From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Jan Beulich" Subject: RE: [patch 14/33] xen: xen time implementation Date: Wed, 06 Jun 2007 14:21:32 +0200 Message-ID: <4666C2EC.76E4.0078.0@novell.com> References: <907625E08839C4409CE5768403633E0B02561DCD@sefsexmb1.amd.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <907625E08839C4409CE5768403633E0B02561DCD@sefsexmb1.amd.com> Content-Disposition: inline List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Mats Petersson , Keir Fraser Cc: Jeremy Fitzhardinge , Xen-devel List-Id: xen-devel@lists.xenproject.org >> Yes, this could be an issue. Is there any way to get an=20 >> interrupt or MCE >> when thermal throttling occurs? > >As far as I'm aware, no. The normal way to "throttle" a procoessor is a >Pulse-Width-modulation on the STOPCLK pin, and the duration is in the >order of microseconds, so an interrupt every few microseconds may not >leave much room for the processor to actually do ANYTHING at that point. > > >Of course, it would be possible to design such hardware that does >clock-throttling to indicate "Clock throttling started" and "clock >throttling stopped" or some such, but as far as I'm aware, this is not >how it works.=20 There already is such an interrupt being generated (and used) on newer = Intel CPUs - see xen/arch/x86/cpu/mcheck/p4.c. Jan