{"id":"CVE-2026-90305","published":"2026-09-17T17:17:28.163","lastModified":"2026-09-17T17:17:28.163","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nARM: 9483/1: select HAVE_POSIX_CPU_TIMERS_TASK_WORK\n\nCommit c6e61c06d606 (\"ARM: 9463/1: Allow to enable RT\") enabled PREEMPT_RT\non ARM but did not select HAVE_POSIX_CPU_TIMERS_TASK_WORK. This leaves\nCONFIG_POSIX_CPU_TIMERS_TASK_WORK disabled, so CPU timers expire in hard\nIRQ context.\n\nOn PREEMPT_RT this makes run_posix_cpu_timers() take the sleeping\nsighand->siglock:\n\n  BUG: sleeping function called from invalid context at spinlock_rt.c:48\n    rt_spin_lock from lock_task_sighand\n    lock_task_sighand from run_posix_cpu_timers\n    run_posix_cpu_timers from update_process_times\n\nARM handles TIF_NOTIFY_RESUME on all return-to-user paths, including v7-M.\nARM32 KVM host support was removed by commit 541ad0150ca4 (\"arm: Remove\n32bit KVM host support\"), so the select need not be conditional on KVM.\n\nSelect it to defer POSIX CPU timer expiry to task context.\n\nReproduced with setrlimit(RLIMIT_CPU, ...) and a busy loop. The same path\nis used by setitimer(ITIMER_PROF or ITIMER_VIRTUAL) and POSIX CPU timers\ncreated with timer_create().","cvssScore":null,"cvssSeverity":null,"cvssVector":null,"cwes":[],"vendors":[],"products":[],"references":[{"url":"https://git.kernel.org/stable/c/8a58a41100ea377e978d99600ec24a9bd0273662","tags":[]},{"url":"https://git.kernel.org/stable/c/8de56782d6e5ba7a9f8c820342dccde65502f93f","tags":[]}],"exploitRefs":[],"hasPoc":false,"ai":null}