Re: i2c-s3c2410: fix check for being in suspend.

[Jean Delvare <khali-PUYAD+kWke1g9hUCZPvPmw@public.gmane.org>]

Hi Ben,

Moving to the i2c list...

On Fri, 31 Oct 2008 12:09:14 +0000, Ben Dooks wrote:
> As noted by Julia Lawall <julia-dAYI7NvHqcQ@public.gmane.org>, we can never
> trigger the check for being in suspend due to the result
> of !readl(i2c->regs + S3C2410_IICCON) & S3C2410_IICCON_IRQEN
> always being 0.
> 
> Add suspend/resume hooks to stop i2c transactions happening
> until the driver has been resumed.
> 
> Signed-off-by: Ben Dooks <ben-linux-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org>
> 
> Index: linux.git/drivers/i2c/busses/i2c-s3c2410.c
> ===================================================================
> --- linux.git.orig/drivers/i2c/busses/i2c-s3c2410.c	2008-10-31 11:51:55.000000000 +0000
> +++ linux.git/drivers/i2c/busses/i2c-s3c2410.c	2008-10-31 12:06:01.000000000 +0000
> @@ -56,6 +56,7 @@ enum s3c24xx_i2c_state {
>  struct s3c24xx_i2c {
>  	spinlock_t		lock;
>  	wait_queue_head_t	wait;
> +	unsigned int		suspended:1;
>  
>  	struct i2c_msg		*msg;
>  	unsigned int		msg_num;
> @@ -507,7 +508,7 @@ static int s3c24xx_i2c_doxfer(struct s3c
>  	unsigned long timeout;
>  	int ret;
>  
> -	if (!readl(i2c->regs + S3C2410_IICCON) & S3C2410_IICCON_IRQEN)
> +	if (i2c->suspended)
>  		return -EIO;
>  
>  	ret = s3c24xx_i2c_set_master(i2c);

This hunk will not apply, ad Julia's fix is already upstream. You'll
have to update your patch a bit.

> @@ -986,17 +987,26 @@ static int s3c24xx_i2c_remove(struct pla
>  }
>  
>  #ifdef CONFIG_PM
> +static int s3c24xx_i2c_suspend_late(struct platform_device *dev,
> +				    pm_message_t msg)
> +{
> +	struct s3c24xx_i2c *i2c = platform_get_drvdata(dev);
> +	i2c->suspended = 1;
> +	return 0;
> +}
> +
>  static int s3c24xx_i2c_resume(struct platform_device *dev)
>  {
>  	struct s3c24xx_i2c *i2c = platform_get_drvdata(dev);
>  
> -	if (i2c != NULL)
> -		s3c24xx_i2c_init(i2c);
> +	i2c->suspended = 0;
> +	s3c24xx_i2c_init(i2c);
>  
>  	return 0;
>  }
>  
>  #else
> +#define s3c24xx_i2c_suspend_late NULL
>  #define s3c24xx_i2c_resume NULL
>  #endif
>  
> @@ -1005,6 +1015,7 @@ static int s3c24xx_i2c_resume(struct pla
>  static struct platform_driver s3c2410_i2c_driver = {
>  	.probe		= s3c24xx_i2c_probe,
>  	.remove		= s3c24xx_i2c_remove,
> +	.suspend_late	= s3c24xx_i2c_suspend_late,
>  	.resume		= s3c24xx_i2c_resume,
>  	.driver		= {
>  		.owner	= THIS_MODULE,
> @@ -1015,6 +1026,7 @@ static struct platform_driver s3c2410_i2
>  static struct platform_driver s3c2440_i2c_driver = {
>  	.probe		= s3c24xx_i2c_probe,
>  	.remove		= s3c24xx_i2c_remove,
> +	.suspend_late	= s3c24xx_i2c_suspend_late,
>  	.resume		= s3c24xx_i2c_resume,
>  	.driver		= {
>  		.owner	= THIS_MODULE,
> 

Will you take care of pushing this upstream, or should I?

Thanks,
-- 
Jean Delvare