On Thu, Oct 21, 2021 at 07:38:08PM +0100, Mark Brown wrote:
On Thu, Oct 21, 2021 at 12:36:27PM +0200, Uwe Kleine-König wrote:
a bit unsure if regulator_bulk_force_disable() is indeed the right function here, its documentation specifies a different usecase.
My motivation for this change was to make it obvious in cs42l51_i2c_remove() that there is no error to handle to eventually make i2c remove callbacks return void, too.
It would be better to just ignore the errors on disable.
Yeah, and part of this is if regulator #1 fails to disable still disable regulators #2 and #3. (I don't know how many there are.) That was why I picked regulator_bulk_force_disable() which has this behaviour.
Realistically you'd have to really be trying to trigger an error here and it's most likely that the system is in enough trouble if one is triggered that it's just not worrying about. I'm not sure how likely it is that anyone would ever remove one of these devices in production though.
So compared to my patch you'd just drop the warning?!
Best regards Uwe