Various discussions on LKML have pointed out that many subsystem/drivers for
devices that may be attached to a genpd and which manages DVFS/OPP though the
genpd performance states, would need very similar updates.
More precisely, they would likely have to call dev_pm_opp_set_rate|opp() to
drop and restore OPPs (which propagates upwards into performance states votes
in genpd), every time their devices should enter/exit a low power state, via
their device PM callbacks.
Rather than having to add the boilerplate code for these things into the
subsystems/drivers, this series implements the logic internally into genpd.
Concerns have been raised about this approach, mostly by myself, around that it
limits flexibility. On the other hand, it starts to look like more and more
people are requesting this to be manged internally in genpd, for good reasons.
So, I think it's worth to give this a try.
In the long run, if it turns out that the flexibility was indeed needed, we can
always deal with that as special cases on top.
Test and reviews are of course greatly appreciated!
Kind regards
Ulf Hansson
Ulf Hansson (4):
PM: domains: Split code in dev_pm_genpd_set_performance_state()
PM: domains: Return early if perf state is already set for the device
PM: domains: Drop/restore performance state votes for devices at
runtime PM
PM: domains: Drop/restore performance state votes for devices at
system PM
drivers/base/power/domain.c | 70 +++++++++++++++++++++++++++++--------
include/linux/pm_domain.h | 2 ++
2 files changed, 58 insertions(+), 14 deletions(-)
--
2.25.1
To prepare some of the code in dev_pm_genpd_set_performance_state() to be
re-used from subsequent changes, let's split it up into two functions.
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- A small cosmetic update and added Viresh's tag.
---
drivers/base/power/domain.c | 31 +++++++++++++++++++------------
1 file changed, 19 insertions(+), 12 deletions(-)
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
For the similar reasons, but to avoid the same kind of boilerplate code in
device PM callbacks for system sleep in subsystems/drivers, let's drop and
restore performance states votes in genpd for the attached devices during
system sleep.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few cosmetic changes.
---
drivers/base/power/domain.c | 9 +++++++++
include/linux/pm_domain.h | 1 +
2 files changed, 10 insertions(+)
When dev_pm_genpd_set_performance_state() gets called to set a new
performance state for the device, let's take a quicker path by doing an
early return, if it turns out that the new state is already set for the
device.
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- New patch.
---
drivers/base/power/domain.c | 3 +++
1 file changed, 3 insertions(+)
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
---
drivers/base/power/domain.c | 27 +++++++++++++++++++++++++--
include/linux/pm_domain.h | 1 +
2 files changed, 26 insertions(+), 2 deletions(-)
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
---
drivers/base/power/domain.c | 27 +++++++++++++++++++++++++--
include/linux/pm_domain.h | 1 +
2 files changed, 26 insertions(+), 2 deletions(-)
+ Mark Brown, Dmitry Baryshkov
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
For the similar reasons, but to avoid the same kind of boilerplate code in
device PM callbacks for system sleep in subsystems/drivers, let's drop and
restore performance states votes in genpd for the attached devices during
system sleep.
Signed-off-by: Ulf Hansson <redacted>
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface. We likely don't want to drop
the performance state internally in genpd when genpd_suspend_noirq()
gets called, for the corresponding struct device for the regulator.
I guess if genpd should drop performance states like $subject patch
suggest, we need some kind of additional coordination, that allows a
subsystem/driver to inform genpd when it should avoid it. Or something
along those lines.
Kind regards
Uffe
quoted hunk
---
Changes in v2:
- Rebased.
- A few cosmetic changes.
---
drivers/base/power/domain.c | 9 +++++++++
include/linux/pm_domain.h | 1 +
2 files changed, 10 insertions(+)
On Thu, 3 Jun 2021 at 11:55, Viresh Kumar [off-list ref] wrote:
On 03-06-21, 11:34, Ulf Hansson wrote:
quoted
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
---
drivers/base/power/domain.c | 27 +++++++++++++++++++++++++--
include/linux/pm_domain.h | 1 +
2 files changed, 26 insertions(+), 2 deletions(-)
@@ -400,6 +400,23 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state)returnret;}+staticintgenpd_drop_performance_state(structdevice*dev)
What about passing the state pointer here? that will simplify the
callers to just a call.
Not sure I get that. Can you elaborate a bit more?
quoted
+{
+ unsigned int prev_state = dev_gpd_data(dev)->performance_state;
+
+ if (!genpd_set_performance_state(dev, 0))
+ return prev_state;
+
+ return 0;
+}
+
+static void genpd_restore_performance_state(struct device *dev,
+ unsigned int state)
+{
+ if (state)
I will skip this check, as we are checking it in
genpd_set_performance_state() anyway ?
I don't want us to override OPP votes made by the subsystem/driver
level runtime PM callbacks. For example, if the drivers manage this
thing themselves, that should be preserved.
That said, by the check above I want to avoid setting the state to
zero internally by genpd, if the driver level ->runtime_resume()
callback has already restored the state.
quoted
+ genpd_set_performance_state(dev, state);
+}
+
/**
* dev_pm_genpd_set_performance_state- Set performance state of device's power
* domain.
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-03 11:13:15
On Thu, Jun 03, 2021 at 11:34:34AM +0200, Ulf Hansson wrote:
Various discussions on LKML have pointed out that many subsystem/drivers for
devices that may be attached to a genpd and which manages DVFS/OPP though the
genpd performance states, would need very similar updates.
More precisely, they would likely have to call dev_pm_opp_set_rate|opp() to
drop and restore OPPs (which propagates upwards into performance states votes
in genpd), every time their devices should enter/exit a low power state, via
their device PM callbacks.
Rather than having to add the boilerplate code for these things into the
subsystems/drivers, this series implements the logic internally into genpd.
Concerns have been raised about this approach, mostly by myself, around that it
limits flexibility. On the other hand, it starts to look like more and more
people are requesting this to be manged internally in genpd, for good reasons.
So, I think it's worth to give this a try.
In the long run, if it turns out that the flexibility was indeed needed, we can
always deal with that as special cases on top.
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Stephan
[1]: https://lore.kernel.org/linux-pm/20200826093328.88268-1-stephan@gerhold.net/
From: Mark Brown <broonie@kernel.org> Date: 2021-06-03 11:15:43
On Thu, Jun 03, 2021 at 12:20:57PM +0200, Ulf Hansson wrote:
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
quoted
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface. We likely don't want to drop
the performance state internally in genpd when genpd_suspend_noirq()
gets called, for the corresponding struct device for the regulator.
I guess if genpd should drop performance states like $subject patch
suggest, we need some kind of additional coordination, that allows a
subsystem/driver to inform genpd when it should avoid it. Or something
along those lines.
I'm not sure what you're looking for from me here - was there a concrete
question or somehing?
On Thu, 3 Jun 2021 at 12:31, Ulf Hansson [off-list ref] wrote:
On Thu, 3 Jun 2021 at 11:55, Viresh Kumar [off-list ref] wrote:
quoted
On 03-06-21, 11:34, Ulf Hansson wrote:
quoted
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
---
drivers/base/power/domain.c | 27 +++++++++++++++++++++++++--
include/linux/pm_domain.h | 1 +
2 files changed, 26 insertions(+), 2 deletions(-)
@@ -400,6 +400,23 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state)returnret;}+staticintgenpd_drop_performance_state(structdevice*dev)
What about passing the state pointer here? that will simplify the
callers to just a call.
Not sure I get that. Can you elaborate a bit more?
quoted
quoted
+{
+ unsigned int prev_state = dev_gpd_data(dev)->performance_state;
+
+ if (!genpd_set_performance_state(dev, 0))
+ return prev_state;
+
+ return 0;
+}
+
+static void genpd_restore_performance_state(struct device *dev,
+ unsigned int state)
+{
+ if (state)
I will skip this check, as we are checking it in
genpd_set_performance_state() anyway ?
I don't want us to override OPP votes made by the subsystem/driver
level runtime PM callbacks. For example, if the drivers manage this
thing themselves, that should be preserved.
That said, by the check above I want to avoid setting the state to
zero internally by genpd, if the driver level ->runtime_resume()
callback has already restored the state.
Ehh, forget about what I said about the ->runtime_resume() callback.
I am mostly trying to avoid restoring a state that is zero, just to be
sure nobody else on some different level outside gendp, have decided
to set a new OPP in-between our calls to
genpd_drop|restore_performance state.
[...]
Kind regards
Uffe
On Thu, 3 Jun 2021 at 13:15, Mark Brown [off-list ref] wrote:
On Thu, Jun 03, 2021 at 12:20:57PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
quoted
quoted
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
quoted
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
quoted
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface. We likely don't want to drop
the performance state internally in genpd when genpd_suspend_noirq()
gets called, for the corresponding struct device for the regulator.
quoted
I guess if genpd should drop performance states like $subject patch
suggest, we need some kind of additional coordination, that allows a
subsystem/driver to inform genpd when it should avoid it. Or something
along those lines.
I'm not sure what you're looking for from me here - was there a concrete
question or somehing?
Nope, not really, sorry if that was not clear.
I just wanted to loop you in, as to make sure that we don't change
something at the PM domain level, which may not fit well with the
regulator implementation.
Kind regards
Uffe
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
On Thu, Jun 03, 2021 at 11:34:34AM +0200, Ulf Hansson wrote:
quoted
Various discussions on LKML have pointed out that many subsystem/drivers for
devices that may be attached to a genpd and which manages DVFS/OPP though the
genpd performance states, would need very similar updates.
More precisely, they would likely have to call dev_pm_opp_set_rate|opp() to
drop and restore OPPs (which propagates upwards into performance states votes
in genpd), every time their devices should enter/exit a low power state, via
their device PM callbacks.
Rather than having to add the boilerplate code for these things into the
subsystems/drivers, this series implements the logic internally into genpd.
Concerns have been raised about this approach, mostly by myself, around that it
limits flexibility. On the other hand, it starts to look like more and more
people are requesting this to be manged internally in genpd, for good reasons.
So, I think it's worth to give this a try.
In the long run, if it turns out that the flexibility was indeed needed, we can
always deal with that as special cases on top.
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
When it comes to the system suspend/resume path, being implemented in
patch4, we should probably defer that patch from being merged. It
turned out that we probably need to think more about that approach.
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-03 17:18:36
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
Stephan
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
I tested this on NVIDIA Tegra by removing the boilerplate code from
drivers' RPM and haven't noticed any problems, the performance state is
dropped/restored as expected. Thank you.
