/* * All Video Processing kernels * Copyright © <2010>, Intel Corporation. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the * "Software"), to deal in the Software without restriction, including * without limitation the rights to use, copy, modify, merge, publish, * distribute, sub license, and/or sell copies of the Software, and to * permit persons to whom the Software is furnished to do so, subject to * the following conditions: * * The above copyright notice and this permission notice (including the * next paragraph) shall be included in all copies or substantial portions * of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. * IN NO EVENT SHALL PRECISION INSIGHT AND/OR ITS SUPPLIERS BE LIABLE FOR * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. * * This file was originally licensed under the following license * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. * */ // Module name: RGB16x8_Save_Y416.asm // // Save packed ARGB 444 frame data block of size 16x8 // // To save 16x8 block (128x8 byte layout for ARGB 16bit per component) we need 4 send instructions // ----------------- // | 1 | 2 | 3 | 4 | // ----------------- #include "RGB16x8_Save_RGB.inc" shl (1) rMSGSRC.0<1>:d wORIX<0;1,0>:w 3:w { NoDDClr } // H. block origin need to become 8 times mov (1) rMSGSRC.1<1>:d wORIY<0;1,0>:w { NoDDClr, NoDDChk } // Block origin (1st quadrant) mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_ARGB:ud { NoDDChk } // Block width and height (32x8) mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud /* Not needed for validation kernels for now -vK //Use the mask to determine which pixels shouldn't be over-written and (1) acc0.0<1>:ud udBLOCK_MASK<0;1,0>:ud 0x00FFFFFF:ud cmp.ge.f0.0(1) dNULLREG acc0.0<0;1,0>:ud 0x00FFFFFF:ud //Check if all pixels in the block need to be modified (f0.0) jmpi WriteARGBToDataPort //If mask is not all 1's, then load the entire 64x8 block //so that only those bytes may be modified that need to be (using the mask) // Load first block 16x8 packed ARGB 444 --------------------------------------- or (1) acc0.0<1>:ud udBLOCK_MASK<0;1,0>:ud 0xFF00FF00:ud //Check first block cmp.e.f0.0 (1) dNULLREG acc0.0<0;1,0>:ud 0xFFFFFFFF:ud (f0.0) jmpi SkipFirstBlockMerge //If full mask then skip this block send (8) udSRC_ARGB(0)<1> mMSGHDR udDUMMY_NULL nDATAPORT_READ nDPMR_MSGDSC+nDPR_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud //Merge the data mov (1) f0.0:uw ubBLOCK_MASK_V:ub //Load the mask on flag reg (f0.0) mov (8) rMASK_TEMP<1>:uw uwBLOCK_MASK_H:uw //use sel instruction - vK (-f0.0) mov (8) rMASK_TEMP<1>:uw 0:uw $for(0, 0; (-f0.1) mov (8) udDEST_ARGB(%2)<1> udSRC_ARGB(%1) } SkipFirstBlockMerge: // Load second block 16x8 packed ARGB 444 --------------------------------------- or (1) acc0.0<1>:ud udBLOCK_MASK<0;1,0>:ud 0xFF0000FF:ud //Check second block cmp.e.f0.0 (1) dNULLREG acc0.0<0;1,0>:ud 0xFFFFFFFF:ud (f0.0) jmpi WriteARGBToDataPort //If full mask then skip this block add (1) mMSGHDR.0<1>:d rMSGSRC.0<0;1,0>:d 32:d // Point to 2nd part send (8) udSRC_ARGB(0)<1> mMSGHDR udDUMMY_NULL nDATAPORT_READ nDPMR_MSGDSC+nDPR_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud // Point to 1st part again //Merge the data mov (1) f0.0:uw ubBLOCK_MASK_V:ub //Load the mask on flag reg (f0.0) shr (8) rMASK_TEMP<1>:uw uwBLOCK_MASK_H:uw 8:uw //load the mask for second block (-f0.0) mov (8) rMASK_TEMP<1>:uw 0:uw $for(0, 1; (-f0.1) mov (8) udDEST_ARGB(%2)<1> udSRC_ARGB(%1) } */ WriteARGBToDataPort: // Move packed data to MRF and output //Write 1st 4X8 pixels $for(0; udDEST_ARGB(%1*4) } send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud //Write 2nd 4X8 pixels mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud add (1) mMSGHDR.0<1>:d rMSGSRC.0<0;1,0>:d 32:d // Point to 2nd part $for(0; udDEST_ARGB(%1*4+1) } send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud //Write 3rd 4X8 pixels mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud add (1) mMSGHDR.0<1>:d rMSGSRC.0<0;1,0>:d 64:d // Point to 2nd part $for(0; udDEST_ARGB(%1*4+2) } send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud //Write 4th 4X8 pixels mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud add (1) mMSGHDR.0<1>:d rMSGSRC.0<0;1,0>:d 96:d // Point to 2nd part $for(0; udDEST_ARGB(%1*4+3) } send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_ARGB+nBI_DESTINATION_RGB:ud // End of RGB16x8_Save_Y416