summaryrefslogtreecommitdiffstats
path: root/firmware/target/arm/lcd-as-memframe.S
blob: 4532bab2212c1fd4f464b53422c455b6d01ff870 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
/***************************************************************************
 *             __________               __   ___.
 *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
 *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
 *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
 *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
 *                     \/            \/     \/    \/            \/
 * $Id$
 *
 * Copyright (C) 2007 by Michael Sevakis
 *
 * ARM code for memory framebuffer LCDs
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version 2
 * of the License, or (at your option) any later version.
 *
 * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
 * KIND, either express or implied.
 *
 ****************************************************************************/

#include "config.h"
#include "cpu.h"

/****************************************************************************
 * void lcd_copy_buffer_rect(fb_data *dst, fb_data *src, int width,
 *                           int height);
 */
     .section    .icode, "ax", %progbits
     .align      2
     .global     lcd_copy_buffer_rect
     .type       lcd_copy_buffer_rect, %function
                                        @ r0 = dst
                                        @ r1 = src
                                        @ r2 = width
                                        @ r3 = height
lcd_copy_buffer_rect:                   @
    stmfd   sp!, { r4-r11, lr }         @ save non-scratch regs
    mov     r5, r2                      @ r5 = cached width
    rsb     r4, r2, #LCD_WIDTH          @ r4 = LCD_WIDTH - width
10: @ copy line                         @
    subs    r2, r5, #1                  @ r2 = width - 1
    beq     40f @ finish line           @ one halfword? skip to trailing copy
    tst     r0, #2                      @ word aligned?
    beq     20f @ rem copy              @ yes? skip to word copy
    ldrh    r6, [r1], #2                @ copy leading halfword
    subs    r2, r2, #1                  @
    strh    r6, [r0], #2                @
    ble     40f @ finish line           @ next line if lt or finish
                                        @ trailing halfword if eq
20: @ rem copy                          @
    add     r14, r2, #1                 @ get remaining width mod 16 after word
                                        @ align (rw)
    and     r14, r14, #0xe              @ r14 = 0 (16), 2, 4, 6, 8, 10, 12, 14
    add     pc, pc, r14, lsl #3         @ branch to 32-byte align
    nop                                 @ 
    b       30f                         @ rw % 16 = 0 or 1? use octword loop
    nop                                 @
    nop                                 @
    nop                                 @
    ldr     r6, [r1], #4                @ rw % 16 = 2 or 3
    subs    r2, r2, #2                  @
    str     r6, [r0], #4                @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r7 }              @ rw % 16 = 4 or 5
    subs    r2, r2, #4                  @
    stmia   r0!, { r6-r7 }              @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r8 }              @ rw % 16 = 6 or 7
    subs    r2, r2, #6                  @
    stmia   r0!, { r6-r8 }              @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r9 }              @ rw % 16 = 8 or 9
    subs    r2, r2, #8                  @
    stmia   r0!, { r6-r9 }              @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r10 }             @ rw % 16 = 10 or 11
    subs    r2, r2, #10                 @
    stmia   r0!, { r6-r10 }             @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r11 }             @ rw % 16 = 12 or 13
    subs    r2, r2, #12                 @
    stmia   r0!, { r6-r11 }             @
    b       25f @ copy up done          @
    ldmia   r1!, { r6-r12 }             @ rw % 16 = 14 or 15
    subs    r2, r2, #14                 @
    stmia   r0!, { r6-r12 }             @
25: @ copy up done                      @
    ble     40f @ finish line           @ no 32-byte segments remaining?
