Re: [Xen-devel] No noht?

[Mark Williamson <mark.williamson@xxxxxxxxxxxx>]

> Perhaps that's because the common use case is "try
> noht and see if it makes any difference".  Guess what?
> Since "noht" is silently ignored, it *doesn't* make any
> difference. ;-) There are many xen-devel and xen-users
> postings (since it disappeared) that indicate that noht
> is assumed to work.  Even Mark Williamson has a 2007
> posting that implies that it should work.

Hrmmm.  I often poke into the sourcecode to check regarding options I'm about 
to tell people to use; I probably didn't do that for noht so it may still 
have been around at the time.

I can't remember it going away; I don't imagine it would be very difficult to 
reintroduce, if we figure out where / why it went away in the first place!  
Failing that, it should get de-documented from anywhere it's currently 
mentioned...

> Also, I'll bet the fraction of machines running Xen
> today that both support hyperthreading (and have it
> turned on in the BIOS) is very small.  But
> if rumors are true, that may change over the
> next couple of years.

Yup, it sounds like HT is on its way back to Intel machines.

Cheers,
Mark

> Any comments or concerns from CPU vendors?
>
> > -----Original Message-----
> > From: Keir Fraser [mailto:keir.fraser@xxxxxxxxxxxxx]
> > Sent: Monday, April 21, 2008 4:54 PM
> > To: dan.magenheimer@xxxxxxxxxx; Xen-Devel (E-mail)
> > Subject: Re: [Xen-devel] No noht?
> >
> >
> > It disappeared a long time ago, possibly by accident, but
> > noone has yet
> > complained.
> >
> >  -- Keir
> >
> > On 21/4/08 22:42, "Dan Magenheimer"
> >
> > <dan.magenheimer@xxxxxxxxxx> wrote:
> > > Am I correct in reading the code that there is no "noht"
> > > (turn off hyperthreading) option in Xen (except in powerpc)?
> > > This was a surprise to me as I've seen it documented in
> > > various places and I recently suggested it as a diagnostic
> > > to a customer.
> > >
> > > ===================================
> > > If Xen could save time in a bottle / then clocks wouldn't
> >
> > virtually skew /
> >
> > > It would save every tick / for VMs that aren't quick /
> > > and Xen then would send them anew
> > > (with apologies to the late great Jim Croce)
> > > _______________________________________________
> > > Xen-devel mailing list
> > > Xen-devel@xxxxxxxxxxxxxxxxxxx
> > > http://lists.xensource.com/xen-devel
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-devel



-- 
Push Me Pull You - Distributed SCM tool (http://www.cl.cam.ac.uk/~maw48/pmpu/)

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel