Re: your mail

Alan Cox (alan@lxorguk.ukuu.org.uk)
Wed, 13 Mar 2002 22:35:07 +0000 (GMT)


> +/*
> + * Deal with CONFIG_MODVERSIONS
> + */
> +#if 0 /* Pb with MODVERSIONS */
> +#if CONFIG_MODVERSIONS==1
> +#define MODVERSIONS
> +#include <linux/modversions.h>
> +#endif
> +#endif

[modversions.h is magically included by the kernel for you when its in
kernel if you haven't worked that one out yet]

> +#define PP_NO 3
> +struct tipar_struct table[PP_NO];
static ?

> + for(i=0; i < delay; i++) {
> + inbyte(minor);
> + }
> + schedule();

Oh random tip

if(current->need_resched)
schedule();

will just give up the CPU when you are out of time

> + if(table[minor].opened)
> + return -EBUSY;
> + table[minor].opened++;

Think about open/close at the same moment or SMP - the watchdog drivers all
had this problem and now do

unsigned long opened = 0;

if(test_and_set_bit(0, &opened))
return -EBUSY;

clear_bit(0, &opened)

[this generates atomic operations so is always safe]

> + if(!table[minor].opened)
> + return -EFAULT;

BUG() may be better - it can't happen so BUG() will get a backtrace
and actually get it reported 8)

> +static long long tipar_lseek(struct file * file, long long offset, int origin)
> +{
> + return -ESPIPE;
> +}

Can go (you now use no_llseek)

Basically except for the open/close one I'm now just picking holes.
For the device major/minors see http://www.lanana.org.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/