On Fri, Jul 14, 2000 at 09:27:15PM -0700, Hans Reiser wrote:
> Alexander Viro wrote:
>
> > * Probably inode methods for directories are going to be called without
> > BKL. Again, I'll do the work for in-kernel filesystems, but 3rd-party ones
> > will have to be fixed by their maintainers.
>
> I encourage you to do this one, it really affects performance significantly for
> squid. Reiserfs would put everything in one directory for squid and gain 10-24%
> if not for this lock. Whether it should be done pre or post 2.4 I leave to
> others..... except to say that we will be happy to scramble a bit to adapt to
> your fix of this.
I think you're talking about different locks. Squid is harmed by the
inode semaphore on the directory and possibly the super block lock, Al
is talking about the SMP BKL. On UP it should make no difference
at all. If the BKL lock was to blame for squid multiple directories would
not have helped (and I guess it was on UP anyways)
-Andi
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Jul 15 2000 - 21:00:18 EST