[PATCH] regulator: pwm: Fix regulator ramp delay for continuous mode

[Boris Brezillon]

+Thierry

Hi Doug,

On Fri, 1 Jul 2016 13:17:48 -0700
Doug Anderson <dianders at chromium.org> wrote:

> Mark,
> 
> On Fri, Jul 1, 2016 at 1:06 AM, Mark Brown <broonie at kernel.org> wrote:
> > On Mon, Jun 27, 2016 at 09:53:11PM -0700, Douglas Anderson wrote:
> >  
> >> Note that this patch is atop Boris's recent PWM regulator fixes.  If
> >> desired it wouldn't be too hard to write it atop the old code, though
> >> quite honestly anyone using a PWM regulator should probably be using his
> >> new code.  
> >
> > Given that those don't seem likely to go in any time soon and aren't
> > going to go in as a fix can you please respin against current code?  
> 
> I went to write this patch and it was pretty easy to write.  ...but
> then I realized that I had no real way to test it.  I can compile test
> it, but that's it.  Do you still want it?
> 
> Specifically without Boris's patches then I don't believe the devices
> I have access to are bootable when I tell Linux about the PWM
> regulator, especially when I configure the PWM regulator to use
> continuous mode (which is where the bug is).  As I understand from
> <https://patchwork.kernel.org/patch/9175279/> the current PWM
> regulator assumes that 0% duty cycle is the lowest voltage and 100% is
> the highest voltage.  My system is backwards (0% is highest, 100%
> lowest).  Also: last I tried even the non-continuous mode without
> Boris's patches the regulator would glitch enough at bootup that my
> system would crash (the PWM regulator affects the memory controller).
> 
> Anyway, let me know if you want me to post the untested, rebased
> patch...  Personally I'd prefer to wait for Boris's patches and the
> land the tested version.
> 
> Boris: any idea for what the next steps on your series are?

I'm waiting for reviews or acks. AFAIR, I addressed all the comments
Thierry made on my v2, so I guess we're good on the PWM side, but that
would be good to have a review from Mark now that we agreed on the PWM
APIs.

That shouldn't prevent you from rebasing this patch on Linus' tree and
tag it for stable (I'll rebase my series on top of the regulator tree if
needed). 

Regards,

Boris