summaryrefslogtreecommitdiffstats
path: root/baseline/source/cjpeg_transupp/cjpeg_transupp.c
blob: e77d15b970026dca1f8514a7eb4ac8ac811ae628 (plain) (blame)
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
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
/*

  This program is part of the TACLeBench benchmark suite.
  Version V 2.0

  Name: cjpeg_transupp

  Author: Thomas G. Lane

  Function: This file contains image transformation routines and other utility
    code used by the jpegtran sample application. These are NOT part of the core
    JPEG library. But we keep these routines separate from jpegtran.c to ease
    the task of maintaining jpegtran-like programs that have other user
    interfaces.

  Source: MediaBench II
          http://euler.slu.edu/~fritts/mediabench (mirror)

  Original name: cjpeg

  Changes: No major functional changes.

  License: See the accompanying README file.

*/


/*
  Include section
*/

#include "../extra.h"
#include "jpeglib.h"


/*
  Forward declaration of functions
*/

void cjpeg_transupp_initSeed( void );
signed char cjpeg_transupp_randomInteger( void );
void cjpeg_transupp_init( void );
int cjpeg_transupp_return( void );
void cjpeg_transupp_do_flip_v( j_compress_ptr );
void cjpeg_transupp_do_rot_90( j_compress_ptr );
void cjpeg_transupp_do_rot_180( j_compress_ptr );
void cjpeg_transupp_do_rot_270( j_compress_ptr );
void cjpeg_transupp_do_transverse( j_compress_ptr );
void cjpeg_transupp_main( void );
//int main( void );


/*
  Declaration of global variables
*/

volatile int cjpeg_transupp_seed;

signed char cjpeg_transupp_input[ 256 ];
signed char cjpeg_transupp_input2[ 80 ];
signed char cjpeg_transupp_input3[ 65 ];
signed char cjpeg_transupp_input3_2[ 65 ];
signed char cjpeg_transupp_input4[ 64 ];
signed char cjpeg_transupp_input5[ 65 ];
signed char cjpeg_transupp_input5_2[ 65 ];

/* Output arrays replace writing of results into a file. */
signed char cjpeg_transupp_output_data[ 512 ];
signed char cjpeg_transupp_output_data2[ 512 ];
signed char cjpeg_transupp_output_data3[ 512 ];
signed char cjpeg_transupp_output_data4[ 512 ];
signed char cjpeg_transupp_output_data5[ 512 ];

struct jpeg_compress_struct cjpeg_transupp_dstinfo;


/*
  Initialization- and return-value-related functions
*/

void cjpeg_transupp_initSeed( void )
{
  cjpeg_transupp_seed = 0;
}


/*
  cjpeg_transupp_RandomInteger generates random integers between -128 and 127.
*/
signed char cjpeg_transupp_randomInteger( void )
{
  cjpeg_transupp_seed = ( ( ( cjpeg_transupp_seed * 133 ) + 81 ) % 256 ) - 128;
  return( cjpeg_transupp_seed );
}


void cjpeg_transupp_init( void )
{
  register int i;


  cjpeg_transupp_dstinfo.max_h_samp_factor = 2;
  cjpeg_transupp_dstinfo.max_v_samp_factor = 2;
  cjpeg_transupp_dstinfo.num_components = 3;

  cjpeg_transupp_initSeed();

  _Pragma( "loopbound min 256 max 256" )
  for ( i = 0; i < 256; i++ )
    cjpeg_transupp_input[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 80 max 80" )
  for ( i = 0; i < 80; i++ )
    cjpeg_transupp_input2[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 65 max 65" )
  for ( i = 0; i < 65; i++ )
    cjpeg_transupp_input3[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 65 max 65" )
  for ( i = 0; i < 65; i++ )
    cjpeg_transupp_input3_2[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 64 max 64" )
  for ( i = 0; i < 64; i++ )
    cjpeg_transupp_input4[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 65 max 65" )
  for ( i = 0; i < 65; i++ )
    cjpeg_transupp_input5[ i ] = cjpeg_transupp_randomInteger();

  _Pragma( "loopbound min 65 max 65" )
  for ( i = 0; i < 65; i++ )
    cjpeg_transupp_input5_2[ i ] = cjpeg_transupp_randomInteger();
}


int cjpeg_transupp_return( void )
{
  int checksum = 0;
  unsigned int i;


