From mboxrd@z Thu Jan  1 00:00:00 1970
From: Anthony Liguori <anthony@codemonkey.ws>
Date: Wed, 29 Oct 2008 13:01:01 +0000
Subject: Re: [patch] v4 - fold struct vcpu_info into CPUState
Message-Id: <49085E8D.4050305@codemonkey.ws>
List-Id: <kvm-ia64.vger.kernel.org>
References: <49073D0C.6030706@sgi.com>
In-Reply-To: <49073D0C.6030706@sgi.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: kvm-ia64@vger.kernel.org

Jes Sorensen wrote:
> Hi,
>
> Here's an updated version of the patch. It should fix the problems
> Hollis ran into, and also compile on x86_64 again :-)
>
> I managed to get rid of all the runtime use of qemu_kvm_cpu_env(),
> except for the hotplug code. But I think it's reasonable to do the
> walk of the linked list in that case. However, the more I have looked
> at this, the more obvious to me it becomes that it is right<tm> to
> expose struct CPUState to libkvm, and avoid passing around int vcpu.
>
> Comments and test reports very welcome!

FWIW, vcpu_info seems to have nothing to do with KVM.  It's entirely 
build around the IO thread.  The IO thread really isn't KVM specific.

I would just stick these fields in CPU_COMMON unconditionally if you're 
going to move them at all.

Regards,

Anthony Liguori

> Cheers,
> Jes
>