sched/accounting: Re-use scheduler statistics for the root cgroup
authorGlauber Costa <glommer@parallels.com>
Mon, 28 Nov 2011 16:45:19 +0000 (14:45 -0200)
committerIngo Molnar <mingo@elte.hu>
Tue, 6 Dec 2011 19:51:21 +0000 (20:51 +0100)
commit54c707e98de9ca899e6552a47c797c62c45885ee
tree61ec9be62b2b3db0201aca7c8eddb5e49239baf5
parentb39e66eaf9c573f38133e894256caeaf9fd2a528
sched/accounting: Re-use scheduler statistics for the root cgroup

Right now, after we collect tick statistics for user and system and store them
in a well known location, we keep the same statistics again for cpuacct.
Since cpuacct is hierarchical, the numbers for the root cgroup should be
absolutely equal to the system-wide numbers.

So it would be better to just use it: this patch changes cpuacct accounting
in a way that the cpustat statistics are kept in a struct kernel_cpustat percpu
array. In the root cgroup case, we just point it to the main array. The rest of
the hierarchy walk can be totally disabled later with a static branch - but I am
not doing it here.

Signed-off-by: Glauber Costa <glommer@parallels.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Tuner <pjt@google.com>
Link: http://lkml.kernel.org/r/1322498719-2255-4-git-send-email-glommer@parallels.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
kernel/sched/core.c
kernel/sched/sched.h