Message ID | 20190405091647.4169-2-huntbag@linux.vnet.ibm.com (mailing list archive) |
---|---|
State | Changes Requested, archived |
Headers | show |
Series | Auto-promotion logic for cpuidle states | expand |
Hi Abhishek, > Currently, the cpuidle governors (menu /ladder) determine what idle state > an idling CPU should enter into based on heuristics that depend on the > idle history on that CPU. Given that no predictive heuristic is perfect, > there are cases where the governor predicts a shallow idle state, hoping > that the CPU will be busy soon. However, if no new workload is scheduled > on that CPU in the near future, the CPU will end up in the shallow state. > > In case of POWER, this is problematic, when the predicted state in the > aforementioned scenario is a lite stop state, as such lite states will > inhibit SMT folding, thereby depriving the other threads in the core from > using the core resources. > > To address this, such lite states need to be autopromoted. The cpuidle- > core can queue timer to correspond with the residency value of the next > available state. Thus leading to auto-promotion to a deeper idle state as > soon as possible. > This sounds sensible to me, although I'm not really qualified to offer a full power-management opinion on it. I have some general code questions and comments, however, which are below: > Signed-off-by: Abhishek Goel <huntbag@linux.vnet.ibm.com> > --- > > v1->v2 : Removed timeout_needed and rebased to current upstream kernel > > drivers/cpuidle/cpuidle.c | 68 +++++++++++++++++++++++++++++- > drivers/cpuidle/governors/ladder.c | 3 +- > drivers/cpuidle/governors/menu.c | 22 +++++++++- > include/linux/cpuidle.h | 10 ++++- > 4 files changed, 99 insertions(+), 4 deletions(-) > > diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c > index 7f108309e..11ce43f19 100644 > --- a/drivers/cpuidle/cpuidle.c > +++ b/drivers/cpuidle/cpuidle.c > @@ -36,6 +36,11 @@ static int enabled_devices; > static int off __read_mostly; > static int initialized __read_mostly; > > +struct auto_promotion { > + struct hrtimer hrtimer; > + unsigned long timeout_us; > +}; > + > int cpuidle_disabled(void) > { > return off; > @@ -188,6 +193,54 @@ int cpuidle_enter_s2idle(struct cpuidle_driver *drv, struct cpuidle_device *dev) > } > #endif /* CONFIG_SUSPEND */ > > +enum hrtimer_restart auto_promotion_hrtimer_callback(struct hrtimer *hrtimer) > +{ > + return HRTIMER_NORESTART; > +} > + > +#ifdef CONFIG_CPU_IDLE_AUTO_PROMOTION As far as I can tell, this config flag isn't defined until the next patch, making this dead code for now. Is this intentional? > +DEFINE_PER_CPU(struct auto_promotion, ap); A quick grep suggests that most per-cpu variable have more descriptive names, perhaps this one should too. > + > +static void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state *state) > +{ > + struct auto_promotion *this_ap = &per_cpu(ap, cpu); > + > + if (state->flags & CPUIDLE_FLAG_AUTO_PROMOTION) > + hrtimer_start(&this_ap->hrtimer, ns_to_ktime(this_ap->timeout_us > + * 1000), HRTIMER_MODE_REL_PINNED); Would it be clearer to have both sides of the multiplication on the same line? i.e. + hrtimer_start(&this_ap->hrtimer, + ns_to_ktime(this_ap->timeout_us * 1000), + HRTIMER_MODE_REL_PINNED); > +} > + > +static void cpuidle_auto_promotion_cancel(int cpu) > +{ > + struct hrtimer *hrtimer; > + > + hrtimer = &per_cpu(ap, cpu).hrtimer; > + if (hrtimer_is_queued(hrtimer)) > + hrtimer_cancel(hrtimer); > +} > + > +static void cpuidle_auto_promotion_update(int cpu, unsigned long timeout) > +{ > + per_cpu(ap, cpu).timeout_us = timeout; > +} > + > +static void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver *drv) > +{ > + struct auto_promotion *this_ap = &per_cpu(ap, cpu); > + > + hrtimer_init(&this_ap->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); > + this_ap->hrtimer.function = auto_promotion_hrtimer_callback; > +} > +#else > +static inline void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state > + *state) { } > +static inline void cpuidle_auto_promotion_cancel(int cpu) { } > +static inline void cpuidle_auto_promotion_update(int cpu, unsigned long > + timeout) { } > +static inline void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver > + *drv) { } Several of these have the type, then a line break, and then the name (unsigned long\n timeout). This is a bit harder to read, they should probably all be on the same line. > +#endif > + > /** > * cpuidle_enter_state - enter the state and update stats > * @dev: cpuidle device for this cpu > @@ -225,12 +278,17 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, > trace_cpu_idle_rcuidle(index, dev->cpu); > time_start = ns_to_ktime(local_clock()); > > + cpuidle_auto_promotion_start(dev->cpu, target_state); > + > stop_critical_timings(); > entered_state = target_state->enter(dev, drv, index); > start_critical_timings(); > > sched_clock_idle_wakeup_event(); > time_end = ns_to_ktime(local_clock()); > + > + cpuidle_auto_promotion_cancel(dev->cpu); > + > trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); > > /* The cpu is no longer idle or about to enter idle. */ > @@ -312,7 +370,13 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, > int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > bool *stop_tick) > { > - return cpuidle_curr_governor->select(drv, dev, stop_tick); > + unsigned long timeout_us, ret; > + > + timeout_us = UINT_MAX; > + ret = cpuidle_curr_governor->select(drv, dev, stop_tick, &timeout_us); > + cpuidle_auto_promotion_update(dev->cpu, timeout_us); > + > + return ret; > } > > /** > @@ -658,6 +722,8 @@ int cpuidle_register(struct cpuidle_driver *drv, > device = &per_cpu(cpuidle_dev, cpu); > device->cpu = cpu; > > + cpuidle_auto_promotion_init(cpu, drv); > + > #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED > /* > * On multiplatform for ARM, the coupled idle states could be > diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c > index f0dddc66a..65b518dd7 100644 > --- a/drivers/cpuidle/governors/ladder.c > +++ b/drivers/cpuidle/governors/ladder.c > @@ -64,7 +64,8 @@ static inline void ladder_do_selection(struct ladder_device *ldev, > * @dummy: not used I think you need an addition to the docstring for your new variable. > */ > static int ladder_select_state(struct cpuidle_driver *drv, > - struct cpuidle_device *dev, bool *dummy) > + struct cpuidle_device *dev, bool *dummy, > + unsigned long *unused) > { > struct ladder_device *ldev = this_cpu_ptr(&ladder_devices); > struct ladder_device_state *last_state; > diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c > index 5951604e7..835e337de 100644 > --- a/drivers/cpuidle/governors/menu.c > +++ b/drivers/cpuidle/governors/menu.c > @@ -276,7 +276,7 @@ static unsigned int get_typical_interval(struct menu_device *data, > * @stop_tick: indication on whether or not to stop the tick Likewise here. > */ > static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > - bool *stop_tick) > + bool *stop_tick, unsigned long *timeout) > { > struct menu_device *data = this_cpu_ptr(&menu_devices); > int latency_req = cpuidle_governor_latency_req(dev->cpu); > @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > } > } > > +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION > + if (drv->states[idx].flags & CPUIDLE_FLAG_AUTO_PROMOTION) { > + /* > + * Timeout is intended to be defined as sum of target residency > + * of next available state, entry latency and exit latency. If > + * time interval equal to timeout is spent in current state, > + * and if it is a shallow lite state, we may want to auto- > + * promote from such state. This comment makes sense if you already understand auto-promotion. That's fair enough - you wrote it and you presumably understand what your code does :) But for me it's a bit confusing! I think you want to start with a sentence about what autopromotion is (preferably not using power-specific terminology) and then explain the calculation of the timeouts. > + */ > + for (i = idx + 1; i < drv->state_count; i++) { > + if (drv->states[i].disabled || > + dev->states_usage[i].disable) > + continue; > + *timeout = drv->states[i].target_residency + > + 2 * drv->states[i].exit_latency; > + break; > + } > + } > +#endif > + > return idx; > } > > diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h > index 3b3947232..84d76d1ec 100644 > --- a/include/linux/cpuidle.h > +++ b/include/linux/cpuidle.h > @@ -72,6 +72,13 @@ struct cpuidle_state { > #define CPUIDLE_FLAG_POLLING BIT(0) /* polling state */ > #define CPUIDLE_FLAG_COUPLED BIT(1) /* state applies to multiple cpus */ > #define CPUIDLE_FLAG_TIMER_STOP BIT(2) /* timer is stopped on this state */ > +/* > + * State with only and only fast state bit set don't even lose user context. "only and only"? > + * But such states prevent other sibling threads from thread folding benefits. > + * And hence we don't want to stay for too long in such states and want to > + * auto-promote from it. I think this comment mixes Power-specific and generic concepts. (But I'm not a PM expert so tell me if I'm wrong here.) I think, if I've understood correctly: in the generic code, the bit represents a state that we do not want to linger in, which we want to definitely leave after some time. On Power, we have a state that doesn't lose user context but which prevents thread folding, so this is an example of a state where we want to auto-promote. > + */ > +#define CPUIDLE_FLAG_AUTO_PROMOTION BIT(3) > > struct cpuidle_device_kobj; > struct cpuidle_state_kobj; > @@ -243,7 +250,8 @@ struct cpuidle_governor { > > int (*select) (struct cpuidle_driver *drv, > struct cpuidle_device *dev, > - bool *stop_tick); > + bool *stop_tick, unsigned long > + *timeout); > void (*reflect) (struct cpuidle_device *dev, int index); > }; > > -- > 2.17.1
Hi, Sorry, just realised another thing I wanted to ask: > @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > } > } > > > +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION Why is this based on CPUIDLE_FLAG_ rather than CONFIG_CPU_IDLE_? Won't this always be true, given that the flag is defined regardless of the config option in the header? > + if (drv->states[idx].flags & CPUIDLE_FLAG_AUTO_PROMOTION) { > + /* > + * Timeout is intended to be defined as sum of target residency > + * of next available state, entry latency and exit latency. If > + * time interval equal to timeout is spent in current state, > + * and if it is a shallow lite state, we may want to auto- > + * promote from such state. > + */ Regards, Daniel > + for (i = idx + 1; i < drv->state_count; i++) { > + if (drv->states[i].disabled || > + dev->states_usage[i].disable) > + continue; > + *timeout = drv->states[i].target_residency + > + 2 * drv->states[i].exit_latency; > + break; > + } > + } > +#endif > + > return idx; > } > > diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h > index 3b3947232..84d76d1ec 100644 > --- a/include/linux/cpuidle.h > +++ b/include/linux/cpuidle.h > @@ -72,6 +72,13 @@ struct cpuidle_state { > #define CPUIDLE_FLAG_POLLING BIT(0) /* polling state */ > #define CPUIDLE_FLAG_COUPLED BIT(1) /* state applies to multiple cpus */ > #define CPUIDLE_FLAG_TIMER_STOP BIT(2) /* timer is stopped on this state */ > +/* > + * State with only and only fast state bit set don't even lose user context. > + * But such states prevent other sibling threads from thread folding benefits. > + * And hence we don't want to stay for too long in such states and want to > + * auto-promote from it. > + */ > +#define CPUIDLE_FLAG_AUTO_PROMOTION BIT(3) > > struct cpuidle_device_kobj; > struct cpuidle_state_kobj; > @@ -243,7 +250,8 @@ struct cpuidle_governor { > > int (*select) (struct cpuidle_driver *drv, > struct cpuidle_device *dev, > - bool *stop_tick); > + bool *stop_tick, unsigned long > + *timeout); > void (*reflect) (struct cpuidle_device *dev, int index); > }; > > -- > 2.17.1
