Re: Success with new futex code from 2.6.0-test5

Bill Soudan wrote:
> The .text.lock.futex hits make me think I can squeeze out a bit more 
> performance by porting Boris' patch to split up the global lock into one 
> per hash bucket.  I'm not sure if there's much that can be done to 
> optimize futex_wake & futex_wait - though the body of wake is mainly 
> searching through the hash table, so I'll try tweaking the size.

You could also try using a lock per N buckets - i.e. just shift the
masked hash a few bits.

Also, we're not sure if the hash function is adequate.  Looking into a
better one.  Glad you're testing the patch; I don't have the resources
to test it :)

-- Jamie