Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[vdr] Re: Try to fix "unknown picture error"



Am Donnerstag, 11. September 2003 23:45 schrieben Sie:
> Klaus Schmidinger wrote:
> > Tilo Renkl wrote:
> > > Hi,
> > >
> > > I searched my LOG's, and found that I've never had a VDR-Installation
> > > without sc-Plugin, so as you know that sc prevents this bug (at
> > > least ut to version 0.1.11), I can not say whitch "vanilla"-version
> > > of VDR introduced that bug. *SORRY*
> > >
> > > In my eyes it must have something to do with CAM-handling, because
> > > sc-Plugin prevents it.
> >
> > Just an idea (since I can't be of much help otherwise because I don't
> > have this problem): The CAM handling in VDR does quite a bit of section
> > filtering, so if you don't use any CAM, you might want to try commenting
> > out line 1167 of VDR/eit.c:
> >
> >       AddFilter(0x00, 0x00);  // PAT
> >
> > Whitout this line, VDR won't set any filters for CA information.
>
> But this doesn't explain, why the combination vdr 1.1.26 and DVB from
> 27.4.03 throws the "unknown picture error" and the same vdr 1.1.26 with DVB
> from 11.6.03 don't cause this error (until now). And as I said, I had vdr
> 1.1.26 (with driver from 11.6.03 as I remember) for many weeks without
> block artefacts and the "unknown picture error". But if you have any
> further idea... let me know.
>
> Martin

I wonder, if we are looking at the right place for the reason of this errors.

1. Many people never had this error on their systems with one or more cards 
1a. Klaus never saw this error on his system on a suse 7.3

2. Some people have the error with different versions of driver and vdr in 
various combinations;
2a. on "infected systems" the error occurs on 1-card systems (I had it some 
days ago) and 2-card systems with different types of cards
2b. Some of the now "infected" had a errorfree system earlier

If I see this points one is clear for me: the reason cannot be vdr or driver 
alone.If the error would be only in vdr, why have we 1 and 1a?
I think it is an interaction with the running system (kernel, libraries) with 
vdr and driver. What shared libraries do guys of 2 and 2a have?
What did the guys with the "infected" systems (2b) do except updating vdr and 
driver, before the error ocours?

For me, one big thing was the update from Suse 7.3 to 8.2 and now I have on a 
1-card and 2-card system these errors.

So my conclusion, lets compare our systems and libraries that are loaded by 
vdr.
ldd ./vdr1.2.25pre2 gives on my suse 8.2:
        libjpeg.so.62 => /usr/lib/libjpeg.so.62 (0x40025000)
        libpthread.so.0 => /lib/libpthread.so.0 (0x40044000)
        libdl.so.2 => /lib/libdl.so.2 (0x40095000)
        libstdc++.so.5 => /usr/lib/libstdc++.so.5 (0x40098000)
        libm.so.6 => /lib/libm.so.6 (0x40150000)
        libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x40172000)
        libc.so.6 => /lib/libc.so.6 (0x4017a000)
        /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
Maybe some of you know, how to show the libs loaded by the driver too.

Please give me some feedback, does anybody with the same system and libs have 
a stable system?

Harald


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe vdr" as subject.



Home | Main Index | Thread Index