30: @ octword loop                      @ copy 16 pixels per loop
    ldmia   r1!, { r6-r12, r14 }        @
    subs    r2, r2, #16                 @
    stmia   r0!, { r6-r12, r14 }        @
    bgt     30b @ octword loop          @ 
40: @ finish line                       @
    ldreqh  r6, [r1], #2                @ finish last halfword if eq ...
    add     r1, r1, r4, lsl #1          @
    streqh  r6, [r0], #2                @ ...
    add     r0, r0, r4, lsl #1          @
    subs    r3, r3, #1                  @ next line
    bgt     10b @ copy line             @
    ldmfd   sp!, { r4-r11, pc }         @ restore regs and return
    .ltorg                              @ dump constant pool
    .size   lcd_copy_buffer_rect, .-lcd_copy_buffer_rect

/****************************************************************************
 * void lcd_write_yuv_420_lines(fb_data *dst,
 *                              unsigned char const * const src[3],
 *                              int width,
 *                              int stride);
 *
 *   |R|   |1.000000 -0.000001  1.402000| |Y'|
 *   |G| = |1.000000 -0.334136 -0.714136| |Pb|
 *   |B|   |1.000000  1.772000  0.000000| |Pr|
 *   Scaled, normalized, rounded and tweaked to yield RGB 565:
 *   |R|   |74   0 101| |Y' -  16| >> 9
 *   |G| = |74 -24 -51| |Cb - 128| >> 8
 *   |B|   |74 128   0| |Cr - 128| >> 9
 *
 * Write four RGB565 pixels in the following order on each loop:
 * 1 3 + > down
 * 2 4 \/ left
 */
    .section    .icode, "ax", %progbits
    .align      2
    .global     lcd_write_yuv420_lines
    .type       lcd_write_yuv420_lines, %function
lcd_write_yuv420_lines:
                                        @ r0 = dst
                                        @ r1 = yuv_src
                                        @ r2 = width
                                        @ r3 = stride
    stmfd       sp!, { r4-r10, lr }     @ save non-scratch
    ldmia       r1, { r4, r5, r6 }      @ r4 = yuv_src[0] = Y'_p
                                        @ r5 = yuv_src[1] = Cb_p
                                        @ r6 = yuv_src[2] = Cr_p
                                        @ r1 = scratch
    sub         r3, r3, #1              @
10: @ loop line                         @
    ldrb        r7, [r4], #1            @ r7 = *Y'_p++;
    ldrb        r8, [r5], #1            @ r8 = *Cb_p++;
    ldrb        r9, [r6], #1            @ r9 = *Cr_p++;
                                        @
    sub         r7, r7, #16             @ r7 = Y = (Y' - 16)*74
    add         r12, r7, r7, asl #2     @ actually (Y' - 16)*37 and shift right
    add         r7, r12, r7, asl #5     @ by one less when adding - same for all
                                        @
    sub         r8, r8, #128            @ Cb -= 128
    sub         r9, r9, #128            @ Cr -= 128
                                        @
    add         r10, r9, r9, asl #1     @ r10 = Cr*51 + Cb*24
    add         r10, r10, r10, asl #4   @
    add         r10, r10, r8, asl #3    @
    add         r10, r10, r8, asl #4    @
                                        @
    add         lr, r9, r9, asl #2      @ r9 = Cr*101
    add         lr, lr, r9, asl #5      @
    add         r9, lr, r9, asl #6      @
                                        @
    add         r8, r8, #2              @ r8 = bu = (Cb*128 + 128) >> 8
    mov         r8, r8, asr #2          @
    add         r9, r9, #256            @ r9 = rv = (r9 + 256) >> 9
    mov         r9, r9, asr #9          @
    rsb         r10, r10, #128          @ r10 = guv = (-r10 + 128) >> 8
    mov         r10, r10, asr #8        @
                                        @ compute R, G, and B
    add         r1, r8, r7, asr #8      @ r1  = b = (Y >> 9) + bu
    add         lr, r9, r7, asr #8      @ lr = r = (Y >> 9) + rv
    add         r7, r10, r7, asr #7     @ r7  = g = (Y >> 8) + guv
                                        @
#if ARM_ARCH >= 6
    usat        r1, #5, r1              @ clamp b
    usat        lr, #5, lr              @ clamp r
    usat        r7, #6, r7              @ clamp g
#else
    orr         r12, r1, lr             @ check if clamping is needed...
