Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sun, 9 Feb 2020 03:44:22 -0800
From:   Olof Johansson <olof@lixom.net>
To:     Krzysztof Kozlowski <krzk@kernel.org>
Cc:     Joe Perches <joe@perches.com>,
        Russell King <linux@armlinux.org.uk>,
        Alexander Shiyan <shc_work@mail.ru>,
        Uwe Kleine-K??nig <u.kleine-koenig@pengutronix.de>,
        Pengutronix Kernel Team <kernel@pengutronix.de>,
        Shawn Guo <shawnguo@kernel.org>,
        Sascha Hauer <s.hauer@pengutronix.de>,
        Fabio Estevam <festevam@gmail.com>,
        NXP Linux Team <linux-imx@nxp.com>,
        Vladimir Zapolskiy <vz@mleia.com>,
        Aaro Koskinen <aaro.koskinen@iki.fi>,
        Tony Lindgren <tony@atomide.com>,
        Maxime Coquelin <mcoquelin.stm32@gmail.com>,
        Alexandre Torgue <alexandre.torgue@st.com>,
        Liviu Dudau <liviu.dudau@arm.com>,
        Sudeep Holla <sudeep.holla@arm.com>,
        Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
        Avi Fishman <avifishman70@gmail.com>,
        Tomer Maimon <tmaimon77@gmail.com>,
        Tali Perry <tali.perry1@gmail.com>,
        Patrick Venture <venture@google.com>,
        Nancy Yuen <yuenn@google.com>,
        Benjamin Fair <benjaminfair@google.com>,
        Arnd Bergmann <arnd@arndb.de>,
        linux-arm-kernel@lists.infradead.org,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        linux-omap@vger.kernel.org,
        linux-stm32@st-md-mailman.stormreply.com, openbmc@lists.ozlabs.org,
        arm@kernel.org, soc@kernel.org
Subject: Re: [PATCH 2/2] ARM: configs: Cleanup old Kconfig options
Message-ID: <20200209114422.as5xpytakhaa3vur@localhost>
References: <20200130195525.4525-1-krzk@kernel.org>
 <20200130195525.4525-2-krzk@kernel.org>
 <9f8a0a8e09893e7087d2212fb0eeb94a908b7be1.camel@perches.com>
 <CAJKOXPf5Mf4FCmtME5yJsBZeP8BkYJgcxkKzS2hd-gp-mq3nag@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAJKOXPf5Mf4FCmtME5yJsBZeP8BkYJgcxkKzS2hd-gp-mq3nag@mail.gmail.com>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Feb 03, 2020 at 12:52:45PM +0100, Krzysztof Kozlowski wrote:
> On Thu, 30 Jan 2020 at 23:06, Joe Perches <joe@perches.com> wrote:
> >
> > On Thu, 2020-01-30 at 20:55 +0100, Krzysztof Kozlowski wrote:
> > > CONFIG_MMC_BLOCK_BOUNCE is gone since commit c3dccb74be28 ("mmc: core:
> > > Delete bounce buffer Kconfig option").
> > >
> > > CONFIG_LBDAF is gone since commit 72deb455b5ec ("block: remove
> > > CONFIG_LBDAF").
> > >
> > > CONFIG_IOSCHED_DEADLINE and CONFIG_IOSCHED_CFQ are gone since
> > > commit f382fb0bcef4 ("block: remove legacy IO schedulers").
> > >
> > > The IOSCHED_DEADLINE was replaced by MQ_IOSCHED_DEADLINE and it will be
> > > now enabled by default (along with MQ_IOSCHED_KYBER).
> > >
> > > The IOSCHED_BFQ seems to replace IOSCHED_CFQ so select it in configs
> > > previously choosing the latter.
> > >
> > > CONFIG_CROSS_COMPILE is gone since commit f1089c92da79 ("kbuild: remove
> > > CONFIG_CROSS_COMPILE support").
> >
> > Hi Krzysztof.
> >
> > There seems there are a lot more of these unused CONFIG_<foo>
> > symbols in various defconfigs. (just for arm and treewide below)
> >
> > ARM defconfigs:
> 
> Hi Joe,
> 
> Nice finding! The trickier point is to nicely remove them because:
> 1. The easiest is 'savedefconfig' but then some valuable options might
> disappear (like recently happened with DEBUG_FS),

Note that while they disappear from the defconfig, they were already not part
of the build. So kernels have been built without them for a while. It's a good
way to surface the problem, but it's pretty clear that trees fall in the forest
here all the time and nobody is noticing.

> 2. They could be removed in automated way with a script. However in
> such case what about replacements? If some symbol was replaced with
> other (or just renamed), maybe we should enable the other one to
> restore the desired functionality?
> 3. Or maybe let's don't care about keeping defconfigs stable and just
> clean them up automatically.

Churning defconfigs is just noise, and a source of annoying needless
conflicts when people do it at the same time. If an option is no longer
in-tree, it doesn't do any harm. But it makes sense to clean up every
now and then like the original patch here.


-Olof