Hi Grygorii,
On Mon, Sep 29, 2014 at 4:38 PM, Grygorii Strashko
<grygorii.strashko@xxxxxx> wrote:
> This patch switches Keystone 2 PM code to use Generic PM domains
> instead of PM clock domains because of the lack of DT support
> for the last.
Thanks, this looks interesting, as today I've been digging deeper into the
!CONFIG_PM_RUNTIME case for my patch series, and ended up with something
similar...
> --- a/arch/arm/mach-keystone/pm_domain.c
> +++ b/arch/arm/mach-keystone/pm_domain.c
> @@ -12,69 +12,110 @@
> * version 2, as published by the Free Software Foundation.
> */
>
> +#include <linux/clk.h>
> #include <linux/init.h>
> -#include <linux/pm_runtime.h>
> #include <linux/pm_clock.h>
> +#include <linux/pm_domain.h>
> #include <linux/platform_device.h>
> -#include <linux/clk-provider.h>
> #include <linux/of.h>
>
> -#ifdef CONFIG_PM_RUNTIME
> -static int keystone_pm_runtime_suspend(struct device *dev)
> +#ifdef CONFIG_PM_GENERIC_DOMAINS
> +
> +struct keystone_domain {
> + struct generic_pm_domain base;
> + struct device *dev;
> +};
> +
> +void keystone_pm_domain_attach_dev(struct device *dev)
> {
> + struct clk *clk;
> int ret;
> + int i = 0;
>
> dev_dbg(dev, "%s\n", __func__);
>
> - ret = pm_generic_runtime_suspend(dev);
> - if (ret)
> - return ret;
> -
> - ret = pm_clk_suspend(dev);
> + ret = pm_clk_create(dev);
> if (ret) {
> - pm_generic_runtime_resume(dev);
> - return ret;
> + dev_err(dev, "pm_clk_create failed %d\n", ret);
> + return;
> + };
> +
> + while ((clk = of_clk_get(dev->of_node, i++)) && !IS_ERR(clk)) {
> + ret = pm_clk_add_clk(dev, clk);
This is an important difference compared to the non-DT pm_clk_notify()
version for !CONFIG_PM_RUNTIME.
You do call pm_clk_create() and pm_clk_add_clk(), while
pm_clk_notify() doesn't. Hence for the latter, the clocklist is always empty.
> + if (ret) {
> + dev_err(dev, "pm_clk_add_clk failed %d\n", ret);
> + goto clk_err;
> + };
> }
>
> - return 0;
> + if (!IS_ENABLED(CONFIG_PM_RUNTIME)) {
> + ret = pm_clk_resume(dev);
As a consequence of the above, calling pm_clk_resume() here just works,
and there's no need for the separate enable_clock()/disable_clock()
functions, like in drivers/base/power/clock_ops.c.
> + if (ret) {
> + dev_err(dev, "pm_clk_resume failed %d\n", ret);
> + goto clk_err;
> + };
> + }
> + return;
> +
> +clk_err:
> + pm_clk_destroy(dev);
> }
>
> -static int keystone_pm_runtime_resume(struct device *dev)
> +void keystone_pm_domain_detach_dev(struct device *dev)
> {
> dev_dbg(dev, "%s\n", __func__);
> -
> - pm_clk_resume(dev);
> -
> - return pm_generic_runtime_resume(dev);
> + pm_clk_destroy(dev);
> }
> -#endif
>
> -static struct dev_pm_domain keystone_pm_domain = {
> - .ops = {
> - SET_RUNTIME_PM_OPS(keystone_pm_runtime_suspend,
> - keystone_pm_runtime_resume, NULL)
> - USE_PLATFORM_PM_SLEEP_OPS
> +static const struct keystone_domain keystone_domain = {
> + .base = {
> + .name = "keystone",
> + .attach_dev = keystone_pm_domain_attach_dev,
> + .detach_dev = keystone_pm_domain_detach_dev,
> + .dev_ops = {
> + .stop = pm_clk_suspend,
> + .start = pm_clk_resume,
Same here: the clocks will be disabled/enabled on system suspend/resume,
which is not the case for the non-DT case.
Rafael: shouldn't pm_clk_notify() in drivers/base/power/clock_ops.c
behave the same as above?
Or is there a good reason the non-PM runtime version doesn't call pm_clk_add()?
Then the two versions (CONFIG_PM_RUNTIME enabled vs. disabled) can become
more similar, and perhaps be merged.
> + },
> + .power_off_latency_ns = 25000,
> + .power_on_latency_ns = 2000000,
> },
> };
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
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
