From: Javier Martinez Canillas <javier@osg.samsung.com> To: Krzysztof Kozlowski <k.kozlowski@samsung.com> Cc: Alim Akhtar <alim.akhtar@samsung.com>, linux-samsung-soc@vger.kernel.org, kgene@kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH 7/7] ARM: EXYNOS: Remove code for restart and poweroff for exynos SoCs Date: Mon, 19 Oct 2015 16:07:46 +0200 [thread overview] Message-ID: <5624F932.7030001@osg.samsung.com> (raw) In-Reply-To: <CAJKOXPcvKeDNwof5eyD5XjzUiZrHoyShD8gOPGo7uUxEyypQbw@mail.gmail.com> Hello Krzysztof, On 10/19/2015 03:28 PM, Krzysztof Kozlowski wrote: > 2015-10-19 18:56 GMT+09:00 Javier Martinez Canillas <javier@osg.samsung.com>: >> Hello, >> >> On 10/19/2015 09:00 AM, Krzysztof Kozlowski wrote: >>> On 19.10.2015 15:03, Alim Akhtar wrote: >>>> Now we can use the generic syscon-{reboot/poweroff} drivers, >>>> so we don't need special handling for reboot/poweroff in >>>> exynos pmu driver. This patch remove the same. >>>> >>>> Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com> >>>> --- >>>> arch/arm/mach-exynos/pmu.c | 43 ------------------------------------------- >>>> 1 file changed, 43 deletions(-) >>> >>> I think that removal of this stuff will effectively remove the >>> restart/poweroff handlers from: >>> 1. Other defconfigs, like multi_v7 >>> 2. Custom configs. >>> >> >> This will also break old DTBs that don't have a "syscon-poweroff" device >> node that contains the necessary PMU regmap, offset and mask information. > > I am not sure whether this is ABI break issue. There was no compatible > mentioning that "reset works" which now would be replaced. The > existing PMU compatible (like samsung,exynos4412-pmu) does not mention > "reset" as a feature coming with this compatible. > > So no ABI break. > > I deliberately didn't use the "DT ABI break" expression since as you said is not part of the documented DT bindings. But what I said is that this change will break old DTBs with newer kernels since reboot and power off will stop working after $SUBJECT. I'm not a particular fan of the stable DT idea since in practice it seems to do more harm than good but since that was decided, the expectation for users is that booting a new kernel with an old DT should not cause any regression. So I think that at least a comment in the commit message is needed so if there are people really using old DTs with newer kernels on Exynos boards, they can know that the commit causes such an issue instead of having to figure it out themselves. >> >>> Previously this code was always compiled in for ARCH_EXYNOS. Now it is >>> not so I am thinking about selecting necessary drivers from main exynos >>> Kconfig symbol. That could be tricky though, because "select" should be >>> used only for non-visible symbols. >>> >>> Any ideas how to solve that? >>> >> >> Is true that select should only be used for non-visible symbols but there >> are others user visible symbols that are selected by ARCH_EXYNOS such as >> EXYNOS_THERMAL. So I think selecting the regmap syscon reset stuff there >> is a sensible option. > > Selecting from defconfig is not sufficient... since I do not have > other idea than selecting then ovak, but Alim please check it whether > it does not create circular dependencies on various configs. > Agreed, Kconfig circular dependencies is the reason why select is avoided. Fortunately now the 0-day bot analyzes even posted patches so it's possible that such an issue could be found even before these patches are merged. > Best regards, > Krzysztof > Best regards, -- Javier Martinez Canillas Open Source Group Samsung Research America
WARNING: multiple messages have this Message-ID (diff)
