Jump to content


Photo

OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05


  • Please log in to reply
172 replies to this topic

Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #41 habab

  • Member
  • 4 posts

0
Neutral

Posted 9 January 2016 - 11:55

@MastaG

same for Enfinity? :blink:



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #42 MastaG

  • Senior Member
  • 1,531 posts

+118
Excellent

Posted 9 January 2016 - 13:41

Thanks again for your dedication Chris :)

To everyone else, I'm a bit busy this weekend.

But I'll try my best to solve all the problems asap.



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #43 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 9 January 2016 - 16:34

Hi Chris

I have at least 3 times and each time completely rebuilt again had the spinner, if you have found a solution that would be great and I am you very grateful.

Do you already have fixed the bug in the Commit? then
I build my smooth again a new image.

https://github.com/c...64a6131fbe90704


@MastaG

It's the weekend, spend your free time with your family, there is no hurry for me.

Greeting Akki

Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #44 yuriy1975h

  • Member
  • 7 posts

0
Neutral

Posted 9 January 2016 - 17:20

@yuriy1975h

Would it be possible for you to do the following?

Login using telnet and type: lsmod | grep rt

Then paste the output here.

I'd like to know which wifi kernel module is loaded.

 

There are a few that ship with Taapat's drivers and a few from OE-A.

I need to know which one has been loaded :)

 

MastaG the answer to the command lsmod | grep rt from your image.

 

[H[Jroot@spark:~# [6n
lsmod | grep rt


rt5370sta             546700  0
smartcard              25168  0



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #45 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 9 January 2016 - 22:27

@ Chris

Just for info, a new image have just built with your changes to the gstdvbvideosink.c,
unfortunately comes with me still at the start of
Streams of Spinner.

Why is this happening at the VU + Boxes and
other Boxes go.


Greeting Akki

Edited by Akki, 9 January 2016 - 22:28.


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #46 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 10 January 2016 - 10:14

@ Chris

Just for info, a new image have just built with your changes to the gstdvbvideosink.c,
unfortunately comes with me still at the start of
Streams of Spinner.

Why is this happening at the VU + Boxes and
other Boxes go.


Greeting Akki

Well gues you do something wrong with the build.

 

Just try the here included mediasink for vuduo2 gst-1.7.1

 

unzip the file and copy the 3 files in map /usr/lib/gstreamer-1.0/

libgstdvbaudiosink.so
libgstdvbvideosink.so
libgstmpeg4p2unpack.so

to You're stb , use ftp for that (overwrite the existing ones)

 

Just reboot the box.

Attached Files



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #47 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 10 January 2016 - 14:29

Thank you for your help Chris, I build again
new and then comes also still the spinner, I'm going to the 3 files from you
replace with the originals and see what happens.

So I habs just tested, although I your 3
Files use comes after selecting the streams above the buffer display, stops at 1%, then goes after 3sec. away and then comes the spinner.

#DESCRIPTION Test3
#SERVICE 4097:0:0:0:0:0:0:0:0:0:http%3a//xg5tbqm52h1pot9.babahhcdn.com/witch/season1/304680.mp4:Dougnut test

Thank you again for your effort to help me.

Greeting Akki

Edited by Akki, 10 January 2016 - 14:31.


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #48 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 10 January 2016 - 15:12

By me it all goes 100 % so something else in you're build is really wrong. What I don't .

 

By those streams I Never a spinner that's on vuduo2 and dm8000 .

 

Looks You have a ghost in you're stb. :P

 

I'm well using master-next, last gstreamer head.



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #49 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 10 January 2016 - 15:37

Hi Chris


I doubt also slow to me, either I do what is completely wrong when building the image, or as you already write in my Duo2 a spirit of wants to annoy me.

I'll send you times ne PN, with all the steps and Entries in the files.

Somewhere but must be the gremlins have hidden.

Greeting Akki

Edited by Akki, 10 January 2016 - 15:38.


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #50 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 10 January 2016 - 20:01

@ALL DEVELOPERS

 

I just found out a so what super CRAZY issue. 

 

Like some perhaps already saw that on vuplus series with some media some persons reported sand keeper problems. Example off those media tracks :

#SERVICE 4097:0:0:0:0:0:0:0:0:0:http%3A//xg5tbqm52h1pot9.babahhcdn.com/witch/season1/304680.mp4:Dougnut test
#DESCRIPTION Dougnut test
#SERVICE 4097:0:0:0:0:0:0:0:0:0:http%3a//cdn.phoenix.intergi.com/19850/videos/4357110/video-sd.mp4:Test MP4
#DESCRIPTION Test MP4
#SERVICE 4097:0:0:0:0:0:0:0:0:0:https%3a//cdn.video.playwire.com/19850/videos/4369729/video-sd.mp4?hosting_id%3d19850%2522%2520bitrate=%25221200%2522%2520height=%2522480%2522%2520width=%2522854%2522%0d:playwire test%0d
#DESCRIPTION playwire test

I tried to check and by me (master-next now) gst head off 4 days old .(1.7.1.1)

 

I actually could not reproduce the issue on my vuduo2 (but I well had seen that at a certain time ) since I do work in general with a slight modified e2 at base, more specific it's just one rule i commented in servicemp3.cpp.  The rule I comment is..

		case GST_MESSAGE_BUFFERING:
			if (m_sourceinfo.is_streaming)
			{
				GstBufferingMode mode;
				gst_message_parse_buffering(msg, &(m_bufferInfo.bufferPercent));
				//eDebug("[eServiceMP3] Buffering %u percent done", m_bufferInfo.bufferPercent); //this rule is commented by me
				gst_message_parse_buffering_stats(msg, &mode, &(m_bufferInfo.avgInRate), &(m_bufferInfo.avgOutRate), &(m_bufferInfo.bufferingLeft));
				m_event((iPlayableService*)this, evBuffering);

I just do that to be rid off the pretty intensive logging about the buffering into logs when I study sequencing.

