On Fri, 28 Feb 2014 11:04:17 +0100, Janne Grunau <[email protected]> wrote: > On 2014-02-27 21:40:09 +0100, Anton Khirnov wrote: > > avcodec_flush_buffers() must release all internally held references > > according to its documentation, for which all the threads need to be > > flushed. > > > > CC:[email protected] > > Bug-Id: vlc/9665 > > --- > > libavcodec/pthread_frame.c | 5 +++-- > > 1 file changed, 3 insertions(+), 2 deletions(-) > > > > diff --git a/libavcodec/pthread_frame.c b/libavcodec/pthread_frame.c > > index 36b4fff..1af8ff5 100644 > > --- a/libavcodec/pthread_frame.c > > +++ b/libavcodec/pthread_frame.c > > @@ -672,8 +672,6 @@ void ff_thread_flush(AVCodecContext *avctx) > > if (fctx->prev_thread) { > > if (fctx->prev_thread != &fctx->threads[0]) > > update_context_from_thread(fctx->threads[0].avctx, > > fctx->prev_thread->avctx, 0); > > - if (avctx->codec->flush) > > - avctx->codec->flush(fctx->threads[0].avctx); > > } > > > > fctx->next_decoding = fctx->next_finished = 0; > > @@ -686,6 +684,9 @@ void ff_thread_flush(AVCodecContext *avctx) > > av_frame_unref(p->frame); > > > > release_delayed_buffers(p); > > + > > + if (avctx->codec->flush) > > + avctx->codec->flush(p->avctx); > > } > > } > > > > ok if tested with THREADS=x; x > 1 >
I don't think flush gets called in the fate tests. In any case, the VLC reporter said this fixed the bug without apparent side effects. -- Anton Khirnov _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
