avfilter/dctdnoiz: add slice threading

This commit is contained in:
Clément Bœsch 2014-08-10 21:49:55 +02:00
parent cf24e10a09
commit c85e0ca68a
1 changed files with 84 additions and 30 deletions

View File

@ -36,18 +36,22 @@
static const char *const var_names[] = { "c", NULL }; static const char *const var_names[] = { "c", NULL };
enum { VAR_C, VAR_VARS_NB }; enum { VAR_C, VAR_VARS_NB };
#define MAX_THREADS 8
typedef struct DCTdnoizContext { typedef struct DCTdnoizContext {
const AVClass *class; const AVClass *class;
/* coefficient factor expression */ /* coefficient factor expression */
char *expr_str; char *expr_str;
AVExpr *expr; AVExpr *expr[MAX_THREADS];
double var_values[VAR_VARS_NB]; double var_values[MAX_THREADS][VAR_VARS_NB];
int nb_threads;
int pr_width, pr_height; // width and height to process int pr_width, pr_height; // width and height to process
float sigma; // used when no expression are st float sigma; // used when no expression are st
float th; // threshold (3*sigma) float th; // threshold (3*sigma)
float *cbuf[2][3]; // two planar rgb color buffers float *cbuf[2][3]; // two planar rgb color buffers
float *slices[MAX_THREADS]; // slices buffers (1 slice buffer per thread)
float *weights; // dct coeff are cumulated with overlapping; these values are used for averaging float *weights; // dct coeff are cumulated with overlapping; these values are used for averaging
int p_linesize; // line sizes for color and weights int p_linesize; // line sizes for color and weights
int overlap; // number of block overlapping pixels int overlap; // number of block overlapping pixels
@ -56,7 +60,8 @@ typedef struct DCTdnoizContext {
int bsize; // block size, 1<<n int bsize; // block size, 1<<n
void (*filter_freq_func)(struct DCTdnoizContext *s, void (*filter_freq_func)(struct DCTdnoizContext *s,
const float *src, int src_linesize, const float *src, int src_linesize,
float *dst, int dst_linesize); float *dst, int dst_linesize,
int thread_id);
void (*color_decorrelation)(float **dst, int dst_linesize, void (*color_decorrelation)(float **dst, int dst_linesize,
const uint8_t *src, int src_linesize, const uint8_t *src, int src_linesize,
int w, int h); int w, int h);
@ -377,16 +382,17 @@ static av_always_inline void filter_freq_##bsize(const float *src, int src_lines
\ \
static void filter_freq_sigma_##bsize(DCTdnoizContext *s, \ static void filter_freq_sigma_##bsize(DCTdnoizContext *s, \
const float *src, int src_linesize, \ const float *src, int src_linesize, \
float *dst, int dst_linesize) \ float *dst, int dst_linesize, int thread_id) \
{ \ { \
filter_freq_##bsize(src, src_linesize, dst, dst_linesize, NULL, NULL, s->th); \ filter_freq_##bsize(src, src_linesize, dst, dst_linesize, NULL, NULL, s->th); \
} \ } \
\ \
static void filter_freq_expr_##bsize(DCTdnoizContext *s, \ static void filter_freq_expr_##bsize(DCTdnoizContext *s, \
const float *src, int src_linesize, \ const float *src, int src_linesize, \
float *dst, int dst_linesize) \ float *dst, int dst_linesize, int thread_id) \
{ \ { \
filter_freq_##bsize(src, src_linesize, dst, dst_linesize, s->expr, s->var_values, 0); \ filter_freq_##bsize(src, src_linesize, dst, dst_linesize, \
s->expr[thread_id], s->var_values[thread_id], 0); \
} }
DEF_FILTER_FREQ_FUNCS(8) DEF_FILTER_FREQ_FUNCS(8)
@ -475,7 +481,7 @@ static int config_input(AVFilterLink *inlink)
{ {