Hi Daniel, Thanks for such a descriptive review. I will include all the suggestions made in my next iteration. --Abhishek On 04/08/2019 07:42 PM, Daniel Axtens wrote: > Hi Abhishek, > >> Currently, the cpuidle governors (menu /ladder) determine what idle state >> an idling CPU should enter into based on heuristics that depend on the >> idle history on that CPU. Given that no predictive heuristic is perfect, >> there are cases where the governor predicts a shallow idle state, hoping >> that the CPU will be busy soon. However, if no new workload is scheduled >> on that CPU in the near future, the CPU will end up in the shallow state. >> >> In case of POWER, this is problematic, when the predicted state in the >> aforementioned scenario is a lite stop state, as such lite states will >> inhibit SMT folding, thereby depriving the other threads in the core from >> using the core resources. >> >> To address this, such lite states need to be autopromoted. The cpuidle- >> core can queue timer to correspond with the residency value of the next >> available state. Thus leading to auto-promotion to a deeper idle state as >> soon as possible. >> > This sounds sensible to me, although I'm not really qualified to offer a > full power-management opinion on it. I have some general code questions > and comments, however, which are below: > >> Signed-off-by: Abhishek Goel <huntbag@linux.vnet.ibm.com> >> --- >> >> v1->v2 : Removed timeout_needed and rebased to current upstream kernel >> >> drivers/cpuidle/cpuidle.c | 68 +++++++++++++++++++++++++++++- >> drivers/cpuidle/governors/ladder.c | 3 +- >> drivers/cpuidle/governors/menu.c | 22 +++++++++- >> include/linux/cpuidle.h | 10 ++++- >> 4 files changed, 99 insertions(+), 4 deletions(-) >> >> diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c >> index 7f108309e..11ce43f19 100644 >> --- a/drivers/cpuidle/cpuidle.c >> +++ b/drivers/cpuidle/cpuidle.c >> @@ -36,6 +36,11 @@ static int enabled_devices; >> static int off __read_mostly; >> static int initialized __read_mostly; >> >> +struct auto_promotion { >> + struct hrtimer hrtimer; >> + unsigned long timeout_us; >> +}; >> + >> int cpuidle_disabled(void) >> { >> return off; >> @@ -188,6 +193,54 @@ int cpuidle_enter_s2idle(struct cpuidle_driver *drv, struct cpuidle_device *dev) >> } >> #endif /* CONFIG_SUSPEND */ >> >> +enum hrtimer_restart auto_promotion_hrtimer_callback(struct hrtimer *hrtimer) >> +{ >> + return HRTIMER_NORESTART; >> +} >> + >> +#ifdef CONFIG_CPU_IDLE_AUTO_PROMOTION > As far as I can tell, this config flag isn't defined until the next > patch, making this dead code for now. Is this intentional? > >> +DEFINE_PER_CPU(struct auto_promotion, ap); > A quick grep suggests that most per-cpu variable have more descriptive > names, perhaps this one should too. > >> + >> +static void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state *state) >> +{ >> + struct auto_promotion *this_ap = &per_cpu(ap, cpu); >> + >> + if (state->flags & CPUIDLE_FLAG_AUTO_PROMOTION) >> + hrtimer_start(&this_ap->hrtimer, ns_to_ktime(this_ap->timeout_us >> + * 1000), HRTIMER_MODE_REL_PINNED); > Would it be clearer to have both sides of the multiplication on the same > line? i.e. > + hrtimer_start(&this_ap->hrtimer, > + ns_to_ktime(this_ap->timeout_us * 1000), > + HRTIMER_MODE_REL_PINNED); > >> +} >> + >> +static void cpuidle_auto_promotion_cancel(int cpu) >> +{ >> + struct hrtimer *hrtimer; >> + >> + hrtimer = &per_cpu(ap, cpu).hrtimer; >> + if (hrtimer_is_queued(hrtimer)) >> + hrtimer_cancel(hrtimer); >> +} >> + >> +static void cpuidle_auto_promotion_update(int cpu, unsigned long timeout) >> +{ >> + per_cpu(ap, cpu).timeout_us = timeout; >> +} >> + >> +static void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver *drv) >> +{ >> + struct auto_promotion *this_ap = &per_cpu(ap, cpu); >> + >> + hrtimer_init(&this_ap->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); >> + this_ap->hrtimer.function = auto_promotion_hrtimer_callback; >> +} >> +#else >> +static inline void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state >> + *state) { } >> +static inline void cpuidle_auto_promotion_cancel(int cpu) { } >> +static inline void cpuidle_auto_promotion_update(int cpu, unsigned long >> + timeout) { } >> +static inline void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver >> + *drv) { } > Several of these have the type, then a line break, and then the name > (unsigned long\n timeout). This is a bit harder to read, they should > probably all be on the same line. > >> +#endif >> + >> /** >> * cpuidle_enter_state - enter the state and update stats >> * @dev: cpuidle device for this cpu >> @@ -225,12 +278,17 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, >> trace_cpu_idle_rcuidle(index, dev->cpu); >> time_start = ns_to_ktime(local_clock()); >> >> + cpuidle_auto_promotion_start(dev->cpu, target_state); >> + >> stop_critical_timings(); >> entered_state = target_state->enter(dev, drv, index); >> start_critical_timings(); >> >> sched_clock_idle_wakeup_event(); >> time_end = ns_to_ktime(local_clock()); >> + >> + cpuidle_auto_promotion_cancel(dev->cpu); >> + >> trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); >> >> /* The cpu is no longer idle or about to enter idle. */ >> @@ -312,7 +370,13 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, >> int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, >> bool *stop_tick) >> { >> - return cpuidle_curr_governor->select(drv, dev, stop_tick); >> + unsigned long timeout_us, ret; >> + >> + timeout_us = UINT_MAX; >> + ret = cpuidle_curr_governor->select(drv, dev, stop_tick, &timeout_us); >> + cpuidle_auto_promotion_update(dev->cpu, timeout_us); >> + >> + return ret; >> } >> >> /** >> @@ -658,6 +722,8 @@ int cpuidle_register(struct cpuidle_driver *drv, >> device = &per_cpu(cpuidle_dev, cpu); >> device->cpu = cpu; >> >> + cpuidle_auto_promotion_init(cpu, drv); >> + >> #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED >> /* >> * On multiplatform for ARM, the coupled idle states could be >> diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c >> index f0dddc66a..65b518dd7 100644 >> --- a/drivers/cpuidle/governors/ladder.c >> +++ b/drivers/cpuidle/governors/ladder.c >> @@ -64,7 +64,8 @@ static inline void ladder_do_selection(struct ladder_device *ldev, >> * @dummy: not used > I think you need an addition to the docstring for your new variable. > >> */ >> static int ladder_select_state(struct cpuidle_driver *drv, >> - struct cpuidle_device *dev, bool *dummy) >> + struct cpuidle_device *dev, bool *dummy, >> + unsigned long *unused) >> { >> struct ladder_device *ldev = this_cpu_ptr(&ladder_devices); >> struct ladder_device_state *last_state; >> diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c >> index 5951604e7..835e337de 100644 >> --- a/drivers/cpuidle/governors/menu.c >> +++ b/drivers/cpuidle/governors/menu.c >> @@ -276,7 +276,7 @@ static unsigned int get_typical_interval(struct menu_device *data, >> * @stop_tick: indication on whether or not to stop the tick > Likewise here. > >> */ >> static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, >> - bool *stop_tick) >> + bool *stop_tick, unsigned long *timeout) >> { >> struct menu_device *data = this_cpu_ptr(&menu_devices); >> int latency_req = cpuidle_governor_latency_req(dev->cpu); >> @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, >> } >> } >> >> +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION >> + if (drv->states[idx].flags & CPUIDLE_FLAG_AUTO_PROMOTION) { >> + /* >> + * Timeout is intended to be defined as sum of target residency >> + * of next available state, entry latency and exit latency. If >> + * time interval equal to timeout is spent in current state, >> + * and if it is a shallow lite state, we may want to auto- >> + * promote from such state. > This comment makes sense if you already understand auto-promotion. That's > fair enough - you wrote it and you presumably understand what your code > does :) But for me it's a bit confusing! I think you want to start with > a sentence about what autopromotion is (preferably not using > power-specific terminology) and then explain the calculation of the > timeouts. > >> + */ >> + for (i = idx + 1; i < drv->state_count; i++) { >> + if (drv->states[i].disabled || >> + dev->states_usage[i].disable) >> + continue; >> + *timeout = drv->states[i].target_residency + >> + 2 * drv->states[i].exit_latency; >> + break; >> + } >> + } >> +#endif >> + >> return idx; >> } >> >> diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h >> index 3b3947232..84d76d1ec 100644 >> --- a/include/linux/cpuidle.h >> +++ b/include/linux/cpuidle.h >> @@ -72,6 +72,13 @@ struct cpuidle_state { >> #define CPUIDLE_FLAG_POLLING BIT(0) /* polling state */ >> #define CPUIDLE_FLAG_COUPLED BIT(1) /* state applies to multiple cpus */ >> #define CPUIDLE_FLAG_TIMER_STOP BIT(2) /* timer is stopped on this state */ >> +/* >> + * State with only and only fast state bit set don't even lose user context. > "only and only"? >> + * But such states prevent other sibling threads from thread folding benefits. >> + * And hence we don't want to stay for too long in such states and want to >> + * auto-promote from it. > I think this comment mixes Power-specific and generic concepts. (But I'm > not a PM expert so tell me if I'm wrong here.) I think, if I've > understood correctly: in the generic code, the bit represents a state > that we do not want to linger in, which we want to definitely leave > after some time. On Power, we have a state that doesn't lose user > context but which prevents thread folding, so this is an example of a > state where we want to auto-promote. > >> + */ >> +#define CPUIDLE_FLAG_AUTO_PROMOTION BIT(3) >> >> struct cpuidle_device_kobj; >> struct cpuidle_state_kobj; >> @@ -243,7 +250,8 @@ struct cpuidle_governor { >> >> int (*select) (struct cpuidle_driver *drv, >> struct cpuidle_device *dev, >> - bool *stop_tick); >> + bool *stop_tick, unsigned long >> + *timeout); >> void (*reflect) (struct cpuidle_device *dev, int index); >> }; >> >> -- >> 2.17.1
On Tue, Apr 9, 2019 at 11:29 AM Abhishek <huntbag@linux.vnet.ibm.com> wrote: > > Hi Daniel, > > Thanks for such a descriptive review. I will include all the suggestions > made in my next iteration. Please give me some time to send comments before that.
Hi Daniel, On 04/08/2019 07:55 PM, Daniel Axtens wrote: > Hi, > > Sorry, just realised another thing I wanted to ask: > >> @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, >> } >> } >> >> >> +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION > Why is this based on CPUIDLE_FLAG_ rather than CONFIG_CPU_IDLE_? Won't > this always be true, given that the flag is defined regardless of the > config option in the header? Yeah, You are right. This should have been CONFIG_CPU_IDLE_AUTO_PROMOTION. --Abhishek
On 04/09/2019 03:00 PM, Rafael J. Wysocki wrote: > On Tue, Apr 9, 2019 at 11:29 AM Abhishek <huntbag@linux.vnet.ibm.com> wrote: >> Hi Daniel, >> >> Thanks for such a descriptive review. I will include all the suggestions >> made in my next iteration. > Please give me some time to send comments before that. Sure, I will wait for your review.
On Fri, Apr 5, 2019 at 11:17 AM Abhishek Goel <huntbag@linux.vnet.ibm.com> wrote: > > Currently, the cpuidle governors (menu /ladder) determine what idle state There are three governors in 5.1-rc. > an idling CPU should enter into based on heuristics that depend on the > idle history on that CPU. Given that no predictive heuristic is perfect, > there are cases where the governor predicts a shallow idle state, hoping > that the CPU will be busy soon. However, if no new workload is scheduled > on that CPU in the near future, the CPU will end up in the shallow state. > > In case of POWER, this is problematic, when the predicted state in the > aforementioned scenario is a lite stop state, as such lite states will > inhibit SMT folding, thereby depriving the other threads in the core from > using the core resources. > > To address this, such lite states need to be autopromoted. I don't quite agree with this statement and it doesn't even match what the patch does AFAICS. "Autopromotion" would be going from the given state to a deeper one without running state selection in between, but that's not what's going on here. > The cpuidle-core can queue timer to correspond with the residency value of the next > available state. Thus leading to auto-promotion to a deeper idle state as > soon as possible. No, it doesn't automatically cause a deeper state to be used next time. It simply kicks the CPU out of the idle state and one more iteration of the idle loop runs on it. Whether or not a deeper state will be selected in that iteration depends on the governor computations carried out in it. Now, this appears to be almost analogous to the "polling" state used on x86 which uses the next idle state's target residency as a timeout. While generally I'm not a big fan of setting up timers in the idle loop (it sort of feels like pulling your own hair in order to get yourself out of a swamp), if idle states like these are there in your platform, setting up a timer to get out of them in the driver's ->enter() routine might not be particularly objectionable. Doing that in the core is a whole different story, though. Generally, this adds quite a bit of complexity (on the "ugly" side of things IMO) to the core to cover a corner case present in one platform, while IMO it can be covered in the driver for that platform directly. > Signed-off-by: Abhishek Goel <huntbag@linux.vnet.ibm.com> > --- > > v1->v2 : Removed timeout_needed and rebased to current upstream kernel > > drivers/cpuidle/cpuidle.c | 68 +++++++++++++++++++++++++++++- > drivers/cpuidle/governors/ladder.c | 3 +- > drivers/cpuidle/governors/menu.c | 22 +++++++++- > include/linux/cpuidle.h | 10 ++++- > 4 files changed, 99 insertions(+), 4 deletions(-) > > diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c > index 7f108309e..11ce43f19 100644 > --- a/drivers/cpuidle/cpuidle.c > +++ b/drivers/cpuidle/cpuidle.c > @@ -36,6 +36,11 @@ static int enabled_devices; > static int off __read_mostly; > static int initialized __read_mostly; > > +struct auto_promotion { > + struct hrtimer hrtimer; > + unsigned long timeout_us; > +}; > + > int cpuidle_disabled(void) > { > return off; > @@ -188,6 +193,54 @@ int cpuidle_enter_s2idle(struct cpuidle_driver *drv, struct cpuidle_device *dev) > } > #endif /* CONFIG_SUSPEND */ > > +enum hrtimer_restart auto_promotion_hrtimer_callback(struct hrtimer *hrtimer) > +{ > + return HRTIMER_NORESTART; > +} > + > +#ifdef CONFIG_CPU_IDLE_AUTO_PROMOTION > +DEFINE_PER_CPU(struct auto_promotion, ap); > + > +static void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state *state) > +{ > + struct auto_promotion *this_ap = &per_cpu(ap, cpu); > + > + if (state->flags & CPUIDLE_FLAG_AUTO_PROMOTION) > + hrtimer_start(&this_ap->hrtimer, ns_to_ktime(this_ap->timeout_us > + * 1000), HRTIMER_MODE_REL_PINNED); > +} > + > +static void cpuidle_auto_promotion_cancel(int cpu) > +{ > + struct hrtimer *hrtimer; > + > + hrtimer = &per_cpu(ap, cpu).hrtimer; > + if (hrtimer_is_queued(hrtimer)) > + hrtimer_cancel(hrtimer); > +} > + > +static void cpuidle_auto_promotion_update(int cpu, unsigned long timeout) > +{ > + per_cpu(ap, cpu).timeout_us = timeout; > +} > + > +static void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver *drv) > +{ > + struct auto_promotion *this_ap = &per_cpu(ap, cpu); > + > + hrtimer_init(&this_ap->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); > + this_ap->hrtimer.function = auto_promotion_hrtimer_callback; > +} > +#else > +static inline void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state > + *state) { } > +static inline void cpuidle_auto_promotion_cancel(int cpu) { } > +static inline void cpuidle_auto_promotion_update(int cpu, unsigned long > + timeout) { } > +static inline void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver > + *drv) { } > +#endif > + > /** > * cpuidle_enter_state - enter the state and update stats > * @dev: cpuidle device for this cpu > @@ -225,12 +278,17 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, > trace_cpu_idle_rcuidle(index, dev->cpu); > time_start = ns_to_ktime(local_clock()); > > + cpuidle_auto_promotion_start(dev->cpu, target_state); First off, I wouldn't call it "auto-promotion", because it just adds a timeout to trigger if the CPU spends too much time in the target state. Second, and more important, I don't see why this cannot be done in target_state->enter() just for the state in which it is needed (in analogy with the "polling" state). > + > stop_critical_timings(); > entered_state = target_state->enter(dev, drv, index); > start_critical_timings(); > > sched_clock_idle_wakeup_event(); > time_end = ns_to_ktime(local_clock()); > + > + cpuidle_auto_promotion_cancel(dev->cpu); > + > trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); > > /* The cpu is no longer idle or about to enter idle. */ > @@ -312,7 +370,13 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, > int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > bool *stop_tick) > { > - return cpuidle_curr_governor->select(drv, dev, stop_tick); > + unsigned long timeout_us, ret; > + > + timeout_us = UINT_MAX; > + ret = cpuidle_curr_governor->select(drv, dev, stop_tick, &timeout_us); > + cpuidle_auto_promotion_update(dev->cpu, timeout_us); > + > + return ret; > } > > /** > @@ -658,6 +722,8 @@ int cpuidle_register(struct cpuidle_driver *drv, > device = &per_cpu(cpuidle_dev, cpu); > device->cpu = cpu; > > + cpuidle_auto_promotion_init(cpu, drv); > + > #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED > /* > * On multiplatform for ARM, the coupled idle states could be > diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c > index f0dddc66a..65b518dd7 100644 > --- a/drivers/cpuidle/governors/ladder.c > +++ b/drivers/cpuidle/governors/ladder.c > @@ -64,7 +64,8 @@ static inline void ladder_do_selection(struct ladder_device *ldev, > * @dummy: not used > */ > static int ladder_select_state(struct cpuidle_driver *drv, > - struct cpuidle_device *dev, bool *dummy) > + struct cpuidle_device *dev, bool *dummy, > + unsigned long *unused) > { > struct ladder_device *ldev = this_cpu_ptr(&ladder_devices); > struct ladder_device_state *last_state; > diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c > index 5951604e7..835e337de 100644 > --- a/drivers/cpuidle/governors/menu.c > +++ b/drivers/cpuidle/governors/menu.c > @@ -276,7 +276,7 @@ static unsigned int get_typical_interval(struct menu_device *data, > * @stop_tick: indication on whether or not to stop the tick > */ > static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > - bool *stop_tick) > + bool *stop_tick, unsigned long *timeout) > { > struct menu_device *data = this_cpu_ptr(&menu_devices); > int latency_req = cpuidle_governor_latency_req(dev->cpu); > @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, > } > } > > +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION > + if (drv->states[idx].flags & CPUIDLE_FLAG_AUTO_PROMOTION) { > + /* > + * Timeout is intended to be defined as sum of target residency > + * of next available state, entry latency and exit latency. If > + * time interval equal to timeout is spent in current state, > + * and if it is a shallow lite state, we may want to auto- > + * promote from such state. > + */ > + for (i = idx + 1; i < drv->state_count; i++) { > + if (drv->states[i].disabled || > + dev->states_usage[i].disable) > + continue; > + *timeout = drv->states[i].target_residency + > + 2 * drv->states[i].exit_latency; > + break; > + } > + } > +#endif Why do you need to do the above in the governor at all? The driver's ->enter() callback knows what state has been selected, it can check what the next available state is and set up the timer accordingly. I don't see the need to pass the timeout from the governor to the core anyway when the next thing the governor does is to return the selected state index. > + > return idx; > } > > diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h > index 3b3947232..84d76d1ec 100644 > --- a/include/linux/cpuidle.h > +++ b/include/linux/cpuidle.h > @@ -72,6 +72,13 @@ struct cpuidle_state { > #define CPUIDLE_FLAG_POLLING BIT(0) /* polling state */ > #define CPUIDLE_FLAG_COUPLED BIT(1) /* state applies to multiple cpus */ > #define CPUIDLE_FLAG_TIMER_STOP BIT(2) /* timer is stopped on this state */ > +/* > + * State with only and only fast state bit set don't even lose user context. > + * But such states prevent other sibling threads from thread folding benefits. > + * And hence we don't want to stay for too long in such states and want to > + * auto-promote from it. > + */ > +#define CPUIDLE_FLAG_AUTO_PROMOTION BIT(3) > > struct cpuidle_device_kobj; > struct cpuidle_state_kobj; > @@ -243,7 +250,8 @@ struct cpuidle_governor { > > int (*select) (struct cpuidle_driver *drv, > struct cpuidle_device *dev, > - bool *stop_tick); > + bool *stop_tick, unsigned long > + *timeout); > void (*reflect) (struct cpuidle_device *dev, int index); > }; > > -- > 2.17.1 >
Hi Rafael, Thanks for the Review. Few inline replies below. On 04/09/2019 03:31 PM, Rafael J. Wysocki wrote: > On Fri, Apr 5, 2019 at 11:17 AM Abhishek Goel > <huntbag@linux.vnet.ibm.com> wrote: >> Currently, the cpuidle governors (menu /ladder) determine what idle state > There are three governors in 5.1-rc. > >> an idling CPU should enter into based on heuristics that depend on the >> idle history on that CPU. Given that no predictive heuristic is perfect, >> there are cases where the governor predicts a shallow idle state, hoping >> that the CPU will be busy soon. However, if no new workload is scheduled >> on that CPU in the near future, the CPU will end up in the shallow state. >> >> In case of POWER, this is problematic, when the predicted state in the >> aforementioned scenario is a lite stop state, as such lite states will >> inhibit SMT folding, thereby depriving the other threads in the core from >> using the core resources. >> >> To address this, such lite states need to be autopromoted. > I don't quite agree with this statement and it doesn't even match what > the patch does AFAICS. "Autopromotion" would be going from the given > state to a deeper one without running state selection in between, but > that's not what's going on here. Thinking to call it "timed-exit". Is that good? >> The cpuidle-core can queue timer to correspond with the residency value of the next >> available state. Thus leading to auto-promotion to a deeper idle state as >> soon as possible. > No, it doesn't automatically cause a deeper state to be used next > time. It simply kicks the CPU out of the idle state and one more > iteration of the idle loop runs on it. Whether or not a deeper state > will be selected in that iteration depends on the governor > computations carried out in it. I did not mean that next state is chosen automatically. I should have been more descriptive here instead of just using "as soon as possible" > Now, this appears to be almost analogous to the "polling" state used > on x86 which uses the next idle state's target residency as a timeout. > > While generally I'm not a big fan of setting up timers in the idle > loop (it sort of feels like pulling your own hair in order to get > yourself out of a swamp), if idle states like these are there in your > platform, setting up a timer to get out of them in the driver's > ->enter() routine might not be particularly objectionable. Doing that > in the core is a whole different story, though. > > Generally, this adds quite a bit of complexity (on the "ugly" side of > things IMO) to the core to cover a corner case present in one > platform, while IMO it can be covered in the driver for that platform > directly. As of now, since this code doesn't add any benefit to the other platform, I will post a patch with this implementation covered in platform-specific driver code. You are right that all the information needed for this implementation are also available there in platform driver code, so we should be good to go.
diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index 7f108309e..11ce43f19 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -36,6 +36,11 @@ static int enabled_devices; static int off __read_mostly; static int initialized __read_mostly; +struct auto_promotion { + struct hrtimer hrtimer; + unsigned long timeout_us; +}; + int cpuidle_disabled(void) { return off; @@ -188,6 +193,54 @@ int cpuidle_enter_s2idle(struct cpuidle_driver *drv, struct cpuidle_device *dev) } #endif /* CONFIG_SUSPEND */ +enum hrtimer_restart auto_promotion_hrtimer_callback(struct hrtimer *hrtimer) +{ + return HRTIMER_NORESTART; +} + +#ifdef CONFIG_CPU_IDLE_AUTO_PROMOTION +DEFINE_PER_CPU(struct auto_promotion, ap); + +static void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state *state) +{ + struct auto_promotion *this_ap = &per_cpu(ap, cpu); + + if (state->flags & CPUIDLE_FLAG_AUTO_PROMOTION) + hrtimer_start(&this_ap->hrtimer, ns_to_ktime(this_ap->timeout_us + * 1000), HRTIMER_MODE_REL_PINNED); +} + +static void cpuidle_auto_promotion_cancel(int cpu) +{ + struct hrtimer *hrtimer; + + hrtimer = &per_cpu(ap, cpu).hrtimer; + if (hrtimer_is_queued(hrtimer)) + hrtimer_cancel(hrtimer); +} + +static void cpuidle_auto_promotion_update(int cpu, unsigned long timeout) +{ + per_cpu(ap, cpu).timeout_us = timeout; +} + +static void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver *drv) +{ + struct auto_promotion *this_ap = &per_cpu(ap, cpu); + + hrtimer_init(&this_ap->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + this_ap->hrtimer.function = auto_promotion_hrtimer_callback; +} +#else +static inline void cpuidle_auto_promotion_start(int cpu, struct cpuidle_state + *state) { } +static inline void cpuidle_auto_promotion_cancel(int cpu) { } +static inline void cpuidle_auto_promotion_update(int cpu, unsigned long + timeout) { } +static inline void cpuidle_auto_promotion_init(int cpu, struct cpuidle_driver + *drv) { } +#endif + /** * cpuidle_enter_state - enter the state and update stats * @dev: cpuidle device for this cpu @@ -225,12 +278,17 