Sk4px

src/opts/SkNx_sse.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_sse_DEFINED
 #define SkNx_sse_DEFINED
 
 // This file may assume <= SSE2, but must check SK_CPU_SSE_LEVEL for anything more recent.
 #include <immintrin.h>
 
 template <>
 class SkNb<2, 4> {
 public:
     SkNb(const __m128i& vec) : fVec(vec) {}
 
     SkNb() {}
     bool allTrue() const { return 0xff == (_mm_movemask_epi8(fVec) & 0xff); }
     bool anyTrue() const { return 0x00 != (_mm_movemask_epi8(fVec) & 0xff); }
 
-private:
     __m128i fVec;
 };
 
 template <>
 class SkNb<4, 4> {
 public:
     SkNb(const __m128i& vec) : fVec(vec) {}
 
     SkNb() {}
     bool allTrue() const { return 0xffff == _mm_movemask_epi8(fVec); }
     bool anyTrue() const { return 0x0000 != _mm_movemask_epi8(fVec); }
 
-private:
     __m128i fVec;
 };
 
 template <>
 class SkNb<2, 8> {
 public:
     SkNb(const __m128i& vec) : fVec(vec) {}
 
     SkNb() {}
     bool allTrue() const { return 0xffff == _mm_movemask_epi8(fVec); }
     bool anyTrue() const { return 0x0000 != _mm_movemask_epi8(fVec); }
 
-private:
     __m128i fVec;
 };
 
 
 template <>
 class SkNf<2, float> {
     typedef SkNb<2, 4> Nb;
 public:
     SkNf(const __m128& vec) : fVec(vec) {}
 
@@ skipping 28 matching lines @@
 
     SkNf       invert() const { return SkNf(1) / *this; }
     SkNf approxInvert() const { return _mm_rcp_ps(fVec); }
 
     template <int k> float kth() const {
         SkASSERT(0 <= k && k < 2);
         union { __m128 v; float fs[4]; } pun = {fVec};
         return pun.fs[k&1];
     }
 
-private:
     __m128 fVec;
 };
 
 template <>
 class SkNf<2, double> {
     typedef SkNb<2, 8> Nb;
 public:
     SkNf(const __m128d& vec) : fVec(vec) {}
 
     SkNf() {}
@@ skipping 25 matching lines @@
 
     SkNf       invert() const { return SkNf(1) / *this; }
     SkNf approxInvert() const { return _mm_cvtps_pd(_mm_rcp_ps(_mm_cvtpd_ps(fVec))); }
 
     template <int k> double kth() const {
         SkASSERT(0 <= k && k < 2);
         union { __m128d v; double ds[2]; } pun = {fVec};
         return pun.ds[k&1];
     }
 
-private:
     __m128d fVec;
 };
 
 template <>
 class SkNi<4, int> {
 public:
     SkNi(const __m128i& vec) : fVec(vec) {}
 
     SkNi() {}
     explicit SkNi(int val) : fVec(_mm_set1_epi32(val)) {}
@@ skipping 17 matching lines @@
     template <int k> int kth() const {
         SkASSERT(0 <= k && k < 4);
         switch (k) {
             case 0: return _mm_cvtsi128_si32(fVec);
             case 1: return _mm_cvtsi128_si32(_mm_srli_si128(fVec,  4));
             case 2: return _mm_cvtsi128_si32(_mm_srli_si128(fVec,  8));
             case 3: return _mm_cvtsi128_si32(_mm_srli_si128(fVec, 12));
             default: SkASSERT(false); return 0;
         }
     }
-protected:
+
     __m128i fVec;
 };
 
 template <>
 class SkNf<4, float> {
     typedef SkNb<4, 4> Nb;
 public:
     SkNf(const __m128& vec) : fVec(vec) {}
 
     SkNf() {}
@@ skipping 27 matching lines @@
 
     SkNf       invert() const { return SkNf(1) / *this; }
     SkNf approxInvert() const { return _mm_rcp_ps(fVec); }
 
     template <int k> float kth() const {
         SkASSERT(0 <= k && k < 4);
         union { __m128 v; float fs[4]; } pun = {fVec};
         return pun.fs[k&3];
     }
 
-protected:
     __m128 fVec;
 };
 
 template <>
 class SkNi<4, uint16_t> {
 public:
     SkNi(const __m128i& vec) : fVec(vec) {}
 
