Re: -mm seems significanty slower than mainline on kernbench

[Con Kolivas]

On Thu, 12 Jan 2006 01:26 pm, Martin Bligh wrote:
> Peter Williams wrote:
> > Con Kolivas wrote:
> >> On Thu, 12 Jan 2006 12:29 pm, Peter Williams wrote:
> >>> Con Kolivas wrote:
> >>>> This is a shot in the dark. We haven't confirmed 1. there is a
> >>>> problem 2.
> >>>> that this is the problem nor 3. that this patch will fix the problem.
> >>>
> >>> I disagree.  I think that there is a clear mistake in my original patch
> >>> that this patch fixes.
> >>
> >> I agree with you on that. The real concern is that we were just about
> >> to push it upstream. So where does this leave us?  I propose we delay
> >> merging the
> >> "improved smp nice handling" patch into mainline pending your further
> >> changes.
> >
> > I think that they're already in 2.6.15 minus my "move load not tasks"
> > modification which I was expecting to go into 2.6.16.  Is that what you
> > meant?
> >
> > If so I think this is a small and obvious fix that shouldn't delay the
> > merging of "move load not tasks" into the mainline.  But it's not my
> > call.
>
> BTW, in reference to a question from last night ... -git7 seems fine. So
> if you merged it already, it wasn't that ;-)

Thanks and looks like the results are in from 2.6.14-rc2-mm1 with the patch 
backed out.

Drumroll....

http://test.kernel.org/perf/kernbench.moe.png

The performance goes back to a range similar to 2.6.14-rc1-mm1 (see 
2.6.14-rc2-mm1 + 20328 in blue). Unfortunately this does implicate this 
patch. Can we put it back into -mm only and allow Peter's tweaks/fixes to go 
on top and have it tested some more before going upstream?

Peter, Ingo?

Cheers,
Con
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/