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H","source":"Linux"},"assigner":"Linux","products":["Linux Linux"],"affected":[{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"1cba30bf9fdd6c982708f3587f609a30c370d889","status":"affected","lessThan":"ba21ab247a5be5382da7464b95afbe5f0e9aa503","versionType":"git"},{"version":"1cba30bf9fdd6c982708f3587f609a30c370d889","status":"affected","lessThan":"5170efd9c344c68a8075dcb8ed38d3f8a60e7ed4","versionType":"git"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["io_uring/fdinfo.c"]},{"vendor":"Linux","product":"Linux","defaultStatus":"affected","versions":[{"version":"6.19","status":"affected"},{"version":"0","status":"unaffected","lessThan":"6.19","versionType":"semver"},{"version":"6.19.11","status":"unaffected","lessThanOrEqual":"6.19.*","versionType":"semver"},{"version":"7.0","status":"unaffected","lessThanOrEqual":"*","versionType":"original_commit_for_fix"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["io_uring/fdinfo.c"]}],"references":["https://git.kernel.org/stable/c/ba21ab247a5be5382da7464b95afbe5f0e9aa503","https://git.kernel.org/stable/c/5170efd9c344c68a8075dcb8ed38d3f8a60e7ed4"],"weaknesses":[],"ingestedAt":"2026-10-29T20:11:40.391Z","source":{"repository":"CVEProject/cvelistV5","path":"cves/2026/31xxx/CVE-2026-31484.json","commit":"556c9d3a8cd17e218bb8e452ba7cd619a384ff47"},"raw":{"dataType":"CVE_RECORD","dataVersion":"5.2","cveMetadata":{"cveId":"CVE-2026-31484","state":"PUBLISHED","datePublished":"2026-04-22T13:54:10.223Z","dateUpdated":"2026-08-05T12:22:59.333Z","assignerShortName":"Linux","assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","dateReserved":"2026-03-09T15:48:24.101Z"},"containers":{"cna":{"title":"io_uring/fdinfo: fix OOB read in SQE_MIXED wrap check","descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nio_uring/fdinfo: fix OOB read in SQE_MIXED wrap check\n\n__io_uring_show_fdinfo() iterates over pending SQEs and, for 128-byte\nSQEs on an IORING_SETUP_SQE_MIXED ring, needs to detect when the second\nhalf of the SQE would be past the end of the sq_sqes array. The current\ncheck tests (++sq_head & sq_mask) == 0, but sq_head is only incremented\nwhen a 128-byte SQE is encountered, not on every iteration. The actual\narray index is sq_idx = (i + sq_head) & sq_mask, which can be sq_mask\n(the last slot) while the wrap check passes.\n\nFix by checking sq_idx directly. Keep the sq_head increment so the loop\nstill skips the second half of the 128-byte SQE on the next iteration."}],"affected":[{"vendor":"Linux","product":"Linux","defaultStatus":"unaffected","versions":[{"version":"1cba30bf9fdd6c982708f3587f609a30c370d889","status":"affected","lessThan":"ba21ab247a5be5382da7464b95afbe5f0e9aa503","versionType":"git"},{"version":"1cba30bf9fdd6c982708f3587f609a30c370d889","status":"affected","lessThan":"5170efd9c344c68a8075dcb8ed38d3f8a60e7ed4","versionType":"git"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["io_uring/fdinfo.c"]},{"vendor":"Linux","product":"Linux","defaultStatus":"affected","versions":[{"version":"6.19","status":"affected"},{"version":"0","status":"unaffected","lessThan":"6.19","versionType":"semver"},{"version":"6.19.11","status":"unaffected","lessThanOrEqual":"6.19.*","versionType":"semver"},{"version":"7.0","status":"unaffected","lessThanOrEqual":"*","versionType":"original_commit_for_fix"}],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["io_uring/fdinfo.c"]}],"references":[{"url":"https://git.kernel.org/stable/c/ba21ab247a5be5382da7464b95afbe5f0e9aa503"},{"url":"https://git.kernel.org/stable/c/5170efd9c344c68a8075dcb8ed38d3f8a60e7ed4"}],"metrics":[{"cvssV3_1":{"version":"3.1","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H","baseScore":7.1,"baseSeverity":"HIGH"},"scenarios":[{"lang":"en","value":"AV:L - The vulnerable path is reached locally by creating an io_uring instance and reading its procfs fdinfo entry; there is no network or adjacent attack surface.\nAC:L - The attacker controls the ring setup flags, SQ tail, SQ array mapping, and 128-byte SQE opcode needed to place a mixed SQE at the last physical slot. No race or external timing condition is required, and transient lock contention can be retried.\nPR:L - A normal unprivileged local user can use io_uring when the default io_uring sysctl allows it and can read /proc/self/fdinfo for its own fd. No real root or init-namespace capability is required.\nUI:N - The attacker triggers the bug by reading their own fdinfo file. No victim action is needed.\nS:U - The issue is a local kernel memory access bug within the same kernel security authority. It does not cross a VM, IOMMU, or other separate security scope boundary.\nC:H - The bug prints the second half of a 128-byte SQE by dereferencing up to 64 bytes past the SQE array, exposing adjacent kernel memory through procfs. The leak is attacker-triggerable and repeatable, so the highest defensible confidentiality impact is high.\nI:N - The vulnerable operation is an out-of-bounds read used for fdinfo printing and does not write to kernel memory or directly modify protected state.\nA:H - The same out-of-bounds kernel dereference can cross the end of the SQE allocation or a vmap-backed region and fault/oops depending on allocation layout. An unprivileged attacker can repeatedly trigger the path, so availability impact is high."}]}],"providerMetadata":{"shortName":"Linux","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","dateUpdated":"2026-08-05T12:22:59.333Z"},"cpeApplicability":[{"nodes":[{"operator":"OR","negate":false,"cpeMatch":[{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"6.19","versionEndExcluding":"6.19.11"},{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"6.19","versionEndExcluding":"7.0"}]}]}],"x_generator":{"engine":"bippy-1.2.0"}}}}}