Start using GrGpuCommandBuffer in GrDrawTarget.

src/gpu/batches/GrCopySurfaceBatch.h

Index: src/gpu/batches/GrCopySurfaceBatch.h
diff --git a/src/gpu/batches/GrCopySurfaceBatch.h b/src/gpu/batches/GrCopySurfaceBatch.h
index 3926643f8aedff441e2c36aece565a267c3d9ba8..d0269b79bf8c795494f551edf39d0b21e91bcd4d 100644
--- a/src/gpu/batches/GrCopySurfaceBatch.h
+++ b/src/gpu/batches/GrCopySurfaceBatch.h
@@ -36,7 +36,7 @@ public:
         GrRenderTarget* rt = fDst.get()->asRenderTarget();
         return rt ? rt->getUniqueID() : 0;
     }
-    GrRenderTarget* renderTarget() const override { return fDst.get()->asRenderTarget(); }
+    GrRenderTarget* renderTarget() const override { return nullptr; }
 
     SkString dumpInfo() const override {
         SkString string;
@@ -63,8 +63,13 @@ private:
 
     void onPrepare(GrBatchFlushState*) override {}
 
-    void onDraw(GrBatchFlushState* state) override {
-        state->gpu()->copySurface(fDst.get(), fSrc.get(), fSrcRect, fDstPoint);
+    void onDraw(GrBatchFlushState* state, GrGpuCommandBuffer* buffer) override {
+        if (!buffer) {
+            state->gpu()->copySurface(fDst.get(), fSrc.get(), fSrcRect, fDstPoint);
+        } else {
+            // currently we are not sending copies through the GrGpuCommandBuffer

[bsalomon, 2016/06/20 16:51:14]

Does this break anything with GL?

[egdaniel, 2016/06/22 15:26:57]

it does not. All copySurfaceBatches return a nullptr for render target. Before
this was only used to draw the debug batch bounds if the RT existed which really
isn't that interesting with copies anyways.

In GrDrawTarget, if we don't have a render target then we don't create a command
buffer. This means we get here and always call the copy directly on the gpu. See
comments in previous patch to why this is also safe for vulkan

[bsalomon, 2016/06/22 15:41:15]

This feels a little gnarly. Is there a different long term plan here? Do we
expect post-MDB that we will always have a command buffer?

[egdaniel, 2016/06/22 21:13:35]

As per discussion, copies should be moved away from Batches and thus not go
through command buffers.

+            SkASSERT(false);
+        }
     }
 
     GrPendingIOResource<GrSurface, kWrite_GrIOType> fDst;