From: javier@osg.samsung.com (Javier Martinez Canillas) To: linux-arm-kernel@lists.infradead.org Subject: [PATCH 7/7] ARM: EXYNOS: Remove code for restart and poweroff for exynos SoCs Date: Mon, 19 Oct 2015 16:07:46 +0200 [thread overview] Message-ID: <5624F932.7030001@osg.samsung.com> (raw) In-Reply-To: <CAJKOXPcvKeDNwof5eyD5XjzUiZrHoyShD8gOPGo7uUxEyypQbw@mail.gmail.com> Hello Krzysztof, On 10/19/2015 03:28 PM, Krzysztof Kozlowski wrote: > 2015-10-19 18:56 GMT+09:00 Javier Martinez Canillas <javier@osg.samsung.com>: >> Hello, >> >> On 10/19/2015 09:00 AM, Krzysztof Kozlowski wrote: >>> On 19.10.2015 15:03, Alim Akhtar wrote: >>>> Now we can use the generic syscon-{reboot/poweroff} drivers, >>>> so we don't need special handling for reboot/poweroff in >>>> exynos pmu driver. This patch remove the same. >>>> >>>> Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com> >>>> --- >>>> arch/arm/mach-exynos/pmu.c | 43 ------------------------------------------- >>>> 1 file changed, 43 deletions(-) >>> >>> I think that removal of this stuff will effectively remove the >>> restart/poweroff handlers from: >>> 1. Other defconfigs, like multi_v7 >>> 2. Custom configs. >>> >> >> This will also break old DTBs that don't have a "syscon-poweroff" device >> node that contains the necessary PMU regmap, offset and mask information. > > I am not sure whether this is ABI break issue. There was no compatible > mentioning that "reset works" which now would be replaced. The > existing PMU compatible (like samsung,exynos4412-pmu) does not mention > "reset" as a feature coming with this compatible. > > So no ABI break. > > I deliberately didn't use the "DT ABI break" expression since as you said is not part of the documented DT bindings. But what I said is that this change will break old DTBs with newer kernels since reboot and power off will stop working after $SUBJECT. I'm not a particular fan of the stable DT idea since in practice it seems to do more harm than good but since that was decided, the expectation for users is that booting a new kernel with an old DT should not cause any regression. So I think that at least a comment in the commit message is needed so if there are people really using old DTs with newer kernels on Exynos boards, they can know that the commit causes such an issue instead of having to figure it out themselves. >> >>> Previously this code was always compiled in for ARCH_EXYNOS. Now it is >>> not so I am thinking about selecting necessary drivers from main exynos >>> Kconfig symbol. That could be tricky though, because "select" should be >>> used only for non-visible symbols. >>> >>> Any ideas how to solve that? >>> >> >> Is true that select should only be used for non-visible symbols but there >> are others user visible symbols that are selected by ARCH_EXYNOS such as >> EXYNOS_THERMAL. So I think selecting the regmap syscon reset stuff there >> is a sensible option. > > Selecting from defconfig is not sufficient... since I do not have > other idea than selecting then ovak, but Alim please check it whether > it does not create circular dependencies on various configs. > Agreed, Kconfig circular dependencies is the reason why select is avoided. Fortunately now the 0-day bot analyzes even posted patches so it's possible that such an issue could be found even before these patches are merged. > Best regards, > Krzysztof > Best regards, -- Javier Martinez Canillas Open Source Group Samsung Research America
