From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christian Hoelbling Subject: Re: [PATCH 2.6.8-rc2] speedstep SMT support Date: Thu, 29 Jul 2004 12:19:00 +0000 Sender: cpufreq-bounces@www.linux.org.uk Message-ID: <4108EB34.9060505@cern.ch> References: <4107B8DE.5020401@cern.ch> <20040728172028.GB8678@dominikbrodowski.de> Reply-To: holbling@cpt.univ-mrs.fr Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20040728172028.GB8678@dominikbrodowski.de> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: cpufreq-bounces+glkc-cpufreq=gmane.org@www.linux.org.uk Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Dominik Brodowski Cc: cpufreq@www.linux.org.uk hi and thanks alot for your comments! Dominik Brodowski wrote: > >This I'd prefer not to be merged: >a) CONFIG_SMP doesn't mean cpu_sibling_map exist, e.g. on non-x86 architectures >b) it "emulates" sibling-awareness, though the governors are not. > > agree on both of these >Instead, I'd prefer if we keep up with the "only-register-one-CPU" in the >speedstep-ich case until a proper SMT patchkit is applied. > >However, for the other changes: could you consider my opinions, maybe update it >and resend it, please? > > > i will gladly do. however, i am a bit afraid of missing some patch - could you send me a diff of your source tree to a recent vanilla source so i know where to start from? or is there some repository somewhere that i don't know about where the cpufreq development sources are kept? thanks alot, chris