command_buffer: Support instanced path rendering in gpu command buffer

command_buffer: Support instanced path rendering in gpu command buffer

Implement support for instanced path rendering in gpu command buffer.
Exposes following new functions through command buffer:
glStencilFillPathInstancedCHROMIUM
glStencilStrokePathInstancedCHROMIUM
glCoverFillPathInstancedCHROMIUM
glCoverStrokePathInstancedCHROMIUM
glStencilThenCoverFillPathInstancedCHROMIUM
glStencilThenCoverStrokePathInstancedCHROMIUM

These functions are exposed as part of the CHROMIUM_path_rendering
GL ES extension.

Converts the NV_path_rendering functions that Skia calls to these
functions, passes the calls through the command buffer and then
calls the corresponding NV_path_rendering functions.

BUG=344330
Committed: https://crrev.com/fb3f117d605fcba79a9a6e85dad021a6d77d57cc
Cr-Commit-Position: refs/heads/master@{#356526}

[Kimmo Kinnunen, 2014-08-25 12:39:37]

Depends on:
https://codereview.chromium.org/169603002/ Add initial support for
NV_path_rendering extension to gpu command buffer

https://codereview.chromium.org/169403005/ Support path rendering functions with
CHROMIUM_path_rendering in gpu command buffer

[Kimmo Kinnunen, 2014-09-08 11:19:48]

ptal if you have time..

[vmiura, 2015-06-19 20:26:09]

vmiura@chromium.org changed reviewers:
+ dyen@chromium.org, sievers@chromium.org

[vmiura, 2015-08-28 20:02:32]

https://codereview.chromium.org/477623004/diff/730001/gpu/GLES2/extensions/CH...
File gpu/GLES2/extensions/CHROMIUM/CHROMIUM_path_rendering.txt (right):

https://codereview.chromium.org/477623004/diff/730001/gpu/GLES2/extensions/CH...
gpu/GLES2/extensions/CHROMIUM/CHROMIUM_path_rendering.txt:74:
BOUNDING_BOX_OF_BOUNDING_BOXES_CHROMIUM
nit: BOUNDING_BOX_OF_BOUNDING_BOXES_CHROMIUM  0x909C ?

https://codereview.chromium.org/477623004/diff/730001/gpu/command_buffer/clie...
File gpu/command_buffer/client/gles2_implementation.cc (right):

https://codereview.chromium.org/477623004/diff/730001/gpu/command_buffer/clie...
gpu/command_buffer/client/gles2_implementation.cc:5923: return true;
Why don't we set GL_ERROR here?

[Kimmo Kinnunen, 2015-08-31 07:36:53]

https://codereview.chromium.org/477623004/diff/730001/gpu/GLES2/extensions/CH...
File gpu/GLES2/extensions/CHROMIUM/CHROMIUM_path_rendering.txt (right):

https://codereview.chromium.org/477623004/diff/730001/gpu/GLES2/extensions/CH...
gpu/GLES2/extensions/CHROMIUM/CHROMIUM_path_rendering.txt:74:
BOUNDING_BOX_OF_BOUNDING_BOXES_CHROMIUM
On 2015/08/28 20:02:32, vmiura wrote:
> nit: BOUNDING_BOX_OF_BOUNDING_BOXES_CHROMIUM  0x909C ?

Done.

https://codereview.chromium.org/477623004/diff/730001/gpu/command_buffer/clie...
File gpu/command_buffer/client/gles2_implementation.cc (right):

https://codereview.chromium.org/477623004/diff/730001/gpu/command_buffer/clie...
gpu/command_buffer/client/gles2_implementation.cc:5923: return true;
On 2015/08/28 20:02:32, vmiura wrote:
> Why don't we set GL_ERROR here?

It is not an error to draw 0 paths?
Returning 'true' means that the call is made to the service side.
We do not check any conditions below, because we will anyway check those in
service side.
We want validate all the parameters since the call with 0 paths might be valid
or invalid. We need to call to the service side because some of the validation
is only available at the service side.

[Kimmo Kinnunen, 2015-09-21 12:12:48]

On 2015/08/31 07:36:53, Kimmo Kinnunen wrote:


Ping: Would it be possible to get a review on this? This is the first patch that
would go in, so it's kind of blocking merging the other patches.

