[64studio-devel] IRQ-23 process

[Frieder Bürzele]

Steve A wrote:
> hello Lee
>
> Thanks for the pointer to help trace the problem!
>
>
> tux at 64studio:~$ cat /proc/interrupts
>           CPU0
>  0:        201   IO-APIC-edge      timer
>  1:          2   IO-APIC-edge      i8042
>  7:          0   IO-APIC-edge      parport0
>  8:          0   IO-APIC-edge      rtc
>  9:          0   IO-APIC-fasteoi   acpi
> 12:       9134   IO-APIC-edge      i8042
> 15:       3412   IO-APIC-edge      ide1
> 18:          0   IO-APIC-fasteoi   ICE1712
> 19:          3   IO-APIC-fasteoi   ohci1394
> 20:       7056   IO-APIC-fasteoi   libata
> 21:          0   IO-APIC-fasteoi   libata
> 22:          3   IO-APIC-fasteoi   ehci_hcd:usb2, NVidia CK804
> 23:    2638656   IO-APIC-fasteoi   ohci_hcd:usb1, eth1
> NMI:          0
> LOC:     395083
> ERR:          0
>
> I've rebooted a few times, and it appears that the IRQ associated with 
> eth1 (either 23 or 21) is always 3% CPU busy.  Sometimes the machine 
> boots with eth1 on its own interrupt, which is why I'm so sure.  Also 
> I'm getting no network traffic showing on the router.
>
> This does not happen with the earlier Kernel 2.16.19.1
>
> I'm out of my depth now, I'll create a ticket because I can create 
> this every time, and it is new for this kernel.
>

We had the same problem with ingo already released a fix for testing:
quoted from our ml:
<snip>

>
> Thanks again. I reported it with a copy to Ingo and got a patch back
> that fixes the issue. I don't know if this will result in a change to
> rt-sources down the road. (I'm not that smart.) Anyway, it's in the
> forcedeth.c driver should the question come up. Here's the patch I
> applied:
>
> http://people.redhat.com/mingo/realtime-preempt/testing/forcedeth-rt-tweak.patch 
>
>
> Cheers,
> Mark
>

greetz Frieder (gentoo proaudio)