[linux-dvb] Re: IRQ issues with Nova-S CI when using CI Module
Andrew de Quincey
adq_dvb at lidskialf.net
Mon Jul 25 17:24:24 CEST 2005
On Monday 25 July 2005 15:08, Michael Ditum wrote:
> Thanks for that Andrew, I'll try the latest CVS version in a while,
> I'm currently having trouble getting 2.6.13 booting at the moment...
> I'll have a look to see if your modification are ok once I've got that
> Back to the other IRQ problem...
> I've done some more testing. I've now tested it on 2 servers which
> have similar, but not exactly the same, hardware. It occurs on our HP
> Proliant ML350 G4 and our HP Proliant ML370 G4.
> The problem only seems to occur on bootup from a poweroff. If I
> reboot, it then starts working correctly until I power it off.
> 2 different print outs to /var/log/messages depending on if the card
> has a CI or not.
> Jul 13 16:27:34 bloodhound kernel: Uhhuh. NMI received. Dazed and
> confused, but trying to continue
> Jul 13 16:27:34 bloodhound kernel: You probably have a hardware
> problem with your RAM chips
> Jul 13 16:27:34 bloodhound kernel: irq 11: nobody cared!
> Jul 13 16:27:34 bloodhound kernel: [<c014efc4>] __report_bad_irq+0x24/0x7d
> Jul 13 16:27:34 bloodhound kernel: [<c014f0a6>] note_interrupt+0x6b/0x89
> 10: 11805 XT-PIC ioc0, ioc1, saa7146 (0), saa7146 (1),
> uhci_hcd:usb3, uhci_hcd:usb4
> 11: 3900000 XT-PIC ehci_hcd:usb1, uhci_hcd:usb2,
> uhci_hcd:usb5, eth0
Bear with me - its been a year or so since I messed about with ACPI.
So you're getting IRQ 11 when you shouldn't be effectively. Since its
happening on two servers, this is likely either a kernel bug or a BIOS bug. I
take it the cards work fine on other (non-HP) machines?
AFAIR, the motherboard chipset of modern machines can usually be reprogrammed
to route IRQs to a variety of destinations. This looks like linux thinks the
IRQ for your DVB cards is 10 when actually your motherboard chipset is set up
for IRQ 11.
In legacy (non-ACPI) mode linux has a set of functions for each chipset for
configuring IRQ routing. It may be that there is a bug in this, or else this
is a newer chipset it doesn't quite know everything about (some manufactuers
don't release this documentation :( ). Or else your BIOS is setting it up
wrongly... I'm sorry I can't remember the exact details.
You're running your motherboard in legacy (non-ACPI) mode. Have you tried
using ACPI under linux? Under ACPI, linux executes a set of virtual machine
instructions on startup. This configures the hardware without linux having to
have specific support for that chipset.
Another possibility might be to upgrade the bios.
More information about the linux-dvb