| 1 | /* |
| 2 | * Copyright (C) 2013 Wei Gao <weigao@multicorewareinc.com> |
| 3 | * Copyright (C) 2013 Lenny Wang |
| 4 | * |
| 5 | * This file is part of FFmpeg. |
| 6 | * |
| 7 | * FFmpeg is free software; you can redistribute it and/or |
| 8 | * modify it under the terms of the GNU Lesser General Public |
| 9 | * License as published by the Free Software Foundation; either |
| 10 | * version 2.1 of the License, or (at your option) any later version. |
| 11 | * |
| 12 | * FFmpeg is distributed in the hope that it will be useful, |
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | * Lesser General Public License for more details. |
| 16 | * |
| 17 | * You should have received a copy of the GNU Lesser General Public |
| 18 | * License along with FFmpeg; if not, write to the Free Software |
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | */ |
| 21 | |
| 22 | /** |
| 23 | * @file |
| 24 | * unsharp input video |
| 25 | */ |
| 26 | |
| 27 | #include "unsharp_opencl.h" |
| 28 | #include "libavutil/common.h" |
| 29 | #include "libavutil/opencl_internal.h" |
| 30 | |
| 31 | #define PLANE_NUM 3 |
| 32 | #define ROUND_TO_16(a) (((((a) - 1)/16)+1)*16) |
| 33 | |
| 34 | static inline void add_mask_counter(uint32_t *dst, uint32_t *counter1, uint32_t *counter2, int len) |
| 35 | { |
| 36 | int i; |
| 37 | for (i = 0; i < len; i++) { |
| 38 | dst[i] = counter1[i] + counter2[i]; |
| 39 | } |
| 40 | } |
| 41 | |
| 42 | static int compute_mask(int step, uint32_t *mask) |
| 43 | { |
| 44 | int i, z, ret = 0; |
| 45 | int counter_size = sizeof(uint32_t) * (2 * step + 1); |
| 46 | uint32_t *temp1_counter, *temp2_counter, **counter; |
| 47 | temp1_counter = av_mallocz(counter_size); |
| 48 | if (!temp1_counter) { |
| 49 | ret = AVERROR(ENOMEM); |
| 50 | goto end; |
| 51 | } |
| 52 | temp2_counter = av_mallocz(counter_size); |
| 53 | if (!temp2_counter) { |
| 54 | ret = AVERROR(ENOMEM); |
| 55 | goto end; |
| 56 | } |
| 57 | counter = av_mallocz_array(2 * step + 1, sizeof(uint32_t *)); |
| 58 | if (!counter) { |
| 59 | ret = AVERROR(ENOMEM); |
| 60 | goto end; |
| 61 | } |
| 62 | for (i = 0; i < 2 * step + 1; i++) { |
| 63 | counter[i] = av_mallocz(counter_size); |
| 64 | if (!counter[i]) { |
| 65 | ret = AVERROR(ENOMEM); |
| 66 | goto end; |
| 67 | } |
| 68 | } |
| 69 | for (i = 0; i < 2 * step + 1; i++) { |
| 70 | memset(temp1_counter, 0, counter_size); |
| 71 | temp1_counter[i] = 1; |
| 72 | for (z = 0; z < step * 2; z += 2) { |
| 73 | add_mask_counter(temp2_counter, counter[z], temp1_counter, step * 2); |
| 74 | memcpy(counter[z], temp1_counter, counter_size); |
| 75 | add_mask_counter(temp1_counter, counter[z + 1], temp2_counter, step * 2); |
| 76 | memcpy(counter[z + 1], temp2_counter, counter_size); |
| 77 | } |