  _Pragma( "loopbound min 512 max 512" )
  for ( i = 0; i < 512; i++ )
    checksum += cjpeg_transupp_output_data[ i ];

  _Pragma( "loopbound min 512 max 512" )
  for ( i = 0; i < 512; i++ )
    checksum += cjpeg_transupp_output_data2[ i ];

  _Pragma( "loopbound min 512 max 512" )
  for ( i = 0; i < 512; i++ )
    checksum += cjpeg_transupp_output_data3[ i ];

  _Pragma( "loopbound min 512 max 512" )
  for ( i = 0; i < 512; i++ )
    checksum += cjpeg_transupp_output_data4[ i ];

  _Pragma( "loopbound min 512 max 512" )
  for ( i = 0; i < 512; i++ )
    checksum += cjpeg_transupp_output_data5[ i ];

  return( checksum );
}


/*
  Algorithm core functions
*/

/*
  Vertical flip
*/
void cjpeg_transupp_do_flip_v( j_compress_ptr dstinfo )
{
  unsigned int MCU_rows, comp_height, dst_blk_x, dst_blk_y;
  int ci, i, j, offset_y;
  JCOEFPTR src_ptr, dst_ptr;


  /*
    We output into a separate array because we can't touch different rows of the
    source virtual array simultaneously. Otherwise, this is a pretty
    straightforward analog of horizontal flip.
    Within a DCT block, vertical mirroring is done by changing the signs of
    odd-numbered rows.
    Partial iMCUs at the bottom edge are copied verbatim.
  */
  MCU_rows = dstinfo->image_height / ( dstinfo->max_v_samp_factor * DCTSIZE );

  int compptr_v_samp_factor = 8;
  unsigned int compptr_height_in_blocks = 19;
  unsigned int compptr_width_in_blocks = 29;

  _Pragma( "loopbound min 3 max 3" )
  for ( ci = 0; ci < dstinfo->num_components;
        ci++, compptr_v_samp_factor = 1, compptr_width_in_blocks = 15 ) {
    comp_height = MCU_rows * compptr_v_samp_factor;

    compptr_height_in_blocks = 10;
    _Pragma( "loopbound min 2 max 10" )
    for ( dst_blk_y = 0; dst_blk_y < compptr_height_in_blocks;
          dst_blk_y += compptr_v_samp_factor ) {

      _Pragma( "loopbound min 1 max 8" )
      for ( offset_y = 0; offset_y < compptr_v_samp_factor; offset_y++ ) {
        if ( dst_blk_y < comp_height ) {

          /* Row is within the mirrorable area. */
          _Pragma( "loopbound min 15 max 29" )
          for ( dst_blk_x = 0; dst_blk_x < compptr_width_in_blocks;
               dst_blk_x++ ) {

            src_ptr = cjpeg_transupp_input;
            dst_ptr = cjpeg_transupp_output_data;

            _Pragma( "loopbound min 4 max 4" )
            for ( i = 0; i < DCTSIZE; i += 2 ) {

              /* copy even row */
              j = 0;
              _Pragma( "loopbound min 8 max 8" )
              do {
                if ( dst_blk_x < comp_height )
                  *dst_ptr++ = *src_ptr++;
                j++;
              } while ( j < DCTSIZE );

              /* copy odd row with sign change */
              j = 0;
              _Pragma( "loopbound min 8 max 8" )
              do {
                if ( dst_blk_x < comp_height )
                  *dst_ptr++ = - *src_ptr++;
                j++;
              } while ( j < DCTSIZE );
            }
          }
        }
      }
    }
  }
}


/*
  90 degree rotation is equivalent to
   1. Transposing the image;
   2. Horizontal mirroring.
  These two steps are merged into a single processing routine.
*/
void cjpeg_transupp_do_rot_90( j_compress_ptr dstinfo )
{
  unsigned int MCU_cols, comp_width, dst_blk_x, dst_blk_y;
  int ci, i, j, offset_x, offset_y;
  JCOEFPTR src_ptr, dst_ptr;


