[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] Xen-unstable: Regression (on AMD) host boot stuck on (XEN) testing the IO APIC.......................

Tuesday, August 26, 2014, 9:34:30 AM, you wrote:

>>>> On 25.08.14 at 20:25, <linux@xxxxxxxxxxxxxx> wrote:
>> Monday, August 25, 2014, 7:54:05 PM, you wrote:
>>> Hmm - Does reverting e13b3203990706db1313ec2aadd9a30b249ee793 fix the issue?
>> Hmm i completely overlooked that commit (it's quite evident when you look at 
>> the 
>> commitdiff :-) )
>> Just tested reverting that single commit and that indeed lets the machine 
>> boot again.

> Considering how early this is, the only theory I have is that
> rcu_pending() may end up constantly returning true, thus always
> causing RCU_SOFTIRQ to get raised, not allowing the loop in
> __do_softirq() to ever exit. If that's the case, we seem to have
> two options:
> - pass into __print_IO_APIC() whether it's being called from a key
>   handler
> - only call process_pending_softirqs() when system_state >=
>   SYS_STATE_smp_boot

> Personally I'd prefer the former option.

> But Sander, a fundamental question: Why would you _always_
> boot your system with "apic=debug" and those various other
> debugging options? They are, as you may guess, for debugging,
> not for day to day use. (That said I'm glad you have this in place
> for this particular case, as it allowed spotting a regression early.)

> Jan

Hmm just a left over from previous endeavours :-)

Don't know if they also incur a (large) performance penalty ?
(they primarily seemed to add some debug info to the log)

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.