DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 3B8BAC01BD57
From: Bandan Das <bsd@redhat.com>
To: David Hildenbrand <david@redhat.com>
Cc: kvm@vger.kernel.org, pbonzini@redhat.com, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v4 3/3] KVM: nVMX: Emulate EPTP switching for the L1 hypervisor
References: <20170710204936.4001-1-bsd@redhat.com>
        <20170710204936.4001-4-bsd@redhat.com>
        <2d50ebc4-9328-ce08-b55b-6a331ee13cc3@redhat.com>
        <jpgy3ruam9i.fsf@linux.bootlegged.copy>
Date: Tue, 11 Jul 2017 14:24:08 -0400
In-Reply-To: <jpgy3ruam9i.fsf@linux.bootlegged.copy> (Bandan Das's message of
        "Tue, 11 Jul 2017 13:58:33 -0400")
Message-ID: <jpg1spmsugn.fsf@linux.bootlegged.copy>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.2 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1315
Lines: 42

Bandan Das <bsd@redhat.com> writes:
....
>>> +	/*
>>> +	 * If the (L2) guest does a vmfunc to the currently
>>> +	 * active ept pointer, we don't have to do anything else
>>> +	 */
>>> +	if (vmcs12->ept_pointer != address) {
>>> +		if (address >> cpuid_maxphyaddr(vcpu) ||
>>> +		    !IS_ALIGNED(address, 4096))
>>
>> Couldn't the pfn still be invalid and make kvm_mmu_reload() fail?
>> (triggering a KVM_REQ_TRIPLE_FAULT)
>
> If there's a triple fault, I think it's a good idea to inject it
> back. Basically, there's no need to take care of damage control
> that L1 is intentionally doing.
>
>>> +			goto fail;
>>> +		kvm_mmu_unload(vcpu);
>>> +		vmcs12->ept_pointer = address;
>>> +		kvm_mmu_reload(vcpu);
>>
>> I was thinking about something like this:
>>
>> kvm_mmu_unload(vcpu);
>> old = vmcs12->ept_pointer;
>> vmcs12->ept_pointer = address;
>> if (kvm_mmu_reload(vcpu)) {
>> 	/* pointer invalid, restore previous state */
>> 	kvm_clear_request(KVM_REQ_TRIPLE_FAULT, vcpu);
>> 	vmcs12->ept_pointer = old;
>> 	kvm_mmu_reload(vcpu);
>> 	goto fail;
>> }
>>
>> The you can inherit the checks from mmu_check_root().

Actually, thinking about this a bit more, I agree with you. Any fault
with a vmfunc operation should end with a vmfunc vmexit, so this
is a good thing to have. Thank you for this idea! :)

Bandan