/src/libjpeg-turbo.main/jcdctmgr.c
| Line | Count | Source (jump to first uncovered line) | 
| 1 |  | /* | 
| 2 |  |  * jcdctmgr.c | 
| 3 |  |  * | 
| 4 |  |  * This file was part of the Independent JPEG Group's software: | 
| 5 |  |  * Copyright (C) 1994-1996, Thomas G. Lane. | 
| 6 |  |  * libjpeg-turbo Modifications: | 
| 7 |  |  * Copyright (C) 1999-2006, MIYASAKA Masaru. | 
| 8 |  |  * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB | 
| 9 |  |  * Copyright (C) 2011, 2014-2015, 2022, D. R. Commander. | 
| 10 |  |  * For conditions of distribution and use, see the accompanying README.ijg | 
| 11 |  |  * file. | 
| 12 |  |  * | 
| 13 |  |  * This file contains the forward-DCT management logic. | 
| 14 |  |  * This code selects a particular DCT implementation to be used, | 
| 15 |  |  * and it performs related housekeeping chores including coefficient | 
| 16 |  |  * quantization. | 
| 17 |  |  */ | 
| 18 |  |  | 
| 19 |  | #define JPEG_INTERNALS | 
| 20 |  | #include "jinclude.h" | 
| 21 |  | #include "jpeglib.h" | 
| 22 |  | #include "jdct.h"               /* Private declarations for DCT subsystem */ | 
| 23 |  | #include "jsimddct.h" | 
| 24 |  |  | 
| 25 |  |  | 
| 26 |  | /* Private subobject for this module */ | 
| 27 |  |  | 
| 28 |  | typedef void (*forward_DCT_method_ptr) (DCTELEM *data); | 
| 29 |  | typedef void (*float_DCT_method_ptr) (FAST_FLOAT *data); | 
| 30 |  |  | 
| 31 |  | typedef void (*convsamp_method_ptr) (_JSAMPARRAY sample_data, | 
| 32 |  |                                      JDIMENSION start_col, | 
| 33 |  |                                      DCTELEM *workspace); | 
| 34 |  | typedef void (*float_convsamp_method_ptr) (_JSAMPARRAY sample_data, | 
| 35 |  |                                            JDIMENSION start_col, | 
| 36 |  |                                            FAST_FLOAT *workspace); | 
| 37 |  |  | 
| 38 |  | typedef void (*quantize_method_ptr) (JCOEFPTR coef_block, DCTELEM *divisors, | 
| 39 |  |                                      DCTELEM *workspace); | 
| 40 |  | typedef void (*float_quantize_method_ptr) (JCOEFPTR coef_block, | 
| 41 |  |                                            FAST_FLOAT *divisors, | 
| 42 |  |                                            FAST_FLOAT *workspace); | 
| 43 |  |  | 
| 44 |  | METHODDEF(void) quantize(JCOEFPTR, DCTELEM *, DCTELEM *); | 
| 45 |  |  | 
| 46 |  | typedef struct { | 
| 47 |  |   struct jpeg_forward_dct pub;  /* public fields */ | 
| 48 |  |  | 
| 49 |  |   /* Pointer to the DCT routine actually in use */ | 
| 50 |  |   forward_DCT_method_ptr dct; | 
| 51 |  |   convsamp_method_ptr convsamp; | 
| 52 |  |   quantize_method_ptr quantize; | 
| 53 |  |  | 
| 54 |  |   /* The actual post-DCT divisors --- not identical to the quant table | 
| 55 |  |    * entries, because of scaling (especially for an unnormalized DCT). | 
| 56 |  |    * Each table is given in normal array order. | 
| 57 |  |    */ | 
| 58 |  |   DCTELEM *divisors[NUM_QUANT_TBLS]; | 
| 59 |  |  | 
| 60 |  |   /* work area for FDCT subroutine */ | 
| 61 |  |   DCTELEM *workspace; | 
| 62 |  |  | 
| 63 |  | #ifdef DCT_FLOAT_SUPPORTED | 
| 64 |  |   /* Same as above for the floating-point case. */ | 
| 65 |  |   float_DCT_method_ptr float_dct; | 
| 66 |  |   float_convsamp_method_ptr float_convsamp; | 
| 67 |  |   float_quantize_method_ptr float_quantize; | 
| 68 |  |   FAST_FLOAT *float_divisors[NUM_QUANT_TBLS]; | 
| 69 |  |   FAST_FLOAT *float_workspace; | 
| 70 |  | #endif | 
| 71 |  | } my_fdct_controller; | 
| 72 |  |  | 
| 73 |  | typedef my_fdct_controller *my_fdct_ptr; | 
| 74 |  |  | 
| 75 |  |  | 
| 76 |  | #if BITS_IN_JSAMPLE == 8 | 
| 77 |  |  | 
| 78 |  | /* | 
| 79 |  |  * Find the highest bit in an integer through binary search. | 
| 80 |  |  */ | 
| 81 |  |  | 
| 82 |  | LOCAL(int) | 
| 83 |  | flss(UINT16 val) | 
| 84 | 0 | { | 
| 85 | 0 |   int bit; | 
| 86 |  | 
 | 
| 87 | 0 |   bit = 16; | 
| 88 |  | 
 | 
| 89 | 0 |   if (!val) | 
| 90 | 0 |     return 0; | 
| 91 |  |  | 
| 92 | 0 |   if (!(val & 0xff00)) { | 
| 93 | 0 |     bit -= 8; | 
| 94 | 0 |     val <<= 8; | 
| 95 | 0 |   } | 
| 96 | 0 |   if (!(val & 0xf000)) { | 
| 97 | 0 |     bit -= 4; | 
