1 /* 2 * Copyright (c) 2019, Intel Corporation 3 * 4 * Permission is hereby granted, free of charge, to any person obtaining a 5 * copy of this software and associated documentation files (the "Software"), 6 * to deal in the Software without restriction, including without limitation 7 * the rights to use, copy, modify, merge, publish, distribute, sublicense, 8 * and/or sell copies of the Software, and to permit persons to whom the 9 * Software is furnished to do so, subject to the following conditions: 10 * 11 * The above copyright notice and this permission notice shall be included 12 * in all copies or substantial portions of the Software. 13 * 14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS 15 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 17 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR 18 * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, 19 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR 20 * OTHER DEALINGS IN THE SOFTWARE. 21 */ 22 { 23 matrix<ushort, 1, 16> Temp; 24 ushort Alpha = (ushort)((ColorFill(3)) << 8); 25 Alpha = cm_add<ushort>(Alpha, 256, SAT); 26 ushort Alpha1 = cm_add<ushort>(0xff00, - Alpha, SAT); 27 28 #pragma unroll 29 for (int i = 0; i < 4; i++) 30 { 31 /* 32 Buffer layout after shuffle 33 _________________________________________________ 34 |_______Block0__________|_______Block1__________| 35 |_______Block2__________|_______Block3__________| 36 |_______Block4__________|_______Block5__________| 37 |_______Block6__________|_______Block7__________| 38 39 Write back buffer layout correlate to the block number#, each box stands for 1 GRF 40 _______________________________________________ 41 |____R0_________R1_____|____R2_________R3_____| 42 |____G0_________G1_____|____G2_________G3_____| 43 |____B0_________B1_____|____B2_________B3_____| 44 |____A0_________A1_____|____A2_________A3_____| 45 |____R4_________R5_____|____R6_________R7_____| 46 |____G4_________G5_____|____G6_________G7_____| 47 |____B4_________B5_____|____B6_________B7_____| 48 |____A4_________A5_____|____A6_________A7_____| 49 */ 50 51 // R0/G0/B0/A0 52 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_0, 0) * Alpha)) + (((ColorFill(0) << 8 )* Alpha1))) >> 16; 53 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_0, 0).merge(Temp, (ColorFill(0) << 8), TempMask0[0][4 * i]); 54 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_0, 0) * Alpha)) + (((ColorFill(1) << 8)* Alpha1))) >> 16; 55 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_0, 0).merge(Temp, (ColorFill(1) << 8), TempMask0[0][4 * i]); 56 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_0, 0) * Alpha)) + (((ColorFill(2) << 8)* Alpha1))) >> 16; 57 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_0, 0).merge(Temp, (ColorFill(2) << 8), TempMask0[0][4 * i]); 58 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_0, 0) * Alpha)) + (((ColorFill(3) << 8)* Alpha1))) >> 16; 59 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_0, 0).merge(Temp, (ColorFill(3) << 8), TempMask0[0][4 * i]); 60 61 // R1/G1/B1/A1 62 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_1, 0) * Alpha)) + (((ColorFill(0) << 8)* Alpha1))) >> 16; 63 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_1, 0).merge(Temp, (ColorFill(0) << 8), TempMask0[0][4 * i + 1]); 64 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_1, 0) * Alpha)) + (((ColorFill(1) << 8)* Alpha1))) >> 16; 65 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_1, 0).merge(Temp, (ColorFill(1) << 8), TempMask0[0][4 * i + 1]); 66 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_1, 0) * Alpha)) + (((ColorFill(2) << 8)* Alpha1))) >> 16; 67 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_1, 0).merge(Temp, (ColorFill(2) << 8), TempMask0[0][4 * i + 1]); 68 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_1, 0) * Alpha)) + (((ColorFill(3) << 8)* Alpha1))) >> 16; 69 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_1, 0).merge(Temp, (ColorFill(3) << 8), TempMask0[0][4 * i + 1]); 70 71 // R2/G2/B2/A2 72 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_2, 0) * Alpha)) + (((ColorFill(0) << 8)* Alpha1))) >> 16; 73 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_2, 0).merge(Temp, (ColorFill(0) << 8), TempMask0[0][4 * i + 2]); 74 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_2, 0) * Alpha)) + (((ColorFill(1) << 8)* Alpha1))) >> 16; 75 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_2, 0).merge(Temp, (ColorFill(1) << 8), TempMask0[0][4 * i + 2]); 76 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_2, 0) * Alpha)) + (((ColorFill(2) << 8)* Alpha1))) >> 16; 77 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_2, 0).merge(Temp, (ColorFill(2) << 8), TempMask0[0][4 * i + 2]); 78 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_2, 0) * Alpha)) + (((ColorFill(3) << 8)* Alpha1))) >> 16; 79 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_2, 0).merge(Temp, (ColorFill(3) << 8), TempMask0[0][4 * i + 2]); 80 81 // R3/G3/B3/A3 82 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_3, 0) * Alpha)) + (((ColorFill(0) << 8)* Alpha1))) >> 16; 83 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_RV_3, 0).merge(Temp, (ColorFill(0) << 8), TempMask0[0][4 * i + 3]); 84 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_3, 0) * Alpha)) + (((ColorFill(1) << 8)* Alpha1))) >> 16; 85 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_GY_3, 0).merge(Temp, (ColorFill(1) << 8), TempMask0[0][4 * i + 3]); 86 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_3, 0) * Alpha)) + (((ColorFill(2) << 8)* Alpha1))) >> 16; 87 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_BU_3, 0).merge(Temp, (ColorFill(2) << 8), TempMask0[0][4 * i + 3]); 88 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_3, 0) * Alpha)) + (((ColorFill(3) << 8)* Alpha1))) >> 16; 89 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(16 * i + Channel_Offset_A_3, 0).merge(Temp, (ColorFill(3) << 8), TempMask0[0][4 * i + 3]); 90 } 91 92 TempMask0 = 0xFFFFFFFF; 93 }