Improve GL helpers.

Improve GL helpers.

Use ApplicationConnector instead of Shell to obtain an offscreen
GL context.  This makes it easier to create GL contexts from
separate threads since the Shell interface is usually owned by
the main thread.

Eagerly unbind GL textures from the context after creation to prevent
spurious circularity warnings from the GPU Service caused by textures
being bound while not actually in use.

BUG=

Committed: https://chromium.googlesource.com/external/mojo/+/0d386d89f313c941182389ddb3c773dd937fa240

[jeffbrown, 2015-12-17 20:46:16]

jeffbrown@google.com changed reviewers:
+ abarth@google.com, jamesr@chromium.org

[jeffbrown, 2015-12-17 20:46:17]

TBR

[jeffbrown, 2015-12-17 20:50:07]

Description was changed from

==========
Improve GL helpers.

Use ApplicationConnector instead of Shell to obtain an offscreen
GL context.  This makes it easier to create GL contexts from
separate threads since the Shell interface is usually owned by
the main thread.

Eagerly unbind GL textures from the context after creation to prevent
spurious circularity warnings from the GPU Service caused by textures
being bound while not actually in use.

BUG=
==========

to

==========
Improve GL helpers.

Use ApplicationConnector instead of Shell to obtain an offscreen
GL context.  This makes it easier to create GL contexts from
separate threads since the Shell interface is usually owned by
the main thread.

Eagerly unbind GL textures from the context after creation to prevent
spurious circularity warnings from the GPU Service caused by textures
being bound while not actually in use.

BUG=

Committed:
https://chromium.googlesource.com/external/mojo/+/0d386d89f313c941182389ddb3c...
==========

[jeffbrown, 2015-12-17 20:50:08]

Committed patchset #1 (id:1) manually as
0d386d89f313c941182389ddb3c773dd937fa240.

[vtl, 2015-12-17 22:30:35]

vtl@google.com changed reviewers:
+ vtl@google.com

[vtl, 2015-12-17 22:30:35]

LGTM w/very minor nits.

(I'm not the expert on the unbinding part, but it seems like a good idea to me
in general.)

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.cc
File mojo/gpu/gl_context_owner.cc (right):

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.c...
mojo/gpu/gl_context_owner.cc:11:
GLContextOwner::GLContextOwner(mojo::ApplicationConnector* connector)
Ditto (here and below).

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.h
File mojo/gpu/gl_context_owner.h (right):

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.h...
mojo/gpu/gl_context_owner.h:17: explicit
GLContextOwner(mojo::ApplicationConnector* connector);
nit (pre-existing condition): "mojo::" not necessary

(whether this should just be in the mojo namespace is another question <shrug>)

[jeffbrown, 2015-12-17 22:33:23]

Done.  FWIW, I'm thinking about moving all of this stuff into the mojo::gfx
namespace later on.

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.cc
File mojo/gpu/gl_context_owner.cc (right):

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.c...
mojo/gpu/gl_context_owner.cc:11:
GLContextOwner::GLContextOwner(mojo::ApplicationConnector* connector)
On 2015/12/17 22:30:35, vtl wrote:
> Ditto (here and below).

Done.

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.h
File mojo/gpu/gl_context_owner.h (right):

https://codereview.chromium.org/1537783002/diff/1/mojo/gpu/gl_context_owner.h...
mojo/gpu/gl_context_owner.h:17: explicit
GLContextOwner(mojo::ApplicationConnector* connector);
On 2015/12/17 22:30:35, vtl wrote:
> nit (pre-existing condition): "mojo::" not necessary
> 
> (whether this should just be in the mojo namespace is another question
<shrug>)

Done.