• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /*
2  * The copyright in this software is being made available under the 2-clauses
3  * BSD License, included below. This software may be subject to other third
4  * party and contributor rights, including patent rights, and no such rights
5  * are granted under this license.
6  *
7  * Copyright (c) 2002-2014, Universite catholique de Louvain (UCL), Belgium
8  * Copyright (c) 2002-2014, Professor Benoit Macq
9  * Copyright (c) 2001-2003, David Janssens
10  * Copyright (c) 2002-2003, Yannick Verschueren
11  * Copyright (c) 2003-2007, Francois-Olivier Devaux
12  * Copyright (c) 2003-2014, Antonin Descampe
13  * Copyright (c) 2005, Herve Drolon, FreeImage Team
14  * Copyright (c) 2008, 2011-2012, Centre National d'Etudes Spatiales (CNES), FR
15  * Copyright (c) 2012, CS Systemes d'Information, France
16  * All rights reserved.
17  *
18  * Redistribution and use in source and binary forms, with or without
19  * modification, are permitted provided that the following conditions
20  * are met:
21  * 1. Redistributions of source code must retain the above copyright
22  *    notice, this list of conditions and the following disclaimer.
23  * 2. Redistributions in binary form must reproduce the above copyright
24  *    notice, this list of conditions and the following disclaimer in the
25  *    documentation and/or other materials provided with the distribution.
26  *
27  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS `AS IS'
28  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30  * ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37  * POSSIBILITY OF SUCH DAMAGE.
38  */
39 
40 #ifdef __SSE__
41 #include <xmmintrin.h>
42 #endif
43 #ifdef __SSE2__
44 #include <emmintrin.h>
45 #endif
46 #ifdef __SSE4_1__
47 #include <smmintrin.h>
48 #endif
49 
50 #include "opj_includes.h"
51 
52 /* <summary> */
53 /* This table contains the norms of the basis function of the reversible MCT. */
54 /* </summary> */
55 static const OPJ_FLOAT64 opj_mct_norms[3] = { 1.732, .8292, .8292 };
56 
57 /* <summary> */
58 /* This table contains the norms of the basis function of the irreversible MCT. */
59 /* </summary> */
60 static const OPJ_FLOAT64 opj_mct_norms_real[3] = { 1.732, 1.805, 1.573 };
61 
opj_mct_get_mct_norms()62 const OPJ_FLOAT64 * opj_mct_get_mct_norms ()
63 {
64 	return opj_mct_norms;
65 }
66 
opj_mct_get_mct_norms_real()67 const OPJ_FLOAT64 * opj_mct_get_mct_norms_real ()
68 {
69 	return opj_mct_norms_real;
70 }
71 
72 /* <summary> */
73 /* Foward reversible MCT. */
