Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
To:     Sean Christopherson <sean.j.christopherson@intel.com>,
        Cathy Avery <cavery@redhat.com>
Cc:     linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
        vkuznets@redhat.com, wei.huang2@amd.com, mlevitsk@redhat.com
References: <20201011184818.3609-1-cavery@redhat.com>
 <20201011184818.3609-3-cavery@redhat.com>
 <20201013013349.GB10366@linux.intel.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH v2 2/2] KVM: SVM: Use a separate vmcb for the nested L2
 guest
Message-ID: <f3268301-6cbd-904e-949c-7ccc4a2e5d36@redhat.com>
Date:   Fri, 13 Nov 2020 18:36:02 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.4.0
MIME-Version: 1.0
In-Reply-To: <20201013013349.GB10366@linux.intel.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 13/10/20 03:33, Sean Christopherson wrote:
>> +	svm->vmcb = svm->vmcb01;
>> +	svm->vmcb_pa = svm->vmcb01_pa;
> I very highly recommend adding a helper to switch VMCB.  Odds are very good
> there will be more than just these two lines of boilerplate code for changing
> the active VMCB.

Yes, probably we can make svm->vmcb01 and svm->vmcb02 something like 
VMX's struct loaded_vmcs:

struct kvm_vmcb {
	void *vmcb;
	unsigned long pa;
}

I don't expect a lot more to happen due to SVM having no need for 
caching, so for now I think it's okay.

I have other comments for which I'll reply to the patch itself.

Paolo