    orr         r12, r12, r7, asr #1    @ ...at all
    cmp         r12, #31                @
    bls         15f @ no clamp          @
    cmp         r1, #31                 @ clamp b
    mvnhi       r1, r1, asr #31         @
    andhi       r1, r1, #31             @
    cmp         lr, #31                 @ clamp r
    mvnhi       lr, lr, asr #31         @
    andhi       lr, lr, #31             @
    cmp         r7, #63                 @ clamp g
    mvnhi       r7, r7, asr #31         @
    andhi       r7, r7, #63             @
15: @ no clamp                          @
#endif
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    orr         r1, r1, r7, lsl #5      @ r4 |= (g << 5)
    orr         r1, r1, lr, lsl #11     @ r4 = b | (r << 11)

#if LCD_WIDTH >= LCD_HEIGHT
    strh        r1, [r0]                @
#elif LCD_WIDTH < 256
    strh        r1, [r0], #LCD_WIDTH    @ store pixel
#else
    strh        r1, [r0]                @
#endif
                                        @
    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*74
    add         r12, r7, r7, asl #2     @
    add         r7, r12, r7, asl #5     @
                                        @ compute R, G, and B
    add         r1, r8, r7, asr #8      @ r1  = b = (Y >> 9) + bu
    add         lr, r9, r7, asr #8      @ lr = r = (Y >> 9) + rv
    add         r7, r10, r7, asr #7     @ r7  = g = (Y >> 8) + guv
                                        @
#if ARM_ARCH >= 6
    usat        r1, #5, r1              @ clamp b
    usat        lr, #5, lr              @ clamp r
    usat        r7, #6, r7              @ clamp g
#else
    orr         r12, r1, lr             @ check if clamping is needed...
    orr         r12, r12, r7, asr #1    @ ...at all
    cmp         r12, #31                @
    bls         15f @ no clamp          @
    cmp         r1, #31                 @ clamp b
    mvnhi       r1, r1, asr #31         @
    andhi       r1, r1, #31             @
    cmp         lr, #31                 @ clamp r
    mvnhi       lr, lr, asr #31         @
    andhi       lr, lr, #31             @
    cmp         r7, #63                 @ clamp g
    mvnhi       r7, r7, asr #31         @
    andhi       r7, r7, #63             @
15: @ no clamp                          @
#endif
                                        @
    ldrb        r12, [r4], #1           @ r12 = Y' = *(Y'_p++)
                                        @
    orr         r1, r1, lr, lsl #11     @ r1 = b | (r << 11)
    orr         r1, r1, r7, lsl #5      @ r1 |= (g << 5)

#if LCD_WIDTH >= LCD_HEIGHT
    add         r0, r0, #2*LCD_WIDTH    @
    strh        r1, [r0]        @ store pixel
    sub         r0, r0, #2*LCD_WIDTH    @
#elif LCD_WIDTH < 256
    strh        r1, [r0, #-LCD_WIDTH-2] @ store pixel
#else
    strh        r1, [r0, #-2]           @
    add         r0, r0, #LCD_WIDTH      @
#endif
                                        @
    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*74
    add         r12, r7, r7, asl #2     @
    add         r7, r12, r7, asl #5     @
                                        @ compute R, G, and B
    add         r1, r8, r7, asr #8      @ r1  = b = (Y >> 9) + bu
    add         lr, r9, r7, asr #8      @ lr = r = (Y >> 9) + rv
    add         r7, r10, r7, asr #7     @ r7  = g = (Y >> 8) + guv
                                        @
#if ARM_ARCH >= 6
    usat        r1, #5, r1              @ clamp b
    usat        lr, #5, lr              @ clamp r
    usat        r7, #6, r7              @ clamp g
#else
    orr         r12, r1, lr             @ check if clamping is needed...
