• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 98 subscribers
  • Views 743 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Linux/DRA746: fastforward video update is not proper with waylandsink

Ikshwaku Chauhan
Expert 2630 points
Part Number: DRA746

Tool/software: Linux

Hello Team,

We are using custom h/w and software based on DRA746 and SDK 3.03.

In one or our use case for fastforward playback, it is not fast-forwarding but playing the previous frame.And getting the following logs gstreamer logs:

456479 2018/05/18 21:58:51.000000 245.1340 155 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(753):codec_process:<decoder> push failed flushing
456480 2018/05/18 21:58:51.000000 245.1342 156 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(1472):gst_ducati_viddec_chain:<decoder> push from codec_process failed flushing
457536 2018/05/18 21:58:52.000000 245.1476 157 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs WARN ducati gstducatividdec.c(570):codec_process:<decoder> err=-1, extendedError=00040000
457537 2018/05/18 21:58:52.000000 245.1477 158 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs ERROR ducati gstducati.c(61):gst_ducati_log_extended_error_info: Bit 18 (00040000): stream end
457540 2018/05/18 21:58:52.000000 245.1524 161 ECU1 CrM HMPE 2092 log info verbose 1 T4174 timeThread: seekpos: 6051700680 rate: 2000

458659 2018/05/18 21:58:53.000000 246.1590 195 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(753):codec_process:<decoder> push failed flushing
458660 2018/05/18 21:58:53.000000 246.1592 196 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(1472):gst_ducati_viddec_chain:<decoder> push from codec_process failed flushing
458663 2018/05/18 21:58:53.000000 246.1776 198 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs WARN ducati gstducatividdec.c(570):codec_process:<decoder> err=-1, extendedError=00040000
458664 2018/05/18 21:58:53.000000 246.1781 199 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs ERROR ducati gstducati.c(61):gst_ducati_log_extended_error_info: Bit 18 (00040000): stream end
458668 2018/05/18 21:58:53.000000 246.1838 202 ECU1 CrM HMPE 2092 log info verbose 1 T4174 timeThread: seekpos: 1929886621 rate: 2000

459442 2018/05/18 21:58:53.000000 246.9903 231 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(753):codec_process:<decoder> push failed flushing
459443 2018/05/18 21:58:53.000000 246.9904 232 ECU1 CrM HMPE 2092 log info verbose 1 T2926 GSTLogs WARN ducati gstducatividdec.c(1472):gst_ducati_viddec_chain:<decoder> push from codec_process failed flushing
459458 2018/05/18 21:58:53.000000 247.0140 233 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs WARN ducati gstducatividdec.c(570):codec_process:<decoder> err=-1, extendedError=00040000
459459 2018/05/18 21:58:53.000000 247.0141 234 ECU1 CrM HMPE 2092 log info verbose 1 T4174 GSTLogs ERROR ducati gstducati.c(61):gst_ducati_log_extended_error_info: Bit 18 (00040000): stream end
459475 2018/05/18 21:58:53.000000 247.0169 235 ECU1 CrM HMPE 2092 log info verbose 1 T4174 timeThread: seekpos: 1764693877 rate: 2000

Could you please help us in identifying what went wrong here..? 

Thanks,

Ikshwaku 

  • 0
    TI__Mastermind 30480 points
    Hi Ikshwaku,
    Are you using HMI to do fast-forward/rewind?
    TI has a way to test trick-play with an application gsttestplayer.
    Refer this processors.wiki.ti.com/index.php
    but this doesn't use playbin.

    Is there a way to reproduce the issue with any sample application on EVM?
    Can you please share an observation video to understand the issue ?

    Thanks
  • 0 in reply to Ramprasad
    Expert 2630 points
    Hello Ram,

    Yes we are using HMI for fast forward.

    Share you the observation videos through our portal.

    Thanks
  • 0 in reply to Ikshwaku Chauhan
    TI__Mastermind 30480 points
    Hi Ikshwaku,
    Can you confirm if this issue is observed for all streams or all formats (h264, mpeg4, wmv etc) and all resolutions or for a specific stream?

    I am not able to get any clue on which plugin is making the issue here.
    To narrow down if the issue is with ducati or not you can try remove libgstducati.so from /usr/bin/gstramer1-.0 to force playbin to
    choose software decoder and set use-drm=true for waylandsink and check the behaviour. Let me know the result.
  • 0 in reply to Ramprasad
    TI__Mastermind 30480 points
    Hi Ikshwaku,
    Did you get chance to test with software decoders?

    Ram
  • 0 in reply to Ramprasad
    Expert 2630 points

    Hello Ram,

    We are not able to reproduce this Fast Forward issue with our current builds.
    But, we have another fast forward issue specific to WMV files, where when we do fast forward with WMV files then after some time video hangs. Currently we are putting things in order for VPE on A15 in our builds, then we will test it if Fast Forward for WMV files is resolved or not.
    Will update you in couple of days.

    Thanks for our support.

    Regards,
    Ikshwaku

  • 0 in reply to Ikshwaku Chauhan
    TI__Mastermind 30480 points
    Hi Ikshwaku,
    I will close this issue.

    Can you let me know the VC1-decoder version used in your system. This can you get by
    cat /sys/kernel/debug/remoteproc/remoteproc1/trace0 on target. It will print all codec versions.

    We have fixed hang issue with VC1 decoder with fast-forward.
    If the VC1 version is VC1D-01.00.02.05, then your system doesn't have the fix . VC1D-01.00.02.06 has this fix for this hang issue with VC1 decoder with trick-play.

    Refer the commits of git://git.ti.com/ivimm/ipumm.git to know all fixes required.
    We have fixed hang issue with other decoders also. I would recommend you to take the fix by updating the ipumm recipe.

    Let me know if you need any help here.

    Ram
  • 0 in reply to Ramprasad
    Expert 2630 points
    Hello Ram,

    Could you please help in identifying the patches, that we can take for the the updated version of VC1 .

    Currently, we are using this ipumm-fw version " 3.00.13.00-r5" and this SRCREV = "fd441443a4289c801a0c8d9f00b6966f7fe3476c".

    Thanks,
    Ikshwaku
  • 0 in reply to Ikshwaku Chauhan
    TI__Mastermind 30480 points
    Hi Ikshwaku,
    This commit-id more than a year old and there are many codec fixes.
    If you are looking for only for this bugfix, I suggest if you can take the patch from commit-id 59ec47f1236cca02c155199a748da5d1af9e2745.
    There are more than 50 commits after fd441443a4289c801a0c8d9f00b6966f7fe3476c. This includes IPC/BIOS vversion changes also.

    If you want to get all the codec fixes , then you will need to migrate IPC, BIOS etc also. IPC version is migrated in 3.04.

    Please check if you can apply only this patch with commit-id 59ec47f1236cca02c155199a748da5d1af9e2745 from git://git.ti.com/ivimm/ipumm.git

    Ram