Re: [PATCH 1/3] iio: frequency: Remove 'out of memory' message

[Lars-Peter Clausen <lars@metafoo.de>]

On 12/23/2014 09:21 PM, Hartmut Knaack wrote:
> Lars-Peter Clausen schrieb am 23.12.2014 um 10:07:
>> On 12/23/2014 09:51 AM, Daniel Baluta wrote:
>>> On Tue, Dec 16, 2014 at 11:11 AM, Roberta Dobrescu
>>> <roberta.dobrescu@gmail.com> wrote:
>>>> This patch fixes the following checkpatch.pl warning:
>>>> WARNING: Possible unnecessary 'out of memory' message
>>>>
>>>> Signed-off-by: Roberta Dobrescu <roberta.dobrescu@gmail.com>
>>>> ---
>>>>    drivers/iio/frequency/adf4350.c | 4 +---
>>>>    1 file changed, 1 insertion(+), 3 deletions(-)
>>>>
>>>> diff --git a/drivers/iio/frequency/adf4350.c b/drivers/iio/frequency/adf4350.c
>>>> index 63a25d9..2b301eb 100644
>>>> --- a/drivers/iio/frequency/adf4350.c
>>>> +++ b/drivers/iio/frequency/adf4350.c
>>>> @@ -387,10 +387,8 @@ static struct adf4350_platform_data *adf4350_parse_dt(struct device *dev)
>>>>           int ret;
>>>>
>>>>           pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
>>>> -       if (!pdata) {
>>>> -               dev_err(dev, "could not allocate memory for platform data\n");
>>>> +       if (!pdata)
>>>>                   return NULL;
>>>> -       }
>>>>
>>> Based on Harmut's review we can return ERR_PTR(-ENOMEM) here
>>> and then use IS_ERR and PTR_ERR inside adf4350_probe()
>>
>> This patch is fine as it is. It removes the redundant error message. Every
>> thing else should be done in a second patch.
>
> I can't follow on this one. From what I see, if memory allocation fails here,
> we will get an appropriate error message here, and _probe will pass up -EINVAL,
> causing a misleading error message. Where do you see redundancy?

Unless you pass __GFP_NOWARN to k*alloc it will print a error message on its 
own if the allocation fails.

- Lars