Debian Codegraph @

src/shaders/post_processing/gen7/RGB_to_YUV.g4a

raw
1 /*
2 * Copyright 2000-2011 Intel Corporation All Rights Reserved
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the
6 * "Software"), to deal in the Software without restriction, including
7 * without limitation the rights to use, copy, modify, merge, publish,
8 * distribute, sub license, and/or sell copies of the Software, and to
9 * permit persons to whom the Software is furnished to do so, subject to
10 * the following conditions:
11 *
12 * The above copyright notice and this permission notice (including the
13 * next paragraph) shall be included in all copies or substantial portions
14 * of the Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
17 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
18 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
19 * IN NO EVENT SHALL PRECISION INSIGHT AND/OR ITS SUPPLIERS BE LIABLE FOR
20 * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
21 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
22 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
23 *
24 * This file was originally licensed under the following license
25 *
26 * Licensed under the Apache License, Version 2.0 (the "License");
27 * you may not use this file except in compliance with the License.
28 * You may obtain a copy of the License at
29 *
30 * http://www.apache.org/licenses/LICENSE-2.0
31 *
32 * Unless required by applicable law or agreed to in writing, software
33 * distributed under the License is distributed on an "AS IS" BASIS,
34 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
35 * See the License for the specific language governing permissions and
36 * limitations under the License.
37 * Authors:
38 * Zhao Yakui <[email protected]>
39 */
40
41
42
43 // Module name: common.inc
44 //
45 // Common header file for all Video-Processing kernels
46 //
47
48 .default_execution_size (16)
49 .default_register_type :ub
50
51 .reg_count_total 128
52 .reg_count_payload 7
53
54 //========== Common constants ==========
55
56
57 //========== Macros ==========
58
59
60 //Fast Jump, For more details see "Set_Layer_N.asm"
61
62
63 //========== Defines ====================
64
65 //========== Static Parameters (Common To All) ==========
66 //r1
67
68
69 //r2
70
71 // e.g. byte0 byte1 byte2
72 // YUYV 0 1 3
73 // YVYU 0 3 1
74
75 //Color Pipe (IECP) parameters
76
77
78 //ByteCopy
79
80
81 //r4
82
83 // e.g. byte0 byte1 byte2
84 // YUYV 0 1 3
85 // YVYU 0 3 1
86
87
88 //========== Inline parameters (Common To All) ===========
89
90
91 //============== Binding Index Table===========
92 //Common between DNDI and DNUV
93
94
95 //================= Common Message Descriptor =====
96 // Message descriptor for thread spawning
97 // Message Descriptors
98 // = 000 0001 (min message len 1 ) 0,0000 (resp len 0 -add later)
99 // 0000,0000,0000
100 // 0001(Spawn a root thread),0001 (Root thread spawn thread)
101 // = 0x02000011
102 // Thread Spawner Message Descriptor
103
104
105 // Message descriptor for atomic operation add
106 // Message Descriptors
107 // = 000 0110 (min message len 6 ) 0,0000 (resp len 0 -add later)
108 // 1(header present)001,10(typed atomic operation)0(return enabled)0(slot group, low 8 bits),0111 (AOP_Add)
109 // 0000,0000 (Binding table index, added later)
110 // = 0x02000011
111
112 // Atomic Operation Add Message Descriptor
113
114
115 // Message descriptor for dataport media write
116 // Message Descriptors
117 // = 000 0001 (min message len 1 - add later) 00000 (resp len 0)
118 // 1 (header present 1) 0 1010 (media block write) 000000
119 // 00000000 (binding table index - set later)
120 // = 0x020A8000
121
122
123 // Message Length defines
124
125
126 // Response Length defines
