Re: [PATCH v2 1/3] serial: mctrl_gpio: Avoid probe failures in case of missing gpiolib

[Schrempf Frieder]

On 01.08.19 22:33, Uwe Kleine-KÃnig wrote:
> On Thu, Aug 01, 2019 at 06:45:21PM +0000, Schrempf Frieder wrote:
>> From: Frieder Schrempf <frieder.schrempf@xxxxxxxxxx>
>>
>> If CONFIG_GPIOLIB is not enabled, mctrl_gpio_init() and
>> mctrl_gpio_init_noauto() will currently return an error pointer with
>> -ENOSYS. As the mctrl GPIOs are usually optional, drivers need to
>> check for this condition to allow continue probing.
>>
>> To avoid the need for this check in each driver, we return NULL
>> instead, as all the mctrl_gpio_*() functions are skipped anyway.
>> We also adapt mctrl_gpio_to_gpiod() to be in line with this change.
>>
>> Signed-off-by: Frieder Schrempf <frieder.schrempf@xxxxxxxxxx>
>> Reviewed-by: Fabio Estevam <festevam@xxxxxxxxx>
> 
> nitpick: put your S-o-b last.

Ok.

>> ---
>> Changes in v2
>> =============
>> * Move the sh_sci changes to a separate patch
>> * Add a patch for the 8250 driver
>> * Add Fabio's R-b tag
>> ---
>>   drivers/tty/serial/serial_mctrl_gpio.c | 3 +++
>>   drivers/tty/serial/serial_mctrl_gpio.h | 6 +++---
>>   2 files changed, 6 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/tty/serial/serial_mctrl_gpio.c b/drivers/tty/serial/serial_mctrl_gpio.c
>> index 2b400189be91..54c43e02e375 100644
>> --- a/drivers/tty/serial/serial_mctrl_gpio.c
>> +++ b/drivers/tty/serial/serial_mctrl_gpio.c
>> @@ -61,6 +61,9 @@ EXPORT_SYMBOL_GPL(mctrl_gpio_set);
>>   struct gpio_desc *mctrl_gpio_to_gpiod(struct mctrl_gpios *gpios,
>>   				      enum mctrl_gpio_idx gidx)
>>   {
>> +	if (gpios == NULL)
>> +		return NULL;
>> +
> 
> I wonder why you need this. If GPIOLIB is off this code isn't active and
> with GPIOLIB calling mctrl_gpio_to_gpiod with a gpios == NULL is a bug
> that IMHO should not be silently ignored.
> 
> Am I missing something (again)?

No, you're right. My thoughts were, that if the mctrl_gpio functions are 
allowed to be passed a NULL pointer in general, they all should have a 
NULL check, even if in the current context (GPIOLIB disabled) this code 
is not active. Apparently there are other cases when a NULL pointer is 
passed, see [1]. So you can't really consider gpios == NULL to be a bug 
as this seems to be allowed in general.

[1] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit?id=434be0ae7aa746f481c3a22139e472dbc3f4f817

> 
>>   	return gpios->gpio[gidx];
>>   }
> 
> Best regards
> Uwe
>