| 98 | 0 |     val <<= 4; | 
| 99 | 0 |   } | 
| 100 | 0 |   if (!(val & 0xc000)) { | 
| 101 | 0 |     bit -= 2; | 
| 102 | 0 |     val <<= 2; | 
| 103 | 0 |   } | 
| 104 | 0 |   if (!(val & 0x8000)) { | 
| 105 | 0 |     bit -= 1; | 
| 106 | 0 |     val <<= 1; | 
| 107 | 0 |   } | 
| 108 |  | 
 | 
| 109 | 0 |   return bit; | 
| 110 | 0 | } | 
| 111 |  |  | 
| 112 |  |  | 
| 113 |  | /* | 
| 114 |  |  * Compute values to do a division using reciprocal. | 
| 115 |  |  * | 
| 116 |  |  * This implementation is based on an algorithm described in | 
| 117 |  |  *   "How to optimize for the Pentium family of microprocessors" | 
| 118 |  |  *   (http://www.agner.org/assem/). | 
| 119 |  |  * More information about the basic algorithm can be found in | 
| 120 |  |  * the paper "Integer Division Using Reciprocals" by Robert Alverson. | 
| 121 |  |  * | 
| 122 |  |  * The basic idea is to replace x/d by x * d^-1. In order to store | 
| 123 |  |  * d^-1 with enough precision we shift it left a few places. It turns | 
| 124 |  |  * out that this algoright gives just enough precision, and also fits | 
| 125 |  |  * into DCTELEM: | 
| 126 |  |  * | 
| 127 |  |  *   b = (the number of significant bits in divisor) - 1 | 
| 128 |  |  *   r = (word size) + b | 
| 129 |  |  *   f = 2^r / divisor | 
| 130 |  |  * | 
| 131 |  |  * f will not be an integer for most cases, so we need to compensate | 
| 132 |  |  * for the rounding error introduced: | 
| 133 |  |  * | 
| 134 |  |  *   no fractional part: | 
| 135 |  |  * | 
| 136 |  |  *       result = input >> r | 
| 137 |  |  * | 
| 138 |  |  *   fractional part of f < 0.5: | 
| 139 |  |  * | 
| 140 |  |  *       round f down to nearest integer | 
| 141 |  |  *       result = ((input + 1) * f) >> r | 
| 142 |  |  * | 
| 143 |  |  *   fractional part of f > 0.5: | 
| 144 |  |  * | 
| 145 |  |  *       round f up to nearest integer | 
| 146 |  |  *       result = (input * f) >> r | 
| 147 |  |  * | 
| 148 |  |  * This is the original algorithm that gives truncated results. But we | 
| 149 |  |  * want properly rounded results, so we replace "input" with | 
| 150 |  |  * "input + divisor/2". | 
| 151 |  |  * | 
| 152 |  |  * In order to allow SIMD implementations we also tweak the values to | 
| 153 |  |  * allow the same calculation to be made at all times: | 
| 154 |  |  * | 
| 155 |  |  *   dctbl[0] = f rounded to nearest integer | 
| 156 |  |  *   dctbl[1] = divisor / 2 (+ 1 if fractional part of f < 0.5) | 
| 157 |  |  *   dctbl[2] = 1 << ((word size) * 2 - r) | 
| 158 |  |  *   dctbl[3] = r - (word size) | 
| 159 |  |  * | 
| 160 |  |  * dctbl[2] is for stupid instruction sets where the shift operation | 
| 161 |  |  * isn't member wise (e.g. MMX). | 
| 162 |  |  * | 
| 163 |  |  * The reason dctbl[2] and dctbl[3] reduce the shift with (word size) | 
| 164 |  |  * is that most SIMD implementations have a "multiply and store top | 
| 165 |  |  * half" operation. | 
| 166 |  |  * | 
| 167 |  |  * Lastly, we store each of the values in their own table instead | 
| 168 |  |  * of in a consecutive manner, yet again in order to allow SIMD | 
| 169 |  |  * routines. | 
| 170 |  |  */ | 
| 171 |  |  | 
| 172 |  | LOCAL(int) | 
| 173 |  | compute_reciprocal(UINT16 divisor, DCTELEM *dtbl) | 
| 174 | 0 | { | 
| 175 | 0 |   UDCTELEM2 fq, fr; | 
| 176 | 0 |   UDCTELEM c; | 
| 177 | 0 |   int b, r; | 
| 178 |  | 
 | 
| 179 | 0 |   if (divisor == 1) { | 
| 180 |  |     /* divisor == 1 means unquantized, so these reciprocal/correction/shift | 
| 181 |  |      * values will cause the C quantization algorithm to act like the | 
| 182 |  |      * identity function.  Since only the C quantization algorithm is used in | 
| 183 |  |      * these cases, the scale value is irrelevant. | 
| 184 |  |      */ | 
| 185 | 0 |     dtbl[DCTSIZE2 * 0] = (DCTELEM)1;                        /* reciprocal */ | 
| 186 | 0 |     dtbl[DCTSIZE2 * 1] = (DCTELEM)0;                        /* correction */ | 
| 187 | 0 |     dtbl[DCTSIZE2 * 2] = (DCTELEM)1;                        /* scale */ | 
| 188 | 0 |     dtbl[DCTSIZE2 * 3] = -(DCTELEM)(sizeof(DCTELEM) * 8);   /* shift */ | 
| 189 | 0 |     return 0; | 
| 190 | 0 |   } | 
| 191 |  |  | 
