lkml.org 
[lkml]   [2020]   [Oct]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH] x86/msr: do not warn on writes to OC_MAILBOX
From
Date
On Tue, 2020-10-20 at 12:40 -0700, Dave Hansen wrote:
> On 10/20/20 11:40 AM, Srinivas Pandruvada wrote:
> > On Tue, 2020-10-20 at 19:47 +0200, Borislav Petkov wrote:
> > > On Tue, Oct 20, 2020 at 10:21:48AM -0700, Srinivas Pandruvada
> > > wrote:
> > > > These command id are model specific. There is no guarantee that
> > > > even
> > > > meaning changes. So I don't think we should write any code in
> > > > kernel
> > > > which can't stick.
> > > Ok, is there a common *set* of values present on all models
> > Sorry, don't know.
>
> So, the question is: Is Intel willing to document this on a
> sufficient
> number of models that folks can make a sane driver out of it?
>
> Srinivas, that seems like a pretty sane thing for the community to
> ask.
> We've got random folks poking at MSRs and we don't know whether
> they're
> nuts or not and whether we should spew warnings of disdain. Seems
> like
> it would be in Intel's best interests to understand what users are
> doing
> with this MSR and to try to make sure they're not doing stuff which
> is
> too nutty, or at least give them the chance of avoiding warnings if
> they're being nice.
We are all for it. We have added several sysfs interfaces and adding
more.

>
> Sounds like Borislav is willing to help give Intel's customers a
> nicer
> interface. Mostly we from Intel would have to go dig out the docs
> for
> as many models as we can, and make sure we're allowed to talk about
> it
> publicly.
>
That is the problem. There is no public document.

> I dunno. Maybe we should try it for *one* model and see how it goes.
> Maybe start with the one we're already poking from inside the kernel.

\
 
 \ /
  Last update: 2020-10-21 15:11    [W:0.114 / U:4.480 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site