firmware-testing-team team mailing list archive

[Colin Ian King <colin.king@xxxxxxxxxxxxx>]

On 15/11/12 15:53, Brendan Donegan wrote:
> On 15/11/12 10:33, Colin Ian King wrote:
> > On 15/11/12 10:26, Brendan Donegan wrote:
> > > Hi,
> > >
> > > I've come across this critical FWTS failure on one system which
> > > otherwise looks fine for certification:
> > >
> > > FAILED [CRITICAL] KlogDisablingIRQ: Test 1, CRITICAL Kernel message: [
> > > 12.502739] Disabling IRQ #16
> > > ADVICE: The kernel detected an irq storm. This is most probably an IRQ
> > > routing bug.
> > >
> > > (Full Log)
> > > http://paste.ubuntu.com/1359992/
> > >
> > > Shall I raise a kernel bug for this and can anyone shed any light on the
> > > possible effects of this error/what it means?
> >
> > So, this kernel error comes from kernel/irq/spurious.c,
> > note_interrupt() which has detected that IRQ 16 has generated > 99900
> > interrupts but it seems that it's not being responded to so the IRQ
> > has been disabled.  We should have a bug filed and see why this is
> > misbehaving.
> >
> > I'd like to see what /proc/interrupts contains too.
> http://paste.ubuntu.com/1360502/
>
> Pastebined just after running fwts and observing the failure. Anything
> telling?
> >

Well, IRQ 16 is for eth0, so it's strange we're seeing that message, but 
if it really is disabling IRQ 16 then ethernet will stop working, so 
that needs looking at for sure.

Colin