Hi Geert,

Some more review comments.

+
+
+#ifdef CONFIG_PM_RUNTIME
+
+static int of_clk_pm_runtime_suspend(struct device *dev)
+{
+       int ret;
+
+       ret = pm_generic_runtime_suspend(dev);
+       if (ret)
+               return ret;
+
+       ret = pm_clk_suspend(dev);

What about slow clocks? Those aren't handled with pm_clk_suspend().

+       if (ret) {
+               pm_generic_runtime_resume(dev);
+               return ret;
+       }
+
+       return 0;
+}
+
+static int of_clk_pm_runtime_resume(struct device *dev)
+{
+       pm_clk_resume(dev);

What about slow clocks? Those aren't handled with pm_clk_resume().

+       return pm_generic_runtime_resume(dev);
+}
+
+static struct dev_pm_domain of_clk_pm_domain = {
+       .ops = {
+               .runtime_suspend = of_clk_pm_runtime_suspend,
+               .runtime_resume = of_clk_pm_runtime_resume,

Drivers/subsystems may invoke pm_runtime_force_suspend|resume() from
some of their system PM callbacks, which requires the runtime PM
callbacks to be defined for CONFIG_PM instead of CONFIG_PM_RUNTIME, I
believe that should be changed here as well.

+               USE_PLATFORM_PM_SLEEP_OPS

What about other buses beside the platfrom bus. Certainly we need to
handle devices attached to any other subsystem type as well.

Kind regards
Ulf Hansson