| 192 | 0 |   b = flss(divisor) - 1; | 
| 193 | 0 |   r  = sizeof(DCTELEM) * 8 + b; | 
| 194 |  | 
 | 
| 195 | 0 |   fq = ((UDCTELEM2)1 << r) / divisor; | 
| 196 | 0 |   fr = ((UDCTELEM2)1 << r) % divisor; | 
| 197 |  | 
 | 
| 198 | 0 |   c = divisor / 2;                      /* for rounding */ | 
| 199 |  | 
 | 
| 200 | 0 |   if (fr == 0) {                        /* divisor is power of two */ | 
| 201 |  |     /* fq will be one bit too large to fit in DCTELEM, so adjust */ | 
| 202 | 0 |     fq >>= 1; | 
| 203 | 0 |     r--; | 
| 204 | 0 |   } else if (fr <= (divisor / 2U)) {    /* fractional part is < 0.5 */ | 
| 205 | 0 |     c++; | 
| 206 | 0 |   } else {                              /* fractional part is > 0.5 */ | 
| 207 | 0 |     fq++; | 
| 208 | 0 |   } | 
| 209 |  | 
 | 
| 210 | 0 |   dtbl[DCTSIZE2 * 0] = (DCTELEM)fq;     /* reciprocal */ | 
| 211 | 0 |   dtbl[DCTSIZE2 * 1] = (DCTELEM)c;      /* correction + roundfactor */ | 
| 212 | 0 | #ifdef WITH_SIMD | 
| 213 | 0 |   dtbl[DCTSIZE2 * 2] = (DCTELEM)(1 << (sizeof(DCTELEM) * 8 * 2 - r)); /* scale */ | 
| 214 |  | #else | 
| 215 |  |   dtbl[DCTSIZE2 * 2] = 1; | 
| 216 |  | #endif | 
| 217 | 0 |   dtbl[DCTSIZE2 * 3] = (DCTELEM)r - sizeof(DCTELEM) * 8; /* shift */ | 
| 218 |  | 
 | 
| 219 | 0 |   if (r <= 16) return 0; | 
| 220 | 0 |   else return 1; | 
| 221 | 0 | } | 
| 222 |  |  | 
| 223 |  | #endif | 
| 224 |  |  | 
| 225 |  |  | 
| 226 |  | /* | 
| 227 |  |  * Initialize for a processing pass. | 
| 228 |  |  * Verify that all referenced Q-tables are present, and set up | 
| 229 |  |  * the divisor table for each one. | 
| 230 |  |  * In the current implementation, DCT of all components is done during | 
| 231 |  |  * the first pass, even if only some components will be output in the | 
| 232 |  |  * first scan.  Hence all components should be examined here. | 
| 233 |  |  */ | 
| 234 |  |  | 
| 235 |  | METHODDEF(void) | 
| 236 |  | start_pass_fdctmgr(j_compress_ptr cinfo) | 
| 237 | 0 | { | 
| 238 | 0 |   my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct; | 
| 239 | 0 |   int ci, qtblno, i; | 
| 240 | 0 |   jpeg_component_info *compptr; | 
| 241 | 0 |   JQUANT_TBL *qtbl; | 
| 242 | 0 |   DCTELEM *dtbl; | 
| 243 |  | 
 | 
| 244 | 0 |   for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; | 
| 245 | 0 |        ci++, compptr++) { | 
| 246 | 0 |     qtblno = compptr->quant_tbl_no; | 
| 247 |  |     /* Make sure specified quantization table is present */ | 
| 248 | 0 |     if (qtblno < 0 || qtblno >= NUM_QUANT_TBLS || | 
| 249 | 0 |         cinfo->quant_tbl_ptrs[qtblno] == NULL) | 
| 250 | 0 |       ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, qtblno); | 
| 251 | 0 |     qtbl = cinfo->quant_tbl_ptrs[qtblno]; | 
| 252 |  |     /* Compute divisors for this quant table */ | 
| 253 |  |     /* We may do this more than once for same table, but it's not a big deal */ | 
| 254 | 0 |     switch (cinfo->dct_method) { | 
| 255 | 0 | #ifdef DCT_ISLOW_SUPPORTED | 
| 256 | 0 |     case JDCT_ISLOW: | 
| 257 |  |       /* For LL&M IDCT method, divisors are equal to raw quantization | 
| 258 |  |        * coefficients multiplied by 8 (to counteract scaling). | 
| 259 |  |        */ | 
| 260 | 0 |       if (fdct->divisors[qtblno] == NULL) { | 
| 261 | 0 |         fdct->divisors[qtblno] = (DCTELEM *) | 
| 262 | 0 |           (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 263 | 0 |                                       (DCTSIZE2 * 4) * sizeof(DCTELEM)); | 
| 264 | 0 |       } | 
| 265 | 0 |       dtbl = fdct->divisors[qtblno]; | 
| 266 | 0 |       for (i = 0; i < DCTSIZE2; i++) { | 
| 267 |  | #if BITS_IN_JSAMPLE == 8 | 
| 268 |  | #ifdef WITH_SIMD | 
| 269 |  |         if (!compute_reciprocal(qtbl->quantval[i] << 3, &dtbl[i]) && | 
| 270 |  |             fdct->quantize == jsimd_quantize) | 
| 271 |  |           fdct->quantize = quantize; | 
| 272 |  | #else | 
| 273 |  |         compute_reciprocal(qtbl->quantval[i] << 3, &dtbl[i]); | 
| 274 |  | #endif | 
| 275 |  | #else | 
| 276 | 0 |         dtbl[i] = ((DCTELEM)qtbl->quantval[i]) << 3; | 
| 277 | 0 | #endif | 
| 278 | 0 |       } | 
| 279 | 0 |       break; | 
| 280 | 0 | #endif | 
| 281 | 0 | #ifdef DCT_IFAST_SUPPORTED | 
