On Thu, Aug 28, 2014 at 12:44:18AM +0200, Javier Martinez Canillas wrote: > On Wed, Aug 27, 2014 at 11:03 PM, Tomasz Figa <tomasz.figa@xxxxxxxxx> wrote: > This is the case for Chromebooks as well but the solution implemented > in the downstream Chrome OS 3.8 kernel is what Tomasz suggested *sigh* This is not what you were describing, though most of what you then go on to say doesn't correspond to the issue Tomasz was describing. > Yes, AFAIK the bootloader (none of them because Chromebooks use two > chained U-boots) change the regulators default opmode so once is set > to OFF on .disable, that value is preserved on warm reboot. This made > sense with the Chrome OS kernel since the kernel always set the opmode > defined in the "regulator-op-mode" DT property and did not relied on > the bootloader to set the most efficient default opmode. I'm sorry but this is just really unclear. Does the bootloader change the mode or does it not change the mode? You start off saying that the bootloader does change the mode, then go on to say that it doesn't and that the system relies on Linux changing the mode (which is consistent with what you were originally saying and what you say for much of the rest of this mail but not with what Tomasz was talking about). > Also, other drivers have customs operating mode DT properties, please > take a look at Documentation/devicetree/bindings/regulator/s5m8767-regulator.txt: Per driver things can potentially be fine, trying to define generic properties is much harder.
Attachment:
signature.asc
Description: Digital signature
