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 | /**************************************************************************/
/* image_compress_cvtt.cpp */
/**************************************************************************/
/* This file is part of: */
/* GODOT ENGINE */
/* https://godotengine.org */
/**************************************************************************/
/* Copyright (c) 2014-present Godot Engine contributors (see AUTHORS.md). */
/* Copyright (c) 2007-2014 Juan Linietsky, Ariel Manzur. */
/* */
/* Permission is hereby granted, free of charge, to any person obtaining */
/* a copy of this software and associated documentation files (the */
/* "Software"), to deal in the Software without restriction, including */
/* without limitation the rights to use, copy, modify, merge, publish, */
/* distribute, sublicense, and/or sell copies of the Software, and to */
/* permit persons to whom the Software is furnished to do so, subject to */
/* the following conditions: */
/* */
/* The above copyright notice and this permission notice shall be */
/* included in all copies or substantial portions of the Software. */
/* */
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, */
/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF */
/* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. */
/* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY */
/* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, */
/* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE */
/* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
/**************************************************************************/
#include "image_compress_cvtt.h"
#include "core/os/os.h"
#include "core/os/thread.h"
#include "core/print_string.h"
#include "core/safe_refcount.h"
#include <ConvectionKernels.h>
struct CVTTCompressionJobParams {
bool is_hdr;
bool is_signed;
int bytes_per_pixel;
cvtt::Options options;
};
struct CVTTCompressionRowTask {
const uint8_t *in_mm_bytes;
uint8_t *out_mm_bytes;
int y_start;
int width;
int height;
};
struct CVTTCompressionJobQueue {
CVTTCompressionJobParams job_params;
const CVTTCompressionRowTask *job_tasks;
uint32_t num_tasks;
SafeNumeric<uint32_t> current_task;
};
static void _digest_row_task(const CVTTCompressionJobParams &p_job_params, const CVTTCompressionRowTask &p_row_task) {
const uint8_t *in_bytes = p_row_task.in_mm_bytes;
uint8_t *out_bytes = p_row_task.out_mm_bytes;
int w = p_row_task.width;
int h = p_row_task.height;
int y_start = p_row_task.y_start;
int y_end = y_start + 4;
int bytes_per_pixel = p_job_params.bytes_per_pixel;
bool is_hdr = p_job_params.is_hdr;
bool is_signed = p_job_params.is_signed;
cvtt::PixelBlockU8 input_blocks_ldr[cvtt::NumParallelBlocks];
cvtt::PixelBlockF16 input_blocks_hdr[cvtt::NumParallelBlocks];
for (int x_start = 0; x_start < w; x_start += 4 * cvtt::NumParallelBlocks) {
int x_end = x_start + 4 * cvtt::NumParallelBlocks;
for (int y = y_start; y < y_end; y++) {
int first_input_element = (y - y_start) * 4;
const uint8_t *row_start;
if (y >= h) {
row_start = in_bytes + (h - 1) * (w * bytes_per_pixel);
} else {
row_start = in_bytes + y * (w * bytes_per_pixel);
}
for (int x = x_start; x < x_end; x++) {
const uint8_t *pixel_start;
if (x >= w) {
pixel_start = row_start + (w - 1) * bytes_per_pixel;
} else {
pixel_start = row_start + x * bytes_per_pixel;
}
int block_index = (x - x_start) / 4;
int block_element = (x - x_start) % 4 + first_input_element;
if (is_hdr) {
memcpy(input_blocks_hdr[block_index].m_pixels[block_element], pixel_start, bytes_per_pixel);
input_blocks_hdr[block_index].m_pixels[block_element][3] = 0x3c00; // 1.0 (unused)
} else {
memcpy(input_blocks_ldr[block_index].m_pixels[block_element], pixel_start, bytes_per_pixel);
}
}
}
uint8_t output_blocks[16 * cvtt::NumParallelBlocks];
if (is_hdr) {
if (is_signed) {
cvtt::Kernels::EncodeBC6HS(output_blocks, input_blocks_hdr, p_job_params.options);
} else {
cvtt::Kernels::EncodeBC6HU(output_blocks, input_blocks_hdr, p_job_params.options);
}
} else {
cvtt::Kernels::EncodeBC7(output_blocks, input_blocks_ldr, p_job_params.options);
}
unsigned int num_real_blocks = ((w - x_start) + 3) / 4;
if (num_real_blocks > cvtt::NumParallelBlocks) {