  /*
    Because of the horizontal mirror step, we can't process partial iMCUs at the
    (output) right edge properly. They just get transposed and not mirrored.
  */
  MCU_cols = dstinfo->image_width / ( dstinfo->max_h_samp_factor * DCTSIZE );

  int compptr_h_samp_factor = 2;
  int compptr_v_samp_factor = 8;
  unsigned int compptr_height_in_blocks = 29;
  unsigned int compptr_width_in_blocks = 19;


  _Pragma( "loopbound min 3 max 3" )
  for ( ci = 0; ci < dstinfo->num_components;
        ci++, compptr_h_samp_factor = compptr_v_samp_factor = 1,
        compptr_height_in_blocks = 15, compptr_width_in_blocks = 10 ) {

    comp_width = MCU_cols * compptr_h_samp_factor;

    _Pragma( "loopbound min 4 max 15" )
    for ( dst_blk_y = 0; dst_blk_y < compptr_height_in_blocks;
          dst_blk_y += compptr_v_samp_factor ) {

      offset_y = 0;
      _Pragma( "loopbound min 1 max 8" )
      for ( ; offset_y < compptr_v_samp_factor; offset_y++ ) {
        dst_blk_x = 0;
        _Pragma( "loopbound min 10 max 10" )
        for ( ; dst_blk_x < compptr_width_in_blocks;
              dst_blk_x += compptr_h_samp_factor ) {

          offset_x = 0;
          _Pragma( "loopbound min 1 max 2" )
          for ( ; offset_x < compptr_h_samp_factor; offset_x++ ) {

            src_ptr = cjpeg_transupp_input2;

            if ( dst_blk_x < comp_width ) {

              /* Block is within the mirrorable area. */
              dst_ptr = cjpeg_transupp_output_data2;

              _Pragma( "loopbound min 4 max 4" )
              for ( i = 0; i < DCTSIZE; i++ ) {
                j = 0;
                _Pragma( "loopbound min 8 max 8" )
                for ( ; j < DCTSIZE; j++ )
                  dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];

                i++;

                _Pragma( "loopbound min 8 max 8" )
                for ( j = 0; j < DCTSIZE; j++ )
                  dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
              }
            } else {
              /* Edge blocks are transposed but not mirrored. */
              dst_ptr = cjpeg_transupp_output_data2;

              _Pragma( "loopbound min 8 max 8" )
              for ( i = 0; i < DCTSIZE; i++ )
                j = 0;
              _Pragma( "loopbound min 8 max 8" )
              for ( ; j < DCTSIZE; j++ ) {
                if ( dst_blk_y < comp_width )
                  dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
              }
            }
          }
        }
      }
    }
  }
}


/*
  270 degree rotation is equivalent to
    1. Horizontal mirroring;
    2. Transposing the image.
  These two steps are merged into a single processing routine.
*/
void cjpeg_transupp_do_rot_270( j_compress_ptr dstinfo )
{
  unsigned int MCU_rows, comp_height, dst_blk_x, dst_blk_y;
  int ci, i, j, offset_x, offset_y;
  JCOEFPTR src_ptr, dst_ptr;


  /*
    Because of the horizontal mirror step, we can't process partial iMCUs at the
    (output) bottom edge properly. They just get transposed and not mirrored.
  */
  MCU_rows = dstinfo->image_height / ( dstinfo->max_v_samp_factor * DCTSIZE );

  int compptr_h_samp_factor = 2;
  int compptr_v_samp_factor = 8;
  unsigned int compptr_height_in_blocks = 29;
  unsigned int compptr_width_in_blocks = 19;

  _Pragma( "loopbound min 3 max 3" )
  for ( ci = 0; ci < dstinfo->num_components;
        ci++, compptr_h_samp_factor = compptr_v_samp_factor = 1,
        compptr_height_in_blocks = 15, compptr_width_in_blocks = 10 ) {

    comp_height = MCU_rows * compptr_v_samp_factor;

    _Pragma( "loopbound min 4 max 15" )
    for ( dst_blk_y = 0; dst_blk_y < compptr_height_in_blocks;
          dst_blk_y += compptr_v_samp_factor ) {

      offset_y = 0;
      _Pragma( "loopbound min 1 max 8" )
      for ( ; offset_y < compptr_v_samp_factor; offset_y++ ) {
        dst_blk_x = 0;
        _Pragma( "loopbound min 10 max 10" )
        for ( ; dst_blk_x < compptr_width_in_blocks;
              dst_blk_x += compptr_h_samp_factor ) {

          offset_x = 0;
          _Pragma( "loopbound min 1 max 2" )
          for ( ; offset_x < compptr_h_samp_factor; offset_x++ ) {

            dst_ptr = cjpeg_transupp_output_data3;

            if ( dst_blk_y < comp_height ) {

              /* Block is within the mirrorable area. */
              src_ptr = cjpeg_transupp_input3;

              _Pragma( "loopbound min 8 max 8" )
              for ( i = 0; i < DCTSIZE; i++ ) {
                j = 0;
                _Pragma( "loopbound min 4 max 4" )
                for ( ; j < DCTSIZE; j++ ) {
                  dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
                  j++;
                  dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
                }
              }
            } else {

              /* Edge blocks are transposed but not mirrored. */
              src_ptr = cjpeg_transupp_input3_2;

              _Pragma( "loopbound min 8 max 8" )
              for ( i = 0; i < DCTSIZE; i++ )
                j = 0;
              _Pragma( "loopbound min 8 max 8" )
              for ( ; j < DCTSIZE; j++ ) {
                if ( dst_blk_y < comp_height )
                  dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
              }
            }
          }
        }
      }
    }
  }
}


/*
  180 degree rotation is equivalent to
    1. Vertical mirroring;
    2. Horizontal mirroring.
  These two steps are merged into a single processing routine.
*/
void cjpeg_transupp_do_rot_180( j_compress_ptr dstinfo )
{
  unsigned int MCU_cols, MCU_rows, comp_width, comp_height, dst_blk_x,
               dst_blk_y;
  int ci, i, j, offset_y;
  JCOEFPTR src_ptr, dst_ptr;

  int compptr_h_samp_factor = 2;
  int compptr_v_samp_factor = 8;
  unsigned int compptr_width_in_blocks = 29;
  unsigned int compptr_height_in_blocks = 19;


  MCU_cols = dstinfo->image_width / ( dstinfo->max_h_samp_factor * DCTSIZE );
  MCU_rows = dstinfo->image_height / ( dstinfo->max_v_samp_factor * DCTSIZE );

  _Pragma( "loopbound min 3 max 3" )
  for ( ci = 0; ci < dstinfo->num_components; ci++,
        compptr_h_samp_factor = compptr_v_samp_factor = 1,
        compptr_width_in_blocks  = 15, compptr_height_in_blocks = 10 ) {

    comp_width  = MCU_cols * compptr_h_samp_factor;
    comp_height = MCU_rows * compptr_v_samp_factor;

    _Pragma( "loopbound min 3 max 10" )
    for ( dst_blk_y = 0; dst_blk_y < compptr_height_in_blocks;
          dst_blk_y += compptr_v_samp_factor ) {
      offset_y = 0;
      _Pragma( "loopbound min 1 max 8" )
      for ( ; offset_y < compptr_v_samp_factor; offset_y++ ) {
        if ( dst_blk_y < comp_height ) {

          /* Row is within the mirrorable area. */

          /* Process the blocks that can be mirrored both ways. */
          _Pragma( "loopbound min 14 max 28" )
          for ( dst_blk_x = 0; dst_blk_x < comp_width; dst_blk_x++ ) {
            dst_ptr = cjpeg_transupp_output_data4;
            src_ptr = cjpeg_transupp_input4;

            _Pragma( "loopbound min 4 max 4" )
            for ( i = 0; i < DCTSIZE; i += 2 ) {
              j = 0;
              /* For even row, negate every odd column. */
              _Pragma( "loopbound min 4 max 4" )
              for ( ; j < DCTSIZE; j += 2 ) {
                if ( dst_blk_x < comp_height ) {
                  *dst_ptr++ = *src_ptr++;
                  *dst_ptr++ = - *src_ptr++;
                }
              }

              j = 0;
              /* For odd row, negate every even column. */
              _Pragma( "loopbound min 4 max 4" )
              for ( ; j < DCTSIZE; j += 2 ) {
                if ( dst_blk_x < comp_height ) {
                  *dst_ptr++ = - *src_ptr++;
                  *dst_ptr++ = *src_ptr++;
                }
              }
            }
          }