    orr         r12, r12, r7, asr #1    @ ...at all
    cmp         r12, #31                @
    bls         15f @ no clamp          @
    cmp         r1, #31                 @ clamp b
    mvnhi       r1, r1, asr #31         @
    andhi       r1, r1, #31             @
    cmp         lr, #31                 @ clamp r
    mvnhi       lr, lr, asr #31         @
    andhi       lr, lr, #31             @
    cmp         r7, #63                 @ clamp g
    mvnhi       r7, r7, asr #31         @
    andhi       r7, r7, #63             @
15: @ no clamp                          @
#endif
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    orr         r1, r1, r7, lsl #5      @ r1 = b | (g << 5)
    orr         r1, r1, lr, lsl #11     @ r1 |= (r << 11)

#if LCD_WIDTH >= LCD_HEIGHT
    strh        r1, [r0, #2]
#elif LCD_WIDTH <  256
    strh        r1, [r0, #LCD_WIDTH]!   @ store pixel
#else
    strh        r1, [r0]                @
#endif
                                        @
    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*74
    add         r12, r7, r7, asl #2     @
    add         r7, r12, r7, asl #5     @
                                        @ compute R, G, and B
    add         r1, r8, r7, asr #8      @ r1  = b = (Y >> 9) + bu
    add         lr, r9, r7, asr #8      @ lr = r = (Y >> 9) + rv
    add         r7, r10, r7, asr #7     @ r7  = g = (Y >> 8) + guv
                                        @
#if ARM_ARCH >= 6
    usat        r1, #5, r1              @ clamp b
    usat        lr, #5, lr              @ clamp r
    usat        r7, #6, r7              @ clamp g
#else
    orr         r12, r1, lr             @ check if clamping is needed...
    orr         r12, r12, r7, asr #1    @ ...at all
    cmp         r12, #31                @
    bls         15f @ no clamp          @
    cmp         r1, #31                 @ clamp b
    mvnhi       r1, r1, asr #31         @
    andhi       r1, r1, #31             @
    cmp         lr, #31                 @ clamp r
    mvnhi       lr, lr, asr #31         @
    andhi       lr, lr, #31             @
    cmp         r7, #63                 @ clamp g
    mvnhi       r7, r7, asr #31         @
    andhi       r7, r7, #63             @
15: @ no clamp                          @
#endif
                                        @
    orr         r12, r1, lr, lsl #11    @ r12 = b | (r << 11)
    orr         r12, r12, r7, lsl #5    @ r12 |= (g << 5)

#if LCD_WIDTH >= LCD_HEIGHT
    add         r0, r0, #2*LCD_WIDTH
    strh        r12, [r0, #2]
#if LCD_WIDTH <= 512
    sub         r0, r0, #(2*LCD_WIDTH)-4
#else
    sub         r0, r0, #(2*LCD_WIDTH)
    add         r0, r0, #4
#endif
#else
    strh        r12, [r0, #-2]          @ store pixel
#if LCD_WIDTH < 256
    add         r0, r0, #2*LCD_WIDTH    @
#else
    add         r0, r0, #LCD_WIDTH      @
#endif
#endif
                                        @
    subs        r2, r2, #2              @ subtract block from width
    bgt         10b @ loop line         @
                                        @
    ldmfd       sp!, { r4-r10, pc }     @ restore registers and return
    .ltorg                              @ dump constant pool
    .size   lcd_write_yuv420_lines, .-lcd_write_yuv420_lines


/****************************************************************************
 * void lcd_write_yuv_420_lines_odither(fb_data *dst,
 *                                      unsigned char const * const src[3],
 *                                      int width,
 *                                      int stride,
 *                                      int x_screen,
 *                                      int y_screen);
 *
 *   |R|   |1.000000 -0.000001  1.402000| |Y'|
 *   |G| = |1.000000 -0.334136 -0.714136| |Pb|
 *   |B|   |1.000000  1.772000  0.000000| |Pr|
 *   Red scaled at twice g & b but at same precision to place it in correct
 *   bit position after multiply and leave instruction count lower.