| 282 | 0 |     case JDCT_IFAST: | 
| 283 | 0 |       { | 
| 284 |  |         /* For AA&N IDCT method, divisors are equal to quantization | 
| 285 |  |          * coefficients scaled by scalefactor[row]*scalefactor[col], where | 
| 286 |  |          *   scalefactor[0] = 1 | 
| 287 |  |          *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7 | 
| 288 |  |          * We apply a further scale factor of 8. | 
| 289 |  |          */ | 
| 290 | 0 | #define CONST_BITS  14 | 
| 291 | 0 |         static const INT16 aanscales[DCTSIZE2] = { | 
| 292 |  |           /* precomputed values scaled up by 14 bits */ | 
| 293 | 0 |           16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520, | 
| 294 | 0 |           22725, 31521, 29692, 26722, 22725, 17855, 12299,  6270, | 
| 295 | 0 |           21407, 29692, 27969, 25172, 21407, 16819, 11585,  5906, | 
| 296 | 0 |           19266, 26722, 25172, 22654, 19266, 15137, 10426,  5315, | 
| 297 | 0 |           16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520, | 
| 298 | 0 |           12873, 17855, 16819, 15137, 12873, 10114,  6967,  3552, | 
| 299 | 0 |            8867, 12299, 11585, 10426,  8867,  6967,  4799,  2446, | 
| 300 | 0 |            4520,  6270,  5906,  5315,  4520,  3552,  2446,  1247 | 
| 301 | 0 |         }; | 
| 302 | 0 |         SHIFT_TEMPS | 
| 303 |  | 
 | 
| 304 | 0 |         if (fdct->divisors[qtblno] == NULL) { | 
| 305 | 0 |           fdct->divisors[qtblno] = (DCTELEM *) | 
| 306 | 0 |             (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 307 | 0 |                                         (DCTSIZE2 * 4) * sizeof(DCTELEM)); | 
| 308 | 0 |         } | 
| 309 | 0 |         dtbl = fdct->divisors[qtblno]; | 
| 310 | 0 |         for (i = 0; i < DCTSIZE2; i++) { | 
| 311 |  | #if BITS_IN_JSAMPLE == 8 | 
| 312 |  | #ifdef WITH_SIMD | 
| 313 |  |           if (!compute_reciprocal( | 
| 314 |  |                 DESCALE(MULTIPLY16V16((JLONG)qtbl->quantval[i], | 
| 315 |  |                                       (JLONG)aanscales[i]), | 
| 316 |  |                         CONST_BITS - 3), &dtbl[i]) && | 
| 317 |  |               fdct->quantize == jsimd_quantize) | 
| 318 |  |             fdct->quantize = quantize; | 
| 319 |  | #else | 
| 320 |  |           compute_reciprocal( | 
| 321 |  |             DESCALE(MULTIPLY16V16((JLONG)qtbl->quantval[i], | 
| 322 |  |                                   (JLONG)aanscales[i]), | 
| 323 |  |                     CONST_BITS-3), &dtbl[i]); | 
| 324 |  | #endif | 
| 325 |  | #else | 
| 326 | 0 |           dtbl[i] = (DCTELEM) | 
| 327 | 0 |             DESCALE(MULTIPLY16V16((JLONG)qtbl->quantval[i], | 
| 328 | 0 |                                   (JLONG)aanscales[i]), | 
| 329 | 0 |                     CONST_BITS - 3); | 
| 330 | 0 | #endif | 
| 331 | 0 |         } | 
| 332 | 0 |       } | 
| 333 | 0 |       break; | 
| 334 | 0 | #endif | 
| 335 | 0 | #ifdef DCT_FLOAT_SUPPORTED | 
| 336 | 0 |     case JDCT_FLOAT: | 
| 337 | 0 |       { | 
| 338 |  |         /* For float AA&N IDCT method, divisors are equal to quantization | 
| 339 |  |          * coefficients scaled by scalefactor[row]*scalefactor[col], where | 
| 340 |  |          *   scalefactor[0] = 1 | 
| 341 |  |          *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7 | 
| 342 |  |          * We apply a further scale factor of 8. | 
| 343 |  |          * What's actually stored is 1/divisor so that the inner loop can | 
| 344 |  |          * use a multiplication rather than a division. | 
| 345 |  |          */ | 
| 346 | 0 |         FAST_FLOAT *fdtbl; | 
| 347 | 0 |         int row, col; | 
| 348 | 0 |         static const double aanscalefactor[DCTSIZE] = { | 
| 349 | 0 |           1.0, 1.387039845, 1.306562965, 1.175875602, | 
| 350 | 0 |           1.0, 0.785694958, 0.541196100, 0.275899379 | 
| 351 | 0 |         }; | 
| 352 |  | 
 | 
| 353 | 0 |         if (fdct->float_divisors[qtblno] == NULL) { | 
| 354 | 0 |           fdct->float_divisors[qtblno] = (FAST_FLOAT *) | 
| 355 | 0 |             (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 356 | 0 |                                         DCTSIZE2 * sizeof(FAST_FLOAT)); | 