74 /* </summary> */
75 #ifdef __SSE2__
opj_mct_encode(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)76 void opj_mct_encode(
77 		OPJ_INT32* restrict c0,
78 		OPJ_INT32* restrict c1,
79 		OPJ_INT32* restrict c2,
80 		OPJ_UINT32 n)
81 {
82 	OPJ_SIZE_T i;
83 	const OPJ_SIZE_T len = n;
84 
85 	for(i = 0; i < (len & ~3U); i += 4) {
86 		__m128i y, u, v;
87 		__m128i r = _mm_load_si128((const __m128i *)&(c0[i]));
88 		__m128i g = _mm_load_si128((const __m128i *)&(c1[i]));
89 		__m128i b = _mm_load_si128((const __m128i *)&(c2[i]));
90 		y = _mm_add_epi32(g, g);
91 		y = _mm_add_epi32(y, b);
92 		y = _mm_add_epi32(y, r);
93 		y = _mm_srai_epi32(y, 2);
94 		u = _mm_sub_epi32(b, g);
95 		v = _mm_sub_epi32(r, g);
96 		_mm_store_si128((__m128i *)&(c0[i]), y);
97 		_mm_store_si128((__m128i *)&(c1[i]), u);
98 		_mm_store_si128((__m128i *)&(c2[i]), v);
99 	}
100 
101 	for(; i < len; ++i) {
102 		OPJ_INT32 r = c0[i];
103 		OPJ_INT32 g = c1[i];
104 		OPJ_INT32 b = c2[i];
105 		OPJ_INT32 y = (r + (g * 2) + b) >> 2;
106 		OPJ_INT32 u = b - g;
107 		OPJ_INT32 v = r - g;
108 		c0[i] = y;
109 		c1[i] = u;
110 		c2[i] = v;
111 	}
112 }
113 #else
opj_mct_encode(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)114 void opj_mct_encode(
115 		OPJ_INT32* restrict c0,
116 		OPJ_INT32* restrict c1,
117 		OPJ_INT32* restrict c2,
118 		OPJ_UINT32 n)
119 {
120 	OPJ_SIZE_T i;
121 	const OPJ_SIZE_T len = n;
122 
123 	for(i = 0; i < len; ++i) {
124 		OPJ_INT32 r = c0[i];
125 		OPJ_INT32 g = c1[i];
126 		OPJ_INT32 b = c2[i];
127 		OPJ_INT32 y = (r + (g * 2) + b) >> 2;
128 		OPJ_INT32 u = b - g;
129 		OPJ_INT32 v = r - g;
130 		c0[i] = y;
131 		c1[i] = u;
132 		c2[i] = v;
133 	}
134 }
135 #endif
136 
137 /* <summary> */
138 /* Inverse reversible MCT. */
139 /* </summary> */
140 #ifdef __SSE2__
opj_mct_decode(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)141 void opj_mct_decode(
142 		OPJ_INT32* restrict c0,
143 		OPJ_INT32* restrict c1,
144 		OPJ_INT32* restrict c2,
145 		OPJ_UINT32 n)
146 {
147 	OPJ_SIZE_T i;
148 	const OPJ_SIZE_T len = n;
149 
150 	for(i = 0; i < (len & ~3U); i += 4) {
151 		__m128i r, g, b;
152 		__m128i y = _mm_load_si128((const __m128i *)&(c0[i]));
153 		__m128i u = _mm_load_si128((const __m128i *)&(c1[i]));
154 		__m128i v = _mm_load_si128((const __m128i *)&(c2[i]));
155 		g = y;
156 		g = _mm_sub_epi32(g, _mm_srai_epi32(_mm_add_epi32(u, v), 2));
157 		r = _mm_add_epi32(v, g);
158 		b = _mm_add_epi32(u, g);
159 		_mm_store_si128((__m128i *)&(c0[i]), r);
160 		_mm_store_si128((__m128i *)&(c1[i]), g);
161 		_mm_store_si128((__m128i *)&(c2[i]), b);
162 	}
163 	for (; i < len; ++i) {
164 		OPJ_INT32 y = c0[i];
165 		OPJ_INT32 u = c1[i];
166 		OPJ_INT32 v = c2[i];
167 		OPJ_INT32 g = y - ((u + v) >> 2);
168 		OPJ_INT32 r = v + g;
169 		OPJ_INT32 b = u + g;
170 		c0[i] = r;
171 		c1[i] = g;
172 		c2[i] = b;
173 	}
174 }
175 #else
opj_mct_decode(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)176 void opj_mct_decode(
177 		OPJ_INT32* restrict c0,
178 		OPJ_INT32* restrict c1,
179 		OPJ_INT32* restrict c2,
180 		OPJ_UINT32 n)
181 {
182 	OPJ_UINT32 i;
183 	for (i = 0; i < n; ++i) {
184 		OPJ_INT32 y = c0[i];
185 		OPJ_INT32 u = c1[i];
186 		OPJ_INT32 v = c2[i];
187 		OPJ_INT32 g = y - ((u + v) >> 2);
188 		OPJ_INT32 r = v + g;
189 		OPJ_INT32 b = u + g;
190 		c0[i] = r;
191 		c1[i] = g;
192 		c2[i] = b;
193 	}
194 }
195 #endif
196 
197 /* <summary> */
198 /* Get norm of basis function of reversible MCT. */
199 /* </summary> */
opj_mct_getnorm(OPJ_UINT32 compno)200 OPJ_FLOAT64 opj_mct_getnorm(OPJ_UINT32 compno) {
201 	return opj_mct_norms[compno];
202 }
203 
204 /* <summary> */
205 /* Foward irreversible MCT. */
206 /* </summary> */
207 #ifdef __SSE4_1__
opj_mct_encode_real(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)208 void opj_mct_encode_real(
209 												 OPJ_INT32* restrict c0,
210 												 OPJ_INT32* restrict c1,
211 												 OPJ_INT32* restrict c2,
212 												 OPJ_UINT32 n)
213 {
214 	OPJ_SIZE_T i;
215 	const OPJ_SIZE_T len = n;
216 
217 	const __m128i ry = _mm_set1_epi32(2449);
218 	const __m128i gy = _mm_set1_epi32(4809);
219 	const __m128i by = _mm_set1_epi32(934);
220 	const __m128i ru = _mm_set1_epi32(1382);
221 	const __m128i gu = _mm_set1_epi32(2714);
222 	/* const __m128i bu = _mm_set1_epi32(4096); */
223 	/* const __m128i rv = _mm_set1_epi32(4096); */
224 	const __m128i gv = _mm_set1_epi32(3430);
225 	const __m128i bv = _mm_set1_epi32(666);
226 	const __m128i mulround = _mm_shuffle_epi32(_mm_cvtsi32_si128(4096), _MM_SHUFFLE(1, 0, 1, 0));
227 
228 	for(i = 0; i < (len & ~3U); i += 4) {
229 		__m128i lo, hi;
230 		__m128i y, u, v;
231 		__m128i r = _mm_load_si128((const __m128i *)&(c0[i]));
232 		__m128i g = _mm_load_si128((const __m128i *)&(c1[i]));
233 		__m128i b = _mm_load_si128((const __m128i *)&(c2[i]));
234 
235 		lo = r;
236 		hi = _mm_shuffle_epi32(r, _MM_SHUFFLE(3, 3, 1, 1));
237 		lo = _mm_mul_epi32(lo, ry);
238 		hi = _mm_mul_epi32(hi, ry);
239 		lo = _mm_add_epi64(lo, mulround);
240 		hi = _mm_add_epi64(hi, mulround);
241 		lo = _mm_srli_epi64(lo, 13);
