@@ -393,6 +393,83 @@ static int formats_declared(AVFilterContext *f)
return 1;
}
+static int insert_auto_filter(AVFilterContext **convert, AVFilterGraph *graph,
+ AVFilterLink *link, const AVFilterNegotiation *neg,
+ unsigned conv_step, int *converter_count, void *log_ctx)
+{
+ int ret;
+ const AVFilter *filter;
+ AVFilterContext *ctx;
+ AVFilterLink *inlink, *outlink;
+ char inst_name[30];
+ const char *opts = neg->conversion_filters[conv_step].conversion_opts_offset == 0 ? NULL :
+ FF_FIELD_AT(char *, neg->conversion_filters[conv_step].conversion_opts_offset, *graph);
+ const char *name = neg->conversion_filters[conv_step].conversion_filter;
+
+ if (graph->disable_auto_convert) {
+ av_log(log_ctx, AV_LOG_ERROR,
+ "The filters '%s' and '%s' do not have a common format "
+ "and automatic conversion is disabled.\n",
+ link->src->name, link->dst->name);
+ return AVERROR(EINVAL);
+ }
+
+ if (!(filter = avfilter_get_by_name(name))) {
+ av_log(log_ctx, AV_LOG_ERROR,
+ "'%s' filter not present, cannot convert formats.\n",
+ name);
+ return AVERROR(EINVAL);
+ }
+ snprintf(inst_name, sizeof(inst_name), "auto_%s_%d",
+ name, (*converter_count)++);
+ ret = avfilter_graph_create_filter(&ctx, filter, inst_name, opts, NULL, graph);
+ if (ret < 0)
+ return ret;
+
+ if ((ret = avfilter_insert_filter(link, ctx, 0, 0)) < 0)
+ return ret;
+
+ if ((ret = filter_query_formats(ctx)) < 0)
+ return ret;
+
+ inlink = ctx->inputs[0];
+ outlink = ctx->outputs[0];
+ av_assert0( inlink->incfg.formats->refcount > 0);
+ av_assert0( inlink->outcfg.formats->refcount > 0);
+ av_assert0(outlink->incfg.formats->refcount > 0);
+ av_assert0(outlink->outcfg.formats->refcount > 0);
+ if (outlink->type == AVMEDIA_TYPE_AUDIO) {
+ av_assert0( inlink-> incfg.samplerates->refcount > 0);
+ av_assert0( inlink->outcfg.samplerates->refcount > 0);
+ av_assert0(outlink-> incfg.samplerates->refcount > 0);
+ av_assert0(outlink->outcfg.samplerates->refcount > 0);
+ av_assert0( inlink-> incfg.channel_layouts->refcount > 0);
+ av_assert0( inlink->outcfg.channel_layouts->refcount > 0);
+ av_assert0(outlink-> incfg.channel_layouts->refcount > 0);
+ av_assert0(outlink->outcfg.channel_layouts->refcount > 0);
+ }
+
+ *convert = ctx;
+ return 0;
+}
+
+static int merge_auto_filter(AVFilterContext *convert, const AVFilterNegotiation *neg)
+{
+ int neg_step, ret;
+ AVFilterLink *inlink = convert->inputs[0];
+ AVFilterLink *outlink = convert->outputs[0];
+#define MERGE(merger, link) \
+ ((merger)->merge(FF_FIELD_AT(void *, (merger)->offset, (link)->incfg), \
+ FF_FIELD_AT(void *, (merger)->offset, (link)->outcfg)))
+ for (neg_step = 0; neg_step < neg->nb_mergers; neg_step++) {
+ const AVFilterFormatsMerger *m = &neg->mergers[neg_step];
+ if ((ret = MERGE(m, inlink)) <= 0 ||
+ (ret = MERGE(m, outlink)) <= 0)
+ break;
+ }
+ return ret;
+}
+
/**
* Perform one round of query_formats() and merging formats lists on the
* filter graph.
