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) 2017, IntoPix SA <contact@intopix.com>
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS `AS IS'
20 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
23 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29 * POSSIBILITY OF SUCH DAMAGE.
30 */
31
32#include "opj_includes.h"
33
34
35struct opj_sparse_array_int32 {
36 OPJ_UINT32 width;
37 OPJ_UINT32 height;
38 OPJ_UINT32 block_width;
39 OPJ_UINT32 block_height;
40 OPJ_UINT32 block_count_hor;
41 OPJ_UINT32 block_count_ver;
42 OPJ_INT32** data_blocks;
43};
44
45opj_sparse_array_int32_t* opj_sparse_array_int32_create(OPJ_UINT32 width,
46 OPJ_UINT32 height,
47 OPJ_UINT32 block_width,
48 OPJ_UINT32 block_height)
49{
50 opj_sparse_array_int32_t* sa;
51
52 if (width == 0 || height == 0 || block_width == 0 || block_height == 0) {
53 return NULL;
54 }
55 if (block_width > ((OPJ_UINT32)~0U) / block_height / sizeof(OPJ_INT32)) {
56 return NULL;
57 }
58
59 sa = (opj_sparse_array_int32_t*) opj_calloc(1,
60 sizeof(opj_sparse_array_int32_t));
61 sa->width = width;
62 sa->height = height;
63 sa->block_width = block_width;
64 sa->block_height = block_height;
65 sa->block_count_hor = opj_uint_ceildiv(width, block_width);
66 sa->block_count_ver = opj_uint_ceildiv(height, block_height);
67 if (sa->block_count_hor > ((OPJ_UINT32)~0U) / sa->block_count_ver) {
68 opj_free(sa);
69 return NULL;
70 }
71 sa->data_blocks = (OPJ_INT32**) opj_calloc(sizeof(OPJ_INT32*),
72 sa->block_count_hor * sa->block_count_ver);
73 if (sa->data_blocks == NULL) {
74 opj_free(sa);
75 return NULL;
76 }
77
78 return sa;
79}
80
81void opj_sparse_array_int32_free(opj_sparse_array_int32_t* sa)
82{
83 if (sa) {
84 OPJ_UINT32 i;
85 for (i = 0; i < sa->block_count_hor * sa->block_count_ver; i++) {
86 if (sa->data_blocks[i]) {
87 opj_free(sa->data_blocks[i]);
88 }
89 }
90 opj_free(sa->data_blocks);
91 opj_free(sa);
92 }
93}
94
95OPJ_BOOL opj_sparse_array_is_region_valid(const opj_sparse_array_int32_t* sa,
96 OPJ_UINT32 x0,
97 OPJ_UINT32 y0,
98 OPJ_UINT32 x1,
99 OPJ_UINT32 y1)
100{
101 return !(x0 >= sa->width || x1 <= x0 || x1 > sa->width ||
102 y0 >= sa->height || y1 <= y0 || y1 > sa->height);
103}
104
105static OPJ_BOOL opj_sparse_array_int32_read_or_write(
106 const opj_sparse_array_int32_t* sa,
107 OPJ_UINT32 x0,
108 OPJ_UINT32 y0,
109 OPJ_UINT32 x1,
110 OPJ_UINT32 y1,
111 OPJ_INT32* buf,
112 OPJ_UINT32 buf_col_stride,
113 OPJ_UINT32 buf_line_stride,
114 OPJ_BOOL forgiving,
115 OPJ_BOOL is_read_op)
116{
117 OPJ_UINT32 y, block_y;
118 OPJ_UINT32 y_incr = 0;
119 const OPJ_UINT32 block_width = sa->block_width;
120
121 if (!opj_sparse_array_is_region_valid(sa, x0, y0, x1, y1)) {
122 return forgiving;
123 }
124
125 block_y = y0 / sa->block_height;
126 for (y = y0; y < y1; block_y ++, y += y_incr) {
127 OPJ_UINT32 x, block_x;
128 OPJ_UINT32 x_incr = 0;
129 OPJ_UINT32 block_y_offset;
130 y_incr = (y == y0) ? sa->block_height - (y0 % sa->block_height) :
131 sa->block_height;
132 block_y_offset = sa->block_height - y_incr;
133 y_incr = opj_uint_min(y_incr, y1 - y);
134 block_x = x0 / block_width;
135 for (x = x0; x < x1; block_x ++, x += x_incr) {
136 OPJ_UINT32 j;
137 OPJ_UINT32 block_x_offset;
