From patchwork Mon Nov 11 16:23:58 2013 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Nishanth Menon X-Patchwork-Id: 3168371 Return-Path: X-Original-To: patchwork-linux-omap@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork1.web.kernel.org (Postfix) with ESMTP id EAB589F3AE for ; Mon, 11 Nov 2013 16:24:53 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 3107620259 for ; Mon, 11 Nov 2013 16:24:49 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id C91A620256 for ; Mon, 11 Nov 2013 16:24:47 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754719Ab3KKQYb (ORCPT ); Mon, 11 Nov 2013 11:24:31 -0500 Received: from bear.ext.ti.com ([192.94.94.41]:43852 "EHLO bear.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754710Ab3KKQYZ (ORCPT ); Mon, 11 Nov 2013 11:24:25 -0500 Received: from dflxv15.itg.ti.com ([128.247.5.124]) by bear.ext.ti.com (8.13.7/8.13.7) with ESMTP id rABGNxi1030118; Mon, 11 Nov 2013 10:23:59 -0600 Received: from DLEE71.ent.ti.com (dlee71.ent.ti.com [157.170.170.114]) by dflxv15.itg.ti.com (8.14.3/8.13.8) with ESMTP id rABGNxUL028946; Mon, 11 Nov 2013 10:23:59 -0600 Received: from dflp32.itg.ti.com (10.64.6.15) by DLEE71.ent.ti.com (157.170.170.114) with Microsoft SMTP Server id 14.2.342.3; Mon, 11 Nov 2013 10:23:58 -0600 Received: from localhost (ileax41-snat.itg.ti.com [10.172.224.153]) by dflp32.itg.ti.com (8.14.3/8.13.8) with ESMTP id rABGNwT7002850; Mon, 11 Nov 2013 10:23:58 -0600 Date: Mon, 11 Nov 2013 10:23:58 -0600 From: Nishanth Menon To: Kevin Hilman CC: Tony Lindgren , , , , , Subject: Re: [PATCH] ARM: OMAP2+: omap_device: maintain sane runtime pm status around suspend/resume Message-ID: <20131111162358.GA6451@kahuna> References: <1383857942-23630-1-git-send-email-nm@ti.com> <87iow3rd3y.fsf@linaro.org> <527C2B78.2070201@ti.com> <874n7npw1e.fsf@linaro.org> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <874n7npw1e.fsf@linaro.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-omap-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-omap@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP On 16:38-20131107, Kevin Hilman wrote: [...] > That's debatable I guess. The ideal world is that runtime PM hides all > of this, but I'm not sure it's achievable in all cases. > Agreed. some drivers like edma need to save and restore context around suspend. [...] > No, that sysfs knob is for disabling runtime PM. We still want the > device to hit low-power state in system suspend. Solving that problem > is half the reason we have this omap_device noirq mess in the first > place. > > You need to test this by disabling runtime PM from userspace and ensure > that the low power state is still hit during suspend. > Done and it still does work, makes sense since it just ensures that runtime PM's dev->power.runtime_status is set to RPM_SUSPENDED instead of RPM_ACTIVE for devices that depend on autosuspend. Logs (based on vendor kernel which has relevant out of tree patches to enable suspend resume - still in the works): AM335x-BBB: http://pastie.org/8472182 OMAP5-uEVM: http://pastie.org/8472183 > >> > >>> + /* NOTE: *might* indicate driver race */ > >> > >> Yes, a driver race which should then be fixed in the driver. > > > > true if this is a non-autosuspend device, in auto suspend devices, > > this could be a regular phenomenon if timeout is pretty large.. but > > atleast that should allow debug. > > Agreed. I wasn't thinking about the autosuspend case. Thanks for > clarifying. > > >> > >>> + dev_dbg(dev, "%s: Force suspending\n", > >>> + __func__); > >>> + pm_runtime_set_suspended(dev); > >>> + od->flags |= OMAP_DEVICE_SUSPEND_FORCED; > >> > >> Not sure why you need an additonal flag. Why not just always do this > >> and use the existin OMAP_DEVICE_SUSPENDED flag. > > > > restore of runtime data structure state is only needed for specific > > devices - not all.. > > The question is why do you a flag in addition to OMAP_DEVICE_SUSPEND. > Whenever that flag is set, omap_device has gone behind your back, and > the runtime PM status should be kept in sync. Yes, you are right, originally, I had intended this to indicate devices that needed to be runtime_status updated, but then, now I realize that it is true for all devices that have OMAP_DEVICE_SUSPEND set. It can be applied without an additional flag. Do see if the updated patch is more sensible: -- >8 -- From 96b5a7b89fef4ba55bca48bae83e5536d697c6c4 Mon Sep 17 00:00:00 2001 From: Nishanth Menon Date: Thu, 24 Oct 2013 09:12:42 -0500 Subject: [PATCH V2] ARM: OMAP2+: omap_device: maintain sane runtime pm status around suspend/resume OMAP device hooks around suspend|resume_noirq ensures that hwmod devices are forced to idle using omap_device_idle/enable as part of the last stage of suspend activity. For a device such as i2c who uses autosuspend, it is possible to enter the suspend path with dev->power.runtime_status = RPM_ACTIVE. As part of the suspend flow, the generic runtime logic would increment it's dev->power.disable_depth to 1. This should prevent further pm_runtime_get_sync from succeeding once the runtime_status has been set to RPM_SUSPENDED. Now, as part of the suspend_noirq handler in omap_device, we force the following: if the device status is !suspended, we force the device to idle using omap_device_idle (clocks are cut etc..). This ensures that from a hardware perspective, the device is "suspended". However, runtime_status is left to be active. *if* an operation is attempted after this point to pm_runtime_get_sync, runtime framework depends on runtime_status to indicate accurately the device status, and since it sees it to be ACTIVE, it assumes the module is functional and returns a non-error value. As a result the user will see pm_runtime_get succeed, however a register access will crash due to the lack of clocks. To prevent this from happening, we should ensure that runtime_status exactly indicates the device status. As a result of this change any further calls to pm_runtime_get* would return -EACCES (since disable_depth is 1). On resume, we restore the clocks and runtime status exactly as we suspended with. Reported-by: J Keerthy Signed-off-by: Nishanth Menon Acked-by: Rajendra Nayak Acked-by: Kevin Hilman --- arch/arm/mach-omap2/omap_device.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/arm/mach-omap2/omap_device.c b/arch/arm/mach-omap2/omap_device.c index b69dd9a..f97b34b 100644 --- a/arch/arm/mach-omap2/omap_device.c +++ b/arch/arm/mach-omap2/omap_device.c @@ -621,6 +621,7 @@ static int _od_suspend_noirq(struct device *dev) if (!ret && !pm_runtime_status_suspended(dev)) { if (pm_generic_runtime_suspend(dev) == 0) { + pm_runtime_set_suspended(dev); omap_device_idle(pdev); od->flags |= OMAP_DEVICE_SUSPENDED; } @@ -634,10 +635,10 @@ static int _od_resume_noirq(struct device *dev) struct platform_device *pdev = to_platform_device(dev); struct omap_device *od = to_omap_device(pdev); - if ((od->flags & OMAP_DEVICE_SUSPENDED) && - !pm_runtime_status_suspended(dev)) { + if (od->flags & OMAP_DEVICE_SUSPENDED) { od->flags &= ~OMAP_DEVICE_SUSPENDED; omap_device_enable(pdev); + pm_runtime_set_active(dev); pm_generic_runtime_resume(dev); }