@@ -433,7 +510,7 @@ static int query_formats(AVFilterGraph *graph, void *log_ctx)
for (j = 0; j < filter->nb_inputs; j++) {
AVFilterLink *link = filter->inputs[j];
const AVFilterNegotiation *neg;
- unsigned neg_step;
+ unsigned neg_step, conv_step;
int convert_needed = 0;
if (!link)
@@ -469,68 +546,50 @@ static int query_formats(AVFilterGraph *graph, void *log_ctx)
}
if (convert_needed) {
- AVFilterContext *convert;
- const AVFilter *filter;
- AVFilterLink *inlink, *outlink;
- char inst_name[30];
- const char *opts;
-
- if (graph->disable_auto_convert) {
- av_log(log_ctx, AV_LOG_ERROR,
- "The filters '%s' and '%s' do not have a common format "
- "and automatic conversion is disabled.\n",
- link->src->name, link->dst->name);
- return AVERROR(EINVAL);
- }
-
- /* couldn't merge format lists. auto-insert conversion filter */
- if (!(filter = avfilter_get_by_name(neg->conversion_filter))) {
- av_log(log_ctx, AV_LOG_ERROR,
- "'%s' filter not present, cannot convert formats.\n",
- neg->conversion_filter);
- return AVERROR(EINVAL);
- }
- snprintf(inst_name, sizeof(inst_name), "auto_%s_%d",
- neg->conversion_filter, converter_count++);
- opts = FF_FIELD_AT(char *, neg->conversion_opts_offset, *graph);
- ret = avfilter_graph_create_filter(&convert, filter, inst_name, opts, NULL, graph);
- if (ret < 0)
- return ret;
- if ((ret = avfilter_insert_filter(link, convert, 0, 0)) < 0)
- return ret;
-
- if ((ret = filter_query_formats(convert)) < 0)
- return ret;
+ for (conv_step = 0; conv_step < neg->nb_conversion_filters; conv_step++) {
+ AVFilterContext *convert;
+ ret = insert_auto_filter(&convert, graph, link, neg,
+ conv_step, &converter_count, log_ctx);
+ if (ret < 0) {
+ av_log(log_ctx, AV_LOG_ERROR, "Failed to insert an auto filter.\n");
+ return ret;
+ }
- inlink = convert->inputs[0];
- outlink = convert->outputs[0];
- av_assert0( inlink->incfg.formats->refcount > 0);
- av_assert0( inlink->outcfg.formats->refcount > 0);
- av_assert0(outlink->incfg.formats->refcount > 0);
- av_assert0(outlink->outcfg.formats->refcount > 0);
- if (outlink->type == AVMEDIA_TYPE_AUDIO) {
- av_assert0( inlink-> incfg.samplerates->refcount > 0);
- av_assert0( inlink->outcfg.samplerates->refcount > 0);
- av_assert0(outlink-> incfg.samplerates->refcount > 0);
- av_assert0(outlink->outcfg.samplerates->refcount > 0);
- av_assert0( inlink-> incfg.channel_layouts->refcount > 0);
- av_assert0( inlink->outcfg.channel_layouts->refcount > 0);
- av_assert0(outlink-> incfg.channel_layouts->refcount > 0);
- av_assert0(outlink->outcfg.channel_layouts->refcount > 0);
- }
-#define MERGE(merger, link) \
- ((merger)->merge(FF_FIELD_AT(void *, (merger)->offset, (link)->incfg), \
- FF_FIELD_AT(void *, (merger)->offset, (link)->outcfg)))
- for (neg_step = 0; neg_step < neg->nb_mergers; neg_step++) {
- const AVFilterFormatsMerger *m = &neg->mergers[neg_step];
- if ((ret = MERGE(m, inlink)) <= 0 ||
- (ret = MERGE(m, outlink)) <= 0) {
- if (ret < 0)
- return ret;
- av_log(log_ctx, AV_LOG_ERROR,
- "Impossible to convert between the formats supported by the filter "
- "'%s' and the filter '%s'\n", link->src->name, link->dst->name);
- return AVERROR(ENOSYS);
+ ret = merge_auto_filter(convert, neg);
+ if (ret < 0)
+ return ret;
+ else if (ret > 0)
+ break;
+ else if (conv_step < neg->nb_conversion_filters - 1) {
+ AVFilterLink *inlink = convert->inputs[0];
+ AVFilterLink *outlink = convert->outputs[0];
+ av_log(log_ctx, AV_LOG_VERBOSE,
+ "Impossible to convert between the formats supported by the filter "
+ "'%s' and the filter '%s', try another conversion filter.\n",
