Implemented support for GL constext share groups in the renderer process.

Implemented support for GL context share groups in the renderer process.I put all compositor and canvas contexts in the same share group so they can share IDs.
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=95345

[apatrick_chromium, 2011-08-02 20:47:03]

This isn't ready for review yet; I'm still testing it. If you want to integrate
it with your changes though, here it is.

[apatrick_chromium, 2011-08-03 21:44:10]

Ready for review.

[jamesr, 2011-08-03 21:51:20]

lgtm w/ comments

http://codereview.chromium.org/7554015/diff/16001/content/renderer/gpu/render...
File content/renderer/gpu/renderer_gl_context.cc (right):

http://codereview.chromium.org/7554015/diff/16001/content/renderer/gpu/render...
content/renderer/gpu/renderer_gl_context.cc:204:
parent_->gles2_implementation_->Flush();
why is this doing a flush on the parent? shouldn't it be doing a flush on this
context, since it's really creating a texture here (and not on the parent)?

http://codereview.chromium.org/7554015/diff/16001/gpu/command_buffer/common/i...
File gpu/command_buffer/common/id_allocator.h (right):

http://codereview.chromium.org/7554015/diff/16001/gpu/command_buffer/common/i...
gpu/command_buffer/common/id_allocator.h:76: // implementation does not track
which IDs are currently used. It is usedul for
typo: usedul -> useful

http://codereview.chromium.org/7554015/diff/16001/webkit/gpu/webgraphicsconte...
File webkit/gpu/webgraphicscontext3d_in_process_command_buffer_impl.cc (right):

http://codereview.chromium.org/7554015/diff/16001/webkit/gpu/webgraphicsconte...
webkit/gpu/webgraphicscontext3d_in_process_command_buffer_impl.cc:318:
CHECK(state.error == ::gpu::error::kNoError);
On second thought, can you get rid of the RunAllPending() call here?  I'd be
much happier without it, even if that means we had to change it if we later made
the command buffer scheduler use PostTask().  Right now this is just super scary
code :/