 *   |R|   |258   0  408| |Y' -  16|
 *   |G| = |149 -49 -104| |Cb - 128|
 *   |B|   |149 258    0| |Cr - 128|
 *
 * Write four RGB565 pixels in the following order on each loop:
 * 1 3 + > down
 * 2 4 \/ left
 *
 * Kernel pattern (raw|rotated|use order):
 * 5 3 4 2   2 6 3 7     row0   row2          > down
 * 1 7 0 6 | 4 0 5 1 | 2 4 6 0 3 5 7 1 col0     left
 * 4 2 5 3 | 3 7 2 6 | 3 5 7 1 2 4 6 0 col2  \/
 * 0 6 1 7   5 1 4 0
 */
    .section    .icode, "ax", %progbits
    .align      2
    .global     lcd_write_yuv420_lines_odither
    .type       lcd_write_yuv420_lines_odither, %function
lcd_write_yuv420_lines_odither:
                                        @ r0 = dst
                                        @ r1 = yuv_src
                                        @ r2 = width
                                        @ r3 = stride
                                        @ [sp]   = x_screen
                                        @ [sp+4] = y_screen
    stmfd       sp!, { r4-r11, lr }     @ save non-scratch
    ldmia       r1, { r4, r5, r6 }      @ r4 = yuv_src[0] = Y'_p
                                        @ r5 = yuv_src[1] = Cb_p
                                        @ r6 = yuv_src[2] = Cr_p
                                        @
    sub         r3, r3, #1              @
    add         r1, sp, #36             @ Line up pattern and kernel quadrant
    ldmia       r1, { r12, r14 }        @
    eor         r14, r14, r12           @
    and         r14, r14, #0x2          @
    mov         r14, r14, lsl #6        @ 0x00 or 0x80
10: @ loop line                         @
                                        @
    ldrb        r7, [r4], #1            @ r7 = *Y'_p++;
    ldrb        r8, [r5], #1            @ r8 = *Cb_p++;
    ldrb        r9, [r6], #1            @ r9 = *Cr_p++;
                                        @
    eor         r14, r14, #0x80         @ flip pattern quadrant
                                        @
    sub         r7, r7, #16             @ r7 = Y = (Y' - 16)*149
    add         r12, r7, r7, asl #2     @
    add         r12, r12, r12, asl #4   @
    add         r7, r12, r7, asl #6     @
                                        @    
    sub         r8, r8, #128            @ Cb -= 128
    sub         r9, r9, #128            @ Cr -= 128
                                        @
    add         r10, r8, r8, asl #4     @ r10 = guv = Cr*104 + Cb*49
    add         r10, r10, r8, asl #5    @
    add         r10, r10, r9, asl #3    @
    add         r10, r10, r9, asl #5    @
    add         r10, r10, r9, asl #6    @
                                        @
    mov         r8, r8, asl #1          @ r8 = bu = Cb*258
    add         r8, r8, r8, asl #7      @
                                        @
    add         r9, r9, r9, asl #1      @ r9 = rv = Cr*408
    add         r9, r9, r9, asl #4      @
    mov         r9, r9, asl #3          @
                                        @
                                        @ compute R, G, and B
    add         r1, r8, r7              @ r1  = b' = Y + bu
    add         r11, r9, r7, asl #1     @ r11 = r' = Y*2 + rv
    rsb         r7, r10, r7             @ r7  = g' = Y + guv
                                        @
                                        @ r8 = bu, r9 = rv, r10 = guv
                                        @
    sub         r12, r1, r1, lsr #5     @ r1 = 31/32*b + b/256
    add         r1, r12, r1, lsr #8     @
                                        @
    sub         r12, r11, r11, lsr #5   @ r11 = 31/32*r + r/256
    add         r11, r12, r11, lsr #8   @
                                        @
    sub         r12, r7, r7, lsr #6     @ r7 = 63/64*g + g/256
    add         r7, r12, r7, lsr #8     @
                                        @
    add         r12, r14, #0x100        @
                                        @
    add         r1, r1, r12             @ b = r1 + delta
    add         r11, r11, r12, lsl #1   @ r = r11 + delta*2
    add         r7, r7, r12, lsr #1     @ g = r7 + delta/2
                                        @
#if ARM_ARCH >= 6
    usat        r11, #5, r11, asr #11   @ clamp r
    usat        r7, #6, r7, asr #9      @ clamp g
    usat        r1, #5, r1, asr #10     @ clamp b
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    orr         r1, r1, r11, lsl #11    @ r1 = b | (r << 11)
    orr         r1, r1, r7, lsl #5      @ r1 |= (g << 5)
#else
    orr         r12, r1, r11, asr #1    @ check if clamping is needed...