+ link->src->name, link->dst->name);
+ unsigned dstpad_idx = outlink->dstpad - outlink->dst->input_pads;
+ converter_count--;
+ /* re-hookup the link */
+ inlink->dst = outlink->dst;
+ inlink->dstpad = &outlink->dst->input_pads[dstpad_idx];
+ outlink->dst->inputs[dstpad_idx] = inlink;
+ if (outlink->outcfg.formats)
+ ff_formats_changeref(&outlink->outcfg.formats,
+ &inlink->outcfg.formats);
+ if (outlink->outcfg.samplerates)
+ ff_formats_changeref(&outlink->outcfg.samplerates,
+ &inlink->outcfg.samplerates);
+ if (outlink->outcfg.channel_layouts)
+ ff_channel_layouts_changeref(&outlink->outcfg.channel_layouts,
+ &inlink->outcfg.channel_layouts);
+ /* remove the previous auto filter */
+ convert->inputs[0] = NULL;
+ convert->outputs[0]->dst = NULL;
+ avfilter_free(convert);
+ } else {
+ av_log(log_ctx, AV_LOG_ERROR,
+ "Impossible to convert between the formats supported by the filter "
+ "'%s' and the filter '%s'\n", link->src->name, link->dst->name);
}
}
}
@@ -326,18 +326,36 @@ static const AVFilterFormatsMerger mergers_audio[] = {
},
};
+static const AVFilterFormatsFilter filters_video[] = {
+ {
+ .conversion_filter = "scale",
+ .conversion_opts_offset = offsetof(AVFilterGraph, scale_sws_opts),
+ },
+ {
+ .conversion_filter = "hwmap",
+ .conversion_opts_offset = 0,
+ }
+};
+
+static const AVFilterFormatsFilter filters_audio[] = {
+ {
+ .conversion_filter = "aresample",
+ .conversion_opts_offset = offsetof(AVFilterGraph, aresample_swr_opts),
+ }
+};
+
static const AVFilterNegotiation negotiate_video = {
.nb_mergers = FF_ARRAY_ELEMS(mergers_video),
.mergers = mergers_video,
- .conversion_filter = "scale",
- .conversion_opts_offset = offsetof(AVFilterGraph, scale_sws_opts),
+ .nb_conversion_filters = FF_ARRAY_ELEMS(filters_video),
+ .conversion_filters = filters_video,
};
static const AVFilterNegotiation negotiate_audio = {
.nb_mergers = FF_ARRAY_ELEMS(mergers_audio),
.mergers = mergers_audio,
- .conversion_filter = "aresample",
- .conversion_opts_offset = offsetof(AVFilterGraph, aresample_swr_opts),
+ .nb_conversion_filters = FF_ARRAY_ELEMS(filters_audio),
+ .conversion_filters = filters_audio,
};
const AVFilterNegotiation *ff_filter_get_negotiation(AVFilterLink *link)
@@ -330,6 +330,11 @@ typedef struct AVFilterFormatMerger {
int (*can_merge)(const void *a, const void *b);
} AVFilterFormatsMerger;
+typedef struct AVFilterFormatFilter {
+ const char *conversion_filter;
+ unsigned conversion_opts_offset;
+} AVFilterFormatsFilter;
+
/**
* Callbacks and properties to describe the steps of a format negotiation.
*
@@ -418,8 +423,8 @@ typedef struct AVFilterFormatMerger {
typedef struct AVFilterNegotiation {
unsigned nb_mergers;
const AVFilterFormatsMerger *mergers;
- const char *conversion_filter;
- unsigned conversion_opts_offset;
+ unsigned nb_conversion_filters;
+ const AVFilterFormatsFilter *conversion_filters;
} AVFilterNegotiation;
const AVFilterNegotiation *ff_filter_get_negotiation(AVFilterLink *link);
When two formats lists cannot be merged, a scale filter is auto-inserted. However, when it comes to hardware map, we have to manually add a hwmap filter to do the conversion. This patch introduces an auto hwmap filter to do the hwmap conversion automatically and modifies the negotiation structures, making it extensible for more auto filters. Signed-off-by: Tong Wu <tong1.wu@intel.com> --- libavfilter/avfiltergraph.c | 183 ++++++++++++++++++++++++------------ libavfilter/formats.c | 26 ++++- libavfilter/formats.h | 9 +- 3 files changed, 150 insertions(+), 68 deletions(-)