On Wed, 19 Dec 2018 at 11:11, Vincent Guittot
[off-list ref] wrote:

On Wed, 19 Dec 2018 at 10:58, Ulf Hansson [off-list ref] wrote:

quoted

On Tue, 18 Dec 2018 at 15:55, Vincent Guittot
[off-list ref] wrote:

quoted

Some drivers (like i915/drm) need to get the accounted suspended time.
pm_runtime_accounted_time_get() will return the suspended or active
accounted time until now.

I suggest to leave the active accounted time out for now. At least
until we have some users.

This is needed to keep same feature level for i915/drm

I don't follow. According to the changes in the drm driver in patch2,
we are only calling the new pm_runtime interface with RPM_SUSPENDED?

quoted

That said, perhaps rename the function to something along the lines
of, pm_runtime_last_suspended_time(), to make it more clear.

quoted

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
---
 drivers/base/power/runtime.c | 26 ++++++++++++++++++++++++++
 include/linux/pm_runtime.h   |  2 ++
 2 files changed, 28 insertions(+)

diff --git a/drivers/base/power/runtime.c b/drivers/base/power/runtime.c
index 7062469..6461469 100644
--- a/drivers/base/power/runtime.c
+++ b/drivers/base/power/runtime.c

@@ -88,6 +88,32 @@ static void __update_runtime_status(struct device *dev, enum rpm_status status)
        dev->power.runtime_status = status;
 }

+u64 pm_runtime_accounted_time_get(struct device *dev, enum rpm_status status, bool update)
+{
+       u64 now = ktime_to_ns(ktime_get());

I think you should stay on jiffies here - and then switch to ktime in patch 3.

quoted

+       u64 delta = 0, time = 0;
+       unsigned long flags;
+
+       spin_lock_irqsave(&dev->power.lock, flags);
+
+       if (dev->power.disable_depth > 0)
+               goto unlock;
+
+       /* Add ongoing state  if requested */
+       if (update && dev->power.runtime_status == status)
+               delta = now - dev->power.accounting_timestamp;
+

Hmm. Do we really need to update the accounting timestamp? I would
rather avoid it if possible.

i915/drm uses this to track ongoing suspended state. In fact they are
mainly interested by this part

Again, sorry I don't follow.

My suggested changes below, would do exactly that; track the ongoing
suspended state.

The user can call the function several times while the device remains
RPM_SUSPENDED, and if needed the user could then compute the delta
in-between the calls, for whatever reason that may be needed.

quoted

It seems like it should be sufficient to return the delta between
"now" and the "dev->power.accounting_timestamp", when
"dev->power.runtime_status == RPM_SUSPENDED".

In other case, just return 0, because we are not in RPM_SUSPENDED state.

only the RPM_SUSPENDED is used by i915/drm but wanted to provide a
generic interface to get
suspended or not suspend state

I see.

Although, unless Rafael thinks different, I would rather try to keep
this as simple as possible and expose only what is needed and nothing
more.

quoted

quoted

+       if (status == RPM_SUSPENDED)
+               time = dev->power.suspended_time + delta;
+       else
+               time = dev->power.active_time + delta;
+
+unlock:
+       spin_unlock_irqrestore(&dev->power.lock, flags);
+
+       return time;
+}
+
 /**
  * pm_runtime_deactivate_timer - Deactivate given device's suspend timer.
  * @dev: Device to handle.

diff --git a/include/linux/pm_runtime.h b/include/linux/pm_runtime.h
index 54af4ee..86f21f9 100644
--- a/include/linux/pm_runtime.h
+++ b/include/linux/pm_runtime.h

@@ -113,6 +113,8 @@ static inline bool pm_runtime_is_irq_safe(struct device *dev)
        return dev->power.irq_safe;
 }

+extern u64 pm_runtime_accounted_time_get(struct device *dev, enum rpm_status status, bool update);
+
 #else /* !CONFIG_PM */

 static inline bool queue_pm_work(struct work_struct *work) { return false; }
--

2.7.4

Kind regards
Uffe

Kind regards
Uffe