Re: [PATCH 1/6] kvm/x86/svm: force new asid on vcpu migration

From: Marcelo Tosatti <mtosatti@redhat.com>
To: Joerg Roedel <joerg.roedel@amd.com>
Cc: Avi Kivity <avi@redhat.com>,
	kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/6] kvm/x86/svm: force new asid on vcpu migration
Date: Thu, 5 Mar 2009 14:53:33 -0300	[thread overview]
Message-ID: <20090305175333.GA2169@amt.cnet> (raw)
In-Reply-To: <1236255153-4432-2-git-send-email-joerg.roedel@amd.com>

On Thu, Mar 05, 2009 at 01:12:28PM +0100, Joerg Roedel wrote:
> Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
> ---
>  arch/x86/kvm/svm.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> index 1821c20..0e66bca 100644
> --- a/arch/x86/kvm/svm.c
> +++ b/arch/x86/kvm/svm.c
> @@ -180,7 +180,7 @@ static inline void kvm_write_cr2(unsigned long val)
>  
>  static inline void force_new_asid(struct kvm_vcpu *vcpu)
>  {
> -	to_svm(vcpu)->asid_generation--;
> +	to_svm(vcpu)->asid_generation = 0;
>  }

Won't the per-cpu asig_generation overflow at some point? 

And can this comparison in pre_svm_run ever be true:

        if (svm->vcpu.cpu != cpu

?

>  static inline void flush_guest_tlb(struct kvm_vcpu *vcpu)
> @@ -716,6 +716,7 @@ static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
>  		svm->vmcb->control.tsc_offset += delta;
>  		vcpu->cpu = cpu;
>  		kvm_migrate_timers(vcpu);
> +		force_new_asid(vcpu);
>  	}
>  
>  	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)