Change protoc build rule to not use cygwin bash.

Change protoc build rule to not use cygwin bash.

Now protoc.exe will be directly invoked.

This rolls cacheinvalidation from r218:r219
------------------------------------------------------------------------
r219 | ghcooper@gmail.com | 2012-07-20 17:44:14 -0700 (Fri, 20 Jul 2012) | 1 line

move protos out of v2/; use relative paths for proto imports
------------------------------------------------------------------------

This also rolls gyp from r1440:r1441
------------------------------------------------------------------------
r1441 | iannucci@chromium.org | 2012-07-26 17:54:17 -0700 (Thu, 26 Jul 2012) | 8 lines

Fix bug: non-cygwin actions incorrectly run options through _FixPath.

This adds an additional heuristic to the existing one. Now if an option begins
with either a '/' OR a '-', it assumes it's not a strict path.

Added a TODO which notes that the existing (and new) behavior will miss options
such as '--opt=a/posix/path', and '/opt:a/posix/path'.
Review URL: https://chromiumcodereview.appspot.com/10829049
------------------------------------------------------------------------


BUG=123026, 113339
R=scottmg@chromium.org,cmp@chromium.org,nsylvain@chromium.org,jar@chromium.org,akalin@chromium.org,maruel@chromium.org

TEST=

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=149100

[akalin, 2012-07-24 04:35:40]

lgtm

mention the cacheinvalidation roll in the CL description

[Ryan Sleevi, 2012-07-30 06:54:50]

On 2012/07/27 22:35:09, iannucci wrote:

Please note cacheinvalidation change -> http://crbug.com/113339

[iannucci, 2012-07-30 18:35:08]

On 2012/07/30 06:54:50, Ryan Sleevi wrote:
> On 2012/07/27 22:35:09, iannucci wrote:
> 
> Please note cacheinvalidation change -> http://crbug.com/113339

Noted :)

[iannucci, 2012-07-30 18:35:40]

PTAL, when you have a minute. Subject says it all :)

Thanks,
Robbie

[scottmg, 2012-07-30 19:03:55]

Looks fine to me, and akalin already said so? In general try to only specify 1
reviewer because no one knows who you want to review what, or at least mention
in the mail what subsets you want people to look at.

[iannucci, 2012-07-30 20:20:33]

On 2012/07/30 19:03:55, scottmg wrote:
> Looks fine to me, and akalin already said so? In general try to only specify 1
> reviewer because no one knows who you want to review what, or at least mention
> in the mail what subsets you want people to look at.

Sorry, akalin had reviewed before I actually published the review (patches 2+3
came after the lgtm).

I'm not actually sure who is the most appropriate reviewer(s) of this change.

[scottmg, 2012-07-30 20:25:04]

OK, lgtm

[commit-bot: I haz the power, 2012-07-30 20:32:39]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/10796051/11001

[commit-bot: I haz the power, 2012-07-30 22:36:00]

Try job failure for 10796051-11001 (retry) on win_rel for step "compile"
(clobber build).
It's a second try, previously, step "compile" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2012-07-30 23:26:18]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/10796051/21004

[commit-bot: I haz the power, 2012-07-31 01:07:34]

Change committed as 149100