242 		hi = _mm_slli_epi64(hi, 32-13);
243 		y = _mm_blend_epi16(lo, hi, 0xCC);
244 
245 		lo = g;
246 		hi = _mm_shuffle_epi32(g, _MM_SHUFFLE(3, 3, 1, 1));
247 		lo = _mm_mul_epi32(lo, gy);
248 		hi = _mm_mul_epi32(hi, gy);
249 		lo = _mm_add_epi64(lo, mulround);
250 		hi = _mm_add_epi64(hi, mulround);
251 		lo = _mm_srli_epi64(lo, 13);
252 		hi = _mm_slli_epi64(hi, 32-13);
253 		y = _mm_add_epi32(y, _mm_blend_epi16(lo, hi, 0xCC));
254 
255 		lo = b;
256 		hi = _mm_shuffle_epi32(b, _MM_SHUFFLE(3, 3, 1, 1));
257 		lo = _mm_mul_epi32(lo, by);
258 		hi = _mm_mul_epi32(hi, by);
259 		lo = _mm_add_epi64(lo, mulround);
260 		hi = _mm_add_epi64(hi, mulround);
261 		lo = _mm_srli_epi64(lo, 13);
262 		hi = _mm_slli_epi64(hi, 32-13);
263 		y = _mm_add_epi32(y, _mm_blend_epi16(lo, hi, 0xCC));
264 		_mm_store_si128((__m128i *)&(c0[i]), y);
265 
266 		/*lo = b;
267 		hi = _mm_shuffle_epi32(b, _MM_SHUFFLE(3, 3, 1, 1));
268 		lo = _mm_mul_epi32(lo, mulround);
269 		hi = _mm_mul_epi32(hi, mulround);*/
270 		lo = _mm_cvtepi32_epi64(_mm_shuffle_epi32(b, _MM_SHUFFLE(3, 2, 2, 0)));
271 		hi = _mm_cvtepi32_epi64(_mm_shuffle_epi32(b, _MM_SHUFFLE(3, 2, 3, 1)));
272 		lo = _mm_slli_epi64(lo, 12);
273 		hi = _mm_slli_epi64(hi, 12);
274 		lo = _mm_add_epi64(lo, mulround);
275 		hi = _mm_add_epi64(hi, mulround);
276 		lo = _mm_srli_epi64(lo, 13);
277 		hi = _mm_slli_epi64(hi, 32-13);
278 		u = _mm_blend_epi16(lo, hi, 0xCC);
279 
280 		lo = r;
281 		hi = _mm_shuffle_epi32(r, _MM_SHUFFLE(3, 3, 1, 1));
282 		lo = _mm_mul_epi32(lo, ru);
283 		hi = _mm_mul_epi32(hi, ru);
284 		lo = _mm_add_epi64(lo, mulround);
285 		hi = _mm_add_epi64(hi, mulround);
286 		lo = _mm_srli_epi64(lo, 13);
287 		hi = _mm_slli_epi64(hi, 32-13);
288 		u = _mm_sub_epi32(u, _mm_blend_epi16(lo, hi, 0xCC));
289 
290 		lo = g;
291 		hi = _mm_shuffle_epi32(g, _MM_SHUFFLE(3, 3, 1, 1));
292 		lo = _mm_mul_epi32(lo, gu);
293 		hi = _mm_mul_epi32(hi, gu);
294 		lo = _mm_add_epi64(lo, mulround);
295 		hi = _mm_add_epi64(hi, mulround);
296 		lo = _mm_srli_epi64(lo, 13);
297 		hi = _mm_slli_epi64(hi, 32-13);
298 		u = _mm_sub_epi32(u, _mm_blend_epi16(lo, hi, 0xCC));
299 		_mm_store_si128((__m128i *)&(c1[i]), u);
300 
301 		/*lo = r;
302 		hi = _mm_shuffle_epi32(r, _MM_SHUFFLE(3, 3, 1, 1));
303 		lo = _mm_mul_epi32(lo, mulround);
304 		hi = _mm_mul_epi32(hi, mulround);*/
305 		lo = _mm_cvtepi32_epi64(_mm_shuffle_epi32(r, _MM_SHUFFLE(3, 2, 2, 0)));
306 		hi = _mm_cvtepi32_epi64(_mm_shuffle_epi32(r, _MM_SHUFFLE(3, 2, 3, 1)));
307 		lo = _mm_slli_epi64(lo, 12);
