From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dmitry Osipenko Subject: Re: [PATCH v3 1/3] clk: tegra: Mark HCLK, SCLK and EMC as critical Date: Mon, 12 Mar 2018 15:37:28 +0300 Message-ID: <5fc86cc6-947f-ec7c-0c57-905f7aecdfbb@gmail.com> References: <699ce67980d71fd315085ea9785ee6213e0772cb.1515589507.git.digetx@gmail.com> <1487f4a9-d924-3366-c86a-f1bd33ebb1bf@gmail.com> <20180308144437.GJ3529@ulmo> <92084f1b-848a-39e6-7d77-7ceaa6a6f1d3@gmail.com> <20180312071548.GB23060@ulmo> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Return-path: In-Reply-To: <20180312071548.GB23060@ulmo> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: Thierry Reding Cc: Michael Turquette , Stephen Boyd , Peter De Schrijver , Prashant Gaikwad , Jonathan Hunter , linux-clk@vger.kernel.org, linux-tegra@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-tegra@vger.kernel.org On 12.03.2018 10:15, Thierry Reding wrote: > On Fri, Mar 09, 2018 at 05:35:46PM +0300, Dmitry Osipenko wrote: >> On 08.03.2018 17:44, Thierry Reding wrote: >>> On Thu, Mar 01, 2018 at 04:33:29PM +0300, Dmitry Osipenko wrote: >>>> On 15.01.2018 13:56, Dmitry Osipenko wrote: >>>>> On 10.01.2018 16:59, Dmitry Osipenko wrote: >>>>>> Machine dies if HCLK, SCLK or EMC is disabled. Hence mark these clocks >>>>>> as critical. >>>>>> >>>>>> Signed-off-by: Dmitry Osipenko >>>>>> Acked-by: Peter De Schrijver >>>>>> --- >>>>>> >>>>>> Change log: >>>>>> v2: Fixed accidentally missed marking EMC as critical on Tegra30 and >>>>>> Tegra124. Switched to a use of common EMC gate definition on Tegra20 >>>>>> and Tegra30. >>>>>> >>>>>> v3: Dropped marking PLL_P outputs as critical, because seems they are >>>>>> not so critical. Although, I still haven't got a definitive answer >>>>>> about what exact HW functions are affected by the fixed-clocks. >>>>>> Anyway it should be cleaner to correct the actual drivers. >>>>> >>>>> Stephen / Michael, would it be possible to schedule these patches for 4.16? My >>>>> T20 and T30 devices aren't working without the 'critical clocks' patch. Things >>>>> happen to work with the opensource u-boot, but not with the proprietary >>>>> bootloader. It's probably not a big deal that out-of-tree devices are broken, >>>>> although would be nice to have one problem less. >>>> >>>> Guys, is there anything I could do to get these patches in linux-next? >>> >>> I've picked these up into the for-4.17/clk branch in the Tegra tree. I >>> already have that branch for the MBIST patches which are a dependency >>> for the for-4.17/soc branch. >> >> Thank you very much! Could you please add stable tag to this ("Mark HCLK, SCLK >> and EMC as critical") patch? It would be nice to have 4.16 unbroken eventually >> for those who (have to) use downstream android bootloader. >> >> Cc: # v4.16 > > Should we add a Fixes: line instead? That way it will get automatically > backported to all applicable stable releases. > > This is a little complicated because the clocks were introduced in > different commits, most of them a very long time ago: > > a7c8485a0ebbdce303c6709e208bb4fd08aff8ad (sclk) > 2db04f16b589c6c96bd07df3f1ef8558bfdb6810 (emc) > 76ebc134d45d7e6e1dc29fdcef4e539c5bc76eb8 (emc) > ... > > So Fixes: is perhaps not desirable after all, but then, so isn't tagging > v4.16 specifically because this is a bug since almost forever, so v4.16 > is fairly arbitrary. Shouldn't we at least get this fixed for the last > couple of LTS releases? That issue was masked in earlier kernels and only appears in 4.16, there is no need to backport it further. I thought that adding of 'stable' tag is enough to get patch backported automatically, so let's add 'fixes' instead: Fixes: 109eba2eb61a ("clk: tegra: Mark APB clock as critical") This patch applies cleanly to 4.16, everything should be fine.