127
128
129 // Block Width and Height Size defines
130
131
132 // Extended Message Descriptors
133
134
135 // Common message descriptors:
136
137
138 //===================== Math Function Control ===================================
139
140
141 //============ Message Registers ===============
142 // buf4 starts from r28
143
144
145 //#define mMSGHDR_EOT r43 // Dummy Message Register for EOT
146
147
148 .declare mubMSGPAYLOAD Base=r30 ElementSize=1 SrcRegion=<16;16,1> Type=ub
149 .declare muwMSGPAYLOAD Base=r30 ElementSize=2 SrcRegion=<16;16,1> Type=uw
150 .declare mudMSGPAYLOAD Base=r30 ElementSize=4 SrcRegion=<8;8,1> Type=ud
151 .declare mfMSGPAYLOAD Base=r30 ElementSize=4 SrcRegion=<8;8,1> Type=f
152
153 //=================== End of thread instruction ===========================
154
155
156 //=====================Pointers Used=====================================
157
158
159 //=======================================================================
160
161
162 //r11-r17
163 // Define temp space for any usages
164
165
166 // Common Buffers
167
168
169 // temp space for rotation
170
171 .declare fROBUF Base=r11.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=f
172
173 .declare udROBUF Base=r11.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=ud
174
175 .declare uwROBUF Base=r11.0 ElementSize=2 SrcRegion=<16;16,1> DstRegion=<1> Type=uw
176
177 .declare ubROBUF Base=r11.0 ElementSize=1 SrcRegion=<16;16,1> DstRegion=<1> Type=ub
178
179 .declare ub4ROBUF Base=r11.0 ElementSize=1 SrcRegion=<32;8,4> DstRegion=<4> Type=ub
180
181
182 // End of common.inc
183
184
185 // Module name: YUV_to_RGB.asm
186 //
187 // Convert YUV to RGB, handle it by 16x4 block
188 //
189
190
191 // Description: Includes all definitions explicit to Fast Composite.
192
193
194 //========== GRF partition ==========
195 // r0 header : r0 (1 GRF)
196 // Static parameters : r1 - r6 (6 GRFS)
197 // Inline parameters : r7 - r8 (2 GRFs)
198 // MSGSRC : r27 (1 GRF)
199 //===================================
200
201 //Interface:
202 //========== Static Parameters (Explicit To Fast Composite) ==========
203 //r1
204 //CSC Set 0
205
206
207 .declare udCSC_CURBE Base=r1.0 ElementSize=4 Type=ud
208
209 //Constant alpha
210
211
212 //r2
213
214
215 // Gen7 AVS WA
216
217
218 // WiDi Definitions
219
220
221 //Colorfill
222
223
224 // 0: 0-degree, 1: 90, 2: 180, 3: 270-degree, clockwise.
225
226 .declare ubCOLOR_PIXEL_VAL Base=r2.20 ElementSize=1 SrcRegion=<0;1,0> DstRegion=<1> Type=ub
227
228 //r3
229 //Normalised Ratio of Horizontal step size with main video for all layers
230
231
232 //Normalised Ratio of Horizontal step size with main video for all layers becomes
233 //Normalised Horizontal step size for all layers in VP_Setup.asm
234
235
236 //r4
237 //Normalised Vertical step size for all layers
238
239
240 //r5
241 //Normalised Vertical Frame Origin for all layers
242
243
244 //r6
245 //Normalised Horizontal Frame Origin for all layers
246
247
248 //========== Inline Parameters (Explicit To Fast Composite) ==========
249
250
251 //Main video Step X
252
253
254 //====================== Binding table (Explicit To Fast Composite)=========================================
255
256
257 //Used by Interlaced Scaling Kernels
258
259
260 //========== Sampler State Table Index (Explicit To Fast Composite)==========
261 //Sampler Index for AVS/IEF messages
262
263
264 //Sampler Index for SIMD16 sampler messages
265
266
267 //=============================================================================
268
269 .declare fBUFFER_0 Base=r64.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=f
270 .declare fBUFFER_1 Base=r80.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=f
271 .declare fBUFFER_2 Base=r96.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=f
