From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH] clocksource: Allow toggling between runtime and persistent clocksource for idle Date: Tue, 7 Jul 2015 03:24:58 -0700 Message-ID: <20150707102458.GR10705@atomide.com> References: <1436166725-18353-1-git-send-email-tony@atomide.com> <20150706152201.GJ10705@atomide.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: John Stultz Cc: Thomas Gleixner , lkml , "linux-omap@vger.kernel.org" , Felipe Balbi , Nishanth Menon , Yingjoe Chen , "Rafael J. Wysocki" , Peter Zijlstra List-Id: linux-omap@vger.kernel.org * John Stultz [150706 10:53]: > On Mon, Jul 6, 2015 at 8:46 AM, Thomas Gleixner wrote: > > On Mon, 6 Jul 2015, Tony Lindgren wrote: > >> * Thomas Gleixner [150706 07:20]: > >> > On Mon, 6 Jul 2015, Tony Lindgren wrote: > >> The timekeeping accuracy issue certainly needs some thinking, and > >> also the resolution between the clocksources can be different.. In the > >> test case I have the slow timer is always on and of a lower resolution > >> than the ARM global timer being used during runtime. > >> > >> Got some handy timer test in mind you want me to run to provide data > >> on the accuracy? > > > > John Stultz might have something. > > You can turn on ntp stats in your ntp.conf and chart the loopstats > data w/ gnuplot: > > set terminal png > set output "loopstat.png" > plot "/var/log/ntpstats/loopstats." using 2:3 with linespoints > > > I also have the drift-log.py and graph-log.py scripts I use here: > https://github.com/johnstultz-work/timetests > > But those aren't really ready for mass consumption, as they're > probably not very cross-distro compatible. Great thanks, I'll play a bit with those to see how bad the jitter gets. Regrds, Tony