          /* Any remaining right-edge blocks are only mirrored vertically. */
          _Pragma( "loopbound min 1 max 1" )
          for ( ; dst_blk_x < compptr_width_in_blocks; dst_blk_x++ ) {

            dst_ptr = cjpeg_transupp_output_data4;
            src_ptr = cjpeg_transupp_input4;
            _Pragma( "loopbound min 4 max 4" )
            for ( i = 0; i < DCTSIZE; i += 2 ) {
              j = 0;
              _Pragma( "loopbound min 8 max 8" )
              for ( ; j < DCTSIZE; j++ )
                *dst_ptr++ = *src_ptr++;
              j = 0;
              _Pragma( "loopbound min 8 max 8" )
              for ( ; j < DCTSIZE; j++ )
                *dst_ptr++ = - *src_ptr++;
            }
          }
        } else {

          /* Remaining rows are just mirrored horizontally. */
          dst_blk_x = 0;
          /* Process the blocks that can be mirrored. */
          _Pragma( "loopbound min 14 max 28" )
          do {
            dst_ptr = cjpeg_transupp_output_data4;
            src_ptr = cjpeg_transupp_input4;

            i = 0;
            _Pragma( "loopbound min 32 max 32" )
            while ( i < DCTSIZE2 ) {
              *dst_ptr++ = *src_ptr++;
              *dst_ptr++ = - *src_ptr++;
              i += 2;
              dst_ptr += 0;
            }
            dst_blk_x++;
            dst_ptr += 0;
          } while ( dst_blk_x < comp_width );

          /* Any remaining right-edge blocks are only copied. */
          _Pragma( "loopbound min 1 max 1" )
          for ( ; dst_blk_x < compptr_width_in_blocks; dst_blk_x++ ) {

            dst_ptr = cjpeg_transupp_output_data4;
            src_ptr = cjpeg_transupp_input4;
            _Pragma( "loopbound min 64 max 64" )
            do {
              *dst_ptr++ = *src_ptr++;
              i++;
            } while ( i < DCTSIZE2 );
          }
        }
      }
    }
  }
}


/*
  Transverse transpose is equivalent to
    1. 180 degree rotation;
    2. Transposition;
  or
    1. Horizontal mirroring;
    2. Transposition;
    3. Horizontal mirroring.
  These steps are merged into a single processing routine.
*/
void cjpeg_transupp_do_transverse( j_compress_ptr dstinfo )
{
  unsigned int MCU_cols, MCU_rows, comp_width, comp_height, dst_blk_x,
               dst_blk_y;
  int ci, i, j, offset_x, offset_y;
  JCOEFPTR src_ptr, dst_ptr;

  int compptr_h_samp_factor = 2;
  int compptr_v_samp_factor = 8;
  unsigned int compptr_height_in_blocks = 29;
  unsigned int compptr_width_in_blocks = 19;


  MCU_cols = dstinfo->image_width / ( dstinfo->max_h_samp_factor * DCTSIZE );
  MCU_rows = dstinfo->image_height / ( dstinfo->max_v_samp_factor * DCTSIZE );

  _Pragma( "loopbound min 3 max 3" )
  for ( ci = 0; ci < dstinfo->num_components; ci++,
        compptr_h_samp_factor = compptr_v_samp_factor = 1,
        compptr_height_in_blocks = 15, compptr_width_in_blocks  = 10 ) {

    comp_width  = MCU_cols * compptr_h_samp_factor;
    comp_height = MCU_rows * compptr_v_samp_factor;

    _Pragma( "loopbound min 4 max 15" )
    for ( dst_blk_y = 0; dst_blk_y < compptr_height_in_blocks;
          dst_blk_y += compptr_v_samp_factor ) {
      offset_y = 0;
      _Pragma( "loopbound min 1 max 8" )
      do {
        dst_blk_x = 0;
        _Pragma( "loopbound min 10 max 10" )
        do {
          offset_x = 0;
          _Pragma( "loopbound min 1 max 2" )
          for ( ; offset_x < compptr_h_samp_factor; offset_x++ ) {

            if ( dst_blk_y < comp_height ) {
              src_ptr = cjpeg_transupp_input5;

              if ( dst_blk_x < comp_width ) {
                /* Block is within the mirrorable area. */
                dst_ptr = cjpeg_transupp_output_data5;