| 357 | 0 |         } | 
| 358 | 0 |         fdtbl = fdct->float_divisors[qtblno]; | 
| 359 | 0 |         i = 0; | 
| 360 | 0 |         for (row = 0; row < DCTSIZE; row++) { | 
| 361 | 0 |           for (col = 0; col < DCTSIZE; col++) { | 
| 362 | 0 |             fdtbl[i] = (FAST_FLOAT) | 
| 363 | 0 |               (1.0 / (((double)qtbl->quantval[i] * | 
| 364 | 0 |                        aanscalefactor[row] * aanscalefactor[col] * 8.0))); | 
| 365 | 0 |             i++; | 
| 366 | 0 |           } | 
| 367 | 0 |         } | 
| 368 | 0 |       } | 
| 369 | 0 |       break; | 
| 370 | 0 | #endif | 
| 371 | 0 |     default: | 
| 372 | 0 |       ERREXIT(cinfo, JERR_NOT_COMPILED); | 
| 373 | 0 |       break; | 
| 374 | 0 |     } | 
| 375 | 0 |   } | 
| 376 | 0 | } | 
| 377 |  |  | 
| 378 |  |  | 
| 379 |  | /* | 
| 380 |  |  * Load data into workspace, applying unsigned->signed conversion. | 
| 381 |  |  */ | 
| 382 |  |  | 
| 383 |  | METHODDEF(void) | 
| 384 |  | convsamp(_JSAMPARRAY sample_data, JDIMENSION start_col, DCTELEM *workspace) | 
| 385 | 0 | { | 
| 386 | 0 |   register DCTELEM *workspaceptr; | 
| 387 | 0 |   register _JSAMPROW elemptr; | 
| 388 | 0 |   register int elemr; | 
| 389 |  | 
 | 
| 390 | 0 |   workspaceptr = workspace; | 
| 391 | 0 |   for (elemr = 0; elemr < DCTSIZE; elemr++) { | 
| 392 | 0 |     elemptr = sample_data[elemr] + start_col; | 
| 393 |  | 
 | 
| 394 | 0 | #if DCTSIZE == 8                /* unroll the inner loop */ | 
| 395 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 396 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 397 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 398 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 399 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 400 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 401 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 402 | 0 |     *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 403 |  | #else | 
| 404 |  |     { | 
| 405 |  |       register int elemc; | 
| 406 |  |       for (elemc = DCTSIZE; elemc > 0; elemc--) | 
| 407 |  |         *workspaceptr++ = (*elemptr++) - _CENTERJSAMPLE; | 
| 408 |  |     } | 
| 409 |  | #endif | 
| 410 | 0 |   } | 
| 411 | 0 | } | 
| 412 |  |  | 
| 413 |  |  | 
| 414 |  | /* | 
| 415 |  |  * Quantize/descale the coefficients, and store into coef_blocks[]. | 
| 416 |  |  */ | 
| 417 |  |  | 
| 418 |  | METHODDEF(void) | 
| 419 |  | quantize(JCOEFPTR coef_block, DCTELEM *divisors, DCTELEM *workspace) | 
| 420 | 0 | { | 
| 421 | 0 |   int i; | 
| 422 | 0 |   DCTELEM temp; | 
| 423 | 0 |   JCOEFPTR output_ptr = coef_block; | 
| 424 |  | 
 | 
| 425 |  | #if BITS_IN_JSAMPLE == 8 | 
| 426 |  |  | 
| 427 |  |   UDCTELEM recip, corr; | 
| 428 |  |   int shift; | 
| 429 |  |   UDCTELEM2 product; | 
| 430 |  |  | 
| 431 |  |   for (i = 0; i < DCTSIZE2; i++) { | 
| 432 |  |     temp = workspace[i]; | 
| 433 |  |     recip = divisors[i + DCTSIZE2 * 0]; | 
| 434 |  |     corr =  divisors[i + DCTSIZE2 * 1]; | 
| 435 |  |     shift = divisors[i + DCTSIZE2 * 3]; | 
| 436 |  |  | 
| 437 |  |     if (temp < 0) { | 
| 438 |  |       temp = -temp; | 
| 439 |  |       product = (UDCTELEM2)(temp + corr) * recip; | 
| 440 |  |       product >>= shift + sizeof(DCTELEM) * 8; | 
| 441 |  |       temp = (DCTELEM)product; | 
| 442 |  |       temp = -temp; | 
| 443 |  |     } else { | 
| 444 |  |       product = (UDCTELEM2)(temp + corr) * recip; | 
| 445 |  |       product >>= shift + sizeof(DCTELEM) * 8; | 
| 446 |  |       temp = (DCTELEM)product; | 
| 447 |  |     } | 
| 448 |  |     output_ptr[i] = (JCOEF)temp; | 
| 449 |  |   } | 
| 450 |  |  | 
| 451 |  | #else | 
| 452 |  | 
 | 
| 453 | 0 |   register DCTELEM qval; | 
| 454 |  | 
 | 
| 455 | 0 |   for (i = 0; i < DCTSIZE2; i++) { | 
| 456 | 0 |     qval = divisors[i]; | 
| 457 | 0 |     temp = workspace[i]; | 
| 458 |  |     /* Divide the coefficient value by qval, ensuring proper rounding. | 
| 459 |  |      * Since C does not specify the direction of rounding for negative | 
| 460 |  |      * quotients, we have to force the dividend positive for portability. | 
| 461 |  |      * | 
| 462 |  |      * In most files, at least half of the output values will be zero | 
