{"api_version":"1","generated_at":"2026-08-24T20:41:46+00:00","cve":"CVE-2026-68384","urls":{"html":"https://cve.report/CVE-2026-68384","api":"https://cve.report/api/cve/CVE-2026-68384.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-68384","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-68384"},"summary":{"title":"drm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves","description":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves\n\nxe_bo_move() attaches VF CCS read/write batch buffers (BBs) to a BO\nafter it transitions NULL/SYSTEM -> TT, and detaches them after it\ntransitions TT -> SYSTEM. Both operations were done synchronously on\nthe CPU immediately after building the move's copy/clear fence,\nwithout waiting for that fence to signal. This creates two races with\nVF migration:\n\n- Attach happens too late relative to the copy job it is meant to\n  protect. If the copy job is submitted before the CCS BBs are\n  attached, a VF migration event that pauses execution mid-copy can\n  observe partially copied CCS metadata without the attach state\n  needed to correctly save/restore it.\n\n- Detach happens too early relative to the copy job that moves data\n  out of TT. The CCS BBs are torn down right after the copy fence is\n  obtained, while the actual blit may still be in flight. A VF\n  migration event that pauses execution mid-copy can then race the\n  save/restore path against the still-running blit, and the CCS BBs\n  it would need to make sense of the paused state have already been\n  removed.\n\nFix both races:\n\n- Move the attach call to before the copy/clear job is submitted, so\n  the CCS BBs are already registered by the time the copy runs. On\n  attach failure, unwind and bail out of the move. xe_migrate_ccs_rw_copy()\n  now takes the destination resource explicitly, since bo->ttm.resource\n  is not updated to the new resource until after the move commits.\n\n- Detach only after explicitly waiting for the copy fence to signal,\n  instead of tearing down the CCS BBs immediately after obtaining it.\n\nWhile here, also fix xe_sriov_vf_ccs_attach_bo() to properly unwind and\npropagate errors: the per-context loop previously never broke out on\nerror, silently discarding earlier failures. Unwind by clearing each\nattached context directly via xe_migrate_ccs_rw_copy_clear() instead of\nreusing xe_sriov_vf_ccs_detach_bo(), which requires both contexts to be\nattached before it will clean up either one.\n\n(cherry picked from commit d45ad0aa7a1eb5d7288b5ed948b05695611dc39e)","state":"PUBLISHED","assigner":"Linux","published_at":"2026-08-10 13:20:31","updated_at":"2026-08-17 06:17:46"},"problem_types":[],"metrics":[{"version":"3.1","source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","score":"7.8","severity":"HIGH","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","data":{"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","attackVector":"LOCAL","attackComplexity":"LOW","privilegesRequired":"LOW","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"}},{"version":"3.1","source":"CNA","type":"DECLARED","score":"7.8","severity":"HIGH","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","data":{"baseScore":7.8,"baseSeverity":"HIGH","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","version":"3.1"}}],"references":[{"url":"https://git.kernel.org/stable/c/35ba43b541117bfb595e4b807ba447cf4335cc7d","name":"https://git.kernel.org/stable/c/35ba43b541117bfb595e4b807ba447cf4335cc7d","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/56441f9e08ad68697295b8835266d2bc48ab59b5","name":"https://git.kernel.org/stable/c/56441f9e08ad68697295b8835266d2bc48ab59b5","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/f2ebfd5cc87f1393a30c8b8b0a6c20cb22cffa97","name":"https://git.kernel.org/stable/c/f2ebfd5cc87f1393a30c8b8b0a6c20cb22cffa97","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-68384","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-68384","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 864690cf4dd62482b6dd049d82c509886c904303 35ba43b541117bfb595e4b807ba447cf4335cc7d git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 864690cf4dd62482b6dd049d82c509886c904303 f2ebfd5cc87f1393a30c8b8b0a6c20cb22cffa97 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 864690cf4dd62482b6dd049d82c509886c904303 56441f9e08ad68697295b8835266d2bc48ab59b5 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 6.18","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.18 semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.18.42 6.18.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.1.6 7.1.