[Kimmo Kinnunen, 2015-09-22 12:19:57]

kkinnunen@nvidia.com changed reviewers:
+ jam@chromium.org

[Kimmo Kinnunen, 2015-09-22 12:19:58]

jam@: could you take a look at
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h?
Thanks!
Since 
https://codereview.chromium.org/1349233002 , all extension functions are
available in mojo.

[jam, 2015-09-22 15:52:13]

On 2015/09/22 12:19:58, Kimmo Kinnunen wrote:
> jam@: could you take a look at
>
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h?
> Thanks!
> Since 
> https://codereview.chromium.org/1349233002 , all extension functions are
> available in mojo.

piman would be a better reviewer than me

[piman, 2015-09-22 19:16:44]

piman@chromium.org changed reviewers:
+ zmo@chromium.org

[piman, 2015-09-22 19:16:46]

+zmo to do the review, I can rubber stamp once he's happy.

[Zhenyao Mo, 2015-09-30 02:39:44]

Kimmo, can you do another rebase?  I'll start reviewing this CL.  It would be
easier if it's based on the latest top of tree.

[Kimmo Kinnunen, 2015-10-01 11:17:39]

On 2015/09/30 02:39:44, Zhenyao Mo wrote:
> Kimmo, can you do another rebase?  I'll start reviewing this CL.  It would be
> easier if it's based on the latest top of tree.

Done

[Zhenyao Mo, 2015-10-01 20:43:12]

On 2015/10/01 11:17:39, Kimmo Kinnunen wrote:
> On 2015/09/30 02:39:44, Zhenyao Mo wrote:
> > Kimmo, can you do another rebase?  I'll start reviewing this CL.  It would
be
> > easier if it's based on the latest top of tree.
> 
> Done

Thanks.  I am on it, but it may take me a day or two to finish this large CL.

[Zhenyao Mo, 2015-10-06 22:51:01]

Mostly looks good.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:227: !GLES2Util::IsNPOT(mask +
1);
overflow check for mask + 1

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15106: if (path_name_type ==
GL_BYTE) {
A switch (path_name_type) is much cleaner here.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15175: T* paths = NULL;
const T*

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15210: if (sizeof(T) ==
sizeof(GLuint)) {
I think this optimization is unnecessary and makes the code much uglier.

It's much cleaner if you just get rid of the out_temp_path_buffer arg, and just
pass in out_paths as a GLuint array pointer, and the caller should be
responsible for allocating and deleting that array.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15221: *out_error =
error::kOutOfBounds;
This is incorrect.  You should generate an INVALID_OPERATIONS GL error according
to the spec, not a command buffer error.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15253:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
no GL error, but return error::kUnknownCommand instead

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15259: if (num_paths < 0) {
It looks to me that the validation of num_paths, path_name_type, mask,
transform_type can be folded into PrepareInstancedPathCommand, so each Handle*
command doesn't need to duplicate the code.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15311:
glStencilFillPathInstancedNV(num_paths, path_name_type, paths, 0, fill_mode,
Should you always use GL_UNSIGNED_INT here as the path_name_type? Here and all
other functions.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15325:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
Same here, no GL error, but return error::kUnknownCommand.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15350: GLuint ref =
static_cast<GLuint>(c.reference);
In the spec, reference is defined as GLint, not GLuint.  Please be consistent.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15385:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
Same here, no GL error, but return error::kUnknownCommand.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15450:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
Same here, no GL error, but return error::kUnknownCommand.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15517:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
Same here, no GL error, but return error::kUnknownCommand.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15599:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
Same here, no GL error, but return error::kUnknownCommand.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15631: GLuint reference =
static_cast<GLuint>(c.reference);
GLint instead of GLuint

[Kimmo Kinnunen, 2015-10-08 09:28:29]

Thanks for the thorough inspection.
A new version, ptal.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:227: !GLES2Util::IsNPOT(mask +
1);
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> overflow check for mask + 1

While not terribly useful, I think overflow is intended. This is in context of
bit twiddling, after all. I left this as is, but I can change it if needed.
(Also note, it's not new code, it's existing code)

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15106: if (path_name_type ==
GL_BYTE) {
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> A switch (path_name_type) is much cleaner here.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15175: T* paths = NULL;
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> const T*

Still left this as is, see below. (The optimization discussion)

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15210: if (sizeof(T) ==
sizeof(GLuint)) {
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> I think this optimization is unnecessary and makes the code much uglier.
> 
> It's much cleaner if you just get rid of the out_temp_path_buffer arg, and
just
> pass in out_paths as a GLuint array pointer, and the caller should be
> responsible for allocating and deleting that array.

