Go to most recent revision | Details | Last modification | View Log | RSS feed
Rev | Author | Line No. | Line |
---|---|---|---|
4349 | Serge | 1 | /* |
2 | * Copyright (C) 2013 Wei Gao |
||
3 | * |
||
4 | * This file is part of FFmpeg. |
||
5 | * |
||
6 | * FFmpeg is free software; you can redistribute it and/or |
||
7 | * modify it under the terms of the GNU Lesser General Public |
||
8 | * License as published by the Free Software Foundation; either |
||
9 | * version 2.1 of the License, or (at your option) any later version. |
||
10 | * |
||
11 | * FFmpeg is distributed in the hope that it will be useful, |
||
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
||
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
||
14 | * Lesser General Public License for more details. |
||
15 | * |
||
16 | * You should have received a copy of the GNU Lesser General Public |
||
17 | * License along with FFmpeg; if not, write to the Free Software |
||
18 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
||
19 | */ |
||
20 | |||
21 | /** |
||
22 | * @file |
||
23 | * unsharp input video |
||
24 | */ |
||
25 | |||
26 | #include "unsharp_opencl.h" |
||
27 | #include "libavutil/common.h" |
||
28 | #include "libavutil/opencl_internal.h" |
||
29 | |||
30 | #define PLANE_NUM 3 |
||
31 | |||
32 | static inline void add_mask_counter(uint32_t *dst, uint32_t *counter1, uint32_t *counter2, int len) |
||
33 | { |
||
34 | int i; |
||
35 | for (i = 0; i < len; i++) { |
||
36 | dst[i] = counter1[i] + counter2[i]; |
||
37 | } |
||
38 | } |
||
39 | |||
40 | static int compute_mask(int step, uint32_t *mask) |
||
41 | { |
||
42 | int i, z, ret = 0; |
||
43 | int counter_size = sizeof(uint32_t) * (2 * step + 1); |
||
44 | uint32_t *temp1_counter, *temp2_counter, **counter; |
||
45 | temp1_counter = av_mallocz(counter_size); |
||
46 | if (!temp1_counter) { |
||
47 | ret = AVERROR(ENOMEM); |
||
48 | goto end; |
||
49 | } |
||
50 | temp2_counter = av_mallocz(counter_size); |
||
51 | if (!temp2_counter) { |
||
52 | ret = AVERROR(ENOMEM); |
||
53 | goto end; |
||
54 | } |
||
55 | counter = av_mallocz(sizeof(uint32_t *) * (2 * step + 1)); |
||
56 | if (!counter) { |
||
57 | ret = AVERROR(ENOMEM); |
||
58 | goto end; |
||
59 | } |
||
60 | for (i = 0; i < 2 * step + 1; i++) { |
||
61 | counter[i] = av_mallocz(counter_size); |
||
62 | if (!counter[i]) { |
||
63 | ret = AVERROR(ENOMEM); |
||
64 | goto end; |
||
65 | } |
||
66 | } |
||
67 | for (i = 0; i < 2 * step + 1; i++) { |
||
68 | memset(temp1_counter, 0, counter_size); |
||
69 | temp1_counter[i] = 1; |
||
70 | for (z = 0; z < step * 2; z += 2) { |
||
71 | add_mask_counter(temp2_counter, counter[z], temp1_counter, step * 2); |
||
72 | memcpy(counter[z], temp1_counter, counter_size); |
||
73 | add_mask_counter(temp1_counter, counter[z + 1], temp2_counter, step * 2); |
||
74 | memcpy(counter[z + 1], temp2_counter, counter_size); |
||
75 | } |
||
76 | } |
||
77 | memcpy(mask, temp1_counter, counter_size); |
||
78 | end: |
