Re: [PATCH 1/1] mm: prevent a race between process_mrelease and exit_mmap

[Michal Hocko]

On Tue 09-11-21 20:26:56, Michal Hocko wrote:
> On Tue 09-11-21 11:01:02, Suren Baghdasaryan wrote:
> [...]
> > Discussing how the patch I want to post works for maple trees that
> > Matthew is working on, I've got a question:
> > 
> > IIUC, according to Michal's post here:
> > https://lore.kernel.org/all/20170725154514.GN26723@xxxxxxxxxxxxxx,
> > unmap_vmas() can race with other mmap_lock read holders (including
> > oom_reap_task_mm()) with no issues.
> > Maple tree patchset requires rcu read lock or the mmap semaphore be
> > held (read or write side) when walking the tree, including inside
> > unmap_vmas(). When asked, he told me that he is not sure why it's
> > currently "safe" to walk the vma->vm_next list in unmap_vmas() while
> > another thread is reaping the mm.
> > Michal (or maybe someone else), could you please clarify why
> > unmap_vmas() can safely race with oom_reap_task_mm()? Or maybe my
> > understanding was wrong?
> 
> I cannot really comment on the mapple tree part. But the existing
> synchronization between oom reaper and exit_mmap is based on
> - oom_reaper takes mmap_sem for reading
> - exit_mmap sets MMF_OOM_SKIP and takes the exclusive mmap_sem before 
>   unmap_vmas.
> 
> The oom_reaper therefore can either unmap the address space if the lock
> is taken before exit_mmap or it would it would bale out on MMF_OOM_SKIP
> if it takes the lock afterwards. So the reaper cannot race with
> unmap_vmas.

Forgot to mention, that _if_ we can get rid of the nasty unlock;lock
pattern in exit_mmap and simply take the exclusive mmap_sem there for
unmap_vmas onward then we could get rid of the MMF_OOM_SKIP as well
because oom_reaper would simply have no vmas to iterate through so the
whole thing would become much more easier to follow.
-- 
Michal Hocko
SUSE Labs