| 463 |  |      * (at default quantization settings, more like three-quarters...) | 
| 464 |  |      * so we should ensure that this case is fast.  On many machines, | 
| 465 |  |      * a comparison is enough cheaper than a divide to make a special test | 
| 466 |  |      * a win.  Since both inputs will be nonnegative, we need only test | 
| 467 |  |      * for a < b to discover whether a/b is 0. | 
| 468 |  |      * If your machine's division is fast enough, define FAST_DIVIDE. | 
| 469 |  |      */ | 
| 470 |  | #ifdef FAST_DIVIDE | 
| 471 |  | #define DIVIDE_BY(a, b)  a /= b | 
| 472 |  | #else | 
| 473 | 0 | #define DIVIDE_BY(a, b)  if (a >= b) a /= b;  else a = 0 | 
| 474 | 0 | #endif | 
| 475 | 0 |     if (temp < 0) { | 
| 476 | 0 |       temp = -temp; | 
| 477 | 0 |       temp += qval >> 1;        /* for rounding */ | 
| 478 | 0 |       DIVIDE_BY(temp, qval); | 
| 479 | 0 |       temp = -temp; | 
| 480 | 0 |     } else { | 
| 481 | 0 |       temp += qval >> 1;        /* for rounding */ | 
| 482 | 0 |       DIVIDE_BY(temp, qval); | 
| 483 | 0 |     } | 
| 484 | 0 |     output_ptr[i] = (JCOEF)temp; | 
| 485 | 0 |   } | 
| 486 |  | 
 | 
| 487 | 0 | #endif | 
| 488 |  | 
 | 
| 489 | 0 | } | 
| 490 |  |  | 
| 491 |  |  | 
| 492 |  | /* | 
| 493 |  |  * Perform forward DCT on one or more blocks of a component. | 
| 494 |  |  * | 
| 495 |  |  * The input samples are taken from the sample_data[] array starting at | 
| 496 |  |  * position start_row/start_col, and moving to the right for any additional | 
| 497 |  |  * blocks. The quantized coefficients are returned in coef_blocks[]. | 
| 498 |  |  */ | 
| 499 |  |  | 
| 500 |  | METHODDEF(void) | 
| 501 |  | forward_DCT(j_compress_ptr cinfo, jpeg_component_info *compptr, | 
| 502 |  |             _JSAMPARRAY sample_data, JBLOCKROW coef_blocks, | 
| 503 |  |             JDIMENSION start_row, JDIMENSION start_col, JDIMENSION num_blocks) | 
| 504 |  | /* This version is used for integer DCT implementations. */ | 
| 505 | 0 | { | 
| 506 |  |   /* This routine is heavily used, so it's worth coding it tightly. */ | 
| 507 | 0 |   my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct; | 
| 508 | 0 |   DCTELEM *divisors = fdct->divisors[compptr->quant_tbl_no]; | 
| 509 | 0 |   DCTELEM *workspace; | 
| 510 | 0 |   JDIMENSION bi; | 
| 511 |  |  | 
| 512 |  |   /* Make sure the compiler doesn't look up these every pass */ | 
| 513 | 0 |   forward_DCT_method_ptr do_dct = fdct->dct; | 
| 514 | 0 |   convsamp_method_ptr do_convsamp = fdct->convsamp; | 
| 515 | 0 |   quantize_method_ptr do_quantize = fdct->quantize; | 
| 516 | 0 |   workspace = fdct->workspace; | 
| 517 |  | 
 | 
| 518 | 0 |   sample_data += start_row;     /* fold in the vertical offset once */ | 
| 519 |  | 
 | 
| 520 | 0 |   for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) { | 
| 521 |  |     /* Load data into workspace, applying unsigned->signed conversion */ | 
| 522 | 0 |     (*do_convsamp) (sample_data, start_col, workspace); | 
| 523 |  |  | 
| 524 |  |     /* Perform the DCT */ | 
| 525 | 0 |     (*do_dct) (workspace); | 
| 526 |  |  | 
| 527 |  |     /* Quantize/descale the coefficients, and store into coef_blocks[] */ | 
| 528 | 0 |     (*do_quantize) (coef_blocks[bi], divisors, workspace); | 
| 529 | 0 |   } | 
| 530 | 0 | } | 
| 531 |  |  | 
| 532 |  |  | 
| 533 |  | #ifdef DCT_FLOAT_SUPPORTED | 
| 534 |  |  | 
| 535 |  | METHODDEF(void) | 
| 536 |  | convsamp_float(_JSAMPARRAY sample_data, JDIMENSION start_col, | 
| 537 |  |                FAST_FLOAT *workspace) | 
| 538 | 0 | { | 
| 539 | 0 |   register FAST_FLOAT *workspaceptr; | 
| 540 | 0 |   register _JSAMPROW elemptr; | 
| 541 | 0 |   register int elemr; | 
| 542 |  | 
 | 
| 543 | 0 |   workspaceptr = workspace; | 
| 544 | 0 |   for (elemr = 0; elemr < DCTSIZE; elemr++) { | 
| 545 | 0 |     elemptr = sample_data[elemr] + start_col; | 
| 546 | 0 | #if DCTSIZE == 8                /* unroll the inner loop */ | 