||
79 | av_freep(&temp1_counter); |
||
80 | av_freep(&temp2_counter); |
||
81 | for (i = 0; i < 2 * step + 1; i++) { |
||
82 | av_freep(&counter[i]); |
||
83 | } |
||
84 | av_freep(&counter); |
||
85 | return ret; |
||
86 | } |
||
87 | |||
88 | static int compute_mask_matrix(cl_mem cl_mask_matrix, int step_x, int step_y) |
||
89 | { |
||
90 | int i, j, ret = 0; |
||
91 | uint32_t *mask_matrix, *mask_x, *mask_y; |
||
92 | size_t size_matrix = sizeof(uint32_t) * (2 * step_x + 1) * (2 * step_y + 1); |
||
93 | mask_x = av_mallocz(sizeof(uint32_t) * (2 * step_x + 1)); |
||
94 | if (!mask_x) { |
||
95 | ret = AVERROR(ENOMEM); |
||
96 | goto end; |
||
97 | } |
||
98 | mask_y = av_mallocz(sizeof(uint32_t) * (2 * step_y + 1)); |
||
99 | if (!mask_y) { |
||
100 | ret = AVERROR(ENOMEM); |
||
101 | goto end; |
||
102 | } |
||
103 | mask_matrix = av_mallocz(size_matrix); |
||
104 | if (!mask_matrix) { |
||
105 | ret = AVERROR(ENOMEM); |
||
106 | goto end; |
||
107 | } |
||
108 | ret = compute_mask(step_x, mask_x); |
||
109 | if (ret < 0) |
||
110 | goto end; |
||
111 | ret = compute_mask(step_y, mask_y); |
||
112 | if (ret < 0) |
||
113 | goto end; |
||
114 | for (j = 0; j < 2 * step_y + 1; j++) { |
||
115 | for (i = 0; i < 2 * step_x + 1; i++) { |
||
116 | mask_matrix[i + j * (2 * step_x + 1)] = mask_y[j] * mask_x[i]; |
||
117 | } |
||
118 | } |
||
119 | ret = av_opencl_buffer_write(cl_mask_matrix, (uint8_t *)mask_matrix, size_matrix); |
||
120 | end: |
||
121 | av_freep(&mask_x); |
||
122 | av_freep(&mask_y); |
||
123 | av_freep(&mask_matrix); |
||
124 | return ret; |
||
125 | } |
||
126 | |||
127 | static int generate_mask(AVFilterContext *ctx) |
||
128 | { |
||
129 | UnsharpContext *unsharp = ctx->priv; |
||
130 | int i, ret = 0, step_x[2], step_y[2]; |
||
131 | cl_mem mask_matrix[2]; |
||
132 | mask_matrix[0] = unsharp->opencl_ctx.cl_luma_mask; |
||
133 | mask_matrix[1] = unsharp->opencl_ctx.cl_chroma_mask; |
||
134 | step_x[0] = unsharp->luma.steps_x; |
||
135 | step_x[1] = unsharp->chroma.steps_x; |
||
136 | step_y[0] = unsharp->luma.steps_y; |
||
137 | step_y[1] = unsharp->chroma.steps_y; |
||
138 | if (!mask_matrix[0] || !mask_matrix[1]) { |
||
139 | av_log(ctx, AV_LOG_ERROR, "Luma mask and chroma mask should not be NULL\n"); |
||
140 | return AVERROR(EINVAL); |
||
141 | } |
||
142 | for (i = 0; i < 2; i++) { |
||
143 | ret = compute_mask_matrix(mask_matrix[i], step_x[i], step_y[i]); |
||
144 | if (ret < 0) |
||
145 | return ret; |
||
146 | } |
||
147 | return ret; |
||
148 | } |
||
149 | |||
150 | int ff_opencl_apply_unsharp(AVFilterContext *ctx, AVFrame *in, AVFrame *out) |
||
151 | { |
||
152 | int ret; |
||
153 | AVFilterLink *link = ctx->inputs[0]; |
||
154 | UnsharpContext *unsharp = ctx->priv; |
||
155 | cl_int status; |
||
156 | int cw = FF_CEIL_RSHIFT(link->w, unsharp->hsub); |
||
157 | int ch = FF_CEIL_RSHIFT(link->h, unsharp->vsub); |
||
158 | const size_t global_work_size = link->w * link->h + 2 * ch * cw; |
||
159 | FFOpenclParam opencl_param = {0}; |