138 OPJ_INT32* src_block;
139 x_incr = (x == x0) ? block_width - (x0 % block_width) : block_width;
140 block_x_offset = block_width - x_incr;
141 x_incr = opj_uint_min(x_incr, x1 - x);
142 src_block = sa->data_blocks[block_y * sa->block_count_hor + block_x];
143 if (is_read_op) {
144 if (src_block == NULL) {
145 if (buf_col_stride == 1) {
146 OPJ_INT32* dest_ptr = buf + (y - y0) * (OPJ_SIZE_T)buf_line_stride +
147 (x - x0) * buf_col_stride;
148 for (j = 0; j < y_incr; j++) {
149 memset(dest_ptr, 0, sizeof(OPJ_INT32) * x_incr);
150 dest_ptr += buf_line_stride;
151 }
152 } else {
153 OPJ_INT32* dest_ptr = buf + (y - y0) * (OPJ_SIZE_T)buf_line_stride +
154 (x - x0) * buf_col_stride;
155 for (j = 0; j < y_incr; j++) {
156 OPJ_UINT32 k;
157 for (k = 0; k < x_incr; k++) {
158 dest_ptr[k * buf_col_stride] = 0;
159 }
160 dest_ptr += buf_line_stride;
161 }
162 }
163 } else {
164 const OPJ_INT32* OPJ_RESTRICT src_ptr = src_block + block_y_offset *
165 (OPJ_SIZE_T)block_width + block_x_offset;
166 if (buf_col_stride == 1) {
167 OPJ_INT32* OPJ_RESTRICT dest_ptr = buf + (y - y0) * (OPJ_SIZE_T)buf_line_stride
168 +
169 (x - x0) * buf_col_stride;
170 if (x_incr == 4) {
171 /* Same code as general branch, but the compiler */
172 /* can have an efficient memcpy() */
173 (void)(x_incr); /* trick to silent cppcheck duplicateBranch warning */
174 for (j = 0; j < y_incr; j++) {
175 memcpy(dest_ptr, src_ptr, sizeof(OPJ_INT32) * x_incr);
176 dest_ptr += buf_line_stride;
177 src_ptr += block_width;
178 }
179 } else {
180 for (j = 0; j < y_incr; j++) {
181 memcpy(dest_ptr, src_ptr, sizeof(OPJ_INT32) * x_incr);
182 dest_ptr += buf_line_stride;
183 src_ptr += block_width;
184 }
185 }
186 } else {
187 OPJ_INT32* OPJ_RESTRICT dest_ptr = buf + (y - y0) * (OPJ_SIZE_T)buf_line_stride
188 +
189 (x - x0) * buf_col_stride;
190 if (x_incr == 1) {
191 for (j = 0; j < y_incr; j++) {
192 *dest_ptr = *src_ptr;
193 dest_ptr += buf_line_stride;
194 src_ptr += block_width;
195 }
196 } else if (y_incr == 1 && buf_col_stride == 2) {
197 OPJ_UINT32 k;
198 for (k = 0; k < (x_incr & ~3U); k += 4) {
199 dest_ptr[k * buf_col_stride] = src_ptr[k];
200 dest_ptr[(k + 1) * buf_col_stride] = src_ptr[k + 1];
201 dest_ptr[(k + 2) * buf_col_stride] = src_ptr[k + 2];
202 dest_ptr[(k + 3) * buf_col_stride] = src_ptr[k + 3];
203 }
204 for (; k < x_incr; k++) {
205 dest_ptr[k * buf_col_stride] = src_ptr[k];
206 }
207 } else if (x_incr >= 8 && buf_col_stride == 8) {
208 for (j = 0; j < y_incr; j++) {
209 OPJ_UINT32 k;
210 for (k = 0; k < (x_incr & ~3U); k += 4) {
211 dest_ptr[k * buf_col_stride] = src_ptr[k];
212 dest_ptr[(k + 1) * buf_col_stride] = src_ptr[k + 1];
213 dest_ptr[(k + 2) * buf_col_stride] = src_ptr[k + 2];
214 dest_ptr[(k + 3) * buf_col_stride] = src_ptr[k + 3];
215 }
216 for (; k < x_incr; k++) {
217 dest_ptr[k * buf_col_stride] = src_ptr[k];
218 }
219 dest_ptr += buf_line_stride;
220 src_ptr += block_width;
221 }
222 } else {
223 /* General case */
224 for (j = 0; j < y_incr; j++) {
225 OPJ_UINT32 k;
226 for (k = 0; k < x_incr; k++) {
227 dest_ptr[k * buf_col_stride] = src_ptr[k];
228 }
229 dest_ptr += buf_line_stride;
230 src_ptr += block_width;
231 }
232 }
233 }
234 }
235 } else {
236 if (src_block == NULL) {
237 src_block = (OPJ_INT32*) opj_calloc(1,
238 sa->block_width * sa->block_height * sizeof(OPJ_INT32));
239 if (src_block == NULL) {
240 return OPJ_FALSE;
241 }
242 sa->data_blocks[block_y * sa->block_count_hor + block_x] = src_block;
243 }
244
245 if (buf_col_stride == 1) {
246 OPJ_INT32* OPJ_RESTRICT dest_ptr = src_block + block_y_offset *
247 (OPJ_SIZE_T)block_width + block_x_offset;
248 const OPJ_INT32* OPJ_RESTRICT src_ptr = buf + (y - y0) *
249 (OPJ_SIZE_T)buf_line_stride + (x - x0) * buf_col_stride;
250 if (x_incr == 4) {
251 /* Same code as general branch, but the compiler */
252 /* can have an efficient memcpy() */
253 (void)(x_incr); /* trick to silent cppcheck duplicateBranch warning */
254 for (j = 0; j < y_incr; j++) {
255 memcpy(dest_ptr, src_ptr, sizeof(OPJ_INT32) * x_incr);
256 dest_ptr += block_width;
257 src_ptr += buf_line_stride;
258 }
259 } else {
260 for (j = 0; j < y_incr; j++) {
261 memcpy(dest_ptr, src_ptr, sizeof(OPJ_INT32) * x_incr);
262 dest_ptr += block_width;
263 src_ptr += buf_line_stride;
264 }
265 }
266 } else {
267 OPJ_INT32* OPJ_RESTRICT dest_ptr = src_block + block_y_offset *
268 (OPJ_SIZE_T)block_width + block_x_offset;
269 const OPJ_INT32* OPJ_RESTRICT src_ptr = buf + (y - y0) *
270 (OPJ_SIZE_T)buf_line_stride + (x - x0) * buf_col_stride;
271 if (x_incr == 1) {
272 for (j = 0; j < y_incr; j++) {
273 *dest_ptr = *src_ptr;
274 src_ptr += buf_line_stride;
275 dest_ptr += block_width;
276 }
277 } else if (x_incr >= 8 && buf_col_stride == 8) {
278 for (j = 0; j < y_incr; j++) {
279 OPJ_UINT32 k;
280 for (k = 0; k < (x_incr & ~3U); k += 4) {
281 dest_ptr[k] = src_ptr[k * buf_col_stride];
282 dest_ptr[k + 1] = src_ptr[(k + 1) * buf_col_stride];
283 dest_ptr[k + 2] = src_ptr[(k + 2) * buf_col_stride];
284 dest_ptr[k + 3] = src_ptr[(k + 3) * buf_col_stride];
285 }
286 for (; k < x_incr; k++) {
287 dest_ptr[k] = src_ptr[k * buf_col_stride];
288 }
289 src_ptr += buf_line_stride;
290 dest_ptr += block_width;
291 }
292 } else {
293 /* General case */
294 for (j = 0; j < y_incr; j++) {
295 OPJ_UINT32 k;
296 for (k = 0; k < x_incr; k++) {
297 dest_ptr[k] = src_ptr[k * buf_col_stride];
298 }
299 src_ptr += buf_line_stride;
300 dest_ptr += block_width;
301 }
302 }
303 }
304 }
305 }
306 }
307
308 return OPJ_TRUE;
309}
310
311OPJ_BOOL opj_sparse_array_int32_read(const opj_sparse_array_int32_t* sa,
312 OPJ_UINT32 x0,
313 OPJ_UINT32 y0,
314 OPJ_UINT32 x1,
315 OPJ_UINT32 y1,
316 OPJ_INT32* dest,
317 OPJ_UINT32 dest_col_stride,
318 OPJ_UINT32 dest_line_stride,
319 OPJ_BOOL forgiving)
320{
321 return opj_sparse_array_int32_read_or_write(
322 (opj_sparse_array_int32_t*)sa, x0, y0, x1, y1,
323 dest,
324 dest_col_stride,
325 dest_line_stride,
326 forgiving,
327 OPJ_TRUE);
328}
329
330OPJ_BOOL opj_sparse_array_int32_write(opj_sparse_array_int32_t* sa,
331 OPJ_UINT32 x0,
332 OPJ_UINT32 y0,
333 OPJ_UINT32 x1,
334 OPJ_UINT32 y1,
335 const OPJ_INT32* src,
336 OPJ_UINT32 src_col_stride,
337 OPJ_UINT32 src_line_stride,
338 OPJ_BOOL forgiving)
339{
340 return opj_sparse_array_int32_read_or_write(sa, x0, y0, x1, y1,
341 (OPJ_INT32*)src,
342 src_col_stride,
343 src_line_stride,
344 forgiving,
345 OPJ_FALSE);
346}
347