Re: [PATCH] driver-core: platform: Avoid to return IRQ 0 in platform_get_irq()

[Arvind Yadav]

Hi Dmitry,


On Tuesday 05 December 2017 11:50 PM, Dmitry Torokhov wrote:
> On Tue, Dec 05, 2017 at 11:41:52PM +0530, Arvind Yadav wrote:
> > Function platform_get_irq() can return 0. Which means NO_IRQ.
> > So this change will not allow to return 0.
> >
> > This change is help to use platform_get_irq() without this,
> >
> > 	val = platform_get_irq();
> > 	if (val <= 0)
> > 		ret = val ? val : -ENODEV;
> >
> > Signed-off-by: Arvind Yadav <arvind.yadav.cs@xxxxxxxxx>
> You need to audit the drivers and make sure you are not breaking them
> with this. I believe at least i2c designware controller on certain
> boards has DEFINE_RES_IRQ(0) which will be broken by this patch.
Yes, you are right. Few driver is using  DEFINE_RES_IRQ(0).
Example:
drivers/mfd/intel-lpss.c
static const struct resource intel_lpss_dev_resources[] = {
        DEFINE_RES_MEM_NAMED(LPSS_DEV_OFFSET, LPSS_DEV_SIZE, "lpss_dev"),
        DEFINE_RES_MEM_NAMED(LPSS_PRIV_OFFSET, LPSS_PRIV_SIZE, 
"lpss_priv"),
        DEFINE_RES_IRQ(0),
};
There's a bunch of platforms in the kernel that still use IRQ0. They can 
suffer because
of this change. I should have mark this patch for testing.
> > ---
> >   drivers/base/platform.c | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/base/platform.c b/drivers/base/platform.c
> > index c203fb9..7b3079c 100644
> > --- a/drivers/base/platform.c
> > +++ b/drivers/base/platform.c
> > @@ -127,7 +127,7 @@ int platform_get_irq(struct platform_device *dev, unsigned int num)
> >   		irqd_set_trigger_type(irqd, r->flags & IORESOURCE_BITS);
> >   	}
> >   
> > -	return r ? r->start : -ENXIO;
> > +	return r && r->start ? r->start : -ENXIO;
> >   #endif
> >   }
> >   EXPORT_SYMBOL_GPL(platform_get_irq);
> > --
> > 2.7.4
> Thanks.
Thanks,