Reland https://codereview.chromium.org/286903025

Reland https://codereview.chromium.org/286903025

BUG=skia:

Committed: http://code.google.com/p/skia/source/detail?r=14966

[kelvinly, 2014-05-28 00:09:12]

All the trybots succeeded, can I resubmit? (Also, sorry for grabbing a couple of
the chromium bots, I somehow forgot we had our own trybots).

[epoger, 2014-05-28 01:33:28]

LGTM with some suggestions for next time...

On 2014/05/28 00:09:12, kelvinly wrote:
> All the trybots succeeded, can I resubmit? (Also, sorry for grabbing a couple
of
> the chromium bots, I somehow forgot we had our own trybots).

I have just launched a trybot on Build-Win7-VS2010-x86-Release-ANGLE-Trybot,
because that corresponds to the builder that we noted failed in
https://code.google.com/p/skia/issues/detail?id=2609#c1 .  If that builder
succeeds, I think it's fine to commit (in the morning).

Next time, I think it's a good idea to stage re-land CLs in a way similar to how
I did in https://codereview.chromium.org/297703006/ ('reland 'upload SKP
renderings that did not match expectations''):

- Create initial patchset with the original CL, without modification
- Send that patchset to some trybots, showing that they fail
- Create patchset 2 with fixes
- Send that patchset to the same trybots, showing that they now succeed

Benefits of this approach include:
- Demonstration that the bots which failed with the original CL should now
succeed
- Allows reviewer to look at just the changes between patchsets 1 and 2 to see
exactly what the fix was

[epoger, 2014-05-28 02:03:15]

I retract my approval... as you can see,
Build-Win7-VS2010-x86-Release-ANGLE-Trybot fails.

[21:33:52.898000] FAILED: ninja -t msvc -e environment.x86 -- "C:\Program Files
(x86)\Microsoft Visual Studio 10.0\VC\bin\cl.exe" /nologo /showIncludes /FC
@obj\bench\bbh_shootout.TimerData.obj.rsp /c ..\..\bench\TimerData.cpp
/Foobj\bench\bbh_shootout.TimerData.obj /Fdbbh_shootout.pdb 
[21:33:52.898000] c:\program files (x86)\microsoft visual studio
10.0\vc\include\xlocale(323) : error C2220: warning treated as error - no
'object' file generated
[21:33:52.898000] c:\program files (x86)\microsoft visual studio
10.0\vc\include\xlocale(323) : warning C4530: C++ exception handler used, but
unwind semantics are not enabled. Specify /EHsc

[kelvinly, 2014-05-29 15:15:50]

Looks like it's working! Approval to commit?

[benchen, 2014-05-29 16:28:39]

Hi Kelvin, so on Patch 11 the trybots succeeded? If so this lgtm. Elliot, what
do you think?

[epoger, 2014-05-29 16:52:44]

lgtm

[kelvinly, 2014-05-29 16:56:17]

The CQ bit was checked by kelvinly@google.com

[commit-bot: I haz the power, 2014-05-29 16:57:16]

CQ is trying da patch. Follow status at
 https://skia-tree-status.appspot.com/cq/kelvinly@google.com/304613002/190001

[commit-bot: I haz the power, 2014-05-29 17:10:29]

Change committed as 14966