From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-97620][MODERATE 7.0] drm/xe: Flush LSC untyped L1 dataport cache after rcs/ccs batches [ Upstream Date: Fri, 25 Sep 2026 09:40:43 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-97620 X-AL-KERNEL-Priority: MODERATE 7.0 X-AL-KERNEL-Severity: MODERATE 7.0 X-AL-KERNEL-Base-Severity: MODERATE X-AL-KERNEL-KPANIC: YES X-AL-KERNEL-ActionableScore: 5 X-AL-KERNEL-ActionableScore-Lower: 4 X-AL-KERNEL-Commit: fab569a69c3ae9beced68307e36048efa2709bec List-Id: CVE: CVE-2026-97620 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: drm/xe: Flush LSC untyped L1 dataport cache after rcs/ccs batches [ Upstream Commit: fab569a69c3ae9beced68307e36048efa2709bec Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fab569a69c3ae9beced68307e36048efa2709bec Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97620 Analysis date: Fri, 25 Sep 2026 09:40:43 -0400 ActionableScore: 5 ActionableScore lower bound: 4 Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review Manual review required: YES Summary: No one-sentence report phrase was found. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97620 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97620 The original announcement does not normally provide a security severity estimate, CVSS assessment, or enough information to determine whether the reported kernel bug represents a practically relevant security issue. This report was generated by AL-KERNEL, an AI-assisted Linux kernel vulnerability analysis system developed by Alexander Larkin. It combines an autonomous classifier with LLM-assisted technical analysis and a separate ActionableScore mechanism. The purpose of this report is to prioritize Linux kernel CVEs before manual review, identify cases that require prompt investigation, and support automatic closure of issues that are unlikely to have meaningful security impact. Published priority for this report: MODERATE 7.0 Manual review required: YES A detailed explanation of the methodology and priority rules is included at the end of this message. ====================================================================== AL-KERNEL CLASSIFICATION RESULT ====================================================================== CVE-2026-97620 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-97620 CHECK WITH IMPACT FROM ORIG NN LOW Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N';CWE-226;*CWE-200;CWE-524;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N';BEST CVSS score: '5.5';DESCR 'A missing flush of the untyped L1 dataport cache in the DRM Xe driver can leave stale GPU cached data visible after memory is reclaimed or reused by another GPU context. A local process with permitted GPU access may potentially observe data left by a previous context through reused GPU memory. For the CVSS the PR:L is used because an attacker needs local access to a DRM render node or another authorized GPU interface. The issue is not directly network reachable. Impact is primarily confidentiality exposure. No memory corruption, integrity impact, privilege escalation, or direct denial of service primitive is demonstrated by the patch. Xe2 and later platforms are relevant to this kernel side mitigation. MTL era platforms require separate validation because the described cache flush coupling is not reliable there.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) YES INIT LEAK HARDWARE KPANIC INCREASED_FROM_LOW_BASED_ON_REQUIREMANUALCHECK YES NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=5 ActionableScoreLower=4 PrimitiveClass=CLASS_B RealUAF=NO HostAdminRequired=NO NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO 1. ActionableScore * Raw conservative score: 4 * Conservative score: 4 * Raw paranoid score: 5 * Paranoid score: 5 * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: 2. Signal breakdown * +1 Local unprivileged trigger. A local process permitted to access an Intel DRM Xe render node can submit GPU workloads and participate in GPU memory allocation and reuse without host administrator privilege. * +1 Confidentiality plausible. The commit explicitly states that stale untyped L1 cache data can leak after memory is reclaimed or evicted. * +1 Cross-context memory ownership and isolation violation. Data associated with a previous GPU context can remain observable after the backing memory is reused by another context. * +1 Broad active execution path. The affected render and compute submission path is core DRM Xe GPU functionality on affected Xe2 hardware, not a diagnostic or optional debug path. * +1 Paranoid only. The cache data may contain a significant portion of a previously used GPU allocation. A later authorized GPU context may plausibly recover more than incidental stale bytes, supporting a higher confidentiality interpretation. Not awarded: * No Real UAF. This is stale cache visibility after memory reuse, not dereference of a freed kernel