From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Machek Subject: Re: [RFC][PATCH] PM / sleep: Rename symbols, functions and variables related to sleep Date: Wed, 30 Jul 2014 08:50:34 +0200 Message-ID: <20140730065034.GB8024@amd.pavel.ucw.cz> References: <5215142.eaz1Ajyth6@vostro.rjw.lan> <5425640.ztbaPas56b@vostro.rjw.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <5425640.ztbaPas56b@vostro.rjw.lan> Sender: linux-kernel-owner@vger.kernel.org To: "Rafael J. Wysocki" Cc: Linux PM list , Linux Kernel Mailing List , Zhang Rui List-Id: linux-pm@vger.kernel.org On Wed 2014-07-30 01:41:25, Rafael J. Wysocki wrote: > On Monday, July 28, 2014 01:13:56 AM Rafael J. Wysocki wrote: > > From: Rafael J. Wysocki > > > > The names of several symbols, data types, functions and variables > > related to system sleep states are confusing and don't reflect the > > real behavior of those states correctly. > > > > First of all, there generally are two sleep states that require > > platform support and one sleep state that is platform-independent. > > The first two of them are currently known as MEM and STANDBY, > > although these names really only match what the states do on full > > hardware ACPI compliant systems. MEM in particular is supposed > > to mean "suspend-to-RAM", but in fact it means "the deepest sleep > > state available with platform support". The definition of STANDBY > > is even more arbitrary. > > > > Moreover, the remaining sleep state that doesn't need platform support > > is currently called FREEZE, which leads to double confusion with the > > process freezer (used during transitions to all sleep states) and > > with the freeze stage of processing devices during hibernation. > > > > For these reasons, rename the PM_SUSPEND_MEM, PM_SUSPEND_STANDBY > > and PM_SUSPEND_FREEZE symbols to PM_SUSPEND_PLATFORM_DEEP, > > PM_SUSPEND_PLATFORM_SHALLOW and PM_SUSPEND_IDLE_SLEEP, respectively, > > everywhere and rename data types, functions and variables related to > > those states to match the new names of the symbols. > > > > This is a semi-mechanical replacement of names and it should not lead > > to any functional differences. > > > > Signed-off-by: Rafael J. Wysocki > > OK, any comments or objections anyone? > > Or should I just make it a non-RFC and apply it? I don't really like the new names, and I'm not sure if I like direction this is heading. At least on ACPI machines, S1 and S3 is different -- different parts of hardware will stay powered on. So if you want to charge your cellphone with machine suspended, you want S1... and you don't want S1/S3 to change names. (Ok, this is not what you are doing now, but what you plan AFAICT.) > > For these reasons, rename the PM_SUSPEND_MEM, PM_SUSPEND_STANDBY > > and PM_SUSPEND_FREEZE symbols to PM_SUSPEND_PLATFORM_DEEP, > > PM_SUSPEND_PLATFORM_SHALLOW and PM_SUSPEND_IDLE_SLEEP, respectively, I'm not sure why you want to mention "platform there", AFAICT better names would be PM_SUSPEND_DEEP, PM_SUSPEND_SHALLOW, PM_IDLE_SLEEP. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html