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;
Message-ID: <dff229a39e30f84dcf8cc8caffc41f83bd5ece73.camel@redhat.com>
Subject: Re: [PATCH v2 12/12] kvm/x86: Remove APICV activate mode
 inconsistency check
From:   Maxim Levitsky <mlevitsk@redhat.com>
To:     Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
        linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc:     pbonzini@redhat.com, seanjc@google.com, joro@8bytes.org,
        jon.grimm@amd.com, wei.huang2@amd.com, terry.bowman@amd.com
Date:   Mon, 18 Apr 2022 15:55:34 +0300
In-Reply-To: <20220412115822.14351-13-suravee.suthikulpanit@amd.com>
References: <20220412115822.14351-1-suravee.suthikulpanit@amd.com>
         <20220412115822.14351-13-suravee.suthikulpanit@amd.com>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.36.5 (3.36.5-2.fc32) 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Tue, 2022-04-12 at 06:58 -0500, Suravee Suthikulpanit wrote:
> When launching a VM with x2APIC and specify more than 255 vCPUs,
> the guest kernel can disable x2APIC (e.g. specify nox2apic kernel option).
> The VM fallbacks to xAPIC mode, and disable the vCPU ID 255.
> 
> In this case, APICV should be disabled for the vCPU ID 255.
> Therefore, the APICV mode consisency check is no longer valid.
> 
> Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
> ---
>  arch/x86/kvm/x86.c | 13 +++++--------
>  1 file changed, 5 insertions(+), 8 deletions(-)
> 
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 0c0ca599a353..d0fac57e9996 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -9765,6 +9765,11 @@ void kvm_vcpu_update_apicv(struct kvm_vcpu *vcpu)
>  	down_read(&vcpu->kvm->arch.apicv_update_lock);
>  
>  	activate = kvm_apicv_activated(vcpu->kvm);
> +
> +	/* Do not activate AVIC when APIC is disabled */
> +	if (kvm_get_apic_mode(vcpu) == LAPIC_MODE_DISABLED)
> +		activate = false;
> +
>  	if (vcpu->arch.apicv_active == activate)
>  		goto out;
>  
> @@ -10159,14 +10164,6 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
>  	guest_timing_enter_irqoff();
>  
>  	for (;;) {
> -		/*
> -		 * Assert that vCPU vs. VM APICv state is consistent.  An APICv
> -		 * update must kick and wait for all vCPUs before toggling the
> -		 * per-VM state, and responsing vCPUs must wait for the update
> -		 * to complete before servicing KVM_REQ_APICV_UPDATE.
> -		 */
> -		WARN_ON_ONCE(kvm_apicv_activated(vcpu->kvm) != kvm_vcpu_apicv_active(vcpu));
> -
>  		exit_fastpath = static_call(kvm_x86_vcpu_run)(vcpu);
>  		if (likely(exit_fastpath != EXIT_FASTPATH_REENTER_GUEST))
>  			break;

That warning catches bugs, please don't remove it.

It can be made conditional on this vCPU having APIC enabled instead.

Best regards,
	Maxim Levitsky