From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753882AbbBTIsX (ORCPT ); Fri, 20 Feb 2015 03:48:23 -0500 Received: from casper.infradead.org ([85.118.1.10]:50317 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752962AbbBTIsV (ORCPT ); Fri, 20 Feb 2015 03:48:21 -0500 Date: Fri, 20 Feb 2015 09:48:07 +0100 From: Peter Zijlstra To: Ingo Molnar Cc: Viresh Kumar , Thomas Gleixner , Ingo Molnar , linaro-kernel@lists.linaro.org, Kevin Hilman , Preeti U Murthy , Daniel Lezcano , linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Frederic Weisbecker , linaro-networking@linaro.org, Steven Miao , Mark Salter , Michal Simek , Ralf Baechle , Ley Foon Tan , Jonas Bonn , "David S. Miller" , Jeff Dike , Guan Xuetao Subject: Re: [PATCH] clockevents: Add (missing) default case for switch blocks Message-ID: <20150220084807.GJ21418@twins.programming.kicks-ass.net> References: <6291e308ab77a480c6b1732e16108c5fe6f66afa.1424412815.git.viresh.kumar@linaro.org> <20150220083842.GA20387@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20150220083842.GA20387@gmail.com> User-Agent: Mutt/1.5.21 (2012-12-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Feb 20, 2015 at 09:38:42AM +0100, Ingo Molnar wrote: > > * Viresh Kumar wrote: > > > Many clockevent drivers are using a switch block for handling modes in their > > ->set_mode() callback. Some of these do not have a 'default' case and adding a > > new mode in the 'enum clock_event_mode', starts giving following warnings for > > these platforms about unhandled modes (e.g. XXX). > > > > warning: enumeration value ‘XXX’ not handled in switch [-Wswitch] > > > > This patch adds default cases for them. > > > > In order to keep things simple, add following to the switch blocks: > > > > default: > > break; > > > > This can lead to different behavior for individual cases. > > > > 1) Some of the drivers don't have any special stuff in their ->set_mode() > > callback before or after the switch blocks. And so this default case would > > simply return for them without any updates to the clockevent device. > > > > 2) But in some cases, the clockevent device is disabled/stopped as soon as we > > enter the ->set_mode() callback and is enabled within the switch block or > > after it. And the clockevent device *may* stay disabled for default case. > > So this whole approach looks fragile for several reasons: > > - 'mode setting' callbacks are just bad by design > because they mix several functions into the same entry > point, complicating the handler functions > unnecessarily. We should reduce complexity, not expand > on it. > > - now by adding 'default' you hide from drivers the > ability to easily discover whether it has been updated > to some new core clockevents mode setting feature or > not. So this patch was a follow on from bd624d75db21 ("clockevents: Introduce mode specific callbacks"). That patch changes the set_mode() interface; and provides per mode functions. New (and updated) drivers should not use ->set_mode() anymore, but it was felt that we do not want to go do flag day changes. And it allows for adding optional modes; not every driver needs to go implement all mode functions if there is a sane default action. But it does mean we need to be able to add values to the enum.