     SkNi() {}
     explicit SkNi(uint16_t val) : fVec(_mm_set1_epi16(val)) {}
     static SkNi Load(const uint16_t vals[4]) { return _mm_loadl_epi64((const __m128i*)vals); }
     SkNi(uint16_t a, uint16_t b, uint16_t c, uint16_t d) : fVec(_mm_setr_epi16(a,b,c,d,0,0,0,0)) {}
 
     void store(uint16_t vals[4]) const { _mm_storel_epi64((__m128i*)vals, fVec); }
 
     SkNi operator + (const SkNi& o) const { return _mm_add_epi16(fVec, o.fVec); }
     SkNi operator - (const SkNi& o) const { return _mm_sub_epi16(fVec, o.fVec); }
     SkNi operator * (const SkNi& o) const { return _mm_mullo_epi16(fVec, o.fVec); }
 
     SkNi operator << (int bits) const { return _mm_slli_epi16(fVec, bits); }
     SkNi operator >> (int bits) const { return _mm_srli_epi16(fVec, bits); }
 
     template <int k> uint16_t kth() const {
         SkASSERT(0 <= k && k < 4);
         return _mm_extract_epi16(fVec, k);
     }
-protected:
+
     __m128i fVec;
 };
 
 template <>
 class SkNi<8, uint16_t> {
 public:
     SkNi(const __m128i& vec) : fVec(vec) {}
 
     SkNi() {}
     explicit SkNi(uint16_t val) : fVec(_mm_set1_epi16(val)) {}
     static SkNi Load(const uint16_t vals[8]) { return _mm_loadu_si128((const __m128i*)vals); }
     SkNi(uint16_t a, uint16_t b, uint16_t c, uint16_t d,
          uint16_t e, uint16_t f, uint16_t g, uint16_t h) : fVec(_mm_setr_epi16(a,b,c,d,e,f,g,h)) {}
 
     void store(uint16_t vals[8]) const { _mm_storeu_si128((__m128i*)vals, fVec); }
 
     SkNi operator + (const SkNi& o) const { return _mm_add_epi16(fVec, o.fVec); }
     SkNi operator - (const SkNi& o) const { return _mm_sub_epi16(fVec, o.fVec); }
     SkNi operator * (const SkNi& o) const { return _mm_mullo_epi16(fVec, o.fVec); }
 
     SkNi operator << (int bits) const { return _mm_slli_epi16(fVec, bits); }
     SkNi operator >> (int bits) const { return _mm_srli_epi16(fVec, bits); }
 
     template <int k> uint16_t kth() const {
         SkASSERT(0 <= k && k < 8);
         return _mm_extract_epi16(fVec, k);
     }
-protected:
+
     __m128i fVec;
 };
 
+template <>
+class SkNi<16, uint8_t> {
+public:
+    SkNi(const __m128i& vec) : fVec(vec) {}
+
+    SkNi() {}
+    explicit SkNi(uint8_t val) : fVec(_mm_set1_epi8(val)) {}
+    static SkNi Load(const uint8_t vals[16]) { return _mm_loadu_si128((const __m128i*)vals); }
+    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(_mm_setr_epi8(a,b,c,d, e,f,g,h, i,j,k,l, m,n,o,p)) {}
+
+    void store(uint8_t vals[16]) const { _mm_storeu_si128((__m128i*)vals, fVec); }
+
+    SkNi operator + (const SkNi& o) const { return _mm_add_epi8(fVec, o.fVec); }
+    SkNi operator - (const SkNi& o) const { return _mm_sub_epi8(fVec, o.fVec); }
+
+    // SSE cannot multiply or shift vectors of uint8_t.
+    SkNi operator * (const SkNi& o) const { SkASSERT(false); return fVec; }
+    SkNi operator << (int bits) const { SkASSERT(false); return fVec; }
+    SkNi operator >> (int bits) const { SkASSERT(false); return fVec; }
+
+    template <int k> uint8_t kth() const {
+        SkASSERT(0 <= k && k < 16);
+        // SSE4.1 would just `return _mm_extract_epi8(fVec, k)`.  We have to read 16-bits instead.
+        int pair = _mm_extract_epi16(fVec, k/2);
+        return k % 2 == 0 ? pair : (pair >> 8);
+    }
+
+    __m128i fVec;
+};
+
 #endif//SkNx_sse_DEFINED