| 547 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 548 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 549 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 550 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 551 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 552 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 553 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 554 | 0 |     *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 555 |  | #else | 
| 556 |  |     { | 
| 557 |  |       register int elemc; | 
| 558 |  |       for (elemc = DCTSIZE; elemc > 0; elemc--) | 
| 559 |  |         *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - _CENTERJSAMPLE); | 
| 560 |  |     } | 
| 561 |  | #endif | 
| 562 | 0 |   } | 
| 563 | 0 | } | 
| 564 |  |  | 
| 565 |  |  | 
| 566 |  | METHODDEF(void) | 
| 567 |  | quantize_float(JCOEFPTR coef_block, FAST_FLOAT *divisors, | 
| 568 |  |                FAST_FLOAT *workspace) | 
| 569 | 0 | { | 
| 570 | 0 |   register FAST_FLOAT temp; | 
| 571 | 0 |   register int i; | 
| 572 | 0 |   register JCOEFPTR output_ptr = coef_block; | 
| 573 |  | 
 | 
| 574 | 0 |   for (i = 0; i < DCTSIZE2; i++) { | 
| 575 |  |     /* Apply the quantization and scaling factor */ | 
| 576 | 0 |     temp = workspace[i] * divisors[i]; | 
| 577 |  |  | 
| 578 |  |     /* Round to nearest integer. | 
| 579 |  |      * Since C does not specify the direction of rounding for negative | 
| 580 |  |      * quotients, we have to force the dividend positive for portability. | 
| 581 |  |      * The maximum coefficient size is +-16K (for 12-bit data), so this | 
| 582 |  |      * code should work for either 16-bit or 32-bit ints. | 
| 583 |  |      */ | 
| 584 | 0 |     output_ptr[i] = (JCOEF)((int)(temp + (FAST_FLOAT)16384.5) - 16384); | 
| 585 | 0 |   } | 
| 586 | 0 | } | 
| 587 |  |  | 
| 588 |  |  | 
| 589 |  | METHODDEF(void) | 
| 590 |  | forward_DCT_float(j_compress_ptr cinfo, jpeg_component_info *compptr, | 
| 591 |  |                   _JSAMPARRAY sample_data, JBLOCKROW coef_blocks, | 
| 592 |  |                   JDIMENSION start_row, JDIMENSION start_col, | 
| 593 |  |                   JDIMENSION num_blocks) | 
| 594 |  | /* This version is used for floating-point DCT implementations. */ | 
| 595 | 0 | { | 
| 596 |  |   /* This routine is heavily used, so it's worth coding it tightly. */ | 
| 597 | 0 |   my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct; | 
| 598 | 0 |   FAST_FLOAT *divisors = fdct->float_divisors[compptr->quant_tbl_no]; | 
| 599 | 0 |   FAST_FLOAT *workspace; | 
| 600 | 0 |   JDIMENSION bi; | 
| 601 |  |  | 
| 602 |  |  | 
| 603 |  |   /* Make sure the compiler doesn't look up these every pass */ | 
| 604 | 0 |   float_DCT_method_ptr do_dct = fdct->float_dct; | 
| 605 | 0 |   float_convsamp_method_ptr do_convsamp = fdct->float_convsamp; | 
| 606 | 0 |   float_quantize_method_ptr do_quantize = fdct->float_quantize; | 
| 607 | 0 |   workspace = fdct->float_workspace; | 
| 608 |  | 
 | 
| 609 | 0 |   sample_data += start_row;     /* fold in the vertical offset once */ | 
| 610 |  | 
 | 
| 611 | 0 |   for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) { | 
| 612 |  |     /* Load data into workspace, applying unsigned->signed conversion */ | 
| 613 | 0 |     (*do_convsamp) (sample_data, start_col, workspace); | 
| 614 |  |  | 
| 615 |  |     /* Perform the DCT */ | 
| 616 | 0 |     (*do_dct) (workspace); | 
| 617 |  |  | 
| 618 |  |     /* Quantize/descale the coefficients, and store into coef_blocks[] */ | 
| 619 | 0 |     (*do_quantize) (coef_blocks[bi], divisors, workspace); | 
| 620 | 0 |   } | 
| 621 | 0 | } | 
| 622 |  |  | 
| 623 |  | #endif /* DCT_FLOAT_SUPPORTED */ | 
| 624 |  |  | 
| 625 |  |  | 
| 626 |  | /* | 
| 627 |  |  * Initialize FDCT manager. | 
| 628 |  |  */ | 
| 629 |  |  | 
| 630 |  | GLOBAL(void) | 
| 631 |  | _jinit_forward_dct(j_compress_ptr cinfo) | 
| 632 | 0 | { | 
| 633 | 0 |   my_fdct_ptr fdct; | 
| 634 | 0 |   int i; | 
| 635 |  | 
 | 
| 636 | 0 |   if (cinfo->data_precision != BITS_IN_JSAMPLE) | 
| 637 | 0 |     ERREXIT1(cinfo, JERR_BAD_PRECISION, cinfo->data_precision); | 
| 638 |  | 
 | 
| 639 | 0 |   fdct = (my_fdct_ptr) | 
| 640 | 0 |     (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 641 | 0 |                                 sizeof(my_fdct_controller)); | 
| 642 | 0 |   cinfo->fdct = (struct jpeg_forward_dct *)fdct; | 
