mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 1 | /* |
| 2 | * Copyright 2015 Google Inc. |
| 3 | * |
| 4 | * Use of this source code is governed by a BSD-style license that can be |
| 5 | * found in the LICENSE file. |
| 6 | */ |
| 7 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 8 | namespace { // See Sk4px.h |
mtklein | aa999cb | 2015-05-22 17:18:21 -0700 | [diff] [blame] | 9 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 10 | inline Sk4px Sk4px::DupPMColor(SkPMColor px) { return Sk16b((uint8x16_t)vdupq_n_u32(px)); } |
| 11 | |
| 12 | inline Sk4px Sk4px::Load4(const SkPMColor px[4]) { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 13 | return Sk16b((uint8x16_t)vld1q_u32(px)); |
| 14 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 15 | inline Sk4px Sk4px::Load2(const SkPMColor px[2]) { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 16 | uint32x2_t px2 = vld1_u32(px); |
| 17 | return Sk16b((uint8x16_t)vcombine_u32(px2, px2)); |
| 18 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 19 | inline Sk4px Sk4px::Load1(const SkPMColor px[1]) { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 20 | return Sk16b((uint8x16_t)vdupq_n_u32(*px)); |
| 21 | } |
| 22 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 23 | inline void Sk4px::store4(SkPMColor px[4]) const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 24 | vst1q_u32(px, (uint32x4_t)this->fVec); |
| 25 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 26 | inline void Sk4px::store2(SkPMColor px[2]) const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 27 | vst1_u32(px, (uint32x2_t)vget_low_u8(this->fVec)); |
| 28 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 29 | inline void Sk4px::store1(SkPMColor px[1]) const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 30 | vst1q_lane_u32(px, (uint32x4_t)this->fVec, 0); |
| 31 | } |
| 32 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 33 | inline Sk4px::Wide Sk4px::widenLo() const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 34 | return Sk16h(vmovl_u8(vget_low_u8 (this->fVec)), |
| 35 | vmovl_u8(vget_high_u8(this->fVec))); |
| 36 | } |
| 37 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 38 | inline Sk4px::Wide Sk4px::widenHi() const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 39 | return Sk16h(vshll_n_u8(vget_low_u8 (this->fVec), 8), |
| 40 | vshll_n_u8(vget_high_u8(this->fVec), 8)); |
| 41 | } |
| 42 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 43 | inline Sk4px::Wide Sk4px::widenLoHi() const { |
mtklein | 4be181e | 2015-07-14 10:54:19 -0700 | [diff] [blame] | 44 | auto zipped = vzipq_u8(this->fVec, this->fVec); |
| 45 | return Sk16h((uint16x8_t)zipped.val[0], |
| 46 | (uint16x8_t)zipped.val[1]); |
| 47 | } |
| 48 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 49 | inline Sk4px::Wide Sk4px::mulWiden(const Sk16b& other) const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 50 | return Sk16h(vmull_u8(vget_low_u8 (this->fVec), vget_low_u8 (other.fVec)), |
| 51 | vmull_u8(vget_high_u8(this->fVec), vget_high_u8(other.fVec))); |
| 52 | } |
| 53 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 54 | inline Sk4px Sk4px::Wide::addNarrowHi(const Sk16h& other) const { |
mtklein | d2ffd36 | 2015-05-12 06:11:21 -0700 | [diff] [blame] | 55 | const Sk4px::Wide o(other); // Should be no code, but allows us to access fLo, fHi. |
| 56 | return Sk16b(vcombine_u8(vaddhn_u16(this->fLo.fVec, o.fLo.fVec), |
| 57 | vaddhn_u16(this->fHi.fVec, o.fHi.fVec))); |