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
I tested this on NVIDIA Tegra by removing the boilerplate code from
drivers' RPM and haven't noticed any problems, the performance state is
dropped/restored as expected. Thank you.
On Thu, 3 Jun 2021 at 12:31, Ulf Hansson [off-list ref] wrote:
quoted
quoted
quoted
+static int genpd_drop_performance_state(struct device *dev)
+{
+ unsigned int prev_state = dev_gpd_data(dev)->performance_state;
+
+ if (!genpd_set_performance_state(dev, 0))
+ return prev_state;
+
+ return 0;
+}
+
+static void genpd_restore_performance_state(struct device *dev,
+ unsigned int state)
+{
+ if (state)
I will skip this check, as we are checking it in
genpd_set_performance_state() anyway ?
I don't want us to override OPP votes made by the subsystem/driver
level runtime PM callbacks. For example, if the drivers manage this
thing themselves, that should be preserved.
That said, by the check above I want to avoid setting the state to
zero internally by genpd, if the driver level ->runtime_resume()
callback has already restored the state.
Ehh, forget about what I said about the ->runtime_resume() callback.
I am mostly trying to avoid restoring a state that is zero, just to be
sure nobody else on some different level outside gendp, have decided
to set a new OPP in-between our calls to
genpd_drop|restore_performance state.
What stops the core to call genpd_drop_performance_state() in the
first place here, if the driver was doing its own thing ? If that gets
called, then restore should be without any checks IMO. The state
should already be 0 at this point of time, I don't know why this will
get called again with state 0, but it will have no effect.
Can you give some sort of flow sequence where I can see the problem a
bit more clearly ?
--
viresh
On Thu, 3 Jun 2021 at 19:16, Stephan Gerhold [off-list ref] wrote:
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
Right. I understand your point and I am open to discuss an
implementation. Although, I suggest we continue that separately from
the $subject series.
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
Well, that's what most subsystems/drivers need to do.
Moreover, we have specific devices that only use one default OPP [1].
quoted
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
What do you mean by "next time"?
My main point is, if the device enters runtime suspend state, why
should we keep the vote for an OPP for the device? I mean, the device
isn't going to be used anyway.
On Thu, 3 Jun 2021 at 21:08, Dmitry Osipenko [off-list ref] wrote:
03.06.2021 22:02, Dmitry Osipenko пишет:
quoted
03.06.2021 12:34, Ulf Hansson пишет:
quoted
A subsystem/driver that need to manage OPPs for its device, should
typically drop its vote for the OPP when the device becomes runtime
suspended. In this way, the corresponding aggregation of the performance
state votes that is managed in genpd for the attached PM domain, may find
that the aggregated vote can be decreased. Hence, it may allow genpd to set
the lower performance state for the PM domain, thus avoiding to waste
energy.
To accomplish this, typically a subsystem/driver would need to call
dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend()
callback, to drop the vote for the OPP. Accordingly, it needs another call
to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its
->runtime_resume() callback.
To avoid boilerplate code in subsystems/driver to deal with these things,
let's instead manage this internally in genpd.
Signed-off-by: Ulf Hansson <redacted>
---
Changes in v2:
- Rebased.
- A few minor cosmetic changes.
- Deal with the error path in genpd_runtime_resume().
I tested this on NVIDIA Tegra by removing the boilerplate code from
drivers' RPM and haven't noticed any problems, the performance state is
dropped/restored as expected. Thank you.
On Fri, 4 Jun 2021 at 05:53, Viresh Kumar [off-list ref] wrote:
On 03-06-21, 13:17, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 12:31, Ulf Hansson [off-list ref] wrote:
quoted
quoted
quoted
+static int genpd_drop_performance_state(struct device *dev)
+{
+ unsigned int prev_state = dev_gpd_data(dev)->performance_state;
+
+ if (!genpd_set_performance_state(dev, 0))
+ return prev_state;
+
+ return 0;
+}
+
+static void genpd_restore_performance_state(struct device *dev,
+ unsigned int state)
+{
+ if (state)
I will skip this check, as we are checking it in
genpd_set_performance_state() anyway ?
I don't want us to override OPP votes made by the subsystem/driver
level runtime PM callbacks. For example, if the drivers manage this
thing themselves, that should be preserved.
That said, by the check above I want to avoid setting the state to
zero internally by genpd, if the driver level ->runtime_resume()
callback has already restored the state.
Ehh, forget about what I said about the ->runtime_resume() callback.
I am mostly trying to avoid restoring a state that is zero, just to be
sure nobody else on some different level outside gendp, have decided
to set a new OPP in-between our calls to
genpd_drop|restore_performance state.
What stops the core to call genpd_drop_performance_state() in the
first place here, if the driver was doing its own thing ? If that gets
called, then restore should be without any checks IMO. The state
should already be 0 at this point of time, I don't know why this will
get called again with state 0, but it will have no effect.
Can you give some sort of flow sequence where I can see the problem a
bit more clearly ?
Starting calls from the subsystem/driver:
------
dev_pm_genpd_set_performance_state(dev, 100);
"run a use case with device runtime resumed"
...
"use case ends"
dev_pm_genpd_set_performance_state(dev, 0);
pm_runtime_put()
->genpd_runtime_suspend()
gpd_data->performance_state == 0, -> gpd_data->rpm_pstate = 0;
...
