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;
Subject: Re: [PATCH v2 2/9] KVM: X86: Introduce KVM_HC_PAGE_ENC_STATUS
 hypercall
To:     Ashish Kalra <ashish.kalra@amd.com>,
        "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Cc:     tglx@linutronix.de, mingo@redhat.com, hpa@zytor.com,
        joro@8bytes.org, bp@suse.de, thomas.lendacky@amd.com,
        x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
        srutherford@google.com, brijesh.singh@amd.com,
        dovmurik@linux.vnet.ibm.com, tobin@ibm.com, jejb@linux.ibm.com,
        frankeh@us.ibm.com
References: <cover.1606782580.git.ashish.kalra@amd.com>
 <40acca4b49cd904ea73038309908151508fb555c.1606782580.git.ashish.kalra@amd.com>
 <20201202165420.GI3226@work-vm>
 <20201202212232.GB14672@ashkalra_ubuntu_server>
From:   Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <5813902a-f272-a6fe-b32c-c470cb18c1ff@redhat.com>
Date:   Sun, 6 Dec 2020 11:25:18 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.4.0
MIME-Version: 1.0
In-Reply-To: <20201202212232.GB14672@ashkalra_ubuntu_server>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 02/12/20 22:22, Ashish Kalra wrote:
> Hello Dave,
> 
> On Wed, Dec 02, 2020 at 04:54:20PM +0000, Dr. David Alan Gilbert wrote:
>> * Ashish Kalra (Ashish.Kalra@amd.com) wrote:
>>> From: Brijesh Singh <brijesh.singh@amd.com>
>>>
>>> This hypercall is used by the SEV guest to notify a change in the page
>>> encryption status to the hypervisor. The hypercall should be invoked
>>> only when the encryption attribute is changed from encrypted -> decrypted
>>> and vice versa. By default all guest pages are considered encrypted.
>>
>> Is it defined whether these are supposed to be called before or after
>> the the page type has been changed; is it change the type and then
>> notify or the other way around?

It doesn't matter.  However, you have do it before writing to the page, 
and the content of the page is unspecified between the hypercall and the 
write to the page.

So you cannot for example encrypt a page in place (using the same PFN 
but different settings of the C bit), you need to do:

	hypercall(); /* mark enc_data as encrypted */
	/*
	 * The contents of enc_data is now undefined as it can change
	 * across migration.
	 */
	memset(enc_data, unenc_data, PAGE_SIZE);

Paolo