Hi,
This is a follow on thread from
this.
I fixed the problem initially by having the BIOS re-scan the hardware. This forced the SCSI card to have an interrupt all to itself, instead of sharing with the nVidia video card. However, I now have hooked up a UPS to one one the USB ports. Previously, I didn't have any USB devices connected. This caused the USB to now use the same IRQ as the SCSI card, so I now have a bunch of errors again:
Jan 8 13:53:02 The-Tardis kernel: scsi0: PCI error Interrupt at seqaddr = 0x9
Jan 8 13:53:02 The-Tardis kernel: scsi0: Data Parity Error Detected during address or writ
e data phase
Jan 8 13:53:02 The-Tardis kernel: scsi0: WARNING WARNING WARNING WARNING
Jan 8 13:53:02 The-Tardis kernel: scsi0: Too many PCI parity errors observed as a target.
Jan 8 13:53:02 The-Tardis kernel: scsi0: Some device on this bus is generating bad parity.
Jan 8 13:53:02 The-Tardis kernel: scsi0: This is an error *observed by*, not *generated by
*, this controller.
Jan 8 13:53:02 The-Tardis kernel: scsi0: PCI parity error checking has been disabled.
Jan 8 13:53:02 The-Tardis kernel: scsi0: WARNING WARNING WARNING WARNING
Here's how the IRQs are assigned:
CPU0
0: 1124175 XT-PIC timer
1: 4093 XT-PIC i8042
2: 0 XT-PIC cascade
3: 55530 XT-PIC NE2000
4: 127 XT-PIC serial
5: 6029 XT-PIC ES1688
8: 2 XT-PIC rtc
9: 1 XT-PIC acpi
10: 324629 XT-PIC eth1, nvidia
11: 14517 XT-PIC aic7xxx, uhci_hcd:usb1
12: 36672 XT-PIC i8042
14: 18176 XT-PIC ide0
15: 160280 XT-PIC ide1
NMI: 0
ERR: 5
I haven't tried switching the card slots around yet, but I don't think that will help, because it always seems to "double" them up.
Is there a way to "force" the SCSI card to use an IRQ all by itself. Or, to make the USB share IRQ 10 instead.
Cheers,
Eddie