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.2 * original_commit_for_fix","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":{"cve_year":"2026","cve_id":"68384","cve":"CVE-2026-68384","epss":"0.001340000","percentile":"0.033540000","score_date":"2026-08-17","updated_at":"2026-08-18 00:11:47"},"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"defaultStatus":"unaffected","product":"Linux","programFiles":["drivers/gpu/drm/xe/xe_bo.c","drivers/gpu/drm/xe/xe_migrate.c","drivers/gpu/drm/xe/xe_migrate.h","drivers/gpu/drm/xe/xe_sriov_vf_ccs.c","drivers/gpu/drm/xe/xe_sriov_vf_ccs.h"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"lessThan":"35ba43b541117bfb595e4b807ba447cf4335cc7d","status":"affected","version":"864690cf4dd62482b6dd049d82c509886c904303","versionType":"git"},{"lessThan":"f2ebfd5cc87f1393a30c8b8b0a6c20cb22cffa97","status":"affected","version":"864690cf4dd62482b6dd049d82c509886c904303","versionType":"git"},{"lessThan":"56441f9e08ad68697295b8835266d2bc48ab59b5","status":"affected","version":"864690cf4dd62482b6dd049d82c509886c904303","versionType":"git"}]},{"defaultStatus":"affected","product":"Linux","programFiles":["drivers/gpu/drm/xe/xe_bo.c","drivers/gpu/drm/xe/xe_migrate.c","drivers/gpu/drm/xe/xe_migrate.h","drivers/gpu/drm/xe/xe_sriov_vf_ccs.c","drivers/gpu/drm/xe/xe_sriov_vf_ccs.h"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"status":"affected","version":"6.18"},{"lessThan":"6.18","status":"unaffected","version":"0","versionType":"semver"},{"lessThanOrEqual":"6.18.*","status":"unaffected","version":"6.18.42","versionType":"semver"},{"lessThanOrEqual":"7.1.*","status":"unaffected","version":"7.1.6","versionType":"semver"},{"lessThanOrEqual":"*","status":"unaffected","version":"7.2","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"cpeMatch":[{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"6.18.42","versionStartIncluding":"6.18","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.1.6","versionStartIncluding":"6.18","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.2","versionStartIncluding":"6.18","vulnerable":true}],"negate":false,"operator":"OR"}]}],"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves\n\nxe_bo_move() attaches VF CCS read/write batch buffers (BBs) to a BO\nafter it transitions NULL/SYSTEM -> TT, and detaches them after it\ntransitions TT -> SYSTEM. Both operations were done synchronously on\nthe CPU immediately after building the move's copy/clear fence,\nwithout waiting for that fence to signal. This creates two races with\nVF migration:\n\n- Attach happens too late relative to the copy job it is meant to\n  protect. If the copy job is submitted before the CCS BBs are\n  attached, a VF migration event that pauses execution mid-copy can\n  observe partially copied CCS metadata without the attach state\n  needed to correctly save/restore it.\n\n- Detach happens too early relative to the copy job that moves data\n  out of TT. The CCS BBs are torn down right after the copy fence is\n  obtained, while the actual blit may still be in flight. A VF\n  migration event that pauses execution mid-copy can then race the\n  save/restore path against the still-running blit, and the CCS BBs\n  it would need to make sense of the paused state have already been\n  removed.\n\nFix both races:\n\n- Move the attach call to before the copy/clear job is submitted, so\n  the CCS BBs are already registered by the time the copy runs. On\n  attach failure, unwind and bail out of the move. xe_migrate_ccs_rw_copy()\n  now takes the destination resource explicitly, since bo->ttm.resource\n  is not updated to the new resource until after the move commits.\n\n- Detach only after explicitly waiting for the copy fence to signal,\n  instead of tearing down the CCS BBs immediately after obtaining it.\n\nWhile here, also fix xe_sriov_vf_ccs_attach_bo() to properly unwind and\npropagate errors: the per-context loop previously never broke out on\nerror, silently discarding earlier failures. Unwind by clearing each\nattached context directly via xe_migrate_ccs_rw_copy_clear() instead of\nreusing xe_sriov_vf_ccs_detach_bo(), which requires both contexts to be\nattached before it will clean up either one.\n\n(cherry picked from commit d45ad0aa7a1eb5d7288b5ed948b05695611dc39e)"}],"metrics":[{"cvssV3_1":{"baseScore":7.8,"baseSeverity":"HIGH","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"scenarios":[{"lang":"en","value":"AV:L - The affected code is reached through DRM/TTM buffer-object migration in the xe driver, driven by local ioctls on a render node (BO create/exec/eviction pressure). No network or remote input reaches xe_bo_move(), so local access to the GPU device node is required.\nAC:L - A local user can continuously drive TT<->SYSTEM BO moves, keeping copy blits in flight indefinitely and making the attach/detach window essentially always open when a VF migration occurs, so the race is reliably winnable from the attacker's side.\nPR:L - Only an unprivileged local account with access to /dev/dri/renderD* is needed - render nodes are routinely accessible to normal users, graphical sessions and containers, and no capability check guards xe_bo_move() or the CCS attach/detach paths.