From mboxrd@z Thu Jan 1 00:00:00 1970 From: Krzysztof Kozlowski Subject: Re: [PATCH v4 2/3] pinctrl: exynos: Fix GPIO setup failure because domain clock being gated Date: Mon, 05 Jan 2015 15:15:30 +0100 Message-ID: <1420467330.16219.4.camel@AMDC1943> References: <1417777254-26579-1-git-send-email-k.kozlowski@samsung.com> <1417777254-26579-3-git-send-email-k.kozlowski@samsung.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-reply-to: Sender: linux-kernel-owner@vger.kernel.org To: Linus Walleij Cc: Tomasz Figa , Mike Turquette , Sylwester Nawrocki , Kukjin Kim , "linux-kernel@vger.kernel.org" , linux-samsung-soc , "linux-arm-kernel@lists.infradead.org" , Javier Martinez Canillas , "linux-gpio@vger.kernel.org" , "devicetree@vger.kernel.org" , Vivek Gautam , Kevin Hilman , Kyungmin Park , Marek Szyprowski , Bartlomiej Zolnierkiewicz List-Id: linux-gpio@vger.kernel.org On pon, 2015-01-05 at 15:05 +0100, Linus Walleij wrote: > On Fri, Dec 5, 2014 at 12:00 PM, Krzysztof Kozlowski > wrote: > > > The audio subsystem on Exynos 5420 has separate clocks and GPIO. To > > operate properly on GPIOs the main block clock 'mau_epll' must be > > enabled. > > > > This was observed on Peach Pi/Pit and Arndale Octa (after enabling i2s0) > > after introducing runtime PM to pl330 DMA driver. After that commit the > > 'mau_epll' was gated, because the "amba" clock was disabled and there > > were no more users of mau_epll. > > > > The system hang just before probing i2s0 because > > samsung_pinmux_setup() tried to access memory from audss block which was > > gated. > > > > Add a clock property to the pinctrl driver and enable the clock during > > GPIO setup. During normal GPIO operations (set, get, set_direction) the > > clock is not enabled. > > > > Signed-off-by: Krzysztof Kozlowski > > Tested-by: Javier Martinez Canillas > > Tomasz, is this OK and should I apply it for fixes or next? Ha! That is a good question. The issue is fixed for now by the workaround (merged) [1]. The workaround just enables the clock for entire runtime of Exynos 5420-like device. This has some energy impact - around 1.5% in idle [2]. So actually this is question which way we want to solve it: 1. Stick with the workaround (small piece of code, energy impact when not used). 2. Full clock enable on use (big chunk of code, no energy impact when not used). [1] https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=f1e9203e2366164b832d8a6ce10134de8c575178 [2] http://www.spinics.net/lists/linux-samsung-soc/msg39827.html Best regards, Krzysztof