Re: [PATCH 01/05] sh: Add GPIO and pinmux base code

[Paul Mundt <lethal@linux-sh.org>]

On Sun, Sep 28, 2008 at 03:10:16AM +0900, Magnus Damm wrote:
> From: Magnus Damm <damm@igel.co.jp>
> 
> This patch adds the pinmux table parser and gpiolib glue.
> The old SH3 header code is removed.
> 
> Signed-off-by: Magnus Damm <damm@igel.co.jp>

For starters, I don't know why you decided to kill off the SH-3 gpio
header. It's still in active use by the magicpanelr2 board. Once 7720 and
mpr2 are converted over, then we can kill off the header, but not before
that. Although I suppose we can just fix up the board header to use
cpu-sh3/cpu/gpio.h explicitly.

> --- 0001/arch/sh/include/asm/gpio.h
> +++ work/arch/sh/include/asm/gpio.h	2008-09-28 02:03:40.000000000 +0900
> @@ -1,19 +1,119 @@
>  /*
> - *  include/asm-sh/gpio.h
> + * Generic GPIO API using gpiolib and pinmux table support for SuperH.
>   *
> - *  Copyright (C) 2007 Markus Brunner, Mark Jonas
> + * Copyright (c) 2008 Magnus Damm
>   *
> - *  Addresses for the Pin Function Controller
> + * Generic GPIO derived from x86 version:
> + * Copyright (c) 2007-2008  MontaVista Software, Inc.
> + * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
>   *
> - * This file is subject to the terms and conditions of the GNU General Public
> - * License.  See the file "COPYING" in the main directory of this archive
> - * for more details.
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
>   */
>  #ifndef __ASM_SH_GPIO_H
>  #define __ASM_SH_GPIO_H
>  
I wouldn't bother with the header comments at all, there's nothing
profound here and it's almost all pinmux stuff anyways.

> +static int read_write_reg(unsigned long reg, unsigned long reg_width,
> +			  unsigned long field_width, unsigned long in_pos,
> +			  unsigned long value, int do_write)
> +{
> +	unsigned long flags, data, mask, pos;
> +
> +	flags = 0; /* kill silly warning */
> +	data = 0;
> +	mask = (1 << field_width) - 1;
> +	pos = reg_width - ((in_pos + 1) * field_width);
> +
> +#ifdef DEBUG
> +	pr_info("sh pinmux: %s, addr = %lx, value = %ld, pos = %ld, "
> +		"r_width = %ld, f_width = %ld\n",
> +		do_write ? "write" : "read", reg, value, pos,
> +		reg_width, field_width);
> +#endif
> +
> +	if (do_write)
> +		spin_lock_irqsave(&gpio_register_lock, flags);
> +
The locking here is a bit non-obvious, and should be commented. I had to
read through this a couple of times to figure out what it was supposed to
be doing. You seem to be trying to use a spinlock as a rwlock, which is a
bit error prone to say the least.