From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Zijlstra Subject: Re: [RFC/RFT][PATCH 4/7] cpuidle: menu: Split idle duration prediction from state selection Date: Mon, 5 Mar 2018 12:38:03 +0100 Message-ID: <20180305113803.GO25201@hirez.programming.kicks-ass.net> References: <1657351.s4RTvEoqBQ@aspire.rjw.lan> <2332986.m9oRvTSu8E@aspire.rjw.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <2332986.m9oRvTSu8E@aspire.rjw.lan> Sender: linux-kernel-owner@vger.kernel.org To: "Rafael J. Wysocki" Cc: Thomas Gleixner , Frederic Weisbecker , Paul McKenney , Thomas Ilsche , Doug Smythies , Rik van Riel , Aubrey Li , Mike Galbraith , LKML , Linux PM List-Id: linux-pm@vger.kernel.org On Sun, Mar 04, 2018 at 11:26:24PM +0100, Rafael J. Wysocki wrote: > From: Rafael J. Wysocki > > In order to address the issue with short idle duration predictions > by the idle governor after the tick has been stopped, prepare the > menu governor code for reordering with respect to the timekeeping > code that stops the tick. > > Use the observation that menu_select() can be split into two > functions, one predicting the idle duration and one selecting the > idle state, and rework it accordingly. I actually think this is the wrong way around. We really should be predicting state not duration. Yes the duration thing is an intermediate value, but I don't think it makes any sense what so ever to preserve that in the predictor. The end result is the idle state, we should aim for that. As per: https://lkml.org/lkml/2017/7/18/615 there are definite advantages to _not_ preserving duration information beyond the state boundaries.