Messages in this thread | | | Date | Tue, 19 Oct 2010 14:54:15 +1100 | From | Nick Piggin <> | Subject | Re: [PATCH 17/19] fs: Reduce inode I_FREEING and factor inode disposal |
| |
On Mon, Oct 18, 2010 at 02:27:02PM -0700, Sage Weil wrote: > I took a look at this code and was able to remove the i_lock -> > dcache_lock nesting. To be honest, I'm not sure why I did it that way in > the first place. The only point (now) where i_lock is used/needed is > while checking the ceph inode flags, and that's done without holding > dcache_lock (presumably soon to be d_lock).
Great, it was a bit nasty of ceph to add that in without documenting it, but I guess it got past reviewers so it was up to them to ack it or point it out.
I think we should probably enforce a rule that dcache_lock not nest inside i_lock :) Anybody wanting to do that in future had better have a really good reason.
> The patch is 622386be in the for-next branch of > git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client.git > > Anyway, hope this helps!
It does, thanks.
| |