num_real_blocks = cvtt::NumParallelBlocks;
}
memcpy(out_bytes, output_blocks, 16 * num_real_blocks);
out_bytes += 16 * num_real_blocks;
}
}
static void _digest_job_queue(void *p_job_queue) {
CVTTCompressionJobQueue *job_queue = static_cast<CVTTCompressionJobQueue *>(p_job_queue);
for (uint32_t next_task = job_queue->current_task.increment(); next_task <= job_queue->num_tasks; next_task = job_queue->current_task.increment()) {
_digest_row_task(job_queue->job_params, job_queue->job_tasks[next_task - 1]);
}
}
void image_compress_cvtt(Image *p_image, float p_lossy_quality, Image::CompressSource p_source) {
if (p_image->get_format() >= Image::FORMAT_BPTC_RGBA) {
return; //do not compress, already compressed
}
int w = p_image->get_width();
int h = p_image->get_height();
bool is_ldr = (p_image->get_format() <= Image::FORMAT_RGBA8);
bool is_hdr = (p_image->get_format() >= Image::FORMAT_RH) && (p_image->get_format() <= Image::FORMAT_RGBE9995);
if (!is_ldr && !is_hdr) {
return; // Not a usable source format
}
cvtt::Options options;
uint32_t flags = cvtt::Flags::Fastest;
if (p_lossy_quality > 0.85) {
flags = cvtt::Flags::Ultra;
} else if (p_lossy_quality > 0.75) {
flags = cvtt::Flags::Better;
} else if (p_lossy_quality > 0.55) {
flags = cvtt::Flags::Default;
} else if (p_lossy_quality > 0.35) {
flags = cvtt::Flags::Fast;
} else if (p_lossy_quality > 0.15) {
flags = cvtt::Flags::Faster;
}
flags |= cvtt::Flags::BC7_RespectPunchThrough;
if (p_source == Image::COMPRESS_SOURCE_NORMAL) {
flags |= cvtt::Flags::Uniform;
}
options.flags = flags;
Image::Format target_format = Image::FORMAT_BPTC_RGBA;
bool is_signed = false;
if (is_hdr) {
if (p_image->get_format() != Image::FORMAT_RGBH) {
p_image->convert(Image::FORMAT_RGBH);
}
PoolVector<uint8_t>::Read rb = p_image->get_data().read();
const uint16_t *source_data = reinterpret_cast<const uint16_t *>(&rb[0]);
int pixel_element_count = w * h * 3;
for (int i = 0; i < pixel_element_count; i++) {
if ((source_data[i] & 0x8000) != 0 && (source_data[i] & 0x7fff) != 0) {
is_signed = true;
break;
}
}
target_format = is_signed ? Image::FORMAT_BPTC_RGBF : Image::FORMAT_BPTC_RGBFU;
} else {
p_image->convert(Image::FORMAT_RGBA8); //still uses RGBA to convert
}
PoolVector<uint8_t>::Read rb = p_image->get_data().read();
PoolVector<uint8_t> data;
int target_size = Image::get_image_data_size(w, h, target_format, p_image->has_mipmaps());
int mm_count = p_image->has_mipmaps() ? Image::get_image_required_mipmaps(w, h, target_format) : 0;
data.resize(target_size);
int shift = Image::get_format_pixel_rshift(target_format);
PoolVector<uint8_t>::Write wb = data.write();
int dst_ofs = 0;
CVTTCompressionJobQueue job_queue;
job_queue.job_params.is_hdr = is_hdr;
job_queue.job_params.is_signed = is_signed;
job_queue.job_params.options = options;
job_queue.job_params.bytes_per_pixel = is_hdr ? 6 : 4;
#ifdef NO_THREADS
int num_job_threads = 0;<--- Assignment 'num_job_threads=0', assigned value is 0<--- 'num_job_threads' is assigned value '0' here.
#else
int num_job_threads = OS::get_singleton()->can_use_threads() ? (OS::get_singleton()->get_processor_count() - 1) : 0;
#endif
PoolVector<CVTTCompressionRowTask> tasks;
for (int i = 0; i <= mm_count; i++) {
int bw = w % 4 != 0 ? w + (4 - w % 4) : w;
int bh = h % 4 != 0 ? h + (4 - h % 4) : h;
int src_ofs = p_image->get_mipmap_offset(i);
const uint8_t *in_bytes = &rb[src_ofs];
uint8_t *out_bytes = &wb[dst_ofs];
for (int y_start = 0; y_start < h; y_start += 4) {
CVTTCompressionRowTask row_task;
row_task.width = w;
row_task.height = h;
row_task.y_start = y_start;
row_task.in_mm_bytes = in_bytes;
row_task.out_mm_bytes = out_bytes;
if (num_job_threads > 0) {<--- Condition 'num_job_threads>0' is always false
tasks.push_back(row_task);
} else {
_digest_row_task(job_queue.job_params, row_task);
}
out_bytes += 16 * (bw / 4);
}
dst_ofs += (MAX(4, bw) * MAX(4, bh)) >> shift;
w = MAX(w / 2, 1);
h = MAX(h / 2, 1);
}
if (num_job_threads > 0) {<--- The comparison 'num_job_threads > 0' is always false. [+]Finding the same expression on both sides of an operator is suspicious and might indicate a cut and paste or logic error. Please examine this code carefully to determine if it is correct.
PoolVector<Thread *> threads;
threads.resize(num_job_threads);
PoolVector<Thread *>::Write threads_wb = threads.write();
PoolVector<CVTTCompressionRowTask>::Read tasks_rb = tasks.read();
job_queue.job_tasks = &tasks_rb[0];
job_queue.current_task.set(0);
job_queue.num_tasks = static_cast<uint32_t>(tasks.size());
for (int i = 0; i < num_job_threads; i++) {
threads_wb[i] = memnew(Thread);
threads_wb[i]->start(_digest_job_queue, &job_queue);
}
_digest_job_queue(&job_queue);
for (int i = 0; i < num_job_threads; i++) {
threads_wb[i]->wait_to_finish();
memdelete(threads_wb[i]);
}
}
p_image->create(p_image->get_width(), p_image->get_height(), p_image->has_mipmaps(), target_format, data);
}
void image_decompress_cvtt(Image *p_image) {
Image::Format target_format;
bool is_signed = false;
bool is_hdr = false;
Image::Format input_format = p_image->get_format();
switch (input_format) {
case Image::FORMAT_BPTC_RGBA:
target_format = Image::FORMAT_RGBA8;
break;
case Image::FORMAT_BPTC_RGBF:
case Image::FORMAT_BPTC_RGBFU:
target_format = Image::FORMAT_RGBH;
is_signed = (input_format == Image::FORMAT_BPTC_RGBF);
is_hdr = true;
break;
default:
return; // Invalid input format
};
int w = p_image->get_width();
int h = p_image->get_height();
PoolVector<uint8_t>::Read rb = p_image->get_data().read();
PoolVector<uint8_t> data;
int target_size = Image::get_image_data_size(w, h, target_format, p_image->has_mipmaps());
int mm_count = p_image->get_mipmap_count();
data.resize(target_size);
PoolVector<uint8_t>::Write wb = data.write();
int bytes_per_pixel = is_hdr ? 6 : 4;
int dst_ofs = 0;
for (int i = 0; i <= mm_count; i++) {
int src_ofs = p_image->get_mipmap_offset(i);
const uint8_t *in_bytes = &rb[src_ofs];
uint8_t *out_bytes = &wb[dst_ofs];
cvtt::PixelBlockU8 output_blocks_ldr[cvtt::NumParallelBlocks];
cvtt::PixelBlockF16 output_blocks_hdr[cvtt::NumParallelBlocks];
for (int y_start = 0; y_start < h; y_start += 4) {
int y_end = y_start + 4;
for (int x_start = 0; x_start < w; x_start += 4 * cvtt::NumParallelBlocks) {
int x_end = x_start + 4 * cvtt::NumParallelBlocks;
uint8_t input_blocks[16 * cvtt::NumParallelBlocks];
memset(input_blocks, 0, sizeof(input_blocks));
unsigned int num_real_blocks = ((w - x_start) + 3) / 4;
if (num_real_blocks > cvtt::NumParallelBlocks) {
num_real_blocks = cvtt::NumParallelBlocks;
}
memcpy(input_blocks, in_bytes, 16 * num_real_blocks);
in_bytes += 16 * num_real_blocks;
if (is_hdr) {
if (is_signed) {
cvtt::Kernels::DecodeBC6HS(output_blocks_hdr, input_blocks);
} else {
cvtt::Kernels::DecodeBC6HU(output_blocks_hdr, input_blocks);
}
} else {
cvtt::Kernels::DecodeBC7(output_blocks_ldr, input_blocks);
}
for (int y = y_start; y < y_end; y++) {
int first_input_element = (y - y_start) * 4;
uint8_t *row_start;
if (y >= h) {
row_start = out_bytes + (h - 1) * (w * bytes_per_pixel);
} else {
row_start = out_bytes + y * (w * bytes_per_pixel);
}
for (int x = x_start; x < x_end; x++) {
uint8_t *pixel_start;
if (x >= w) {
pixel_start = row_start + (w - 1) * bytes_per_pixel;
} else {
pixel_start = row_start + x * bytes_per_pixel;
}
int block_index = (x - x_start) / 4;
int block_element = (x - x_start) % 4 + first_input_element;
if (is_hdr) {
memcpy(pixel_start, output_blocks_hdr[block_index].m_pixels[block_element], bytes_per_pixel);
} else {
memcpy(pixel_start, output_blocks_ldr[block_index].m_pixels[block_element], bytes_per_pixel);
}
}
}
}
}
dst_ofs += w * h * bytes_per_pixel;
w >>= 1;
h >>= 1;
}
rb.release();
wb.release();
p_image->create(p_image->get_width(), p_image->get_height(), p_image->has_mipmaps(), target_format, data);
}
|