308 		hi = _mm_slli_epi64(hi, 12);
309 		lo = _mm_add_epi64(lo, mulround);
310 		hi = _mm_add_epi64(hi, mulround);
311 		lo = _mm_srli_epi64(lo, 13);
312 		hi = _mm_slli_epi64(hi, 32-13);
313 		v = _mm_blend_epi16(lo, hi, 0xCC);
314 
315 		lo = g;
316 		hi = _mm_shuffle_epi32(g, _MM_SHUFFLE(3, 3, 1, 1));
317 		lo = _mm_mul_epi32(lo, gv);
318 		hi = _mm_mul_epi32(hi, gv);
319 		lo = _mm_add_epi64(lo, mulround);
320 		hi = _mm_add_epi64(hi, mulround);
321 		lo = _mm_srli_epi64(lo, 13);
322 		hi = _mm_slli_epi64(hi, 32-13);
323 		v = _mm_sub_epi32(v, _mm_blend_epi16(lo, hi, 0xCC));
324 
325 		lo = b;
326 		hi = _mm_shuffle_epi32(b, _MM_SHUFFLE(3, 3, 1, 1));
327 		lo = _mm_mul_epi32(lo, bv);
328 		hi = _mm_mul_epi32(hi, bv);
329 		lo = _mm_add_epi64(lo, mulround);
330 		hi = _mm_add_epi64(hi, mulround);
331 		lo = _mm_srli_epi64(lo, 13);
332 		hi = _mm_slli_epi64(hi, 32-13);
333 		v = _mm_sub_epi32(v, _mm_blend_epi16(lo, hi, 0xCC));
334 		_mm_store_si128((__m128i *)&(c2[i]), v);
335 	}
336 	for(; i < len; ++i) {
337 		OPJ_INT32 r = c0[i];
338 		OPJ_INT32 g = c1[i];
339 		OPJ_INT32 b = c2[i];
340 		OPJ_INT32 y =  opj_int_fix_mul(r, 2449) + opj_int_fix_mul(g, 4809) + opj_int_fix_mul(b, 934);
341 		OPJ_INT32 u = -opj_int_fix_mul(r, 1382) - opj_int_fix_mul(g, 2714) + opj_int_fix_mul(b, 4096);
342 		OPJ_INT32 v =  opj_int_fix_mul(r, 4096) - opj_int_fix_mul(g, 3430) - opj_int_fix_mul(b, 666);
343 		c0[i] = y;
344 		c1[i] = u;
345 		c2[i] = v;
346 	}
347 }
348 #else
opj_mct_encode_real(OPJ_INT32 * restrict c0,OPJ_INT32 * restrict c1,OPJ_INT32 * restrict c2,OPJ_UINT32 n)349 void opj_mct_encode_real(
350 		OPJ_INT32* restrict c0,
351 		OPJ_INT32* restrict c1,
352 		OPJ_INT32* restrict c2,
353 		OPJ_UINT32 n)
354 {
355 	OPJ_UINT32 i;
356 	for(i = 0; i < n; ++i) {
357 		OPJ_INT32 r = c0[i];
358 		OPJ_INT32 g = c1[i];
359 		OPJ_INT32 b = c2[i];
360 		OPJ_INT32 y =  opj_int_fix_mul(r, 2449) + opj_int_fix_mul(g, 4809) + opj_int_fix_mul(b, 934);
361 		OPJ_INT32 u = -opj_int_fix_mul(r, 1382) - opj_int_fix_mul(g, 2714) + opj_int_fix_mul(b, 4096);
362 		OPJ_INT32 v =  opj_int_fix_mul(r, 4096) - opj_int_fix_mul(g, 3430) - opj_int_fix_mul(b, 666);
363 		c0[i] = y;
364 		c1[i] = u;
365 		c2[i] = v;
366 	}
367 }
368 #endif
369 
370 /* <summary> */
371 /* Inverse irreversible MCT. */
372 /* </summary> */
opj_mct_decode_real(OPJ_FLOAT32 * restrict c0,OPJ_FLOAT32 * restrict c1,OPJ_FLOAT32 * restrict c2,OPJ_UINT32 n)373 void opj_mct_decode_real(
374 		OPJ_FLOAT32* restrict c0,
375 		OPJ_FLOAT32* restrict c1,
376 		OPJ_FLOAT32* restrict c2,