"new use case start"
dev_pm_genpd_set_performance_state(dev, 100);
pm_runtime_get_sync()
->genpd_runtime_resume()
gpd_data->performance_state == 100, -> gpd_data->rpm_pstate = 0;
(This is where we need to check for "zero" to not override the value)
.....
------
I wouldn't say that the above is the way how I see the calls to
dev_pm_genpd_set_performance_state (or actually
dev_pm_opp_set_rate|opp()) being deployed. The calls should rather be
done from the subsystem/driver's ->runtime_suspend|resume() callback,
then the path above would work in the way you suggest.
Although, as we currently treat performance states and power states in
genpd orthogonally, I wanted to make sure we could cope with both
situations.
Did this help? :-)
Kind regards
Uffe
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-04 08:23:34
On Fri, Jun 04, 2021 at 09:18:45AM +0200, Ulf Hansson wrote:
On Thu, 3 Jun 2021 at 19:16, Stephan Gerhold [off-list ref] wrote:
quoted
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
Right. I understand your point and I am open to discuss an
implementation. Although, I suggest we continue that separately from
the $subject series.
quoted
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
Well, that's what most subsystems/drivers need to do.
Moreover, we have specific devices that only use one default OPP [1].
quoted
quoted
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
What do you mean by "next time"?
Basically just like:
<device runtime-suspended>
driver does dev_pm_genpd_set_performance_state(...)
- performance state is applied immediately, even though device does
apparently not actually want the power domain to be powered on
<device runtime resumes>
- performance state is kept
<device runtime suspends>
- performance state is dropped
...
I'm not saying this example makes sense (it doesn't for me). It doesn't
make sense to vote for a performance state while runtime suspended.
But with this patch series we still allow that, and it will kind of
produce inconsistent behavior that the performance state is applied
immediately, even though the device is currently runtime-suspended.
But once it runtime suspends again, suddenly it is dropped.
And when you say:
My main point is, if the device enters runtime suspend state, why
should we keep the vote for an OPP for the device? I mean, the device
isn't going to be used anyway.
A very similar point would be: "If the device *is* in runtime suspend
state, why should we take a vote for an OPP for the device?"
But I understand that this might be something we should address
separately in a follow-up patch/discussion. Don't get me wrong, I agree
this patch set is good, I just think we should go one step further and
finally make this consistent and less prone to side effects.
A good first step might be something like a WARN_ON_ONCE(...) if a
device tries to vote for a performance state while runtime suspended.
Then we might get a clearer picture which drivers do that currently.
Stephan
On Fri, 4 Jun 2021 at 10:23, Stephan Gerhold [off-list ref] wrote:
On Fri, Jun 04, 2021 at 09:18:45AM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 19:16, Stephan Gerhold [off-list ref] wrote:
quoted
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
Right. I understand your point and I am open to discuss an
implementation. Although, I suggest we continue that separately from
the $subject series.
quoted
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
Well, that's what most subsystems/drivers need to do.
Moreover, we have specific devices that only use one default OPP [1].
quoted
quoted
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
What do you mean by "next time"?
Basically just like:
<device runtime-suspended>
driver does dev_pm_genpd_set_performance_state(...)
- performance state is applied immediately, even though device does
apparently not actually want the power domain to be powered on
<device runtime resumes>
- performance state is kept
<device runtime suspends>
- performance state is dropped
Yep, this is what would happen.
...
I'm not saying this example makes sense (it doesn't for me). It doesn't
make sense to vote for a performance state while runtime suspended.
But with this patch series we still allow that, and it will kind of
produce inconsistent behavior that the performance state is applied
immediately, even though the device is currently runtime-suspended.
But once it runtime suspends again, suddenly it is dropped.
Yes.
Note that, I have been looking at the existing callers of
dev_pm_genpd_set_performance_state() in the kernel as of today. It
should not be an issue, at least as far as I can tell.
And when you say:
quoted
My main point is, if the device enters runtime suspend state, why
should we keep the vote for an OPP for the device? I mean, the device
isn't going to be used anyway.
A very similar point would be: "If the device *is* in runtime suspend
state, why should we take a vote for an OPP for the device?"
But I understand that this might be something we should address
separately in a follow-up patch/discussion. Don't get me wrong, I agree
this patch set is good, I just think we should go one step further and
finally make this consistent and less prone to side effects.
I agree. We should look into how to change the behaviour. I intend to
have a look at it in a while.
A good first step might be something like a WARN_ON_ONCE(...) if a
device tries to vote for a performance state while runtime suspended.
Then we might get a clearer picture which drivers do that currently.
That's an idea we could try, even if the number of users are quite
limited today. I can try the "git grep" analyze-method, I will
probably find most of them.
Stephan
That said, are you okay that we move forward with the $subject series
(except patch4)?
Kind regards
Uffe
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-04 11:51:08
On Fri, Jun 04, 2021 at 12:57:59PM +0200, Ulf Hansson wrote:
On Fri, 4 Jun 2021 at 10:23, Stephan Gerhold [off-list ref] wrote:
quoted
On Fri, Jun 04, 2021 at 09:18:45AM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 19:16, Stephan Gerhold [off-list ref] wrote:
quoted
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
Right. I understand your point and I am open to discuss an
implementation. Although, I suggest we continue that separately from
the $subject series.
quoted
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
Well, that's what most subsystems/drivers need to do.
