From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-53374][IMPORTANT] drm/amdgpu: zero-initialize GART table on allocation Date: Sun, 19 Jul 2026 06:43:46 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-53374 X-AL-KERNEL-Priority: IMPORTANT X-AL-KERNEL-Severity: IMPORTANT X-AL-KERNEL-Base-Severity: IMPORTANT X-AL-KERNEL-KPANIC: YES X-AL-KERNEL-ActionableScore: 7 X-AL-KERNEL-ActionableScore-Lower: 5 X-AL-KERNEL-Commit: 40df11255d71b02e20e70579f1b12b687e396e26 List-Id: CVE: CVE-2026-53374 Priority: IMPORTANT AL-KERNEL base severity: IMPORTANT KPANIC flag: YES Patch: drm/amdgpu: zero-initialize GART table on allocation Commit: 40df11255d71b02e20e70579f1b12b687e396e26 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=40df11255d71b02e20e70579f1b12b687e396e26 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-53374 Analysis date: Sun, 19 Jul 2026 06:43:46 -0400 ActionableScore: 7 ActionableScore lower bound: 5 Actionable bucket: Strong Important candidate / Actionable Moderate at minimum Manual review required: YES Summary: AMDGPU GART table memory was not zero-initialized on allocation, allowing stale uninitialized PTEs to be speculatively cached by the GPU TLB and potentially causing GPU faults, kernel DoS, or unintended GPU access to memory outside the intended mapping. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-53374 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-53374 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: IMPORTANT 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-2026-53374 IMPORTANT CHECK WITH IMPACT FROM ORIG NN LOW Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-908;CWE-665;CWE-824;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'AMDGPU GART page table memory in VRAM was not zero-initialized after allocation, so uninitialized entries from the same cacheline could be speculatively loaded into the GPU TLB as apparently valid PTEs. A local user able to submit GPU work may be able to influence GART mappings and later cause the GPU to use a stale garbage PTE, which can lead to incorrect DMA style memory access, data corruption, GPU fault, or kernel level denial of service. For the CVSS the PR:L value is used because reliable triggering generally requires local access to the DRM GPU device or render node, which is commonly available to non-root users in desktop and compute deployments. The issue is not network reachable and requires local GPU access. Impact is at least local denial of service and in the paranoid case may include confidentiality and integrity impact because stale valid translation entries can point the GPU at memory outside the intended mapping context. YES REQUIRES MANUAL CHECK';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 7) SKIP CVE-2026-53374 UNKNOWN SKIP No affected files built, so skip this CVE NO - - unknown MAYBE SIMPLEFIX HARDWARE KPANIC MEMORY KPANIC INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN7 - - checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=7 ActionableScoreLower=5 ## 1. ActionableScore * Conservative score: 5 * Paranoid score: 7 * Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**:: ## 2. Signal breakdown Conservative signals: * Local unprivileged trigger: +1. A local user with access to `/dev/dri/renderD*` may submit GPU work on many desktop or compute deployments. * Memory corruption, weak or indirect corruption candidate: +1. The issue involves uninitialized GART PTEs that may be interpreted as valid GPU translations, but the exact target and payload are not clearly attacker-controlled. * Privileged kernel/device-management memory corruption path: +1. The corrupted state is GPU address-translation metadata in a trusted DRM driver path. * Reliable kernel crash / strong DoS: +1. A stale garbage PTE can plausibly cause GPU faults, device hangs, or kernel-level DoS. * Confidentiality impact plausible: +1. A stale valid translation may expose memory outside the intended GPU mapping. * Integrity impact plausible: +1. A stale valid translation may allow writes to unintended memory through GPU access. * Broad/common deployed exposure: +1. AMDGPU is a common DRM driver on systems with AMD GPUs. * Hard or unreliable conditions: -1. Triggering depends on uninitialized VRAM contents, cacheline behavior, speculative GPU TLB loading, and timing/order of PTE writes. * No clearly attacker-controlled payload or target: -1. The garbage PTE value is not shown to be directly controlled. Paranoid adjustment: * The paranoid score treats the stale GART PTE as a stronger DMA-style memory corruption primitive: +1 additional over the weak corruption case. * Weak LPE concern: +1. Privilege escalation is not demonstrated, but stale GPU translations into unintended memory make LPE a plausible manual-review concern. ## 3. Reachability analysis The bug is local, not network reachable. It requires an AMDGPU device and a path where the GART table is allocated in VRAM and then partially populated while uninitialized entries may be speculatively loaded by the GPU TLB. Typical triggering requires local access to the DRM GPU device or render node. On many systems, render nodes are available to non-root users through `render` or `video` group membership, so PR:L is the practical CVSS-style privilege assumption. Containers only matter if the host passes through `/dev/dri` devices or GPU access into the container. This is not a pure admin-only initialization bug because the dangerous state can affect later GPU mappings used by user-submitted GPU workloads. Call-site confidence: medium, because the patch shows the GART allocation path and the commit explains the TLB/PTE interaction, but does not include a concrete reproducer or exploit path. ## 4. Severity interpretation This is stronger than an ordinary Moderate because the affected object is a GPU page table, not a simple driver-local buffer. The realistic demonstrated impact is local DoS or GPU/kernel fault from stale invalid mappings. The theoretical but defensible higher-risk interpretation is unintended GPU DMA-style access to memory outside the intended mapping, which can imply confidentiality and integrity impact. It should not be treated as proven arbitrary code execution or proven LPE. However, because the primitive involves stale valid PTEs in GPU address translation, it should not be auto-closed as a low-risk crash-only issue. ## 5. One-sentence report phrase AMDGPU GART table memory was not zero-initialized on allocation, allowing stale uninitialized PTEs to be speculatively cached by the GPU TLB and potentially causing GPU faults, kernel DoS, or unintended GPU access to memory outside the intended mapping. ## 6. Manual review recommendation MANUAL CHECK REQUIRED Reason: this is a GPU page-table initialization bug with possible stale PTE memory access effects. The practical exploitability is uncertain, but the affected state is security-sensitive enough to require manual review rather than automatic closure. ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: drm/amdgpu: zero-initialize GART table on allocation Commit: 40df11255d71b02e20e70579f1b12b687e396e26 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=40df11255d71b02e20e70579f1b12b687e396e26 Commit description: GART TLB is flushed after unmapping but not after mapping. Since amdgpu_bo_create_kernel() does not zero-initialize the buffer, when a single PTE is written the TLB may speculatively load other uninitialized entries from the same cacheline. Those garbage entries can appear valid, and a subsequent write to another PTE in the same cacheline may cause the GPU to use a stale garbage PTE from the TLB. Fix this by calling memset_io() to zero-initialize the GART table with gart_pte_flags immediately after allocation. Using AMDGPU_GEM_CREATE_VRAM_CLEARED, SDMA-based clear will not work since SDMA needs GART to be initialized to work. Suggested-by: Felix Kuehling Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit d9af826 ) Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman Changed files: drivers/gpu/drm/amd/amdgpu/amdgpu_gart.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=40df11255d71b02e20e70579f1b12b687e396e26 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-53374 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-53374 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:H/PR:L/UI:N/S:U/C:L/I:L/A:H The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H The Best / paranoid CVSS score: 7 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: 5 Paranoid ActionableScore: 7 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: IMPORTANT KPANIC detected for this report: YES Published priority for this report (same as in Subject): IMPORTANT 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).