On Mon, 2011-06-06 at 21:10 +0200, Ramon Hofer wrote:
> On 06/06/2011 01:31 PM, Andy Walls wrote: 
> > On Mon, 2011-06-06 at 13:09 +0200, Ramon Hofer wrote: 
> > > Hi all 
> > > 
> > > I have had a problem with my PVR-500 which I was able to solve by 
> > > patching the kernel. 
> > > The messages about "[ivtv-users] WinTV PVR-500 stops recording
> > > after 
> > > some minutes    (ivtv0: DMA TIMEOUT)" here on the list describe
> > > my 
> > > problem and Andy's solutions. Thanks again! 
> > > 
> > > Unfortunately my system still has problems when recording. 
> > > It crashes from time to time and reboots automatically or just
> > > Freezes 
> > > after writing some error messages (I could take a photo of them if
> > > you'd 
> > > want to know them). 
> > I don't have time for a proper reply, but here is some reading for
> > you: 
> > 
> > http://ivtvdriver.org/pipermail/ivtv-users/2009-August/009528.html 
> > http://ivtvdriver.org/pipermail/ivtv-users/2009-August/009532.html 
> > http://ivtvdriver.org/pipermail/ivtv-users/2009-August/009539.html 
> > 
> > It's all very disheartening.  You essentially have to experiment
> > your 
> > way through an enormous amount of unkonws to figure out why the PCI 
> > chipset on your motherboard is unhappy enough to reboot the system.
> > All 
> > with no error logging about the triggering event.  The labor cost
> > alone 
> > is probably more than the price of another Desktp PC with a new
> > video 
> > card. 
> > 
> > In the end, Jeroen did actaully find the problem on his system.  It
> > was 
> > caused by a change in the kernel that modified the way his
> > motherboard 
> > PCI chipset was set up. 
> Thanks for the reply! 
> 
> Unfortunately I can't spent that much time trying to locate the
> error. 
> 
> Is it correct that the problem is caused by the card and that there
> should be no problems with the PVR-150? 

No, the problem is caused by the system and the interactions of its
subsystems, components, firmware and software.

I can't guarantee you any success because:

- Like all great products rushed to market, the CX23415 and CX23416 have
bugs in both the hardware and firmware.  I'll speculate that a
manifestation of those bugs can possibly be the CX2341[56] DMA-ing data
to some address where it shouldn't, or maybe causing an unusual
condition on the PCI bus.

- PCI motherboard chipset and PCI bridge chip manufacturers are all
implementing to the PCI specifications.  The implementers can interpret
things differently and/or decide to do things differently within the
language of the specifications.

- From what I understand, Internext Compression really only tested the
CX2341[56] chips against one specific vendor's PCI chipsets during
development.  

- The Linux kernel, including the ivtv driver, is never bug free.  The
Linux kernel can also set up motherboard PCI chipsets in a manner that
is quite different from the way Windows sets them up.


The best rule of thumb I can give you, for building a system with
CX2341[56] based devices, is to use a motherboard with an Intel PCI
chipset and specifically avoid VIA PCI chipsets.


> How are the chances that the PVR-500 works on a different mainboard? 

Overall, very good.  The random PCI reboot problem is rare; although the
PLX PCI bridge chip on the PVR-500 board appears to be a contributor.

Unfortunately, there is no absolute way to know until you test an
assembled system.

Regards,
Andy

> Best regards 
> Ramon 




_______________________________________________
ivtv-users mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-users

Reply via email to