From 8b0510b52478a4ede8a7537da7faa66242bd0ae4 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Thu, 11 Aug 2016 11:40:05 +0100 Subject: PM / Domains: Always enable debugfs support if available Debugfs support for PM domains is only enabled if both CONFIG_PM_DEBUG and CONFIG_PM_ADVANCED_DEBUG are enabled. CONFIG_PM_ADVANCED_DEBUG is described as "extra PM attributes in sysfs for low-level debugging/testing" which does not seem related. Given that the debugfs for PM domains only allows users to view the state of the PM domains, always enable debugfs support for PM domains if PM domains and debugfs support is enabled. Signed-off-by: Jon Hunter Acked-by: Geert Uytterhoeven Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index a1f2aff..058c8b6 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1636,7 +1636,7 @@ EXPORT_SYMBOL_GPL(genpd_dev_pm_attach); /*** debugfs support ***/ -#ifdef CONFIG_PM_ADVANCED_DEBUG +#ifdef CONFIG_DEBUG_FS #include #include #include @@ -1784,4 +1784,4 @@ static void __exit pm_genpd_debug_exit(void) debugfs_remove_recursive(pm_genpd_debugfs_dir); } __exitcall(pm_genpd_debug_exit); -#endif /* CONFIG_PM_ADVANCED_DEBUG */ +#endif /* CONFIG_DEBUG_FS */ -- cgit v0.10.2 From af48d7bc3756a0cd882d65bff14ab39746ba57fe Mon Sep 17 00:00:00 2001 From: Christophe Jaillet Date: Thu, 11 Aug 2016 15:02:30 +0200 Subject: ARM: cpuidle: Fix error return code We know that 'ret = 0' because it has been tested a few lines above. So, if 'kzalloc' fails, 0 will be returned instead of an error code. Return -ENOMEM instead. Fixes: a0d46a3dfdc3 ("ARM: cpuidle: Register per cpuidle device") Signed-off-by: Christophe Jaillet Acked-by: Lorenzo Pieralisi Cc: 4.1+ # 4.1+ Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpuidle/cpuidle-arm.c b/drivers/cpuidle/cpuidle-arm.c index e342565e..1855b9e 100644 --- a/drivers/cpuidle/cpuidle-arm.c +++ b/drivers/cpuidle/cpuidle-arm.c @@ -135,6 +135,7 @@ static int __init arm_idle_init(void) dev = kzalloc(sizeof(*dev), GFP_KERNEL); if (!dev) { pr_err("Failed to allocate cpuidle device\n"); + ret = -ENOMEM; goto out_fail; } dev->cpu = cpu; -- cgit v0.10.2 From c4b4057267f258e6d823080205f47227fb2abfd7 Mon Sep 17 00:00:00 2001 From: Chanwoo Choi Date: Tue, 16 Aug 2016 15:27:19 +0900 Subject: cpufreq: dt: Add exynos5433 compatible to use generic cpufreq driver This patch adds the exynos5433 compatible string for supporting the generic cpufreq driver on Exynos5433. Signed-off-by: Chanwoo Choi Reviewed-by: Krzysztof Kozlowski Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq-dt-platdev.c b/drivers/cpufreq/cpufreq-dt-platdev.c index 0bb44d5..470db30 100644 --- a/drivers/cpufreq/cpufreq-dt-platdev.c +++ b/drivers/cpufreq/cpufreq-dt-platdev.c @@ -40,6 +40,7 @@ static const struct of_device_id machines[] __initconst = { { .compatible = "samsung,exynos5250", }, #ifndef CONFIG_BL_SWITCHER { .compatible = "samsung,exynos5420", }, + { .compatible = "samsung,exynos5433", }, { .compatible = "samsung,exynos5800", }, #endif -- cgit v0.10.2 From 58919e83c85c3a3c5fb34025dc0e95ddd998c478 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Tue, 16 Aug 2016 22:14:55 +0200 Subject: cpufreq / sched: Pass flags to cpufreq_update_util() It is useful to know the reason why cpufreq_update_util() has just been called and that can be passed as flags to cpufreq_update_util() and to the ->func() callback in struct update_util_data. However, doing that in addition to passing the util and max arguments they already take would be clumsy, so avoid it. Instead, use the observation that the schedutil governor is part of the scheduler proper, so it can access scheduler data directly. This allows the util and max arguments of cpufreq_update_util() and the ->func() callback in struct update_util_data to be replaced with a flags one, but schedutil has to be modified to follow. Thus make the schedutil governor obtain the CFS utilization information from the scheduler and use the "RT" and "DL" flags instead of the special utilization value of ULONG_MAX to track updates from the RT and DL sched classes. Make it non-modular too to avoid having to export scheduler variables to modules at large. Next, update all of the other users of cpufreq_update_util() and the ->func() callback in struct update_util_data accordingly. Suggested-by: Peter Zijlstra Signed-off-by: Rafael J. Wysocki Acked-by: Peter Zijlstra (Intel) Acked-by: Viresh Kumar diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig index 74919aa..4dc9525 100644 --- a/drivers/cpufreq/Kconfig +++ b/drivers/cpufreq/Kconfig @@ -194,7 +194,7 @@ config CPU_FREQ_GOV_CONSERVATIVE If in doubt, say N. config CPU_FREQ_GOV_SCHEDUTIL - tristate "'schedutil' cpufreq policy governor" + bool "'schedutil' cpufreq policy governor" depends on CPU_FREQ && SMP select CPU_FREQ_GOV_ATTR_SET select IRQ_WORK @@ -208,9 +208,6 @@ config CPU_FREQ_GOV_SCHEDUTIL frequency tipping point is at utilization/capacity equal to 80% in both cases. - To compile this driver as a module, choose M here: the module will - be called cpufreq_schedutil. - If in doubt, say N. comment "CPU frequency scaling drivers" diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c index e415349..642dd0f 100644 --- a/drivers/cpufreq/cpufreq_governor.c +++ b/drivers/cpufreq/cpufreq_governor.c @@ -260,7 +260,7 @@ static void dbs_irq_work(struct irq_work *irq_work) } static void dbs_update_util_handler(struct update_util_data *data, u64 time, - unsigned long util, unsigned long max) + unsigned int flags) { struct cpu_dbs_info *cdbs = container_of(data, struct cpu_dbs_info, update_util); struct policy_dbs_info *policy_dbs = cdbs->policy_dbs; diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index be9eade..bdbe936 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1329,7 +1329,7 @@ static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu) } static void intel_pstate_update_util(struct update_util_data *data, u64 time, - unsigned long util, unsigned long max) + unsigned int flags) { struct cpudata *cpu = container_of(data, struct cpudata, update_util); u64 delta_ns = time - cpu->sample.time; diff --git a/include/linux/sched.h b/include/linux/sched.h index 62c68e5..b0fa726 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -3469,15 +3469,19 @@ static inline unsigned long rlimit_max(unsigned int limit) return task_rlimit_max(current, limit); } +#define SCHED_CPUFREQ_RT (1U << 0) +#define SCHED_CPUFREQ_DL (1U << 1) + +#define SCHED_CPUFREQ_RT_DL (SCHED_CPUFREQ_RT | SCHED_CPUFREQ_DL) + #ifdef CONFIG_CPU_FREQ struct update_util_data { - void (*func)(struct update_util_data *data, - u64 time, unsigned long util, unsigned long max); + void (*func)(struct update_util_data *data, u64 time, unsigned int flags); }; void cpufreq_add_update_util_hook(int cpu, struct update_util_data *data, - void (*func)(struct update_util_data *data, u64 time, - unsigned long util, unsigned long max)); + void (*func)(struct update_util_data *data, u64 time, + unsigned int flags)); void cpufreq_remove_update_util_hook(int cpu); #endif /* CONFIG_CPU_FREQ */ diff --git a/kernel/sched/cpufreq.c b/kernel/sched/cpufreq.c index 1141954..dbc5144 100644 --- a/kernel/sched/cpufreq.c +++ b/kernel/sched/cpufreq.c @@ -33,7 +33,7 @@ DEFINE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); */ void cpufreq_add_update_util_hook(int cpu, struct update_util_data *data, void (*func)(struct update_util_data *data, u64 time, - unsigned long util, unsigned long max)) + unsigned int flags)) { if (WARN_ON(!data || !func)) return; diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index a84641b..60d985f 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -12,7 +12,6 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include -#include #include #include @@ -53,6 +52,7 @@ struct sugov_cpu { unsigned long util; unsigned long max; u64 last_update; + unsigned int flags; }; static DEFINE_PER_CPU(struct sugov_cpu, sugov_cpu); @@ -144,24 +144,39 @@ static unsigned int get_next_freq(struct sugov_cpu *sg_cpu, unsigned long util, return cpufreq_driver_resolve_freq(policy, freq); } +static void sugov_get_util(unsigned long *util, unsigned long *max) +{ + struct rq *rq = this_rq(); + unsigned long cfs_max = rq->cpu_capacity_orig; + + *util = min(rq->cfs.avg.util_avg, cfs_max); + *max = cfs_max; +} + static void sugov_update_single(struct update_util_data *hook, u64 time, - unsigned long util, unsigned long max) + unsigned int flags) { struct sugov_cpu *sg_cpu = container_of(hook, struct sugov_cpu, update_util); struct sugov_policy *sg_policy = sg_cpu->sg_policy; struct cpufreq_policy *policy = sg_policy->policy; + unsigned long util, max; unsigned int next_f; if (!sugov_should_update_freq(sg_policy, time)) return; - next_f = util == ULONG_MAX ? policy->cpuinfo.max_freq : - get_next_freq(sg_cpu, util, max); + if (flags & SCHED_CPUFREQ_RT_DL) { + next_f = policy->cpuinfo.max_freq; + } else { + sugov_get_util(&util, &max); + next_f = get_next_freq(sg_cpu, util, max); + } sugov_update_commit(sg_policy, time, next_f); } static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, - unsigned long util, unsigned long max) + unsigned long util, unsigned long max, + unsigned int flags) { struct sugov_policy *sg_policy = sg_cpu->sg_policy; struct cpufreq_policy *policy = sg_policy->policy; @@ -169,7 +184,7 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, u64 last_freq_update_time = sg_policy->last_freq_update_time; unsigned int j; - if (util == ULONG_MAX) + if (flags & SCHED_CPUFREQ_RT_DL) return max_f; for_each_cpu(j, policy->cpus) { @@ -192,10 +207,10 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, if (delta_ns > TICK_NSEC) continue; - j_util = j_sg_cpu->util; - if (j_util == ULONG_MAX) + if (j_sg_cpu->flags & SCHED_CPUFREQ_RT_DL) return max_f; + j_util = j_sg_cpu->util; j_max = j_sg_cpu->max; if (j_util * max > j_max * util) { util = j_util; @@ -207,20 +222,24 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, } static void sugov_update_shared(struct update_util_data *hook, u64 time, - unsigned long util, unsigned long max) + unsigned int flags) { struct sugov_cpu *sg_cpu = container_of(hook, struct sugov_cpu, update_util); struct sugov_policy *sg_policy = sg_cpu->sg_policy; + unsigned long util, max; unsigned int next_f; + sugov_get_util(&util, &max); + raw_spin_lock(&sg_policy->update_lock); sg_cpu->util = util; sg_cpu->max = max; + sg_cpu->flags = flags; sg_cpu->last_update = time; if (sugov_should_update_freq(sg_policy, time)) { - next_f = sugov_next_freq_shared(sg_cpu, util, max); + next_f = sugov_next_freq_shared(sg_cpu, util, max, flags); sugov_update_commit(sg_policy, time, next_f); } @@ -444,8 +463,9 @@ static int sugov_start(struct cpufreq_policy *policy) sg_cpu->sg_policy = sg_policy; if (policy_is_shared(policy)) { - sg_cpu->util = ULONG_MAX; + sg_cpu->util = 0; sg_cpu->max = 0; + sg_cpu->flags = SCHED_CPUFREQ_RT; sg_cpu->last_update = 0; sg_cpu->cached_raw_freq = 0; cpufreq_add_update_util_hook(cpu, &sg_cpu->update_util, @@ -495,28 +515,15 @@ static struct cpufreq_governor schedutil_gov = { .limits = sugov_limits, }; -static int __init sugov_module_init(void) -{ - return cpufreq_register_governor(&schedutil_gov); -} - -static void __exit sugov_module_exit(void) -{ - cpufreq_unregister_governor(&schedutil_gov); -} - -MODULE_AUTHOR("Rafael J. Wysocki "); -MODULE_DESCRIPTION("Utilization-based CPU frequency selection"); -MODULE_LICENSE("GPL"); - #ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL struct cpufreq_governor *cpufreq_default_governor(void) { return &schedutil_gov; } - -fs_initcall(sugov_module_init); -#else -module_init(sugov_module_init); #endif -module_exit(sugov_module_exit); + +static int __init sugov_register(void) +{ + return cpufreq_register_governor(&schedutil_gov); +} +fs_initcall(sugov_register); diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index 1ce8867..4464cc3 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -735,9 +735,9 @@ static void update_curr_dl(struct rq *rq) return; } - /* kick cpufreq (see the comment in linux/cpufreq.h). */ + /* kick cpufreq (see the comment in kernel/sched/sched.h). */ if (cpu_of(rq) == smp_processor_id()) - cpufreq_trigger_update(rq_clock(rq)); + cpufreq_update_util(rq_clock(rq), SCHED_CPUFREQ_DL); schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 039de34..f91fa57 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -2875,11 +2875,8 @@ static inline void update_tg_load_avg(struct cfs_rq *cfs_rq, int force) {} static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) { - struct rq *rq = rq_of(cfs_rq); - int cpu = cpu_of(rq); - - if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) { - unsigned long max = rq->cpu_capacity_orig; + if (&this_rq()->cfs == cfs_rq) { + struct rq *rq = rq_of(cfs_rq); /* * There are a few boundary cases this might miss but it should @@ -2897,8 +2894,7 @@ static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) * * See cpu_util(). */ - cpufreq_update_util(rq_clock(rq), - min(cfs_rq->avg.util_avg, max), max); + cpufreq_update_util(rq_clock(rq), 0); } } @@ -3162,7 +3158,7 @@ static inline void update_load_avg(struct sched_entity *se, int not_used) struct cfs_rq *cfs_rq = cfs_rq_of(se); struct rq *rq = rq_of(cfs_rq); - cpufreq_trigger_update(rq_clock(rq)); + cpufreq_update_util(rq_clock(rq), 0); } static inline void diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index d5690b7..8a9cd9b 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -957,9 +957,9 @@ static void update_curr_rt(struct rq *rq) if (unlikely((s64)delta_exec <= 0)) return; - /* Kick cpufreq (see the comment in linux/cpufreq.h). */ + /* Kick cpufreq (see the comment in kernel/sched/sched.h). */ if (cpu_of(rq) == smp_processor_id()) - cpufreq_trigger_update(rq_clock(rq)); + cpufreq_update_util(rq_clock(rq), SCHED_CPUFREQ_RT); schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index c64fc51..82fc554 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1764,26 +1764,12 @@ DECLARE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); /** * cpufreq_update_util - Take a note about CPU utilization changes. * @time: Current time. - * @util: Current utilization. - * @max: Utilization ceiling. + * @flags: Update reason flags. * - * This function is called by the scheduler on every invocation of - * update_load_avg() on the CPU whose utilization is being updated. + * This function is called by the scheduler on the CPU whose utilization is + * being updated. * * It can only be called from RCU-sched read-side critical sections. - */ -static inline void cpufreq_update_util(u64 time, unsigned long util, unsigned long max) -{ - struct update_util_data *data; - - data = rcu_dereference_sched(*this_cpu_ptr(&cpufreq_update_util_data)); - if (data) - data->func(data, time, util, max); -} - -/** - * cpufreq_trigger_update - Trigger CPU performance state evaluation if needed. - * @time: Current time. * * The way cpufreq is currently arranged requires it to evaluate the CPU * performance state (frequency/voltage) on a regular basis to prevent it from @@ -1797,13 +1783,16 @@ static inline void cpufreq_update_util(u64 time, unsigned long util, unsigned lo * but that really is a band-aid. Going forward it should be replaced with * solutions targeted more specifically at RT and DL tasks. */ -static inline void cpufreq_trigger_update(u64 time) +static inline void cpufreq_update_util(u64 time, unsigned int flags) { - cpufreq_update_util(time, ULONG_MAX, 0); + struct update_util_data *data; + + data = rcu_dereference_sched(*this_cpu_ptr(&cpufreq_update_util_data)); + if (data) + data->func(data, time, flags); } #else -static inline void cpufreq_update_util(u64 time, unsigned long util, unsigned long max) {} -static inline void cpufreq_trigger_update(u64 time) {} +static inline void cpufreq_update_util(u64 time, unsigned int flags) {} #endif /* CONFIG_CPU_FREQ */ #ifdef arch_scale_freq_capacity -- cgit v0.10.2 From 12bde33dbb3eadd60343a8a71c39766073c1d752 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 10 Aug 2016 03:11:17 +0200 Subject: cpufreq / sched: Pass runqueue pointer to cpufreq_update_util() All of the callers of cpufreq_update_util() pass rq_clock(rq) to it as the time argument and some of them check whether or not cpu_of(rq) is equal to smp_processor_id() before calling it, so rework it to take a runqueue pointer as the argument and move the rq_clock(rq) evaluation into it. Additionally, provide a wrapper checking cpu_of(rq) against smp_processor_id() for the cpufreq_update_util() callers that need it. Signed-off-by: Rafael J. Wysocki Acked-by: Peter Zijlstra (Intel) Acked-by: Viresh Kumar diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index 4464cc3..9747796 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -736,8 +736,7 @@ static void update_curr_dl(struct rq *rq) } /* kick cpufreq (see the comment in kernel/sched/sched.h). */ - if (cpu_of(rq) == smp_processor_id()) - cpufreq_update_util(rq_clock(rq), SCHED_CPUFREQ_DL); + cpufreq_update_this_cpu(rq, SCHED_CPUFREQ_DL); schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index f91fa57..5d558cc 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -2876,8 +2876,6 @@ static inline void update_tg_load_avg(struct cfs_rq *cfs_rq, int force) {} static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) { if (&this_rq()->cfs == cfs_rq) { - struct rq *rq = rq_of(cfs_rq); - /* * There are a few boundary cases this might miss but it should * get called often enough that that should (hopefully) not be @@ -2894,7 +2892,7 @@ static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) * * See cpu_util(). */ - cpufreq_update_util(rq_clock(rq), 0); + cpufreq_update_util(rq_of(cfs_rq), 0); } } @@ -3155,10 +3153,7 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq, bool update_freq) static inline void update_load_avg(struct sched_entity *se, int not_used) { - struct cfs_rq *cfs_rq = cfs_rq_of(se); - struct rq *rq = rq_of(cfs_rq); - - cpufreq_update_util(rq_clock(rq), 0); + cpufreq_update_util(rq_of(cfs_rq_of(se)), 0); } static inline void diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 8a9cd9b..2516b8d 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -958,8 +958,7 @@ static void update_curr_rt(struct rq *rq) return; /* Kick cpufreq (see the comment in kernel/sched/sched.h). */ - if (cpu_of(rq) == smp_processor_id()) - cpufreq_update_util(rq_clock(rq), SCHED_CPUFREQ_RT); + cpufreq_update_this_cpu(rq, SCHED_CPUFREQ_RT); schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 82fc554..b7fc1ce 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1763,7 +1763,7 @@ DECLARE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); /** * cpufreq_update_util - Take a note about CPU utilization changes. - * @time: Current time. + * @rq: Runqueue to carry out the update for. * @flags: Update reason flags. * * This function is called by the scheduler on the CPU whose utilization is @@ -1783,16 +1783,23 @@ DECLARE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); * but that really is a band-aid. Going forward it should be replaced with * solutions targeted more specifically at RT and DL tasks. */ -static inline void cpufreq_update_util(u64 time, unsigned int flags) +static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) { struct update_util_data *data; data = rcu_dereference_sched(*this_cpu_ptr(&cpufreq_update_util_data)); if (data) - data->func(data, time, flags); + data->func(data, rq_clock(rq), flags); +} + +static inline void cpufreq_update_this_cpu(struct rq *rq, unsigned int flags) +{ + if (cpu_of(rq) == smp_processor_id()) + cpufreq_update_util(rq, flags); } #else -static inline void cpufreq_update_util(u64 time, unsigned int flags) {} +static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) {} +static inline void cpufreq_update_this_cpu(struct rq *rq, unsigned int flags) {} #endif /* CONFIG_CPU_FREQ */ #ifdef arch_scale_freq_capacity -- cgit v0.10.2 From b0d8a69d08069f4f4b1d9670d57a7e22da6e6c6a Mon Sep 17 00:00:00 2001 From: Markus Elfring Date: Tue, 16 Aug 2016 23:00:07 +0200 Subject: cpufreq-SCPI: Delete unnecessary assignment for the field "owner" The field "owner" is set by the core. Thus delete an unneeded initialisation. Generated by scripts/coccinelle/api/platform_no_drv_owner.cocci Signed-off-by: Markus Elfring Acked-by: Viresh Kumar Acked-by: Sudeep Holla Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/scpi-cpufreq.c b/drivers/cpufreq/scpi-cpufreq.c index e8a7bf5..ea7a4e1 100644 --- a/drivers/cpufreq/scpi-cpufreq.c +++ b/drivers/cpufreq/scpi-cpufreq.c @@ -105,7 +105,6 @@ static int scpi_cpufreq_remove(struct platform_device *pdev) static struct platform_driver scpi_cpufreq_platdrv = { .driver = { .name = "scpi-cpufreq", - .owner = THIS_MODULE, }, .probe = scpi_cpufreq_probe, .remove = scpi_cpufreq_remove, -- cgit v0.10.2 From 5f16d59cb97d24bafd172b7d935aff0f6c10d8ba Mon Sep 17 00:00:00 2001 From: Javier Martinez Canillas Date: Fri, 19 Aug 2016 11:43:12 -0400 Subject: ARM: exynos_defconfig: Don't attempt to enable schedutil governor as module The schedutil CPUFreq governor could be built as a module but the change "cpufreq / sched: Pass flags to cpufreq_update_util()" made it non-modular so attempting to enable as module leads to a warning: warning: symbol value 'm' invalid for CPU_FREQ_GOV_SCHEDUTIL Signed-off-by: Javier Martinez Canillas Signed-off-by: Krzysztof Kozlowski diff --git a/arch/arm/configs/exynos_defconfig b/arch/arm/configs/exynos_defconfig index 01986de..36cc7cc 100644 --- a/arch/arm/configs/exynos_defconfig +++ b/arch/arm/configs/exynos_defconfig @@ -28,7 +28,7 @@ CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND=y CONFIG_CPU_FREQ_GOV_POWERSAVE=m CONFIG_CPU_FREQ_GOV_USERSPACE=m CONFIG_CPU_FREQ_GOV_CONSERVATIVE=m -CONFIG_CPU_FREQ_GOV_SCHEDUTIL=m +CONFIG_CPU_FREQ_GOV_SCHEDUTIL=y CONFIG_CPUFREQ_DT=y CONFIG_CPU_IDLE=y CONFIG_ARM_EXYNOS_CPUIDLE=y -- cgit v0.10.2 From 08023fb9cec03ee2992df51fa728c3f14a1ba0ea Mon Sep 17 00:00:00 2001 From: Javier Martinez Canillas Date: Fri, 19 Aug 2016 11:43:13 -0400 Subject: ARM: multi_v7_defconfig: Don't attempt to enable schedutil governor as module The schedutil CPUFreq governor could be built as a module but the change "cpufreq / sched: Pass flags to cpufreq_update_util()" made it non-modular so attempting to enable as module leads to a warning: warning: symbol value 'm' invalid for CPU_FREQ_GOV_SCHEDUTIL Signed-off-by: Javier Martinez Canillas Signed-off-by: Krzysztof Kozlowski diff --git a/arch/arm/configs/multi_v7_defconfig b/arch/arm/configs/multi_v7_defconfig index 2c8665c..ab5fea7 100644 --- a/arch/arm/configs/multi_v7_defconfig +++ b/arch/arm/configs/multi_v7_defconfig @@ -135,7 +135,7 @@ CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND=y CONFIG_CPU_FREQ_GOV_POWERSAVE=m CONFIG_CPU_FREQ_GOV_USERSPACE=m CONFIG_CPU_FREQ_GOV_CONSERVATIVE=m -CONFIG_CPU_FREQ_GOV_SCHEDUTIL=m +CONFIG_CPU_FREQ_GOV_SCHEDUTIL=y CONFIG_QORIQ_CPUFREQ=y CONFIG_CPU_IDLE=y CONFIG_ARM_CPUIDLE=y -- cgit v0.10.2 From 3689ad7ed6a836c4eec5e7bdd17a11a79591bef9 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 31 Aug 2016 03:11:31 +0200 Subject: cpufreq: Drop unnecessary check from cpufreq_policy_alloc() Since cpufreq_policy_alloc() doesn't use its dev variable for anything useful, drop that variable from there along with the NULL check against it. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 3dd4884..13fb589 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -1073,13 +1073,9 @@ static void handle_update(struct work_struct *work) static struct cpufreq_policy *cpufreq_policy_alloc(unsigned int cpu) { - struct device *dev = get_cpu_device(cpu); struct cpufreq_policy *policy; int ret; - if (WARN_ON(!dev)) - return NULL; - policy = kzalloc(sizeof(*policy), GFP_KERNEL); if (!policy) return NULL; -- cgit v0.10.2 From 8314bc83f6a33958a033955e9bdc48e8dd4d5fb0 Mon Sep 17 00:00:00 2001 From: Steve Muckle Date: Fri, 26 Aug 2016 11:40:47 -0700 Subject: cpufreq / sched: ignore SMT when determining max cpu capacity PELT does not consider SMT when scaling its utilization values via arch_scale_cpu_capacity(). The value in rq->cpu_capacity_orig does take SMT into consideration though and therefore may be smaller than the utilization reported by PELT. On an Intel i7-3630QM for example rq->cpu_capacity_orig is 589 but util_avg scales up to 1024. This means that a 50% utilized CPU will show up in schedutil as ~86% busy. Fix this by using the same CPU scaling value in schedutil as that which is used by PELT. Signed-off-by: Steve Muckle Signed-off-by: Rafael J. Wysocki diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index 60d985f..cb8a77b 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -147,7 +147,9 @@ static unsigned int get_next_freq(struct sugov_cpu *sg_cpu, unsigned long util, static void sugov_get_util(unsigned long *util, unsigned long *max) { struct rq *rq = this_rq(); - unsigned long cfs_max = rq->cpu_capacity_orig; + unsigned long cfs_max; + + cfs_max = arch_scale_cpu_capacity(NULL, smp_processor_id()); *util = min(rq->cfs.avg.util_avg, cfs_max); *max = cfs_max; -- cgit v0.10.2 From 0b38ed36a1ac5e8b8c6bee6f10aada45111a0658 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Fri, 8 Jul 2016 13:44:04 +0000 Subject: PM / devfreq: exynos-ppmu: remove unneeded of_node_put() for_each_child_of_node() performs an of_node_put() on each iteration, so putting an of_node_put() before a continue results in a double put. Generated by: scripts/coccinelle/iterators/device_node_continue.cocci Signed-off-by: Wei Yongjun Signed-off-by: MyungJoo Ham diff --git a/drivers/devfreq/event/exynos-ppmu.c b/drivers/devfreq/event/exynos-ppmu.c index 845bf25..f55cf0e 100644 --- a/drivers/devfreq/event/exynos-ppmu.c +++ b/drivers/devfreq/event/exynos-ppmu.c @@ -406,8 +406,6 @@ static int of_get_devfreq_events(struct device_node *np, of_property_read_string(node, "event-name", &desc[j].name); j++; - - of_node_put(node); } info->desc = desc; -- cgit v0.10.2 From 797da5598f3a6fd92a514352d67078a682a432cb Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Fri, 19 Aug 2016 08:36:55 +0200 Subject: PM / devfreq: Add COMPILE_TEST for build coverage The SoC-specific devfreq and devfreq-event drivers can be build tested on all architectures. Signed-off-by: Krzysztof Kozlowski Reviewed-by: Chanwoo Choi Signed-off-by: MyungJoo Ham diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig index a5be56e..8766cba 100644 --- a/drivers/devfreq/Kconfig +++ b/drivers/devfreq/Kconfig @@ -76,7 +76,7 @@ comment "DEVFREQ Drivers" config ARM_EXYNOS_BUS_DEVFREQ tristate "ARM EXYNOS Generic Memory Bus DEVFREQ Driver" - depends on ARCH_EXYNOS + depends on ARCH_EXYNOS || COMPILE_TEST select DEVFREQ_GOV_SIMPLE_ONDEMAND select DEVFREQ_GOV_PASSIVE select DEVFREQ_EVENT_EXYNOS_PPMU @@ -92,7 +92,7 @@ config ARM_EXYNOS_BUS_DEVFREQ config ARM_TEGRA_DEVFREQ tristate "Tegra DEVFREQ Driver" - depends on ARCH_TEGRA_124_SOC + depends on ARCH_TEGRA_124_SOC || COMPILE_TEST select DEVFREQ_GOV_SIMPLE_ONDEMAND select PM_OPP help diff --git a/drivers/devfreq/event/Kconfig b/drivers/devfreq/event/Kconfig index eb6f74a..1d1cd78 100644 --- a/drivers/devfreq/event/Kconfig +++ b/drivers/devfreq/event/Kconfig @@ -15,7 +15,7 @@ if PM_DEVFREQ_EVENT config DEVFREQ_EVENT_EXYNOS_NOCP tristate "EXYNOS NoC (Network On Chip) Probe DEVFREQ event Driver" - depends on ARCH_EXYNOS + depends on ARCH_EXYNOS || COMPILE_TEST select PM_OPP help This add the devfreq-event driver for Exynos SoC. It provides NoC @@ -23,7 +23,7 @@ config DEVFREQ_EVENT_EXYNOS_NOCP config DEVFREQ_EVENT_EXYNOS_PPMU tristate "EXYNOS PPMU (Platform Performance Monitoring Unit) DEVFREQ event Driver" - depends on ARCH_EXYNOS + depends on ARCH_EXYNOS || COMPILE_TEST select PM_OPP help This add the devfreq-event driver for Exynos SoC. It provides PPMU -- cgit v0.10.2 From 989a0fc7412fe2a6f327b2e38858d6ac08ac6fdd Mon Sep 17 00:00:00 2001 From: Jisheng Zhang Date: Thu, 25 Aug 2016 20:06:14 +0800 Subject: PM / devfreq: fix Kconfig indent style Use tab rather than space to indent, and tab + two spaces to indent help message. Signed-off-by: Jisheng Zhang Signed-off-by: MyungJoo Ham diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig index 8766cba..f08502c 100644 --- a/drivers/devfreq/Kconfig +++ b/drivers/devfreq/Kconfig @@ -91,14 +91,14 @@ config ARM_EXYNOS_BUS_DEVFREQ This does not yet operate with optimal voltages. config ARM_TEGRA_DEVFREQ - tristate "Tegra DEVFREQ Driver" - depends on ARCH_TEGRA_124_SOC || COMPILE_TEST - select DEVFREQ_GOV_SIMPLE_ONDEMAND - select PM_OPP - help - This adds the DEVFREQ driver for the Tegra family of SoCs. - It reads ACTMON counters of memory controllers and adjusts the - operating frequencies and voltages with OPP support. + tristate "Tegra DEVFREQ Driver" + depends on ARCH_TEGRA_124_SOC || COMPILE_TEST + select DEVFREQ_GOV_SIMPLE_ONDEMAND + select PM_OPP + help + This adds the DEVFREQ driver for the Tegra family of SoCs. + It reads ACTMON counters of memory controllers and adjusts the + operating frequencies and voltages with OPP support. source "drivers/devfreq/event/Kconfig" -- cgit v0.10.2 From c8a9a6daccad495c48d5435d3487956ce01bc6a1 Mon Sep 17 00:00:00 2001 From: Lin Huang Date: Thu, 4 Aug 2016 19:32:33 +0900 Subject: PM / devfreq: event: remove duplicate devfreq_event_get_drvdata() there define two devfreq_event_get_drvdata() function in devfreq-event.h when disable CONFIG_PM_DEVFREQ_EVENT, it will lead to build fail. So remove devfreq_event_get_drvdata() function. Fixes: f262f28c1470 ("PM / devfreq: event: Add devfreq_event class") Cc: stable@vger.kernel.org Signed-off-by: Lin Huang Signed-off-by: Chanwoo Choi Signed-off-by: MyungJoo Ham diff --git a/include/linux/devfreq-event.h b/include/linux/devfreq-event.h index 0a83a1e..4db00b0 100644 --- a/include/linux/devfreq-event.h +++ b/include/linux/devfreq-event.h @@ -148,11 +148,6 @@ static inline int devfreq_event_reset_event(struct devfreq_event_dev *edev) return -EINVAL; } -static inline void *devfreq_event_get_drvdata(struct devfreq_event_dev *edev) -{ - return ERR_PTR(-EINVAL); -} - static inline struct devfreq_event_dev *devfreq_event_get_edev_by_phandle( struct device *dev, int index) { -- cgit v0.10.2 From d3d81969e38023265dfe5de18ea6a4d898dc6375 Mon Sep 17 00:00:00 2001 From: Lin Huang Date: Mon, 5 Sep 2016 13:06:07 +0800 Subject: Documentation: bindings: add dt documentation for dfi controller This patch adds the documentation for rockchip dfi devfreq-event driver. Signed-off-by: Lin Huang Acked-by: Chanwoo Choi Signed-off-by: MyungJoo Ham diff --git a/Documentation/devicetree/bindings/devfreq/event/rockchip-dfi.txt b/Documentation/devicetree/bindings/devfreq/event/rockchip-dfi.txt new file mode 100644 index 0000000..f223313 --- /dev/null +++ b/Documentation/devicetree/bindings/devfreq/event/rockchip-dfi.txt @@ -0,0 +1,19 @@ + +* Rockchip rk3399 DFI device + +Required properties: +- compatible: Must be "rockchip,rk3399-dfi". +- reg: physical base address of each DFI and length of memory mapped region +- rockchip,pmu: phandle to the syscon managing the "pmu general register files" +- clocks: phandles for clock specified in "clock-names" property +- clock-names : the name of clock used by the DFI, must be "pclk_ddr_mon"; + +Example: + dfi: dfi@0xff630000 { + compatible = "rockchip,rk3399-dfi"; + reg = <0x00 0xff630000 0x00 0x4000>; + rockchip,pmu = <&pmugrf>; + clocks = <&cru PCLK_DDR_MON>; + clock-names = "pclk_ddr_mon"; + status = "disabled"; + }; -- cgit v0.10.2 From b9d1262bca0afcbb67fdb309ed45b34f0226e964 Mon Sep 17 00:00:00 2001 From: Lin Huang Date: Mon, 5 Sep 2016 13:06:08 +0800 Subject: PM / devfreq: event: support rockchip dfi controller on rk3399 platform, there is dfi conroller can monitor ddr load, base on this result, we can do ddr freqency scaling. Signed-off-by: Lin Huang Signed-off-by: MyungJoo Ham Acked-by: Chanwoo Choi diff --git a/drivers/devfreq/event/Kconfig b/drivers/devfreq/event/Kconfig index 1d1cd78..0fdae86 100644 --- a/drivers/devfreq/event/Kconfig +++ b/drivers/devfreq/event/Kconfig @@ -30,4 +30,11 @@ config DEVFREQ_EVENT_EXYNOS_PPMU (Platform Performance Monitoring Unit) counters to estimate the utilization of each module. +config DEVFREQ_EVENT_ROCKCHIP_DFI + tristate "ROCKCHIP DFI DEVFREQ event Driver" + depends on ARCH_ROCKCHIP + help + This add the devfreq-event driver for Rockchip SoC. It provides DFI + (DDR Monitor Module) driver to count ddr load. + endif # PM_DEVFREQ_EVENT diff --git a/drivers/devfreq/event/Makefile b/drivers/devfreq/event/Makefile index 3d6afd3..dda7090 100644 --- a/drivers/devfreq/event/Makefile +++ b/drivers/devfreq/event/Makefile @@ -2,3 +2,4 @@ obj-$(CONFIG_DEVFREQ_EVENT_EXYNOS_NOCP) += exynos-nocp.o obj-$(CONFIG_DEVFREQ_EVENT_EXYNOS_PPMU) += exynos-ppmu.o +obj-$(CONFIG_DEVFREQ_EVENT_ROCKCHIP_DFI) += rockchip-dfi.o diff --git a/drivers/devfreq/event/rockchip-dfi.c b/drivers/devfreq/event/rockchip-dfi.c new file mode 100644 index 0000000..43fcc5a --- /dev/null +++ b/drivers/devfreq/event/rockchip-dfi.c @@ -0,0 +1,256 @@ +/* + * Copyright (c) 2016, Fuzhou Rockchip Electronics Co., Ltd + * Author: Lin Huang + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define RK3399_DMC_NUM_CH 2 + +/* DDRMON_CTRL */ +#define DDRMON_CTRL 0x04 +#define CLR_DDRMON_CTRL (0x1f0000 << 0) +#define LPDDR4_EN (0x10001 << 4) +#define HARDWARE_EN (0x10001 << 3) +#define LPDDR3_EN (0x10001 << 2) +#define SOFTWARE_EN (0x10001 << 1) +#define SOFTWARE_DIS (0x10000 << 1) +#define TIME_CNT_EN (0x10001 << 0) + +#define DDRMON_CH0_COUNT_NUM 0x28 +#define DDRMON_CH0_DFI_ACCESS_NUM 0x2c +#define DDRMON_CH1_COUNT_NUM 0x3c +#define DDRMON_CH1_DFI_ACCESS_NUM 0x40 + +/* pmu grf */ +#define PMUGRF_OS_REG2 0x308 +#define DDRTYPE_SHIFT 13 +#define DDRTYPE_MASK 7 + +enum { + DDR3 = 3, + LPDDR3 = 6, + LPDDR4 = 7, + UNUSED = 0xFF +}; + +struct dmc_usage { + u32 access; + u32 total; +}; + +/* + * The dfi controller can monitor DDR load. It has an upper and lower threshold + * for the operating points. Whenever the usage leaves these bounds an event is + * generated to indicate the DDR frequency should be changed. + */ +struct rockchip_dfi { + struct devfreq_event_dev *edev; + struct devfreq_event_desc *desc; + struct dmc_usage ch_usage[RK3399_DMC_NUM_CH]; + struct device *dev; + void __iomem *regs; + struct regmap *regmap_pmu; + struct clk *clk; +}; + +static void rockchip_dfi_start_hardware_counter(struct devfreq_event_dev *edev) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + void __iomem *dfi_regs = info->regs; + u32 val; + u32 ddr_type; + + /* get ddr type */ + regmap_read(info->regmap_pmu, PMUGRF_OS_REG2, &val); + ddr_type = (val >> DDRTYPE_SHIFT) & DDRTYPE_MASK; + + /* clear DDRMON_CTRL setting */ + writel_relaxed(CLR_DDRMON_CTRL, dfi_regs + DDRMON_CTRL); + + /* set ddr type to dfi */ + if (ddr_type == LPDDR3) + writel_relaxed(LPDDR3_EN, dfi_regs + DDRMON_CTRL); + else if (ddr_type == LPDDR4) + writel_relaxed(LPDDR4_EN, dfi_regs + DDRMON_CTRL); + + /* enable count, use software mode */ + writel_relaxed(SOFTWARE_EN, dfi_regs + DDRMON_CTRL); +} + +static void rockchip_dfi_stop_hardware_counter(struct devfreq_event_dev *edev) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + void __iomem *dfi_regs = info->regs; + + writel_relaxed(SOFTWARE_DIS, dfi_regs + DDRMON_CTRL); +} + +static int rockchip_dfi_get_busier_ch(struct devfreq_event_dev *edev) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + u32 tmp, max = 0; + u32 i, busier_ch = 0; + void __iomem *dfi_regs = info->regs; + + rockchip_dfi_stop_hardware_counter(edev); + + /* Find out which channel is busier */ + for (i = 0; i < RK3399_DMC_NUM_CH; i++) { + info->ch_usage[i].access = readl_relaxed(dfi_regs + + DDRMON_CH0_DFI_ACCESS_NUM + i * 20) * 4; + info->ch_usage[i].total = readl_relaxed(dfi_regs + + DDRMON_CH0_COUNT_NUM + i * 20); + tmp = info->ch_usage[i].access; + if (tmp > max) { + busier_ch = i; + max = tmp; + } + } + rockchip_dfi_start_hardware_counter(edev); + + return busier_ch; +} + +static int rockchip_dfi_disable(struct devfreq_event_dev *edev) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + + rockchip_dfi_stop_hardware_counter(edev); + clk_disable_unprepare(info->clk); + + return 0; +} + +static int rockchip_dfi_enable(struct devfreq_event_dev *edev) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + int ret; + + ret = clk_prepare_enable(info->clk); + if (ret) { + dev_err(&edev->dev, "failed to enable dfi clk: %d\n", ret); + return ret; + } + + rockchip_dfi_start_hardware_counter(edev); + return 0; +} + +static int rockchip_dfi_set_event(struct devfreq_event_dev *edev) +{ + return 0; +} + +static int rockchip_dfi_get_event(struct devfreq_event_dev *edev, + struct devfreq_event_data *edata) +{ + struct rockchip_dfi *info = devfreq_event_get_drvdata(edev); + int busier_ch; + + busier_ch = rockchip_dfi_get_busier_ch(edev); + + edata->load_count = info->ch_usage[busier_ch].access; + edata->total_count = info->ch_usage[busier_ch].total; + + return 0; +} + +static const struct devfreq_event_ops rockchip_dfi_ops = { + .disable = rockchip_dfi_disable, + .enable = rockchip_dfi_enable, + .get_event = rockchip_dfi_get_event, + .set_event = rockchip_dfi_set_event, +}; + +static const struct of_device_id rockchip_dfi_id_match[] = { + { .compatible = "rockchip,rk3399-dfi" }, + { }, +}; + +static int rockchip_dfi_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct rockchip_dfi *data; + struct resource *res; + struct devfreq_event_desc *desc; + struct device_node *np = pdev->dev.of_node, *node; + + data = devm_kzalloc(dev, sizeof(struct rockchip_dfi), GFP_KERNEL); + if (!data) + return -ENOMEM; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + data->regs = devm_ioremap_resource(&pdev->dev, res); + if (IS_ERR(data->regs)) + return PTR_ERR(data->regs); + + data->clk = devm_clk_get(dev, "pclk_ddr_mon"); + if (IS_ERR(data->clk)) { + dev_err(dev, "Cannot get the clk dmc_clk\n"); + return PTR_ERR(data->clk); + }; + + /* try to find the optional reference to the pmu syscon */ + node = of_parse_phandle(np, "rockchip,pmu", 0); + if (node) { + data->regmap_pmu = syscon_node_to_regmap(node); + if (IS_ERR(data->regmap_pmu)) + return PTR_ERR(data->regmap_pmu); + } + data->dev = dev; + + desc = devm_kzalloc(dev, sizeof(*desc), GFP_KERNEL); + if (!desc) + return -ENOMEM; + + desc->ops = &rockchip_dfi_ops; + desc->driver_data = data; + desc->name = np->name; + data->desc = desc; + + data->edev = devm_devfreq_event_add_edev(&pdev->dev, desc); + if (IS_ERR(data->edev)) { + dev_err(&pdev->dev, + "failed to add devfreq-event device\n"); + return PTR_ERR(data->edev); + } + + platform_set_drvdata(pdev, data); + + return 0; +} + +static struct platform_driver rockchip_dfi_driver = { + .probe = rockchip_dfi_probe, + .driver = { + .name = "rockchip-dfi", + .of_match_table = rockchip_dfi_id_match, + }, +}; +module_platform_driver(rockchip_dfi_driver); + +MODULE_LICENSE("GPL v2"); +MODULE_AUTHOR("Lin Huang "); +MODULE_DESCRIPTION("Rockchip DFI driver"); -- cgit v0.10.2 From c1ceb8f7c167eeae79e083417cce8995de3b9238 Mon Sep 17 00:00:00 2001 From: Lin Huang Date: Mon, 5 Sep 2016 13:06:09 +0800 Subject: Documentation: bindings: add dt documentation for rk3399 dmc This patch adds the documentation for rockchip rk3399 dmc driver. Signed-off-by: Lin Huang Reviewed-by: Chanwoo Choi Signed-off-by: MyungJoo Ham diff --git a/Documentation/devicetree/bindings/devfreq/rk3399_dmc.txt b/Documentation/devicetree/bindings/devfreq/rk3399_dmc.txt new file mode 100644 index 0000000..7a9e860 --- /dev/null +++ b/Documentation/devicetree/bindings/devfreq/rk3399_dmc.txt @@ -0,0 +1,209 @@ +* Rockchip rk3399 DMC(Dynamic Memory Controller) device + +Required properties: +- compatible: Must be "rockchip,rk3399-dmc". +- devfreq-events: Node to get DDR loading, Refer to + Documentation/devicetree/bindings/devfreq/ + rockchip-dfi.txt +- interrupts: The interrupt number to the CPU. The interrupt + specifier format depends on the interrupt controller. + It should be DCF interrupts, when DDR dvfs finish, + it will happen. +- clocks: Phandles for clock specified in "clock-names" property +- clock-names : The name of clock used by the DFI, must be + "pclk_ddr_mon"; +- operating-points-v2: Refer to Documentation/devicetree/bindings/power/opp.txt + for details. +- center-supply: DMC supply node. +- status: Marks the node enabled/disabled. + +Following properties are ddr timing: + +- rockchip,dram_speed_bin : Value reference include/dt-bindings/clock/ddr.h, + it select ddr3 cl-trp-trcd type, default value + "DDR3_DEFAULT".it must selected according to + "Speed Bin" in ddr3 datasheet, DO NOT use + smaller "Speed Bin" than ddr3 exactly is. + +- rockchip,pd_idle : Config the PD_IDLE value, defined the power-down + idle period, memories are places into power-down + mode if bus is idle for PD_IDLE DFI clocks. + +- rockchip,sr_idle : Configure the SR_IDLE value, defined the + selfrefresh idle period, memories are places + into self-refresh mode if bus is idle for + SR_IDLE*1024 DFI clocks (DFI clocks freq is + half of dram's clocks), defaule value is "0". + +- rockchip,sr_mc_gate_idle : Defined the self-refresh with memory and + controller clock gating idle period, memories + are places into self-refresh mode and memory + controller clock arg gating if bus is idle for + sr_mc_gate_idle*1024 DFI clocks. + +- rockchip,srpd_lite_idle : Defined the self-refresh power down idle + period, memories are places into self-refresh + power down mode if bus is idle for + srpd_lite_idle*1024 DFI clocks. This parameter + is for LPDDR4 only. + +- rockchip,standby_idle : Defined the standby idle period, memories are + places into self-refresh than controller, pi, + phy and dram clock will gating if bus is idle + for standby_idle * DFI clocks. + +- rockchip,dram_dll_disb_freq : It's defined the DDR3 dll bypass frequency in + MHz, when ddr freq less than DRAM_DLL_DISB_FREQ, + ddr3 dll will bypssed note: if dll was bypassed, + the odt also stop working. + +- rockchip,phy_dll_disb_freq : Defined the PHY dll bypass frequency in + MHz (Mega Hz), when ddr freq less than + DRAM_DLL_DISB_FREQ, phy dll will bypssed. + note: phy dll and phy odt are independent. + +- rockchip,ddr3_odt_disb_freq : When dram type is DDR3, this parameter defined + the odt disable frequency in MHz (Mega Hz), + when ddr frequency less then ddr3_odt_disb_freq, + the odt on dram side and controller side are + both disabled. + +- rockchip,ddr3_drv : When dram type is DDR3, this parameter define + the dram side driver stength in ohm, default + value is DDR3_DS_40ohm. + +- rockchip,ddr3_odt : When dram type is DDR3, this parameter define + the dram side ODT stength in ohm, default value + is DDR3_ODT_120ohm. + +- rockchip,phy_ddr3_ca_drv : When dram type is DDR3, this parameter define + the phy side CA line(incluing command line, + address line and clock