    orr         r12, r12, r7            @ ...at all
    movs        r12, r12, asr #15       @
    beq         15f @ no clamp          @
    movs        r12, r1, asr #15        @ clamp b
    mvnne       r1, r12, lsr #15        @
    andne       r1, r1, #0x7c00         @ mask b only if clamped
    movs        r12, r11, asr #16       @ clamp r
    mvnne       r11, r12, lsr #16       @
    movs        r12, r7, asr #15        @ clamp g
    mvnne       r7, r12, lsr #15        @
15: @ no clamp                          @
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    and         r11, r11, #0xf800       @ pack pixel
    and         r7, r7, #0x7e00         @ r1 = pixel = (r & 0xf800) |
    orr         r11, r11, r7, lsr #4    @              ((g & 0x7e00) >> 4) |
    orr         r1, r11, r1, lsr #10    @              (b >> 10)
#endif
                                        @
#if LCD_WIDTH >= LCD_HEIGHT
    strh        r1, [r0]        @
#elif LCD_WIDTH < 256
    strh        r1, [r0], #LCD_WIDTH    @ store pixel
#else
    strh        r1, [r0]                @
#endif
                                        @
    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*149
    add         r12, r7, r7, asl #2     @
    add         r12, r12, r12, asl #4   @
    add         r7, r12, r7, asl #6     @
                                        @ compute R, G, and B
    add         r1, r8, r7              @ r1  = b' = Y + bu
    add         r11, r9, r7, asl #1     @ r11 = r' = Y*2 + rv
    rsb         r7, r10, r7             @ r7  = g' = Y + guv
                                        @
    sub         r12, r1, r1, lsr #5     @ r1  = 31/32*b' + b'/256
    add         r1, r12, r1, lsr #8     @
                                        @
    sub         r12, r11, r11, lsr #5   @ r11 = 31/32*r' + r'/256
    add         r11, r12, r11, lsr #8   @
                                        @
    sub         r12, r7, r7, lsr #6     @ r7  = 63/64*g' + g'/256
    add         r7, r12, r7, lsr #8     @
                                        @
    add         r12, r14, #0x200        @
                                        @
    add         r1, r1, r12             @ b = r1 + delta
    add         r11, r11, r12, lsl #1   @ r = r11 + delta*2
    add         r7, r7, r12, lsr #1     @ g = r7 + delta/2
                                        @
#if ARM_ARCH >= 6
    usat        r11, #5, r11, asr #11   @ clamp r
    usat        r7, #6, r7, asr #9      @ clamp g
    usat        r1, #5, r1, asr #10     @ clamp b
                                        @
    ldrb        r12, [r4], #1           @ r12 = Y' = *(Y'_p++)
                                        @
    orr         r1, r1, r11, lsl #11    @ r1 = b | (r << 11)
    orr         r1, r1, r7, lsl #5      @ r1 |= (g << 5)
#else
    orr         r12, r1, r11, asr #1    @ check if clamping is needed...
