[mythtv-users] PVR-250 ivtv hang on capture?

[Chris Delis]

Could it be related to the VIA chipset bug?? (kt400)

http://ivtv.sourceforge.net/FAQ.html#_jump21

On Tue, 21 Oct 2003, Joel Conover wrote:

> Cross posting from ivtv list --  any hunches on what could be causing this? KT400 chipset issue perhaps?  Any suggestions on a fix and or workaround?
> 
> 
> 
> 
> I've been plunking around with a new PVR-250 to build a myth box, and have successfully set up a functional system on a 1.7 Ghz Athlon machine with a KT400 chipset. The install is verbatim from Jarod Wilson's instructions for myth, using Axel Thimm's ATRPMS packages....
> 
> 
> 
> I've been trying to 'burn the system in', and set up a live tv 'capture' (live play) and let it run for a few days.  Trial run #1 hung up in the middle of the night.  Trial run #2 ran for 36 hours straight, but then I hopped out and jumped back in (via the myth menus), and found nothing but a black screen. I'm running debug=1, so I don't get a lot in the message log, but I did get this....
> 
> 
> 
> Oct 20 23:23:58 mythtv kernel: saa7115[0]: decoder set size
> Oct 20 23:23:58 mythtv kernel: Hpsc: 0x00001, Hfsc: 0x00600
> Oct 20 23:23:58 mythtv kernel: Setting full NTSC height
> Oct 20 23:27:45 mythtv kernel: ivtv: DMA Error. Result: 0x0000000b
> 
> 
> Any suggestions?  Hitting escape to go back to the myth menu took me back once, but dumping back into capture spelled the end of that... I could still access the machine (mouse down to the taskbar) and kill mythfrontend, but I couldn't do a cat /dev/video0 > /tmp/test.mpg any more. No errors, just sits for ~40 sec and dumps out with a 0 byte file.
> 
> 
> 
> I got the impression that this driver was pretty stable (for alpha) and haven't seen any similar posts, so I'm throwing it out here.  Feel free to ask questions/make suggestions for further debugging -- I'm new to IVTV and Myth, but UNIX competent; I know there is  a lot more info that could be pertinent here, but I don't know what's useful, so I'll wait for questions back.
> 
> 
> 
> Thanks 
> 
> Joel
>