line) driver strength. + Default value is PHY_DRV_ODT_40. + +- rockchip,phy_ddr3_dq_drv : When dram type is DDR3, this parameter define + the phy side DQ line(incluing DQS/DQ/DM line) + driver strength. default value is PHY_DRV_ODT_40. + +- rockchip,phy_ddr3_odt : When dram type is DDR3, this parameter define the + phy side odt strength, default value is + PHY_DRV_ODT_240. + +- rockchip,lpddr3_odt_disb_freq : When dram type is LPDDR3, this parameter defined + then odt disable frequency in MHz (Mega Hz), + when ddr frequency less then ddr3_odt_disb_freq, + the odt on dram side and controller side are + both disabled. + +- rockchip,lpddr3_drv : When dram type is LPDDR3, this parameter define + the dram side driver stength in ohm, default + value is LP3_DS_34ohm. + +- rockchip,lpddr3_odt : When dram type is LPDDR3, this parameter define + the dram side ODT stength in ohm, default value + is LP3_ODT_240ohm. + +- rockchip,phy_lpddr3_ca_drv : When dram type is LPDDR3, this parameter define + the phy side CA line(incluing command line, + address line and clock line) driver strength. + default value is PHY_DRV_ODT_40. + +- rockchip,phy_lpddr3_dq_drv : When dram type is LPDDR3, this parameter define + the phy side DQ line(incluing DQS/DQ/DM line) + driver strength. default value is + PHY_DRV_ODT_40. + +- rockchip,phy_lpddr3_odt : When dram type is LPDDR3, this parameter define + the phy side odt strength, default value is + PHY_DRV_ODT_240. + +- rockchip,lpddr4_odt_disb_freq : When dram type is LPDDR4, this parameter + defined the odt disable frequency in + MHz (Mega Hz), when ddr frequency less then + ddr3_odt_disb_freq, the odt on dram side and + controller side are both disabled. + +- rockchip,lpddr4_drv : When dram type is LPDDR4, this parameter define + the dram side driver stength in ohm, default + value is LP4_PDDS_60ohm. + +- rockchip,lpddr4_dq_odt : When dram type is LPDDR4, this parameter define + the dram side ODT on dqs/dq line stength in ohm, + default value is LP4_DQ_ODT_40ohm. + +- rockchip,lpddr4_ca_odt : When dram type is LPDDR4, this parameter define + the dram side ODT on ca line stength in ohm, + default value is LP4_CA_ODT_40ohm. + +- rockchip,phy_lpddr4_ca_drv : When dram type is LPDDR4, this parameter define + the phy side CA line(incluing command address + line) driver strength. default value is + PHY_DRV_ODT_40. + +- rockchip,phy_lpddr4_ck_cs_drv : When dram type is LPDDR4, this parameter define + the phy side clock line and cs line driver + strength. default value is PHY_DRV_ODT_80. + +- rockchip,phy_lpddr4_dq_drv : When dram type is LPDDR4, this parameter define + the phy side DQ line(incluing DQS/DQ/DM line) + driver strength. default value is PHY_DRV_ODT_80. + +- rockchip,phy_lpddr4_odt : When dram type is LPDDR4, this parameter define + the phy side odt strength, default value is + PHY_DRV_ODT_60. + +Example: + dmc_opp_table: dmc_opp_table { + compatible = "operating-points-v2"; + + opp00 { + opp-hz = /bits/ 64 <300000000>; + opp-microvolt = <900000>; + }; + opp01 { + opp-hz = /bits/ 64 <666000000>; + opp-microvolt = <900000>; + }; + }; + + dmc: dmc { + compatible = "rockchip,rk3399-dmc"; + devfreq-events = <&dfi>; + interrupts = ; + clocks = <&cru SCLK_DDRCLK>; + clock-names = "dmc_clk"; + operating-points-v2 = <&dmc_opp_table>; + center-supply = <&ppvar_centerlogic>; + upthreshold = <15>; + downdifferential = <10>; + rockchip,ddr3_speed_bin = <21>; + rockchip,pd_idle = <0x40>; + rockchip,sr_idle = <0x2>; + rockchip,sr_mc_gate_idle = <0x3>; + rockchip,srpd_lite_idle = <0x4>; + rockchip,standby_idle = <0x2000>; + rockchip,dram_dll_dis_freq = <300>; + rockchip,phy_dll_dis_freq = <125>; + rockchip,auto_pd_dis_freq = <666>; + rockchip,ddr3_odt_dis_freq = <333>; + rockchip,ddr3_drv = ; + rockchip,ddr3_odt = ; + rockchip,phy_ddr3_ca_drv = ; + rockchip,phy_ddr3_dq_drv = ; + rockchip,phy_ddr3_odt = ; + rockchip,lpddr3_odt_dis_freq = <333>; + rockchip,lpddr3_drv = ; + rockchip,lpddr3_odt = ; + rockchip,phy_lpddr3_ca_drv = ; + rockchip,phy_lpddr3_dq_drv = ; + rockchip,phy_lpddr3_odt = ; + rockchip,lpddr4_odt_dis_freq = <333>; + rockchip,lpddr4_drv = ; + rockchip,lpddr4_dq_odt = ; + rockchip,lpddr4_ca_odt = ; + rockchip,phy_lpddr4_ca_drv = ; + rockchip,phy_lpddr4_ck_cs_drv = ; + rockchip,phy_lpddr4_dq_drv = ; + rockchip,phy_lpddr4_odt = ; + status = "disabled"; + }; -- cgit v0.10.2 From 5a893e31a636cca3798af2db5aee8d3d144b1e1e Mon Sep 17 00:00:00 2001 From: Lin Huang Date: Mon, 5 Sep 2016 13:06:10 +0800 Subject: PM / devfreq: rockchip: add devfreq driver for rk3399 dmc base on dfi result, we do ddr frequency scaling, register dmc driver to devfreq framework, and use simple-ondemand policy. Signed-off-by: Lin Huang Signed-off-by: MyngJoo Ham Reviewed-by: Chanwoo Choi diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig index f08502c..cadd56e 100644 --- a/drivers/devfreq/Kconfig +++ b/drivers/devfreq/Kconfig @@ -100,6 +100,17 @@ config ARM_TEGRA_DEVFREQ It reads ACTMON counters of memory controllers and adjusts the operating frequencies and voltages with OPP support. +config ARM_RK3399_DMC_DEVFREQ + tristate "ARM RK3399 DMC DEVFREQ Driver" + depends on ARCH_ROCKCHIP + select DEVFREQ_EVENT_ROCKCHIP_DFI + select DEVFREQ_GOV_SIMPLE_ONDEMAND + select PM_OPP + help + This adds the DEVFREQ driver for the RK3399 DMC(Dynamic Memory Controller). + It sets the frequency for the memory controller and reads the usage counts + from hardware. + source "drivers/devfreq/event/Kconfig" endif # PM_DEVFREQ diff --git a/drivers/devfreq/Makefile b/drivers/devfreq/Makefile index 09f11d9..fbff40a 100644 --- a/drivers/devfreq/Makefile +++ b/drivers/devfreq/Makefile @@ -8,6 +8,7 @@ obj-$(CONFIG_DEVFREQ_GOV_PASSIVE) += governor_passive.o # DEVFREQ Drivers obj-$(CONFIG_ARM_EXYNOS_BUS_DEVFREQ) += exynos-bus.o +obj-$(CONFIG_ARM_RK3399_DMC_DEVFREQ) += rk3399_dmc.o obj-$(CONFIG_ARM_TEGRA_DEVFREQ) += tegra-devfreq.o # DEVFREQ Event Drivers diff --git a/drivers/devfreq/rk3399_dmc.c b/drivers/devfreq/rk3399_dmc.c new file mode 100644 index 0000000..54d65f2 --- /dev/null +++ b/drivers/devfreq/rk3399_dmc.c @@ -0,0 +1,480 @@ +/* + * Copyright (c) 2016, Fuzhou Rockchip Electronics Co., Ltd. + * Author: Lin Huang + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +struct dram_timing { + unsigned int ddr3_speed_bin; + unsigned int pd_idle; + unsigned int sr_idle; + unsigned int sr_mc_gate_idle; + unsigned int srpd_lite_idle; + unsigned int standby_idle; + unsigned int auto_pd_dis_freq; + unsigned int dram_dll_dis_freq; + unsigned int phy_dll_dis_freq; + unsigned int ddr3_odt_dis_freq; + unsigned int ddr3_drv; + unsigned int ddr3_odt; + unsigned int phy_ddr3_ca_drv; + unsigned int phy_ddr3_dq_drv; + unsigned int phy_ddr3_odt; + unsigned int lpddr3_odt_dis_freq; + unsigned int lpddr3_drv; + unsigned int lpddr3_odt; + unsigned int phy_lpddr3_ca_drv; + unsigned int phy_lpddr3_dq_drv; + unsigned int phy_lpddr3_odt; + unsigned int lpddr4_odt_dis_freq; + unsigned int lpddr4_drv; + unsigned int lpddr4_dq_odt; + unsigned int lpddr4_ca_odt; + unsigned int phy_lpddr4_ca_drv; + unsigned int phy_lpddr4_ck_cs_drv; + unsigned int phy_lpddr4_dq_drv; + unsigned int phy_lpddr4_odt; +}; + +struct rk3399_dmcfreq { + struct device *dev; + struct devfreq *devfreq; + struct devfreq_simple_ondemand_data ondemand_data; + struct clk *dmc_clk; + struct devfreq_event_dev *edev; + struct mutex lock; + struct dram_timing timing; + + /* + * DDR Converser of Frequency (DCF) is used to implement DDR frequency + * conversion without the participation of CPU, we will implement and + * control it in arm trust firmware. + */ + wait_queue_head_t wait_dcf_queue; + int irq; + int wait_dcf_flag; + struct regulator *vdd_center; + unsigned long rate, target_rate; + unsigned long volt, target_volt; + struct dev_pm_opp *curr_opp; +}; + +static int rk3399_dmcfreq_target(struct device *dev, unsigned long *freq, + u32 flags) +{ + struct rk3399_dmcfreq *dmcfreq = dev_get_drvdata(dev); + struct dev_pm_opp *opp; + unsigned long old_clk_rate = dmcfreq->rate; + unsigned long target_volt, target_rate; + int err; + + rcu_read_lock(); + opp = devfreq_recommended_opp(dev, freq, flags); + if (IS_ERR(opp)) { + rcu_read_unlock(); + return PTR_ERR(opp); + } + + target_rate = dev_pm_opp_get_freq(opp); + target_volt = dev_pm_opp_get_voltage(opp); + + dmcfreq->rate = dev_pm_opp_get_freq(dmcfreq->curr_opp); + dmcfreq->volt = dev_pm_opp_get_voltage(dmcfreq->curr_opp); + + rcu_read_unlock(); + + if (dmcfreq->rate == target_rate) + return 0; + + mutex_lock(&dmcfreq->lock); + + /* + * If frequency scaling from low to high, adjust voltage first. + * If frequency scaling from high to low, adjust frequency first. + */ + if (old_clk_rate < target_rate) { + err = regulator_set_voltage(dmcfreq->vdd_center, target_volt, + target_volt); + if (err) { + dev_err(dev, "Cannot to set voltage %lu uV\n", + target_volt); + goto out; + } + } + dmcfreq->wait_dcf_flag = 1; + + err = clk_set_rate(dmcfreq->dmc_clk, target_rate); + if (err) { + dev_err(dev, "Cannot to set frequency %lu (%d)\n", + target_rate, err); + regulator_set_voltage(dmcfreq->vdd_center, dmcfreq->volt, + dmcfreq->volt); + goto out; + } + + /* + * Wait until bcf irq happen, it means freq scaling finish in + * arm trust firmware, use 100ms as timeout time. + */ + if (!wait_event_timeout(dmcfreq->wait_dcf_queue, + !dmcfreq->wait_dcf_flag, HZ / 10)) + dev_warn(dev, "Timeout waiting for dcf interrupt\n"); + + /* + * Check the dpll rate, + * There only two result we will get, + * 1. Ddr frequency scaling fail, we still get the old rate. + * 2. Ddr frequency scaling sucessful, we get the rate we set. + */ + dmcfreq->rate = clk_get_rate(dmcfreq->dmc_clk); + + /* If get the incorrect rate, set voltage to old value. */ + if (dmcfreq->rate != target_rate) { + dev_err(dev, "Get wrong ddr frequency, Request frequency %lu,\ + Current frequency %lu\n", target_rate, dmcfreq->rate); + regulator_set_voltage(dmcfreq->vdd_center, dmcfreq->volt, + dmcfreq->volt); + goto out; + } else if (old_clk_rate > target_rate) + err = regulator_set_voltage(dmcfreq->vdd_center, target_volt, + target_volt); + if (err) + dev_err(dev, "Cannot to set vol %lu uV\n", target_volt); + + dmcfreq->curr_opp = opp; +out: + mutex_unlock(&dmcfreq->lock); + return err; +} + +static int rk3399_dmcfreq_get_dev_status(struct device *dev, + struct devfreq_dev_status *stat) +{ + struct rk3399_dmcfreq *dmcfreq = dev_get_drvdata(dev); + struct devfreq_event_data edata; + int ret = 0; + + ret = devfreq_event_get_event(dmcfreq->edev, &edata); + if (ret < 0) + return ret; + + stat->current_frequency = dmcfreq->rate; + stat->busy_time = edata.load_count; + stat->total_time = edata.total_count; + + return ret; +} + +static