HIGH Introduced in 4.0
arm ShowPte UAF
CVE-2026-90303
CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
01Description
In the Linux kernel, the following vulnerability has been resolved: ARM: 9485/1: mm: acquire mmap write lock around show_pte() for user faults When CONFIG_DEBUG_USER=y, and cmdline "user_debug=31" is set, a user fault may trigger show_pte() without any lock. If another thread in the same process concurrently calls munmap(), the page table pages may be freed while show_pte() is still traversing them, causing a use-after-free in show_pte(). If CONFIG_ARM_LPAE=y, this may cause a kernel panic if the pages table of PMD are freed when show_pte() is running. Acquire mmap_write_lock() around show_pte() for user faults to fix the contention. For user faults, additionally restrict that show_pte() is called only when the addr is a user-space address (addr < TASK_SIZE). This is because the lock of tsk->mm only protects the virtual memory of user address space, furthermore, dumping the page tables of a kernel-space address for user faults is unnecessary and may have security implications. Keep everything unchanged for kernel faults, because the kernel is already in the "oops" state, acquiring a lock may risk a deadlock.
02KernelScan AI Analysis
Risk summary
On 32-bit ARM systems compiled with CONFIG_DEBUG_USER and booted with user_debug=31, a local unprivileged user can race a concurrent munmap against the kernel's page-table diagnostic dump triggered by a user fault, causing a use-after-free on freed page table pages. This can crash the kernel and potentially corrupt memory. The vulnerability is gated behind a non-default debug configuration, limiting exposure to development or debug-enabled deployments.
Vulnerability analysis
When ARM user-fault debug reporting is enabled via a specific kernel command-line option, the kernel dumps page table entries on unhandled user-space faults without holding the necessary lock. If another thread in the same process concurrently unmaps memory, the page table pages being traversed can be freed mid-walk, causing a use-after-free. The fix acquires the proper lock before dumping page tables for user faults and restricts the dump to user-space addresses only, since that lock protects only user virtual memory and dumping kernel-space page tables from a user fault context is unnecessary. Any local unprivileged user on an ARM system with this debug configuration enabled can trigger the race by running a multithreaded process that faults in one thread while unmapping in another.
Lifecycle
03Fix Versions
| Branch | Introduced | Fixed in | Patch commit |
|---|---|---|---|
| 5.15 | 4.0 | 5.15.221 | 07e4d5380f2a |
| mainline | 4.0 | 7.3-rc1 | 1039bffd6ae9 |
| 6.1 | 4.0 | 6.1.188 | 2a14d7797a49 |
| 6.12 | 4.0 | 6.12.110 | 59bbf86d0ff9 |
| 6.6 | 4.0 | 6.6.157 | 63e3c958a602 |
| 7.2 | 4.0 | 7.2.6 | 720408d98d9f |
| 5.10 | 4.0 | 5.10.270 | ab14f07952ad |
| 6.18 | 4.0 | 6.18.52 | c71f9a56520b |
| cip/6.12 | 4.0 | 6.12.111-cip32 | 59bbf86d0ff9 |