| 643 | 0 |   fdct->pub.start_pass = start_pass_fdctmgr; | 
| 644 |  |  | 
| 645 |  |   /* First determine the DCT... */ | 
| 646 | 0 |   switch (cinfo->dct_method) { | 
| 647 | 0 | #ifdef DCT_ISLOW_SUPPORTED | 
| 648 | 0 |   case JDCT_ISLOW: | 
| 649 | 0 |     fdct->pub._forward_DCT = forward_DCT; | 
| 650 |  | #ifdef WITH_SIMD | 
| 651 | 0 |     if (jsimd_can_fdct_islow()) | 
| 652 | 0 |       fdct->dct = jsimd_fdct_islow; | 
| 653 | 0 |     else | 
| 654 | 0 | #endif | 
| 655 | 0 |       fdct->dct = _jpeg_fdct_islow; | 
| 656 | 0 |     break; | 
| 657 | 0 | #endif | 
| 658 | 0 | #ifdef DCT_IFAST_SUPPORTED | 
| 659 | 0 |   case JDCT_IFAST: | 
| 660 | 0 |     fdct->pub._forward_DCT = forward_DCT; | 
| 661 |  | #ifdef WITH_SIMD | 
| 662 | 0 |     if (jsimd_can_fdct_ifast()) | 
| 663 | 0 |       fdct->dct = jsimd_fdct_ifast; | 
| 664 | 0 |     else | 
| 665 | 0 | #endif | 
| 666 | 0 |       fdct->dct = _jpeg_fdct_ifast; | 
| 667 | 0 |     break; | 
| 668 | 0 | #endif | 
| 669 | 0 | #ifdef DCT_FLOAT_SUPPORTED | 
| 670 | 0 |   case JDCT_FLOAT: | 
| 671 | 0 |     fdct->pub._forward_DCT = forward_DCT_float; | 
| 672 |  | #ifdef WITH_SIMD | 
| 673 | 0 |     if (jsimd_can_fdct_float()) | 
| 674 | 0 |       fdct->float_dct = jsimd_fdct_float; | 
| 675 | 0 |     else | 
| 676 | 0 | #endif | 
| 677 | 0 |       fdct->float_dct = jpeg_fdct_float; | 
| 678 | 0 |     break; | 
| 679 | 0 | #endif | 
| 680 | 0 |   default: | 
| 681 | 0 |     ERREXIT(cinfo, JERR_NOT_COMPILED); | 
| 682 | 0 |     break; | 
| 683 | 0 |   } | 
| 684 |  |  | 
| 685 |  |   /* ...then the supporting stages. */ | 
| 686 | 0 |   switch (cinfo->dct_method) { | 
| 687 | 0 | #ifdef DCT_ISLOW_SUPPORTED | 
| 688 | 0 |   case JDCT_ISLOW: | 
| 689 | 0 | #endif | 
| 690 | 0 | #ifdef DCT_IFAST_SUPPORTED | 
| 691 | 0 |   case JDCT_IFAST: | 
| 692 | 0 | #endif | 
| 693 | 0 | #if defined(DCT_ISLOW_SUPPORTED) || defined(DCT_IFAST_SUPPORTED) | 
| 694 |  | #ifdef WITH_SIMD | 
| 695 | 0 |     if (jsimd_can_convsamp()) | 
| 696 | 0 |       fdct->convsamp = jsimd_convsamp; | 
| 697 | 0 |     else | 
| 698 | 0 | #endif | 
| 699 | 0 |       fdct->convsamp = convsamp; | 
| 700 |  | #ifdef WITH_SIMD | 
| 701 | 0 |     if (jsimd_can_quantize()) | 
| 702 | 0 |       fdct->quantize = jsimd_quantize; | 
| 703 | 0 |     else | 
| 704 | 0 | #endif | 
| 705 | 0 |       fdct->quantize = quantize; | 
| 706 | 0 |     break; | 
| 707 | 0 | #endif | 
| 708 | 0 | #ifdef DCT_FLOAT_SUPPORTED | 
| 709 | 0 |   case JDCT_FLOAT: | 
| 710 |  | #ifdef WITH_SIMD | 
| 711 | 0 |     if (jsimd_can_convsamp_float()) | 
| 712 | 0 |       fdct->float_convsamp = jsimd_convsamp_float; | 
| 713 | 0 |     else | 
| 714 | 0 | #endif | 
| 715 | 0 |       fdct->float_convsamp = convsamp_float; | 
| 716 |  | #ifdef WITH_SIMD | 
| 717 | 0 |     if (jsimd_can_quantize_float()) | 
| 718 | 0 |       fdct->float_quantize = jsimd_quantize_float; | 
| 719 | 0 |     else | 
| 720 | 0 | #endif | 
| 721 | 0 |       fdct->float_quantize = quantize_float; | 
| 722 | 0 |     break; | 
| 723 | 0 | #endif | 
| 724 | 0 |   default: | 
| 725 | 0 |     ERREXIT(cinfo, JERR_NOT_COMPILED); | 
| 726 | 0 |     break; | 
| 727 | 0 |   } | 
| 728 |  |  | 
| 729 |  |   /* Allocate workspace memory */ | 
| 730 | 0 | #ifdef DCT_FLOAT_SUPPORTED | 
| 731 | 0 |   if (cinfo->dct_method == JDCT_FLOAT) | 
| 732 | 0 |     fdct->float_workspace = (FAST_FLOAT *) | 
| 733 | 0 |       (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 734 | 0 |                                   sizeof(FAST_FLOAT) * DCTSIZE2); | 
| 735 | 0 |   else | 
| 736 | 0 | #endif | 
| 737 | 0 |     fdct->workspace = (DCTELEM *) | 
| 738 | 0 |       (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE, | 
| 739 | 0 |                                   sizeof(DCTELEM) * DCTSIZE2); | 
| 740 |  |  | 
| 741 |  |   /* Mark divisor tables unallocated */ | 
| 742 | 0 |   for (i = 0; i < NUM_QUANT_TBLS; i++) { | 
| 743 | 0 |     fdct->divisors[i] = NULL; | 
| 744 | 0 | #ifdef DCT_FLOAT_SUPPORTED | 
| 745 | 0 |     fdct->float_divisors[i] = NULL; | 
| 746 | 0 | #endif | 
| 747 | 0 |   } | 
| 748 | 0 | } Unexecuted instantiation: j12init_forward_dctUnexecuted instantiation: jinit_forward_dct |