On 4/7/22 23:12, Andrew Morton wrote: > On Thu, 7 Apr 2022 23:01:37 -0400 Nico Pache <npache@xxxxxxxxxx> wrote: > >> The pthread struct is allocated on PRIVATE|ANONYMOUS memory [1] which can >> be targeted by the oom reaper. This mapping is used to store the futex >> robust list head; the kernel does not keep a copy of the robust list and >> instead references a userspace address to maintain the robustness during >> a process death. A race can occur between exit_mm and the oom reaper that >> allows the oom reaper to free the memory of the futex robust list before >> the exit path has handled the futex death: >> >> CPU1 CPU2 >> ------------------------------------------------------------------------ >> page_fault >> do_exit "signal" >> wake_oom_reaper >> oom_reaper >> oom_reap_task_mm (invalidates mm) >> exit_mm >> exit_mm_release >> futex_exit_release >> futex_cleanup >> exit_robust_list >> get_user (EFAULT- can't access memory) >> >> If the get_user EFAULT's, the kernel will be unable to recover the >> waiters on the robust_list, leaving userspace mutexes hung indefinitely. >> >> Use the robust_list address stored in the kernel to skip the VMA that holds >> it, allowing a successful futex_cleanup. >> >> Theoretically a failure can still occur if there are locks mapped as >> PRIVATE|ANON; however, the robust futexes are a best-effort approach. >> This patch only strengthens that best-effort. >> >> The following case can still fail: >> robust head (skipped) -> private lock (reaped) -> shared lock (skipped) >> >> Reproducer: https://gitlab.com/jsavitz/oom_futex_reproducer >> >> [1] https://elixir.bootlin.com/glibc/latest/source/nptl/allocatestack.c#L370 >> >> ... >> >> --- a/include/linux/oom.h >> +++ b/include/linux/oom.h >> @@ -106,7 +106,7 @@ static inline vm_fault_t check_stable_address_space(struct mm_struct *mm) >> return 0; >> } >> >> -bool __oom_reap_task_mm(struct mm_struct *mm); >> +bool __oom_reap_task_mm(struct mm_struct *mm, void *robust_list); >> >> long oom_badness(struct task_struct *p, >> unsigned long totalpages); >> diff --git a/mm/mmap.c b/mm/mmap.c >> index 3aa839f81e63..d5af1b83cbb2 100644 >> --- a/mm/mmap.c >> +++ b/mm/mmap.c >> @@ -3109,6 +3109,11 @@ void exit_mmap(struct mm_struct *mm) >> struct mmu_gather tlb; >> struct vm_area_struct *vma; >> unsigned long nr_accounted = 0; >> + void *robust_list; >> + >> +#ifdef CONFIG_FUTEX >> + robust_list = current->robust_list; >> +#endif >> >> /* mm's last user has gone, and its about to be pulled down */ >> mmu_notifier_release(mm); >> @@ -3126,7 +3131,8 @@ void exit_mmap(struct mm_struct *mm) >> * to mmu_notifier_release(mm) ensures mmu notifier callbacks in >> * __oom_reap_task_mm() will not block. >> */ >> - (void)__oom_reap_task_mm(mm); >> + (void)__oom_reap_task_mm(mm, robust_list); > > uninitialized var warning when CONFIG_FUTEX=n? > >> + >> set_bit(MMF_OOM_SKIP, &mm->flags); >> } >> >> --- a/mm/oom_kill.c >> +++ b/mm/oom_kill.c >> @@ -509,9 +509,10 @@ static DECLARE_WAIT_QUEUE_HEAD(oom_reaper_wait); >> static struct task_struct *oom_reaper_list; >> static DEFINE_SPINLOCK(oom_reaper_lock); >> >> -bool __oom_reap_task_mm(struct mm_struct *mm) >> +bool __oom_reap_task_mm(struct mm_struct *mm, void *robust_list) > > Well, this is no longer necessarily a robust_list*. It's just an > address to skip and the name should reflect that? ok fair point. I left it as robust list to be more clean what is being skipped and renamed the head variable to skip_vma so its more verbose. But I dont have that strong of an opinion on it so I'll change it real quick. > >> { >> struct vm_area_struct *vma; >> + unsigned long skip_vma = (unsigned long) robust_list; >> bool ret = true; >> >> /* >> @@ -526,6 +527,20 @@ bool __oom_reap_task_mm(struct mm_struct *mm) >> if (vma->vm_flags & (VM_HUGETLB|VM_PFNMAP)) >> continue; >> >> +#ifdef CONFIG_FUTEX >> + /* >> + * The OOM reaper runs concurrently with do_exit. >> + * The robust_list_head is stored in userspace and is required >> + * by the exit path to recover the robust futex waiters. >> + * Skip the VMA that contains the robust_list to allow for >> + * proper cleanup. >> + */ >> + if (vma->vm_start <= skip_vma && vma->vm_end > skip_vma) { >> + pr_info("oom_reaper: skipping vma, contains robust_list"); >> + continue; >> + } >> +#endif >> + >> /* >> * Only anonymous pages have a good chance to be dropped >> * without additional steps which we cannot afford as we >> @@ -567,6 +582,7 @@ bool __oom_reap_task_mm(struct mm_struct *mm) >> static bool oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) >> { >> bool ret = true; >> + void *robust_list; >> >> if (!mmap_read_trylock(mm)) { >> trace_skip_task_reaping(tsk->pid); >> @@ -586,8 +602,11 @@ static bool oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) >> >> trace_start_task_reaping(tsk->pid); >> >> +#ifdef CONFIG_FUTEX >> + robust_list = tsk->robust_list; >> +#endif >> /* failed to reap part of the address space. Try again later */ >> - ret = __oom_reap_task_mm(mm); >> + ret = __oom_reap_task_mm(mm, robust_list); > > unintialized var when CONFIG_FUTEX=n? > >> if (!ret) >> goto out_finish; >> >> @@ -1149,6 +1168,7 @@ SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) >> unsigned int f_flags; >> bool reap = false; >> long ret = 0; >> + void *robust_list; >> >> if (flags) >> return -EINVAL; >> @@ -1186,11 +1206,16 @@ SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) >> ret = -EINTR; >> goto drop_mm; >> } >> + >> +#ifdef CONFIG_FUTEX >> + robust_list = p->robust_list; >> +#endif >> /* >> * Check MMF_OOM_SKIP again under mmap_read_lock protection to ensure >> * possible change in exit_mmap is seen >> */ >> - if (!test_bit(MMF_OOM_SKIP, &mm->flags) && !__oom_reap_task_mm(mm)) >> + if (!test_bit(MMF_OOM_SKIP, &mm->flags) && >> + !__oom_reap_task_mm(mm, robust_list)) > > again Good catch. Ill fix those real quick. > >> ret = -EAGAIN; >> mmap_read_unlock(mm); >