int rk3399_dmcfreq_get_cur_freq(struct device *dev, unsigned long *freq) +{ + struct rk3399_dmcfreq *dmcfreq = dev_get_drvdata(dev); + + *freq = dmcfreq->rate; + + return 0; +} + +static struct devfreq_dev_profile rk3399_devfreq_dmc_profile = { + .polling_ms = 200, + .target = rk3399_dmcfreq_target, + .get_dev_status = rk3399_dmcfreq_get_dev_status, + .get_cur_freq = rk3399_dmcfreq_get_cur_freq, +}; + +static __maybe_unused int rk3399_dmcfreq_suspend(struct device *dev) +{ + struct rk3399_dmcfreq *dmcfreq = dev_get_drvdata(dev); + int ret = 0; + + ret = devfreq_event_disable_edev(dmcfreq->edev); + if (ret < 0) { + dev_err(dev, "failed to disable the devfreq-event devices\n"); + return ret; + } + + ret = devfreq_suspend_device(dmcfreq->devfreq); + if (ret < 0) { + dev_err(dev, "failed to suspend the devfreq devices\n"); + return ret; + } + + return 0; +} + +static __maybe_unused int rk3399_dmcfreq_resume(struct device *dev) +{ + struct rk3399_dmcfreq *dmcfreq = dev_get_drvdata(dev); + int ret = 0; + + ret = devfreq_event_enable_edev(dmcfreq->edev); + if (ret < 0) { + dev_err(dev, "failed to enable the devfreq-event devices\n"); + return ret; + } + + ret = devfreq_resume_device(dmcfreq->devfreq); + if (ret < 0) { + dev_err(dev, "failed to resume the devfreq devices\n"); + return ret; + } + return ret; +} + +static SIMPLE_DEV_PM_OPS(rk3399_dmcfreq_pm, rk3399_dmcfreq_suspend, + rk3399_dmcfreq_resume); + +static irqreturn_t rk3399_dmc_irq(int irq, void *dev_id) +{ + struct rk3399_dmcfreq *dmcfreq = dev_id; + struct arm_smccc_res res; + + dmcfreq->wait_dcf_flag = 0; + wake_up(&dmcfreq->wait_dcf_queue); + + /* Clear the DCF interrupt */ + arm_smccc_smc(ROCKCHIP_SIP_DRAM_FREQ, 0, 0, + ROCKCHIP_SIP_CONFIG_DRAM_CLR_IRQ, + 0, 0, 0, 0, &res); + + return IRQ_HANDLED; +} + +static int of_get_ddr_timings(struct dram_timing *timing, + struct device_node *np) +{ + int ret = 0; + + ret = of_property_read_u32(np, "rockchip,ddr3_speed_bin", + &timing->ddr3_speed_bin); + ret |= of_property_read_u32(np, "rockchip,pd_idle", + &timing->pd_idle); + ret |= of_property_read_u32(np, "rockchip,sr_idle", + &timing->sr_idle); + ret |= of_property_read_u32(np, "rockchip,sr_mc_gate_idle", + &timing->sr_mc_gate_idle); + ret |= of_property_read_u32(np, "rockchip,srpd_lite_idle", + &timing->srpd_lite_idle); + ret |= of_property_read_u32(np, "rockchip,standby_idle", + &timing->standby_idle); + ret |= of_property_read_u32(np, "rockchip,auto_pd_dis_freq", + &timing->auto_pd_dis_freq); + ret |= of_property_read_u32(np, "rockchip,dram_dll_dis_freq", + &timing->dram_dll_dis_freq); + ret |= of_property_read_u32(np, "rockchip,phy_dll_dis_freq", + &timing->phy_dll_dis_freq); + ret |= of_property_read_u32(np, "rockchip,ddr3_odt_dis_freq", + &timing->ddr3_odt_dis_freq); + ret |= of_property_read_u32(np, "rockchip,ddr3_drv", + &timing->ddr3_drv); + ret |= of_property_read_u32(np, "rockchip,ddr3_odt", + &timing->ddr3_odt); + ret |= of_property_read_u32(np, "rockchip,phy_ddr3_ca_drv", + &timing->phy_ddr3_ca_drv); + ret |= of_property_read_u32(np, "rockchip,phy_ddr3_dq_drv", + &timing->phy_ddr3_dq_drv); + ret |= of_property_read_u32(np, "rockchip,phy_ddr3_odt", + &timing->phy_ddr3_odt); + ret |= of_property_read_u32(np, "rockchip,lpddr3_odt_dis_freq", + &timing->lpddr3_odt_dis_freq); + ret |= of_property_read_u32(np, "rockchip,lpddr3_drv", + &timing->lpddr3_drv); + ret |= of_property_read_u32(np, "rockchip,lpddr3_odt", + &timing->lpddr3_odt); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr3_ca_drv", + &timing->phy_lpddr3_ca_drv); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr3_dq_drv", + &timing->phy_lpddr3_dq_drv); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr3_odt", + &timing->phy_lpddr3_odt); + ret |= of_property_read_u32(np, "rockchip,lpddr4_odt_dis_freq", + &timing->lpddr4_odt_dis_freq); + ret |= of_property_read_u32(np, "rockchip,lpddr4_drv", + &timing->lpddr4_drv); + ret |= of_property_read_u32(np, "rockchip,lpddr4_dq_odt", + &timing->lpddr4_dq_odt); + ret |= of_property_read_u32(np, "rockchip,lpddr4_ca_odt", + &timing->lpddr4_ca_odt); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr4_ca_drv", + &timing->phy_lpddr4_ca_drv); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr4_ck_cs_drv", + &timing->phy_lpddr4_ck_cs_drv); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr4_dq_drv", + &timing->phy_lpddr4_dq_drv); + ret |= of_property_read_u32(np, "rockchip,phy_lpddr4_odt", + &timing->phy_lpddr4_odt); + + return ret; +} + +static int rk3399_dmcfreq_probe(struct platform_device *pdev) +{ + struct arm_smccc_res res; + struct device *dev = &pdev->dev; + struct device_node *np = pdev->dev.of_node; + struct rk3399_dmcfreq *data; + int ret, irq, index, size; + uint32_t *timing; + struct dev_pm_opp *opp; + + irq = platform_get_irq(pdev, 0); + if (irq < 0) { + dev_err(&pdev->dev, "Cannot get the dmc interrupt resource\n"); + return -EINVAL; + } + data = devm_kzalloc(dev, sizeof(struct rk3399_dmcfreq), GFP_KERNEL); + if (!data) + return -ENOMEM; + + mutex_init(&data->lock); + + data->vdd_center = devm_regulator_get(dev, "center"); + if (IS_ERR(data->vdd_center)) { + dev_err(dev, "Cannot get the regulator \"center\"\n"); + return PTR_ERR(data->vdd_center); + } + + data->dmc_clk = devm_clk_get(dev, "dmc_clk"); + if (IS_ERR(data->dmc_clk)) { + dev_err(dev, "Cannot get the clk dmc_clk\n"); + return PTR_ERR(data->dmc_clk); + }; + + data->irq = irq; + ret = devm_request_irq(dev, irq, rk3399_dmc_irq, 0, + dev_name(dev), data); + if (ret) { + dev_err(dev, "Failed to request dmc irq: %d\n", ret); + return ret; + } + + init_waitqueue_head(&data->wait_dcf_queue); + data->wait_dcf_flag = 0; + + data->edev = devfreq_event_get_edev_by_phandle(dev, 0); + if (IS_ERR(data->edev)) + return -EPROBE_DEFER; + + ret = devfreq_event_enable_edev(data->edev); + if (ret < 0) { + dev_err(dev, "failed to enable devfreq-event devices\n"); + return ret; + } + + /* + * Get dram timing and pass it to arm trust firmware, + * the dram drvier in arm trust firmware will get these + * timing and to do dram initial. + */ + if (!of_get_ddr_timings(&data->timing, np)) { + timing = &data->timing.ddr3_speed_bin; + size = sizeof(struct dram_timing) / 4; + for (index = 0; index < size; index++) { + arm_smccc_smc(ROCKCHIP_SIP_DRAM_FREQ, *timing++, index, + ROCKCHIP_SIP_CONFIG_DRAM_SET_PARAM, + 0, 0, 0, 0, &res); + if (res.a0) { + dev_err(dev, "Failed to set dram param: %ld\n", + res.a0); + return -EINVAL; + } + } + } + + arm_smccc_smc(ROCKCHIP_SIP_DRAM_FREQ, 0, 0, + ROCKCHIP_SIP_CONFIG_DRAM_INIT, + 0, 0, 0, 0, &res); + + /* + * We add a devfreq driver to our parent since it has a device tree node + * with operating points. + */ + if (dev_pm_opp_of_add_table(dev)) { + dev_err(dev, "Invalid operating-points in device tree.\n"); + rcu_read_unlock(); + return -EINVAL; + } + + of_property_read_u32(np, "upthreshold", + &data->ondemand_data.upthreshold); + of_property_read_u32(np, "downdifferential", + &data->ondemand_data.downdifferential); + + data->rate = clk_get_rate(data->dmc_clk); + + rcu_read_lock(); + opp = devfreq_recommended_opp(dev, &data->rate, 0); + if (IS_ERR(opp)) { + rcu_read_unlock(); + return PTR_ERR(opp); + } + rcu_read_unlock(); + data->curr_opp = opp; + + rk3399_devfreq_dmc_profile.initial_freq = data->rate; + + data->devfreq = devfreq_add_device(dev, + &rk3399_devfreq_dmc_profile, + "simple_ondemand", + &data->ondemand_data); + if (IS_ERR(data->devfreq)) + return PTR_ERR(data->devfreq); + devm_devfreq_register_opp_notifier(dev, data->devfreq); + + data->dev = dev; + platform_set_drvdata(pdev, data); + + return 0; +} + +static int rk3399_dmcfreq_remove(struct platform_device *pdev) +{ + struct rk3399_dmcfreq *dmcfreq = platform_get_drvdata(pdev); + + regulator_put(dmcfreq->vdd_center); + + return 0; +} + +static const struct of_device_id rk3399dmc_devfreq_of_match[] = { + { .compatible = "rockchip,rk3399-dmc" }, + { }, +}; + +static struct platform_driver rk3399_dmcfreq_driver = { + .probe = rk3399_dmcfreq_probe, + .remove = rk3399_dmcfreq_remove, + .driver = { + .name = "rk3399-dmc-freq", + .pm = &rk3399_dmcfreq_pm, + .of_match_table = rk3399dmc_devfreq_of_match, + }, +}; +module_platform_driver(rk3399_dmcfreq_driver); + +MODULE_LICENSE("GPL v2"); +MODULE_AUTHOR("Lin Huang "); +MODULE_DESCRIPTION("RK3399 dmcfreq driver with devfreq framework"); -- cgit v0.10.2 From 5b3f249c94ce1f46bacd9814385b0ee2d1ae52f3 Mon Sep 17 00:00:00 2001 From: Chen Yu Date: Fri, 19 Aug 2016 12:37:23 +0800 Subject: PM / sleep: Increase default DPM watchdog timeout to 120 Recently we have a new report that, the harddisk can not resume on time due to firmware issues, and got a kernel panic because of DPM watchdog timeout. So adjust the default timeout from 60 to 120 to survive on this platform, and make DPM_WATCHDOG depending on EXPERT. Link: https://bugzilla.kernel.org/show_bug.cgi?id=117971 Suggested-by: Pavel Machek Suggested-by: Rafael J. Wysocki Reported-by: Higuita Signed-off-by: Chen Yu Acked-by: Pavel Machek Signed-off-by: Rafael J. Wysocki diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 68d3ebc..e8517b6 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -186,7 +186,7 @@ config PM_SLEEP_DEBUG config DPM_WATCHDOG bool "Device suspend/resume watchdog" - depends on PM_DEBUG && PSTORE + depends on PM_DEBUG && PSTORE && EXPERT ---help--- Sets up a watchdog timer to capture drivers that are locked up attempting to suspend/resume a device. @@ -197,7 +197,7 @@ config DPM_WATCHDOG config DPM_WATCHDOG_TIMEOUT int "Watchdog timeout in seconds" range 1 120 - default 60 + default 120 depends on DPM_WATCHDOG config PM_TRACE -- cgit v0.10.2 From fa7fd6fa38e36d88bc9f2d0e45e5b9bd0387079f Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Fri, 19 Aug 2016 14:41:00 +0100 Subject: PM / sleep: enable suspend-to-idle even without registered suspend_ops Suspend-to-idle (aka the "freeze" sleep state) is a system sleep state in which all of the processors enter deepest possible idle state and wait for interrupts right after suspending all the devices. There is no hard requirement for a platform to support and register platform specific suspend_ops to enter suspend-to-idle/freeze state. Only deeper system sleep states like PM_SUSPEND_STANDBY and PM_SUSPEND_MEM rely on such low level support/implementation. suspend-to-idle can be entered as along as all the devices can be suspended. This patch enables the support for suspend-to-idle even on systems that don't have any low level support for deeper system sleep states and/or don't register any platform specific suspend_ops. Signed-off-by: Sudeep Holla Tested-by: Andy Gross Signed-off-by: Rafael J. Wysocki diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 