| Index: src/opts/SkPMFloat_SSSE3.h
|
| diff --git a/src/opts/SkPMFloat_SSSE3.h b/src/opts/SkPMFloat_SSSE3.h
|
| index 9ff73568fcaf24d87eee06bedb203a95a0dd7b17..c1c6d67446e0fb385220c0b53a0e7e1c512ffd67 100644
|
| --- a/src/opts/SkPMFloat_SSSE3.h
|
| +++ b/src/opts/SkPMFloat_SSSE3.h
|
| @@ -18,13 +18,13 @@
|
| const int _ = 255; // _ means to zero that byte.
|
| __m128i fix8 = _mm_set_epi32(0,0,0,c),
|
| fix8_32 = _mm_shuffle_epi8(fix8, _mm_set_epi8(_,_,_,3, _,_,_,2, _,_,_,1, _,_,_,0));
|
| - fVec = _mm_cvtepi32_ps(fix8_32);
|
| + fColors = _mm_cvtepi32_ps(fix8_32);
|
| SkASSERT(this->isValid());
|
| }
|
|
|
| inline SkPMColor SkPMFloat::trunc() const {
|
| const int _ = 255; // _ means to zero that byte.
|
| - __m128i fix8_32 = _mm_cvttps_epi32(fVec),
|
| + __m128i fix8_32 = _mm_cvttps_epi32(fColors.vec()),
|
| fix8 = _mm_shuffle_epi8(fix8_32, _mm_set_epi8(_,_,_,_, _,_,_,_, _,_,_,_, 12,8,4,0));
|
| SkPMColor c = _mm_cvtsi128_si32(fix8);
|
| SkPMColorAssert(c);
|
| @@ -37,7 +37,7 @@
|
|
|
| inline SkPMColor SkPMFloat::roundClamp() const {
|
| // We don't use _mm_cvtps_epi32, because we want precise control over how 0.5 rounds (up).
|
| - __m128i fix8_32 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), fVec)),
|
| + __m128i fix8_32 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), fColors.vec())),
|
| fix8_16 = _mm_packus_epi16(fix8_32, fix8_32),
|
| fix8 = _mm_packus_epi16(fix8_16, fix8_16);
|
| SkPMColor c = _mm_cvtsi128_si32(fix8);
|
| @@ -69,10 +69,10 @@
|
| SkPMColor colors[4]) {
|
| // Same as _SSE2.h's. We use 3 _mm_packus_epi16() where the naive loop uses 8.
|
| // We don't use _mm_cvtps_epi32, because we want precise control over how 0.5 rounds (up).
|
| - __m128i c0 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), a.fVec)),
|
| - c1 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), b.fVec)),
|
| - c2 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), c.fVec)),
|
| - c3 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), d.fVec));
|
| + __m128i c0 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), a.fColors.vec())),
|
| + c1 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), b.fColors.vec())),
|
| + c2 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), c.fColors.vec())),
|
| + c3 = _mm_cvttps_epi32(_mm_add_ps(_mm_set1_ps(0.5f), d.fColors.vec()));
|
| __m128i c3210 = _mm_packus_epi16(_mm_packus_epi16(c0, c1),
|
| _mm_packus_epi16(c2, c3));
|
| _mm_storeu_si128((__m128i*)colors, c3210);
|
|
|