| 78 | } |
| 79 | memcpy(mask, temp1_counter, counter_size); |
| 80 | end: |
| 81 | av_freep(&temp1_counter); |
| 82 | av_freep(&temp2_counter); |
| 83 | for (i = 0; i < 2 * step + 1; i++) { |
| 84 | av_freep(&counter[i]); |
| 85 | } |
| 86 | av_freep(&counter); |
| 87 | return ret; |
| 88 | } |
| 89 | |
| 90 | static int compute_mask_matrix(cl_mem cl_mask_matrix, int step_x, int step_y) |
| 91 | { |
| 92 | int i, j, ret = 0; |
| 93 | uint32_t *mask_matrix, *mask_x, *mask_y; |
| 94 | size_t size_matrix = sizeof(uint32_t) * (2 * step_x + 1) * (2 * step_y + 1); |
| 95 | mask_x = av_mallocz_array(2 * step_x + 1, sizeof(uint32_t)); |
| 96 | if (!mask_x) { |
| 97 | ret = AVERROR(ENOMEM); |
| 98 | goto end; |
| 99 | } |
| 100 | mask_y = av_mallocz_array(2 * step_y + 1, sizeof(uint32_t)); |
| 101 | if (!mask_y) { |
| 102 | ret = AVERROR(ENOMEM); |
| 103 | goto end; |
| 104 | } |
| 105 | mask_matrix = av_mallocz(size_matrix); |
| 106 | if (!mask_matrix) { |
| 107 | ret = AVERROR(ENOMEM); |
| 108 | goto end; |
| 109 | } |
| 110 | ret = compute_mask(step_x, mask_x); |
| 111 | if (ret < 0) |
| 112 | goto end; |
| 113 | ret = compute_mask(step_y, mask_y); |
| 114 | if (ret < 0) |
| 115 | goto end; |
| 116 | for (j = 0; j < 2 * step_y + 1; j++) { |
| 117 | for (i = 0; i < 2 * step_x + 1; i++) { |
| 118 | mask_matrix[i + j * (2 * step_x + 1)] = mask_y[j] * mask_x[i]; |
| 119 | } |
| 120 | } |
| 121 | ret = av_opencl_buffer_write(cl_mask_matrix, (uint8_t *)mask_matrix, size_matrix); |
| 122 | end: |
| 123 | av_freep(&mask_x); |
| 124 | av_freep(&mask_y); |
| 125 | av_freep(&mask_matrix); |
| 126 | return ret; |
| 127 | } |
| 128 | |
| 129 | static int generate_mask(AVFilterContext *ctx) |
| 130 | { |
| 131 | UnsharpContext *unsharp = ctx->priv; |
| 132 | int i, ret = 0, step_x[2], step_y[2]; |
| 133 | cl_mem mask_matrix[2]; |
| 134 | mask_matrix[0] = unsharp->opencl_ctx.cl_luma_mask; |
| 135 | mask_matrix[1] = unsharp->opencl_ctx.cl_chroma_mask; |
| 136 | step_x[0] = unsharp->luma.steps_x; |
| 137 | step_x[1] = unsharp->chroma.steps_x; |
| 138 | step_y[0] = unsharp->luma.steps_y; |
| 139 | step_y[1] = unsharp->chroma.steps_y; |
| 140 | |
| 141 | /* use default kernel if any matrix dim larger than 8 due to limited local mem size */ |
| 142 | if (step_x[0]>8 || step_x[1]>8 || step_y[0]>8 || step_y[1]>8) |
| 143 | unsharp->opencl_ctx.use_fast_kernels = 0; |
| 144 | else |
| 145 | unsharp->opencl_ctx.use_fast_kernels = 1; |
| 146 | |
| 147 | if (!mask_matrix[0] || !mask_matrix[1]) { |
| 148 | av_log(ctx, AV_LOG_ERROR, "Luma mask and chroma mask should not be NULL\n"); |
| 149 | return AVERROR(EINVAL); |
| 150 | } |
| 151 | for (i = 0; i < 2; i++) { |
| 152 | ret = compute_mask_matrix(mask_matrix[i], step_x[i], step_y[i]); |
| 153 | if (ret < 0) |
