linux-kernel - Re: [PATCH v2 1/3] KVM: VMX: Move vmcs related resetting out of vmx_vcpu

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sat, 19 Oct 2019 02:34:31 +0800
From:   Xiaoyao Li <xiaoyao.li@...el.com>
To:     Sean Christopherson <sean.j.christopherson@...el.com>
Cc:     Paolo Bonzini <pbonzini@...hat.com>,
        Radim Krčmář <rkrcmar@...hat.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        Jim Mattson <jmattson@...gle.com>,
        Joerg Roedel <joro@...tes.org>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 1/3] KVM: VMX: Move vmcs related resetting out of
 vmx_vcpu_reset()

On 10/19/2019 12:57 AM, Sean Christopherson wrote:
> On Fri, Oct 18, 2019 at 05:37:21PM +0800, Xiaoyao Li wrote:
>> Move vmcs related codes into a new function vmx_vmcs_reset() from
>> vmx_vcpu_reset(). So that it's more clearer which data is related with
>> vmcs and can be held in vmcs.
>>
>> Suggested-by: Krish Sadhukhan <krish.sadhukhan@...cle.com>
>> Signed-off-by: Xiaoyao Li <xiaoyao.li@...el.com>
>> ---
>>   arch/x86/kvm/vmx/vmx.c | 65 ++++++++++++++++++++++++------------------
>>   1 file changed, 37 insertions(+), 28 deletions(-)
>>
>> diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
>> index e660e28e9ae0..ef567df344bf 100644
>> --- a/arch/x86/kvm/vmx/vmx.c
>> +++ b/arch/x86/kvm/vmx/vmx.c
>> @@ -4271,33 +4271,11 @@ static void vmx_vcpu_setup(struct vcpu_vmx *vmx)
>>   	}
>>   }
>>   
>> -static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
>> +static void vmx_vmcs_reset(struct kvm_vcpu *vcpu, bool init_event)
> 
> I'd strongly prefer to keep the existing code.  For me, "vmcs_reset" means
> zeroing out the VMCS, i.e. reset the VMCS to a virgin state.  "vcpu_reset"
> means exactly that, stuff vCPU state to emulate RESET/INIT.
> 
> And the split is arbitrary and funky, e.g. EFER is integrated into the
> VMCS on all recent CPUs, but here it's handled in vcpu_reset.
>

I left EFER in vcpu_reset() because it doesn't directly lead to a 
vmcs_write in vmx_set_efer().

OK. I'll drop this patch.