[v4,1/6] Documentation: arm: define DT idle states bindings

ARM based platforms implement a variety of power management schemes that
allow processors to enter idle states at run-time.
The parameters defining these idle states vary on a per-platform basis forcing
the OS to hardcode the state parameters in platform specific static tables
whose size grows as the number of platforms supported in the kernel increases
and hampers device drivers standardization.

Therefore, this patch aims at standardizing idle state device tree bindings for
ARM platforms. Bindings define idle state parameters inclusive of entry methods
and state latencies, to allow operating systems to retrieve the configuration
entries from the device tree and initialize the related power management
drivers, paving the way for common code in the kernel to deal with idle
states and removing the need for static data in current and previous kernel
versions.

Reviewed-by: Sebastian Capella <sebcape@gmail.com>
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
---
 Documentation/devicetree/bindings/arm/cpus.txt     |   8 +
 .../devicetree/bindings/arm/idle-states.txt        | 507 +++++++++++++++++++++
 2 files changed, 515 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/arm/idle-states.txt

On Wed, 11 Jun 2014, Lorenzo Pieralisi wrote:

> ARM based platforms implement a variety of power management schemes that
> allow processors to enter idle states at run-time.
> The parameters defining these idle states vary on a per-platform basis forcing
> the OS to hardcode the state parameters in platform specific static tables
> whose size grows as the number of platforms supported in the kernel increases
> and hampers device drivers standardization.
> 
> Therefore, this patch aims at standardizing idle state device tree bindings for
> ARM platforms. Bindings define idle state parameters inclusive of entry methods
> and state latencies, to allow operating systems to retrieve the configuration
> entries from the device tree and initialize the related power management
> drivers, paving the way for common code in the kernel to deal with idle
> states and removing the need for static data in current and previous kernel
> versions.

Following the offline discussion with Charles, I've some comments.

[...]

> +Idle state parameters (eg entry latency) are platform specific and 
need to be
> +characterized with bindings that provide the required information to OSPM
> +code so that it can build the required tables and use them at runtime.

[...]

> +	- entry-latency-us
> +		Usage: Required
> +		Value type: <prop-encoded-array>
> +		Definition: u32 value representing worst case latency
> +			    in microseconds required to enter the idle state.
> +
> +	- exit-latency-us
> +		Usage: Required
> +		Value type: <prop-encoded-array>
> +		Definition: u32 value representing worst case latency
> +			    in microseconds required to exit the idle state.
> +
> +	- min-residency-us
> +		Usage: Required
> +		Value type: <prop-encoded-array>
> +		Definition: u32 value representing duration in microseconds
> +			    after which this state becomes more energy
> +			    efficient than any shallower states.

I think this would benefit from a clearer definition.  For example, 
should the min-residency-us value include or exclude the entry and exit 
delays?  I think it should since that's what the cpuidle code will have 
to use when testing against expected delay before next wakeup event in 
any case.  Some of your examples don't assume it is the case though, as 
the min-residency-us is smaller than entry+exit delays.

Also I think we'd need a 4th value to fully characterize a state: worst 
case wake-up latency for QoS purposes.

Let's illustrate the different periods on a time line to make it clearer
(hmmm let's see how this can be managed on a braille display :-O ):

EXEC:	Normal CPU execution.

PREP:	Preparation phase before committing the hardware to idle mode
	like cache flushing. This is abortable on pending wake-up 
	event conditions. The abort latency is assumed to be negligible 
	(i.e. less than the ENTRY + EXIT duration). If aborted, we go 
	back to EXEC. This phase is optional. If not abortable, this 
	should be included in the ENTRY phase instead.

ENTRY:	The hardware is committed to idle mode. This period must run to
	completion up to IDLE before anything else can happen.

IDLE:	This is the actual power-saving idle period. This may last 
	between 0 and infinite time, until a wake-up event occurs.

EXIT:	Period during which the CPU is brought back to operational
	mode (EXEC).

...__[EXEC]__|__[PREP]--|__[ENTRY]__|__[IDLE]__|___[EXIT]_--|__[EXEC]__...
             |          |           |          |            |

             |<-- entry-latency --->|

                                               |<- exit-  ->|
                                               |  latency   |

             |<-------------- min-residency --------------->|

                        |<----- worst_wakeup_latency ------>|

entry-latency: Worst case latency required to enter the idle state.  The 
exit_latency may be guaranteed only after entry-latency has passed.

min-residency: Minimum period, including preparation, entry and exit, 
for a given power mode to be worthwhile energy wise.  It must be at 
least equal to entry_latency + exit_latency.

worst_wakeup_latency: Maximum delay between the signaling of a wake-up 
event and the CPU being able to execute normal code again. If not 
specified, this is assumed to be entry-latency + exit_latency.

