[ivtv-devel] System (practically) non-responsive after several hours...no clues
hverkuil at xs4all.nl
Fri Feb 23 10:39:13 CET 2007
On Friday 23 February 2007 00:18, Mr. Myth wrote:
> Brad Barnett wrote:
> > On Thu, 22 Feb 2007 13:51:58 -0800
> > "Rich Kadel" <rich.kadel at appeligo.com> wrote:
> >> The real test would be if all 4 PVR 500's will work (stable) with
> >> a new motherboard, and I don't have a motherboard in my desktop
> >> machine with 4 PCI slots. On a whim, I've crossed my fingers and
> >> purchased a new MSI 945P_Neo3-F and Intel Core 2 Duo 6600E (and
> >> 2Gigs DDR2) to replace the ASUS board and Opteron. Maybe this
> >> will do the trick. I'm recompiling the kernel to include SMP.
> >> Let me know if anything here sounds like a bad idea.
> > There were SMP issues a while back, but I believe that Hans has
> > solved them.
> > I don't know about this motherboard personally, so I can't comment
> > on it. I would check and make sure that the SATA controller and
> > other tidbits are fully supported. These days though, if the
> > motherboard is a few months old (that is, not just new on the
> > market), Linux is generally very good at supporting the mainline
> > items on it.
> I am not completely sure of this statement.
> I recently upgraded from 0.9.1 to 0.10.0 on a single cpu box with
> hyperthreading. I now get kernel panics within 4 hours if I run a
> SMP kernel whether I am recording or not. The box remains up but it
> causes the nic in the machine to stop responding and it stops
> registering interrupts.
> If I boot with maxcpus=1, then the box remains stable.
> Also rolling back to 0.9.1 with HT also gives me a stable box.
> I will try to submit a better bug report this weekend.
Please do. I have no problems at all on my dual core box, but I don't
have a HT motherboard so I cannot test that situation myself.
> ivtv-devel mailing list
> ivtv-devel at ivtvdriver.org
More information about the ivtv-devel