object. * No arbitrary kernel write or directed corruption is demonstrated. * No integrity impact is demonstrated. User-visible application corruption does not establish unauthorized modification of another principals data. * No privilege escalation, callback control, controlled kernel-object reclaim, or kernel arbitrary read or write is established. * No remote reachability is established. * No device-specific negative penalty is applied. Xe2 hardware is hardware-specific but not inherently rare or physical-only. * No race penalty is applied. The issue is a missing cache-maintenance operation, not an externally narrow timing race. 3. Reachability analysis The relevant attacker is a local process that has authorized access to a DRM Xe render node or equivalent GPU submission interface. Such access is commonly granted to desktop users, rendering clients, compute users, or services in the render group and does not require DRM master or display-mode-setting authority. The trigger sequence is local GPU activity. A process submits work that populates untyped L1 cache data, the associated GPU memory is reclaimed or evicted, and another GPU context obtains reused memory before the cache content is reliably flushed. The patch directly identifies this memory-reuse boundary as security relevant. Host administrator privilege is not required for normal render-node command submission once the render device is accessible. However, device-node permissions, cgroup device policy, container device passthrough, and the exact DRM Xe driver deployment determine whether a particular unprivileged user can access the affected interface. NamespaceScopedPrivilege is UNKNOWN. DRM render-node access is not established by the supplied context as a fully namespace-scoped operation. A container may require prior host device exposure or delegated DRM access. The issue is limited to relevant Intel platforms. The kernel-side explicit flush is enabled only for Xe2 and later, where the PIPE_CONTROL flush bit is reliable. The commit states that MTL-era systems have different hardware behavior and require separate validation of Mesa and platform-specific mitigations. 4. Primitive and severity interpretation The primitive is a GPU memory-ownership and cache-coherency failure. The DRM Xe kernel driver may release GPU memory for reuse while stale content remains in the LSC untyped L1 dataport cache. A later GPU context can potentially receive data associated with a prior context instead of only the contents expected for its own allocation. This is neither fixed stale storage nor a kernel heap UAF. It is an active cross-context ownership violation involving cached GPU memory contents after the original allocation has been retired or reassigned. The commit explicitly identifies stale-data leakage as the security consequence. The demonstrated practical consequence is confidentiality exposure between GPU contexts. The patch does not establish a write primitive, active kernel-object corruption, privilege escalation, host compromise, or a direct denial-of-service mechanism. PrimitiveClass is CLASS_B because a meaningful cross-context memory-lifetime and ownership primitive is established, but controlled extraction scope and the sensitivity of affected prior allocations remain unresolved. ImportantCandidate is YES because the difference between incidental stale data exposure and substantial cross-process GPU data disclosure requires platform and workload-specific manual review. AutoDowngradeAllowed is NO because automatically treating this as ordinary graphics corruption could discard a concrete security-relevant stale-data disclosure primitive. The issue is not proven Important severity, but it should not be closed automatically without verifying device exposure, affected GPU generation, render-node access policy, and whether untrusted GPU clients can coexist. 5. One-sentence report phrase A missing DRM Xe untyped L1 dataport cache flush can expose stale GPU data from a previous context after memory reuse, creating a local cross-context confidentiality risk for users authorized to submit GPU workloads. 6. Manual review recommendation MANUAL CHECK REQUIRED The commit explicitly identifies a stale-data leakage condition across GPU memory reuse. Review affected Xe2 hardware, DRM render-node permissions, multi-user or multi-tenant GPU usage, container GPU passthrough, and whether sensitive workloads share the affected GPU. