On Fri, Feb 09, 2024 at 04:36:46PM +0000, Al Viro wrote: > On Fri, Feb 09, 2024 at 03:22:15PM +0100, Christian Brauner wrote: > > On Fri, Feb 09, 2024 at 03:52:19PM +0300, Dmitry Antipov wrote: > > > In 'fasync_remove_entry()', prefer 'kfree_rcu()' over 'call_rcu()' with dummy > > > 'fasync_free_rcu()' callback. This is mostly intended in attempt to fix weird > > > https://syzkaller.appspot.com/bug?id=6a64ad907e361e49e92d1c4c114128a1bda2ed7f, > > > where kmemleak may consider 'fa' as unreferenced during RCU grace period. See > > > https://lore.kernel.org/stable/20230930174657.800551-1-joel@xxxxxxxxxxxxxxxxx > > > as well. Comments are highly appreciated. > > > > > > Signed-off-by: Dmitry Antipov <dmantipov@xxxxxxxxx> > > > --- > > > > Yeah, according to commit ae65a5211d90 ("mm/slab: document kfree() as > > allowed for kmem_cache_alloc() objects") this is now guaranteed to work > > for kmem_cache_alloc() objects since slab is gone. So independent of > > syzbot this seems like a decent enough cleanup. > > Sure, but we'd better make very sure that it does *NOT* get picked by any > -stable prior to 6.4. Yeah, sure. I've not marked it for stable exactly because of that. _But_ we can't exactly control AUTOSEL. Maybe there's a way. In any case, I'll keep an eye out for it and will reply appropriately.
