RE: [PATCH] gpio: Fix test on unsigned in lnw_irq_type()

[Du, Alek]

>-----Original Message-----
>From: Roel Kluin [mailto:roel.kluin@xxxxxxxxx]
>Sent: Thursday, October 15, 2009 1:17 AM
>To: Du, Alek; Andrew Morton; LKML
>Subject: Re: [PATCH] gpio: Fix test on unsigned in lnw_irq_type()
>
>The wrong test was used, gpio is unsigned. Also lnw->chip.ngpio is
>set to 64, so if gpio equals that, then reg (= gpio / 32) becomes 2,
>an index out of bounds for GRER and GFER that have 2 elements.
>
>Signed-off-by: Roel Kluin <roel.kluin@xxxxxxxxx>
>---
>>>>> From: Andrew Morton [mailto:akpm@xxxxxxxxxxxxxxxxxxxx]
>>>>>> -    if (gpio < 0 || gpio > lnw->chip.ngpio)
>>>>>> +    if (gpio > lnw->chip.ngpio)
>>>>>>              return -EINVAL;
>>>>>
>>>>> Should that be >= ?
>
>> Oh, my bad. Andrew is right, should be >=...
>
>Ok, how about this then?
>
>diff --git a/drivers/gpio/langwell_gpio.c b/drivers/gpio/langwell_gpio.c
>index 5711ce5..72af3fc 100644
>--- a/drivers/gpio/langwell_gpio.c
>+++ b/drivers/gpio/langwell_gpio.c
>@@ -123,7 +123,7 @@ static int lnw_irq_type(unsigned irq, unsigned type)
> 	void __iomem *grer = (void __iomem *)(&lnw->reg_base->GRER[reg]);
> 	void __iomem *gfer = (void __iomem *)(&lnw->reg_base->GFER[reg]);
>
>-	if (gpio < 0 || gpio > lnw->chip.ngpio)
>+	if (gpio >= lnw->chip.ngpio)
> 		return -EINVAL;
> 	spin_lock_irqsave(&lnw->lock, flags);
> 	if (type & IRQ_TYPE_EDGE_RISING)

Acked-by: Alek Du <alek.du@xxxxxxxxx>



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