AVFilterContext *ctx = inlink->dst; AVFilterContext *ctx = inlink->dst;
DCTdnoizContext *s = ctx->priv; DCTdnoizContext *s = ctx->priv;
int i, x, y, bx, by, linesize, *iweights; int i, x, y, bx, by, linesize, *iweights, max_slice_h, slice_h;
const int bsize = 1 << s->n; const int bsize = 1 << s->n;
switch (inlink->format) { switch (inlink->format) {
@ -500,6 +506,11 @@ static int config_input(AVFilterLink *inlink)
av_log(ctx, AV_LOG_WARNING, "The last %d vertical pixels won't be denoised\n", av_log(ctx, AV_LOG_WARNING, "The last %d vertical pixels won't be denoised\n",
inlink->h - s->pr_height); inlink->h - s->pr_height);
max_slice_h = s->pr_height / ((s->bsize - 1) * 2);
s->nb_threads = FFMIN3(MAX_THREADS, ctx->graph->nb_threads, max_slice_h);
av_log(ctx, AV_LOG_DEBUG, "threads: [max=%d hmax=%d user=%d] => %d\n",
MAX_THREADS, max_slice_h, ctx->graph->nb_threads, s->nb_threads);
s->p_linesize = linesize = FFALIGN(s->pr_width, 32); s->p_linesize = linesize = FFALIGN(s->pr_width, 32);
for (i = 0; i < 2; i++) { for (i = 0; i < 2; i++) {
s->cbuf[i][0] = av_malloc(linesize * s->pr_height * sizeof(*s->cbuf[i][0])); s->cbuf[i][0] = av_malloc(linesize * s->pr_height * sizeof(*s->cbuf[i][0]));
@ -509,6 +520,27 @@ static int config_input(AVFilterLink *inlink)
return AVERROR(ENOMEM); return AVERROR(ENOMEM);
} }
/* eval expressions are probably not thread safe when the eval internal
* state can be changed (typically through load & store operations) */
if (s->expr_str) {
for (i = 0; i < s->nb_threads; i++) {
int ret = av_expr_parse(&s->expr[i], s->expr_str, var_names,
NULL, NULL, NULL, NULL, 0, ctx);
if (ret < 0)
return ret;
}
}
/* each slice will need to (pre & re)process the top and bottom block of
* the previous one in in addition to its processing area. This is because
* each pixel is averaged by all the surrounding blocks */
slice_h = (int)ceilf(s->pr_height / s->nb_threads) + (s->bsize - 1) * 2;
for (i = 0; i < s->nb_threads; i++) {
s->slices[i] = av_malloc_array(linesize, slice_h * sizeof(*s->slices[i]));
if (!s->slices[i])
return AVERROR(ENOMEM);
}
s->weights = av_malloc(s->pr_height * linesize * sizeof(*s->weights)); s->weights = av_malloc(s->pr_height * linesize * sizeof(*s->weights));
if (!s->weights) if (!s->weights)
return AVERROR(ENOMEM); return AVERROR(ENOMEM);
@ -544,10 +576,6 @@ static av_cold int init(AVFilterContext *ctx)
} }
if (s->expr_str) { if (s->expr_str) {
int ret = av_expr_parse(&s->expr, s->expr_str, var_names,
NULL, NULL, NULL, NULL, 0, ctx);
if (ret < 0)
return ret;
switch (s->n) { switch (s->n) {
case 3: s->filter_freq_func = filter_freq_expr_8; break; case 3: s->filter_freq_func = filter_freq_expr_8; break;
case 4: s->filter_freq_func = filter_freq_expr_16; break; case 4: s->filter_freq_func = filter_freq_expr_16; break;
@ -576,36 +604,56 @@ static int query_formats(AVFilterContext *ctx)
return 0; return 0;
} }
static void filter_plane(AVFilterContext *ctx, typedef struct ThreadData {
float *dst, int dst_linesize, float *src, *dst;
const float *src, int src_linesize, } ThreadData;
int w, int h)
static int filter_slice(AVFilterContext *ctx,
void *arg, int jobnr, int nb_jobs)
{ {
int x, y; int x, y;
DCTdnoizContext *s = ctx->priv; DCTdnoizContext *s = ctx->priv;
float *dst0 = dst; const ThreadData *td = arg;
const float *weights = s->weights; const int w = s->pr_width;
const int h = s->pr_height;
const int slice_start = (h * jobnr ) / nb_jobs;