I tried to make this less ugly. If it is still ugly, I can of course delete it.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15221: *out_error =
error::kOutOfBounds;
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> This is incorrect.  You should generate an INVALID_OPERATIONS GL error
according
> to the spec, not a command buffer error.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15253:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> no GL error, but return error::kUnknownCommand instead

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15259: if (num_paths < 0) {
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> It looks to me that the validation of num_paths, path_name_type, mask,
> transform_type can be folded into PrepareInstancedPathCommand, so each Handle*
> command doesn't need to duplicate the code.

The order of the error checking is significant.
I tried to match:
the spec error order with client order
the client order with service order
(ergo spec order with service order, too).

I tried to reduce the duplication with common code. I hope this would be ok.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15311:
glStencilFillPathInstancedNV(num_paths, path_name_type, paths, 0, fill_mode,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> Should you always use GL_UNSIGNED_INT here as the path_name_type? Here and all
> other functions.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15325:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> Same here, no GL error, but return error::kUnknownCommand.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15350: GLuint ref =
static_cast<GLuint>(c.reference);
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> In the spec, reference is defined as GLint, not GLuint.  Please be consistent.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15385:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> Same here, no GL error, but return error::kUnknownCommand.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15450:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> Same here, no GL error, but return error::kUnknownCommand.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15517:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:01, Zhenyao Mo wrote:
> Same here, no GL error, but return error::kUnknownCommand.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15599:
LOCAL_SET_GL_ERROR(GL_INVALID_OPERATION, kFunctionName,
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> Same here, no GL error, but return error::kUnknownCommand.

Done.

https://codereview.chromium.org/477623004/diff/810001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:15631: GLuint reference =
static_cast<GLuint>(c.reference);
On 2015/10/06 22:51:00, Zhenyao Mo wrote:
> GLint instead of GLuint

Done.

[Zhenyao Mo, 2015-10-08 23:10:59]

Mostly look good.
Please fix these minor comments. Also a question about |paths| with negative
values.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14598: GLuint*
AllocateOrAdopt(GLuint num_paths, T*) {
Please add a comment this default code path covers GLbyte, GLubyte, GLshort,
GLushort.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14611:
path_names_alloc_.reset();
You should DCHECK it's null instead of reset.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14617:
path_names_alloc_.reset();
You should DCHECK it's null instead of reset.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14626: // The Get* functions
will return true if the function call should be stopped.
I feel it's much intuitive if you return true when the function should proceed,
and return false when they should be stopped.  It's like Get* encounters an
error, so return false.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14825: if
(!SafeAddUint32(paths[i], path_base, &client_id)) {
One question: if the paths[i] from client is negative, is that legal or not?

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14828: error_ = error::kNoError;
no need.

[Kimmo Kinnunen, 2015-10-09 11:43:48]

piman@:
Could you review: 
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14598: GLuint*
AllocateOrAdopt(GLuint num_paths, T*) {
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> Please add a comment this default code path covers GLbyte, GLubyte, GLshort,
> GLushort.

Done.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14611:
path_names_alloc_.reset();
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> You should DCHECK it's null instead of reset.

Done.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14617:
path_names_alloc_.reset();
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> You should DCHECK it's null instead of reset.

Done.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14626: // The Get* functions
will return true if the function call should be stopped.
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> I feel it's much intuitive if you return true when the function should
proceed,
> and return false when they should be stopped.  It's like Get* encounters an
> error, so return false.

Done.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14825: if
(!SafeAddUint32(paths[i], path_base, &client_id)) {
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> One question: if the paths[i] from client is negative, is that legal or not?

