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: <59208dd7a83e1369851a2d0c2fe08e3e7639fa59.camel@redhat.com>
Subject: Re: [RFC PATCH 09/13] KVM: SVM: Introduce helper function
 avic_get_apic_id
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:   Thu, 24 Feb 2022 21:46:22 +0200
In-Reply-To: <20220221021922.733373-10-suravee.suthikulpanit@amd.com>
References: <20220221021922.733373-1-suravee.suthikulpanit@amd.com>
         <20220221021922.733373-10-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 Sun, 2022-02-20 at 20:19 -0600, Suravee Suthikulpanit wrote:
> This function returns the currently programmed guest physical
> APIC ID of a vCPU in both xAPIC and x2APIC modes.
> In case of invalid APIC ID based on the current mode,
> the function returns X2APIC_BROADCAST.
> 
> Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
> ---
>  arch/x86/kvm/svm/avic.c | 29 ++++++++++++++++++++++++++---
>  1 file changed, 26 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm/avic.c b/arch/x86/kvm/svm/avic.c
> index 55b3b703b93b..3543b7a4514a 100644
> --- a/arch/x86/kvm/svm/avic.c
> +++ b/arch/x86/kvm/svm/avic.c
> @@ -450,16 +450,35 @@ static void avic_invalidate_logical_id_entry(struct kvm_vcpu *vcpu)
>  		clear_bit(AVIC_LOGICAL_ID_ENTRY_VALID_BIT, (unsigned long *)entry);
>  }
>  
> +static inline u32 avic_get_apic_id(struct kvm_vcpu *vcpu)
> +{
> +	u32 apic_id = kvm_lapic_get_reg(vcpu->arch.apic, APIC_ID);
> +
> +	if (!apic_x2apic_mode(vcpu->arch.apic)) {
> +		/*
> +		 * In case of xAPIC, we do not support
> +		 * APIC ID larger than 254.
> +		 */
> +		if (vcpu->vcpu_id >= APIC_BROADCAST)
> +			return X2APIC_BROADCAST;
This is not a good way to return a error value like that IMHO.

> +		return apic_id >> 24;
> +	} else
> +		return apic_id;
> +}

I don't fully like this to be honest - this should be at least function in lapic.c
and use kvm_xapic_id and kvm_x2apic_id.

And hopefully if I manage to make apic id always read only then we be able
to never use APIC_ID register and always use vcpu->vcpu_id.

Best regards,
	Maxim Levitsky

> +
>  static int avic_handle_ldr_update(struct kvm_vcpu *vcpu)
>  {
>  	int ret = 0;
>  	struct vcpu_svm *svm = to_svm(vcpu);
>  	u32 ldr = kvm_lapic_get_reg(vcpu->arch.apic, APIC_LDR);
> -	u32 id = kvm_xapic_id(vcpu->arch.apic);
> +	u32 id = avic_get_apic_id(vcpu);
>  
>  	if (ldr == svm->ldr_reg)
>  		return 0;
>  
> +	if (id == X2APIC_BROADCAST)
> +		return -EINVAL;
> +
This is what I mean. It is better to check here that we 
vcpu->vcpu_id >= APIC_BROADCAST and fail.


>  	avic_invalidate_logical_id_entry(vcpu);
>  
>  	if (ldr)
> @@ -475,7 +494,10 @@ static int avic_handle_apic_id_update(struct kvm_vcpu *vcpu)
>  {
>  	u64 *old, *new;
>  	struct vcpu_svm *svm = to_svm(vcpu);
> -	u32 id = kvm_xapic_id(vcpu->arch.apic);
> +	u32 id = avic_get_apic_id(vcpu);
> +
> +	if (id == X2APIC_BROADCAST)
> +		return 1;

Same here.
>  
>  	if (vcpu->vcpu_id == id)
>  		return 0;
> @@ -497,7 +519,8 @@ static int avic_handle_apic_id_update(struct kvm_vcpu *vcpu)
>  	 * APIC ID table entry if already setup the LDR.
>  	 */
>  	if (svm->ldr_reg)
> -		avic_handle_ldr_update(vcpu);
> +		if (avic_handle_ldr_update(vcpu))
> +			return 1;
>  
>  	return 0;


Best regards,
	Maxim Levitsky

>  }