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: drm/xe: Flush LSC untyped L1 dataport cache after rcs/ccs batches [ Upstream Commit: fab569a69c3ae9beced68307e36048efa2709bec Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fab569a69c3ae9beced68307e36048efa2709bec Changed files: drivers/gpu/drm/xe/instructions/xe_gpu_commands.h drivers/gpu/drm/xe/xe_ring_ops.c Diff excerpt: Not included in this email. See the upstream URL for the full patch. Full patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fab569a69c3ae9beced68307e36048efa2709bec ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97620 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97620 The original CVE announcement normally does not include a security-level estimate. In particular, it may not contain a CVSS assessment, an impact level, or enough information to determine whether the reported bug is a practically relevant security issue. One purpose of this parallel CVE list is to provide that missing technical and prioritization information. The original goal of the AL-KERNEL project was to prioritize Linux kernel CVE analysis automatically before manual review. The system can also help identify non-security issues that may be suitable for automatic closure. This report was generated by AL-KERNEL, an AI-assisted Linux kernel vulnerability analysis system developed by Alexander Larkin. The first analysis stage combines an autonomous classifier with additional LLM-based analysis. The autonomous classifier runs locally on a CPU and is based on a backpropagation neural network. Together, these mechanisms produce a technical vulnerability description, identify likely weakness types, estimate CVSS severity, and provide input for ActionableScore. Two CVSS estimates are retained because incomplete kernel vulnerability information often permits more than one defensible interpretation: Conservative CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N The Best / paranoid CVSS score: 5.5 The conservative vector represents a lower-impact interpretation. The Best/paranoid vector intentionally represents a plausible upper-bound interpretation and should not automatically be treated as demonstrated real-world impact. CVSS may also need to be adjusted for a particular Linux deployment, because actual reachability, privileges, enabled kernel configuration, hardware, namespaces, exposed device nodes, and other environmental conditions can differ significantly between systems. A separate ActionableScore mechanism evaluates practical remediation urgency. Its analysis may include reachability, attack prerequisites, subsystem exposure, memory-corruption characteristics, denial-of-service reliability, and possible confidentiality, integrity, or privilege-escalation impact. Conservative ActionableScore: 4 Paranoid ActionableScore: 5 The final base severity is taken directly from the second tab-separated field of the AL-KERNEL classification result. ActionableScore does not replace or independently override that final AL-KERNEL decision, and if ActionableScore adjusted impact level of ALKERNEL, then you would see self-readable flags above like INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN7. For an AL-KERNEL result of MODERATE, this report uses the following additional presentation split: ActionableScore below 5 -> MODERATE REGULAR ActionableScore 5 or more -> MODERATE 7.0 The distinction between MODERATE REGULAR and MODERATE 7.0 makes it possible to identify Moderate issues that should receive manual analysis and fixes before lower-priority MODERATE REGULAR issues. In many cases, MODERATE REGULAR fixes may wait for a later rebase or routine update. There is one override in which MODERATE REGULAR becomes MODERATE 7.0 even when the ActionableScore is below 5. When the AL-KERNEL result contains the KPANIC flag, a MODERATE result is always presented as MODERATE 7.0. The KPANIC flag selected with few regexps without usage of AI at all, so it helps to detect cases when Kernel Crash happens and similar (to filter False-Negative results from the LLM usage). KPANIC indicates that a reliable kernel crash, kernel panic, or similarly serious kernel availability impact was identified by the classification workflow. AL-KERNEL base severity for this report: MODERATE KPANIC detected for this report: YES Published priority for this report (same as in Subject): MODERATE 7.0 These results are intended to support engineering triage. They are machine-generated estimates, and cases marked for manual review should be validated by a human security engineer before final disposition. For more info read docs linked from here: https://kernelcve.org/ (and you can submit you own patch there to generate such a report for non-existant CVE-id yet). Note that in many cases this AI tool selects higher severity, than real is (means you can expect Importants instead of Moderate 7.0 or Moderates 7.0 instead of regular Moderates). If you see such cases, please use reply email interface to add additional manual analyses info to this particular CVE. And please, please, let me know when you see Lows instead of Importants or Important instead of Low (because particular for such cases I need to tune this AI tool to make it better for this one and next similar). My contact email for such notifications is alexanjelausa@gmail.com (and both send reply to CVE record itself too and see "reply" button below for howto reply).