From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Rafael J. Wysocki" Subject: Re: [PATCH 3/3][update] PM / sleep: Introduce command line argument for sleep state enumeration Date: Sat, 14 Jun 2014 00:16:26 +0200 Message-ID: <3836835.YaZPGXsEQ2@vostro.rjw.lan> References: <1817539.9KW2IGGbDV@vostro.rjw.lan> <1771964.SepbTGNdtD@vostro.rjw.lan> <20140613214612.GA25712@amd.pavel.ucw.cz> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7Bit Return-path: In-Reply-To: <20140613214612.GA25712@amd.pavel.ucw.cz> Sender: linux-kernel-owner@vger.kernel.org To: Pavel Machek Cc: Linux PM list , Linux Kernel Mailing List List-Id: linux-pm@vger.kernel.org On Friday, June 13, 2014 11:46:12 PM Pavel Machek wrote: > Hi! > > > > > > > From: Rafael J. Wysocki > > > > > > > > > > > > On some systems the platform doesn't support neither > > > > > > PM_SUSPEND_MEM nor PM_SUSPEND_STANDBY, so PM_SUSPEND_FREEZE is the > > > > > > only available system sleep state. However, some user space frameworks > > > > > > only use the "mem" and (sometimes) "standby" sleep state labels, so > > > > > > the users of those systems need to modify user space in order to be > > > > > > able to use system suspend at all and that is not always possible. > > > > > > > > > > I'd say we should fix the frameworks, not add option to change kernel > > > > > interfaces. > > > > > > > > > > Because, as you mentioned, if we add this, we are probably going to > > > > > get stuck with it forever :-(. > > > > > > > > Unfortunately, fixing the frameworks is rather less than realistic in any > > > > reasonable time frame, since Android. :-) > > > > > > Actually, you still have the sources from android, and this issue > > > sounds almost simple enough for binary patch. > > > > > > Android misuses /proc/sys/vm/drop_caches, too, IIRC. Are we going to > > > change interface to match their expectations? They have binder and > > > wakelocks. Are we going to apply those patches just because Android > > > wants that? > > > > That depends on which versions of Android you're talking about. The > > newest ones use the power management interfaces we have upstream. > > Ok, good, so they can fix their code. > > What problem are you solving? Do you have some weird hardware where > suspend to memory is impossible? > > > > Android people usually patch their kernels, anyway, so why not add > > > this one, too? > > > > I'm not talking about Android kernels, but about Android user space. > > I know. Android userspace usually runs on modified kernel, so you can > simply add your patch. But I don't think its suitable for mainline. > > > And this is not only about Android, other distros also have user space that > > uses "mem" only, because nobody has used anything else for a long time anyway. > > For the users of those distros, if they don't want to modify user space, > > having a kernel command line like this is actually helpful. > > Yes, still its wrong place to fix it... This isn't a fix. It's a workaround. > > So I'm really not sure what's the problem? Do you think it's wrong to be > > helpful to users or something? > > It is not wrong to be helpful, but messed up interface is too big a > price. Why? I will have to maintain it after all, right? Rafael