Re: [SMP, USB] UHCI interrupt wrongly routed?

[Guennadi Liakhovetski]

On Wed, 24 Nov 2004, Greg KH wrote:

> On Wed, Nov 24, 2004 at 11:57:15PM +0100, Guennadi Liakhovetski wrote:
> > 2-way running 2.6.9 the onboard UHCI device is configured to IRQ 9 via 
> > XT-PIC???
> > 
> >            CPU0       CPU1       
> >   0:     588054        123    IO-APIC-edge  timer
> >   1:        360          0    IO-APIC-edge  i8042
> >   2:          0          0          XT-PIC  cascade
> >   8:          4          0    IO-APIC-edge  rtc
> >   9:          0          0    IO-APIC-edge  acpi
> >  11:          0          0          XT-PIC  uhci_hcd
> >  15:          2          0    IO-APIC-edge  ide1
> >  16:        173          1   IO-APIC-level  ehci_hcd
> >  17:          0          0   IO-APIC-level  ohci_hcd
> >  18:          0          0   IO-APIC-level  ohci_hcd
> >  19:      99999          1   IO-APIC-level  ohci_hcd
> >  20:        332          0   IO-APIC-level  eth0
> >  21:       2576          0   IO-APIC-level  sym53c8xx
> > NMI:          0          0 
> > LOC:     587726     587988 
> > ERR:          0
> > MIS:          0

...

> > Linux version 2.6.9-rc4-tmscsim (lyakh@xxxxxxxxxxxxx) (gcc version 3.3.2 (Debian)) #1 SMP Mon Oct 25 23:38:23 CEST 2004
> 
> Can you try 2.6.10-rc2 or the latest -bk snapshot?  Hopefully this is
> fixed there.

Unfortunately, the problem is still there with rc2. Do you expect a later 
bk snapshot to be better? I don't really understand how software can fix 
this - it seems to be a BIOS bug - IRQ 11 in irq-pin. Where, looks like, 
it should be 19 (irq 19: nobody cared immediately on USB-insertion). 
Notice, I already run with pci=noacpi, there was one more mis-assigned irq 
in acpi table, don't remember which one though. So, can it really be fixed 
in (stock) kernel, without quirks? Is my only solution PCI rescan with 
dummy-php? Or fix acpi table?... I just still don't understand why irq 11 
is configured as XT-PIC?

Thanks
Guennadi
---
Guennadi Liakhovetski

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/