[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6230401.JFgrqYCo0I@wuerfel>
Date: Wed, 06 Jul 2016 15:45:31 +0200
From: Arnd Bergmann <arnd@...db.de>
To: Paul Gortmaker <paul.gortmaker@...driver.com>
Cc: Fabio Estevam <festevam@...il.com>,
Linus Walleij <linus.walleij@...aro.org>,
Laxman Dewangan <ldewangan@...dia.com>,
"linux-gpio@...r.kernel.org" <linux-gpio@...r.kernel.org>,
linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] pinctrl: imx21: include linux/module.h
On Wednesday, July 6, 2016 9:21:39 AM CEST Paul Gortmaker wrote:
> > --- a/drivers/pinctrl/freescale/pinctrl-imx21.c
> > +++ b/drivers/pinctrl/freescale/pinctrl-imx21.c
> > @@ -325,7 +325,6 @@ static const struct of_device_id imx21_pinctrl_of_match[] =
> > { .compatible = "fsl,imx21-iomuxc", },
> > { }
> > };
> > -MODULE_DEVICE_TABLE(of, imx21_pinctrl_of_match);
> >
> > static struct platform_driver imx21_pinctrl_driver = {
> > .driver = {
> >
> >
> > If you agree I can send a formal patch.
>
> Yes, deleting the MODULE_DEVICE_TABLE is the right fix. I'm curious why
> that never showed up in all the build coverage I did...
I needed several hundred randconfig builds before hitting it, I think
it's just rare because the driver is not enabled most of the time, and
it's possible that the header sometimes gets included implicitly.
Arnd
Powered by blists - more mailing lists