| 58 | } |
mtklein | 8a90edc | 2015-05-13 12:19:42 -0700 | [diff] [blame] | 59 | |
mtklein | cbf4fba | 2015-11-17 14:19:52 -0800 | [diff] [blame] | 60 | inline Sk4px Sk4px::Wide::div255() const { |
mtklein | 9d34406 | 2015-12-07 08:21:11 -0800 | [diff] [blame] | 61 | // Calculated as (x + (x+128)>>8 +128) >> 8. The 'r' in each instruction provides each +128. |
| 62 | return Sk16b(vcombine_u8(vraddhn_u16(this->fLo.fVec, vrshrq_n_u16(this->fLo.fVec, 8)), |
| 63 | vraddhn_u16(this->fHi.fVec, vrshrq_n_u16(this->fHi.fVec, 8)))); |
mtklein | cbf4fba | 2015-11-17 14:19:52 -0800 | [diff] [blame] | 64 | } |
| 65 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 66 | inline Sk4px Sk4px::alphas() const { |
mtklein | 343c7d1 | 2015-06-22 11:00:47 -0700 | [diff] [blame] | 67 | auto as = vshrq_n_u32((uint32x4_t)fVec, SK_A32_SHIFT); // ___3 ___2 ___1 ___0 |
| 68 | return Sk16b((uint8x16_t)vmulq_n_u32(as, 0x01010101)); // 3333 2222 1111 0000 |
mtklein | 8a90edc | 2015-05-13 12:19:42 -0700 | [diff] [blame] | 69 | } |
| 70 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 71 | inline Sk4px Sk4px::Load4Alphas(const SkAlpha a[4]) { |
mtklein | 343c7d1 | 2015-06-22 11:00:47 -0700 | [diff] [blame] | 72 | uint8x16_t a8 = vdupq_n_u8(0); // ____ ____ ____ ____ |
| 73 | a8 = vld1q_lane_u8(a+0, a8, 0); // ____ ____ ____ ___0 |
| 74 | a8 = vld1q_lane_u8(a+1, a8, 4); // ____ ____ ___1 ___0 |
| 75 | a8 = vld1q_lane_u8(a+2, a8, 8); // ____ ___2 ___1 ___0 |
| 76 | a8 = vld1q_lane_u8(a+3, a8, 12); // ___3 ___2 ___1 ___0 |
| 77 | auto a32 = (uint32x4_t)a8; // |
| 78 | return Sk16b((uint8x16_t)vmulq_n_u32(a32, 0x01010101)); // 3333 2222 1111 0000 |
mtklein | 8a90edc | 2015-05-13 12:19:42 -0700 | [diff] [blame] | 79 | } |
| 80 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 81 | inline Sk4px Sk4px::Load2Alphas(const SkAlpha a[2]) { |
mtklein | 343c7d1 | 2015-06-22 11:00:47 -0700 | [diff] [blame] | 82 | uint8x16_t a8 = vdupq_n_u8(0); // ____ ____ ____ ____ |
| 83 | a8 = vld1q_lane_u8(a+0, a8, 0); // ____ ____ ____ ___0 |
| 84 | a8 = vld1q_lane_u8(a+1, a8, 4); // ____ ____ ___1 ___0 |
| 85 | auto a32 = (uint32x4_t)a8; // |
| 86 | return Sk16b((uint8x16_t)vmulq_n_u32(a32, 0x01010101)); // ____ ____ 1111 0000 |
mtklein | 8a90edc | 2015-05-13 12:19:42 -0700 | [diff] [blame] | 87 | } |
mtklein | 0135a41 | 2015-05-15 10:36:21 -0700 | [diff] [blame] | 88 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 89 | inline Sk4px Sk4px::zeroColors() const { |
mtklein | 0135a41 | 2015-05-15 10:36:21 -0700 | [diff] [blame] | 90 | return Sk16b(vandq_u8(this->fVec, (uint8x16_t)vdupq_n_u32(0xFF << SK_A32_SHIFT))); |
| 91 | } |
| 92 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 93 | inline Sk4px Sk4px::zeroAlphas() const { |
mtklein | 0135a41 | 2015-05-15 10:36:21 -0700 | [diff] [blame] | 94 | // vbic(a,b) == a & ~b |
| 95 | return Sk16b(vbicq_u8(this->fVec, (uint8x16_t)vdupq_n_u32(0xFF << SK_A32_SHIFT))); |
| 96 | } |
| 97 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 98 | static inline uint8x16_t widen_to_8888(uint16x4_t v) { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 99 | // RGB565 format: |R....|G.....|B....| |
| 100 | // Bit: 16 11 5 0 |
| 101 | |
| 102 | // First get each pixel into its own 32-bit lane. |
| 103 | // v == rgb3 rgb2 rgb1 rgb0 |
| 104 | // spread == 0000 rgb3 0000 rgb2 0000 rgb1 0000 rgb0 |
| 105 | uint32x4_t spread = vmovl_u16(v); |
| 106 | |
| 107 | // Get each color independently, still in 565 precison but down at bit 0. |
