Re: Patch of a new driver for kernel 2.4.x that need review

["Pekka J Enberg" <penberg@cs.helsinki.fi>]

Mark Gross writes:
> +static int
> +tlclk_open(struct inode *inode, struct file *filp)
> +{
> +	int result;
> +#ifdef MODULE
> +	if (!MOD_IN_USE) {
> +		MOD_INC_USE_COUNT;
> +#endif
> +		/* Make sure there is no interrupt pending will 
> +		   *  initialising interrupt handler */
> +		inb(TLCLK_REG6);
> +
> +		result = request_irq(telclk_interrupt, &tlclk_interrupt,
> +					SA_SHIRQ, "telclock", tlclk_interrupt);

Instead of playing the MOD_IN_USE games, please either (1) grab the irq in 
module init (it's a shared IRQ after all) or (2) do the following: 

static int tlclk_used; 

static int tlclk_open(struct inode *inode, struct file *filp) {
       if (tlclk_used++)
               return 0; 

       // request_irq goes here.
} 

For an example, see the file drivers/input/mouse/amimouse.c (appears in 
2.6.12 at least). 

                        Pekka