Debian Codegraph @

src/shaders/post_processing/gen5_6/Common/RGBX_Save_YUV_Float.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 * Authors:
40 * Halley Zhao <[email protected]>
41 */
42
43 // Module name: RGBX_Save_YUV_Float.asm
44 //----------------------------------------------------------------
45
46 #include "RGBX_Load_16x8.inc"
47
48 #if (0)
49 // 8 grf reg for one row of pixel (2 pixel per grf)
50 #define nTEMP0 34
51 #define nTEMP1 35
52 #define nTEMP2 36
53 #define nTEMP3 37
54 #define nTEMP4 38
55 #define nTEMP5 39
56 #define nTEMP6 40
57 #define nTEMP7 41
58
59 #define nTEMP8 42 // transformation coefficient
60 #define nTEMP10 44 // transformation coefficient
61
62 #define nTEMP12 46 // save Y/U/V in ub format
63 #define nTEMP14 48 // save YUV in ud format
64 #define nTEMP16 50 // dp4 result
65 #define nTEMP17 51
66 #define nTEMP18 52
67
68 #define nTEMP24 58
69 #endif
70
71 $for(0; <nY_NUM_OF_ROWS; 1) {
72 // BGRX | B | G | R | X |
73 // ###### save one row of pixel to temp grf with float format (required by dp4)
74 // mov (8) doesn't work, puzzle
75 mov (4) REG(r, nTEMP0)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 0]<4,1>:ub
76 mov (4) REG(r, nTEMP1)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 8]<4,1>:ub
77 mov (4) REG(r, nTEMP2)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 16]<4,1>:ub
78 mov (4) REG(r, nTEMP3)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 24]<4,1>:ub
79 mov (4) REG(r, nTEMP4)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 0]<4,1>:ub
80 mov (4) REG(r, nTEMP5)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 8]<4,1>:ub
81 mov (4) REG(r, nTEMP6)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 16]<4,1>:ub
82 mov (4) REG(r, nTEMP7)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 24]<4,1>:ub
83 mov (4) REG2(r, nTEMP0, 4)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 4]<4,1>:ub
84 mov (4) REG2(r, nTEMP1, 4)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 12]<4,1>:ub
85 mov (4) REG2(r, nTEMP2, 4)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 20]<4,1>:ub
86 mov (4) REG2(r, nTEMP3, 4)<1>:f r[SRC_RGBA_OFFSET_1,%1*32 + 28]<4,1>:ub
87 mov (4) REG2(r, nTEMP4, 4)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 4]<4,1>:ub
88 mov (4) REG2(r, nTEMP5, 4)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 12]<4,1>:ub
89 mov (4) REG2(r, nTEMP6, 4)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 20]<4,1>:ub
90 mov (4) REG2(r, nTEMP7, 4)<1>:f r[SRC_RGBA_OFFSET_2,%1*32 + 24]<4,1>:ub
91
92 // ###### do one row for Y
93 // ##### dp4(nTEMP16) and save result to uw format(nTEMP12)
94 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(0, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
95 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
96 mov (2) REG2(r, nTEMP12, 0)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
97 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(1, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
98 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
99 mov (2) REG2(r, nTEMP12, 2)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
100 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(2, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
101 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
102 mov (2) REG2(r, nTEMP12, 4)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
103 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(3, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
104 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
105 mov (2) REG2(r, nTEMP12, 6)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
106 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(4, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
107 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
108 mov (2) REG2(r, nTEMP12, 8)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
109 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(5, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
110 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
111 mov (2) REG2(r, nTEMP12, 10)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
112 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(6, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
113 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
114 mov (2) REG2(r, nTEMP12, 12)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
115 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(7, 0)<0;8,1> fRGB_to_Y_Coef_Float<0;4,1>:f
116 mov (2) REG2(r, nTEMP14, 0)<1>:ud REG2(r, nTEMP16, 0)<0;2,4>:f
117 mov (2) REG2(r, nTEMP12, 14)<1>:ub REG2(r, nTEMP14, 0)<0;2,4>:ub
118
119 // #### write Y to the 1 row
120 mov (16) uwDEST_Y(%1)<1> REG2(r,nTEMP12, 0)<0;16,1>:ub
121
122 // ###### do one row for U
123 // ##### dp4(nTEMP16) and save result to uw format(nTEMP12)
124 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(0, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
125 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
126 mov (2) REG2(r, nTEMP12, 0)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
127 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(1, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
128 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
129 mov (2) REG2(r, nTEMP12, 2)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
130 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(2, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
131 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
132 mov (2) REG2(r, nTEMP12, 4)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
133 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(3, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
134 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
135 mov (2) REG2(r, nTEMP12, 6)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
136 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(4, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
137 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
138 mov (2) REG2(r, nTEMP12, 8)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
139 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(5, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
140 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
141 mov (2) REG2(r, nTEMP12, 10)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
142 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(6, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
143 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
144 mov (2) REG2(r, nTEMP12, 12)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
145 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(7, 0)<0;8,1> fRGB_to_U_Coef_Float<0;4,1>:f
146 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
147 mov (2) REG2(r, nTEMP12, 14)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
148 add (16) REG2(r, nTEMP12, 0)<1>:w REG2(r, nTEMP12, 0)<0;16,1>:w 128:w
149 // #### write U to the 1 row
150 mov (16) uwDEST_U(%1)<1> REG2(r,nTEMP12, 0)<0;16,2>:ub
151
152 // ###### do one row for V
153 // ##### dp4(nTEMP16) and save result to uw format(nTEMP12)
154 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(0, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
155 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
156 mov (2) REG2(r, nTEMP12, 0)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
157 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(1, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
158 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
159 mov (2) REG2(r, nTEMP12, 2)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
160 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(2, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
161 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
162 mov (2) REG2(r, nTEMP12, 4)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
163 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(3, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
164 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
165 mov (2) REG2(r, nTEMP12, 6)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
166 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(4, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
167 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
168 mov (2) REG2(r, nTEMP12, 8)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
169 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(5, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
170 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
171 mov (2) REG2(r, nTEMP12, 10)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
172 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(6, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
173 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
174 mov (2) REG2(r, nTEMP12, 12)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
175 dp4 (8) REG2(r, nTEMP16, 0)<1>:f fROW_BGRX(7, 0)<0;8,1> fRGB_to_V_Coef_Float<0;4,1>:f
176 mov (2) REG2(r, nTEMP14, 0)<1>:d REG2(r, nTEMP16, 0)<0;2,4>:f
177 mov (2) REG2(r, nTEMP12, 14)<1>:w REG2(r, nTEMP14, 0)<0;2,2>:w
178 add (16) REG2(r, nTEMP12, 0)<1>:w REG2(r, nTEMP12, 0)<0;16,1>:w 128:w
179
180 // #### write V to the 1 row
181 mov (16) uwDEST_V(%1)<1> REG2(r,nTEMP12, 0)<0;16,2>:ub
182 }