kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* vmexit cycle cost
@ 2012-11-16 16:56 Christoffer Dall
  2012-11-18 14:16 ` Avi Kivity
  0 siblings, 1 reply; 2+ messages in thread
From: Christoffer Dall @ 2012-11-16 16:56 UTC (permalink / raw)
  To: KVM General

Hi all,

just a quick question: using the kvm-unittest vmexit test, it seems
using a VMCALL for exit, a round-trip to the host takes roughly 1000
cycles on a modern processor. This number sounds quite low to me. Can
anyone verify if it's reasonable or have more data on this number has
developed over time given newer processors?

Thanks,
-Christoffer

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: vmexit cycle cost
  2012-11-16 16:56 vmexit cycle cost Christoffer Dall
@ 2012-11-18 14:16 ` Avi Kivity
  0 siblings, 0 replies; 2+ messages in thread
From: Avi Kivity @ 2012-11-18 14:16 UTC (permalink / raw)
  To: Christoffer Dall; +Cc: KVM General

On 11/16/2012 06:56 PM, Christoffer Dall wrote:
> Hi all,
> 
> just a quick question: using the kvm-unittest vmexit test, it seems
> using a VMCALL for exit, a round-trip to the host takes roughly 1000
> cycles on a modern processor. This number sounds quite low to me. Can
> anyone verify if it's reasonable or have more data on this number has
> developed over time given newer processors?

It's reasonable, not low (this number can never be low enough).


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

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2012-11-18 14:16 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-11-16 16:56 vmexit cycle cost Christoffer Dall
2012-11-18 14:16 ` Avi Kivity

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).