st/mesa: always prefer pipe->clear over clear_with_quad (v2)
authorMarek Olšák <marek.olsak@amd.com>
Fri, 6 Dec 2013 17:58:52 +0000 (18:58 +0100)
committerMarek Olšák <marek.olsak@amd.com>
Mon, 13 Jan 2014 14:25:31 +0000 (15:25 +0100)
v2: clear depth and stencil together

src/mesa/state_tracker/st_cb_clear.c

index 887e58bd9b6840b538a40feb048550e21cf28351..363f2f746991899fd9daac6fb91a87e19582340f 100644 (file)
@@ -476,14 +476,23 @@ st_Clear(struct gl_context *ctx, GLbitfield mask)
       }
    }
 
-   /*
-    * If we're going to use clear_with_quad() for any reason, use it for
-    * everything possible.
+   /* Always clear depth and stencil together.
+    * This can only happen when the stencil writemask is not a full mask.
+    */
+   if (quad_buffers & PIPE_CLEAR_DEPTHSTENCIL &&
+       clear_buffers & PIPE_CLEAR_DEPTHSTENCIL) {
+      quad_buffers |= clear_buffers & PIPE_CLEAR_DEPTHSTENCIL;
+      clear_buffers &= ~PIPE_CLEAR_DEPTHSTENCIL;
+   }
+
+   /* Only use quad-based clearing for the renderbuffers which cannot
+    * use pipe->clear. We want to always use pipe->clear for the other
+    * renderbuffers, because it's likely to be faster.
     */
    if (quad_buffers) {
-      quad_buffers |= clear_buffers;
       clear_with_quad(ctx, quad_buffers);
-   } else if (clear_buffers) {
+   }
+   if (clear_buffers) {
       /* We can't translate the clear color to the colorbuffer format,
        * because different colorbuffers may have different formats.
        */