Change SkUnitScalarClampToByte to more accurate implementation.

include/core/SkColorPriv.h

Index: include/core/SkColorPriv.h
diff --git a/include/core/SkColorPriv.h b/include/core/SkColorPriv.h
index bd0e040566cb0408c8ea5697d4e7c2c20781de43..23e82866824ae5e9e393340c740e5bfe829455ba 100644
--- a/include/core/SkColorPriv.h
+++ b/include/core/SkColorPriv.h
@@ -224,13 +224,7 @@ static inline int SkAlphaBlend255(S16CPU src, S16CPU dst, U8CPU alpha) {
 }
 
 static inline U8CPU SkUnitScalarClampToByte(SkScalar x) {
-    if (x < 0) {
-        return 0;
-    }
-    if (x >= SK_Scalar1) {
-        return 255;
-    }
-    return SkScalarToFixed(x) >> 8;
+    return static_cast<U8CPU>(SkScalarPin(x, 0, 1) * 255);

[reed1, 2016/02/22 18:09:32]

In nearly all of our new Sk4f code, we do the following:

value * 255 + 0.5

I wonder if we should to that here as well?

 }
 
 #define SK_R16_BITS     5