On Wed, Apr 26, 2017 at 4:09 AM, wm4 <nfx...@googlemail.com> wrote: > On Tue, 25 Apr 2017 23:52:04 +0700 > Muhammad Faiz <mfc...@gmail.com> wrote: > >> when frame is received, not from other threads. >> >> Should fix fate failure with THREADS>=4: >> make fate-h264-attachment-631 THREADS=4 >> >> Signed-off-by: Muhammad Faiz <mfc...@gmail.com> >> --- >> libavcodec/pthread_frame.c | 4 ++++ >> 1 file changed, 4 insertions(+) >> >> diff --git a/libavcodec/pthread_frame.c b/libavcodec/pthread_frame.c >> index 13d6828..c452ed7 100644 >> --- a/libavcodec/pthread_frame.c >> +++ b/libavcodec/pthread_frame.c >> @@ -547,6 +547,10 @@ int ff_thread_decode_frame(AVCodecContext *avctx, >> >> fctx->next_finished = finished; >> >> + /* if frame is returned, properly set err from the thread that return >> frame */ >> + if (*got_picture_ptr) >> + err = p->result; >> + >> /* return the size of the consumed packet if no error occurred */ >> if (err >= 0) >> err = avpkt->size; > > Well, the logic confuses me. Does this override an earlier set err > value?
Yes, because an earlier set err value may be from a different thread. >Could err be set to the correct value in the first place (inside > of the loop)? No, it was intended on 32a5b631267 _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