272 .declare fBUFFER_3 Base=r112.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=f
273
274 .declare bBUFFER_0 Base=r64.0 ElementSize=1 SrcRegion=<16;16,1> DstRegion=<1> Type=ub
275 .declare bBUFFER_1 Base=r80.0 ElementSize=1 SrcRegion=<16;16,1> DstRegion=<1> Type=ub
276 .declare bBUFFER_2 Base=r96.0 ElementSize=1 SrcRegion=<16;16,1> DstRegion=<1> Type=ub
277 .declare bBUFFER_3 Base=r112.0 ElementSize=1 SrcRegion=<16;16,1> DstRegion=<1> Type=ub
278
279 .declare uwBUFFER_0 Base=r64.0 ElementSize=2 SrcRegion=<16;16,1> DstRegion=<1> Type=uw
280 .declare uwBUFFER_1 Base=r80.0 ElementSize=2 SrcRegion=<16;16,1> DstRegion=<1> Type=uw
281 .declare uwBUFFER_2 Base=r96.0 ElementSize=2 SrcRegion=<16;16,1> DstRegion=<1> Type=uw
282 .declare uwBUFFER_3 Base=r112.0 ElementSize=2 SrcRegion=<16;16,1> DstRegion=<1> Type=uw
283 //Pointer to mask reg
284
285 .declare udBUFFER_0 Base=r64.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=ud
286 .declare udBUFFER_1 Base=r80.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=ud
287 .declare udBUFFER_2 Base=r96.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=ud
288 .declare udBUFFER_3 Base=r112.0 ElementSize=4 SrcRegion=<8;8,1> DstRegion=<1> Type=ud
289 //r18
290
291
292 //Always keep Cannel Pointers and Offsets in same GRF, so that we can use
293 // NODDCLR, NODDCHK flags. -rT
294
295
296 .declare udCSC_COEFF_0 Base=r18.0 ElementSize=4 Type=ud // 1 GRF
297
298 //r19
299
300
301 .declare udCSC_COEFF_1 Base=r19.0 ElementSize=4 Type=ud // 1 GRF
302
303
304 //r20
305
306 .declare uwALPHA_MASK_REG_TEMP Base=r20.0 ElementSize=2 SrcRegion=<16;16,1> Type=uw // 1 GRF
307
308 //r21
309
310 .declare uwALPHA_MASK_REG Base=r21.0 ElementSize=2 SrcRegion=<16;16,1> Type=uw // 1 GRF
311
312 //r22
313
314
315 //Always keep Cannel Pointers and Offsets in same GRF, so that we can use
316 // NODDCLR, NODDCHK flags. -rT
317
318
319 //Keep fORIGIN_X_NLAS, fY_OFFSET_2ND_BLOCK, fSTEP_X_NLAS, pMSGDSC_COPY, ubCONST_ALPHA_COPY as
320 //sub registers of same GRF to enable using NODDCLR NODDCHK. -rT
321
322 //r23
323
324
325 //Lumakey
326
327
328 //r24
329
330
331 //r25
332
333
334 //r26
335
336
337 //defines to generate LABELS during compile time.
338
339
340 //Msg payload buffers; upto 4 full-size messages can be written
341
342 //Unnecessary to use the MSGPayLoad, So it is temporiarily used for conversion of YUV->RGB
343
344 .declare fBUFFER_R Base=r28.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
345 .declare fBUFFER_G Base=r30.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
346 .declare fBUFFER_B Base=r32.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
347
348 .declare fBUFFER_Y Base=r36.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
349 .declare fBUFFER_U Base=r38.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
350 .declare fBUFFER_V Base=r40.0 ElementSize=4 SrcRegion=<8;8,1> Type=f
351
352 .declare wTempY Base=r42.0 ElementSize=2 Type=w
353 .declare wTempU Base=r44.0 ElementSize=2 Type=w
354 .declare wTempV Base=r46.0 ElementSize=2 Type=w
355
356 .declare ubTempY Base=r42.0 ElementSize=1 Type=ub
357 .declare ubTempU Base=r44.0 ElementSize=1 Type=ub
358 .declare ubTempV Base=r46.0 ElementSize=1 Type=ub
359
360 // the r17 register (nTEMP0) is originally defined from "Common.inc"
361 // instead of re-defining a nTEMP0 here, we use "SAVE_RGB" suffix for its naming
362
363 .declare uwTemp0 Base=r17.0 ElementSize=2 Type=uw
364
365
366 //_SAVE_INC_
367 // ITU-R conversion, Now we are using ITU-R conversion
368 // Y = 0.299R + 0.587G + 0.114B
369 // U = -0.169R - 0.331G + 0.499B + 128
370 // V = 0.499R - 0.418G - 0.0813B+ 128
371
372 // At the save module we have all 8 address sub-registers available.
