On Thu, 28 Jul 2011, David Rientjes wrote:
On Thu, 28 Jul 2011, Pekka Enberg wrote:
Christoph, your debugging fix has been in linux-next for few days now
and no problem have been reported. I'm considering sending the series
to Linus. What do you think?
I ran slub/lockless through some stress testing and it seems to be quite
stable on my testing cluster. There is about a 2.3% performance
improvement with the lockless slowpath on the netperf benchmark with
various thread counts on my 16-core 64GB Opterons, so I'd recommend it to
be merged into 3.1.
Great. Could you also test the next stage of patches (not yet even in
Pekka's tree) where we add a per cpu cache of partial allocated slab
pages? This decreases the per node lock contention further. I can repost
the set if the old one does not work for you. Shows significant
improvement here as well.