RE: [Xen-ia64-devel] [PATCH] Fix domain reboot bug

["Tian, Kevin" <kevin.tian@xxxxxxxxx>]

>From: Tristan Gingold [mailto:Tristan.Gingold@xxxxxxxx]
>Sent: 2006年3月30日 19:16
>Le Jeudi 30 Mars 2006 12:36, Tian, Kevin a écrit :
>> From: Tristan Gingold
>[...]
>> >If I understand the patch correctly, you flush all the vhpt when a
>domain
>> >is
>> >destroyed.  Isn't it a little bit too heavy ?
>> >
>> >Tristan.
>>
>> Yes, that's too heavy. We hope the flush can be made more accurately
>> after guest SMP support is fully ready. At that time, people can depend
>on
>> domain-> domain_dirty_cpumask to decide which LPs should receive
>IPI.
>> Fore example, currently domain_dirty_cpumask is even not updated at
>> domain switch. So a most conservative way is chosen here to flush all
>> LPs just like flush_tlb_all at domain destroy, and that can also ensure
>> guest SMP working at early phase before everything is shot down. :-)
>IMHO this is not related to SMP-guest but to VCPU-migration.
>

VCPU migration is one case where vcpu-> vcpu_dirty_cpumask is the hint 
about LPs where specific vcpu has footprint. However even without 
migration, one domain with SMP guest support may have different VCPUs 
running on different LPs even by a static relationship. In that case, 
domain->domain_dirty_cpumask is another hint for LPs where domain 
has footprint. All those information are critical for operations like vtlb 
flush 
which only needs send to related LPs within domain wides.

Thanks,
Kevin

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel