>>>>> "al" == Alexander Viro <[EMAIL PROTECTED]> writes:
Hi al
>> 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
al> What spinlock would it be?
/me begin by eating his own words
s/spin_lock/super block lock/
I mean that I have seen contention in the calls to __wait_on_super()
(i.e. a lot of processes sleeping in __wait_on_supper()).
This was after applying the patch to reduce super block contention by
Stephen (you can get a version that patch against test8-pre6 from the
URL:
http://carpanta.dc.fi.udc.es/~quintela/kernel/2.4.0-test8-pre6/ext2_lock_super_05.patch
>> to use the construction:
>>
>> __mark_buffer_dirty()
>> <...>
>> spinlock_unlock();
>> balance_dirty();
s/spinlock_unlock/unlock_super/
al> Excuse me? superblock lock is _not_ a spinlock. It's worth
al> replacing with a spinlock, true, but then we would be better off making it
al> per-cylinder group...
1- once explained what I intended to tell, what do you think?
2- could you elaborate in the spinlock by cylinder group? I could try
to change that.
Thanks a lot for your comments and your patience to cope with me :)
Later, Juan.
--
In theory, practice and theory are the same, but in practice they
are different -- Larry McVoy
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]
