All of lore.kernel.org
 help / color / mirror / Atom feed
From: arnd@arndb.de (Arnd Bergmann)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 2/3] ARM: mxs: add gpio-mxs platform devices
Date: Fri, 20 May 2011 12:23:00 +0200	[thread overview]
Message-ID: <201105201223.00427.arnd@arndb.de> (raw)
In-Reply-To: <1305885446-27404-3-git-send-email-shawn.guo@linaro.org>

On Friday 20 May 2011 11:57:25 Shawn Guo wrote:

> --- /dev/null
> +++ b/arch/arm/mach-mxs/devices/platform-gpio-mxs.c
> @@ -0,0 +1,92 @@
> +/*
> + * Copyright 2011 Freescale Semiconductor, Inc. All Rights Reserved.
> + *
> + * This program is free software; you can redistribute it and/or modify it under
> + * the terms of the GNU General Public License version 2 as published by the
> + * Free Software Foundation.
> + */

Hmm, I forgot to discuss this in Budapest, but I'm still not convinced about
the code ownership here, I think it should be Copyright Linaro Ltd when a
Linaro assignee writes it, not the member company that you work for.

If your manager thinks it should be copyright Freescale, I would suggest
we discuss it on the Linaro private mailing list so we can find a solution
that everyone is happy with. No need to bother the public with this.

> +#include <linux/compiler.h>
> +#include <linux/err.h>
> +#include <linux/init.h>
> +
> +#include <mach/mx23.h>
> +#include <mach/mx28.h>
> +#include <mach/devices-common.h>
> +
> +struct mxs_gpio_data {
> +	int id;
> +	resource_size_t iobase;
> +	resource_size_t iosize;
> +	resource_size_t irq;
> +};

You don't use iosize anywhere.

> +#define mxs_gpio_data_entry_single(soc, _id)				\
> +	{								\
> +		.id = _id,						\
> +		.iobase = soc ## _PINCTRL ## _BASE_ADDR,		\
> +		.irq = soc ## _INT_GPIO ## _id,				\
> +	}
> +
> +#define mxs_gpio_data_entry(soc, _id)					\
> +	[_id] = mxs_gpio_data_entry_single(soc, _id)
> +
> +#ifdef CONFIG_SOC_IMX23
> +const struct mxs_gpio_data mx23_gpio_data[] __initconst = {
> +#define mx23_gpio_data_entry(_id)					\
> +	mxs_gpio_data_entry(MX23, _id)

I know it's tempting to use macros for these, but I think it obscures
the code a lot, especially when you use them to concatenate identifier
names. Why not just do:

	struct platform_device *gpios;
	gpios = platform_device_register_simple(mxs_host_bus, "mxs-gpio-master", 0, NULL, 0);

	mxs_register_gpio(gpios, 0, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_0);
	mxs_register_gpio(gpios, 1, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_1);
	mxs_register_gpio(gpios, 2, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_2);
	mxs_register_gpio(gpios, 3, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_3);

This is actually shorter and it makes it possible to grep for the
macros you use.

> +struct platform_device *__init mxs_add_gpio(
> +		const struct mxs_gpio_data *data)
> +{
> +	struct resource res[] = {
> +		{
> +			.start = data->iobase,
> +			.end = data->iobase + SZ_8K - 1,
> +			.flags = IORESOURCE_MEM,
> +		}, {
> +			.start = data->irq,
> +			.end = data->irq,
> +			.flags = IORESOURCE_IRQ,
> +		},
> +	};
> +
> +	return mxs_add_platform_device("mxs-gpio", data->id,
> +					res, ARRAY_SIZE(res), NULL, 0);
> +}

mxs_add_platform_device doesn't set the parent pointer correctly, I think you
should either fix that or open-code the platform device creation to do it
right.

	Arnd

WARNING: multiple messages have this Message-ID (diff)
From: Arnd Bergmann <arnd@arndb.de>
To: linux-arm-kernel@lists.infradead.org
Cc: Shawn Guo <shawn.guo@linaro.org>,
	linux-kernel@vger.kernel.org, patches@linaro.org,
	linus.walleij@linaro.org, grant.likely@secretlab.ca,
	kernel@pengutronix.de
Subject: Re: [PATCH 2/3] ARM: mxs: add gpio-mxs platform devices
Date: Fri, 20 May 2011 12:23:00 +0200	[thread overview]
Message-ID: <201105201223.00427.arnd@arndb.de> (raw)
In-Reply-To: <1305885446-27404-3-git-send-email-shawn.guo@linaro.org>

On Friday 20 May 2011 11:57:25 Shawn Guo wrote:

> --- /dev/null
> +++ b/arch/arm/mach-mxs/devices/platform-gpio-mxs.c
> @@ -0,0 +1,92 @@
> +/*
> + * Copyright 2011 Freescale Semiconductor, Inc. All Rights Reserved.
> + *
> + * This program is free software; you can redistribute it and/or modify it under
> + * the terms of the GNU General Public License version 2 as published by the
> + * Free Software Foundation.
> + */

Hmm, I forgot to discuss this in Budapest, but I'm still not convinced about
the code ownership here, I think it should be Copyright Linaro Ltd when a
Linaro assignee writes it, not the member company that you work for.