7693e39..d971837 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -245,6 +245,7 @@ static inline bool idle_should_freeze(void) return unlikely(suspend_freeze_state == FREEZE_STATE_ENTER); } +extern void __init pm_states_init(void); extern void freeze_set_ops(const struct platform_freeze_ops *ops); extern void freeze_wake(void); @@ -279,6 +280,7 @@ static inline bool pm_resume_via_firmware(void) { return false; } static inline void suspend_set_ops(const struct platform_suspend_ops *ops) {} static inline int pm_suspend(suspend_state_t state) { return -ENOSYS; } static inline bool idle_should_freeze(void) { return false; } +static inline void __init pm_states_init(void) {} static inline void freeze_set_ops(const struct platform_freeze_ops *ops) {} static inline void freeze_wake(void) {} #endif /* !CONFIG_SUSPEND */ diff --git a/kernel/power/main.c b/kernel/power/main.c index 5ea50b1..281a697 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -644,6 +644,7 @@ static int __init pm_init(void) return error; hibernate_image_size_init(); hibernate_reserved_size_init(); + pm_states_init(); power_kobj = kobject_create_and_add("power", NULL); if (!power_kobj) return -ENOMEM; diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index 0acab9d..1e7f5da 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -118,10 +118,18 @@ static bool valid_state(suspend_state_t state) */ static bool relative_states; +void __init pm_states_init(void) +{ + /* + * freeze state should be supported even without any suspend_ops, + * initialize pm_states accordingly here + */ + pm_states[PM_SUSPEND_FREEZE] = pm_labels[relative_states ? 0 : 2]; +} + static int __init sleep_states_setup(char *str) { relative_states = !strncmp(str, "1", 1); - pm_states[PM_SUSPEND_FREEZE] = pm_labels[relative_states ? 0 : 2]; return 1; } @@ -211,7 +219,7 @@ static int platform_suspend_begin(suspend_state_t state) { if (state == PM_SUSPEND_FREEZE && freeze_ops && freeze_ops->begin) return freeze_ops->begin(); - else if (suspend_ops->begin) + else if (suspend_ops && suspend_ops->begin) return suspend_ops->begin(state); else return 0; @@ -221,7 +229,7 @@ static void platform_resume_end(suspend_state_t state) { if (state == PM_SUSPEND_FREEZE && freeze_ops && freeze_ops->end) freeze_ops->end(); - else if (suspend_ops->end) + else if (suspend_ops && suspend_ops->end) suspend_ops->end(); } -- cgit v0.10.2 From ffdf8b867b922d2d3150ccfa330d1c6d22716562 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Tue, 6 Sep 2016 14:18:20 +0200 Subject: cpufreq: dt: Add support for r8a7792 Add the compatible string for supporting the generic cpufreq driver on the Renesas R-Car V2H (r8a7792) SoC. Signed-off-by: Geert Uytterhoeven Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq-dt-platdev.c b/drivers/cpufreq/cpufreq-dt-platdev.c index 285ed3e..276378a 100644 --- a/drivers/cpufreq/cpufreq-dt-platdev.c +++ b/drivers/cpufreq/cpufreq-dt-platdev.c @@ -52,6 +52,7 @@ static const struct of_device_id machines[] __initconst = { { .compatible = "renesas,r8a7779", }, { .compatible = "renesas,r8a7790", }, { .compatible = "renesas,r8a7791", }, + { .compatible = "renesas,r8a7792", }, { .compatible = "renesas,r8a7793", }, { .compatible = "renesas,r8a7794", }, { .compatible = "renesas,sh73a0", }, -- cgit v0.10.2 From 1ad1410f632d4141221634308a5e56f339f92009 Mon Sep 17 00:00:00 2001 From: Anisse Astier Date: Fri, 9 Sep 2016 10:43:32 +0200 Subject: PM / Hibernate: allow hibernation with PAGE_POISONING_ZERO PAGE_POISONING_ZERO disables zeroing new pages on alloc, they are poisoned (zeroed) as they become available. In the hibernate use case, free pages will appear in the system without being cleared, left there by the loading kernel. This patch will make sure free pages are cleared on resume when PAGE_POISONING_ZERO is enabled. We free the pages just after resume because we can't do it later: going through any device resume code might allocate some memory and invalidate the free pages bitmap. Thus we don't need to disable hibernation when PAGE_POISONING_ZERO is enabled. Signed-off-by: Anisse Astier Reviewed-by: Kees Cook Acked-by: Pavel Machek Signed-off-by: Rafael J. Wysocki diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index 33c79b6..b26dbc4 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -306,8 +306,10 @@ static int create_image(int platform_mode) if (error) printk(KERN_ERR "PM: Error %d creating hibernation image\n", error); - if (!in_suspend) + if (!in_suspend) { events_check_enabled = false; + clear_free_pages(); + } platform_leave(platform_mode); @@ -1189,22 +1191,6 @@ static int __init nohibernate_setup(char *str) return 1; } -static int __init page_poison_nohibernate_setup(char *str) -{ -#ifdef CONFIG_PAGE_POISONING_ZERO - /* - * The zeroing option for page poison skips the checks on alloc. - * since hibernation doesn't save free pages there's no way to - * guarantee the pages will still be zeroed. - */ - if (!strcmp(str, "on")) { - pr_info("Disabling hibernation due to page poisoning\n"); - return nohibernate_setup(str); - } -#endif - return 1; -} - __setup("noresume", noresume_setup); __setup("resume_offset=", resume_offset_setup); __setup("resume=", resume_setup); @@ -1212,4 +1198,3 @@ __setup("hibernate=", hibernate_setup); __setup("resumewait", resumewait_setup); __setup("resumedelay=", resumedelay_setup); __setup("nohibernate", nohibernate_setup); -__setup("page_poison=", page_poison_nohibernate_setup); diff --git a/kernel/power/power.h b/kernel/power/power.h index 242d8b8..56d1d0d 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -110,6 +110,8 @@ extern int create_basic_memory_bitmaps(void); extern void free_basic_memory_bitmaps(void); extern int hibernate_preallocate_memory(void); +extern void clear_free_pages(void); + /** * Auxiliary structure used for reading the snapshot image data and * metadata from and writing them to the list of page backup entries diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index b022284..4f0f060 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -1132,6 +1132,28 @@ void free_basic_memory_bitmaps(void) pr_debug("PM: Basic memory bitmaps freed\n"); } +void clear_free_pages(void) +{ +#ifdef CONFIG_PAGE_POISONING_ZERO + struct memory_bitmap *bm = free_pages_map; + unsigned long pfn; + + if (WARN_ON(!(free_pages_map))) + return; + + memory_bm_position_reset(bm); + pfn = memory_bm_next_pfn(bm); + while (pfn != BM_END_OF_MAP) { + if (pfn_valid(pfn)) + clear_highpage(pfn_to_page(pfn)); + + pfn = memory_bm_next_pfn(bm); + } + memory_bm_position_reset(bm); + pr_info("PM: free pages cleared after restore\n"); +#endif /* PAGE_POISONING_ZERO */ +} + /** * snapshot_additional_pages - Estimate the number of extra pages needed. * @zone: Memory zone to carry out the computation for. diff --git a/mm/Kconfig.debug b/mm/Kconfig.debug index 22f4cd9..afcc550 100644 --- a/mm/Kconfig.debug +++ b/mm/Kconfig.debug @@ -76,8 +76,6 @@ config PAGE_POISONING_ZERO no longer necessary to write zeros when GFP_ZERO is used on allocation. - Enabling page poisoning with this option will disable hibernation - If unsure, say N bool -- cgit v0.10.2 From dabe73cbeba463e559741355afce010cea7dacaf Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 9 Sep 2016 13:05:24 +0200 Subject: MAINTAINERS: Add Documentation/cpu-freq/ I am told the cpufreq documentation updates should go to the PM list. Signed-off-by: Jean Delvare Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/MAINTAINERS b/MAINTAINERS index db814a8..f88a50c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3281,6 +3281,7 @@ L: linux-pm@vger.kernel.org S: Maintained T: git git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git T: git git://git.linaro.org/people/vireshk/linux.git (For ARM Updates) +F: Documentation/cpu-freq/ F: drivers/cpufreq/ F: include/linux/cpufreq.h -- cgit v0.10.2 From e86eee6bc2aaa6b3637f6497b26beee09a91bde9 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 9 Sep 2016 16:48:06 +0530 Subject: cpufreq: dt: Remove unused code This is leftover from an earlier patch which removed the usage of platform data but forgot to remove this line. Remove it now. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index 3957de8..2bd2053 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -366,8 +366,6 @@ static int dt_cpufreq_probe(struct platform_device *pdev) if (ret) return ret; - dt_cpufreq_driver.driver_data = dev_get_platdata(&pdev->dev); - ret = cpufreq_register_driver(&dt_cpufreq_driver); if (ret) dev_err(&pdev->dev, "failed register driver: %d\n", ret); -- cgit v0.10.2 From 33cc4fc1b2147ee3cf36d02c1106456cb276c043 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 9 Sep 2016 16:48:07 +0530 Subject: cpufreq: dt: Update kconfig description The cpufreq DT driver also supports systems that have multiple clock/voltage domains for CPUs, i.e. multiple policy systems. The description of the Kconfig entry was never updated after the driver was modified to support such systems, fix it. Reported-by: Juri Lelli Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig index 74919aa..9be9013 100644 --- a/drivers/cpufreq/Kconfig +++ b/drivers/cpufreq/Kconfig @@ -225,7 +225,7 @@ config CPUFREQ_DT help This adds a generic DT based cpufreq driver for frequency management. It supports both uniprocessor (UP) and symmetric multiprocessor (SMP) - systems which share clock and voltage across all CPUs. + systems. If in doubt, say N. -- cgit v0.10.2 From 297a66221d2bed61e7d74d11bf071c7b489fc33d Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 9 Sep 2016 16:48:08 +0530 Subject: cpufreq: dt: Support governor tunables per policy The cpufreq-dt driver is also used for systems with multiple clock/voltage domains for CPUs, i.e. multiple cpufreq policies in a system. And in such cases the platform users may want to enable "governor tunables per policy". Support that via platform data, as not all users of the driver would want that behavior. Reported-by: Juri Lelli Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq-dt-platdev.c b/drivers/cpufreq/cpufreq-dt-platdev.c index 276378a..d2637e1 100644 --- a/drivers/cpufreq/cpufreq-dt-platdev.c +++ b/drivers/cpufreq/cpufreq-dt-platdev.c @@ -11,6 +11,8 @@ #include #include +#include "cpufreq-dt.h" + static const struct of_device_id machines[] __initconst = { { .compatible = "allwinner,sun4i-a10", }, { .compatible = "allwinner,sun5i-a10s", }, @@ -93,7 +95,8 @@ static int __init cpufreq_dt_platdev_init(void) if (!match) return -ENODEV; - return PTR_ERR_OR_ZERO(platform_device_register_simple("cpufreq-dt", -1, - NULL, 0)); + return PTR_ERR_OR_ZERO(platform_device_register_data(NULL, "cpufreq-dt", + -1, match->data, + sizeof(struct