Notes:

The cpuidle code would only care about min-residency to select the most 
appropriate mode based on the expected delay before the next event.

The scheduler will care about the following in the near future:

wakeup_delay = exit_latency + max(entry_latency - (now - entry_timestamp), 0)

In other words, the scheduler would wake up the CPU with the shortest 
wake-up latency.  This wake-up latency must take into account the entry 
latency if that period has not expired.  Here the abortable nature of 
the PREP period is ignored on purpose because it cannot be relied upon 
(e.g. if the cache is mostly clean then the PREP deadline may occur much 
sooner than expected).

And pmqos would only care about worst_wakeup_latency.

So... I hope this is useful.  I think the above ascii art could be part 
of your documentation to explain it all.



Nicolas
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Wed, Jun 11, 2014 at 07:15:16PM +0100, Nicolas Pitre wrote:
> On Wed, 11 Jun 2014, Lorenzo Pieralisi wrote:
> 
> > ARM based platforms implement a variety of power management schemes that
> > allow processors to enter idle states at run-time.
> > The parameters defining these idle states vary on a per-platform basis forcing
> > the OS to hardcode the state parameters in platform specific static tables
> > whose size grows as the number of platforms supported in the kernel increases
> > and hampers device drivers standardization.
> > 
> > Therefore, this patch aims at standardizing idle state device tree bindings for
> > ARM platforms. Bindings define idle state parameters inclusive of entry methods
> > and state latencies, to allow operating systems to retrieve the configuration
> > entries from the device tree and initialize the related power management
> > drivers, paving the way for common code in the kernel to deal with idle
> > states and removing the need for static data in current and previous kernel
> > versions.
> 
> Following the offline discussion with Charles, I've some comments.
> 
> [...]

Thank you for summing that discussion up.

> > +Idle state parameters (eg entry latency) are platform specific and 
> need to be
> > +characterized with bindings that provide the required information to OSPM
> > +code so that it can build the required tables and use them at runtime.
> 
> [...]
> 
> > +	- entry-latency-us
> > +		Usage: Required
> > +		Value type: <prop-encoded-array>
> > +		Definition: u32 value representing worst case latency
> > +			    in microseconds required to enter the idle state.
> > +
> > +	- exit-latency-us
> > +		Usage: Required
> > +		Value type: <prop-encoded-array>
> > +		Definition: u32 value representing worst case latency
> > +			    in microseconds required to exit the idle state.
> > +
> > +	- min-residency-us
> > +		Usage: Required
> > +		Value type: <prop-encoded-array>
> > +		Definition: u32 value representing duration in microseconds
> > +			    after which this state becomes more energy
> > +			    efficient than any shallower states.
> 
> I think this would benefit from a clearer definition.  For example, 
> should the min-residency-us value include or exclude the entry and exit 
> delays?  I think it should since that's what the cpuidle code will have 
> to use when testing against expected delay before next wakeup event in 
> any case.  Some of your examples don't assume it is the case though, as 
> the min-residency-us is smaller than entry+exit delays.
> 
> Also I think we'd need a 4th value to fully characterize a state: worst 
> case wake-up latency for QoS purposes.
> 
> Let's illustrate the different periods on a time line to make it clearer
> (hmmm let's see how this can be managed on a braille display :-O ):
> 
> EXEC:	Normal CPU execution.
> 
> PREP:	Preparation phase before committing the hardware to idle mode
> 	like cache flushing. This is abortable on pending wake-up 
> 	event conditions. The abort latency is assumed to be negligible 
> 	(i.e. less than the ENTRY + EXIT duration). If aborted, we go 
> 	back to EXEC. This phase is optional. If not abortable, this 
> 	should be included in the ENTRY phase instead.
> 
> ENTRY:	The hardware is committed to idle mode. This period must run to
> 	completion up to IDLE before anything else can happen.
> 
> IDLE:	This is the actual power-saving idle period. This may last 
> 	between 0 and infinite time, until a wake-up event occurs.
> 
> EXIT:	Period during which the CPU is brought back to operational
> 	mode (EXEC).
> 
> ...__[EXEC]__|__[PREP]--|__[ENTRY]__|__[IDLE]__|___[EXIT]_--|__[EXEC]__...
>              |          |           |          |            |
> 
>              |<-- entry-latency --->|
> 
>                                                |<- exit-  ->|
>                                                |  latency   |
> 
>              |<-------------- min-residency --------------->|
> 
>                         |<----- worst_wakeup_latency ------>|
> 
> entry-latency: Worst case latency required to enter the idle state.  The 
> exit_latency may be guaranteed only after entry-latency has passed.
> 
> min-residency: Minimum period, including preparation, entry and exit, 
> for a given power mode to be worthwhile energy wise.  It must be at 
> least equal to entry_latency + exit_latency.
> 
> worst_wakeup_latency: Maximum delay between the signaling of a wake-up 
> event and the CPU being able to execute normal code again. If not 
> specified, this is assumed to be entry-latency + exit_latency.
> 
> Notes:
> 
> The cpuidle code would only care about min-residency to select the most 
> appropriate mode based on the expected delay before the next event.
> 
> The scheduler will care about the following in the near future:
> 
> wakeup_delay = exit_latency + max(entry_latency - (now - entry_timestamp), 0)
> 
> In other words, the scheduler would wake up the CPU with the shortest 
> wake-up latency.  This wake-up latency must take into account the entry 
> latency if that period has not expired.  Here the abortable nature of 
> the PREP period is ignored on purpose because it cannot be relied upon 
> (e.g. if the cache is mostly clean then the PREP deadline may occur much 
> sooner than expected).
> 
> And pmqos would only care about worst_wakeup_latency.
> 
> So... I hope this is useful.  I think the above ascii art could be part 
> of your documentation to explain it all.

