Re: [PATCH 1/1] ipc/shm: serialize orphan cleanup with shm_nattch updates

[Davidlohr Bueso]

On Thu, 30 Apr 2026, Andrew Morton wrote:

> On Thu, 30 Apr 2026 13:21:34 +0800 Ren Wei <n05ec@xxxxxxxxxx> wrote:
>
> > From: Yilin Zhu <zylzyl2333@xxxxxxxxx>
> >
> > shm_destroy_orphaned() walks the shm idr under shm_ids(ns).rwsem, but
> > that does not serialize all fields tested by shm_may_destroy().  In
> > particular, shm_nattch is updated while holding shm_perm.lock, and attach
> > paths can do that without holding the rwsem.
> >
> > Do not decide that an orphaned segment is unused before taking the object
> > lock.  Move the shm_may_destroy() check under shm_perm.lock, matching the
> > other destroy paths, and unlock the segment when it no longer qualifies
> > for removal.
>
> Thanks.
>
> > Fixes: 4c677e2eefdb ("shm: optimize locking and ipc_namespace getting")
>
> Let's cc more people who were involved in 4c677e2eefdb.
>
> And Davidlohr, who might have opinions.

The same feedback as the previous (duplicate) patch - there is no need to
unconditionally take the lock.

https://lore.kernel.org/all/20260406193322.cu2rrtfmpixlv7yy@offworld/

Thanks,
Davidlohr