377 		OPJ_UINT32 n)
378 {
379 	OPJ_UINT32 i;
380 #ifdef __SSE__
381 	__m128 vrv, vgu, vgv, vbu;
382 	vrv = _mm_set1_ps(1.402f);
383 	vgu = _mm_set1_ps(0.34413f);
384 	vgv = _mm_set1_ps(0.71414f);
385 	vbu = _mm_set1_ps(1.772f);
386 	for (i = 0; i < (n >> 3); ++i) {
387 		__m128 vy, vu, vv;
388 		__m128 vr, vg, vb;
389 
390 		vy = _mm_load_ps(c0);
391 		vu = _mm_load_ps(c1);
392 		vv = _mm_load_ps(c2);
393 		vr = _mm_add_ps(vy, _mm_mul_ps(vv, vrv));
394 		vg = _mm_sub_ps(_mm_sub_ps(vy, _mm_mul_ps(vu, vgu)), _mm_mul_ps(vv, vgv));
395 		vb = _mm_add_ps(vy, _mm_mul_ps(vu, vbu));
396 		_mm_store_ps(c0, vr);
397 		_mm_store_ps(c1, vg);
398 		_mm_store_ps(c2, vb);
399 		c0 += 4;
400 		c1 += 4;
401 		c2 += 4;
402 
403 		vy = _mm_load_ps(c0);
404 		vu = _mm_load_ps(c1);
405 		vv = _mm_load_ps(c2);
406 		vr = _mm_add_ps(vy, _mm_mul_ps(vv, vrv));
407 		vg = _mm_sub_ps(_mm_sub_ps(vy, _mm_mul_ps(vu, vgu)), _mm_mul_ps(vv, vgv));
408 		vb = _mm_add_ps(vy, _mm_mul_ps(vu, vbu));
409 		_mm_store_ps(c0, vr);
410 		_mm_store_ps(c1, vg);
411 		_mm_store_ps(c2, vb);
412 		c0 += 4;
413 		c1 += 4;
414 		c2 += 4;
415 	}
416 	n &= 7;
417 #endif
418 	for(i = 0; i < n; ++i) {
419 		OPJ_FLOAT32 y = c0[i];
420 		OPJ_FLOAT32 u = c1[i];
421 		OPJ_FLOAT32 v = c2[i];
422 		OPJ_FLOAT32 r = y + (v * 1.402f);
423 		OPJ_FLOAT32 g = y - (u * 0.34413f) - (v * (0.71414f));
424 		OPJ_FLOAT32 b = y + (u * 1.772f);
425 		c0[i] = r;
426 		c1[i] = g;
427 		c2[i] = b;
428 	}
429 }
430 
431 /* <summary> */
432 /* Get norm of basis function of irreversible MCT. */
433 /* </summary> */
opj_mct_getnorm_real(OPJ_UINT32 compno)434 OPJ_FLOAT64 opj_mct_getnorm_real(OPJ_UINT32 compno) {
435 	return opj_mct_norms_real[compno];
436 }
437 
438 
opj_mct_encode_custom(OPJ_BYTE * pCodingdata,OPJ_UINT32 n,OPJ_BYTE ** pData,OPJ_UINT32 pNbComp,OPJ_UINT32 isSigned)439 OPJ_BOOL opj_mct_encode_custom(
440 					   OPJ_BYTE * pCodingdata,
441 					   OPJ_UINT32 n,
442 					   OPJ_BYTE ** pData,
443 					   OPJ_UINT32 pNbComp,
444 					   OPJ_UINT32 isSigned)
445 {
446 	OPJ_FLOAT32 * lMct = (OPJ_FLOAT32 *) pCodingdata;
447 	OPJ_UINT32 i;
448 	OPJ_UINT32 j;
449 	OPJ_UINT32 k;
450 	OPJ_UINT32 lNbMatCoeff = pNbComp * pNbComp;
451 	OPJ_INT32 * lCurrentData = 00;
452 	OPJ_INT32 * lCurrentMatrix = 00;
453 	OPJ_INT32 ** lData = (OPJ_INT32 **) pData;
454 	OPJ_UINT32 lMultiplicator = 1 << 13;
455 	OPJ_INT32 * lMctPtr;
456 
457     OPJ_ARG_NOT_USED(isSigned);
458 
459 	lCurrentData = (OPJ_INT32 *) opj_malloc((pNbComp + lNbMatCoeff) * sizeof(OPJ_INT32));
460 	if (! lCurrentData) {
461 		return OPJ_FALSE;
