[PATCH] i2c: sun6i-p2wi: Fix an error message in probe()

Christophe JAILLET christophe.jaillet at wanadoo.fr
Fri Jun 30 12:58:49 PDT 2023


Le 27/06/2023 à 14:08, Dan Carpenter a écrit :
> On Tue, Jun 27, 2023 at 01:59:20PM +0200, Andi Shyti wrote:
>> Hi Dan,
>>
>> On Tue, Jun 27, 2023 at 10:12:36AM +0300, Dan Carpenter wrote:
>>> The "ret" variable is uninitialized.  It was the "p2wi->rstc" variable
>>> that was intended.  We can also use the %pe string format to print the
>>> error code name instead of just the number.
>>>
>>> Fixes: 75ff8a340a81 ("i2c: sun6i-p2wi: Use devm_clk_get_enabled()")
>>> Signed-off-by: Dan Carpenter <dan.carpenter at linaro.org>
>>> ---
>>>   drivers/i2c/busses/i2c-sun6i-p2wi.c | 3 ++-
>>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/i2c/busses/i2c-sun6i-p2wi.c b/drivers/i2c/busses/i2c-sun6i-p2wi.c
>>> index ad8270cdbd3e..fa6020dced59 100644
>>> --- a/drivers/i2c/busses/i2c-sun6i-p2wi.c
>>> +++ b/drivers/i2c/busses/i2c-sun6i-p2wi.c
>>> @@ -250,7 +250,8 @@ static int p2wi_probe(struct platform_device *pdev)
>>>   
>>>   	p2wi->rstc = devm_reset_control_get_exclusive(dev, NULL);
>>>   	if (IS_ERR(p2wi->rstc)) {
>>> -		dev_err(dev, "failed to retrieve reset controller: %d\n", ret);
>>> +		dev_err(dev, "failed to retrieve reset controller: %pe\n",
>>> +			p2wi->rstc);
>>
>> Yes, good catch! Thanks! But I think we want to print the error
>> value here, so I think it should be:
>>
>> -		dev_err(dev, "failed to retrieve reset controller: %d\n", ret);
>> +		dev_err(dev, "failed to retrieve reset controller: %d\n",
>> +			PTR_ERR(p2wi->rstc));
>>
> 
> The %pe which I changed it to is a cool new thing that prints:
> 
> 	failed to retrieve reset controller: -EINVAL\n
> 
> We should create a similar %e printk format that works for ints instead
> of error pointers.  But instead of that you have people who cast error
> codes to pointers just to get the %pe functionality.  And other people
> who make suggestions (this is the catagory that I'm in) but are too lazy
> to do the actual work.

Hi Dan,

why not switch to dev_err_probe()?

CJ

> 
> regards,
> dan carpenter
> 
> 




More information about the linux-arm-kernel mailing list