Histogram first paint PLT's for Cache Sensitivity Analysis.

Histogram first paint PLT's for Cache Sensitivity Analysis.
R=rvargas@chromium.org, jhawkins@chromium.org
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=171091

[rvargas (doing something else), 2012-12-04 22:55:48]

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
File chrome/renderer/page_load_histograms.cc (right):

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
chrome/renderer/page_load_histograms.cc:248: scoped_ptr<TimeDelta>
begin_to_first_paint;
How about TimeDelta foo = TimeDelta::FromInternalValue(-1) (really hacky!) or
keeping a boolean to signal initialization (cleaner).

[tburkard, 2012-12-04 23:09:08]

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
File chrome/renderer/page_load_histograms.cc (right):

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
chrome/renderer/page_load_histograms.cc:248: scoped_ptr<TimeDelta>
begin_to_first_paint;
I thought about both of these, and I find scoped_ptr's more elegant. Unless you
feel really strongly about one of these alternatives.

[James Hawkins, 2012-12-04 23:11:56]

lgtm

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
File chrome/renderer/page_load_histograms.cc (right):

https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
chrome/renderer/page_load_histograms.cc:248: scoped_ptr<TimeDelta>
begin_to_first_paint;
On 2012/12/04 23:09:08, tburkard wrote:
> I thought about both of these, and I find scoped_ptr's more elegant. Unless
you
> feel really strongly about one of these alternatives.

I'm fine w/ scoped_ptr.  NULLness (or non-) indicates readiness of the contained
data.

[rvargas (doing something else), 2012-12-05 00:27:37]

On 2012/12/04 23:09:08, tburkard wrote:
>
https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
> File chrome/renderer/page_load_histograms.cc (right):
> 
>
https://codereview.chromium.org/11419319/diff/1/chrome/renderer/page_load_his...
> chrome/renderer/page_load_histograms.cc:248: scoped_ptr<TimeDelta>
> begin_to_first_paint;
> I thought about both of these, and I find scoped_ptr's more elegant. Unless
you
> feel really strongly about one of these alternatives.

No strong feelings... just dislike for non-needed allocations.

lgtm