Re: [PATCH RESEND 3/3] KVM: Add flush_on_enter before guest enter

[Paolo Bonzini <pbonzini@redhat.com>]

On 09/11/2017 03:02, Wanpeng Li wrote:
> From: Wanpeng Li <wanpeng.li@hotmail.com>
> 
> PV-Flush guest would indicate to flush on enter, flush the TLB before
> entering and exiting the guest.
> 
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Radim Krčmář <rkrcmar@redhat.com>
> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
> ---
>  arch/x86/kvm/x86.c | 12 ++++++++++--
>  1 file changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 1ea28a2..f295360 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -2116,7 +2116,13 @@ static void record_steal_time(struct kvm_vcpu *vcpu)
>  		&vcpu->arch.st.steal, sizeof(struct kvm_steal_time))))
>  		return;
>  
> -	vcpu->arch.st.steal.preempted = KVM_VCPU_NOT_PREEMPTED;
> +	if (xchg(&vcpu->arch.st.steal.preempted, KVM_VCPU_NOT_PREEMPTED) ==
> +			(KVM_VCPU_SHOULD_FLUSH | KVM_VCPU_PREEMPTED))
> +		/*
> +		 * Do TLB_FLUSH before entering the guest, its passed
> +		 * the stage of request checking
> +		 */
> +		kvm_x86_ops->tlb_flush(vcpu);
>  
>  	if (vcpu->arch.st.steal.version & 1)
>  		vcpu->arch.st.steal.version += 1;  /* first time write, random junk */
> @@ -2887,7 +2893,9 @@ static void kvm_steal_time_set_preempted(struct kvm_vcpu *vcpu)
>  	if (!(vcpu->arch.st.msr_val & KVM_MSR_ENABLED))
>  		return;
>  
> -	vcpu->arch.st.steal.preempted = KVM_VCPU_PREEMPTED;
> +	if (xchg(&vcpu->arch.st.steal.preempted, KVM_VCPU_PREEMPTED) ==
> +				KVM_VCPU_SHOULD_FLUSH)
> +		kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu);

This is not necessary.  Instead, you can just OR the KVM_VCPU_PREEMPTED
bit; record_steal_time will pick up the request and do the TLB flush later.

Also, I think this is a case where you should prefer INVVPID to INVEP.
That's because "execution of the INVEPT instruction invalidates
guest-physical mappings and combined mappings" while "execution of the
INVVPID instruction invalidates linear mappings and combined mappings".
In this case, invalidating guest-physical mapping is unnecessary.

So you could add a new bool argument to kvm_x86_ops->tlb_flush.  In
vmx.c, __vmx_flush_tlb can do invept if "enable_ept && (invalidate_gpa
|| !enable_vpid)".

If !enable_ept && !enable_vpid, the feature cannot be made available.

Thanks,

Paolo

>  	kvm_write_guest_offset_cached(vcpu->kvm, &vcpu->arch.st.stime,
>  			&vcpu->arch.st.steal.preempted,
>