div255(x) as ((x+128)*257)>>16 with SSE

src/opts/Sk4px_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.
  */
 
 namespace { // See Sk4px.h
 
 inline Sk4px Sk4px::DupPMColor(SkPMColor px) { return Sk16b((uint8x16_t)vdupq_n_u32(px)); }
@@ skipping 39 matching lines @@
     return Sk16h(vmull_u8(vget_low_u8 (this->fVec), vget_low_u8 (other.fVec)),
                  vmull_u8(vget_high_u8(this->fVec), vget_high_u8(other.fVec)));
 }
 
 inline Sk4px Sk4px::Wide::addNarrowHi(const Sk16h& other) const {
     const Sk4px::Wide o(other);  // Should be no code, but allows us to access fLo, fHi.
     return Sk16b(vcombine_u8(vaddhn_u16(this->fLo.fVec, o.fLo.fVec),
                              vaddhn_u16(this->fHi.fVec, o.fHi.fVec)));
 }
 
+inline Sk4px Sk4px::Wide::div255() const {
+    // Calculated as ((x+128) + ((x+128)>>8)) >> 8.
+    auto v = *this + Sk16h(128);
+    return v.addNarrowHi(v>>8);
+}
+
 inline Sk4px Sk4px::alphas() const {
     auto as = vshrq_n_u32((uint32x4_t)fVec, SK_A32_SHIFT);  // ___3 ___2 ___1 ___0
     return Sk16b((uint8x16_t)vmulq_n_u32(as, 0x01010101));  // 3333 2222 1111 0000
 }
 
 inline Sk4px Sk4px::Load4Alphas(const SkAlpha a[4]) {
     uint8x16_t a8 = vdupq_n_u8(0);                           // ____ ____ ____ ____
     a8 = vld1q_lane_u8(a+0, a8,  0);                         // ____ ____ ____ ___0
     a8 = vld1q_lane_u8(a+1, a8,  4);                         // ____ ____ ___1 ___0
     a8 = vld1q_lane_u8(a+2, a8,  8);                         // ____ ___2 ___1 ___0
@@ skipping 85 matching lines @@
     auto v = narrow_to_565(this->fVec);
     dst[0] = vget_lane_u16(v, 0);
     dst[1] = vget_lane_u16(v, 1);
 }
 inline void Sk4px::store1(SkPMColor16 dst[1]) const {
     dst[0] = vget_lane_u16(narrow_to_565(this->fVec), 0);
 }
 
 } // namespace