373 // So we will use PING-PONG type of scheme to save the data using
374 // pointers pBUF_CHNL_TOP_8x4 and pBUF_CHNL_BOT_8x4. This will help
375 // reduce dependency. - rT
376
377 //wBUFF_CHNL_PTR points to either buffer 0 or buffer 4.
378 //Add appropriate offsets to get pointers for all buffers (1,2,3 or 5).
379 //Offsets are zero for buffer 0 and buffer 4.
380 //It always uses the YUVA layout.
381 //for BUFFER_0
382 mov (4) a0.0<1>:uw r22.0<4;4,1>:uw
383 mov (4) a0.4<1>:uw r22.0<4;4,1>:uw
384 // YUV uses the a0.5,a0.6 and a0.4 as the indirect-register
385 // Y = a0.5, U=a0.6, V=a0.4
386 // if channel swap?
387 // This means that it should be BGRX(B is the LSB) or RGBX
388 // 1 means that it is BGRX.
389 and.nz.f0.0 null<1>:w r2.0<0;1,0>:uw 0x01:w
390 // pointer swap
391 (f0.0) mov (1) uwTemp0<1> a0.0:uw
392 (f0.0) mov (1) a0.0:uw a0.1:uw
393 (f0.0) mov (1) a0.1:uw uwTemp0<0;1,0>
394
395 //the first line in the block 0
396 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 1]<16;8,2>:ub
397 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 17]<16;8,2>:ub
398 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 1]<16;8,2>:ub
399 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 17]<16;8,2>:ub
400 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 1]<16;8,2>:ub
401 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 17]<16;8,2>:ub
402
403 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
404 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
405 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
406
407 mov (16) acc0.0<1>:f 128.0f
408 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
409 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
410 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
411
412 mov (16) acc0.0<1>:f 128.0f
413 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
414 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
415 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
416
417 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
418 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
419 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
420
421 mov (16) r[a0.5, 0]<1>:uw 0:uw
422 mov (16) r[a0.6, 0]<1>:uw 0:uw
423 mov (16) r[a0.4, 0]<1>:uw 0:uw
424 mov (16) r[a0.5,1]<2>:ub ubTempY(0, 0)<32;8,4>
425 mov (16) r[a0.6,1]<2>:ub ubTempU(0, 0)<32;8,4>
426 mov (16) r[a0.4,1]<2>:ub ubTempV(0, 0)<32;8,4>
427
428
429 //the second line in the block 0
430 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 33]<16;8,2>:ub
431 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 49]<16;8,2>:ub
432 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 33]<16;8,2>:ub
433 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 49]<16;8,2>:ub
434 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 33]<16;8,2>:ub
435 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 49]<16;8,2>:ub
436
437 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
438 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
439 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
440
441 mov (16) acc0.0<1>:f 128.0f
442 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
443 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
444 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
445
446 mov (16) acc0.0<1>:f 128.0f
447 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
448 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
449 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
450
451 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
452 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
453 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
454
455 mov (16) r[a0.5, 32]<1>:uw 0:uw
456 mov (16) r[a0.6, 32]<1>:uw 0:uw
457 mov (16) r[a0.4, 32]<1>:uw 0:uw
458 mov (16) r[a0.5,33]<2>:ub ubTempY(0, 0)<32;8,4>
459 mov (16) r[a0.6,33]<2>:ub ubTempU(0, 0)<32;8,4>
460 mov (16) r[a0.4,33]<2>:ub ubTempV(0, 0)<32;8,4>
461
462 //the third line in the block 0
463 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 65]<16;8,2>:ub
464 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 81]<16;8,2>:ub
465 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 65]<16;8,2>:ub
466 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 81]<16;8,2>:ub
467 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 65]<16;8,2>:ub
468 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 81]<16;8,2>:ub
469
470 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
471 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
472 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
473
474 mov (16) acc0.0<1>:f 128.0f
475 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
476 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
477 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
478
479 mov (16) acc0.0<1>:f 128.0f
480 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
481 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
482 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
483
484 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
