On 8 Sep 2000, Juan J. Quintela wrote:
>
> Hi
>
> I have noticed that in ext2 (I haven't read the other FS code) call
> mark_buffer_dirty() with spin_locks held. I think that it is better
What spinlock would it be?
> to use the construction:
>
> __mark_buffer_dirty()
> <...>
> spinlock_unlock();
> balance_dirty();
>
> I attached the patch for ext2 against test8-pre6. Stephen or Tytso,
> do you mean to comment on that??
> I did the patch because I found lock contention in the superblock lock
> when running dbench 48.
Excuse me? superblock lock is _not_ a spinlock. It's worth
replacing with a spinlock, true, but then we would be better off making it
per-cylinder group...
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]
