[RFC PATCH] of/irq: do irq resolution in platform_get_irq_byname()

[grygorii.strashko@ti.com (Grygorii Strashko)]

Hi Grant,

On 05/20/2014 09:17 AM, Grant Likely wrote:
> On Mon, 19 May 2014 14:57:39 +0200, Thierry Reding <thierry.reding@gmail.com> wrote:
>> On Mon, May 19, 2014 at 04:30:59PM +0300, Grygorii Strashko wrote:
>> [...]
>>> diff --git a/drivers/of/irq.c b/drivers/of/irq.c
>> [...]
>>>   /**
>>> + * of_irq_get_byname - Decode a node's IRQ and return it as a Linux irq number
>>> + * @dev: pointer to device tree node
>>> + * @name: zero-based index of the irq
>>
>> This is a name, not an index.
>>
>>> + *
>>> + * Returns Linux irq number on success, or -EPROBE_DEFER if the irq domain
>>> + * is not yet created, or errorno in case of failure.
>>
>> s/errorno/error code/? Also EPROBE_DEFER is also an error code, so I'm
>> not sure if it's worth a special case in the description here.
>>
>>> + *
>>> + */
>>> +int of_irq_get_byname(struct device_node *dev, const char *name)
>>> +{
>>> +	const char *name_irq = NULL;
>>> +	int index = 0;
>>> +
>>> +	if (unlikely(!name))
>>> +		return -EINVAL;
>>> +
>>> +	while (!of_property_read_string_index(dev, "interrupt-names",
>>> +					      index, &name_irq))
>>> +		if (!strcmp(name, name_irq))
>>> +			return of_irq_get(dev, index);
>>
>> Isn't this missing an index++ somewhere? Otherwise it seems like this
>> would loop infinitely if there was no match on the first entry.
>
> Better yet, use of_property_match_string().

yep. Thanks. I've just come to the same idea. Patch re-sent.

Regards,
-grygorii