The PM core was preventing devices from going inactive during system suspend. Remove this constraint and moreover try to inactivate devices by invoking pm_runtime_idle() before proceeding with device's suspend_late callbacks. A great amount of drivers that uses runtime PM will benefit from this. Drivers that implements some of the system ->suspend* callbacks for the only reason of making sure the runtime PM resourses gets inactive, can now solely instead rely on the PM core to handle this. In the case were drivers needs there runtime resourses to be activated to be able to handle their system suspend tasks, those can easily just restore the runtime PM usage count, pm_runtime_put_sync(), once done with it's suspend operations, which then will inactivate the device. Cc: Kevin Hilman <khilman@xxxxxxxxxx> Cc: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> Signed-off-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx> --- There are a quite long history for how the PM core behaves around this topic. Changes has been iterated, back for forward, you might even think that it should have been settled at this point. So, forgive me for bringing this up, again. This patch shall obviously not be merged as is, it will cause regressions for several drivers and subsystems. The intent is instead to try to follow up on previous discussions to see if we might end up closer to the sequence, which my RFC PATCH suggests, or if I am just bothering you completely unnecessary. --- drivers/base/power/main.c | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-) diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index ee039af..670b1d5 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -832,7 +832,7 @@ static void device_complete(struct device *dev, pm_message_t state) device_unlock(dev); - pm_runtime_put(dev); + pm_request_idle(dev); } /** @@ -1004,6 +1004,7 @@ static int device_suspend_late(struct device *dev, pm_message_t state) pm_callback_t callback = NULL; char *info = NULL; + pm_runtime_idle(dev); __pm_runtime_disable(dev, false); if (dev->power.syscore) @@ -1312,14 +1313,6 @@ static int device_prepare(struct device *dev, pm_message_t state) if (dev->power.syscore) return 0; - /* - * If a device's parent goes into runtime suspend at the wrong time, - * it won't be possible to resume the device. To prevent this we - * block runtime suspend here, during the prepare phase, and allow - * it again during the complete phase. - */ - pm_runtime_get_noresume(dev); - device_lock(dev); dev->power.wakeup_path = device_may_wakeup(dev); -- 1.7.9.5 -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