 

Well what happened I just did an upgrade on gstreamer (that's my personal dev versions at this time ) and just also set my self up to date with last next-master . When I do that I always re-flash my box and run in first stage with unmodified e2.

 

I first before anything else did a quick test off media and ..... BINGO.

 

My vuduo2 had the same problem like Akki reported.

 

I spend now so what the whole day in finding out whatever I did it remained.

 

I just first (with the goal to see more into log without endless buffering messages) to comment my line then just updated enigma2.

 

GUESS WHAT ....... ALL sand keeper problems for that media GONE ..

 

To be shure I really first cleaned all recipes concerning gstreamer dvbmediasink e2 and e2 plugins. (with clean in my builds where I could)

 

Then rebuild. flashed the box and yes sand keeper problem. Then just modified e2 and commented the line mentioned above. rebuild and reinstalled e2 on box.

 

NO SAND KEEPER ANY MORE. and this with the box running normally.

 

Looks like eDebug does cause issues when used in messages which are pretty comming on fast and plenty. Even if we run normally



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #51 theparasol

  • Senior Member
  • 4,157 posts

+198
Excellent

Posted 10 January 2016 - 20:31

What happens if you replace "m_bufferInfo.bufferPercent" with a static value?

 

Can it be that "m_bufferInfo.bufferPercent" triggers a "GST_MESSAGE_BUFFERING" so you get basically a loop?


@Camping: ZGemma H.2S, Technisat Multytenne 4-in-1 @Home: Edision Mini 4K, Wave Frontier T55, EMP Centauri EMP DiSEqC 8/1 switch, 4x Inverto Ultra Black single LNB


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #52 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 10 January 2016 - 21:51

Well whit a static value no problem.

 

But The logging off buffering is tremendous I now noticed .

 

It's for shure a error to put eDebug in buffer log.

 

The console log is so full that only with a static value this is een

 

static value set for test

				int test = 3;
				eDebug("[eServiceMP3] Buffering %d percent done", test);


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #53 littlesat

  • PLi® Core member
  • 57,460 posts

+708
Excellent

Posted 10 January 2016 - 23:11

Nice finding.....!
And when you copy the value in an integer?

Edited by littlesat, 10 January 2016 - 23:12.

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #54 Taapat

  • PLi® Core member
  • 2,345 posts

+121
Excellent

Posted 11 January 2016 - 07:57

If we have started here to change something, maybe we need to do more.
I personally disable this message in enigma log.
I think that without this can't be normal watch a live log.
In the log is just buffering messages.
I think that would be more appropriate if in the log to show only a message that the buffering start, and message when bufering ends with the value.


Edited by Taapat, 11 January 2016 - 07:58.


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #55 littlesat

  • PLi® Core member
  • 57,460 posts

+708
Excellent

Posted 11 January 2016 - 08:07

Thie percentage ie also used to display it on the osd.... Could there be some log mechanism if it changes quickly when the value needs also to be displayed on the eDebug...?

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #56 athoik

  • PLi® Core member
  • 8,458 posts

+327
Excellent

Posted 11 January 2016 - 08:59

How about display only when 0, 50, 100?

 

if (!(m_bufferInfo.bufferPercent % 50))
    eDebug("[eServiceMP3] Buffering %u percent done", m_bufferInfo.bufferPercent); 

Wavefield T90: 0.8W - 1.9E - 4.8E - 13E - 16E - 19.2E - 23.5E - 26E - 33E - 39E - 42E - 45E on EMP Centauri DiseqC 16/1
Unamed: 13E Quattro - 9E Quattro on IKUSI MS-0916

Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #57 Taapat

  • PLi® Core member
  • 2,345 posts

+121
Excellent

Posted 11 January 2016 - 09:46

I think it is a very good solution.
Maybe can display every 20, but definitely not as that is now.



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #58 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 11 January 2016 - 10:05

Well The parsing off the logging is used to display on screen.

 

But the eDebug not.

 

If You just remove the line :

eDebug("[eServiceMP3] Buffering %u percent done", m_bufferInfo.bufferPercent);

or just comment it , the sand keeper problem's are gone and , but on screen self I still see the buffering percentage.

 

What I found so weird is that even if you run with just normal init 3 (whitout logging to console or file) The freezeing occurs.



Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #59 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 11 January 2016 - 10:18

By me now I just use a e2 patch to comment the message.

 

I think the parsed info is used in :

ePtr<iStreamBufferInfo> eServiceMP3::getBufferCharge()
{
	return new eStreamBufferInfo(m_bufferInfo.bufferPercent, m_bufferInfo.avgInRate, m_bufferInfo.avgOutRate, m_bufferInfo.bufferingLeft, m_buffer_size);
}

Attached Files


Edited by christophecvr, 11 January 2016 - 10:20.


Re: OpenPLi-4 / Gstreamer 1.7.1 (master git) images 2016-01-05 #60 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 11 January 2016 - 10:37

m_bufferInfo.bufferPercent

 

Is also used in function located servicemp3.c :

int eServiceMP3::getInfo(int w)

Declared in servicemp3.h :

int eServiceMP3::getInfo(int w)

Edited by christophecvr, 11 January 2016 - 10:37.



4 user(s) are reading this topic

0 members, 4 guests, 0 anonymous users