I'm adding Runtime PM support to the SDIO subsystem, and was planning to use the generic runtime pm handlers. Those handlers return -EINVAL if the driver didn't explicitly define the runtime pm handlers: That may result in some drivers defining nop handlers, just to return 0 (in case there's nothing else they need to do). Do we want that ? Alternatively, we may want to allow drivers to enable Runtime PM (by taking the appropriate action for their subsystem, e.g. calling put_noidle in probe and get_noresume in remove), but still not define any runtime pm handlers (implicitly always returning a success), with something like: diff --git a/drivers/base/power/generic_ops.c b/drivers/base/power/generic_ops. index 304c831..531762a 100644 --- a/drivers/base/power/generic_ops.c +++ b/drivers/base/power/generic_ops.c @@ -49,7 +49,7 @@ int pm_generic_runtime_suspend(struct device *dev) const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int ret; - ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : -EINVAL; + ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : 0; return ret; } @@ -68,7 +68,7 @@ int pm_generic_runtime_resume(struct device *dev) const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL; int ret; - ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : -EINVAL; + ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : 0; return ret; } What do you think ? Please note that the generic pm_generic_runtime_idle handler does allow drivers not to explicitly define a runtime_idle handler. Thanks, Ohad. -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
