Re: [PATCH v2 1/2] x86/bugs: Don't fill RSB on VMEXIT with eIBRS+retpoline

["Shah, Amit" <Amit.Shah@amd.com>]

On Thu, 2024-12-05 at 15:12 -0800, Josh Poimboeuf wrote:
> On Tue, Dec 03, 2024 at 12:20:15PM +0100, Borislav Petkov wrote:
> > On Mon, Dec 02, 2024 at 03:35:21PM -0800, Pawan Gupta wrote:
> > > It is in this doc:
> > > 
> > >  
> > > https://www.intel.com/content/www/us/en/developer/articles/technical/software-security-guidance/technical-documentation/indirect-branch-restricted-speculation.html
> > > 
> > 
> > I hope those URLs remain more stable than past experience shows.
> > 
> > >   "Processors with enhanced IBRS still support the usage model
> > > where IBRS is
> > >   set only in the OS/VMM for OSes that enable SMEP. To do this,
> > > such
> > >   processors will ensure that guest behavior cannot control the
> > > RSB after a
> > >   VM exit once IBRS is set, even if IBRS was not set at the time
> > > of the VM
> > >   exit."
> > 
> > ACK, thanks.
> > 
> > Now, can we pls add those excerpts to Documentation/ and point to
> > them from
> > the code so that it is crystal clear why it is ok?
> 
> Ok, I'll try to write up a document.  I'm thinking it should go in
> its
> own return-based-attacks.rst file rather than spectre.rst, which is
> more
> of an outdated historical document at this point.  And we want this
> document to actually be read (and kept up to date) by developers
> instead
> of mostly ignored like the others.
> 

Hey Josh,

Do you plan to submit a v3 with the changes?

Thanks,

		Amit