Re: BUG: soft lockup detected on CPU#5 / xfsdatad

[Christian Røsnes <christian.rosnes@xxxxxxxxx>]

David Chinner wrote:
> FWIW:
>
> > =====================================================
> > Running: cat /proc/interrupts
> >            CPU0       CPU1       CPU2       CPU3       CPU4       CPU5       CPU6       CPU7       
> >   0:        226          0          0          0          0  277717661          0          0    IO-APIC-edge  timer
> >   1:          0          0          0          0          0          9          0          0    IO-APIC-edge  i8042
> >   9:          0          0          0          0          0          0          0          0   IO-APIC-level  acpi
> >  14:          0          0          0          0          0         13          0          0    IO-APIC-edge  ide0
> > 169:          0          0          0          0          0    5649134          0          0   IO-APIC-level  uhci_hcd:usb1, qla2400
> > 185:          0          0          0          0          0  545423724          0          0   IO-APIC-level  eth0
> > 201:          0          0          0          0          0  215363073          0          0   IO-APIC-level  megaraid
> > 209:          0          0          0          0          0         63          0          0   IO-APIC-level  ide2, ehci_hcd:usb4
> > 217:          0          0          0          0          0          0          0          0   IO-APIC-level  uhci_hcd:usb2
> > 225:          0          0          0          0          0          0          0          0   IO-APIC-level  uhci_hcd:usb3
> > NMI:      62452     249384      63225      62699      62361     249362      63203      62677 
> > LOC:  277687412  277301786  277700824  277701522  277687319  277301693  277700731  277701429 
> > ERR:          7
> > MIS:          0
>
> CPU 5 is doing all your interrupt work for timers, usb, ethernet and
> your RAID. You might want to try to spread these interrupts to
> different CPus to reduce the interrupt load on this one CPU. That
> may improve the situation.
>
> Cheers,
>
> Dave.

Thank you for the information.

Is there a preferred way of distributing interrupts across several Xeon 
cpus ?

Also, does anyone know if Intel Xeons behave differently than AMD 
Opterons in this regard
(interrupt cpu distibution on 64-bit kernels) ?

Reason for my asking is that as opposed to the interrupt distribution 
shown in the table above,
where most interrupts on that 2-cpu dual-core hyperthreaded Xeon is 
handled by cpu 5,
my 2-cpu dual-core Opteron rig (kernel 2.6.15, 64-bit) show a more evenly
distribution of interrupts across all its cpus:

opteron# cat /proc/interrupts
         CPU0       CPU1       CPU2       CPU3
0:   33072714   38007277   38475123   37457443    IO-APIC-edge  timer
1:      29837      43445      34925      36416    IO-APIC-edge  i8042
8:  265978968  264597467  333282019  333292177    IO-APIC-edge  rtc
9:          0          0          0          0   IO-APIC-level  acpi
12:     525534     951163     698370     667647    IO-APIC-edge  i8042
14:     982569    1558280    1241299    1200781    IO-APIC-edge  ide0
177:    4015581    6102627    5084144    5415943   IO-APIC-level  
libata, ehci_hcd:usb2
185:    5141100    8979803    6598927    6391688   IO-APIC-level  
libata, NVidia CK804
193:     134874     261894     280336     234794   IO-APIC-level  megaraid
201:          0          0          0          3   IO-APIC-level  ohci1394
209:          0          0          0          0   IO-APIC-level  
ohci_hcd:usb1
217:  192852160          1          3        810   IO-APIC-level  eth0
225:    6879196   11640568    8902256    8644706   IO-APIC-level  nvidia
233:     573425     874257     849441    1650900   IO-APIC-level  arcmsr
NMI:        113         61         79         62
LOC:  147019199  147019177  147019155  147019113
ERR:          0
MIS:          0

I would think that the Xeon could automatically distribute interrupts across
all cpus similar to the Opteron. Maybe some specific 
kernel-compile-option- or
proc-parameter setting is necessary on the Xeon for this to happen ?

Thanks
Christian