Debian Codegraph @

src/shaders/post_processing/gen5_6/Common/PL8x4_Save_IMC3.asm

raw
1 /*
2 * All Video Processing kernels
3 * Copyright © <2010>, Intel Corporation.
4 *
5 * Permission is hereby granted, free of charge, to any person obtaining a
6 * copy of this software and associated documentation files (the
7 * "Software"), to deal in the Software without restriction, including
8 * without limitation the rights to use, copy, modify, merge, publish,
9 * distribute, sub license, and/or sell copies of the Software, and to
10 * permit persons to whom the Software is furnished to do so, subject to
11 * the following conditions:
12 *
13 * The above copyright notice and this permission notice (including the
14 * next paragraph) shall be included in all copies or substantial portions
15 * of the Software.
16 *
17 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
18 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
19 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
20 * IN NO EVENT SHALL PRECISION INSIGHT AND/OR ITS SUPPLIERS BE LIABLE FOR
21 * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
22 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
23 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
24 *
25 * This file was originally licensed under the following license
26 *
27 * Licensed under the Apache License, Version 2.0 (the "License");
28 * you may not use this file except in compliance with the License.
29 * You may obtain a copy of the License at
30 *
31 * http://www.apache.org/licenses/LICENSE-2.0
32 *
33 * Unless required by applicable law or agreed to in writing, software
34 * distributed under the License is distributed on an "AS IS" BASIS,
35 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
36 * See the License for the specific language governing permissions and
37 * limitations under the License.
38 *
39 */
40
41 // Module name: PL8x4_Save_IMC3.asm
42 //
43 // Save planar YUV420 frame data block of size 16x8
44
45 #include "PL8x4_Save_IMC3.inc"
46
47 //Use the mask to determine which pixels shouldn't be over-written
48 and (1) acc0.0<1>:ud udBLOCK_MASK<0;1,0>:ud 0x00FFFFFF:ud
49 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
50 (f0.0) jmpi WritePlanarToDataPort
51
52 //If mask is not all 1's, then load the entire 16x8 block
53 //so that only those bytes may be modified that need to be (using the mask)
54
55 // Load 16x8 planar Y ----------------------------------------------------------
56 mov (2) rMSGSRC.0<1>:d wORIX<2;2,1>:w // Block origin
57 mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_Y:ud // Block width and height (16x8)
58 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
59 send (8) udSRC_Y(0)<1> mMSGHDR udDUMMY_NULL nDATAPORT_READ nDPMR_MSGDSC+nDPR_MSG_SIZE_Y+nBI_DESTINATION_Y:ud
60 // Load 8x4 planar U and V -----------------------------------------------------
61 asr (2) rMSGSRC.0<1>:d wORIX<2;2,1>:w 1:w // U/V block origin should be half of Y's
62 mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_UV:ud // Block width and height (8x4)
63 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
64 send (8) udSRC_U(0)<1> mMSGHDR udDUMMY_NULL nDATAPORT_READ nDPMR_MSGDSC+nDPR_MSG_SIZE_UV+nBI_DESTINATION_U:ud
65 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
66 send (8) udSRC_V(0)<1> mMSGHDR udDUMMY_NULL nDATAPORT_READ nDPMR_MSGDSC+nDPR_MSG_SIZE_UV+nBI_DESTINATION_V:ud
67
68 //expand U and V to be aligned on word boundary - Y remains in bytes
69 $for (nUV_NUM_OF_ROWS/2-1; >-1; -1) {
70 mov (16) uwSRC_U(0, %1*16)<1> ubSRC_U(0, %1*16)
71 mov (16) uwSRC_V(0, %1*16)<1> ubSRC_V(0, %1*16)
72 }
73
74 //Merge the data
75 mov (1) f0.0:uw ubBLOCK_MASK_V:ub //Load the mask on flag reg
76 (f0.0) mov (8) rMASK_TEMP<1>:uw uwBLOCK_MASK_H:uw
77 (-f0.0) mov (8) rMASK_TEMP<1>:uw 0:uw
78
79 // Destination is Word aligned
80 $for(0; <nY_NUM_OF_ROWS; 2) {
81 mov (1) f0.1:uw uwMASK_TEMP(0,%1)<0;1,0>
82 (-f0.1) mov (16) ub2DEST_Y(0, %1*32)<2> ubSRC_Y(0, %1*16)
83 (-f0.1) mov (16) ub2DEST_U(0, %1*8)<1> ubSRC_U(0, %1*8) //only works for Word aligned Byte data
84 (-f0.1) mov (16) ub2DEST_V(0, %1*8)<1> ubSRC_V(0, %1*8) //only works for Word aligned Byte data
85
86 mov (1) f0.1:uw uwMASK_TEMP(0,1+%1)<0;1,0>
87 (-f0.1) mov (16) ub2DEST_Y(0, (1+%1)*32)<2> ubSRC_Y(0, (1+%1)*16)
88 }
89
90 WritePlanarToDataPort:
91 #if !defined(SAVE_UV_ONLY)
92 // Save current planar frame Y block data (16x8) -------------------------------
93 mov (2) rMSGSRC.0<1>:d wORIX<2;2,1>:w // Block origin
94 mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_Y:ud // Block width and height (16x8)
95 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
96 $for(0,0; <nY_NUM_OF_ROWS; 2,1) {
97 mov(16) mubMSGPAYLOAD(%2,0)<1> ub2DEST_Y(%1)REGION(16,2)
98 mov(16) mubMSGPAYLOAD(%2,16)<1> ub2DEST_Y(%1+1)REGION(16,2)
99 }
100 send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_Y+nBI_DESTINATION_Y:ud
101 #endif
102 // Save U/V data block in planar format (8x4) ----------------------------------
103 asr (2) rMSGSRC.0<1>:d wORIX<2;2,1>:w 1:w // U/V block origin should be half of Y's
104 mov (1) rMSGSRC.2<1>:ud nDPW_BLOCK_SIZE_UV:ud // Block width and height (8x4)
105 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
106
107 // Save U picture data ---------------------------------------------------------
108 mov (16) mubMSGPAYLOAD(0,0)<1> ub2DEST_U(0)REGION(16,2) // U rows 0,1
109 mov (16) mubMSGPAYLOAD(0,16)<1> ub2DEST_U(1)REGION(16,2) // U rows 2,3
110 send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_UV+nBI_DESTINATION_U:ud
111 mov (8) mMSGHDR<1>:ud rMSGSRC<8;8,1>:ud
112
113 // Save V picture data ---------------------------------------------------------
114 mov (16) mubMSGPAYLOAD(0,0)<1> ub2DEST_V(0)REGION(16,2) // V rows 0,1
115 mov (16) mubMSGPAYLOAD(0,16)<1> ub2DEST_V(1)REGION(16,2) // V rows 2,3
116 send (8) dNULLREG mMSGHDR udDUMMY_NULL nDATAPORT_WRITE nDPMW_MSGDSC+nDPW_MSG_SIZE_UV+nBI_DESTINATION_V:ud
117
118 // End of PL8x4_Save_IMC3