Hi Chris,
On Fri, Jan 27, 2017 at 4:05 AM, Chris Brandt <Chris.Brandt@xxxxxxxxxxx> wrote:
> On Wednesday, January 25, 2017, Geert Uytterhoeven wrote:
>> >> I think you can handle that in drivers/clk/renesas/clk-mstp.c:
>> >> - in cpg_mstp_attach_dev(), add a call to pm_clk_resume(dev) after
>> the
>> >> call to pm_clk_add_clk(),
>> >> - in cpg_mstp_detach_dev(), add a call to pm_clk_suspend(dev) before
>> the
>> >> call to pm_clk_destroy().
>> >> Yes, that means the module clocks are enabled all the time.
>> >> Of course when running on RZ/A1H ;-)
>> >
>> > That might be OK.
>>
>> Forgot to mention: you should also no longer set GENPD_FLAG_PM_CLK in
>> cpg_mstp_add_clk_domain(). The flag won't hurt, it will just cause extra
>> code to be executed.
>
> So to be clear before I start hacking away, your suggestion here is
> to do this ONLY for RZ/A1 so I don't screw up any other SoCs, right?
Correct.
> For example:
>
> int cpg_mstp_attach_dev()
> {
> ...
>
> error = pm_clk_add_clk(dev, clk);
> if (error) {
> dev_err(dev, "pm_clk_add_clk %pC failed %d\n", clk, error);
> goto fail_destroy;
> }
>
> + if (of_device_is_compatible(np, "renesas,r7s72100-mstp-clocks"))
> + pm_clk_suspend(dev);
That should be pm_clk_resume(dev), as is this the attach function ;-)
And please drop GENPD_FLAG_PM_CLK on RZ/A1, too.
> ...
> }
>
> Then, this would be OK to submit until I find some other way to do stable
> runtime pm for RZ/A1? (which might not be possible...unless I put in the
> silly delay)
>
> A that point, any peripheral I use (status="okay") will stay on, but the
> ones I don't use will stay off, correct?
Correct.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