Moreover, we have specific devices that only use one default OPP [1].
quoted
quoted
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
What do you mean by "next time"?
Basically just like:
<device runtime-suspended>
driver does dev_pm_genpd_set_performance_state(...)
- performance state is applied immediately, even though device does
apparently not actually want the power domain to be powered on
<device runtime resumes>
- performance state is kept
<device runtime suspends>
- performance state is dropped
Yep, this is what would happen.
quoted
...
I'm not saying this example makes sense (it doesn't for me). It doesn't
make sense to vote for a performance state while runtime suspended.
But with this patch series we still allow that, and it will kind of
produce inconsistent behavior that the performance state is applied
immediately, even though the device is currently runtime-suspended.
But once it runtime suspends again, suddenly it is dropped.
Yes.
Note that, I have been looking at the existing callers of
dev_pm_genpd_set_performance_state() in the kernel as of today. It
should not be an issue, at least as far as I can tell.
quoted
And when you say:
quoted
My main point is, if the device enters runtime suspend state, why
should we keep the vote for an OPP for the device? I mean, the device
isn't going to be used anyway.
A very similar point would be: "If the device *is* in runtime suspend
state, why should we take a vote for an OPP for the device?"
But I understand that this might be something we should address
separately in a follow-up patch/discussion. Don't get me wrong, I agree
this patch set is good, I just think we should go one step further and
finally make this consistent and less prone to side effects.
I agree. We should look into how to change the behaviour. I intend to
have a look at it in a while.
Great, thanks!
quoted
A good first step might be something like a WARN_ON_ONCE(...) if a
device tries to vote for a performance state while runtime suspended.
Then we might get a clearer picture which drivers do that currently.
That's an idea we could try, even if the number of users are quite
limited today. I can try the "git grep" analyze-method, I will
probably find most of them.
The current user of "required-opps" for CPU DVFS (just qcom/qcs404.dtsi
with qcom/cpr.c I think?) is definitely broken (never votes to turn on
the power domain). So one requirement for making that change of behavior
is figuring out how to deal with enabling power domains at the OPP core
(or whereever else).
That said, are you okay that we move forward with the $subject series
(except patch4)?
It sounds fine to me. My system doesn't have power domain performance
states set up properly yet (due to various open questions), so I can't
test it properly though. Also, I'm not sure if it's a good idea to omit
patch 4, doesn't that mean drivers that currently drop the performance
states themselves can be only partially cleaned up?
I do have some thoughts about the "regulator-fixed-domain". Not exactly
a solution for the problem you mentioned, just some related thoughts.
Will try to reply there later.
Thanks!
Stephan
Starting calls from the subsystem/driver:
------
dev_pm_genpd_set_performance_state(dev, 100);
"run a use case with device runtime resumed"
...
"use case ends"
dev_pm_genpd_set_performance_state(dev, 0);
pm_runtime_put()
->genpd_runtime_suspend()
gpd_data->performance_state == 0, -> gpd_data->rpm_pstate = 0;
...
"new use case start"
dev_pm_genpd_set_performance_state(dev, 100);
pm_runtime_get_sync()
->genpd_runtime_resume()
gpd_data->performance_state == 100, -> gpd_data->rpm_pstate = 0;
(This is where we need to check for "zero" to not override the value)
.....
------
I wouldn't say that the above is the way how I see the calls to
dev_pm_genpd_set_performance_state (or actually
dev_pm_opp_set_rate|opp()) being deployed. The calls should rather be
done from the subsystem/driver's ->runtime_suspend|resume() callback,
then the path above would work in the way you suggest.
Although, as we currently treat performance states and power states in
genpd orthogonally, I wanted to make sure we could cope with both
situations.
I think letting the drivers to call
dev_pm_genpd_set_performance_state(dev, 0) from suspend/resume makes
it really ugly/racy as both depend on the gpd_data->performance_state
for this. It doesn't look nice. And we shouldn't try to protect such
drivers.
Anyway, your call :)
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-08 12:53:23
Hi,
On Thu, Jun 03, 2021 at 12:20:57PM +0200, Ulf Hansson wrote:
+ Mark Brown, Dmitry Baryshkov
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
quoted
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
For the similar reasons, but to avoid the same kind of boilerplate code in
device PM callbacks for system sleep in subsystems/drivers, let's drop and
restore performance states votes in genpd for the attached devices during
system sleep.
Signed-off-by: Ulf Hansson <redacted>
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface.
Not directly related to your concern, but related to another discussion
we had recently: To me, this looks mostly like another solution for
voting for performance states without doing full DVFS, also known as
assigned-performance-states [1] or required-opps on devices [2]. :)
It's just wrapped in a regulator interface here. Actually, if we
implement [2], the regulator-fixed-domain should mostly just become some
sort of simple wrapper around runtime PM for the regulator device, since
the required-opp might be applied automatically then.
[1]: https://lore.kernel.org/linux-arm-msm/1622095949-2014-1-git-send-email-rnayak@codeaurora.org/
[2]: https://lore.kernel.org/linux-arm-msm/YLYV3ov%2FiBffZMg4@gerhold.net/
We likely don't want to drop the performance state internally in genpd
when genpd_suspend_noirq() gets called, for the corresponding struct
device for the regulator.
So your concern is that the performance state is dropped during suspend
even though the regulator core thinks the regulator stays enabled?
