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 /* 24 Buffer layout after shuffle 25 _________________________________________________ 26 |_______Block0__________|_______Block1__________| 27 |_______Block2__________|_______Block3__________| 28 |_______Block4__________|_______Block5__________| 29 |_______Block6__________|_______Block7__________| 30 31 Write back buffer layout correlate to the block number#, each box stands for 1 GRF 32 _______________________________________________ 33 |____R0_________R1_____|____R2_________R3_____| 34 |____G0_________G1_____|____G2_________G3_____| 35 |____B0_________B1_____|____B2_________B3_____| 36 |____A0_________A1_____|____A2_________A3_____| 37 |____R4_________R5_____|____R6_________R7_____| 38 |____G4_________G5_____|____G6_________G7_____| 39 |____B4_________B5_____|____B6_________B7_____| 40 |____A4_________A5_____|____A6_________A7_____| 41 */ 42 43 uchar Buffer_Index1 = (Buffer_Index & 0x0f) << 4; 44 uchar Buffer_Index2 = (Buffer_Index >> 4) << 4; 45 46 { 47 matrix<ushort, 1, 16> Temp; 48 49 { 50 // R1/G1/B1/A1 51 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_0, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_0, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 52 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]); 53 54 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_0, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_0, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 55 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]); 56 57 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_0, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_0, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 58 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]); 59 } 60 61 { 62 // R2/G2/B2/A2 63 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_1, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_1, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 64 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]); 65 66 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_1, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_1, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 67 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]); 68 69 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_1, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_1, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 70 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]); 71 } 72 73 { 74 // R3/G3/B3/A3 75 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_2, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_2, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 76 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]); 77 78 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_2, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_2, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 79 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]); 80 81 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_2, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_2, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 82 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]); 83 } 84 85 { 86 // R4/G4/B4/A4 87 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_RV_3, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_RV_3, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 88 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]); 89 90 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_GY_3, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_GY_3, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 91 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]); 92 93 Temp = (((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index1 + Channel_Offset_BU_3, 0) * (cm_add<ushort>(0xFF, -ConstAlphaTemp, SAT)))) + ((DataBuffer.format<ushort, 96, 16>().select<1, 1, 16, 1>(Buffer_Index2 + Channel_Offset_BU_3, 0) * (cm_add<uchar>(ConstAlphaTemp, 1, SAT))))) >> 8; 94 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]); 95 } 96 } 97 }