| 108 | auto r5 = vshrq_n_u32(spread, 11), |
| 109 | g6 = vandq_u32(vdupq_n_u32(63), vshrq_n_u32(spread, 5)), |
| 110 | b5 = vandq_u32(vdupq_n_u32(31), spread); |
| 111 | |
| 112 | // Scale 565 precision up to 8-bit each, filling low 323 bits with high bits of each component. |
| 113 | auto r8 = vorrq_u32(vshlq_n_u32(r5, 3), vshrq_n_u32(r5, 2)), |
| 114 | g8 = vorrq_u32(vshlq_n_u32(g6, 2), vshrq_n_u32(g6, 4)), |
| 115 | b8 = vorrq_u32(vshlq_n_u32(b5, 3), vshrq_n_u32(b5, 2)); |
| 116 | |
| 117 | // Now put all the 8-bit components into SkPMColor order. |
| 118 | return (uint8x16_t)vorrq_u32(vshlq_n_u32(r8, SK_R32_SHIFT), // TODO: one shift is zero... |
| 119 | vorrq_u32(vshlq_n_u32(g8, SK_G32_SHIFT), |
| 120 | vorrq_u32(vshlq_n_u32(b8, SK_B32_SHIFT), |
| 121 | vdupq_n_u32(0xFF << SK_A32_SHIFT)))); |
| 122 | } |
| 123 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 124 | static inline uint16x4_t narrow_to_565(uint8x16_t w8x16) { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 125 | uint32x4_t w = (uint32x4_t)w8x16; |
| 126 | |
| 127 | // Extract out top RGB 565 bits of each pixel, with no rounding. |
| 128 | auto r5 = vandq_u32(vdupq_n_u32(31), vshrq_n_u32(w, SK_R32_SHIFT + 3)), |
| 129 | g6 = vandq_u32(vdupq_n_u32(63), vshrq_n_u32(w, SK_G32_SHIFT + 2)), |
| 130 | b5 = vandq_u32(vdupq_n_u32(31), vshrq_n_u32(w, SK_B32_SHIFT + 3)); |
| 131 | |
| 132 | // Now put the bits in place in the low 16-bits of each 32-bit lane. |
| 133 | auto spread = vorrq_u32(vshlq_n_u32(r5, 11), |
| 134 | vorrq_u32(vshlq_n_u32(g6, 5), |
| 135 | b5)); |
| 136 | |
| 137 | // Pack the low 16-bits of our 128-bit register down into a 64-bit register. |
| 138 | // spread == 0000 rgb3 0000 rgb2 0000 rgb1 0000 rgb0 |
| 139 | // v == rgb3 rgb2 rgb1 rgb0 |
| 140 | auto v = vmovn_u32(spread); |
| 141 | return v; |
| 142 | } |
| 143 | |
| 144 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 145 | inline Sk4px Sk4px::Load4(const SkPMColor16 src[4]) { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 146 | return Sk16b(widen_to_8888(vld1_u16(src))); |
| 147 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 148 | inline Sk4px Sk4px::Load2(const SkPMColor16 src[2]) { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 149 | auto src2 = ((uint32_t)src[0] ) |
| 150 | | ((uint32_t)src[1] << 16); |
| 151 | return Sk16b(widen_to_8888(vcreate_u16(src2))); |
| 152 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 153 | inline Sk4px Sk4px::Load1(const SkPMColor16 src[1]) { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 154 | return Sk16b(widen_to_8888(vcreate_u16(src[0]))); |
| 155 | } |
| 156 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 157 | inline void Sk4px::store4(SkPMColor16 dst[4]) const { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 158 | vst1_u16(dst, narrow_to_565(this->fVec)); |
| 159 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 160 | inline void Sk4px::store2(SkPMColor16 dst[2]) const { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 161 | auto v = narrow_to_565(this->fVec); |
| 162 | dst[0] = vget_lane_u16(v, 0); |
| 163 | dst[1] = vget_lane_u16(v, 1); |
| 164 | } |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 165 | inline void Sk4px::store1(SkPMColor16 dst[1]) const { |
mtklein | ced1585 | 2015-07-22 10:52:53 -0700 | [diff] [blame] | 166 | dst[0] = vget_lane_u16(narrow_to_565(this->fVec), 0); |
| 167 | } |
| 168 | |
mtklein | 082e329 | 2015-08-12 11:56:43 -0700 | [diff] [blame] | 169 | } // namespace |
| 170 | |