Re: [PATCH v8 5/5] x86/tdx: Add Quote generation support

From: Sathyanarayanan Kuppuswamy  <sathyanarayanan.kuppuswamy@linux.intel.com>
To: Dave Hansen <dave.hansen@intel.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	x86@kernel.org
Cc: "H . Peter Anvin" <hpa@zytor.com>,
	"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
	Tony Luck <tony.luck@intel.com>, Andi Kleen <ak@linux.intel.com>,
	Kai Huang <kai.huang@intel.com>,
	Wander Lairson Costa <wander@redhat.com>,
	Isaku Yamahata <isaku.yamahata@gmail.com>,
	marcelo.cerri@canonical.com, tim.gardner@canonical.com,
	khalid.elmously@canonical.com, philip.cox@canonical.com,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v8 5/5] x86/tdx: Add Quote generation support
Date: Thu, 21 Jul 2022 10:16:37 -0700	[thread overview]
Message-ID: <b8af932e-13da-ddcd-4ecb-bd8369853242@linux.intel.com> (raw)
In-Reply-To: <a65473db-b307-c076-6d2f-8f2084d81eb5@intel.com>

Hi Dave,

On 7/21/22 10:02 AM, Dave Hansen wrote:
> On 7/21/22 09:54, Sathyanarayanan Kuppuswamy wrote:
>>
>> On 7/21/22 9:49 AM, Dave Hansen wrote:
>>> On 7/21/22 09:42, Sathyanarayanan Kuppuswamy wrote:
>>>> On 7/21/22 9:08 AM, Dave Hansen wrote:
>>>>> On 6/8/22 19:52, Kuppuswamy Sathyanarayanan wrote:
>>>>>> For shared buffer allocation, alternatives like using the DMA API is
>>>>>> also considered. Although it simpler to use, it is not preferred because
>>>>>> dma_alloc_*() APIs require a valid bus device as argument, which would
>>>>>> need converting the attestation driver into a platform device driver.
>>>>>> This is unnecessary, and since the attestation driver does not do real
>>>>>> DMA, there is no need to use real DMA APIs.
>>>>> Let's actually try to walk through the requirements for the memory
>>>>> allocation here.
>>>>>
>>>>>  1. The guest kernel needs to allocate some guest physical memory
>>>>>     for the attestation data buffer
>>>> Physically contiguous memory.
>>> Remind me how large the quote structures are.
>> It depends on the attestation service. In addition to TDREPORT (1K size),
>> during quote generation, additional data can be included in the signed
>> quote. So the spec allows variable length. User agent will communicate with
>> attestation service to identify the appropriate buffer length. Our test
>> uses 8K buffers
> 
> What is this "additional data"?  Is that "REPORTDATA" from the TDX
> module spec?
> 
> 	"Additional REPORTDATA, a 64-byte value, is provided by the
> 	guest TD to be included in the TDG.MR.REPORT."

It is not the REPORTDATA. REPORTDATA is already included in the TDREPORT during
REPORT generation process (using TDG.MR.REPORT).

Spec does not clearly define the details about the additional data. I think it is
related to certificates.

Isaku, do you have more details about it?

Current ABI allows attestation service and agent to decide the quote size. So
we can't make assumptions on what that size will be.

Following is copied from TDX Module specification, sec titled "Measurement
and Attestation"

An Intel SGX Quoting Enclave, written specifically to support quoting Intel TDX TDs,
uses a new ENCLU instruction leaf, EVERIFYREPORT2, to help check the integrity of the
TDG.MR.REPORT. If it passes, the Quoting Enclave can use a certified quote signing key
to sign a quote containing the guest TD’s measurements and the additional data being
quoted.

> 
> 

-- 
Sathyanarayanan Kuppuswamy
Linux Kernel Developer