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 25/30] KVM: nSVM: leave guest mode when clearing EFER.SVME
To:     Krish Sadhukhan <krish.sadhukhan@oracle.com>,
        linux-kernel@vger.kernel.org, kvm@vger.kernel.org
References: <20200529153934.11694-1-pbonzini@redhat.com>
 <20200529153934.11694-26-pbonzini@redhat.com>
 <da854e9e-b305-b938-68f6-995bcc80ffd1@oracle.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <9319ccf9-c318-40e8-7d56-df0e32617430@redhat.com>
Date:   Mon, 1 Jun 2020 09:28:10 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.6.0
MIME-Version: 1.0
In-Reply-To: <da854e9e-b305-b938-68f6-995bcc80ffd1@oracle.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 01/06/20 04:26, Krish Sadhukhan wrote:
> On 5/29/20 8:39 AM, Paolo Bonzini wrote:
>> According to the AMD manual, the effect of turning off EFER.SVME while a
>> guest is running is undefined.  We make it leave guest mode immediately,
>> similar to the effect of clearing the VMX bit in MSR_IA32_FEAT_CTL.
> 
> I see that svm_set_efer() is called in enter_svm_guest_mode() and
> nested_svm_vmexit(). In the VMRUN path, we have already checked
> EFER.SVME in nested_vmcb_checks(). So if it was not set, we wouldn't
> come to enter_svm_guest_mode(). Your fix is only for the #VMEXIT path
> then ?

No, it's for KVM_SET_MSR.

Paolo