Re: dvbinput.cpp CVS - broken

Please send the logs, otherwise I'm not able to deduce where and what 
exactly goes wrong.

Grtz,
Jean-Paul Saman.

Alexander Afanasyev wrote:
> Hi!
> 
> Unfortunately nothing good :(( same error.
> Error goes from normaltrickplay::initwork,
> after
> iRc = m_pConverter->GetNextTsPacket(pPacket)
> it thinks,thinks... after dvbreader says ! TimeOut !
> and everything stops...

Need logs because this could be vital information about where it starts 
to go wrong in contrast to where it crashes.

> 
> I found a lot of changes in this area.
> InitWork() ts2ts skiped some data for syncing,
> but m_pConverter->GetNextTsPacket(pPacket), where default
> value of second parameter is 0, i think nothing skip...
> i tried to add 1 //(pPacket,1)// but nothing good happens...
> 
This parameter "skip" is for doing trickplay and has nothing to do with 
this problem

> :(
> 
> Alexander Afanasyev
> 
> Sunday, July 27, 2003, 4:09:41 PM, you wrote:
> 
> 
>>Alexander,
> 
> 
>>I think that I figured out what goes wrong. Update the CVS version 
>>again. My previous cleanup removed a few lines too much.
> 
> 
>>Alexander Afanasyev wrote:
>>
>>>Hmm, interesting things...
>>>When i debug, i se only printf 'before trickplay' and
>>>program delays for a while and then says that can't
>>>start Trickplay (as i wrote in previous mail).
>>>There is a code cut
>>>-- cut from code --
>>>    dvb->SetTP(dvb->chans[iIndex].tpid, dvb->chans[iIndex].satid);
>>>    dvb->set_front();
>>>
>>>    sleep(1);
>>>    printf( "testing after sleep, before trickplay\n" );
>>>
>>>    //Launch the demux
>>>    m_pTrickPlay->Create();
>>>    printf( "testing after trickplay\n" );
>>>-- end cut --
>>>
>>>also, i try add printf to DoWork() of NormalTrickPlay
>>>but also nothing printed out :( Why thread do not start???
>>
> 
>>Not sure, but I think it has something to do with having the m_pHandler 
>>assigned.
> 
> 
>>>and one more, i try to change place for m_pTrickPlay->Create()
>>>(after SelectPid) and i got a SegmentationFault...
>>
>>This is a race condition. It should be before SelectPID, because they 
>>both work on the m_pHandler.
> 
> 
>>Grtz,
>>Jean-Paul Saman.
> 
> 



-- 
This is the vls-devel mailing-list, see http://www.videolan.org/streaming/
To unsubscribe, please read http://developers.videolan.org/lists.html
If you are in trouble, please contact <[email protected]>