[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87bkewcufi.ffs@tglx>
Date:   Thu, 24 Aug 2023 17:49:05 +0200
From:   Thomas Gleixner <tglx@...utronix.de>
To:     "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
        Dave Hansen <dave.hansen@...el.com>
Cc:     Borislav Petkov <bp@...en8.de>, Andy Lutomirski <luto@...nel.org>,
        Kuppuswamy Sathyanarayanan 
        <sathyanarayanan.kuppuswamy@...ux.intel.com>,
        Elena Reshetova <elena.reshetova@...el.com>,
        Jun Nakajima <jun.nakajima@...el.com>, x86@...nel.org,
        linux-coco@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] x86/tdx: Mark TSC reliable
On Tue, Aug 08 2023 at 23:01, Kirill A. Shutemov wrote:
> On Tue, Aug 08, 2023 at 10:13:05AM -0700, Dave Hansen wrote:
>> I take it this is carved in stone in the TDX specs somewhere.  A
>> reference would be nice.
>
> TDX Module 1.0 spec:
>
> 	5.3.5. Time Stamp Counter (TSC)
>
> 	TDX provides a trusted virtual TSC to the guest TDs. TSC value is
> 	monotonously incrementing, starting from 0 on TD initialization by the
> 	host VMM. The deviation between virtual TSC values read by each VCPU is
> 	small.
Nice weasel wording. What's the definition of "small"?
Any OS needs a guarantee that vCPUs cannot observe time going backwards,
which is obviously possible when the deviation is not small enough.
> Wording in the spec looks okay to me. We can only hope that implementation
> going to be sane.
Hope dies last :)
Powered by blists - more mailing lists