By default, Chromium optimizes for size when compiling on Android.
Forcing the SkAlphaMulQ() function inline can yield as much as a 5-10%
improvement in rasterization time for some of Chromium's telemetry
tests on the Nexus 10, since it's in the inner loop of complex blends.

R=mtklein@google.com, reed@google.com, tomhudson@google.com

Author: tomhudson@chromium.org

Review URL: https://codereview.chromium.org/283753003

git-svn-id: http://skia.googlecode.com/svn/trunk@14723 2bbb7eff-a529-9590-31e7-b0007b416f81
diff --git a/include/core/SkColorPriv.h b/include/core/SkColorPriv.h
index 8fd1e57..21e1783 100644
--- a/include/core/SkColorPriv.h
+++ b/include/core/SkColorPriv.h
@@ -524,7 +524,9 @@
     return SkPackARGB32(a, r, g, b);
 }
 
-static inline uint32_t SkAlphaMulQ(uint32_t c, unsigned scale) {
+// When Android is compiled optimizing for size, SkAlphaMulQ doesn't get
+// inlined; forcing inlining significantly improves performance.
+static SK_ALWAYS_INLINE uint32_t SkAlphaMulQ(uint32_t c, unsigned scale) {
     uint32_t mask = 0xFF00FF;
 
     uint32_t rb = ((c & mask) * scale) >> 8;