    orr         r12, r12, r7            @ ...at all
    movs        r12, r12, asr #15       @
    beq         15f @ no clamp          @
    movs        r12, r1, asr #15        @ clamp b
    mvnne       r1, r12, lsr #15        @
    andne       r1, r1, #0x7c00         @ mask b only if clamped
    movs        r12, r11, asr #16       @ clamp r
    mvnne       r11, r12, lsr #16       @
    movs        r12, r7, asr #15        @ clamp g
    mvnne       r7, r12, lsr #15        @
15: @ no clamp                          @
                                        @
    ldrb        r12, [r4], #1           @ r12 = Y' = *(Y'_p++)
                                        @
    and         r11, r11, #0xf800       @ pack pixel
    and         r7, r7, #0x7e00         @ r1 = pixel = (r & 0xf800) |
    orr         r11, r11, r7, lsr #4    @              ((g & 0x7e00) >> 4) |
    orr         r1, r11, r1, lsr #10    @              (b >> 10)
#endif
                                        @
#if LCD_WIDTH >= LCD_HEIGHT
    add         r0, r0, #2*LCD_WIDTH    @
    strh        r1, [r0]        @ store pixel
    sub         r0, r0, #2*LCD_WIDTH    @
#elif LCD_WIDTH < 256
    strh        r1, [r0, #-LCD_WIDTH-2] @ store pixel
#else
    strh        r1, [r0, #-2]           @ store pixel
    add         r0, r0, #LCD_WIDTH      @
#endif
                                        @
    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*149
    add         r12, r7, r7, asl #2     @
    add         r12, r12, r12, asl #4   @
    add         r7, r12, r7, asl #6     @
                                        @ compute R, G, and B
    add         r1, r8, r7              @ r1  = b' = Y + bu
    add         r11, r9, r7, asl #1     @ r11 = r' = Y*2 + rv
    rsb         r7, r10, r7             @ r7  = g' = Y + guv
                                        @
                                        @ r8 = bu, r9 = rv, r10 = guv
                                        @
    sub         r12, r1, r1, lsr #5     @ r1  = 31/32*b' + b'/256
    add         r1, r12, r1, lsr #8     @
                                        @
    sub         r12, r11, r11, lsr #5   @ r11 = 31/32*r' + r'/256
    add         r11, r12, r11, lsr #8   @
                                        @
    sub         r12, r7, r7, lsr #6     @ r7  = 63/64*g' + g'/256
    add         r7, r12, r7, lsr #8     @
                                        @
    add         r12, r14, #0x300        @
                                        @
    add         r1, r1, r12             @ b = r1 + delta
    add         r11, r11, r12, lsl #1   @ r = r11 + delta*2
    add         r7, r7, r12, lsr #1     @ g = r7 + delta/2
                                        @
#if ARM_ARCH >= 6
    usat        r11, #5, r11, asr #11   @ clamp r
    usat        r7, #6, r7, asr #9      @ clamp g
    usat        r1, #5, r1, asr #10     @ clamp b
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    orr         r1, r1, r11, lsl #11    @ r1 = b | (r << 11)
    orr         r1, r1, r7, lsl #5      @ r1 |= (g << 5)
#else
    orr         r12, r1, r11, asr #1    @ check if clamping is needed...
