lkml.org 
[lkml]   [2013]   [Jan]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [GIT PULL] cputime: Full dynticks task/cputime accounting v7
From
Date
On Mon, 2013-01-28 at 19:02 -0500, Steven Rostedt wrote:

> I've ran these through some basic tests and I don't see any issues. But
> why did you drop the 'comment' patch that you had in v6?
>
> Acked-by: Steven Rostedt <rostedt@goodmis.org>
>

I'm sorry but I need to un-Ack. Due to a test machine setup error, I've
been testing the same config over and over and not other configs.

Everything works great up until I enabled the dynamic tick and force
context tracking.

User time seems to be pretty much the same, but the kernel time is
screwed:

For all my other tests (I finally tested other configs and without your
patches) I had my kernel time test report something like this:

[root@bxtest ~]# time /work/c/kernelspin 10

real 0m10.001s
user 0m1.114s
sys 0m8.886s


But after enabling the force context_tracking I get this:

[root@bxtest ~]# time /work/c/kernelspin 10

real 0m10.001s
user 0m3.769s
sys 0m6.203s


Consistently. All the other runs always had a 1.X 8.X ratio. This seems
to give 3.7x 6.2x ratio consistently. Maybe it's more accurate then the
others. Or something else is broken.

I'll look deeper into it.

-- Steve




\
 
 \ /
  Last update: 2013-01-29 04:01    [W:0.227 / U:0.144 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site