Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:5 as permitted sender) client-ip=2620:137:e000::3:5;
Date:   Wed, 13 Dec 2023 08:32:48 +0000
From:   Sakari Ailus <sakari.ailus@linux.intel.com>
To:     Arnd Bergmann <arnd@kernel.org>
Cc:     Mauro Carvalho Chehab <mchehab@kernel.org>,
        Nathan Chancellor <nathan@kernel.org>,
        Jacopo Mondi <jacopo.mondi@ideasonboard.com>,
        Hans Verkuil <hverkuil-cisco@xs4all.nl>,
        Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
        Arnd Bergmann <arnd@arndb.de>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Bill Wendling <morbo@google.com>,
        Justin Stitt <justinstitt@google.com>,
        Hans de Goede <hdegoede@redhat.com>,
        Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>,
        linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
        llvm@lists.linux.dev
Subject: Re: [PATCH] media: i2c: mt9m114: add CONFIG_COMMON_CLK dependency
Message-ID: <ZXlsMAgkqkGb4kVI@kekkonen.localdomain>
References: <20231212213625.3653558-1-arnd@kernel.org>
 <ZXlmnmGepRnMyNfr@kekkonen.localdomain>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ZXlmnmGepRnMyNfr@kekkonen.localdomain>
Precedence: bulk

On Wed, Dec 13, 2023 at 08:09:03AM +0000, Sakari Ailus wrote:
> Hi Arnd,
> 
> On Tue, Dec 12, 2023 at 10:18:04PM +0100, Arnd Bergmann wrote:
> > From: Arnd Bergmann <arnd@arndb.de>
> > 
> > With clang-16, building without COMMON_CLK triggers a range check on
> > udelay() because of a constant division-by-zero calculation:
> > 
> > ld.lld: error: undefined symbol: __bad_udelay
> > >>> referenced by mt9m114.c
> > >>>               drivers/media/i2c/mt9m114.o:(mt9m114_power_on) in archive vmlinux.a
> > 
> > Avoid this by adding a Kconfig dependency that avoids the broken build.
> 
> This sounds like an odd way to fix an issue with udelay(). On which arch
> does it happen? Wouldn't it be better to fix the udelay() problem in the
> source?
> 
> A quick check reveals there are about 2400 files using udelay.

After discussing with Tomi, I think the driver could be improved, too, by
adding checks for clock frequency and avoiding an obvious potential
division by zero if clk_get_rate() happens to return 0. Switching to
fsleep() would probably address the Clang 16 issue, but that doesn't seem
to be the primary cause here anyway.

-- 
Sakari Ailus