Re: [PATCH v3] gpio: langwell: convert to use irq_domain

[Mika Westerberg <mika.westerberg@linux.intel.com>]

On Tue, May 08, 2012 at 11:44:34AM -0600, Grant Likely wrote:
> On Wed,  2 May 2012 11:15:50 +0300, Mika Westerberg <mika.westerberg@linux.intel.com> wrote:
> > irq_domain already provides a facility to translate from hardware IRQ
> > numbers to Linux IRQ numbers so use that instead of open-coding the logic
> > in the driver.
> > 
> > Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
> 
> Applied, thanks.

Thanks.

> BTW, I noticed a bug below (unrelated to this patch)...
> 
> > ---
> >  static int lnw_irq_type(struct irq_data *d, unsigned type)
> >  {
> >  	struct lnw_gpio *lnw = irq_data_get_irq_chip_data(d);
> > -	u32 gpio = d->irq - lnw->irq_base;
> > +	u32 gpio = irqd_to_hwirq(d);
> >  	unsigned long flags;
> >  	u32 value;
> >  	void __iomem *grer = gpio_reg(&lnw->chip, gpio, GRER);
> > @@ -256,7 +257,8 @@ static void lnw_irq_handler(unsigned irq, struct irq_desc *desc)
> >  			pending &= ~mask;
> >  			/* Clear before handling so we can't lose an edge */
> >  			writel(mask, gedr);
> > -			generic_handle_irq(lnw->irq_base + base + gpio);
> > +			generic_handle_irq(irq_find_mapping(lnw->domain,
> > +							    base + gpio));
> >  		}
> >  	}
> 
> This while loop is actually buggy.  After handling each pending irq
> bit the status register should be re-read.  Otherwise if the same irq
> gets raised again after it is handled, then this loop will ignore it.
> 
> It's not a major issue, but it would mean unnecessary exiting and
> reentering the handler function.

Indeed. I'll take a look.