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;
Message-ID: <05425bd8-46dc-f136-fbe0-52f42351aea1@redhat.com>
Date:   Tue, 28 Sep 2021 18:14:49 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.1.0
Subject: Re: [RFC PATCH 1/2] nSVM: introduce struct vmcb_ctrl_area_cached
Content-Language: en-US
To:     Emanuele Giuseppe Esposito <eesposit@redhat.com>,
        kvm@vger.kernel.org
Cc:     Maxim Levitsky <mlevitsk@redhat.com>,
        Sean Christopherson <seanjc@google.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
        linux-kernel@vger.kernel.org
References: <20210917124956.2042052-1-eesposit@redhat.com>
 <20210917124956.2042052-2-eesposit@redhat.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
In-Reply-To: <20210917124956.2042052-2-eesposit@redhat.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 17/09/21 14:49, Emanuele Giuseppe Esposito wrote:
> +static inline bool vmcb_is_intercept_cached(struct vmcb_ctrl_area_cached *control, u32 bit)
> +{
> +	return vmcb_is_intercept((struct vmcb_control_area *) control,
> +				 bit);
> +}
> +

This is quite dangerous, because it expects that the offset is the same 
between vmcb_control_area and vmcb_ctrl_area_cached.  You can just 
duplicate the implementation (which is essentially just a test_bit), and 
call the function

static inline bool vmcb12_is_intercept(struct kvm_vcpu *vcpu, u32 bit)

Likewise, nested_vmcb_check_controls can just take the vcpu since you 
moved nested_load_control_from_vmcb12 earlier.

Finally, copy_vmcb_control_area can be inlined, and its caller 
nested_load_control_from_vmcb12 can stop copying the ASID.  There is 
only one call to it since commit 4995a3685f1b ("KVM: SVM: Use a separate 
vmcb for the nested L2 guest", 2021-03-15).

Thanks,

Paolo