Video playback freezes/hangs with "<opensles_ringbuffer> No segment available"

[rland]

Hi,
I have a HLS stream that always gets stuck in a random location.
gst-launch-1.0 playbin uri=http://10.9.44.131/a012c3d1-1d3c-4b79-bdd2-0c6dd16aba3b/index.m3u8
I tested on PC and Android, always can reproduce on Android(GStreamer 1.8,1.10.1.11),and each time it stops at the following line log
---------
01-01 10:58:44.084 30006 30357 W : [openslesringbuffer.c:544] <opensles_ringbuffer> No segment available
---------

PC play with GStreamer no problem，ffplay/VLC play no problem.

I don't understand gl, don't know how to investigate it.
This HLS stream is a little big(almost 2G), Is there  a test server can upload HLS  resource to reproduce this issure?

Any help and advice is greatly appreciated.
regards

[Nicolas Dufresne-5]

Le jeudi 09 février 2017 à 01:12 -0800, lucky chou a écrit :

> Hi,
> I have a HLS stream that always gets stuck in a random location.
> gst-launch-1.0 playbin
> uri=http://10.9.44.131/a012c3d1-1d3c-4b79-bdd2-0c6dd16aba3b/index.m3u
> 8
> I tested on PC and Android, always can reproduce on Android(GStreamer
> 1.8,1.10.1.11),and each time it stops at the following line log
> ---------
> 01-01 10:58:44.084 30006 30357 W : [openslesringbuffer.c:544]
> <opensles_ringbuffer> No segment available
> ---------

OpenSLES is the library used to playback audio using native code on
Android. This error is a bit cryptic to me, so I wont comment much. You
may find the associated code in:

  gst-plugins-bad/sys/opensles/openslesringbuffer.c

It's possible this error case is not handled properly. In such pipeline
it's likely the clock is driven by the flow of data in that ring
buffer.

You may also try your luck with using the system clock.

  gst_pipeline_use_clock (pipeline, gst_system_clock_obtain ());

>
> PC play with GStreamer no problem，ffplay/VLC play no problem.
>
> I don't understand gl, don't know how to investigate it.
> This HLS stream is a little big(almost 2G), Is there  a test server
> can
> upload HLS  resource to reproduce this issure?
>
> Any help and advice is greatly appreciated. 
> regards
>
>
>
>
> --
> View this message in context: http://gstreamer-devel.966125.n4.nabble
> .com/Video-playback-freezes-hangs-with-opensles-ringbuffer-No-
> segment-available-tp4681779.html
> Sent from the GStreamer-devel mailing list archive at Nabble.com.
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[rland]

>OpenSLES is the library used to playback audio using native code on
>Android

Thanks for your suggestions.

> This error is a bit cryptic to me, so I wont comment much. You
>may find the associated code in:
>
> gst-plugins-bad/sys/opensles/openslesringbuffer.c
>
>It's possible this error case is not handled properly.

Does it mean that this issure is known,
or that it is only a special case for me?

If this is a known issure, then I mentioned the test video, probably not necessary.

So, is it necessary to manage it on the bugzilla (but I don't  know how to upload the test video, because it is so large)?

> In such pipeline
>it's likely the clock is driven by the flow of data in that ring
>buffer.
>
>You may also try your luck with using the system clock.
>
>  gst_pipeline_use_clock (pipeline, gst_system_clock_obtain ());

I've tried, although not completely solve the problem, but there are improvements.
Previously each time at a random point to stop playing,now  2~3 times will freezes..

[Nicolas Dufresne-5]

Le vendredi 10 février 2017 à 00:58 -0800, lucky chou a écrit :

> > This error is a bit cryptic to me, so I wont comment much. You
> > may find the associated code in:
> >
> > gst-plugins-bad/sys/opensles/openslesringbuffer.c
> >
> > It's possible this error case is not handled properly.
>
> Does it mean that this issure is known, 
> or that it is only a special case for me?
>
> If this is a known issure, then I mentioned the test video, probably
> not
> necessary.
>
> So, is it necessary to manage it on the bugzilla (but I don't  know
> how to
> upload the test video, because it is so large)?

I believe you are the first to report such an issue on Android. Filing
a bug with the problem and your research is probably the next step
indeed.

>
> > In such pipeline
> > it's likely the clock is driven by the flow of data in that ring
> > buffer.
> >
> > You may also try your luck with using the system clock.
> >
> >   gst_pipeline_use_clock (pipeline, gst_system_clock_obtain ());
>
> I've tried, although not completely solve the problem, but there are
> improvements.
> Previously each time at a random point to stop playing,now  2~3 times
> will
> freezes..

This test result shall probably be reported in your bug report.
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[rland]

This same with this bug.