| 154 | return ret; |
| 155 | } |
| 156 | return ret; |
| 157 | } |
| 158 | |
| 159 | int ff_opencl_apply_unsharp(AVFilterContext *ctx, AVFrame *in, AVFrame *out) |
| 160 | { |
| 161 | int ret; |
| 162 | AVFilterLink *link = ctx->inputs[0]; |
| 163 | UnsharpContext *unsharp = ctx->priv; |
| 164 | cl_int status; |
| 165 | FFOpenclParam kernel1 = {0}; |
| 166 | FFOpenclParam kernel2 = {0}; |
| 167 | int width = link->w; |
| 168 | int height = link->h; |
| 169 | int cw = FF_CEIL_RSHIFT(link->w, unsharp->hsub); |
| 170 | int ch = FF_CEIL_RSHIFT(link->h, unsharp->vsub); |
| 171 | size_t globalWorkSize1d = width * height + 2 * ch * cw; |
| 172 | size_t globalWorkSize2dLuma[2]; |
| 173 | size_t globalWorkSize2dChroma[2]; |
| 174 | size_t localWorkSize2d[2] = {16, 16}; |
| 175 | |
| 176 | if (unsharp->opencl_ctx.use_fast_kernels) { |
| 177 | globalWorkSize2dLuma[0] = (size_t)ROUND_TO_16(width); |
| 178 | globalWorkSize2dLuma[1] = (size_t)ROUND_TO_16(height); |
| 179 | globalWorkSize2dChroma[0] = (size_t)ROUND_TO_16(cw); |
| 180 | globalWorkSize2dChroma[1] = (size_t)(2*ROUND_TO_16(ch)); |
| 181 | |
| 182 | kernel1.ctx = ctx; |
| 183 | kernel1.kernel = unsharp->opencl_ctx.kernel_luma; |
| 184 | ret = avpriv_opencl_set_parameter(&kernel1, |
| 185 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_inbuf), |
| 186 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_outbuf), |
| 187 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_luma_mask), |
| 188 | FF_OPENCL_PARAM_INFO(unsharp->luma.amount), |
| 189 | FF_OPENCL_PARAM_INFO(unsharp->luma.scalebits), |
| 190 | FF_OPENCL_PARAM_INFO(unsharp->luma.halfscale), |
| 191 | FF_OPENCL_PARAM_INFO(in->linesize[0]), |
| 192 | FF_OPENCL_PARAM_INFO(out->linesize[0]), |
| 193 | FF_OPENCL_PARAM_INFO(width), |
| 194 | FF_OPENCL_PARAM_INFO(height), |
| 195 | NULL); |
| 196 | if (ret < 0) |
| 197 | return ret; |
| 198 | |
| 199 | kernel2.ctx = ctx; |
| 200 | kernel2.kernel = unsharp->opencl_ctx.kernel_chroma; |
| 201 | ret = avpriv_opencl_set_parameter(&kernel2, |
| 202 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_inbuf), |
| 203 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_outbuf), |
| 204 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_chroma_mask), |
| 205 | FF_OPENCL_PARAM_INFO(unsharp->chroma.amount), |
| 206 | FF_OPENCL_PARAM_INFO(unsharp->chroma.scalebits), |
| 207 | FF_OPENCL_PARAM_INFO(unsharp->chroma.halfscale), |
| 208 | FF_OPENCL_PARAM_INFO(in->linesize[0]), |
| 209 | FF_OPENCL_PARAM_INFO(in->linesize[1]), |
| 210 | FF_OPENCL_PARAM_INFO(out->linesize[0]), |
| 211 | FF_OPENCL_PARAM_INFO(out->linesize[1]), |
| 212 | FF_OPENCL_PARAM_INFO(link->w), |
| 213 | FF_OPENCL_PARAM_INFO(link->h), |
| 214 | FF_OPENCL_PARAM_INFO(cw), |
| 215 | FF_OPENCL_PARAM_INFO(ch), |