I will, it makes perfect sense, let me point out a couple of things:

1) we need 4 properties, 1 optional (worst_wakeup_latency, if not
   present defaults to entry+exit)
2) is everyone ok, given these definitions, in sorting idle states using
   min-residency-us as a rank ?
3) CPUidle:
   idle_state.exit_latency = worst-wakeup-latency
   idle_state.target_residency = min-residency-us
4) PREP (longest period) can be obtained from the other properties, IF it is
   needed
   PREP = (entry + exit) - worst_wakeup (if worst_wakeup omitted, PREP = 0)

If everyone agrees I think these bindings updated with Nico's diagram
and definitions (I will tweak them, not change them because they make
perfect sense to me) are ready to go, if anyone has concerns please
drop a comment.

Thank you Nico !
Lorenzo

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Fri, 13 Jun 2014, Lorenzo Pieralisi wrote:

> On Wed, Jun 11, 2014 at 07:15:16PM +0100, Nicolas Pitre wrote:
> > Let's illustrate the different periods on a time line to make it clearer
> > (hmmm let's see how this can be managed on a braille display :-O ):
> > 
> > EXEC:	Normal CPU execution.
> > 
> > PREP:	Preparation phase before committing the hardware to idle mode
> > 	like cache flushing. This is abortable on pending wake-up 
> > 	event conditions. The abort latency is assumed to be negligible 
> > 	(i.e. less than the ENTRY + EXIT duration). If aborted, we go 
> > 	back to EXEC. This phase is optional. If not abortable, this 
> > 	should be included in the ENTRY phase instead.
> > 
> > ENTRY:	The hardware is committed to idle mode. This period must run to
> > 	completion up to IDLE before anything else can happen.
> > 
> > IDLE:	This is the actual power-saving idle period. This may last 
> > 	between 0 and infinite time, until a wake-up event occurs.
> > 
> > EXIT:	Period during which the CPU is brought back to operational
> > 	mode (EXEC).
> > 
> > ...__[EXEC]__|__[PREP]--|__[ENTRY]__|__[IDLE]__|___[EXIT]_--|__[EXEC]__...
> >              |          |           |          |            |
> > 
> >              |<-- entry-latency --->|
> > 
> >                                                |<- exit-  ->|
> >                                                |  latency   |
> > 
> >              |<-------------- min-residency --------------->|
> > 
> >                         |<----- worst_wakeup_latency ------>|
> > 
> > entry-latency: Worst case latency required to enter the idle state.  The 
> > exit_latency may be guaranteed only after entry-latency has passed.
> > 
> > min-residency: Minimum period, including preparation, entry and exit, 
> > for a given power mode to be worthwhile energy wise.  It must be at 
> > least equal to entry_latency + exit_latency.
> > 
> > worst_wakeup_latency: Maximum delay between the signaling of a wake-up 
> > event and the CPU being able to execute normal code again. If not 
> > specified, this is assumed to be entry-latency + exit_latency.
> > 
> > Notes:
> > 
> > The cpuidle code would only care about min-residency to select the most 
> > appropriate mode based on the expected delay before the next event.
> > 
> > The scheduler will care about the following in the near future:
> > 
> > wakeup_delay = exit_latency + max(entry_latency - (now - entry_timestamp), 0)
> > 
> > In other words, the scheduler would wake up the CPU with the shortest 
> > wake-up latency.  This wake-up latency must take into account the entry 
> > latency if that period has not expired.  Here the abortable nature of 
> > the PREP period is ignored on purpose because it cannot be relied upon 
> > (e.g. if the cache is mostly clean then the PREP deadline may occur much 
> > sooner than expected).
> > 
> > And pmqos would only care about worst_wakeup_latency.
> > 
> > So... I hope this is useful.  I think the above ascii art could be part 
> > of your documentation to explain it all.
> 
> I will, it makes perfect sense, let me point out a couple of things:
> 
> 1) we need 4 properties, 1 optional (worst_wakeup_latency, if not
>    present defaults to entry+exit)
> 2) is everyone ok, given these definitions, in sorting idle states using
>    min-residency-us as a rank ?