I played with regulator-fixed-domain a bit and I would say this is
already broken (unless you rely on one of the side effects I mentioned
in [3]). The power domain gets powered off entirely during system
suspend, and then the performance state won't have any effect either.
I guess we would need some way to say that this device should only be
managed through runtime PM and never automatically suspended during
system suspend?
Stephan
[3]: https://lore.kernel.org/linux-pm/YLkOAyydZMnxkEy+@gerhold.net/
On Tue, 8 Jun 2021 at 14:53, Stephan Gerhold [off-list ref] wrote:
Hi,
On Thu, Jun 03, 2021 at 12:20:57PM +0200, Ulf Hansson wrote:
quoted
+ Mark Brown, Dmitry Baryshkov
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
quoted
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
For the similar reasons, but to avoid the same kind of boilerplate code in
device PM callbacks for system sleep in subsystems/drivers, let's drop and
restore performance states votes in genpd for the attached devices during
system sleep.
Signed-off-by: Ulf Hansson <redacted>
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface.
Not directly related to your concern, but related to another discussion
we had recently: To me, this looks mostly like another solution for
voting for performance states without doing full DVFS, also known as
assigned-performance-states [1] or required-opps on devices [2]. :)
It's just wrapped in a regulator interface here. Actually, if we
implement [2], the regulator-fixed-domain should mostly just become some
sort of simple wrapper around runtime PM for the regulator device, since
the required-opp might be applied automatically then.
Honestly, I am not sure about what the regulator-fixed-domain intends
to model, but I assume it's something that fits well to be modelled as
a plain regulator, to start with.
Perhaps Mark can chime in and spread some light over this?
We likely don't want to drop the performance state internally in genpd
when genpd_suspend_noirq() gets called, for the corresponding struct
device for the regulator.
So your concern is that the performance state is dropped during suspend
even though the regulator core thinks the regulator stays enabled?
Yes.
I played with regulator-fixed-domain a bit and I would say this is
already broken (unless you rely on one of the side effects I mentioned
in [3]). The power domain gets powered off entirely during system
suspend, and then the performance state won't have any effect either.
Right, I get your point.
Although, this isn't a problem, because the on/off and performance
states are today considered as orthogonal in gendp. Well, at least
currently until/if we decide to change this.
I guess we would need some way to say that this device should only be
managed through runtime PM and never automatically suspended during
system suspend?
Yes!
For the on/off state, genpd uses the system wakeup interface to
understand whether the device is used in a wakeup path, see the call
to device_wakeup_path() in genpd_finish_suspend().
If that's the case the PM domain stays powered on during system suspend.
Potentially we could use the same interface (or something similar) to
support these kinds of cases.
From: Mark Brown <broonie@kernel.org> Date: 2021-06-08 14:20:36
On Tue, Jun 08, 2021 at 04:08:55PM +0200, Ulf Hansson wrote:
Honestly, I am not sure about what the regulator-fixed-domain intends
to model, but I assume it's something that fits well to be modelled as
a plain regulator, to start with.
Perhaps Mark can chime in and spread some light over this?
IIRC it's for situations where there's a device that's normally built as
a separate chip that got built into a bigger SoC and wants to rear end
something onto a power domain, I guess especially if the power domain
doesn't cover the whole of a Linux device.
On Tue, 8 Jun 2021 at 16:20, Mark Brown [off-list ref] wrote:
On Tue, Jun 08, 2021 at 04:08:55PM +0200, Ulf Hansson wrote:
quoted
Honestly, I am not sure about what the regulator-fixed-domain intends
to model, but I assume it's something that fits well to be modelled as
a plain regulator, to start with.
quoted
Perhaps Mark can chime in and spread some light over this?
IIRC it's for situations where there's a device that's normally built as
a separate chip that got built into a bigger SoC and wants to rear end
something onto a power domain, I guess especially if the power domain
doesn't cover the whole of a Linux device.
Alright, thanks for explaining!
Certainly something that should not be mixed up with a regular power
rail for shared resources (aka power-domain). Maybe it's worth trying
to extend the description in the DT binding a bit, so it doesn't get
abused?
Kind regards
Uffe
From: Stephan Gerhold <stephan@gerhold.net> Date: 2021-06-08 15:37:23
On Tue, Jun 08, 2021 at 04:08:55PM +0200, Ulf Hansson wrote:
On Tue, 8 Jun 2021 at 14:53, Stephan Gerhold [off-list ref] wrote:
quoted
Hi,
On Thu, Jun 03, 2021 at 12:20:57PM +0200, Ulf Hansson wrote:
quoted
+ Mark Brown, Dmitry Baryshkov
On Thu, 3 Jun 2021 at 11:34, Ulf Hansson [off-list ref] wrote:
quoted
Recent changes in genpd drops and restore performance state votes for
devices during runtime PM.
For the similar reasons, but to avoid the same kind of boilerplate code in
device PM callbacks for system sleep in subsystems/drivers, let's drop and
restore performance states votes in genpd for the attached devices during
system sleep.
Signed-off-by: Ulf Hansson <redacted>
After a second thought, it looks like we maybe should defer to apply
this final patch of the series. At least until we figured out how to
address the below issue:
So, I noticed that we have things like "regulator-fixed-domain", that
uses "required-opps" to enable/disable a regulator through the
dev_pm_set_performance_state() interface.
