Re: [PATCH] KVM: fix calculation of initial value of rdx register

[Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org>]

Hesse, Christian wrote:
> On Thursday 16 November 2006 14:03, Hesse, Christian wrote:
>   
>> It did crash again. The log is attached.
>>     
>
> I don't know if I already mentioned it, but I think it could be interesting: 
> When I reboot the notebook after a crash it starts up normally until X server 
> ist started, then it freezes again. Looks like the second core still has any 
> problem with its interrupts...
>
>   

Very strange.


Something to try:

   boot into runlevel 3 (no X)
  
   # bind all processes to cpu 0
   for i in /proc/[1-9]*/task/*; do taskset -p 1 ${i##*/}; done
   # bind all interrupts to cpu 0
   pkill -9 irqbalance
   for i in  /proc/irq/*/smp_affinity; do echo 1 > $i; done   
   # (may give some errors)

   # run kvm with vnc (don't require X)
   ./qemu ... -vnc 1
> root@revo:~# cat /proc/interrupts
>            CPU0       CPU1
>   0:   57707101          2    IO-APIC-edge  timer
>   1:      26901          0    IO-APIC-edge  i8042
>   8:     521165          0    IO-APIC-edge  rtc
>   9:      46288          1   IO-APIC-level  acpi
>  12:     509781          0    IO-APIC-edge  i8042
>  14:     267769          0    IO-APIC-edge  ide0
>  16:     714844          0   IO-APIC-level  ipw3945
>  17:    4899673    2306254   IO-APIC-level  uhci_hcd:usb5, 
> i915@pci:0000:00:02.0, eth0
>  18:     439507          0   IO-APIC-level  uhci_hcd:usb4
>  19:          0          0   IO-APIC-level  uhci_hcd:usb3
>  20:          0          0   IO-APIC-level  yenta
>  21:          2          0   IO-APIC-level  ohci1394
>  22:          2          0   IO-APIC-level  ehci_hcd:usb1, uhci_hcd:usb2
>  23:    1100203          0   IO-APIC-level  sdhci:slot0, HDA Intel
> NMI:          0          0
> LOC:   57706935   57669244
> ERR:          1
> MIS:          0
> root@revo:~# lspci | grep 00:02.0
> 00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS/940GML 
> Express Integrated Graphics Controller (rev 03)
>
> The second power cycle works without problems though. Do you think that is 
> helpful in any way?
>   


-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV