On Mon, Apr 18, 2016 at 12:36:52PM +0200, Krzysztof Kozlowski wrote: > On 04/18/2016 11:57 AM, Mark Brown wrote: > > That's not the bug, the bug is that the driver is providing the > > operation without providing a non-zero value for it. The device tree > > configuration is to override what the driver is doing. > So your recommendation would be to set the ramp_delay value in > regulator_desc structure? Or at least fix the changelog to describe the problem more clearly. > The problem is that value of 30 mv/us in datasheet looks untrustworthy > (copied from other pages) and vendor kernel (which should be a > reference) uses 12 mv/us. Anyway setting any value greater than 0 seems > like better idea than just sticking to 0... Yup. Probably the more conservative value is going to be safer.
Attachment:
signature.asc
Description: PGP signature
