Re: [ACPI / hotplug / PCI] 1b360f44d00: -80.9% vm-scalability.stddev

[Fengguang Wu]

On Wed, Mar 05, 2014 at 01:58:54PM +0100, Rafael J. Wysocki wrote:
> On Wednesday, March 05, 2014 04:23:36 PM Fengguang Wu wrote:
> > Hi Rafael,
> 
> Hi,
> 
> > I don't know why but we are glad to find the below changes on commit
> 
> Do I understand correcty that you mean you're seeing improvement?

Yes - the random multi-thread malloc test's deviation is reduced noticeably. 

> Either way, if the test below doesn't involve hotplug or device removal via
> sysfs, I have no explanation ...

OK.

Thanks,
Fengguang

> > 1b360f44d009059e446532f29c1a889951e72667 ("ACPI / hotplug / PCI: Fix bridge removal race in handle_hotplug_event()")
> > 
> > test case: brickland2/micro/vm-scalability/300s-anon-r-rand-mt
> > 
> > d42f5da23400833  1b360f44d009059e446532f29  
> > ---------------  -------------------------  
> >       3.54 ~164%     -80.9%       0.68 ~37% TOTAL vm-scalability.stddev
> >      14177 ~ 7%     -22.1%      11040 ~ 7%  TOTAL numa-meminfo.node0.SReclaimable
> >       3543 ~ 7%     -22.1%       2759 ~ 7%  TOTAL numa-vmstat.node0.nr_slab_reclaimable
> >     776196 ~ 6%     -10.0%     698893 ~ 5%  TOTAL numa-numastat.node2.local_node
> >     776196 ~ 6%     -10.0%     698894 ~ 5%  TOTAL numa-numastat.node2.numa_hit
> > 
> > The ~XX% numbers are stddev percent.
> > The [+-]XX% is change percent.
> > 
> > The below graph shows all samples collected during the bisect
> > 
> > [*] bisect-good
> > [O] bisect-bad ("bad" in the sense of bisect)
> > 
> > In which you can see the stableness of the change and bisect.
> > 
> >                                vm-scalability.stddev
> > 
> >    16 ++--------------------------------------------------------------------+
> >       |                                                                 *   |
> >    14 ++                                                                :   |
> >    12 ++                                                                :   |
> >       |                                                                 ::  |
> >    10 ++                                                                ::  |
> >       |                                                                 ::  |
> >     8 ++                                                                ::  |
> >       |                                                                : :  |
> >     6 ++                                                    *          : :  |
> >     4 ++           *      *                                 ::         :  : |
> >       |           +:     : +                       *       : :         :  : |
> >     2 ++*.   *. .*  :    :  *          .*.        + :      :  :   .*   :  : |
> >       |+  * +  *    *. .* O  *.*.*. .**   O.**.O.* O: *.O.* O OO.*  +  :  *.|
> >     0 O+O-OO-O-O-O-OO-O-O---OO-O-O-O-OO-O---OO-*-O---OO-*-O------O-O-*-*----*
> > 
> > 
> > Thanks,
> > Fengguang
> > --
> > 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/
> 
> -- 
> I speak only for myself.
> Rafael J. Wysocki, Intel Open Source Technology Center.
--
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/