If your manager thinks it should be copyright Freescale, I would suggest
we discuss it on the Linaro private mailing list so we can find a solution
that everyone is happy with. No need to bother the public with this.

> +#include <linux/compiler.h>
> +#include <linux/err.h>
> +#include <linux/init.h>
> +
> +#include <mach/mx23.h>
> +#include <mach/mx28.h>
> +#include <mach/devices-common.h>
> +
> +struct mxs_gpio_data {
> +	int id;
> +	resource_size_t iobase;
> +	resource_size_t iosize;
> +	resource_size_t irq;
> +};

You don't use iosize anywhere.

> +#define mxs_gpio_data_entry_single(soc, _id)				\
> +	{								\
> +		.id = _id,						\
> +		.iobase = soc ## _PINCTRL ## _BASE_ADDR,		\
> +		.irq = soc ## _INT_GPIO ## _id,				\
> +	}
> +
> +#define mxs_gpio_data_entry(soc, _id)					\
> +	[_id] = mxs_gpio_data_entry_single(soc, _id)
> +
> +#ifdef CONFIG_SOC_IMX23
> +const struct mxs_gpio_data mx23_gpio_data[] __initconst = {
> +#define mx23_gpio_data_entry(_id)					\
> +	mxs_gpio_data_entry(MX23, _id)

I know it's tempting to use macros for these, but I think it obscures
the code a lot, especially when you use them to concatenate identifier
names. Why not just do:

	struct platform_device *gpios;
	gpios = platform_device_register_simple(mxs_host_bus, "mxs-gpio-master", 0, NULL, 0);

	mxs_register_gpio(gpios, 0, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_0);
	mxs_register_gpio(gpios, 1, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_1);
	mxs_register_gpio(gpios, 2, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_2);
	mxs_register_gpio(gpios, 3, MX23_PINCTRL_BASE_ADDR, MX23_INT_GPIO_3);

This is actually shorter and it makes it possible to grep for the
macros you use.

> +struct platform_device *__init mxs_add_gpio(
> +		const struct mxs_gpio_data *data)
> +{
> +	struct resource res[] = {
> +		{
> +			.start = data->iobase,
> +			.end = data->iobase + SZ_8K - 1,
> +			.flags = IORESOURCE_MEM,
> +		}, {
> +			.start = data->irq,
> +			.end = data->irq,
> +			.flags = IORESOURCE_IRQ,
> +		},
> +	};
> +
> +	return mxs_add_platform_device("mxs-gpio", data->id,
> +					res, ARRAY_SIZE(res), NULL, 0);
> +}

mxs_add_platform_device doesn't set the parent pointer correctly, I think you
should either fix that or open-code the platform device creation to do it
right.

	Arnd

  reply	other threads:[~2011-05-20 10:23 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-20  9:57 [PATCH 0/3] remove mach-mxs gpio driver Shawn Guo
2011-05-20  9:57 ` Shawn Guo
2011-05-20  9:57 ` [PATCH 1/3] ARM: mxs: remove gpio driver from mach-mxs Shawn Guo
2011-05-20  9:57   ` Shawn Guo
2011-05-20 10:04   ` Arnd Bergmann
2011-05-20 10:04     ` Arnd Bergmann
2011-05-20 10:27     ` Shawn Guo
2011-05-20 10:27       ` Shawn Guo
2011-05-20  9:57 ` [PATCH 2/3] ARM: mxs: add gpio-mxs platform devices Shawn Guo
2011-05-20  9:57   ` Shawn Guo
2011-05-20 10:23   ` Arnd Bergmann [this message]
2011-05-20 10:23     ` Arnd Bergmann
2011-05-20 14:03     ` Shawn Guo
2011-05-20 14:03       ` Shawn Guo
2011-05-20 14:23       ` Arnd Bergmann
2011-05-20 14:23         ` Arnd Bergmann
2011-05-20 14:40         ` Shawn Guo
2011-05-20 14:40           ` Shawn Guo
2011-05-27  9:11         ` Shawn Guo
2011-05-27  9:11           ` Shawn Guo
2011-05-27 11:10           ` Arnd Bergmann
2011-05-27 11:10             ` Arnd Bergmann
2011-05-27  8:29     ` Shawn Guo
2011-05-27  8:29       ` Shawn Guo
2011-05-27  8:48       ` Arnd Bergmann
2011-05-27  8:48         ` Arnd Bergmann
2011-05-20  9:57 ` [PATCH 3/3] ARM: mxs: select GPIO_MXS for i.MX23 and i.MX28 Shawn Guo
2011-05-20  9:57   ` Shawn Guo

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=201105201223.00427.arnd@arndb.de \
    --to=arnd@arndb.de \
    --cc=linux-arm-kernel@lists.infradead.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: link
Be 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.