Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Sender: Paolo Bonzini <paolo.bonzini@gmail.com>
Message-ID: <4505b43d-5c33-4199-1259-6d4e8ebac1ec@redhat.com>
Date:   Wed, 20 Apr 2022 17:00:01 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.8.0
Content-Language: en-US
To:     Sean Christopherson <seanjc@google.com>,
        "Maciej S. Szmigiero" <mail@maciej.szmigiero.name>
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
References: <20220402010903.727604-1-seanjc@google.com>
 <20220402010903.727604-2-seanjc@google.com>
 <112c2108-7548-f5bd-493d-19b944701f1b@maciej.szmigiero.name>
 <YkspIjFMwpMYWV05@google.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH 1/8] KVM: nSVM: Sync next_rip field from vmcb12 to vmcb02
In-Reply-To: <YkspIjFMwpMYWV05@google.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 4/4/22 19:21, Sean Christopherson wrote:
> On Mon, Apr 04, 2022, Maciej S. Szmigiero wrote:
>>> @@ -1606,7 +1622,7 @@ static int svm_set_nested_state(struct kvm_vcpu *vcpu,
>>>    	nested_copy_vmcb_control_to_cache(svm, ctl);
>>>    	svm_switch_vmcb(svm, &svm->nested.vmcb02);
>>> -	nested_vmcb02_prepare_control(svm);
>>> +	nested_vmcb02_prepare_control(svm, save->rip);
>>
>> 					   ^
>> I guess this should be "svm->vmcb->save.rip", since
>> KVM_{GET,SET}_NESTED_STATE "save" field contains vmcb01 data,
>> not vmcb{0,1}2 (in contrast to the "control" field).
> 
> Argh, yes.  Is userspace required to set L2 guest state prior to KVM_SET_NESTED_STATE?
> If not, this will result in garbage being loaded into vmcb02.
> 

Let's just require X86_FEATURE_NRIPS, either in general or just to
enable nested virtualiazation, i.e.:

diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index fc1725b7d05f..f8fc8a1b09f1 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -4904,10 +4904,12 @@ static __init int svm_hardware_setup(void)
  			goto err;
  	}
  
-	if (nrips) {
-		if (!boot_cpu_has(X86_FEATURE_NRIPS))
-			nrips = false;
-	}
+	if (!boot_cpu_has(X86_FEATURE_NRIPS))
+		nrips = false;
+	if (nested & !nrips) {
+		pr_warn("Next RIP Save not available, disabling nested virtualization\n");
+		nested = false;
+	}
  
  	enable_apicv = avic = avic && npt_enabled && (boot_cpu_has(X86_FEATURE_AVIC) || force_avic);
  

If I looked it up correctly it was introduced around 2010-2011.

Paolo