{ "dataType": "CVE_RECORD", "dataVersion": "5.2", "cveMetadata": { "cveId": "CVE-2026-68329", "assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "state": "PUBLISHED", "assignerShortName": "Linux", "dateReserved": "2026-07-30T09:28:09.383Z", "datePublished": "2026-08-10T12:03:05.526Z", "dateUpdated": "2026-08-17T05:03:27.689Z" }, "containers": { "cna": { "providerMetadata": { "orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "shortName": "Linux", "dateUpdated": "2026-08-17T05:03:27.689Z" }, "descriptions": [ { "lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\niommu/amd: Wait for completion instead of returning early in iommu_completion_wait()\n\nneed_sync is a per-IOMMU flag shared by all domains and devices behind\nthat IOMMU. It is set whenever a command is queued with sync == true and\ncleared when a completion-wait (CWAIT) command is queued. However, a\ncleared need_sync only means that a covering CWAIT has been queued, not\nthat all previously queued commands have actually completed in hardware.\n\niommu_completion_wait() read need_sync locklessly and returned early\nwhen it was false. This breaks the \"block until all previously queued\ncommands have completed\" contract in a multi-CPU scenario:\n\n CPU2: queue inv-B => need_sync = true\n CPU1: queue CWAIT(N); need_sync = false; then wait_on_sem(N)\n CPU2: read need_sync == false => return 0 (no wait!)\n\nCPU2 returns without waiting for any sequence number even though its\ninv-B may not have completed yet (CWAIT(N), queued after inv-B, has not\nbeen signaled). CPU2 then proceeds to, for example, free page-table\npages while the IOMMU can still walk stale translations, opening a\nuse-after-free window. This is a logical race in the meaning of the\nflag, not a memory-visibility issue, so barriers alone do not help.\n\nFix it without losing the optimization of avoiding redundant CWAIT\ncommands: take iommu->lock before testing need_sync, and when it is\nfalse do not return early but wait for the last allocated sequence\nnumber (cmd_sem_val). Since need_sync == false implies no sync command\nwas queued after the last CWAIT, that CWAIT is FIFO-ordered after every\nnot-yet-completed command, so waiting for its sequence number guarantees\nall prior commands (possibly queued by another CPU) have completed. The\ncommon path with pending work is unchanged and no extra hardware command\nis issued." } ], "metrics": [ { "cvssV3_1": { "version": "3.1", "vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H", "baseScore": 8.8, "baseSeverity": "HIGH" }, "scenarios": [ { "lang": "en", "value": "AV:L - The racy `iommu_completion_wait()` is reached through local DMA map/unmap activity — iommu-dma unmap paths, VFIO/IOMMUFD unmap ioctls, and domain attach/detach — all driven from the local system rather than from network input.\nAC:L - The attacker drives both sides of the race by issuing concurrent unmap/flush operations on multiple CPUs against the same IOMMU, and can retry indefinitely; each hit frees page-table pages while translations are still live, so no condition outside the attacker's influence is required.\nPR:L - An unprivileged local user can generate the concurrent IOMMU invalidation traffic needed via ordinary parallel I/O through DMA-API device paths; no root or capability in the init namespace is required, and privileged VFIO users are only one of several entry points.\nUI:N - The race is triggered purely by the attacker's own concurrent unmap/flush workload; no victim action such as mounting a filesystem or opening a file is needed.\nS:C - The failure defeats IOMMU isolation itself: page-table pages are freed and reused while the IOMMU can still walk them and stale IOTLB entries persist, so a device (including one assigned to a guest or an external peripheral) retains DMA access to host memory outside its domain, crossing the DMA/IOMMU security boundary.\nC:H - Stale translations and use-after-free page tables let a device read host memory that was unmapped, freed, and reallocated to unrelated kernel or user data, yielding effectively arbitrary memory disclosure across the IOMMU boundary.\nI:H - The same stale translations permit device DMA writes into freed and reallocated pages, including page-table pages the IOMMU still walks, giving an arbitrary kernel-memory write primitive usable for control-flow hijacking and privilege escalation.\nA:H - Use-after-free of page-table pages and DMA into reallocated memory corrupt kernel structures and generate IOMMU I/O page faults, leading to oops or panic and full loss of availability." } ] } ], "affected": [ { "product": "Linux", "vendor": "Linux", "defaultStatus": "unaffected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": [ "drivers/iommu/amd/iommu.c" ], "versions": [ { "version": "815b33fdc279d34ab40a8bfe1866623a4cc5669b", "lessThan": "ab7faf5a172ebfdc423ebb3eea4d472740de82f9", "status": "affected", "versionType": "git" }, { "version": "815b33fdc279d34ab40a8bfe1866623a4cc5669b", "lessThan": "93494bd446396c257fb589f59894577e96e406e2", "status": "affected", "versionType": "git" }, { "version": "815b33fdc279d34ab40a8bfe1866623a4cc5669b", "lessThan": "d053eb7e09e10cbdca3fca8b35c1017d438091b2", "status": "affected", "versionType": "git" }, { "version": "815b33fdc279d34ab40a8bfe1866623a4cc5669b", "lessThan": "02f8cefa2ad95ea3754f0cfd6fbae7f866202ccb", "status": "affected", "versionType": "git" }, { "version": "815b33fdc279d34ab40a8bfe1866623a4cc5669b", "lessThan": "1e75a8255f11c81fb07e81e5029cfd75804350a0", "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/iommu/amd/iommu.c" ], "versions": [ { "version": "3.0", "status": "affected" }, { "version": "0", "lessThan": "3.0", "status": "unaffected", "versionType": "semver" }, { "version": "6.6.148", "lessThanOrEqual": "6.6.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.12.101", "lessThanOrEqual": "6.12.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.18.42", "lessThanOrEqual": "6.18.*", "status": "unaffected", "versionType": "semver" }, { "version": "7.1.6", "lessThanOrEqual": "7.1.*", "status": "unaffected", "versionType": "semver" }, { "version": "7.2", "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": "3.0", "versionEndExcluding": "6.6.148" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "3.0", "versionEndExcluding": "6.12.101" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "3.0", "versionEndExcluding": "6.18.42" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "3.0", "versionEndExcluding": "7.1.6" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "3.0", "versionEndExcluding": "7.2" } ] } ] } ], "references": [ { "url": "https://git.kernel.org/stable/c/ab7faf5a172ebfdc423ebb3eea4d472740de82f9" }, { "url": "https://git.kernel.org/stable/c/93494bd446396c257fb589f59894577e96e406e2" }, { "url": "https://git.kernel.org/stable/c/d053eb7e09e10cbdca3fca8b35c1017d438091b2" }, { "url": "https://git.kernel.org/stable/c/02f8cefa2ad95ea3754f0cfd6fbae7f866202ccb" }, { "url": "https://git.kernel.org/stable/c/1e75a8255f11c81fb07e81e5029cfd75804350a0" } ], "title": "iommu/amd: Wait for completion instead of returning early in iommu_completion_wait()", "x_generator": { "engine": "bippy-1.2.0" } } } }