462 	}
463 
464 	lCurrentMatrix = lCurrentData + pNbComp;
465 
466 	for (i =0;i<lNbMatCoeff;++i) {
467 		lCurrentMatrix[i] = (OPJ_INT32) (*(lMct++) * (OPJ_FLOAT32)lMultiplicator);
468 	}
469 
470 	for (i = 0; i < n; ++i)  {
471 		lMctPtr = lCurrentMatrix;
472 		for (j=0;j<pNbComp;++j) {
473 			lCurrentData[j] = (*(lData[j]));
474 		}
475 
476 		for (j=0;j<pNbComp;++j) {
477 			*(lData[j]) = 0;
478 			for (k=0;k<pNbComp;++k) {
479 				*(lData[j]) += opj_int_fix_mul(*lMctPtr, lCurrentData[k]);
480 				++lMctPtr;
481 			}
482 
483 			++lData[j];
484 		}
485 	}
486 
487 	opj_free(lCurrentData);
488 
489 	return OPJ_TRUE;
490 }
491 
opj_mct_decode_custom(OPJ_BYTE * pDecodingData,OPJ_UINT32 n,OPJ_BYTE ** pData,OPJ_UINT32 pNbComp,OPJ_UINT32 isSigned)492 OPJ_BOOL opj_mct_decode_custom(
493 					   OPJ_BYTE * pDecodingData,
494 					   OPJ_UINT32 n,
495 					   OPJ_BYTE ** pData,
496 					   OPJ_UINT32 pNbComp,
497 					   OPJ_UINT32 isSigned)
498 {
499 	OPJ_FLOAT32 * lMct;
500 	OPJ_UINT32 i;
501 	OPJ_UINT32 j;
502 	OPJ_UINT32 k;
503 
504 	OPJ_FLOAT32 * lCurrentData = 00;
505 	OPJ_FLOAT32 * lCurrentResult = 00;
506 	OPJ_FLOAT32 ** lData = (OPJ_FLOAT32 **) pData;
507 
508     OPJ_ARG_NOT_USED(isSigned);
509 
510 	lCurrentData = (OPJ_FLOAT32 *) opj_malloc (2 * pNbComp * sizeof(OPJ_FLOAT32));
511 	if (! lCurrentData) {
512 		return OPJ_FALSE;
513 	}
514 	lCurrentResult = lCurrentData + pNbComp;
515 
516 	for (i = 0; i < n; ++i) {
517 		lMct = (OPJ_FLOAT32 *) pDecodingData;
518 		for (j=0;j<pNbComp;++j) {
519 			lCurrentData[j] = (OPJ_FLOAT32) (*(lData[j]));
520 		}
521 		for (j=0;j<pNbComp;++j) {
522 			lCurrentResult[j] = 0;
523 			for	(k=0;k<pNbComp;++k)	{
524 				lCurrentResult[j] += *(lMct++) * lCurrentData[k];
525 			}
526 			*(lData[j]++) = (OPJ_FLOAT32) (lCurrentResult[j]);
527 		}
528 	}
529 	opj_free(lCurrentData);
530 	return OPJ_TRUE;
531 }
532 
opj_calculate_norms(OPJ_FLOAT64 * pNorms,OPJ_UINT32 pNbComps,OPJ_FLOAT32 * pMatrix)533 void opj_calculate_norms(	OPJ_FLOAT64 * pNorms,
534 							OPJ_UINT32 pNbComps,
535 							OPJ_FLOAT32 * pMatrix)
536 {
537 	OPJ_UINT32 i,j,lIndex;
538 	OPJ_FLOAT32 lCurrentValue;
539 	OPJ_FLOAT64 * lNorms = (OPJ_FLOAT64 *) pNorms;
540 	OPJ_FLOAT32 * lMatrix = (OPJ_FLOAT32 *) pMatrix;
541 
542 	for	(i=0;i<pNbComps;++i) {
543 		lNorms[i] = 0;
544 		lIndex = i;
545 
546 		for	(j=0;j<pNbComps;++j) {
547 			lCurrentValue = lMatrix[lIndex];
548 			lIndex += pNbComps;
549 			lNorms[i] += lCurrentValue * lCurrentValue;
550 		}
551 		lNorms[i] = sqrt(lNorms[i]);
552 	}
553 }
554