The compiler can generate smulbb perfectly well nowadays.

src/core/SkMathPriv.h

 /*
  * Copyright 2012 Google Inc.
  *
  * Use of this source code is governed by a BSD-style license that can be
  * found in the LICENSE file.
  */
 
 #ifndef SkMathPriv_DEFINED
 #define SkMathPriv_DEFINED
 
@@ skipping 39 matching lines @@
 }
 
 ///////////////////////////////////////////////////////////////////////////////
 
 /** Return a*b/255, truncating away any fractional bits. Only valid if both
  a and b are 0..255
  */
 static inline U8CPU SkMulDiv255Trunc(U8CPU a, U8CPU b) {
     SkASSERT((uint8_t)a == a);
     SkASSERT((uint8_t)b == b);
-    unsigned prod = SkMulS16(a, b) + 1;
+    unsigned prod = a*b + 1;
     return (prod + (prod >> 8)) >> 8;
 }
 
 /** Return (a*b)/255, taking the ceiling of any fractional bits. Only valid if
  both a and b are 0..255. The expected result equals (a * b + 254) / 255.
  */
 static inline U8CPU SkMulDiv255Ceiling(U8CPU a, U8CPU b) {
     SkASSERT((uint8_t)a == a);
     SkASSERT((uint8_t)b == b);
-    unsigned prod = SkMulS16(a, b) + 255;
+    unsigned prod = a*b + 255;
     return (prod + (prod >> 8)) >> 8;
 }
 
 /** Just the rounding step in SkDiv255Round: round(value / 255)
  */
 static inline unsigned SkDiv255Round(unsigned prod) {
     prod += 128;
     return (prod + (prod >> 8)) >> 8;
 }
 
 #endif