On Friday 11 August 2006 22:56, Dave Jones wrote:
> On Thu, Aug 10, 2006 at 11:19:13PM +0200, [email protected] wrote:
> > This is a driver for the on-chip watchdog device found on some
> > MIPS RM9000 processors.
> >
> > Signed-off-by: Thomas Koeller <[email protected]>
>
> Mostly same nit-picking comments as your other driver..
Which one?
>
> > +++ b/drivers/char/watchdog/rm9k_wdt.c
> > ...
> > +
> > +#include <linux/config.h>
>
> not needed.
It is, otherwise I do not get CONFIG_WATCHDOG_NOWAYOUT.
>
> > +/* Function prototypes */
> > +static int __init wdt_gpi_probe(struct device *);
> > +static int __exit wdt_gpi_remove(struct device *);
> > +static void wdt_gpi_set_timeout(unsigned int);
> > +static int wdt_gpi_open(struct inode *, struct file *);
> > +static int wdt_gpi_release(struct inode *, struct file *);
> > +static ssize_t wdt_gpi_write(struct file *, const char __user *,
> > size_t, loff_t *);
> > +static long wdt_gpi_ioctl(struct file *, unsigned int, unsigned long);
> > +static const struct resource *wdt_gpi_get_resource(struct
> > platform_device *, const char *, unsigned int);
> > +static int wdt_gpi_notify(struct notifier_block *, unsigned long, void
> > *); +static irqreturn_t wdt_gpi_irqhdl(int, void *, struct pt_regs *);
>
> Can probably (mostly?) go away with some creative reordering.
Probably, but should it? I always considered it good style to have
prototypes for all functions.
>
> > +static int locked = 0;
>
> unneeded initialisation.
Not strictly needed, that's true, but does not do any harm either
and expresses the intention clearly.
>
> > +static int nowayout =
> > +#if defined(CONFIG_WATCHDOG_NOWAYOUT)
> > + 1;
> > +#else
> > + 0;
> > +#endif
>
> static int nowayout = CONFIG_WATCHDOG_NOWAYOUT;
>
> should work.
Does not work. If the option is not selected, CONFIG_WATCHDOG_NOWAYOUT
is undefined, not zero.
>
> > +static void wdt_gpi_set_timeout(unsigned int to)
> > +{
> > + u32 reg;
> > + const u32 wdval = (to * CLOCK) & ~0x0000000f;
> > +
> > + lock_titan_regs();
> > + reg = titan_readl(CPCCR) & ~(0xf << (wd_ctr * 4));
> > + titan_writel(reg, CPCCR);
> > + wmb();
> > + __raw_writel(wdval, wd_regs + 0x0000);
> > + wmb();
> > + titan_writel(reg | (0x2 << (wd_ctr * 4)), CPCCR);
> > + wmb();
> > + titan_writel(reg | (0x5 << (wd_ctr * 4)), CPCCR);
> > + iob();
> > + unlock_titan_regs();
> > +}
>
> As in the previous driver, are these barriers strong enough?
> Or do they need explicit reads of the written addresses to flush the write?
I think they are. Remember, the entire device is integrated in the
processor. No external buses involved.
>
> Dave
Thanks for your comments!
Thomas
--
Thomas Koeller
[email protected]
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
