On Thu, 10 Oct 2013, Ulf Hansson wrote:
> Typically the runtime idle function is triggered after resume and
> probe. Instead of immediately requesting the device to go into
> in-active state we make use of the autosuspend, if we have enabled
> it earlier.
>
> Cc: Len Brown <len.brown@xxxxxxxxx>
> Cc: Pavel Machek <pavel@xxxxxx>
> Cc: Rafael J. Wysocki <rjw@xxxxxxxxxxxxx>
> Cc: Kevin Hilman <khilman@xxxxxxxxxx>
> Cc: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
> Cc: Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx>
> Cc: linux-pm@xxxxxxxxxxxxxxx
> Signed-off-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx>
> ---
> drivers/mmc/core/bus.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/mmc/core/bus.c b/drivers/mmc/core/bus.c
> index cdca8a7..7f0e900 100644
> --- a/drivers/mmc/core/bus.c
> +++ b/drivers/mmc/core/bus.c
> @@ -205,6 +205,12 @@ static int mmc_runtime_resume(struct device *dev)
>
> static int mmc_runtime_idle(struct device *dev)
> {
> + if (pm_runtime_autosuspend_used(dev)) {
> + pm_runtime_mark_last_busy(dev);
> + pm_runtime_autosuspend(dev);
> + return -EBUSY;
> + }
> +
> return 0;
> }
I would greatly prefer to see the core rpm_idle() routine changed
instead. Currently the last line says:
return retval ? retval : rpm_suspend(dev, rpmflags);
The second argument to rpm_suspend should be rpmflags | RPM_AUTO. That
way, if the runtime-idle callback returns 0, we will automatically do
either a normal suspend or an autosuspend, whichever is appropriate.
As an added benefit, with this change there's no need to add the
pm_runtime_autosuspend_used() function.
Alan Stern
--
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