485 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
486 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
487
488 mov (16) r[a0.5, 64]<1>:uw 0:uw
489 mov (16) r[a0.6, 64]<1>:uw 0:uw
490 mov (16) r[a0.4, 64]<1>:uw 0:uw
491 mov (16) r[a0.5,65]<2>:ub ubTempY(0, 0)<32;8,4>
492 mov (16) r[a0.6,65]<2>:ub ubTempU(0, 0)<32;8,4>
493 mov (16) r[a0.4,65]<2>:ub ubTempV(0, 0)<32;8,4>
494
495 //the fourth line in the block 0
496 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 97]<16;8,2>:ub
497 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 113]<16;8,2>:ub
498 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 97]<16;8,2>:ub
499 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 113]<16;8,2>:ub
500 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 97]<16;8,2>:ub
501 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 113]<16;8,2>:ub
502
503 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
504 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
505 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
506
507 mov (16) acc0.0<1>:f 128.0f
508 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
509 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
510 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
511
512 mov (16) acc0.0<1>:f 128.0f
513 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
514 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
515 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
516
517 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
518 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
519 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
520
521 mov (16) r[a0.5, 96]<1>:uw 0:uw
522 mov (16) r[a0.6, 96]<1>:uw 0:uw
523 mov (16) r[a0.4, 96]<1>:uw 0:uw
524 mov (16) r[a0.5,97]<2>:ub ubTempY(0, 0)<32;8,4>
525 mov (16) r[a0.6,97]<2>:ub ubTempU(0, 0)<32;8,4>
526 mov (16) r[a0.4,97]<2>:ub ubTempV(0, 0)<32;8,4>
527
528 //for Buffer_1
529
530 add (8) a0.0<1>:uw a0.0<8;8,1>:uw 512:uw
531 //the first line in the block 1
532 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 1]<16;8,2>:ub
533 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 17]<16;8,2>:ub
534 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 1]<16;8,2>:ub
535 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 17]<16;8,2>:ub
536 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 1]<16;8,2>:ub
537 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 17]<16;8,2>:ub
538
539 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
540 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
541 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
542
543 mov (16) acc0.0<1>:f 128.0f
544 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
545 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
546 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
547
548 mov (16) acc0.0<1>:f 128.0f
549 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
550 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
551 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
552
553 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
554 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
555 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
556
557 mov (16) r[a0.5, 0]<1>:uw 0:uw
558 mov (16) r[a0.6, 0]<1>:uw 0:uw
559 mov (16) r[a0.4, 0]<1>:uw 0:uw
560 mov (16) r[a0.5,1]<2>:ub ubTempY(0, 0)<32;8,4>
561 mov (16) r[a0.6,1]<2>:ub ubTempU(0, 0)<32;8,4>
562 mov (16) r[a0.4,1]<2>:ub ubTempV(0, 0)<32;8,4>
563
564
565 //the second line in the block 1
566 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 33]<16;8,2>:ub
567 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 49]<16;8,2>:ub
568 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 33]<16;8,2>:ub
569 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 49]<16;8,2>:ub
570 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 33]<16;8,2>:ub
571 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 49]<16;8,2>:ub
572
573 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
574 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
575 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
576
577 mov (16) acc0.0<1>:f 128.0f
578 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
579 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
580 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
581
582 mov (16) acc0.0<1>:f 128.0f
583 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
584 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
585 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
586
587 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
588 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
589 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
590
591 mov (16) r[a0.5, 32]<1>:uw 0:uw
592 mov (16) r[a0.6, 32]<1>:uw 0:uw
593 mov (16) r[a0.4, 32]<1>:uw 0:uw
594 mov (16) r[a0.5,33]<2>:ub ubTempY(0, 0)<32;8,4>
595 mov (16) r[a0.6,33]<2>:ub ubTempU(0, 0)<32;8,4>
596 mov (16) r[a0.4,33]<2>:ub ubTempV(0, 0)<32;8,4>
597
598 //the third line in the block 1
