Turn on Sk4px xfermodes when we have NEON too.

src/opts/SkNx_neon.h

 /*
  * Copyright 2015 Google Inc.
  *
  * Use of this source code is governed by a BSD-style license that can be
  * found in the LICENSE file.
  */
 
 #ifndef SkNx_neon_DEFINED
 #define SkNx_neon_DEFINED
 
@@ skipping 337 matching lines @@
 
     SkNi(uint8_t a, uint8_t b, uint8_t c, uint8_t d,
          uint8_t e, uint8_t f, uint8_t g, uint8_t h,
          uint8_t i, uint8_t j, uint8_t k, uint8_t l,
          uint8_t m, uint8_t n, uint8_t o, uint8_t p) {
         fVec = (uint8x16_t) { a,b,c,d, e,f,g,h, i,j,k,l, m,n,o,p };
     }
 
     void store(uint8_t vals[16]) const { vst1q_u8(vals, fVec); }
 
+    SkNi saturatedAdd(const SkNi& o) const { return vqaddq_u8(fVec, o.fVec); }
+
     SkNi operator + (const SkNi& o) const { return vaddq_u8(fVec, o.fVec); }
     SkNi operator - (const SkNi& o) const { return vsubq_u8(fVec, o.fVec); }
     SkNi operator * (const SkNi& o) const { return vmulq_u8(fVec, o.fVec); }
 
     SkNi operator << (int bits) const { SHIFT8(vshlq_n_u8, fVec, bits); }
     SkNi operator >> (int bits) const { SHIFT8(vshrq_n_u8, fVec, bits); }
 
     template <int k> uint8_t kth() const {
         SkASSERT(0 <= k && k < 15);
         return vgetq_lane_u8(fVec, k&16);
     }
 
     uint8x16_t fVec;
 };
 
 #undef SHIFT32
 #undef SHIFT16
 #undef SHIFT8
 
 #endif//SkNx_neon_DEFINED