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, trace_cpu_idle_rcuidle(index, dev->cpu); time_start = ns_to_ktime(local_clock()); + cpuidle_auto_promotion_start(dev->cpu, target_state); + stop_critical_timings(); entered_state = target_state->enter(dev, drv, index); start_critical_timings(); sched_clock_idle_wakeup_event(); time_end = ns_to_ktime(local_clock()); + + cpuidle_auto_promotion_cancel(dev->cpu); + trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); /* The cpu is no longer idle or about to enter idle. */ @@ -312,7 +370,13 @@ int cpuidle_enter_state(struct cpuidle_device *dev, struct cpuidle_driver *drv, int cpuidle_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, bool *stop_tick) { - return cpuidle_curr_governor->select(drv, dev, stop_tick); + unsigned long timeout_us, ret; + + timeout_us = UINT_MAX; + ret = cpuidle_curr_governor->select(drv, dev, stop_tick, &timeout_us); + cpuidle_auto_promotion_update(dev->cpu, timeout_us); + + return ret; } /** @@ -658,6 +722,8 @@ int cpuidle_register(struct cpuidle_driver *drv, device = &per_cpu(cpuidle_dev, cpu); device->cpu = cpu; + cpuidle_auto_promotion_init(cpu, drv); + #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED /* * On multiplatform for ARM, the coupled idle states could be diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c index f0dddc66a..65b518dd7 100644 --- a/drivers/cpuidle/governors/ladder.c +++ b/drivers/cpuidle/governors/ladder.c @@ -64,7 +64,8 @@ static inline void ladder_do_selection(struct ladder_device *ldev, * @dummy: not used */ static int ladder_select_state(struct cpuidle_driver *drv, - struct cpuidle_device *dev, bool *dummy) + struct cpuidle_device *dev, bool *dummy, + unsigned long *unused) { struct ladder_device *ldev = this_cpu_ptr(&ladder_devices); struct ladder_device_state *last_state; diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c index 5951604e7..835e337de 100644 --- a/drivers/cpuidle/governors/menu.c +++ b/drivers/cpuidle/governors/menu.c @@ -276,7 +276,7 @@ static unsigned int get_typical_interval(struct menu_device *data, * @stop_tick: indication on whether or not to stop the tick */ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, - bool *stop_tick) + bool *stop_tick, unsigned long *timeout) { struct menu_device *data = this_cpu_ptr(&menu_devices); int latency_req = cpuidle_governor_latency_req(dev->cpu); @@ -442,6 +442,26 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev, } } +#ifdef CPUIDLE_FLAG_AUTO_PROMOTION + if (drv->states[idx].flags & CPUIDLE_FLAG_AUTO_PROMOTION) { + /* + * Timeout is intended to be defined as sum of target residency + * of next available state, entry latency and exit latency. If + * time interval equal to timeout is spent in current state, + * and if it is a shallow lite state, we may want to auto- + * promote from such state. + */ + for (i = idx + 1; i < drv->state_count; i++) { + if (drv->states[i].disabled || + dev->states_usage[i].disable) + continue; + *timeout = drv->states[i].target_residency + + 2 * drv->states[i].exit_latency; + break; + } + } +#endif + return idx; } diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 3b3947232..84d76d1ec 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -72,6 +72,13 @@ struct cpuidle_state { #define CPUIDLE_FLAG_POLLING BIT(0) /* polling state */ #define CPUIDLE_FLAG_COUPLED BIT(1) /* state applies to multiple cpus */ #define CPUIDLE_FLAG_TIMER_STOP BIT(2) /* timer is stopped on this state */ +/* + * State with only and only fast state bit set don't even lose user context. + * But such states prevent other sibling threads from thread folding benefits. + * And hence we don't want to stay for too long in such states and want to + * auto-promote from it. + */ +#define CPUIDLE_FLAG_AUTO_PROMOTION BIT(3) struct cpuidle_device_kobj; struct cpuidle_state_kobj; @@ -243,7 +250,8 @@ struct cpuidle_governor { int (*select) (struct cpuidle_driver *drv, struct cpuidle_device *dev, - bool *stop_tick); + bool *stop_tick, unsigned long + *timeout); void (*reflect) (struct cpuidle_device *dev, int index); };
Currently, the cpuidle governors (menu /ladder) determine what idle state an idling CPU should enter into based on heuristics that depend on the idle history on that CPU. Given that no predictive heuristic is perfect, there are cases where the governor predicts a shallow idle state, hoping that the CPU will be busy soon. However, if no new workload is scheduled on that CPU in the near future, the CPU will end up in the shallow state. In case of POWER, this is problematic, when the predicted state in the aforementioned scenario is a lite stop state, as such lite states will inhibit SMT folding, thereby depriving the other threads in the core from using the core resources. To address this, such lite states need to be autopromoted. The cpuidle- core can queue timer to correspond with the residency value of the next available state. Thus leading to auto-promotion to a deeper idle state as soon as possible. Signed-off-by: Abhishek Goel <huntbag@linux.vnet.ibm.com> --- v1->v2 : Removed timeout_needed and rebased to current upstream kernel drivers/cpuidle/cpuidle.c | 68 +++++++++++++++++++++++++++++- drivers/cpuidle/governors/ladder.c | 3 +- drivers/cpuidle/governors/menu.c | 22 +++++++++- include/linux/cpuidle.h | 10 ++++- 4 files changed, 99 insertions(+), 4 deletions(-)