Consider all packets in the parser, not just ones with timestamps,

this should greatly simplify handling of packet pos.

Originally committed as revision 17733 to svn://svn.ffmpeg.org/ffmpeg/trunk

libavcodec/parser.c

@@ -141,14 +141,12 @@ int av_parser_parse(AVCodecParserContext *s,
         buf = dummy_buf;
     } else {
         /* add a new packet descriptor */
-        if(pts != AV_NOPTS_VALUE || dts != AV_NOPTS_VALUE){
             i = (s->cur_frame_start_index + 1) & (AV_PARSER_PTS_NB - 1);
             s->cur_frame_start_index = i;
             s->cur_frame_offset[i] = s->cur_offset;
             s->cur_frame_end[i] = s->cur_offset + buf_size;
             s->cur_frame_pts[i] = pts;
             s->cur_frame_dts[i] = dts;
-        }
     }
 
     if (s->fetch_timestamp){