Messages in this thread | | | Date | Sat, 4 Jun 2011 15:19:40 +0100 | From | Al Viro <> | Subject | Re: [PATCH 08/12] superblock: introduce per-sb cache shrinker infrastructure |
| |
On Sat, Jun 04, 2011 at 10:08:48AM -0400, Christoph Hellwig wrote: > On Sat, Jun 04, 2011 at 11:52:12AM +1000, Dave Chinner wrote: > > I wanted to put it early on in the unmount path so that the shrinker > > was guaranteed to be gone before evict_inodes() was called. That > > would mean that it is obviously safe to remove the iprune_sem > > serialisation in that function. > > The iprune_sem removal is fine as soon as you have a per-sb shrinker > for the inodes which keeps an active reference on the superblock until > all the inodes are evicted.
I really don't like that. Stuff keeping active refs, worse yet doing that asynchronously... Shrinkers should *not* do that. Just grab a passive ref (i.e. bump s_count), try grab s_umount (shared) and if that thing still has ->s_root while we hold s_umount, go ahead. Unregister either at the end of generic_shutdown_super() or from deactivate_locked_super(), between the calls of ->kill_sb() and put_filesystem().
| |