CVE-2026-99579: mm/shmem: honour F_SEAL_EXEC when an ACL is set
In the Linux kernel, the following vulnerability has been resolved:
mm/shmem: honour FSEALEXEC when an ACL is set
The owner of a memfd sealed with FSEALEXEC can still change its execute bits by setting a POSIX access ACL. With CONFIGTMPFSPOSIXACL, this lets an unprivileged owner make a non-executable memfd executable, bypassing vm.memfdnoexec=2.
simplesetacl() updates imode directly, bypassing the seal check in shmemsetattr(). Add shmemsetacl() to check the resulting mode. Share the FSEALEXEC check with shmemsetattr() using memfdchecksealsmode() in include/linux/memfd.h.
Fix generated by an LLM and sent by Chris Mason in private.
[ david: use common helper; simplify description ]
Affected Software
Frequently Asked Questions
Who is exposed to this issue?
Systems running the Linux kernel with CONFIG_TMPFS_POSIX_ACL enabled are exposed when an unprivileged user owns a memfd sealed with F_SEAL_EXEC. The issue is relevant where vm.memfd_noexec=2 is relied on to prevent such memfds from becoming executable.
What does an attacker need to exploit it?
The attacker needs to be the owner of a memfd that has F_SEAL_EXEC applied and must be able to set a POSIX access ACL on it. They can then change the effective execute bits through the ACL path.
Are default configurations affected?
The vulnerable behavior requires CONFIG_TMPFS_POSIX_ACL. It also specifically bypasses the protection provided by vm.memfd_noexec=2, so environments not using that setting are not relying on the bypassed control.
How can I determine whether a system is affected?
Check whether the running kernel includes the fix that routes shmem ACL updates through an F_SEAL_EXEC mode check. Also verify whether CONFIG_TMPFS_POSIX_ACL is enabled and whether vm.memfd_noexec is set to 2.