| 216 | NULL); |
| 217 | if (ret < 0) |
| 218 | return ret; |
| 219 | status = clEnqueueNDRangeKernel(unsharp->opencl_ctx.command_queue, |
| 220 | unsharp->opencl_ctx.kernel_luma, 2, NULL, |
| 221 | globalWorkSize2dLuma, localWorkSize2d, 0, NULL, NULL); |
| 222 | status |=clEnqueueNDRangeKernel(unsharp->opencl_ctx.command_queue, |
| 223 | unsharp->opencl_ctx.kernel_chroma, 2, NULL, |
| 224 | globalWorkSize2dChroma, localWorkSize2d, 0, NULL, NULL); |
| 225 | if (status != CL_SUCCESS) { |
| 226 | av_log(ctx, AV_LOG_ERROR, "OpenCL run kernel error occurred: %s\n", av_opencl_errstr(status)); |
| 227 | return AVERROR_EXTERNAL; |
| 228 | } |
| 229 | } else { /* use default kernel */ |
| 230 | kernel1.ctx = ctx; |
| 231 | kernel1.kernel = unsharp->opencl_ctx.kernel_default; |
| 232 | |
| 233 | ret = avpriv_opencl_set_parameter(&kernel1, |
| 234 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_inbuf), |
| 235 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_outbuf), |
| 236 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_luma_mask), |
| 237 | FF_OPENCL_PARAM_INFO(unsharp->opencl_ctx.cl_chroma_mask), |
| 238 | FF_OPENCL_PARAM_INFO(unsharp->luma.amount), |
| 239 | FF_OPENCL_PARAM_INFO(unsharp->chroma.amount), |
| 240 | FF_OPENCL_PARAM_INFO(unsharp->luma.steps_x), |
| 241 | FF_OPENCL_PARAM_INFO(unsharp->luma.steps_y), |
| 242 | FF_OPENCL_PARAM_INFO(unsharp->chroma.steps_x), |
| 243 | FF_OPENCL_PARAM_INFO(unsharp->chroma.steps_y), |
| 244 | FF_OPENCL_PARAM_INFO(unsharp->luma.scalebits), |
| 245 | FF_OPENCL_PARAM_INFO(unsharp->chroma.scalebits), |
| 246 | FF_OPENCL_PARAM_INFO(unsharp->luma.halfscale), |
| 247 | FF_OPENCL_PARAM_INFO(unsharp->chroma.halfscale), |
| 248 | FF_OPENCL_PARAM_INFO(in->linesize[0]), |
| 249 | FF_OPENCL_PARAM_INFO(in->linesize[1]), |
| 250 | FF_OPENCL_PARAM_INFO(out->linesize[0]), |
| 251 | FF_OPENCL_PARAM_INFO(out->linesize[1]), |
| 252 | FF_OPENCL_PARAM_INFO(link->h), |
| 253 | FF_OPENCL_PARAM_INFO(link->w), |
| 254 | FF_OPENCL_PARAM_INFO(ch), |
| 255 | FF_OPENCL_PARAM_INFO(cw), |
| 256 | NULL); |
| 257 | if (ret < 0) |
| 258 | return ret; |
| 259 | status = clEnqueueNDRangeKernel(unsharp->opencl_ctx.command_queue, |
| 260 | unsharp->opencl_ctx.kernel_default, 1, NULL, |
| 261 | &globalWorkSize1d, NULL, 0, NULL, NULL); |
| 262 | if (status != CL_SUCCESS) { |
| 263 | av_log(ctx, AV_LOG_ERROR, "OpenCL run kernel error occurred: %s\n", av_opencl_errstr(status)); |
| 264 | return AVERROR_EXTERNAL; |
| 265 | } |
| 266 | } |
| 267 | clFinish(unsharp->opencl_ctx.command_queue); |
| 268 | return av_opencl_buffer_read_image(out->data, unsharp->opencl_ctx.out_plane_size, |
| 269 | unsharp->opencl_ctx.plane_num, unsharp->opencl_ctx.cl_outbuf, |
| 270 | unsharp->opencl_ctx.cl_outbuf_size); |