Yes.

> 3) CPUidle:
>    idle_state.exit_latency = worst-wakeup-latency
>    idle_state.target_residency = min-residency-us

But exit_latency is not necessarily equal to worst-wakeup-latency.  
We'll need any of those 4 values depending on the context.  So I'd add 
entry_latency and worst_wakeup_latency to struct cpuidle_state.  If a 
driver doesn't initialize entry_latency then it can be left to 0, and if 
worst_wakeup_latency is 0 then it should be set to entry_latency + 
exit_latency by the core code.

> 4) PREP (longest period) can be obtained from the other properties, IF it is
>    needed
>    PREP = (entry + exit) - worst_wakeup (if worst_wakeup omitted, PREP = 0)

Sure.  However I'd avoid documenting it.  As I said this period cannot 
be relied upon because it can vary a lot and if you miss its deadline 
you're up for a much longer delay than expected.  It is useful if a 
wake-up event happens during that period and then the latency can be cut 
short opportunistically. But if we get to the point we need to rely on 
this period to improve things then it would be a good idea to question 
why we need to request and immediately abort a state so often to start 
with.


Nicolas
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

I like these too!  Nice job!

Thanks!

Sebastian
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Fri, Jun 13, 2014 at 06:33:35PM +0100, Nicolas Pitre wrote:
> On Fri, 13 Jun 2014, Lorenzo Pieralisi wrote:
> 
> > On Wed, Jun 11, 2014 at 07:15:16PM +0100, Nicolas Pitre wrote:
> > > Let's illustrate the different periods on a time line to make it clearer
> > > (hmmm let's see how this can be managed on a braille display :-O ):
> > > 
> > > EXEC:	Normal CPU execution.
> > > 
> > > PREP:	Preparation phase before committing the hardware to idle mode
> > > 	like cache flushing. This is abortable on pending wake-up 
> > > 	event conditions. The abort latency is assumed to be negligible 
> > > 	(i.e. less than the ENTRY + EXIT duration). If aborted, we go 
> > > 	back to EXEC. This phase is optional. If not abortable, this 
> > > 	should be included in the ENTRY phase instead.
> > > 
> > > ENTRY:	The hardware is committed to idle mode. This period must run to
> > > 	completion up to IDLE before anything else can happen.
> > > 
> > > IDLE:	This is the actual power-saving idle period. This may last 
> > > 	between 0 and infinite time, until a wake-up event occurs.
> > > 
> > > EXIT:	Period during which the CPU is brought back to operational
> > > 	mode (EXEC).
> > > 
> > > ...__[EXEC]__|__[PREP]--|__[ENTRY]__|__[IDLE]__|___[EXIT]_--|__[EXEC]__...
> > >              |          |           |          |            |
> > > 
> > >              |<-- entry-latency --->|
> > > 
> > >                                                |<- exit-  ->|
> > >                                                |  latency   |
> > > 
> > >              |<-------------- min-residency --------------->|
> > > 
> > >                         |<----- worst_wakeup_latency ------>|
> > > 
> > > entry-latency: Worst case latency required to enter the idle state.  The 
> > > exit_latency may be guaranteed only after entry-latency has passed.
> > > 
> > > min-residency: Minimum period, including preparation, entry and exit, 
> > > for a given power mode to be worthwhile energy wise.  It must be at 
> > > least equal to entry_latency + exit_latency.
> > > 
> > > worst_wakeup_latency: Maximum delay between the signaling of a wake-up 
> > > event and the CPU being able to execute normal code again. If not 
> > > specified, this is assumed to be entry-latency + exit_latency.
> > > 
> > > Notes:
> > > 
> > > The cpuidle code would only care about min-residency to select the most 
> > > appropriate mode based on the expected delay before the next event.
> > > 
> > > The scheduler will care about the following in the near future:
> > > 
> > > wakeup_delay = exit_latency + max(entry_latency - (now - entry_timestamp), 0)
> > > 
> > > In other words, the scheduler would wake up the CPU with the shortest 
> > > wake-up latency.  This wake-up latency must take into account the entry 
> > > latency if that period has not expired.  Here the abortable nature of 
> > > the PREP period is ignored on purpose because it cannot be relied upon 
> > > (e.g. if the cache is mostly clean then the PREP deadline may occur much 
> > > sooner than expected).
> > > 
> > > And pmqos would only care about worst_wakeup_latency.
> > > 
> > > So... I hope this is useful.  I think the above ascii art could be part 
> > > of your documentation to explain it all.
> > 
> > I will, it makes perfect sense, let me point out a couple of things:
> > 
> > 1) we need 4 properties, 1 optional (worst_wakeup_latency, if not
> >    present defaults to entry+exit)
> > 2) is everyone ok, given these definitions, in sorting idle states using
> >    min-residency-us as a rank ?
> 
> Yes.
> 
> > 3) CPUidle:
> >    idle_state.exit_latency = worst-wakeup-latency
> >    idle_state.target_residency = min-residency-us
> 
> But exit_latency is not necessarily equal to worst-wakeup-latency.  
> We'll need any of those 4 values depending on the context.  So I'd add 
> entry_latency and worst_wakeup_latency to struct cpuidle_state.  If a 
> driver doesn't initialize entry_latency then it can be left to 0, and if 
> worst_wakeup_latency is 0 then it should be set to entry_latency + 
> exit_latency by the core code.

Well, that's why I mentioned idle_state.exit_latency, because in CPUidle
today, the struct cpuidle_state.exit_latency field corresponds to our
worst-wakeup-latency property, not to the exit_latency property; I know
it is confusing but at least by defining proper bindings the kernel
structures can be updated with clear semantics (I would not rename them
for the time being though). Fields required by the scheduler (ie
entry_latency) can be added in the patches that rely on them, when we agreed
on the bindings, adding the variables is no big deal.

> > 4) PREP (longest period) can be obtained from the other properties, IF it is
> >    needed
> >    PREP = (entry + exit) - worst_wakeup (if worst_wakeup omitted, PREP = 0)
> 
> Sure.  However I'd avoid documenting it.  As I said this period cannot 
> be relied upon because it can vary a lot and if you miss its deadline 
> you're up for a much longer delay than expected.  It is useful if a 
> wake-up event happens during that period and then the latency can be cut 
> short opportunistically. But if we get to the point we need to rely on 
> this period to improve things then it would be a good idea to question 
> why we need to request and immediately abort a state so often to start 
> with.

Agreed.

Thanks,
Lorenzo

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Mon, 16 Jun 2014, Lorenzo Pieralisi wrote:

> On Fri, Jun 13, 2014 at 06:33:35PM +0100, Nicolas Pitre wrote:
> > >    idle_state.exit_latency = worst-wakeup-latency
> > >    idle_state.target_residency = min-residency-us
> > 
> > But exit_latency is not necessarily equal to worst-wakeup-latency.  
> > We'll need any of those 4 values depending on the context.  So I'd add 
> > entry_latency and worst_wakeup_latency to struct cpuidle_state.  If a 
> > driver doesn't initialize entry_latency then it can be left to 0, and if 
> > worst_wakeup_latency is 0 then it should be set to entry_latency + 
> > exit_latency by the core code.
> 
> Well, that's why I mentioned idle_state.exit_latency, because in CPUidle
> today, the struct cpuidle_state.exit_latency field corresponds to our
> worst-wakeup-latency property, not to the exit_latency property; I know
> it is confusing but at least by defining proper bindings the kernel
> structures can be updated with clear semantics (I would not rename them
> for the time being though).

Why not?  Adding more confusion or even simply keeping the existing one, 
even if it is temporary, doesn't benefit anyone.

> Fields required by the scheduler (ie entry_latency) can be added in 
> the patches that rely on them, when we agreed on the bindings, adding 
> the variables is no big deal.

Sure.


Nicolas
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html