gpu: Refactor GpuMemoryBuffer framework for multi-process support.

gpu/command_buffer/service/gpu_scheduler.h

Index: gpu/command_buffer/service/gpu_scheduler.h
diff --git a/gpu/command_buffer/service/gpu_scheduler.h b/gpu/command_buffer/service/gpu_scheduler.h
index 8fb791cd3fd7fa728f930087cd6e6d17d1837241..c2fd91e4d3f1a1aa9a76546309ce900b24e290ef 100644
--- a/gpu/command_buffer/service/gpu_scheduler.h
+++ b/gpu/command_buffer/service/gpu_scheduler.h
@@ -88,6 +88,7 @@ class GPU_EXPORT GpuScheduler
   virtual bool SetGetBuffer(int32 transfer_buffer_id) OVERRIDE;
   virtual bool SetGetOffset(int32 offset) OVERRIDE;
   virtual int32 GetGetOffset() OVERRIDE;
+  virtual gfx::GLImage* GetImage(int32 image_id) OVERRIDE;

[no sievers, 2013/08/01 00:38:04]

I find plumbing the ImageManager through the Scheduler a little odd.
Maybe the decoder should just keep its own pointer?

Would also make some changes to the unittests and other files unnecessary here.

[reveman, 2013/08/01 13:32:18]

I'd rather not add new state unless necessary. Current patch is consistent with
how transfer buffers and shared memory is handled.

[piman, 2013/08/01 21:19:30]

Transfer buffers and shared memory are an essential part of command buffers.
This is GL-specific and I think it should be kept out of the command processing
layer.

[reveman, 2013/08/08 23:19:00]

Remove this.

 
   void SetCommandProcessedCallback(const base::Closure& callback);