\nUI:N - The attacker only needs to generate ordinary GPU buffer-object traffic; no victim action, file open, or mount is required, and VF migration is a routine automated hypervisor operation rather than a user-initiated one.\nS:U - The corrupted CCS batch buffers, pool nodes and metadata all belong to the guest VF kernel's own security authority; nothing crosses the VF/PF or IOMMU boundary.\nC:H - Detaching frees the sub-allocated CCS batch-buffer node back into the shared pool while the blit and the migration save/restore path may still use it, so recycled command buffers and stale PTEs can cause CCS metadata to be read from or copied out of unrelated buffer objects, exposing other users' GPU buffer contents.\nI:H - A freed-and-reused batch buffer executed by the GPU, plus attach state built against the wrong (old) resource, gives an attacker-influenceable write of CCS/compression metadata into memory that no longer belongs to the BO, corrupting other buffer objects and kernel-managed pool state.\nA:H - Executing torn-down or half-initialized batch buffers corrupts the migration blit and typically ends in a GPU engine hang/reset or failed VF migration, and the unwound-attach bug leaves inconsistent bb_ccs state that trips xe_assert() and leaks pool nodes until the device is unusable."}]}],"providerMetadata":{"dateUpdated":"2026-08-17T05:04:30.813Z","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux"},"references":[{"url":"https://git.kernel.org/stable/c/35ba43b541117bfb595e4b807ba447cf4335cc7d"},{"url":"https://git.kernel.org/stable/c/f2ebfd5cc87f1393a30c8b8b0a6c20cb22cffa97"},{"url":"https://git.kernel.org/stable/c/56441f9e08ad68697295b8835266d2bc48ab59b5"}],"title":"drm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves","x_generator":{"engine":"bippy-1.2.0"}}},"cveMetadata":{"assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","assignerShortName":"Linux","cveId":"CVE-2026-68384","datePublished":"2026-08-10T12:04:03.240Z","dateReserved":"2026-07-30T09:28:09.387Z","dateUpdated":"2026-08-17T05:04:30.813Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-08-10 13:20:31","lastModifiedDate":"2026-08-17 06:17:46","problem_types":[],"metrics":{"cvssMetricV31":[{"source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","cvssData":{"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","attackVector":"LOCAL","attackComplexity":"LOW","privilegesRequired":"LOW","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"},"exploitabilityScore":1.8,"impactScore":5.9}]},"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"68384","Ordinal":"1","Title":"drm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves","CVE":"CVE-2026-68384","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"68384","Ordinal":"1","NoteData":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves\n\nxe_bo_move() attaches VF CCS read/write batch buffers (BBs) to a BO\nafter it transitions NULL/SYSTEM -> TT, and detaches them after it\ntransitions TT -> SYSTEM. Both operations were done synchronously on\nthe CPU immediately after building the move's copy/clear fence,\nwithout waiting for that fence to signal. This creates two races with\nVF migration:\n\n- Attach happens too late relative to the copy job it is meant to\n  protect. If the copy job is submitted before the CCS BBs are\n  attached, a VF migration event that pauses execution mid-copy can\n  observe partially copied CCS metadata without the attach state\n  needed to correctly save/restore it.\n\n- Detach happens too early relative to the copy job that moves data\n  out of TT. The CCS BBs are torn down right after the copy fence is\n  obtained, while the actual blit may still be in flight. A VF\n  migration event that pauses execution mid-copy can then race the\n  save/restore path against the still-running blit, and the CCS BBs\n  it would need to make sense of the paused state have already been\n  removed.\n\nFix both races:\n\n- Move the attach call to before the copy/clear job is submitted, so\n  the CCS BBs are already registered by the time the copy runs. On\n  attach failure, unwind and bail out of the move. xe_migrate_ccs_rw_copy()\n  now takes the destination resource explicitly, since bo->ttm.resource\n  is not updated to the new resource until after the move commits.\n\n- Detach only after explicitly waiting for the copy fence to signal,\n  instead of tearing down the CCS BBs immediately after obtaining it.\n\nWhile here, also fix xe_sriov_vf_ccs_attach_bo() to properly unwind and\npropagate errors: the per-context loop previously never broke out on\nerror, silently discarding earlier failures. Unwind by clearing each\nattached context directly via xe_migrate_ccs_rw_copy_clear() instead of\nreusing xe_sriov_vf_ccs_detach_bo(), which requires both contexts to be\nattached before it will clean up either one.\n\n(cherry picked from commit d45ad0aa7a1eb5d7288b5ed948b05695611dc39e)","Type":"Description","Title":"drm/xe/vf: Fix VF CCS attach/detach race with in-flight BO moves"}]}}}