Re: [PATCH 0/2] Fix ir-rx51 by using PWM pdata

[Tony Lindgren <tony@atomide.com>]

* Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> [160430 10:53]:
> Hi,
> 
> On 29.04.2016 00:27, Tony Lindgren wrote:
> >* Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> [160428 14:21]:
> >>
> >>I didn't test legacy boot, as I don't really see any value of doing it now
> >>the end of the legacy boot is near, the driver does not function correctly,
> >>however the patchset at least allows for the driver to be build and we have
> >>something to improve on. And I am going to send a patch that fixes the
> >>problem with omap_dm_timer_request_specific(). So, for both patches, you may
> >>add:
> >>
> >>Tested-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
> >
> >OK thanks.
> >
> >Mauro, do the driver changes look OK to you?
> >
> >If so, I could queue the driver too for v4.7 because of the
> >dependency with your ack. Or I can provide you an immutable
> >branch with just the pdata changes against v4.6-rc1 if you
> >prefer that.
> >

Mauro, I've applied the pdata patch only into my omap-for-v4.7/legacy branch.
Feel free to merge immutable commit against v4.6-rc1 which is 8453c5cafd32
("ARM: OMAP2+: Add more functions to pwm pdata for ir-rx51") to your tree
also for v4.7 if you decide to apply the ir-rx51 driver changes. Else we need
to wait for v4.7-rc1 :)

> In the meanwhile I was able to make the driver functional (on top of the
> $subject series) - for that purpose I had to fix dmtimer.c - it turns out
> that PM runtime get()/put() is called in almost every function exported by
> dmtimer, which in turn slows down IR transmission to 4-5s instead of 0.5s. I
> also replaced GPT9 dmtimer with PWM framework API (pwm-omap-dmtimer needs a
> patch) and implemented some DT support.
> 
> Now, how shall I proceed with those - wait for the $subject series to be
> accepted or post the patches now?

Best to do incremental patches on what was posted to avoid confusion.

> Tony, I was unable to find the tree on kernel.org your patches are in. Which
> tree to use to base my patches on?

The pdata changes are in Linux next with omap-for-v4.7/legacy, the driver
changes I have not committed into any upstream going tree. Seems like
Mauro can take it after merging in commit 8453c5cafd32.

Regards,

Tony