Let SkBenchmark classes specify that they do no rendering.
Doing this gives us a 15-20% speedup in bench cycle time.
Here again I'm just picking the easy targets.

http://codereview.appspot.com/6500115/



git-svn-id: http://skia.googlecode.com/svn/trunk@5525 2bbb7eff-a529-9590-31e7-b0007b416f81
diff --git a/bench/MathBench.cpp b/bench/MathBench.cpp
index 414036b..df3d3a5 100644
--- a/bench/MathBench.cpp
+++ b/bench/MathBench.cpp
@@ -30,6 +30,8 @@
         for (int i = 0; i < kBuffer; ++i) {
             fSrc[i] = rand.nextSScalar1();
         }
+
+        fIsRendering = false;
     }
 
     virtual void performTest(float* SK_RESTRICT dst,
@@ -265,6 +267,7 @@
             fProc = gRec[index].fProc;
             fName = gRec[index].fName;
         }
+        fIsRendering = false;
     }
 
 protected:
@@ -330,6 +333,7 @@
         } else {
             fName = "floor_std";
         }
+        fIsRendering = false;
     }
 
     virtual void process(float) {}