    orr         r12, r12, r7            @ ...at all
    movs        r12, r12, asr #15       @
    beq         15f @ no clamp          @
    movs        r12, r1, asr #15        @ clamp b
    mvnne       r1, r12, lsr #15        @
    andne       r1, r1, #0x7c00         @ mask b only if clamped
    movs        r12, r11, asr #16       @ clamp r
    mvnne       r11, r12, lsr #16       @
    movs        r12, r7, asr #15        @ clamp g
    mvnne       r7, r12, lsr #15        @
15: @ no clamp                          @
                                        @
    ldrb        r12, [r4, r3]           @ r12 = Y' = *(Y'_p + stride)
                                        @
    and         r11, r11, #0xf800       @ pack pixel
    and         r7, r7, #0x7e00         @ r1 = pixel = (r & 0xf800) |
    orr         r11, r11, r7, lsr #4    @              ((g & 0x7e00) >> 4) |
    orr         r1, r11, r1, lsr #10    @              (b >> 10)
#endif
                                        @
#if LCD_WIDTH >= LCD_HEIGHT
    strh        r1, [r0, #2]
#elif LCD_WIDTH <  256
    strh        r1, [r0, #LCD_WIDTH]!   @ store pixel
#else
    strh        r1, [r0]                @
#endif

    sub         r7, r12, #16            @ r7 = Y = (Y' - 16)*149
    add         r12, r7, r7, asl #2     @
    add         r12, r12, r12, asl #4   @
    add         r7, r12, r7, asl #6     @
                                        @ compute R, G, and B
    add         r1, r8, r7              @ r1  = b' = Y + bu
    add         r11, r9, r7, asl #1     @ r11 = r' = Y*2 + rv
    rsb         r7, r10, r7             @ r7  = g' = Y + guv
                                        @
    sub         r12, r1, r1, lsr #5     @ r1 = 31/32*b + b/256
    add         r1, r12, r1, lsr #8     @
                                        @
    sub         r12, r11, r11, lsr #5   @ r11 = 31/32*r + r/256
    add         r11, r12, r11, lsr #8   @
                                        @
    sub         r12, r7, r7, lsr #6     @ r7 = 63/64*g + g/256
    add         r7, r12, r7, lsr #8     @
                                        @
    @ This element is zero - use r14    @
                                        @
    add         r1, r1, r14             @ b = r1 + delta
    add         r11, r11, r14, lsl #1   @ r = r11 + delta*2
    add         r7, r7, r14, lsr #1     @ g = r7 + delta/2
                                        @
#if ARM_ARCH >= 6
    usat        r11, #5, r11, asr #11   @ clamp r
    usat        r7, #6, r7, asr #9      @ clamp g
    usat        r1, #5, r1, asr #10     @ clamp b
                                        @
    orr         r1, r1, r11, lsl #11    @ r1 = b | (r << 11)
    orr         r1, r1, r7, lsl #5      @ r1 |= (g << 5)
#else
    orr         r12, r1, r11, asr #1    @ check if clamping is needed...
    orr         r12, r12, r7            @ ...at all
    movs        r12, r12, asr #15       @
    beq         15f @ no clamp          @
    movs        r12, r1, asr #15        @ clamp b
    mvnne       r1, r12, lsr #15        @
    andne       r1, r1, #0x7c00         @ mask b only if clamped
    movs        r12, r11, asr #16       @ clamp r
    mvnne       r11, r12, lsr #16       @
    movs        r12, r7, asr #15        @ clamp g
    mvnne       r7, r12, lsr #15        @
15: @ no clamp                          @
                                        @
    and         r11, r11, #0xf800       @ pack pixel
    and         r7, r7, #0x7e00         @ r1 = pixel = (r & 0xf800) |
    orr         r11, r11, r7, lsr #4    @              ((g & 0x7e00) >> 4) |
    orr         r1, r11, r1, lsr #10    @              (b >> 10)
#endif
                                        @
#if LCD_WIDTH >= LCD_HEIGHT
    add         r0, r0, #2*LCD_WIDTH
    strh        r1, [r0, #2]            @ store pixel
#if LCD_WIDTH <= 512
    sub         r0, r0, #(2*LCD_WIDTH)-4
#else
    sub         r0, r0, #(2*LCD_WIDTH)
    add         r0, r0, #4
#endif
#else
    strh        r1, [r0, #-2]           @ store pixel
#if LCD_WIDTH < 256
    add         r0, r0, #2*LCD_WIDTH    @
#else
    add         r0, r0, #LCD_WIDTH      @
#endif
#endif
                                        @
    subs        r2, r2, #2              @ subtract block from width
    bgt         10b @ loop line         @
                                        @
    ldmfd       sp!, { r4-r11, pc }     @ restore registers and return
    .ltorg                              @ dump constant pool
    .size   lcd_write_yuv420_lines_odither, .-lcd_write_yuv420_lines_odither