power_Resume: report time to bring all cpus online

power_Resume: report time to bring all cpus online

look for the last instance of "CPUn is up" and subtract the resume begin from there

BUG=13649
TEST=run power_Resume and check for the presense of the resume_kernel_cpu metric

Change-Id: I3e03e4a3faea7772f5557dcbe21a5d9d0ca836f6

R=bleung@chromium.org,snanda@chromium.org

Committed: http://chrome-svn/viewvc/chromeos?view=rev&revision=30dd1b9

[Sameer Nanda, 2011-03-30 18:12:35]

LGTM.  DId you want to split out the device resume time as well while you are at
it?

[Sonny, 2011-03-30 18:34:23]

On 2011/03/30 18:12:35, Sameer Nanda wrote:
> LGTM.  DId you want to split out the device resume time as well while you are
at
> it?

Yeah, I was thinking of adding something which parses out the 
PM: resume of devices complete after xxx msecs
message.  Sound okay?

[Benson Leung, 2011-03-30 18:47:10]

sounds good to me.

On Wed, Mar 30, 2011 at 11:34 AM,  <sonnyrao@chromium.org> wrote:
> On 2011/03/30 18:12:35, Sameer Nanda wrote:
>>
>> LGTM.  DId you want to split out the device resume time as well while you
>> are
>
> at
>>
>> it?
>
> Yeah, I was thinking of adding something which parses out the
> PM: resume of devices complete after xxx msecs
> message.  Sound okay?
>
> http://codereview.chromium.org/6756015/
>



-- 
Benson Leung
Software Engineer, Chrom* OS
bleung@chromium.org

[Sameer Nanda, 2011-03-30 18:48:12]

On Wed, Mar 30, 2011 at 11:34 AM, <sonnyrao@chromium.org> wrote:

> On 2011/03/30 18:12:35, Sameer Nanda wrote:
>
>> LGTM.  DId you want to split out the device resume time as well while you
>> are
>>
> at
>
>> it?
>>
>
> Yeah, I was thinking of adding something which parses out the
> PM: resume of devices complete after xxx msecs
> message.  Sound okay?


Sounds good.

>
>
> http://codereview.chromium.org/6756015/
>



-- 
Sameer

[Sameer Nanda, 2011-04-04 16:30:49]

LGTM with a nit.

http://codereview.chromium.org/6756015/diff/4001/client/site_tests/power_Resu...
File client/site_tests/power_Resume/power_Resume.py (right):

http://codereview.chromium.org/6756015/diff/4001/client/site_tests/power_Resu...
client/site_tests/power_Resume/power_Resume.py:154:
results['seconds_system_resume_kernel_devices'] = kernel_device_resume_time
 > 80 columns

[Sonny, 2011-04-04 20:37:45]

On 2011/04/04 16:30:49, Sameer Nanda wrote:
> LGTM with a nit.
> 
>
http://codereview.chromium.org/6756015/diff/4001/client/site_tests/power_Resu...
> File client/site_tests/power_Resume/power_Resume.py (right):
> 
>
http://codereview.chromium.org/6756015/diff/4001/client/site_tests/power_Resu...
> client/site_tests/power_Resume/power_Resume.py:154:
> results['seconds_system_resume_kernel_devices'] = kernel_device_resume_time
>  > 80 columns

Yeah the < 80 columns form looks uglier IMHO, but I'll resubmit

[Sameer Nanda, 2011-04-04 20:54:56]

LGTM

[zbehan, 2011-04-15 21:00:07]

I believe this CL breaks the test (and BVT's) on single CPU machines (like
the L13). The message CPUn is up seems to be only printed for non-boot
CPU's.

On Mon, Apr 4, 2011 at 10:54 PM, <snanda@chromium.org> wrote:

> LGTM
>
>
> http://codereview.chromium.org/6756015/
>

[Sonny, 2011-04-15 21:56:10]

Argh, yeah I should have realized that would break on a 1 (logical)
cpu machine.
I can fix that up.  Have you filed a bug on it yet?

Sonny

On Fri, Apr 15, 2011 at 2:00 PM, Zdenek Behan <zbehan@chromium.org> wrote:
> I believe this CL breaks the test (and BVT's) on single CPU machines (like
> the L13). The message CPUn is up seems to be only printed for non-boot
> CPU's.
>
> On Mon, Apr 4, 2011 at 10:54 PM, <snanda@chromium.org> wrote:
>>
>> LGTM
>>
>> http://codereview.chromium.org/6756015/
>
>

[zbehan, 2011-04-15 21:58:55]

Not filed a bug yet, can do.

On Fri, Apr 15, 2011 at 11:55 PM, Sonny Rao <sonnyrao@chromium.org> wrote:

> Argh, yeah I should have realized that would break on a 1 (logical)
> cpu machine.
> I can fix that up.  Have you filed a bug on it yet?
>
> Sonny
>
> On Fri, Apr 15, 2011 at 2:00 PM, Zdenek Behan <zbehan@chromium.org> wrote:
> > I believe this CL breaks the test (and BVT's) on single CPU machines
> (like
> > the L13). The message CPUn is up seems to be only printed for non-boot
> > CPU's.
> >
> > On Mon, Apr 4, 2011 at 10:54 PM, <snanda@chromium.org> wrote:
> >>
> >> LGTM
> >>
> >> http://codereview.chromium.org/6756015/
> >
> >
>