search

Telecommunication-Telemedia-Assessment / bitstream_mode3_p1204_3

Open source reference implementation of ITU-T P.1204.3
Other
58 stars 15 forks source link

Error：Segmentation fault (core dumped) #20

Closed ciwei123 closed 3 years ago

ciwei123 commented 3 years ago

@slhck Thanks for your sharing. When I run: sh /bitstream_mode3_p1204_3-master/p1204_3/bitstream_mode3_videoparser/parser.sh "test_videos/test_video_h264.mkv" --output "./tmp/test_video_h264.json.bz2" I get the error: FrmNum: 0 FrmType: I FrmSize: 77691 POC: 0 Blcky: -nan -nan -nan Segmentation fault (core dumped)

And I continue to run the command: python3 /bitstream_mode3_p1204_3-master/p1204_3/bitstream_mode3_videoparser/PythonInterface/parse.py test_videos/test_video_h264.mkv --output ./tmp/test_video_h264.json.bz2 --dll /bitstream_mode3_p1204_3-master/p1204_3/bitstream_mode3_videoparser/VideoParser/libvideoparser.so I get the error: python3: VideoStat264.c:60: GetVideoStatisticsH264: Assertion `(FrmStat->TrShapes[0] + FrmStat->TrShapes[2] + 4FrmStat->TrShapes[3]) == FrmStat->AnalyzedMBs16' failed. Aborted (core dumped) Could you help me? Thank you very much!!

slhck commented 3 years ago

Seems like an error in the bitstream parser itself. Would it be possible for you to share the source file that causes the issue?

ciwei123 commented 3 years ago

Seems like an error in the bitstream parser itself. Would it be possible for you to share the source file that causes the issue?

@slhck Thanks for your reply. And the video I used is provided by the project, test_video_h265.mkv、test_video_h264.mkv、test_video_vp9.mk. And the " /bitstream_mode3_p1204_3-master/p1204_3/bitstream_mode3_videoparser/PythonInterface/parse.py" is provided by https://github.com/Telecommunication-Telemedia-Assessment/bitstream_mode3_videoparser. Thank you very much!

slhck commented 3 years ago

This certainly should not happen. Could you please tell us which operating system you are using?

ciwei123 commented 3 years ago

This certainly should not happen. Could you please tell us which operating system you are using?

@slhck Thanks for your reply. Ubuntu16.04、a docker, Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, GPU: Tesla M40. Thank you very much!!

stg7 commented 3 years ago

well, that looks a bit weird, could you try to run it on the same CPU without this Tesla GPU? (that's just my educated guess that maybe the ffmpeg based parser has some GPU optimizations that produce such errors)

otherwise, the parser should also now work with Ubuntu 20.04 or 18.04, so you could try this in a docker container.