From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Lezcano Subject: Re: [V4 patch 03/15] cpuidle: make a single register function for all Date: Tue, 23 Apr 2013 14:51:41 +0200 Message-ID: <517683DD.6060805@linaro.org> References: <1366707285-12179-1-git-send-email-daniel.lezcano@linaro.org> <1366707285-12179-4-git-send-email-daniel.lezcano@linaro.org> <51767F57.7060809@ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mail-ee0-f54.google.com ([74.125.83.54]:33997 "EHLO mail-ee0-f54.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755463Ab3DWMvr (ORCPT ); Tue, 23 Apr 2013 08:51:47 -0400 Received: by mail-ee0-f54.google.com with SMTP id e51so249528eek.13 for ; Tue, 23 Apr 2013 05:51:45 -0700 (PDT) In-Reply-To: <51767F57.7060809@ti.com> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Santosh Shilimkar Cc: rjw@sisk.pl, andrew@lunn.ch, kgene.kim@samsung.com, jason@lakedaemon.net, horms@verge.net.au, patches@linaro.org, khilman@deeprootsystems.com, magnus.damm@gmail.com, linux-pm@vger.kernel.org, swarren@wwwdotorg.org, nsekhar@ti.com, rob.herring@calxeda.com, deepthi@linux.vnet.ibm.com, tony@atomide.com, lethal@linux-sh.org, linaro-kernel@lists.linaro.org, kernel@pengutronix.de, josephl@nvidia.com, jkosina@suse.cz, plagnioj@jcrosoft.com, linux@maxim.org.za, linux-arm-kernel@lists.infradead.org On 04/23/2013 02:32 PM, Santosh Shilimkar wrote: > On Tuesday 23 April 2013 02:24 PM, Daniel Lezcano wrote: >> The usual scheme to initialize a cpuidle driver on a SMP is: >> >> cpuidle_register_driver(drv); >> for_each_possible_cpu(cpu) { >> device =3D &per_cpu(cpuidle_dev, cpu); >> cpuidle_register_device(device); >> } >> >> This code is duplicated in each cpuidle driver. >> >> On UP systems, it is done this way: >> >> cpuidle_register_driver(drv); >> device =3D &per_cpu(cpuidle_dev, cpu); >> cpuidle_register_device(device); >> >> On UP, the macro 'for_each_cpu' does one iteration: >> >> #define for_each_cpu(cpu, mask) \ >> for ((cpu) =3D 0; (cpu) < 1; (cpu)++, (void)mask) >> >> Hence, the initialization loop is the same for UP than SMP. >> >> Beside, we saw different bugs / mis-initialization / return code unc= hecked in >> the different drivers, the code is duplicated including bugs. After = fixing all >> these ones, it appears the initialization pattern is the same for ev= eryone. >> >> Please note, some drivers are doing dev->state_count =3D drv->state_= count. This is >> not necessary because it is done by the cpuidle_enable_device functi= on in the >> cpuidle framework. This is true, until you have the same states for = all your >> devices. Otherwise, the 'low level' API should be used instead with = the specific >> initialization for the driver. >> >> Let's add a wrapper function doing this initialization with a cpumas= k parameter >> for the coupled idle states and use it for all the drivers. >> >> That will save a lot of LOC, consolidate the code, and the modificat= ions in the >> future could be done in a single place. Another benefit is the conso= lidation of >> the cpuidle_device variable which is now in the cpuidle framework an= d no longer >> spread accross the different arch specific drivers. >> >> Signed-off-by: Daniel Lezcano >> --- >=20 > I don't see you have addressed the comment on V3 [1] i gave for the s= ubject patch > Any reason ? Yes, sorry for not answering. This modification should be handled in the __cpuidle_register_device function. diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index 49e8d30..936d862 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -372,7 +372,7 @@ static int __cpuidle_register_device(struct cpuidle_device *dev) int ret; struct cpuidle_driver *drv =3D cpuidle_get_cpu_driver(dev); - if (!try_module_get(drv->owner)) + if (!drv || !try_module_get(drv->owner)) return -EINVAL; per_cpu(cpuidle_devices, dev->cpu) =3D dev; Thus, the cpuidle_register function is not impacted by this as it will always do cpuidle_register_driver, followed by cpuidle_register_device. -- Daniel --=20 Linaro.org =E2=94=82 Open source software for= ARM SoCs =46ollow Linaro: Facebook | Twitter | Blog