[FFmpeg-trac] #8430(undetermined:new): Regression in pnm parsing at commit 9fc1031ac2e8691e0140854d727b58cb62431b2b

[FFmpeg]

#8430: Regression in pnm parsing at commit
9fc1031ac2e8691e0140854d727b58cb62431b2b
--------------------------------------+----------------------------------
             Reporter:  symphorien    |                     Type:  defect
               Status:  new           |                 Priority:  normal
            Component:  undetermined  |                  Version:  4.2
             Keywords:                |               Blocked By:
             Blocking:                |  Reproduced by developer:  0
Analyzed by developer:  0             |
--------------------------------------+----------------------------------
 Summary of the bug:
 when using the image2pipe input mode to transform a stream of pnm images
 to mp4, ffmpeg 4.2 and master complains about invalid input format. The
 same file works fine with ffmpeg 3.
 I bisected the regression to commit:
 {{{
 commit 9fc1031ac2e8691e0140854d727b58cb62431b2b (HEAD, tag: bug,
 refs/bisect/bad)
 Author: Michael Niedermayer <michael at niedermayer.cc>
 Date:   Wed Apr 24 21:14:19 2019 +0200

     avcodec/pnm_parser: Remember the size of the image and do not reparse
 the header
 }}}
 which seems related.
 How to reproduce:
 {{{
 $  target/bin/ffmpeg -f image2pipe -vcodec ppm -i pipe: -an -r 25 -y
 -vcodec mpeg4 -s 1920x1080 -aspect 16/9 /tmp/fodsfd.mp4 < small.data
 ffmpeg version N-93685-g9fc1031ac2 Copyright (c) 2000-2019 the FFmpeg
 developers
   built with gcc 8.3.0 (GCC)
   configuration: --prefix=/home/symphorien/src/ffmpeg/target --enable-
 libx264 --enable-gpl
   libavutil      56. 26.100 / 56. 26.100
   libavcodec     58. 52.100 / 58. 52.100
   libavformat    58. 27.103 / 58. 27.103
   libavdevice    58.  7.100 / 58.  7.100
   libavfilter     7. 50.100 /  7. 50.100
   libswscale      5.  4.100 /  5.  4.100
   libswresample   3.  4.100 /  3.  4.100
   libpostproc    55.  4.100 / 55.  4.100
 [image2pipe @ 0x3c72840] Stream #0: not enough frames to estimate rate;
 consider increasing probesize
 Input #0, image2pipe, from 'pipe:':
   Duration: N/A, bitrate: N/A
     Stream #0:0: Video: ppm, rgb24, 1920x1080, 25 tbr, 25 tbn, 25 tbc
 Stream mapping:
   Stream #0:0 -> #0:0 (ppm (native) -> mpeg4 (native))
 Output #0, mp4, to '/tmp/fodsfd.mp4':
   Metadata:
     encoder         : Lavf58.27.103
     Stream #0:0: Video: mpeg4 (mp4v / 0x7634706D), yuv420p, 1920x1080 [SAR
 1:1 DAR 16:9], q=2-31, 200 kb/s, 25 fps, 12800 tbn, 25 tbc
     Metadata:
       encoder         : Lavc58.52.100 mpeg4
     Side data:
       cpb: bitrate max/min/avg: 0/0/200000 buffer size: 0 vbv_delay: -1
 Error while decoding stream #0:0: Invalid data found when processing
 inputkbits/s speed=0.276x
     Last message repeated 1 times
 frame=  240 fps=6.9 q=31.0 Lsize=    1694kB time=00:00:09.56
 bitrate=1451.6kbits/s speed=0.274x
 video:1692kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB
 muxing overhead: 0.111327%
 }}}


 Unfortunately, the smallest reproducer file I could create is 1.4 GB.
 (~140 MB compressed). I tried to upload it to upload.ffmpeg.org as
 explained on https://ffmpeg.org/bugreports.html but I could not connect on
 the server.

 To check the integrity of the input file, I split it into the individual
 frames with the unix split command, and examined them briefly with an
 image viewer and they seem valid.

--
Ticket URL: <https://trac.ffmpeg.org/ticket/8430>
FFmpeg <https://ffmpeg.org>
FFmpeg issue tracker