FlawAtlas
Search the atlas
CVE-2025-38424 Not scored

perf: Fix sample vs do_exit()

In the Linux kernel, the following vulnerability has been resolved: perf: Fix sample vs do_exit() Baisheng Gao reported an ARM64 crash, which Mark decoded as being a synchronous external abort -- most likely due to trying to access MMIO in bad ways. The crash further shows perf trying to do a user stack sample while in exit_mmap()'s tlb_finish_mmu() -- i.e. while tearing down the address space it is trying to access. It turns out that we stop perf after we tear down the userspace mm; a receipie for disaster, since perf likes to access userspace for various reasons. Flip this order by moving up where we stop perf in do_exit(). Additionally, harden PERF_SAMPLE_CALLCHAIN and PERF_SAMPLE_STACK_USER to abort when the current task does not have an mm (exit_mm() makes sure to set current->mm = NULL; before commencing with the actual teardown). Such that CPU wide events don't trip on this same problem.

Exploit probability 0.2%
Published July 25, 2025
Required by Not available
Last source change July 15, 2026

02 / AFFECTED SOFTWARE

Affected packages

Linux Kernel
Unknown Unknown

03 / CONNECTIONS

Connected vulnerabilities

related OPENSUSE-SU-2025:20081-1
related SUSE-SU-2025:02853-1
related SUSE-SU-2025:02923-1
related SUSE-SU-2025:02969-1
related SUSE-SU-2025:02996-1
related SUSE-SU-2025:02997-1
related SUSE-SU-2025:03011-1
related SUSE-SU-2025:03023-1
related SUSE-SU-2025:03204-1
related SUSE-SU-2025:20577-1
related SUSE-SU-2025:20586-1
related SUSE-SU-2025:20601-1
related SUSE-SU-2025:20602-1
related SUSE-SU-2025:21074-1
related SUSE-SU-2025:21139-1
related SUSE-SU-2025:21179-1

04 / EVIDENCE

Source records

Open Source Vulnerabilities CVE-2025-38424

In the Linux kernel, the following vulnerability has been resolved: perf: Fix sample vs do_exit() Baisheng Gao reported an ARM64 crash, which Mark decoded as being a synchronous external abort -- most likely due to trying to access MMIO in bad ways. The crash further shows perf trying to do a user stack sample while in exit_mmap()'s tlb_finish_mmu() -- i.e. while tearing down the address space it is trying to access. It turns out that we stop perf after we tear down the userspace mm; a receipie for disaster, since perf likes to access userspace for various reasons. Flip this order by moving up where we stop perf in do_exit(). Additionally, harden PERF_SAMPLE_CALLCHAIN and PERF_SAMPLE_STACK_USER to abort when the current task does not have an mm (exit_mm() makes sure to set current->mm = NULL; before commencing with the actual teardown). Such that CPU wide events don't trip on this same problem.

View original source

05 / REFERENCES

Further evidence