Not directly related to your concern, but related to another discussion
we had recently: To me, this looks mostly like another solution for
voting for performance states without doing full DVFS, also known as
assigned-performance-states [1] or required-opps on devices [2]. :)
It's just wrapped in a regulator interface here. Actually, if we
implement [2], the regulator-fixed-domain should mostly just become some
sort of simple wrapper around runtime PM for the regulator device, since
the required-opp might be applied automatically then.
Honestly, I am not sure about what the regulator-fixed-domain intends
to model, but I assume it's something that fits well to be modelled as
a plain regulator, to start with.
Perhaps Mark can chime in and spread some light over this?
We likely don't want to drop the performance state internally in genpd
when genpd_suspend_noirq() gets called, for the corresponding struct
device for the regulator.
So your concern is that the performance state is dropped during suspend
even though the regulator core thinks the regulator stays enabled?
Yes.
quoted
I played with regulator-fixed-domain a bit and I would say this is
already broken (unless you rely on one of the side effects I mentioned
in [3]). The power domain gets powered off entirely during system
suspend, and then the performance state won't have any effect either.
Right, I get your point.
Although, this isn't a problem, because the on/off and performance
states are today considered as orthogonal in gendp. Well, at least
currently until/if we decide to change this.
And in practice applying your patch should not be a problem either. :)
The main user so far is arch/arm64/boot/dts/qcom/sm8250.dtsi with the
rpmhpd genpd provider. That one drops the performance states anyway
when the power domain gets powered off.
quoted
I guess we would need some way to say that this device should only be
managed through runtime PM and never automatically suspended during
system suspend?
Yes!
For the on/off state, genpd uses the system wakeup interface to
understand whether the device is used in a wakeup path, see the call
to device_wakeup_path() in genpd_finish_suspend().
If that's the case the PM domain stays powered on during system suspend.
Potentially we could use the same interface (or something similar) to
support these kinds of cases.
Hmm, I wonder if I need something similar for my CPU DVFS case as well.
In my testing back then the power domain for the CPU was powered off
during system suspend. This doesn't make sense, since it's the CPU that
is running all the system suspend code. :)
Actually in my case there is no need to do anything during system
suspend. I use a special MSM8916_VDDMX_AO power domain from the "rpmpd"
genpd provider where _AO means "active-only". I think it's some kind of
firmware mechanism to only apply the performance state vote if the CPU
is actually active and the cluster is not powered down.
(I hope that MSM8916 actually powers down the CPU cluster during
system suspend/s2idle after all your PSCI cpuidle patches :))
So I think I will also need to opt-out from the effects of this patch.
Stephan
On Mon, 7 Jun 2021 at 06:47, Viresh Kumar [off-list ref] wrote:
On 04-06-21, 09:45, Ulf Hansson wrote:
quoted
Starting calls from the subsystem/driver:
------
dev_pm_genpd_set_performance_state(dev, 100);
"run a use case with device runtime resumed"
...
"use case ends"
dev_pm_genpd_set_performance_state(dev, 0);
pm_runtime_put()
->genpd_runtime_suspend()
gpd_data->performance_state == 0, -> gpd_data->rpm_pstate = 0;
...
"new use case start"
dev_pm_genpd_set_performance_state(dev, 100);
pm_runtime_get_sync()
->genpd_runtime_resume()
gpd_data->performance_state == 100, -> gpd_data->rpm_pstate = 0;
(This is where we need to check for "zero" to not override the value)
.....
------
I wouldn't say that the above is the way how I see the calls to
dev_pm_genpd_set_performance_state (or actually
dev_pm_opp_set_rate|opp()) being deployed. The calls should rather be
done from the subsystem/driver's ->runtime_suspend|resume() callback,
then the path above would work in the way you suggest.
Although, as we currently treat performance states and power states in
genpd orthogonally, I wanted to make sure we could cope with both
situations.
I think letting the drivers to call
dev_pm_genpd_set_performance_state(dev, 0) from suspend/resume makes
it really ugly/racy as both depend on the gpd_data->performance_state
for this. It doesn't look nice. And we shouldn't try to protect such
drivers.
Anyway, your call :)
Well, I am not sure we have an option at this point. As long as we
allow performance states to be managed orthogonally to on/off states
in genpd, the check in genpd_restore_performance_state() is needed.
I have started to prepare a new version of the series - and will add a
comment about this in the code to try to clarify this.
[...]
Kind regards
Uffe
From: "Rafael J. Wysocki" <rafael@kernel.org> Date: 2021-06-11 16:43:09
On Fri, Jun 4, 2021 at 1:51 PM Stephan Gerhold [off-list ref] wrote:
On Fri, Jun 04, 2021 at 12:57:59PM +0200, Ulf Hansson wrote:
quoted
On Fri, 4 Jun 2021 at 10:23, Stephan Gerhold [off-list ref] wrote:
quoted
On Fri, Jun 04, 2021 at 09:18:45AM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 19:16, Stephan Gerhold [off-list ref] wrote:
quoted
On Thu, Jun 03, 2021 at 05:27:30PM +0200, Ulf Hansson wrote:
quoted
On Thu, 3 Jun 2021 at 13:13, Stephan Gerhold [off-list ref] wrote:
quoted
I think this might also go into the direction of my problem with the OPP
core for CPU DVFS [1] since the OPP core currently does not "power-on"
the power domains, it just sets a performance state. I got kind of stuck
with all the complexity of power domains in Linux so I think we never
solved that.
