Merge "Reduce hwui CPU time by using glDrawRangeElements" am: 057c7c3164 am: 94f4f5e45f

am: b5facc7516

Change-Id: I9e96bf83ae5ebd4237fc4d78a58e83793de86aad

3 files changed, 15 insertions, 3 deletions

diff --git a/libs/hwui/Glop.h b/libs/hwui/Glop.h
index 34c7934db198..e91c08d5a351 100644
--- a/libs/hwui/Glop.h
+++ b/libs/hwui/Glop.h
@@ -110,6 +110,7 @@ public:
         } vertices;
 
         int elementCount;
+        int vertexCount; // only used for meshes (for glDrawRangeElements)
         TextureVertex mappedVertices[4];
     } mesh;
 
diff --git a/libs/hwui/GlopBuilder.cpp b/libs/hwui/GlopBuilder.cpp
index 931a55a70fd8..2e9a6e895d8a 100644
--- a/libs/hwui/GlopBuilder.cpp
+++ b/libs/hwui/GlopBuilder.cpp
@@ -199,6 +199,7 @@ GlopBuilder& GlopBuilder::setMeshVertexBuffer(const VertexBuffer& vertexBuffer)
             alphaVertex ? kAlphaVertexStride : kVertexStride };
     mOutGlop->mesh.elementCount = indices
                 ? vertexBuffer.getIndexCount() : vertexBuffer.getVertexCount();
+    mOutGlop->mesh.vertexCount = vertexBuffer.getVertexCount(); // used for glDrawRangeElements()
     return *this;
 }
 
diff --git a/libs/hwui/renderstate/RenderState.cpp b/libs/hwui/renderstate/RenderState.cpp
index c8833d2a7489..8b737bb241cb 100644
--- a/libs/hwui/renderstate/RenderState.cpp
+++ b/libs/hwui/renderstate/RenderState.cpp
@@ -413,18 +413,28 @@ void RenderState::render(const Glop& glop, const Matrix4& orthoMatrix) {
         const GLbyte* vertexData = static_cast<const GLbyte*>(vertices.position);
         while (elementsCount > 0) {
             GLsizei drawCount = std::min(elementsCount, (GLsizei) kMaxNumberOfQuads * 6);
+            GLsizei vertexCount = (drawCount / 6) * 4;
             meshState().bindPositionVertexPointer(vertexData, vertices.stride);
             if (vertices.attribFlags & VertexAttribFlags::TextureCoord) {
                 meshState().bindTexCoordsVertexPointer(
                         vertexData + kMeshTextureOffset, vertices.stride);
             }
 
-            glDrawElements(mesh.primitiveMode, drawCount, GL_UNSIGNED_SHORT, nullptr);
+            if (mCaches->extensions().getMajorGlVersion() >= 3) {
+                glDrawRangeElements(mesh.primitiveMode, 0, vertexCount-1, drawCount, GL_UNSIGNED_SHORT, nullptr);
+            } else {
+                glDrawElements(mesh.primitiveMode, drawCount, GL_UNSIGNED_SHORT, nullptr);
+            }
             elementsCount -= drawCount;
-            vertexData += (drawCount / 6) * 4 * vertices.stride;
+            vertexData += vertexCount * vertices.stride;
         }
     } else if (indices.bufferObject || indices.indices) {
-        glDrawElements(mesh.primitiveMode, mesh.elementCount, GL_UNSIGNED_SHORT, indices.indices);
+        if (mCaches->extensions().getMajorGlVersion() >= 3) {
+            // use glDrawRangeElements to reduce CPU overhead (otherwise the driver has to determine the min/max index values)
+            glDrawRangeElements(mesh.primitiveMode, 0, mesh.vertexCount-1, mesh.elementCount, GL_UNSIGNED_SHORT, indices.indices);
+        } else {
+            glDrawElements(mesh.primitiveMode, mesh.elementCount, GL_UNSIGNED_SHORT, indices.indices);
+        }
     } else {
         glDrawArrays(mesh.primitiveMode, 0, mesh.elementCount);
     }