cpuidle: Add a predict callback for the governors
Subsystems (2)
PATCH 1/1 cpuidle: Add a predict callback for the governors
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
--- drivers/cpuidle/cpuidle.c | 16 ++++++++++++++++ include/linux/cpuidle.h | 16 +++++++++++++++- kernel/sched/idle.c | 5 +++++ 3 files changed, 36 insertions(+), 1 deletion(-)
drivers/cpuidle/cpuidle.c +16 -0
@@ -296,6 +296,22 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv,return entered_state; } +/** +* cpuidle_predict - guess estimate the next events on the current CPU +* +* @drv: the cpuidle driver +* @dev: the cpuidle device +* +* Returns 0 on success, < 0 in case of error. The error code depends on +* the governor. +*/ +int cpuidle_predict (struct cpuidle_driver *drv,struct cpuidle_device *dev) +{ +if (cpuidle_curr_governor->predict) +return cpuidle_curr_governor->predict (drv, dev); +return 0 ; +} +/** * cpuidle_select - ask the cpuidle framework to choose an idle state *
include/linux/cpuidle.h +15 -1
@@ -77,6 +77,13 @@ struct cpuidle_device_kobj;struct cpuidle_state_kobj ;struct cpuidle_driver_kobj ; +struct cpuidle_predict { +ktime_t next_hrtimer; +ktime_t next_timer; +ktime_t next_irq; +ktime_t next_resched; +}; +struct cpuidle_device {unsigned int registered:1 ;unsigned int enabled:1 ;
@@ -89,6 +96,7 @@ struct cpuidle_device {struct cpuidle_state_kobj *kobjs[CPUIDLE_STATE_MAX ];struct cpuidle_driver_kobj *kobj_driver;struct cpuidle_device_kobj *kobj_dev; +struct cpuidle_predict predict;struct list_head device_list;#ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED
@@ -124,7 +132,8 @@ struct cpuidle_driver {extern void disable_cpuidle (void );extern bool cpuidle_not_available (struct cpuidle_driver *drv,struct cpuidle_device *dev); - +extern int cpuidle_predict (struct cpuidle_driver *drv, +struct cpuidle_device *dev);extern int cpuidle_select (struct cpuidle_driver *drv,struct cpuidle_device *dev,bool *stop_tick);
@@ -158,6 +167,9 @@ static inline void disable_cpuidle(void) { }static inline bool cpuidle_not_available (struct cpuidle_driver *drv,struct cpuidle_device *dev) {return true; } +static inline int cpuidle_predict (struct cpuidle_driver *drv, +struct cpuidle_device *dev) +{return 0 ; }static inline int cpuidle_select (struct cpuidle_driver *drv,struct cpuidle_device *dev,bool *stop_tick) {return -ENODEV ; }
@@ -241,6 +253,8 @@ struct cpuidle_governor {void (*disable) (struct cpuidle_driver *drv,struct cpuidle_device *dev); +int (*predict) (struct cpuidle_driver *drv, +struct cpuidle_device *dev);int (*select) (struct cpuidle_driver *drv,struct cpuidle_device *dev,bool *stop_tick);
kernel/sched/idle.c +5 -0
@@ -185,6 +185,11 @@ static void cpuidle_idle_call(void) }else {bool stop_tick = true; +/* +* Guess estimate the next events on the current CPU +*/ +cpuidle_predict (drv, dev); +/* * Ask the cpuidle framework to choose a convenient idle state. */
Signature · 1 line
2.17.1
Rafael J. Wysocki
maintainerQuoted 5 lines · Daniel Lezcano
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
Quoted 8 lines
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
Can you explain a bit how you would use this new callback in a governor?
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Daniel Lezcano
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
Quoted 4 lines
Quoted 2 lines
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
Can you explain a bit how you would use this new callback in a governor?
Sure.
Today we have the selection and the prediction tied together. The prediction is modulated with some inputs coming from the governor's policy (eg. performance multiplier).
It is hard to know if the prediction is correct or not, hard to know the duration of the computation for the next event and hard to know if the idle state selection succeeded because of a good prediction or a good governor policy.
I propose to provide the callback where we fill the guess estimated next events on the system, so we can trace them and benchmark the computation time.
The selection of the idle state becomes an separate action where we can apply any specific governor heuristic or policy.
By separating the selection and the prediction, we can identify where the weakness is in our test scenario: the prediction or the governor selection policy.
Signature · 5 lines
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | <http://twitter.com/#!/linaroorg> Twitter | <http://www.linaro.org/linaro-blog/> Blog
Rafael J. Wysocki
maintainerQuoted 35 lines · Daniel Lezcano
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Daniel Lezcano
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
Quoted 4 lines
Quoted 2 lines
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
Can you explain a bit how you would use this new callback in a governor?
Sure.
Today we have the selection and the prediction tied together. The prediction is modulated with some inputs coming from the governor's policy (eg. performance multiplier).
It is hard to know if the prediction is correct or not, hard to know the duration of the computation for the next event and hard to know if the idle state selection succeeded because of a good prediction or a good governor policy.
I propose to provide the callback where we fill the guess estimated next events on the system, so we can trace them and benchmark the computation time.
Quoted 6 lines
The selection of the idle state becomes an separate action where we can apply any specific governor heuristic or policy.
By separating the selection and the prediction, we can identify where the weakness is in our test scenario: the prediction or the governor selection policy.
I'm not quite convinced about the idea that the "prediction" and "selection" parts can be entirely separate.
Besides, this new callback doesn't return anything, it goes before ->select and the latter is invoked unconditionally anyway. That's extra overhead (even if small) for no real gain. I don't see why it is better to do the split in the core rather than in the governor itself. You can always design a governor ->select() to call two separate functions, one after another, internally, so why to do that in the core and inflict that particular design choice on everybody?
For example, there's no "predict" part running before the "select" one in the TEO governor at all. There is something like that in menu, but splitting it off would be rather artificial IMO.
Next, the cpuidle_predict structure. I'm not even sure why it is there at all.
Presumably, the purpose of it is to pass some data from ->predict to ->select, but why does it have to be there in struct cpuidle_device? The governor appears to be the only user of it anyway, so wouldn't it be better to store that information in the governor's internal data structures if really needed?
Moreover, why does it include the next_hrtimer and next_timer fields? It looks like you'd need to do some nohz code surgery to separate the two and would that be useful even? And even so, these values need not be predicted. It is known when the next timer event is going to occur, both with and without the scheduler tick included, and that's why there is tick_nohz_get_sleep_length().
If you want to do a similar thing for non-timer interrupts in order to use the IRQ stats, you can simply provide a function similar to tick_nohz_get_sleep_length() for that, but that's available already, isn't it?
Also, I'm not really sure what next_resched is and how exactly it is going to be computed.
Thanks,
Rafael
Hi Rafael,
Quoted 47 lines · Rafael J. Wysocki
Quoted 35 lines · Daniel Lezcano
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Daniel Lezcano
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
Quoted 4 lines
Quoted 2 lines
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
Can you explain a bit how you would use this new callback in a governor?
Sure.
Today we have the selection and the prediction tied together. The prediction is modulated with some inputs coming from the governor's policy (eg. performance multiplier).
It is hard to know if the prediction is correct or not, hard to know the duration of the computation for the next event and hard to know if the idle state selection succeeded because of a good prediction or a good governor policy.
I propose to provide the callback where we fill the guess estimated next events on the system, so we can trace them and benchmark the computation time.
Quoted 6 lines
The selection of the idle state becomes an separate action where we can apply any specific governor heuristic or policy.
By separating the selection and the prediction, we can identify where the weakness is in our test scenario: the prediction or the governor selection policy.
I'm not quite convinced about the idea that the "prediction" and "selection" parts can be entirely separate.
Quoted 7 lines
Besides, this new callback doesn't return anything, it goes before ->select and the latter is invoked unconditionally anyway. That's extra overhead (even if small) for no real gain. I don't see why it is better to do the split in the core rather than in the governor itself. You can always design a governor ->select() to call two separate functions, one after another, internally, so why to do that in the core and inflict that particular design choice on everybody?
It is a way to clearly identify what is part of the prediction and what is part of the decision in order to easily spot the weakness of the governor. We may be doing good predictions but bad decisions or the opposite.
I agree we are not forced to create a new callback for this and we can create a prediction function directly inside the governor. I don't have a strong preference actually and I'm fine with your proposal.
Quoted 11 lines
For example, there's no "predict" part running before the "select" one in the TEO governor at all. There is something like that in menu, but splitting it off would be rather artificial IMO.
Next, the
cpuidle_predictstructure. I'm not even sure why it is there at all.Presumably, the purpose of it is to pass some data from ->predict to ->select, but why does it have to be there in struct
cpuidle_device? The governor appears to be the only user of it anyway, so wouldn't it be better to store that information in the governor's internal data structures if really needed?
At one moment we will need the prediction information from the scheduler in order to optimize the cpu selection for the wakeup. I thought the cpuidle device structure can be a place where we can store it.
There are a lot of optimizations we can do after knowing when a CPU is expected to wakeup.
Do you have a suggestion where to store the next wakeup for a CPU?
Quoted 4 lines
Moreover, why does it include the
next_hrtimerandnext_timerfields? It looks like you'd need to do some nohz code surgery to separate the two and would that be useful even? And even so, these values need not be predicted.
I agree they don't need to be predicted but they are part of the sources of wake up with a deterministic behavior and fall under the prediction umbrella as they are part of the equation for the next event.
The purpose of filling these fields is to give the select callback all the clues to take its decision.
But regarding your POV, which is valid, we can consider using an internal prediction function and just export the next event without a full description of the different wake up source categories.
Quoted 3 lines
It is known when the next timer event is going to
occur, both with and without the scheduler tick included, and that's why there istick_nohz_get_sleep_length().
The next hrtimer and the next timer are the deadline versions of tick_nohz_get_sleep_length(), respectively the delta_time parameter and the returned value.
I massaged the tick_nohz_get_sleep_length() to return the deadline rather than the remaining time. The changes will come with the patches Ulf is about to send. IMHO, it helps to understand the code by splitting in two functions rather than passing an extra parameter.
Quoted 4 lines
If you want to do a similar thing for non-timer interrupts in order to use the IRQ stats, you can simply provide a function similar to
tick_nohz_get_sleep_length()for that, but that's available already, isn't it?
Actually, I'm interested in deadlines not in relative remaining time because we need later to compare those when we are about to wake up a CPU or enter the cluster state.
Indeed, the irq timings is already available but with the limitation of predicting regular intervals. I rewrote the code to handle both regular and repeating pattern. I'll post the series as soon as I have the numbers.
Quoted 2 lines
Also, I'm not really sure what
next_reschedis and how exactly it is going to be computed.
The next_resched estimation is to handle the situation where you have one CPU handling IO related interrupts for a specific device waking up another CPU with the io blocked task. It is very similar to the avg_idle value.
Signature · 5 lines
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | <http://twitter.com/#!/linaroorg> Twitter | <http://www.linaro.org/linaro-blog/> Blog
Rafael J. Wysocki
maintainerHi Daniel,
Quoted 81 lines · Daniel Lezcano
Hi Rafael,
Quoted 47 lines · Rafael J. Wysocki
Quoted 35 lines · Daniel Lezcano
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Daniel Lezcano
Predicting the next event on the current CPU is implemented in the idle state selection function, thus the selection logic and the prediction are tied together and it is hard to decorrelate both.
The following change introduces the cpuidle function to give the opportunity to the governor to store the guess estimate of the different source of wakeup and then reuse them in the selection process. Consequently we end up with two separate operations clearly identified.
Quoted 4 lines
Quoted 2 lines
As the next events are stored in the cpuidle device structure it is easy to propagate them in the different governor callbacks.
Can you explain a bit how you would use this new callback in a governor?
Sure.
Today we have the selection and the prediction tied together. The prediction is modulated with some inputs coming from the governor's policy (eg. performance multiplier).
It is hard to know if the prediction is correct or not, hard to know the duration of the computation for the next event and hard to know if the idle state selection succeeded because of a good prediction or a good governor policy.
I propose to provide the callback where we fill the guess estimated next events on the system, so we can trace them and benchmark the computation time.
Quoted 6 lines
The selection of the idle state becomes an separate action where we can apply any specific governor heuristic or policy.
By separating the selection and the prediction, we can identify where the weakness is in our test scenario: the prediction or the governor selection policy.
I'm not quite convinced about the idea that the "prediction" and "selection" parts can be entirely separate.
Quoted 7 lines
Besides, this new callback doesn't return anything, it goes before ->select and the latter is invoked unconditionally anyway. That's extra overhead (even if small) for no real gain. I don't see why it is better to do the split in the core rather than in the governor itself. You can always design a governor ->select() to call two separate functions, one after another, internally, so why to do that in the core and inflict that particular design choice on everybody?
It is a way to clearly identify what is part of the prediction and what is part of the decision in order to easily spot the weakness of the governor. We may be doing good predictions but bad decisions or the opposite.
I agree we are not forced to create a new callback for this and we can create a prediction function directly inside the governor. I don't have a strong preference actually and I'm fine with your proposal.
Quoted 11 lines
For example, there's no "predict" part running before the "select" one in the TEO governor at all. There is something like that in menu, but splitting it off would be rather artificial IMO.
Next, the
cpuidle_predictstructure. I'm not even sure why it is there at all.Presumably, the purpose of it is to pass some data from ->predict to ->select, but why does it have to be there in struct
cpuidle_device? The governor appears to be the only user of it anyway, so wouldn't it be better to store that information in the governor's internal data structures if really needed?
Quoted 8 lines
At one moment we will need the prediction information from the scheduler in order to optimize the cpu selection for the wakeup. I thought the cpuidle device structure can be a place where we can store it.
There are a lot of optimizations we can do after knowing when a CPU is expected to wakeup.
Do you have a suggestion where to store the next wakeup for a CPU?
There is no need to store it in the core data structures unless different parts of the framework need it.
For example, if both the governor and the driver used that value, it would make sense to store it in struct cpuidle_device (I believe that Ulf has a use case going it that direction), but if it is used by the governor alone, it should be internal to the governor IMO.
Quoted 11 lines
Quoted 4 lines
Moreover, why does it include the
next_hrtimerandnext_timerfields? It looks like you'd need to do some nohz code surgery to separate the two and would that be useful even? And even so, these values need not be predicted.I agree they don't need to be predicted but they are part of the sources of wake up with a deterministic behavior and fall under the prediction umbrella as they are part of the equation for the next event.
The purpose of filling these fields is to give the select callback all the clues to take its decision.
They still should be internal to the governor as long as it is the only user of them.
Quoted 3 lines
But regarding your POV, which is valid, we can consider using an internal prediction function and just export the next event without a full description of the different wake up source categories.
Right.
Quoted 7 lines
Quoted 3 lines
It is known when the next timer event is going to
occur, both with and without the scheduler tick included, and that's why there istick_nohz_get_sleep_length().The next hrtimer and the next timer are the deadline versions of
tick_nohz_get_sleep_length(), respectively thedelta_timeparameter and the returned value.
I would give them different names then.
The return value is the one without the scheduler tick and the delta_time is the value including it. They both cover all timers (high-res and the other).
Quoted 2 lines
I massaged the
tick_nohz_get_sleep_length()to return the deadline rather than the remaining time.
I'm not sure what you mean by the "deadline". Do you add anything to them before returning? That may not be correct.
Quoted 3 lines
The changes will come with the patches
Ulf is about to send. IMHO, it helps to understand the code by splitting in two functions rather than passing an extra parameter.
OK, so those changes need to be submitted first.
Quoted 8 lines
Quoted 4 lines
If you want to do a similar thing for non-timer interrupts in order to use the IRQ stats, you can simply provide a function similar to
tick_nohz_get_sleep_length()for that, but that's available already, isn't it?Actually, I'm interested in deadlines not in relative remaining time because we need later to compare those when we are about to wake up a CPU or enter the cluster state.
The deadlines can be produced easily by adding the current value of the local CPU clock to the delta. If you return the deadline, however, it may not be comparable with the local clock.
Quoted 5 lines
Indeed, the irq timings is already available but with the limitation of predicting regular intervals. I rewrote the code to handle both regular and repeating pattern. I'll post the series as soon as I have the numbers.
Quoted 8 lines
Quoted 2 lines
Also, I'm not really sure what
next_reschedis and how exactly it is going to be computed.The
next_reschedestimation is to handle the situation where you have one CPU handling IO related interrupts for a specific device waking up another CPU with the io blocked task. It is very similar to theavg_idlevalue.
So you want to estimate when something like that may happen for the CPU going idle at the moment?