Re: [PATCH] sched: smart wake-affine

[Davidlohr Bueso]

On Mon, 2013-07-08 at 10:21 +0200, Peter Zijlstra wrote:
> On Sun, Jul 07, 2013 at 08:43:25AM +0200, Mike Galbraith wrote:
> > On Fri, 2013-07-05 at 14:16 +0800, Michael Wang wrote:
> > 
> > > PeterZ has suggested some optimization which I sent out yesterday, I
> > > suppose they haven't been included into this test yet, correct?
> > 
> > No, that was with both v3 patches applied.  hackbench -l 1000 delta
> > tracked to next pull/build fwiw, but I don't see anything noteworthy
> > elsewhere, so I'm filing it under "tasks/core ~= zillion ~= hohum".
> 
> OK, I'll apply the patches, we'll see what happens. If there significant
> fallout we'll immediately have more information anyway ;-)

So I gave the v2 a spin on my aim7 benchmark on an 80-core 8 socket
DL980. Not much changed, most numbers are in the noise range, however,
with HT off, the high_systime workload suffered in throughput with this
patch with higher concurrency (after 600 users). Image attached.

Thanks,
Davidlohr

Attachment: high_systime-htoff.jpg
Description: JPEG image