CVE-2026-98520

Received Received - Intake

Use-After-Free in AMDKFD Linux Kernel Driver

Publication date: 2026-10-18

Last updated on: 2026-10-19

Assigner: kernel.org

Description

In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: fix use-after-free and multi-container gap in kfd_dev_mapping kfd_dev_mapping caches the address_space of the first /dev/kfd opener so that the GPU reset path can call unmap_mapping_range() to zap all userspace mappings of doorbell and MMIO ranges. This design has two bugs that both manifest under SRIOV with multiple containers: 1. Use-after-free / rwsem deadlock. The cached pointer refers to an inode owned by the first opener's container. When that container exits and its inode is released, kfd_dev_mapping becomes a dangling pointer. A subsequent GPU reset dereferences it inside unmap_mapping_range(), which takes i_mmap_rwsem on the freed inode, causing a hard hang observable as an uninterruptible rwsem wait. 2. Multi-container gap. Only the first opener's address_space is cached; VMAs created by later openers live in a different address_space and are never reached by unmap_mapping_range(). After a GPU reset those stale mappings keep doorbell and MMIO pages accessible to guest userspace with no GPU behind them, risking PCIe transaction timeouts and NMI panics. Fix both bugs with the same approach used by DRM core (drm_drv.c): create a private pseudo-filesystem at module init time and allocate one anonymous inode from it. In kfd_open() redirect every opener's file->f_mapping to that inode's address_space. The inode is module-owned, lives exactly as long as the amdgpu module, and collects VMAs from all openers in one address_space. A single unmap_mapping_range() call in the reset path then correctly reaches every container's mappings with no dangling pointer risk. The hang manifests as an NMI backtrace on the GPU reset workqueue stuck spinning in rwsem_down_read_slowpath() with a corrupted i_mmap_rwsem: Workqueue: amdgpu-reset-dev xgpu_ai_mailbox_flr_work [amdgpu] Call Trace: <TASK> kvm_wait+0x1f/0x40 __pv_queued_spin_lock_slowpath+0x31d/0x3a0 _raw_spin_lock_irq+0x51/0x80 rwsem_down_read_slowpath+0xb3/0x550 down_read+0x48/0xd0 unmap_mapping_range+0x71/0x140 kfd_dev_unmap_mapping_range+0x5b/0x140 [amdgpu] amdgpu_amdkfd_clear_kfd_mapping+0xd8/0x190 [amdgpu] amdgpu_device_gpu_recover+0x232/0x450 [amdgpu] xgpu_ai_mailbox_flr_work+0xb5/0xc0 [amdgpu] process_one_work+0x18e/0x3e0 worker_thread+0x2e3/0x420 kthread+0x10a/0x230 (cherry picked from commit 1128b4a52de1572e87431de837fd9850cb99542c)

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published

2026-10-18

Last Modified

2026-10-19

Generated

2026-10-19

AI Q&A

2026-10-18

EPSS Evaluated

N/A

NVD

CVE-2026-98520

EUVD

EUVD-2026-103760

Affected Vendors & Products

Vendor	Product	Version / Range
Linux	Linux	70cadefcc6160c575b04f763ada34c20e868d577
Linux	Linux	70cadefcc6160c575b04f763ada34c20e868d577
Linux	Linux	7.2

Helpful Resources

Exploitability

CWE

KEV

CWE ID	Description
CWE-UNKNOWN

Attack-Flow Graph

Executive Summary

This vulnerability in the Linux kernel involves a use-after-free and multi-container gap in the kfd_dev_mapping function. It affects AMD GPU devices under SRIOV with multiple containers. The issue causes a dangling pointer and rwsem deadlock when the first container exits, leading to a system hang during GPU reset. The fix involves redirecting all file mappings to a private pseudo-filesystem to avoid dangling pointers.

Detection Guidance

This vulnerability manifests as a hard hang or NMI backtrace during GPU reset operations, specifically in the amdgpu-reset-dev workqueue. Check system logs for entries indicating rwsem_down_read_slowpath hangs or NMI panics during GPU operations. Look for traces involving amdgpu-reset-dev, xgpu_ai_mailbox_flr_work, or kfd_dev_unmap_mapping_range.

Impact Analysis

This vulnerability can cause system hangs or crashes during GPU reset operations, especially in virtualized environments with multiple containers. It may lead to PCIe transaction timeouts and NMI panics, potentially causing data corruption or service disruption.

Compliance Impact

This vulnerability does not directly affect compliance with GDPR or HIPAA as it pertains to a Linux kernel issue involving GPU memory management and potential system hangs. However, if exploited, it could lead to system instability or crashes, potentially causing data loss or unauthorized access in environments where such disruptions violate compliance requirements.

Mitigation Strategies

Apply the Linux kernel patch that fixes the use-after-free and multi-container gap in kfd_dev_mapping. The fix involves redirecting file mappings to a private pseudo-filesystem inode to avoid dangling pointers. Update to a kernel version containing commit 1128b4a52de1572e87431de837fd9850cb99542c or later.

Hi! I’m here to help you understand CVE-2026-98520. Ask me anything about the vulnerability, its impact, or mitigation strategies.

0/70