const int slice_end = (h * (jobnr+1)) / nb_jobs;
const int slice_start_ctx = FFMAX(slice_start - s->bsize + 1, 0);
const int slice_end_ctx = FFMIN(slice_end, h - s->bsize + 1);
const int slice_h = slice_end_ctx - slice_start_ctx;
const int src_linesize = s->p_linesize;
const int dst_linesize = s->p_linesize;
const int slice_linesize = s->p_linesize;
float *dst;
const float *src = td->src + slice_start_ctx * src_linesize;
const float *weights = s->weights + slice_start * dst_linesize;
float *slice = s->slices[jobnr];
// reset block sums // reset block sums
memset(dst, 0, h * dst_linesize * sizeof(*dst)); memset(slice, 0, (slice_h + s->bsize - 1) * dst_linesize * sizeof(*slice));
// block dct sums // block dct sums
for (y = 0; y < h - s->bsize + 1; y += s->step) { for (y = 0; y < slice_h; y += s->step) {
for (x = 0; x < w - s->bsize + 1; x += s->step) for (x = 0; x < w - s->bsize + 1; x += s->step)
s->filter_freq_func(s, src + x, src_linesize, s->filter_freq_func(s, src + x, src_linesize,
dst + x, dst_linesize); slice + x, slice_linesize,
jobnr);
src += s->step * src_linesize; src += s->step * src_linesize;
dst += s->step * dst_linesize; slice += s->step * slice_linesize;
} }
// average blocks // average blocks
dst = dst0; slice = s->slices[jobnr] + (slice_start - slice_start_ctx) * slice_linesize;
for (y = 0; y < h; y++) { dst = td->dst + slice_start * dst_linesize;
for (y = slice_start; y < slice_end; y++) {
for (x = 0; x < w; x++) for (x = 0; x < w; x++)
dst[x] *= weights[x]; dst[x] = slice[x] * weights[x];
slice += slice_linesize;
dst += dst_linesize; dst += dst_linesize;
weights += dst_linesize; weights += dst_linesize;
} }
return 0;
} }
static int filter_frame(AVFilterLink *inlink, AVFrame *in) static int filter_frame(AVFilterLink *inlink, AVFrame *in)
@ -632,10 +680,13 @@ static int filter_frame(AVFilterLink *inlink, AVFrame *in)
s->color_decorrelation(s->cbuf[0], s->p_linesize, s->color_decorrelation(s->cbuf[0], s->p_linesize,
in->data[0], in->linesize[0], in->data[0], in->linesize[0],
s->pr_width, s->pr_height); s->pr_width, s->pr_height);
for (plane = 0; plane < 3; plane++) for (plane = 0; plane < 3; plane++) {
filter_plane(ctx, s->cbuf[1][plane], s->p_linesize, ThreadData td = {
s->cbuf[0][plane], s->p_linesize, .src = s->cbuf[0][plane],
s->pr_width, s->pr_height); .dst = s->cbuf[1][plane],
};
ctx->internal->execute(ctx, filter_slice, &td, NULL, s->nb_threads);
}
s->color_correlation(out->data[0], out->linesize[0], s->color_correlation(out->data[0], out->linesize[0],
s->cbuf[1], s->p_linesize, s->cbuf[1], s->p_linesize,
s->pr_width, s->pr_height); s->pr_width, s->pr_height);
@ -687,7 +738,10 @@ static av_cold void uninit(AVFilterContext *ctx)
av_free(s->cbuf[i][1]); av_free(s->cbuf[i][1]);
av_free(s->cbuf[i][2]); av_free(s->cbuf[i][2]);
} }
av_expr_free(s->expr); for (i = 0; i < s->nb_threads; i++) {
av_free(s->slices[i]);
av_expr_free(s->expr[i]);
}
} }
static const AVFilterPad dctdnoiz_inputs[] = { static const AVFilterPad dctdnoiz_inputs[] = {
@ -718,5 +772,5 @@ AVFilter ff_vf_dctdnoiz = {
.inputs = dctdnoiz_inputs, .inputs = dctdnoiz_inputs,
.outputs = dctdnoiz_outputs, .outputs = dctdnoiz_outputs,
.priv_class = &dctdnoiz_class, .priv_class = &dctdnoiz_class,
.flags = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC, .flags = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC | AVFILTER_FLAG_SLICE_THREADS,
}; };