Good point. Yeah, that is legal.
I think I will remove the overflow check from here and the related spec
language. The over- and underflow can be seen as intended features.

https://codereview.chromium.org/477623004/diff/830001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:14828: error_ = error::kNoError;
On 2015/10/08 23:10:59, Zhenyao Mo wrote:
> no need.

Done.

[piman, 2015-10-09 17:10:30]

RS LGTM for _autogen file. Am I even an OWNER there?

[Kimmo Kinnunen, 2015-10-22 11:10:52]

zmo@: Ping. PTAL when you have time.. The other patches depend on this first
patch.

[Zhenyao Mo, 2015-10-26 22:19:34]

LGTM if bots are green

[Kimmo Kinnunen, 2015-10-27 11:31:04]

On 2015/09/22 15:52:13, jam wrote:
> On 2015/09/22 12:19:58, Kimmo Kinnunen wrote:
> > jam@: could you take a look at
> >
>
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h?
> > Thanks!
> > Since 
> > https://codereview.chromium.org/1349233002 , all extension functions are
> > available in mojo.
> 
> piman would be a better reviewer than me

jam@: piman's review is not enough for the file. Would you have time to look at
the file, or maybe suggest somebody else?

[Zhenyao Mo, 2015-10-27 21:41:27]

zmo@chromium.org changed reviewers:
+ jamesr@chromium.org, yzshen@chromium.org

[Zhenyao Mo, 2015-10-27 21:42:08]

On 2015/10/27 11:31:04, Kimmo Kinnunen wrote:
> On 2015/09/22 15:52:13, jam wrote:
> > On 2015/09/22 12:19:58, Kimmo Kinnunen wrote:
> > > jam@: could you take a look at
> > >
> >
>
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h?
> > > Thanks!
> > > Since 
> > > https://codereview.chromium.org/1349233002 , all extension functions are
> > > available in mojo.
> > 
> > piman would be a better reviewer than me
> 
> jam@: piman's review is not enough for the file. Would you have time to look
at
> the file, or maybe suggest somebody else?

jam@, jamesr@, yzshen@, can one of you stamp the auto-generated file under
third_party/mojo?

[yzshen1, 2015-10-28 07:12:38]

On 2015/10/27 21:42:08, Zhenyao Mo wrote:
> On 2015/10/27 11:31:04, Kimmo Kinnunen wrote:
> > On 2015/09/22 15:52:13, jam wrote:
> > > On 2015/09/22 12:19:58, Kimmo Kinnunen wrote:
> > > > jam@: could you take a look at
> > > >
> > >
> >
>
third_party/mojo/src/mojo/public/c/gles2/gles2_call_visitor_chromium_extension_autogen.h?
> > > > Thanks!
> > > > Since 
> > > > https://codereview.chromium.org/1349233002 , all extension functions are
> > > > available in mojo.
> > > 
> > > piman would be a better reviewer than me
> > 
> > jam@: piman's review is not enough for the file. Would you have time to look
> at
> > the file, or maybe suggest somebody else?
> 
> jam@, jamesr@, yzshen@, can one of you stamp the auto-generated file under
> third_party/mojo?

rubber stamp LGTM for third_party/mojo

[Kimmo Kinnunen, 2015-10-28 07:36:33]

The CQ bit was checked by kkinnunen@nvidia.com

[Kimmo Kinnunen, 2015-10-28 07:36:34]

The patchset sent to the CQ was uploaded after l-g-t-m from piman@chromium.org,
zmo@chromium.org
Link to the patchset: https://codereview.chromium.org/477623004/#ps890001
(title: "fix msvc signedness warning in an unittest")

[commit-bot: I haz the power, 2015-10-28 07:36:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/477623004/890001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/477623004/890001

[commit-bot: I haz the power, 2015-10-28 08:50:09]

Committed patchset #31 (id:890001)

[commit-bot: I haz the power, 2015-10-28 08:51:14]

Patchset 31 (id:??) landed as
https://crrev.com/fb3f117d605fcba79a9a6e85dad021a6d77d57cc
Cr-Commit-Position: refs/heads/master@{#356526}