                _Pragma( "loopbound min 4 max 4" )
                for ( i = 0; i < DCTSIZE; i++ ) {
                  j = 0;
                  _Pragma( "loopbound min 4 max 4" )
                  for ( ; j < DCTSIZE; j++ ) {
                    if ( dst_blk_y < comp_width )
                      dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
                    j++;
                    dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
                  }
                  i++;
                  _Pragma( "loopbound min 4 max 4" )
                  for ( j = 0; j < DCTSIZE; j++ ) {
                    if ( dst_blk_y < comp_width )
                      dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
                    j++;
                    dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
                  }
                }
              } else {
                /* Right-edge blocks are mirrored in y only */
                dst_ptr = cjpeg_transupp_output_data5;
                _Pragma( "loopbound min 8 max 8" )
                for ( i = 0; i < DCTSIZE; i++ ) {
                  j = 0;
                  _Pragma( "loopbound min 4 max 4" )
                  for ( ; j < DCTSIZE; j++ ) {
                    if ( dst_blk_y < comp_width )
                      dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
                    j++;
                    dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
                  }
                }
              }
            } else {
              src_ptr = cjpeg_transupp_input5_2;

              if ( dst_blk_x < comp_width ) {
                /* Bottom-edge blocks are mirrored in x only */
                dst_ptr = cjpeg_transupp_output_data5;

                _Pragma( "loopbound min 4 max 4" )
                for ( i = 0; i < DCTSIZE; i++ ) {
                  j = 0;
                  _Pragma( "loopbound min 8 max 8" )
                  for ( ; j < DCTSIZE; j++ )
                    dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
                  i++;
                  _Pragma( "loopbound min 8 max 8" )
                  for ( j = 0; j < DCTSIZE; j++ )
                    dst_ptr[ j * DCTSIZE + i ] = -src_ptr[ i * DCTSIZE + j ];
                }
              } else {
                /* At lower right corner, just transpose, no mirroring */
                dst_ptr = cjpeg_transupp_output_data5;
                _Pragma( "loopbound min 8 max 8" )
                for ( i = 0; i < DCTSIZE; i++ )
                  j = 0;
                _Pragma( "loopbound min 8 max 8" )
                for ( ; j < DCTSIZE; j++ )
                  dst_ptr[ j * DCTSIZE + i ] = src_ptr[ i * DCTSIZE + j ];
              }
            }
            dst_blk_x += compptr_h_samp_factor;
          }
        } while ( dst_blk_x < compptr_width_in_blocks );
        offset_y++;
      } while ( offset_y < compptr_v_samp_factor );
    }
  }
}


/*
  Main functions
*/

void _Pragma ( "entrypoint" ) cjpeg_transupp_main( void )
{
  cjpeg_transupp_dstinfo.image_width = 227;
  cjpeg_transupp_dstinfo.image_height = 149;

  cjpeg_transupp_do_flip_v( &cjpeg_transupp_dstinfo );

  cjpeg_transupp_dstinfo.image_width = 149;
  cjpeg_transupp_dstinfo.image_height = 227;

  cjpeg_transupp_do_rot_90( &cjpeg_transupp_dstinfo );
  cjpeg_transupp_do_rot_270( &cjpeg_transupp_dstinfo );

  cjpeg_transupp_dstinfo.image_width = 227;
  cjpeg_transupp_dstinfo.image_height = 149;

  cjpeg_transupp_do_rot_180( &cjpeg_transupp_dstinfo );

  cjpeg_transupp_dstinfo.image_width = 149;
  cjpeg_transupp_dstinfo.image_height = 227;

  cjpeg_transupp_do_transverse( &cjpeg_transupp_dstinfo );
}


int main(int argc, char **argv)
{
	SET_UP
	for (jobsComplete=-1; jobsComplete<maxJobs; jobsComplete++){
  		START_LOOP
		cjpeg_transupp_init();
  		cjpeg_transupp_main();
		STOP_LOOP
	}
	WRITE_TO_FILE
  return ( cjpeg_transupp_return() - 660 != 0 );
}