Hmm, that issue is in a way related.
Although, if I understand correctly, that was rather about at what
layer it makes best sense to activate the device (from runtime PM
point of view). And this was needed due to the fact that the
corresponding genpd provider, requires the PM domain to be power on to
allow changing a performance state for it. Did I get that correct?
Yes, mostly. But I guess I keep coming back to the same question:
When/why does it make sense to vote for a "performance state" of
a power domain that is or might be powered off?
"Powered off" sounds like the absolutely lowest possible performance
state to me, it's just not on at all. And if suddenly a device comes and
says "I want performance state X", nothing can change until the power
domain is also "powered on".
I think my "CPU DVFS" problem only exists because in many other
situations it's possible to rely on one of the following side effects:
1. The genpd provider does not care if it's powered on or not.
(i.e. it's always-on or implicitly powers on if state > 0).
2. There is some other device that votes to keep the power domain on.
And that's how the problem relates to my comment for this patch series ...
quoted
quoted
Do I understand your patch set correctly that you basically make the
performance state votes conditional to the "power-on" vote of the device
(which is automatically toggled during runtime/system PM)?
The series can be considered as a step in that direction, but no, this
series doesn't change that behaviour.
Users of dev_pm_genpd_set_performance_state() are still free to set a
performance state, orthogonally to whether the PM domain is powered on
or off.
quoted
If yes, I think that's a good thing. It was always really confusing to me
that a device can make performance state votes if it doesn't actually
want the power domain to be powered on.
I share your view, it's a bit confusing.
Just adding the condition internally to genpd to prevent the caller of
dev_pm_genpd_set_performance() from succeeding to set a new state,
unless the genpd is powered on, should be a rather simple thing to
add.
However, to change this, we first need to double check that all the
callers are making sure they have turned on the PM domain (typically
via runtime PM).
... because if performance state votes would be conditional to the
"power-on" vote of the device, it would no longer be possible
to rely on the side effects mentioned above. So this would most
certainly break some code that (incorrectly?) relies on these side
effects, but would also prevent such code.
Right. I understand your point and I am open to discuss an
implementation. Although, I suggest we continue that separately from
the $subject series.
quoted
My (personal) feeling so far is that just dropping performance votes
during runtime/system suspend just makes the entire situation even more
confusing.
Well, that's what most subsystems/drivers need to do.
Moreover, we have specific devices that only use one default OPP [1].
quoted
quoted
quoted
What happens if a driver calls dev_pm_genpd_set_performance_state(...)
while the device is suspended? Will that mess up the performance state
when the device resumes?
Good question. The idea is:
If genpd in genpd_runtime_suspend() are able to drop an existing vote
for a performance state, it should restore the vote in
genpd_runtime_resume(). This also means, if there is no vote to drop
in genpd_runtime_suspend(), genpd should just leave the vote as is in
genpd_runtime_resume().
But the next time the device enters runtime suspend that vote would be
dropped, wouldn't it? That feels kind of strange to me.
What do you mean by "next time"?
Basically just like:
<device runtime-suspended>
driver does dev_pm_genpd_set_performance_state(...)
- performance state is applied immediately, even though device does
apparently not actually want the power domain to be powered on
<device runtime resumes>
- performance state is kept
<device runtime suspends>
- performance state is dropped
Yep, this is what would happen.
quoted
...
I'm not saying this example makes sense (it doesn't for me). It doesn't
make sense to vote for a performance state while runtime suspended.
But with this patch series we still allow that, and it will kind of
produce inconsistent behavior that the performance state is applied
immediately, even though the device is currently runtime-suspended.
But once it runtime suspends again, suddenly it is dropped.
Yes.
Note that, I have been looking at the existing callers of
dev_pm_genpd_set_performance_state() in the kernel as of today. It
should not be an issue, at least as far as I can tell.
quoted
And when you say:
quoted
My main point is, if the device enters runtime suspend state, why
should we keep the vote for an OPP for the device? I mean, the device
isn't going to be used anyway.
A very similar point would be: "If the device *is* in runtime suspend
state, why should we take a vote for an OPP for the device?"
But I understand that this might be something we should address
separately in a follow-up patch/discussion. Don't get me wrong, I agree
this patch set is good, I just think we should go one step further and
finally make this consistent and less prone to side effects.
I agree. We should look into how to change the behaviour. I intend to
have a look at it in a while.
Great, thanks!
quoted
quoted
A good first step might be something like a WARN_ON_ONCE(...) if a
device tries to vote for a performance state while runtime suspended.
Then we might get a clearer picture which drivers do that currently.
That's an idea we could try, even if the number of users are quite
limited today. I can try the "git grep" analyze-method, I will
probably find most of them.
The current user of "required-opps" for CPU DVFS (just qcom/qcs404.dtsi
with qcom/cpr.c I think?) is definitely broken (never votes to turn on
the power domain). So one requirement for making that change of behavior
is figuring out how to deal with enabling power domains at the OPP core
(or whereever else).
quoted
That said, are you okay that we move forward with the $subject series
(except patch4)?
It sounds fine to me.
All right.
So patches [1-3/4] have been applied as 5.14 material, thanks!