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 uchar Buffer_Index1 = (Buffer_Index & 0x0f) << 4; 24 uchar Buffer_Index2 = (Buffer_Index >> 4) << 4; 25 26 /* 27 Buffer layout after shuffle 28 _________________________________________________ 29 |_______Block0__________|_______Block1__________| 30 |_______Block2__________|_______Block3__________| 31 |_______Block4__________|_______Block5__________| 32 |_______Block6__________|_______Block7__________| 33 34 Write back buffer layout correlate to the block number#, each box stands for 1 GRF 35 _______________________________________________ 36 |____R0_________R1_____|____R2_________R3_____| 37 |____G0_________G1_____|____G2_________G3_____| 38 |____B0_________B1_____|____B2_________B3_____| 39 |____A0_________A1_____|____A2_________A3_____| 40 |____R4_________R5_____|____R6_________R7_____| 41 |____G4_________G5_____|____G6_________G7_____| 42 |____B4_________B5_____|____B6_________B7_____| 43 |____A4_________A5_____|____A6_________A7_____| 44 */ 45 { 46 matrix<ushort, 1, 16> Temp; 47 { 48 // R1/G1/B1/A1 49 matrix_ref<ushort, 1, 16> Alpha = DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + 6, 0); 50 51 { 52 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_0, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_0, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 53 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_0, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_0, 0), TempMask[CalculationMask]); 54 } 55 56 { 57 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_0, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_0, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 58 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_0, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_0, 0), TempMask[CalculationMask]); 59 } 60 61 { 62 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_0, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_0, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 63 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_0, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_0, 0), TempMask[CalculationMask]); 64 } 65 } 66 67 { 68 // R2/G2/B2/A2 69 matrix_ref<ushort, 1, 16> Alpha = DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + 7, 0); 70 71 { 72 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_1, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_1, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 73 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_1, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_1, 0), TempMask[CalculationMask + 1]); 74 } 75 76 { 77 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_1, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_1, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 78 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_1, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_1, 0), TempMask[CalculationMask + 1]); 79 } 80 81 { 82 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_1, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_1, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 83 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_1, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_1, 0), TempMask[CalculationMask + 1]); 84 } 85 } 86 87 { 88 // R3/G3/B3/A3 89 matrix_ref<ushort, 1, 16> Alpha = DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + 14, 0); 90 91 { 92 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_2, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_2, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 93 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_2, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_2, 0), TempMask[CalculationMask + 2]); 94 } 95 96 { 97 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_2, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_2, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 98 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_2, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_2, 0), TempMask[CalculationMask + 2]); 99 } 100 101 { 102 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_2, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_2, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 103 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_2, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_2, 0), TempMask[CalculationMask + 2]); 104 } 105 } 106 107 { 108 // R4/G4/B4/A4 109 matrix_ref<ushort, 1, 16> Alpha = DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + 15, 0); 110 111 { 112 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_3, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_3, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 113 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_3, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_3, 0), TempMask[CalculationMask + 3]); 114 } 115 116 { 117 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_3, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_3, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 118 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_3, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_3, 0), TempMask[CalculationMask + 3]); 119 } 120 121 { 122 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_3, 0) * (cm_add<ushort>(0xFF00, -Alpha, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_3, 0) * (cm_add<ushort>(Alpha, 256, SAT))))) >> 16; 123 DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_3, 0).merge(Temp, DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_3, 0), TempMask[CalculationMask + 3]); 124 } 125 } 126 } 127 128 }