Re: Extcon: support multiple states at a device.

[Chanwoo Choi <cw00.choi@samsung.com>]

Hi Dan,

Thanks for your report. I'll fix it.

Thanks,
Chanwoo Choi

On 05/22/2015 04:28 PM, Dan Carpenter wrote:
> Hello MyungJoo Ham,
> 
> The patch 7eaf7186842c: ('extcon: Use the unique id for external
> connector instead of string') from Apr 24, 2015, leads to the following
> static checker warning:
> 
> 	drivers/extcon/extcon.c:356 extcon_get_cable_state_()
> 	warn: we tested 'index < 0' before and it was 'false'
> 
> drivers/extcon/extcon.c
>    343  /**
>    344   * extcon_get_cable_state_() - Get the status of a specific cable.
>    345   * @edev:       the extcon device that has the cable.
>    346   * @id:         the unique id of each external connector in extcon enumeration.
>    347   */
>    348  int extcon_get_cable_state_(struct extcon_dev *edev, const enum extcon id)
>    349  {
>    350          int index;
>    351  
>    352          index = find_cable_index_by_id(edev, id);
>    353          if (index < 0)
>                     ^^^^^^^^^
> Return error code here.
> 
>    354                  return index;
>    355  
>    356          if (index < 0 || (edev->max_supported && edev->max_supported <= index))
>                     ^^^^^^^^^
> No need to check here.
> 
>    357                  return -EINVAL;
>    358  
>    359          return !!(edev->state & (1 << index));
>    360  }
> 
> But the other issue with this code is that the callers sometimes treat
> errors as found, for example in extcon_get_cable_state().
> 
> regards,
> dan carpenter
>