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;
Subject: Re: [PATCH v2 09/46] KVM: SVM: Drop a redundant init_vmcb() from
 svm_create_vcpu()
To:     Sean Christopherson <seanjc@google.com>
Cc:     Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org, Reiji Watanabe <reijiw@google.com>
References: <20210713163324.627647-1-seanjc@google.com>
 <20210713163324.627647-10-seanjc@google.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <77b78927-77ca-39b8-8882-458fc3ec9ba8@redhat.com>
Date:   Mon, 26 Jul 2021 22:33:22 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.11.0
MIME-Version: 1.0
In-Reply-To: <20210713163324.627647-10-seanjc@google.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 13/07/21 18:32, Sean Christopherson wrote:
> Drop an extra init_vmcb() from svm_create_vcpu(), svm_vcpu_reset() is
> guaranteed to call init_vmcb() and there are no consumers of the VMCB
> data between ->vcpu_create() and ->vcpu_reset().  Keep the call to
> svm_switch_vmcb() as sev_es_create_vcpu() touches the current VMCB, but
> hoist it up a few lines to associate the switch with the allocation of
> vmcb01.
> 
> Reviewed-by: Reiji Watanabe <reijiw@google.com>
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---
>   arch/x86/kvm/svm/svm.c | 4 +---
>   1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
> index 44248548be7d..cef9520fe77f 100644
> --- a/arch/x86/kvm/svm/svm.c
> +++ b/arch/x86/kvm/svm/svm.c
> @@ -1431,15 +1431,13 @@ static int svm_create_vcpu(struct kvm_vcpu *vcpu)
>   
>   	svm->vmcb01.ptr = page_address(vmcb01_page);
>   	svm->vmcb01.pa = __sme_set(page_to_pfn(vmcb01_page) << PAGE_SHIFT);
> +	svm_switch_vmcb(svm, &svm->vmcb01);
>   
>   	if (vmsa_page)
>   		svm->vmsa = page_address(vmsa_page);
>   
>   	svm->guest_state_loaded = false;
>   
> -	svm_switch_vmcb(svm, &svm->vmcb01);
> -	init_vmcb(vcpu);
> -
>   	svm_init_osvw(vcpu);
>   	vcpu->arch.microcode_version = 0x01000065;

While this patch makes sense, I'd rather not include it to reduce the 
part of the code in which svm->vmcb is NULL.   See for example the issue 
reported by the static analyzer in which 
svm_hv_vmcb_dirty_nested_enlightenments (called by svm_vcpu_init_msrpm) 
dereferences svm->vmcb.

Paolo