||
160 | |||
161 | opencl_param.ctx = ctx; |
||
162 | opencl_param.kernel = unsharp->opencl_ctx.kernel_env.kernel; |
||
163 | ret = ff_opencl_set_parameter(&opencl_param, |
||
164 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_inbuf), |
||
165 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_outbuf), |
||
166 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_luma_mask), |
||
167 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_chroma_mask), |
||
168 | FF_OPENCL_PARAM_INFO(unsharp->luma.amount), |
||
169 | FF_OPENCL_PARAM_INFO(unsharp->chroma.amount), |
||
170 | FF_OPENCL_PARAM_INFO(unsharp->luma.steps_x), |
||
171 | FF_OPENCL_PARAM_INFO(unsharp->luma.steps_y), |
||
172 | FF_OPENCL_PARAM_INFO(unsharp->chroma.steps_x), |
||
173 | FF_OPENCL_PARAM_INFO(unsharp->chroma.steps_y), |
||
174 | FF_OPENCL_PARAM_INFO(unsharp->luma.scalebits), |
||
175 | FF_OPENCL_PARAM_INFO(unsharp->chroma.scalebits), |
||
176 | FF_OPENCL_PARAM_INFO(unsharp->luma.halfscale), |
||
177 | FF_OPENCL_PARAM_INFO(unsharp->chroma.halfscale), |
||
178 | FF_OPENCL_PARAM_INFO(in->linesize[0]), |
||
179 | FF_OPENCL_PARAM_INFO(in->linesize[1]), |
||
180 | FF_OPENCL_PARAM_INFO(out->linesize[0]), |
||
181 | FF_OPENCL_PARAM_INFO(out->linesize[1]), |
||
182 | FF_OPENCL_PARAM_INFO(link->h), |
||
183 | FF_OPENCL_PARAM_INFO(link->w), |
||
184 | FF_OPENCL_PARAM_INFO(ch), |
||
185 | FF_OPENCL_PARAM_INFO(cw), |
||
186 | NULL); |
||
187 | if (ret < 0) |
||
188 | return ret; |
||
189 | status = clEnqueueNDRangeKernel(unsharp->opencl_ctx.kernel_env.command_queue, |
||
190 | unsharp->opencl_ctx.kernel_env.kernel, 1, NULL, |
||
191 | &global_work_size, NULL, 0, NULL, NULL); |
||
192 | if (status != CL_SUCCESS) { |
||
193 | av_log(ctx, AV_LOG_ERROR, "OpenCL run kernel error occurred: %s\n", av_opencl_errstr(status)); |
||
194 | return AVERROR_EXTERNAL; |
||
195 | } |
||
196 | clFinish(unsharp->opencl_ctx.kernel_env.command_queue); |
||
197 | return av_opencl_buffer_read_image(out->data, unsharp->opencl_ctx.out_plane_size, |
||
198 | unsharp->opencl_ctx.plane_num, unsharp->opencl_ctx.cl_outbuf, |
||
199 | unsharp->opencl_ctx.cl_outbuf_size); |
||
200 | } |
||
201 | |||
202 | int ff_opencl_unsharp_init(AVFilterContext *ctx) |
||
203 | { |
||
204 | int ret = 0; |
||
205 | UnsharpContext *unsharp = ctx->priv; |
||
206 | ret = av_opencl_init(NULL); |
||
207 | if (ret < 0) |
||
208 | return ret; |
||
209 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_luma_mask, |
||
210 | sizeof(uint32_t) * (2 * unsharp->luma.steps_x + 1) * (2 * unsharp->luma.steps_y + 1), |
||
211 | CL_MEM_READ_ONLY, NULL); |
||
212 | if (ret < 0) |
||
213 | return ret; |
||
214 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_chroma_mask, |
||
215 | sizeof(uint32_t) * (2 * unsharp->chroma.steps_x + 1) * (2 * unsharp->chroma.steps_y + 1), |
||
216 | CL_MEM_READ_ONLY, NULL); |
||
217 | if (ret < 0) |
||
218 | return ret; |
||
219 | ret = generate_mask(ctx); |
||
220 | if (ret < 0) |
||
221 | return ret; |
||
222 | unsharp->opencl_ctx.plane_num = PLANE_NUM; |
||
223 | if (!unsharp->opencl_ctx.kernel_env.kernel) { |
||
224 | ret = av_opencl_create_kernel(&unsharp->opencl_ctx.kernel_env, "unsharp"); |
||
225 | if (ret < 0) { |
||
226 | av_log(ctx, AV_LOG_ERROR, "OpenCL failed to create kernel with name 'unsharp'\n"); |
||
227 | return ret; |
||
228 | } |
||
229 | } |
||
230 | return ret; |
||
231 | } |
||
232 | |||
233 | void ff_opencl_unsharp_uninit(AVFilterContext *ctx) |
||
234 | { |
||
235 | UnsharpContext *unsharp = ctx->priv; |
||
236 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_inbuf); |
||
237 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_outbuf); |
||
238 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_luma_mask); |
||
239 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_chroma_mask); |
||
240 | av_opencl_release_kernel(&unsharp->opencl_ctx.kernel_env); |
||
241 | av_opencl_uninit(); |
||
242 | } |
||
243 | |||
244 | int ff_opencl_unsharp_process_inout_buf(AVFilterContext *ctx, AVFrame *in, AVFrame *out) |
||
245 | { |
||
246 | int ret = 0; |
||
247 | AVFilterLink *link = ctx->inputs[0]; |
||
248 | UnsharpContext *unsharp = ctx->priv; |
||
249 | int ch = FF_CEIL_RSHIFT(link->h, unsharp->vsub); |
||
250 | |||
251 | if ((!unsharp->opencl_ctx.cl_inbuf) || (!unsharp->opencl_ctx.cl_outbuf)) { |
||
252 | unsharp->opencl_ctx.in_plane_size[0] = (in->linesize[0] * in->height); |
||
253 | unsharp->opencl_ctx.in_plane_size[1] = (in->linesize[1] * ch); |
||
254 | unsharp->opencl_ctx.in_plane_size[2] = (in->linesize[2] * ch); |
||
255 | unsharp->opencl_ctx.out_plane_size[0] = (out->linesize[0] * out->height); |
||
256 | unsharp->opencl_ctx.out_plane_size[1] = (out->linesize[1] * ch); |
||
257 | unsharp->opencl_ctx.out_plane_size[2] = (out->linesize[2] * ch); |
||
258 | unsharp->opencl_ctx.cl_inbuf_size = unsharp->opencl_ctx.in_plane_size[0] + |
||
259 | unsharp->opencl_ctx.in_plane_size[1] + |
||
260 | unsharp->opencl_ctx.in_plane_size[2]; |
||
261 | unsharp->opencl_ctx.cl_outbuf_size = unsharp->opencl_ctx.out_plane_size[0] + |
||
262 | unsharp->opencl_ctx.out_plane_size[1] + |
||
263 | unsharp->opencl_ctx.out_plane_size[2]; |
||
264 | if (!unsharp->opencl_ctx.cl_inbuf) { |
||
265 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_inbuf, |
||
266 | unsharp->opencl_ctx.cl_inbuf_size, |
||
267 | CL_MEM_READ_ONLY, NULL); |
||
268 | if (ret < 0) |
||
269 | return ret; |
||
270 | } |
||
271 | if (!unsharp->opencl_ctx.cl_outbuf) { |
||
272 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_outbuf, |
||
273 | unsharp->opencl_ctx.cl_outbuf_size, |
||
274 | CL_MEM_READ_WRITE, NULL); |
||
275 | if (ret < 0) |
||
276 | return ret; |
||
277 | } |
||
278 | } |
||
279 | return av_opencl_buffer_write_image(unsharp->opencl_ctx.cl_inbuf, |
||
280 | unsharp->opencl_ctx.cl_inbuf_size, |
||
281 | 0, in->data, unsharp->opencl_ctx.in_plane_size, |
||
282 | unsharp->opencl_ctx.plane_num); |
||
283 | }>>>>>>>>>>>>>>>>>>> |