Re: [PATCH] watchdog/coh901327: convert to use watchdog core

[Wolfram Sang <w.sang@pengutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 1513 bytes --]

Hi Linus,

> -static long coh901327_ioctl(struct file *file, unsigned int cmd,
> +static long coh901327_ioctl(struct watchdog_device *wdt_dev, unsigned int cmd,
>  			    unsigned long arg)
>  {
>  	int ret = -ENOTTY;
>  	u16 val;
> -	int time;
> -	int new_options;
>  	union {
>  		struct watchdog_info __user *ident;
>  		int __user *i;
>  	} uarg;

'uarg' is nice name for the union ;) You probably don't need that anymore.
If we keep the ioctl, that is, because...

...

>  	case WDIOC_GETTIMELEFT:
>  		clk_enable(clk);
>  		/* Read repeatedly until the value is stable! */

I was wondering if it pays to put this IOCTL to watchdog_dev and add another
callback to watchdog_ops? I'd think so. Wim, Linus, what do you think? We don't
save much (yet?), but since this is part of the kernel-API it might be nice to
have the common part centralized, even if it is mainly put_user (which might be
nice, too, since most users of GETTIMELEFT cast the result to various types).

> +static struct watchdog_device coh901327_wdt = {
> +	.info = &coh901327_ident,
> +	.ops = &coh901327_ops,
> +	/*
> +	 * Max margin is 327 since the 10ms
> +	 * timeout register is max
> +	 * 0x7FFF = 327670ms ~= 327s.
> +	 */

I'd drop that comment, but well...

> +	.min_timeout = 0,
> +	.max_timeout = 327,

Thanks!

   Wolfram

-- 
Pengutronix e.K.                           | Wolfram Sang                |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]