next prev parent reply other threads:[~2015-10-19 14:07 UTC|newest] Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top 2015-10-19 6:03 [PATCH 0/7] Switch to generic syscon regmap based drivers Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 6:03 ` [PATCH 1/7] arm: dts: Add syscon-{reboot, poweroff} nodes for exynos3250 SoCs Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 8:43 ` [1/7] " Pankaj Dubey 2015-10-19 8:43 ` Pankaj Dubey 2015-10-19 13:03 ` [PATCH 1/7] " Javier Martinez Canillas 2015-10-19 13:03 ` Javier Martinez Canillas 2015-10-20 9:00 ` Alim Akhtar 2015-10-20 9:00 ` Alim Akhtar 2015-10-19 6:03 ` [PATCH 2/7] arm: dts: Add syscon-{reboot, poweroff} nodes for exynos4 Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 8:45 ` [2/7] " Pankaj Dubey 2015-10-19 8:45 ` Pankaj Dubey 2015-10-19 13:08 ` [PATCH 2/7] " Javier Martinez Canillas 2015-10-19 13:08 ` Javier Martinez Canillas 2015-10-19 6:03 ` [PATCH 3/7] arm: dts: Add syscon-{reboot, poweroff} nodes for exynos5 Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 8:47 ` [3/7] " Pankaj Dubey 2015-10-19 8:47 ` Pankaj Dubey 2015-10-19 13:09 ` [PATCH 3/7] " Javier Martinez Canillas 2015-10-19 13:09 ` Javier Martinez Canillas 2015-10-19 6:03 ` [PATCH 4/7] arm: dts: Add syscon-{reboot, poweroff} nodes for exynos5410 SoC Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 8:56 ` [4/7] " Pankaj Dubey 2015-10-19 8:56 ` Pankaj Dubey 2015-10-19 13:10 ` [PATCH 4/7] " Javier Martinez Canillas 2015-10-19 13:10 ` Javier Martinez Canillas 2015-10-19 6:03 ` [PATCH 5/7] ARM: exynos_defconfig: Normalize exynos defconfig Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 13:25 ` Javier Martinez Canillas 2015-10-19 13:25 ` Javier Martinez Canillas 2015-10-19 15:58 ` Alim Akhtar 2015-10-19 15:58 ` Alim Akhtar 2015-10-19 16:38 ` Javier Martinez Canillas 2015-10-19 16:38 ` Javier Martinez Canillas 2015-10-19 6:03 ` [PATCH 6/7] ARM: exynos_defconfig: Enable generic syscon-{reboot, poweroff} drivers Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 6:03 ` [PATCH 7/7] ARM: EXYNOS: Remove code for restart and poweroff for exynos SoCs Alim Akhtar 2015-10-19 6:03 ` Alim Akhtar 2015-10-19 7:00 ` Krzysztof Kozlowski 2015-10-19 7:00 ` Krzysztof Kozlowski 2015-10-19 8:27 ` Alim Akhtar 2015-10-19 8:27 ` Alim Akhtar 2015-10-19 9:56 ` Javier Martinez Canillas 2015-10-19 9:56 ` Javier Martinez Canillas 2015-10-19 10:23 ` Alim Akhtar 2015-10-19 10:23 ` Alim Akhtar 2015-10-19 13:28 ` Krzysztof Kozlowski 2015-10-19 13:28 ` Krzysztof Kozlowski 2015-10-19 14:07 ` Javier Martinez Canillas [this message] 2015-10-19 14:07 ` Javier Martinez Canillas 2015-10-19 16:13 ` Alim Akhtar 2015-10-19 16:13 ` Alim Akhtar 2015-10-19 16:42 ` Javier Martinez Canillas 2015-10-19 16:42 ` Javier Martinez Canillas 2015-10-19 16:06 ` Alim Akhtar 2015-10-19 16:06 ` Alim Akhtar 2015-10-20 0:20 ` Krzysztof Kozlowski 2015-10-20 0:20 ` Krzysztof Kozlowski 2015-10-19 6:54 ` [PATCH 0/7] Switch to generic syscon regmap based drivers Krzysztof Kozlowski 2015-10-19 6:54 ` Krzysztof Kozlowski 2015-10-19 7:40 ` Alim Akhtar 2015-10-19 7:40 ` Alim Akhtar 2015-10-19 16:08 ` Tobias Jakobi 2015-10-19 16:08 ` Tobias Jakobi 2015-10-19 16:36 ` Alim Akhtar 2015-10-19 16:36 ` Alim Akhtar 2015-10-19 21:18 ` Moritz Fischer 2015-10-19 21:18 ` Moritz Fischer 2015-10-20 8:56 ` Alim Akhtar 2015-10-20 8:56 ` Alim Akhtar 2015-10-20 13:43 ` Markus Reichl 2015-10-20 13:43 ` Markus Reichl 2015-10-20 13:56 ` Krzysztof Kozlowski 2015-10-20 13:56 ` Krzysztof Kozlowski
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=5624F932.7030001@osg.samsung.com \ --to=javier@osg.samsung.com \ --cc=alim.akhtar@samsung.com \ --cc=devicetree@vger.kernel.org \ --cc=k.kozlowski@samsung.com \ --cc=kgene@kernel.org \ --cc=linux-arm-kernel@lists.infradead.org \ --cc=linux-samsung-soc@vger.kernel.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.