Re: [PATCH 0/8] KVM updates for 2.6.20-rc2

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

Ingo Molnar wrote:
> * Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote:
>
>   
>>> NOTE: this is not a worry for upstream kernel, it is caused by 
>>> PREEMPT_RT scheduling in previously atomic APIs like kunmap(). But 
>>> KVM used to work pretty nicely in -rt and this problem got introduced 
>>> fairly recently, related to some big-page changes IIRC.
>>>       
>> This is not a recent change.  The x86 emulator is fetching an 
>> instruction or a memory operand.
>>     
>
> well, the fact that it's somehow ending up in highmem and thus the kmap 
> becomes nontrivial could be a recent change maybe?
>
>   

Guest memory was always highmem.  If a particular page is highmem is of 
course runtime dependent.

>>> Any suggestions of how to best fix this in -rt? Basically, it would 
>>> be nice to decouple KVM from get_cpu/preempt_disable type of 
>>> interfaces as much as possible, to make it fully preemptible.
>>>       
>> kvm on intel uses hidden cpu registers, so it can't be migrated as a 
>> normal process. [...]
>>     
>
> oh, granted. My suggestion would be to make it more atomic - not to 
> break the physical attachment of a KVM context to a physical CPU (which 
> is unbreakable, obviously).
>
>   
>> [...] We'd need a hook on context switch to check if we migrated to 
>> another cpu, issue an ipi to unload the registers on the previous cpu, 
>> and reload the registers on the current cpu.  The hook would need to 
>> be protected from context switches.  See vmx.c:vmx_vcpu_load() for the 
>> gory details.  The actual launch would require a small non-preemptible 
>> section as well, on both intel and AMD.
>>     
>
> actually, another, possibly better approach would be to attach a KVM 
> context to tasks, automatically loaded/unloaded upon switch_to(). No 
> need to do IPIs. Would this be fast enough?
>   

You don't want or need to unload on switching out, only on migration.  
Since migrations are hopefully few and far between, it's best to let 
them handle the heavy lifting and keep non-migrating context switches cheap.

If we touch the scheduler, maybe we can let the migration threads call a 
kvm callback to unload the vmx context.

-- 
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