| 271 | } |
| 272 | |
| 273 | int ff_opencl_unsharp_init(AVFilterContext *ctx) |
| 274 | { |
| 275 | int ret = 0; |
| 276 | char build_opts[96]; |
| 277 | UnsharpContext *unsharp = ctx->priv; |
| 278 | ret = av_opencl_init(NULL); |
| 279 | if (ret < 0) |
| 280 | return ret; |
| 281 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_luma_mask, |
| 282 | sizeof(uint32_t) * (2 * unsharp->luma.steps_x + 1) * (2 * unsharp->luma.steps_y + 1), |
| 283 | CL_MEM_READ_ONLY, NULL); |
| 284 | if (ret < 0) |
| 285 | return ret; |
| 286 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_chroma_mask, |
| 287 | sizeof(uint32_t) * (2 * unsharp->chroma.steps_x + 1) * (2 * unsharp->chroma.steps_y + 1), |
| 288 | CL_MEM_READ_ONLY, NULL); |
| 289 | if (ret < 0) |
| 290 | return ret; |
| 291 | ret = generate_mask(ctx); |
| 292 | if (ret < 0) |
| 293 | return ret; |
| 294 | unsharp->opencl_ctx.plane_num = PLANE_NUM; |
| 295 | unsharp->opencl_ctx.command_queue = av_opencl_get_command_queue(); |
| 296 | if (!unsharp->opencl_ctx.command_queue) { |
| 297 | av_log(ctx, AV_LOG_ERROR, "Unable to get OpenCL command queue in filter 'unsharp'\n"); |
| 298 | return AVERROR(EINVAL); |
| 299 | } |
| 300 | snprintf(build_opts, 96, "-D LU_RADIUS_X=%d -D LU_RADIUS_Y=%d -D CH_RADIUS_X=%d -D CH_RADIUS_Y=%d", |
| 301 | 2*unsharp->luma.steps_x+1, 2*unsharp->luma.steps_y+1, 2*unsharp->chroma.steps_x+1, 2*unsharp->chroma.steps_y+1); |
| 302 | unsharp->opencl_ctx.program = av_opencl_compile("unsharp", build_opts); |
| 303 | if (!unsharp->opencl_ctx.program) { |
| 304 | av_log(ctx, AV_LOG_ERROR, "OpenCL failed to compile program 'unsharp'\n"); |
| 305 | return AVERROR(EINVAL); |
| 306 | } |
| 307 | if (unsharp->opencl_ctx.use_fast_kernels) { |
| 308 | if (!unsharp->opencl_ctx.kernel_luma) { |
| 309 | unsharp->opencl_ctx.kernel_luma = clCreateKernel(unsharp->opencl_ctx.program, "unsharp_luma", &ret); |
| 310 | if (ret != CL_SUCCESS) { |
| 311 | av_log(ctx, AV_LOG_ERROR, "OpenCL failed to create kernel 'unsharp_luma'\n"); |
| 312 | return ret; |
| 313 | } |
| 314 | } |
| 315 | if (!unsharp->opencl_ctx.kernel_chroma) { |
| 316 | unsharp->opencl_ctx.kernel_chroma = clCreateKernel(unsharp->opencl_ctx.program, "unsharp_chroma", &ret); |
| 317 | if (ret < 0) { |
| 318 | av_log(ctx, AV_LOG_ERROR, "OpenCL failed to create kernel 'unsharp_chroma'\n"); |
| 319 | return ret; |
| 320 | } |
| 321 | } |
| 322 | } |
| 323 | else { |
| 324 | if (!unsharp->opencl_ctx.kernel_default) { |
| 325 | unsharp->opencl_ctx.kernel_default = clCreateKernel(unsharp->opencl_ctx.program, "unsharp_default", &ret); |
| 326 | if (ret < 0) { |
| 327 | av_log(ctx, AV_LOG_ERROR, "OpenCL failed to create kernel 'unsharp_default'\n"); |
| 328 | return ret; |
| 329 | } |
| 330 | } |
| 331 | } |
| 332 | return ret; |
| 333 | } |
| 334 | |
| 335 | void ff_opencl_unsharp_uninit(AVFilterContext *ctx) |
| 336 | { |
| 337 | UnsharpContext *unsharp = ctx->priv; |
| 338 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_inbuf); |
| 339 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_outbuf); |
| 340 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_luma_mask); |
| 341 | av_opencl_buffer_release(&unsharp->opencl_ctx.cl_chroma_mask); |
| 342 | clReleaseKernel(unsharp->opencl_ctx.kernel_default); |
| 343 | clReleaseKernel(unsharp->opencl_ctx.kernel_luma); |
| 344 | clReleaseKernel(unsharp->opencl_ctx.kernel_chroma); |
| 345 | clReleaseProgram(unsharp->opencl_ctx.program); |
| 346 | unsharp->opencl_ctx.command_queue = NULL; |
| 347 | av_opencl_uninit(); |
| 348 | } |
| 349 | |
| 350 | int ff_opencl_unsharp_process_inout_buf(AVFilterContext *ctx, AVFrame *in, AVFrame *out) |
| 351 | { |
| 352 | int ret = 0; |
| 353 | AVFilterLink *link = ctx->inputs[0]; |
| 354 | UnsharpContext *unsharp = ctx->priv; |
| 355 | int ch = FF_CEIL_RSHIFT(link->h, unsharp->vsub); |
| 356 | |
| 357 | if ((!unsharp->opencl_ctx.cl_inbuf) || (!unsharp->opencl_ctx.cl_outbuf)) { |
| 358 | unsharp->opencl_ctx.in_plane_size[0] = (in->linesize[0] * in->height); |
| 359 | unsharp->opencl_ctx.in_plane_size[1] = (in->linesize[1] * ch); |
| 360 | unsharp->opencl_ctx.in_plane_size[2] = (in->linesize[2] * ch); |
| 361 | unsharp->opencl_ctx.out_plane_size[0] = (out->linesize[0] * out->height); |
| 362 | unsharp->opencl_ctx.out_plane_size[1] = (out->linesize[1] * ch); |
| 363 | unsharp->opencl_ctx.out_plane_size[2] = (out->linesize[2] * ch); |
| 364 | unsharp->opencl_ctx.cl_inbuf_size = unsharp->opencl_ctx.in_plane_size[0] + |
| 365 | unsharp->opencl_ctx.in_plane_size[1] + |
| 366 | unsharp->opencl_ctx.in_plane_size[2]; |
| 367 | unsharp->opencl_ctx.cl_outbuf_size = unsharp->opencl_ctx.out_plane_size[0] + |
| 368 | unsharp->opencl_ctx.out_plane_size[1] + |
| 369 | unsharp->opencl_ctx.out_plane_size[2]; |
| 370 | if (!unsharp->opencl_ctx.cl_inbuf) { |
| 371 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_inbuf, |
| 372 | unsharp->opencl_ctx.cl_inbuf_size, |
| 373 | CL_MEM_READ_ONLY, NULL); |
| 374 | if (ret < 0) |
| 375 | return ret; |
| 376 | } |
| 377 | if (!unsharp->opencl_ctx.cl_outbuf) { |
| 378 | ret = av_opencl_buffer_create(&unsharp->opencl_ctx.cl_outbuf, |
| 379 | unsharp->opencl_ctx.cl_outbuf_size, |
| 380 | CL_MEM_READ_WRITE, NULL); |
| 381 | if (ret < 0) |
| 382 | return ret; |
| 383 | } |
| 384 | } |
| 385 | return av_opencl_buffer_write_image(unsharp->opencl_ctx.cl_inbuf, |
| 386 | unsharp->opencl_ctx.cl_inbuf_size, |
| 387 | 0, in->data, unsharp->opencl_ctx.in_plane_size, |
| 388 | unsharp->opencl_ctx.plane_num); |
| 389 | } |