From mboxrd@z Thu Jan 1 00:00:00 1970 From: Martin Cracauer Subject: Need to reuse cpufreq's recalibration code Date: Thu, 9 Nov 2006 16:32:28 -0500 Message-ID: <20061109213228.GA72108@cons.org> Mime-Version: 1.0 Return-path: Content-Disposition: inline List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: cpufreq-bounces@lists.linux.org.uk Errors-To: cpufreq-bounces+glkc-cpufreq=m.gmane.org+glkc-cpufreq=m.gmane.org@lists.linux.org.uk Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: cpufreq@lists.linux.org.uk Hi, I am messing with the CPU clocks outside the cpufreq framework (direct ICS manipulation). While I plan to integrate the code into cpufreq, at this time I cannot since all the math is pocket calculator. What I need to do is recalibrate the CPU clocks after I change the frequency. I observe that the real time clock is correct, but that CPU time is wrongly accounted for and obviously /proc/cpuinfo isn't updated. That happens if I use the TCS, mptmr and hpet. I figured that the best way to proceed is that I reuse some code from cpufreq, namely cpufreq_notify_transition(). I am looking for advise how to achieve this best without messing up too much other code. Maybe I can write an "empty" cpufreq implementation that does nothing but call cpufreq_notify_transition when I touch one of it's /proc files? Platform is x86_64, hardware is Core2 Conroe on D975XBX (975x chipset). Martin -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Martin Cracauer http://www.cons.org/cracauer/