Subject: Re: port-i386/35676: AMD PowerNow! inconsistent probe message
To: None <,,>
From: Quentin Garnier <>
List: netbsd-bugs
Date: 02/18/2007 16:55:03
The following reply was made to PR port-i386/35676; it has been noted by GNATS.

From: Quentin Garnier <>
Cc: matthew green <>,
Subject: Re: port-i386/35676: AMD PowerNow! inconsistent probe message
Date: Sun, 18 Feb 2007 17:52:13 +0100

 Content-Type: text/plain; charset=us-ascii
 Content-Disposition: inline
 Content-Transfer-Encoding: quoted-printable
 On Mon, Feb 19, 2007 at 01:15:08AM +0900, wrote:
 > Now I've glanced the code in x86/x86/powernow_k8.c, and I'm wondering
 > if it is intended to be used in multiprocessor environment.  It looks
 > like writing sysctl "target" sets the core voltage and frequency of
 > the current processor.  I don't know at all about PowerNow!, is it
 > defined per core?  If it is, I think the sysctl should propagate
 > information to all processors, or, it should provide
 > interface to each specific processor.
 > (I dug into but couldn't
 > find where the PowerNow documentation is.)
 That's a known limitation of both our powernow and speedstep code.
 We currently lack an API to run a callback on a specific processor.
 It's probably not that hard to write, but we already have way too
 many APIs to run stuff in a somewhat differed context, I wouldn't know
 which is best to extend...
 Quentin Garnier - -
 "You could have made it, spitting out benchmarks
 Owe it to yourself not to fail"
 Amplifico, Spitting Out Benchmarks, Hometakes Vol. 2, 2005.
 Content-Type: application/pgp-signature
 Content-Disposition: inline
 Version: GnuPG v1.4.6 (NetBSD)