599 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 65]<16;8,2>:ub
600 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 81]<16;8,2>:ub
601 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 65]<16;8,2>:ub
602 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 81]<16;8,2>:ub
603 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 65]<16;8,2>:ub
604 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 81]<16;8,2>:ub
605
606 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
607 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
608 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
609
610 mov (16) acc0.0<1>:f 128.0f
611 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
612 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
613 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
614
615 mov (16) acc0.0<1>:f 128.0f
616 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
617 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
618 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
619
620 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
621 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
622 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
623
624 mov (16) r[a0.5, 64]<1>:uw 0:uw
625 mov (16) r[a0.6, 64]<1>:uw 0:uw
626 mov (16) r[a0.4, 64]<1>:uw 0:uw
627 mov (16) r[a0.5,65]<2>:ub ubTempY(0, 0)<32;8,4>
628 mov (16) r[a0.6,65]<2>:ub ubTempU(0, 0)<32;8,4>
629 mov (16) r[a0.4,65]<2>:ub ubTempV(0, 0)<32;8,4>
630
631 //the fourth line in the block 1
632 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 97]<16;8,2>:ub
633 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 113]<16;8,2>:ub
634 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 97]<16;8,2>:ub
635 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 113]<16;8,2>:ub
636 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 97]<16;8,2>:ub
637 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 113]<16;8,2>:ub
638
639 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
640 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
641 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
642
643 mov (16) acc0.0<1>:f 128.0f
644 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
645 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
646 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
647
648 mov (16) acc0.0<1>:f 128.0f
649 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
650 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
651 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
652
653 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
654 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
655 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
656
657 mov (16) r[a0.5, 96]<1>:uw 0:uw
658 mov (16) r[a0.6, 96]<1>:uw 0:uw
659 mov (16) r[a0.4, 96]<1>:uw 0:uw
660 mov (16) r[a0.5,97]<2>:ub ubTempY(0, 0)<32;8,4>
661 mov (16) r[a0.6,97]<2>:ub ubTempU(0, 0)<32;8,4>
662 mov (16) r[a0.4,97]<2>:ub ubTempV(0, 0)<32;8,4>
663
664 //for Buffer_2
665 add (8) a0.0<1>:uw a0.0<8;8,1>:uw 512:uw
666 //the first line in the block 2
667 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 1]<16;8,2>:ub
668 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 17]<16;8,2>:ub
669 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 1]<16;8,2>:ub
670 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 17]<16;8,2>:ub
671 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 1]<16;8,2>:ub
672 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 17]<16;8,2>:ub
673
674 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
675 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
676 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
677
678 mov (16) acc0.0<1>:f 128.0f
679 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
680 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
681 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
682
683 mov (16) acc0.0<1>:f 128.0f
684 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
685 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
686 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
687
688 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
689 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
690 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
691
692 mov (16) r[a0.5, 0]<1>:uw 0:uw
693 mov (16) r[a0.6, 0]<1>:uw 0:uw
694 mov (16) r[a0.4, 0]<1>:uw 0:uw
695 mov (16) r[a0.5,1]<2>:ub ubTempY(0, 0)<32;8,4>
696 mov (16) r[a0.6,1]<2>:ub ubTempU(0, 0)<32;8,4>
697 mov (16) r[a0.4,1]<2>:ub ubTempV(0, 0)<32;8,4>
698
699 //the second line in the block 2
700 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 33]<16;8,2>:ub
701 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 49]<16;8,2>:ub
702 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 33]<16;8,2>:ub
703 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 49]<16;8,2>:ub
704 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 33]<16;8,2>:ub
705 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 49]<16;8,2>:ub
706
707 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
708 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
709 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
710
711 mov (16) acc0.0<1>:f 128.0f
712 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
713 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
714 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
715
716 mov (16) acc0.0<1>:f 128.0f
717 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
718 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
719 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
720
721 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
722 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
723 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
724
725 mov (16) r[a0.5, 32]<1>:uw 0:uw
726 mov (16) r[a0.6, 32]<1>:uw 0:uw
727 mov (16) r[a0.4, 32]<1>:uw 0:uw
728 mov (16) r[a0.5,33]<2>:ub ubTempY(0, 0)<32;8,4>
729 mov (16) r[a0.6,33]<2>:ub ubTempU(0, 0)<32;8,4>
730 mov (16) r[a0.4,33]<2>:ub ubTempV(0, 0)<32;8,4>
731
732 //the third line in the block 2
733 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 65]<16;8,2>:ub
734 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 81]<16;8,2>:ub
735 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 65]<16;8,2>:ub
736 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 81]<16;8,2>:ub
737 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 65]<16;8,2>:ub
738 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 81]<16;8,2>:ub
739
740 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
741 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
742 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
743
744 mov (16) acc0.0<1>:f 128.0f
745 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
746 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
747 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
748
749 mov (16) acc0.0<1>:f 128.0f
750 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
751 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
752 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
753
754 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
755 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
756 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
757
758 mov (16) r[a0.5, 64]<1>:uw 0:uw
759 mov (16) r[a0.6, 64]<1>:uw 0:uw
760 mov (16) r[a0.4, 64]<1>:uw 0:uw
761 mov (16) r[a0.5,65]<2>:ub ubTempY(0, 0)<32;8,4>
762 mov (16) r[a0.6,65]<2>:ub ubTempU(0, 0)<32;8,4>
763 mov (16) r[a0.4,65]<2>:ub ubTempV(0, 0)<32;8,4>
764
765 //the fourth line in the block 2
766 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 97]<16;8,2>:ub
767 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 113]<16;8,2>:ub
768 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 97]<16;8,2>:ub
769 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 113]<16;8,2>:ub
770 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 97]<16;8,2>:ub
771 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 113]<16;8,2>:ub
772
773 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
774 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
775 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
776
777 mov (16) acc0.0<1>:f 128.0f
778 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
779 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
780 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
781
782 mov (16) acc0.0<1>:f 128.0f
783 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
784 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
785 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
786
787 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
788 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
789 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
790
791 mov (16) r[a0.5, 96]<1>:uw 0:uw
792 mov (16) r[a0.6, 96]<1>:uw 0:uw
793 mov (16) r[a0.4, 96]<1>:uw 0:uw
794 mov (16) r[a0.5,97]<2>:ub ubTempY(0, 0)<32;8,4>
795 mov (16) r[a0.6,97]<2>:ub ubTempU(0, 0)<32;8,4>
796 mov (16) r[a0.4,97]<2>:ub ubTempV(0, 0)<32;8,4>
797
798 //for Buffer_3
799 add (8) a0.0<1>:uw a0.0<8;8,1>:uw 512:uw
800 //the first line in the block 3
801 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 1]<16;8,2>:ub
802 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 17]<16;8,2>:ub
803 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 1]<16;8,2>:ub
804 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 17]<16;8,2>:ub
805 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 1]<16;8,2>:ub
806 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 17]<16;8,2>:ub
807
808 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
809 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
810 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
811
812 mov (16) acc0.0<1>:f 128.0f
813 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
814 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
815 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
816
817 mov (16) acc0.0<1>:f 128.0f
818 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
