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;
MIME-Version: 1.0
References: <20191018154052.1276506-1-arnd@arndb.de> <20191018154201.1276638-9-arnd@arndb.de>
 <bb4713a0-60bd-8d27-874f-e7e3a5adaec8@roeck-us.net>
In-Reply-To: <bb4713a0-60bd-8d27-874f-e7e3a5adaec8@roeck-us.net>
From:   Arnd Bergmann <arnd@arndb.de>
Date:   Tue, 22 Oct 2019 11:44:33 +0200
Message-ID: <CAK8P3a0GPVs+PqUwsL2H8VJqLh=MJnMTwPu9nhX+Nq-xiMS1yQ@mail.gmail.com>
Subject: Re: [PATCH 09/46] watchdog: sa1100: use platform device registration
To:     Guenter Roeck <linux@roeck-us.net>
Cc:     Daniel Mack <daniel@zonque.org>,
        Haojian Zhuang <haojian.zhuang@gmail.com>,
        Robert Jarzmik <robert.jarzmik@free.fr>,
        Linus Walleij <linus.walleij@linaro.org>,
        Wim Van Sebroeck <wim@linux-watchdog.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Linux ARM <linux-arm-kernel@lists.infradead.org>,
        LINUXWATCHDOG <linux-watchdog@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sat, Oct 19, 2019 at 4:07 PM Guenter Roeck <linux@roeck-us.net> wrote:

> > @@ -319,10 +316,13 @@ static struct platform_device *sa11x0_devices[] __initdata = {
> >
> >   static int __init sa1100_init(void)
> >   {
> > +     struct resource wdt_res = DEFINE_RES_MEM(0x90000000, 0x20);
> >       pm_power_off = sa1100_power_off;
> >
> >       regulator_has_full_constraints();
> >
> > +     platform_device_register_simple("sa1100_wdt", -1, &wdt_res, 1);
> > +
> >       return platform_add_devices(sa11x0_devices, ARRAY_SIZE(sa11x0_devices));
>
> Wouldn't it be better to add the watchdog device to sa11x0_devices ?

Generally speaking, platform_device_register_simple() is better than
platform_add_devices(), it does the same thing with fewer source lines
and smaller object code, and it doesn't have the problem of lifetime rules
for statically allocated reference-counted devices.

One day we may want to replace all static platform_device instances with
platform_device_info instead, but right now there are too many of those.

I can change this one to a platform_device for consistency though if you
think it's worth it.

     Arnd