819 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
820 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
821
822 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
823 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
824 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
825
826 mov (16) r[a0.5, 0]<1>:uw 0:uw
827 mov (16) r[a0.6, 0]<1>:uw 0:uw
828 mov (16) r[a0.4, 0]<1>:uw 0:uw
829 mov (16) r[a0.5,1]<2>:ub ubTempY(0, 0)<32;8,4>
830 mov (16) r[a0.6,1]<2>:ub ubTempU(0, 0)<32;8,4>
831 mov (16) r[a0.4,1]<2>:ub ubTempV(0, 0)<32;8,4>
832
833
834 //the second line in the block 3
835 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 33]<16;8,2>:ub
836 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 49]<16;8,2>:ub
837 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 33]<16;8,2>:ub
838 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 49]<16;8,2>:ub
839 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 33]<16;8,2>:ub
840 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 49]<16;8,2>:ub
841
842 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
843 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
844 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
845
846 mov (16) acc0.0<1>:f 128.0f
847 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
848 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
849 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
850
851 mov (16) acc0.0<1>:f 128.0f
852 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
853 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
854 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
855
856 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
857 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
858 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
859
860 mov (16) r[a0.5, 32]<1>:uw 0:uw
861 mov (16) r[a0.6, 32]<1>:uw 0:uw
862 mov (16) r[a0.4, 32]<1>:uw 0:uw
863 mov (16) r[a0.5,33]<2>:ub ubTempY(0, 0)<32;8,4>
864 mov (16) r[a0.6,33]<2>:ub ubTempU(0, 0)<32;8,4>
865 mov (16) r[a0.4,33]<2>:ub ubTempV(0, 0)<32;8,4>
866
867 //the third line in the block 3
868 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 65]<16;8,2>:ub
869 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 81]<16;8,2>:ub
870 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 65]<16;8,2>:ub
871 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 81]<16;8,2>:ub
872 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 65]<16;8,2>:ub
873 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 81]<16;8,2>:ub
874
875 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
876 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
877 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
878
879 mov (16) acc0.0<1>:f 128.0f
880 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
881 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
882 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
883
884 mov (16) acc0.0<1>:f 128.0f
885 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
886 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
887 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
888
889 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
890 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
891 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
892
893 mov (16) r[a0.5, 64]<1>:uw 0:uw
894 mov (16) r[a0.6, 64]<1>:uw 0:uw
895 mov (16) r[a0.4, 64]<1>:uw 0:uw
896 mov (16) r[a0.5,65]<2>:ub ubTempY(0, 0)<32;8,4>
897 mov (16) r[a0.6,65]<2>:ub ubTempU(0, 0)<32;8,4>
898 mov (16) r[a0.4,65]<2>:ub ubTempV(0, 0)<32;8,4>
899
900 //the fourth line in the block 3
901 mov (8) fBUFFER_R(0, 0)<1> r[a0.1, 97]<16;8,2>:ub
902 mov (8) fBUFFER_R(1, 0)<1> r[a0.1, 113]<16;8,2>:ub
903 mov (8) fBUFFER_G(0, 0)<1> r[a0.2, 97]<16;8,2>:ub
904 mov (8) fBUFFER_G(1, 0)<1> r[a0.2, 113]<16;8,2>:ub
905 mov (8) fBUFFER_B(0, 0)<1> r[a0.0, 97]<16;8,2>:ub
906 mov (8) fBUFFER_B(1, 0)<1> r[a0.0, 113]<16;8,2>:ub
907
908 mul (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.299f
909 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> 0.587f
910 mac (16) fBUFFER_Y(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.114f
911
912 mov (16) acc0.0<1>:f 128.0f
913 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> -0.169f
914 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.331f
915 mac (16) fBUFFER_U(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> 0.499f
916
917 mov (16) acc0.0<1>:f 128.0f
918 mac (16) acc0.0<1>:f fBUFFER_R(0, 0)<8;8,1> 0.499f
919 mac (16) acc0.0<1>:f fBUFFER_G(0, 0)<8;8,1> -0.418f
920 mac (16) fBUFFER_V(0, 0)<1> fBUFFER_B(0, 0)<8;8,1> -0.0813f
921
922 mov (16) wTempY(0,0)<2> fBUFFER_Y(0, 0)<8;8,1>
923 mov (16) wTempU(0,0)<2> fBUFFER_U(0, 0)<8;8,1>
924 mov (16) wTempV(0,0)<2> fBUFFER_V(0, 0)<8;8,1>
925
926 mov (16) r[a0.5, 96]<1>:uw 0:uw
927 mov (16) r[a0.6, 96]<1>:uw 0:uw
928 mov (16) r[a0.4, 96]<1>:uw 0:uw
929 mov (16) r[a0.5,97]<2>:ub ubTempY(0, 0)<32;8,4>
930 mov (16) r[a0.6,97]<2>:ub ubTempU(0, 0)<32;8,4>
931 mov (16) r[a0.4,97]<2>:ub ubTempV(0, 0)<32;8,4>
932