cpufreq_dt_platform_data))); } device_initcall(cpufreq_dt_platdev_init); diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index 2bd2053..5c07ae0 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -25,6 +25,8 @@ #include #include +#include "cpufreq-dt.h" + struct private_data { struct device *cpu_dev; struct thermal_cooling_device *cdev; @@ -353,6 +355,7 @@ static struct cpufreq_driver dt_cpufreq_driver = { static int dt_cpufreq_probe(struct platform_device *pdev) { + struct cpufreq_dt_platform_data *data = dev_get_platdata(&pdev->dev); int ret; /* @@ -366,6 +369,9 @@ static int dt_cpufreq_probe(struct platform_device *pdev) if (ret) return ret; + if (data && data->have_governor_per_policy) + dt_cpufreq_driver.flags |= CPUFREQ_HAVE_GOVERNOR_PER_POLICY; + ret = cpufreq_register_driver(&dt_cpufreq_driver); if (ret) dev_err(&pdev->dev, "failed register driver: %d\n", ret); diff --git a/drivers/cpufreq/cpufreq-dt.h b/drivers/cpufreq/cpufreq-dt.h new file mode 100644 index 0000000..54d774e --- /dev/null +++ b/drivers/cpufreq/cpufreq-dt.h @@ -0,0 +1,19 @@ +/* + * Copyright (C) 2016 Linaro + * Viresh Kumar + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __CPUFREQ_DT_H__ +#define __CPUFREQ_DT_H__ + +#include + +struct cpufreq_dt_platform_data { + bool have_governor_per_policy; +}; + +#endif /* __CPUFREQ_DT_H__ */ -- cgit v0.10.2 From 42ce8921cca08ae8d9068968e6658cae856a99df Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sun, 11 Sep 2016 15:06:01 +0200 Subject: intel_pstate: constify local structures For structure types defined in the same file or local header files, find top-level static structure declarations that have the following properties: 1. Never reassigned. 2. Address never taken 3. Not passed to a top-level macro call 4. No pointer or array-typed field passed to a function or stored in a variable. Declare structures having all of these properties as const. Done using Coccinelle. Based on a suggestion by Joe Perches . Signed-off-by: Julia Lawall Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index be9eade..5095b89 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1029,7 +1029,7 @@ static struct cpu_defaults core_params = { }, }; -static struct cpu_defaults silvermont_params = { +static const struct cpu_defaults silvermont_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, @@ -1050,7 +1050,7 @@ static struct cpu_defaults silvermont_params = { }, }; -static struct cpu_defaults airmont_params = { +static const struct cpu_defaults airmont_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, @@ -1071,7 +1071,7 @@ static struct cpu_defaults airmont_params = { }, }; -static struct cpu_defaults knl_params = { +static const struct cpu_defaults knl_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, @@ -1091,7 +1091,7 @@ static struct cpu_defaults knl_params = { }, }; -static struct cpu_defaults bxt_params = { +static const struct cpu_defaults bxt_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, -- cgit v0.10.2 From 26619804e733bbe8a01ab8a438f91e230e91373a Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Mon, 12 Sep 2016 12:07:05 +0530 Subject: cpufreq: create link to policy only for registered CPUs If a cpufreq driver is registered very early in the boot stage (e.g. registered from postcore_initcall()), then cpufreq core may generate kernel warnings for it. In this case, the CPUs are brought online, then the cpufreq driver is registered, and then the CPU topology devices are registered. However, by the time cpufreq_add_dev() gets called, the cpu device isn't stored in the per-cpu variable (cpu_sys_devices,) which is read by get_cpu_device(). So the cpufreq core fails to get device for the CPU, for which cpufreq_add_dev() was called in the first place and we will hit a WARN_ON(!cpu_dev). Even if we reuse the 'dev' parameter passed to cpufreq_add_dev() to avoid that warning, there might be other CPUs online that share the policy with the cpu for which cpufreq_add_dev() is called. Eventually get_cpu_device() will return NULL for them as well, and we will hit the same WARN_ON() again. In order to fix these issues, change cpufreq core to create links to the policy for a cpu only when cpufreq_add_dev() is called for that CPU. Reuse the 'real_cpus' mask to track that as well. Note that cpufreq_remove_dev() already handles removal of the links for individual CPUs and cpufreq_add_dev() has aligned with that now. Reported-by: Russell King Tested-by: Russell King Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 13fb589..3a64136 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -916,58 +916,18 @@ static struct kobj_type ktype_cpufreq = { .release = cpufreq_sysfs_release, }; -static int add_cpu_dev_symlink(struct cpufreq_policy *policy, int cpu) +static int add_cpu_dev_symlink(struct cpufreq_policy *policy, + struct device *dev) { - struct device *cpu_dev; - - pr_debug("%s: Adding symlink for CPU: %u\n", __func__, cpu); - - if (!policy) - return 0; - - cpu_dev = get_cpu_device(cpu); - if (WARN_ON(!cpu_dev)) - return 0; - - return sysfs_create_link(&cpu_dev->kobj, &policy->kobj, "cpufreq"); + dev_dbg(dev, "%s: Adding symlink\n", __func__); + return sysfs_create_link(&dev->kobj, &policy->kobj, "cpufreq"); } -static void remove_cpu_dev_symlink(struct cpufreq_policy *policy, int cpu) +static void remove_cpu_dev_symlink(struct cpufreq_policy *policy, + struct device *dev) { - struct device *cpu_dev; - - pr_debug("%s: Removing symlink for CPU: %u\n", __func__, cpu); - - cpu_dev = get_cpu_device(cpu); - if (WARN_ON(!cpu_dev)) - return; - - sysfs_remove_link(&cpu_dev->kobj, "cpufreq"); -} - -/* Add/remove symlinks for all related CPUs */ -static int cpufreq_add_dev_symlink(struct cpufreq_policy *policy) -{ - unsigned int j; - int ret = 0; - - /* Some related CPUs might not be present (physically hotplugged) */ - for_each_cpu(j, policy->real_cpus) { - ret = add_cpu_dev_symlink(policy, j); - if (ret) - break; - } - - return ret; -} - -static void cpufreq_remove_dev_symlink(struct cpufreq_policy *policy) -{ - unsigned int j; - - /* Some related CPUs might not be present (physically hotplugged) */ - for_each_cpu(j, policy->real_cpus) - remove_cpu_dev_symlink(policy, j); + dev_dbg(dev, "%s: Removing symlink\n", __func__); + sysfs_remove_link(&dev->kobj, "cpufreq"); } static int cpufreq_add_dev_interface(struct cpufreq_policy *policy) @@ -999,7 +959,7 @@ static int cpufreq_add_dev_interface(struct cpufreq_policy *policy) return ret; } - return cpufreq_add_dev_symlink(policy); + return 0; } __weak struct cpufreq_governor *cpufreq_default_governor(void) @@ -1129,7 +1089,6 @@ static void cpufreq_policy_put_kobj(struct cpufreq_policy *policy, bool notify) down_write(&policy->rwsem); cpufreq_stats_free_table(policy); - cpufreq_remove_dev_symlink(policy); kobj = &policy->kobj; cmp = &policy->kobj_unregister; up_write(&policy->rwsem); @@ -1211,8 +1170,8 @@ static int cpufreq_online(unsigned int cpu) if (new_policy) { /* related_cpus should at least include policy->cpus. */ cpumask_copy(policy->related_cpus, policy->cpus); - /* Remember CPUs present at the policy creation time. */ - cpumask_and(policy->real_cpus, policy->cpus, cpu_present_mask); + /* Clear mask of registered CPUs */ + cpumask_clear(policy->real_cpus); } /* @@ -1327,6 +1286,8 @@ out_free_policy: return ret; } +static void cpufreq_offline(unsigned int cpu); + /** * cpufreq_add_dev - the cpufreq interface for a CPU device. * @dev: CPU device. @@ -1336,22 +1297,28 @@ static int cpufreq_add_dev(struct device *dev, struct subsys_interface *sif) { struct cpufreq_policy *policy; unsigned cpu = dev->id; + int ret; dev_dbg(dev, "%s: adding CPU%u\n", __func__, cpu); - if (cpu_online(cpu)) - return cpufreq_online(cpu); + if (cpu_online(cpu)) { + ret = cpufreq_online(cpu); + if (ret) + return ret; + } - /* - * A hotplug notifier will follow and we will handle it as CPU online - * then. For now, just create the sysfs link, unless there is no policy - * or the link is already present. - */ + /* Create sysfs link on CPU registration */ policy = per_cpu(cpufreq_cpu_data, cpu); if (!policy || cpumask_test_and_set_cpu(cpu, policy->real_cpus)) return 0; - return add_cpu_dev_symlink(policy, cpu); + ret = add_cpu_dev_symlink(policy, dev); + if (ret) { + cpumask_clear_cpu(cpu, policy->real_cpus); + cpufreq_offline(cpu); + } + + return ret; } static void cpufreq_offline(unsigned int cpu) @@ -1432,7 +1399,7 @@ static void cpufreq_remove_dev(struct device *dev, struct subsys_interface *sif) cpufreq_offline(cpu); cpumask_clear_cpu(cpu, policy->real_cpus); - remove_cpu_dev_symlink(policy, cpu); + remove_cpu_dev_symlink(policy, dev); if (cpumask_empty(policy->real_cpus)) cpufreq_policy_free(policy, true); -- cgit v0.10.2 From e2cdeaa1232d41b2b8d2da67490f609ac0b7b0dc Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Fri, 19 Aug 2016 09:15:50 -0700 Subject: PM / AVS: SmartReflex: Neaten logging Use a more common logging style. Miscellanea: o Use pr_warn o Coalesce formats adding missing spaces o Argument alignment o Remove unnecessary OOM messages as k.alloc does stack dumps Signed-off-by: Joe Perches Acked-by: Nishanth Menon [khilman: update shortlog] Signed-off-by: Kevin Hilman Signed-off-by: Rafael J. Wysocki diff --git a/drivers/power/avs/smartreflex.c b/drivers/power/avs/smartreflex.c index db9973b..fa0f19b 100644 --- a/drivers/power/avs/smartreflex.c +++ b/drivers/power/avs/smartreflex.c @@ -136,7 +136,7 @@ static void sr_set_clk_length(struct omap_sr *sr) if (IS_ERR(fck)) { dev_err(&sr->pdev->dev, "%s: unable to get fck for device %s\n", - __func__, dev_name(&sr->pdev->dev)); + __func__, dev_name(&sr->pdev->dev)); return; } @@ -170,8 +170,8 @@ static void sr_start_vddautocomp(struct omap_sr *sr) { if (!sr_class || !(sr_class->enable) || !(sr_class->configure)) { dev_warn(&sr->pdev->dev, - "%s: smartreflex class driver not registered\n", - __func__); + "%s: smartreflex class driver not registered\n", + __func__); return; } @@ -183,8 +183,8 @@ static void sr_stop_vddautocomp(struct omap_sr *sr) { if (!sr_class || !(sr_class->disable)) { dev_warn(&sr->pdev->dev, - "%s: smartreflex class driver not registered\n", - __func__); + "%s: smartreflex class driver not registered\n", + __func__); return; } @@ -225,9 +225,8 @@ static int sr_late_init(struct omap_sr *sr_info) error: list_del(&sr_info->node); - dev_err(&sr_info->pdev->dev, "%s: ERROR in registering" - "interrupt handler. Smartreflex will" - "not function as desired\n", __func__); + dev_err(&sr_info->pdev->dev, "%s: ERROR in registering interrupt handler. Smartreflex will not function as desired\n", + __func__); return ret; } @@ -263,7 +262,7 @@ static void sr_v1_disable(struct omap_sr *sr) if (timeout >= SR_DISABLE_TIMEOUT) dev_warn(&sr->pdev->dev, "%s: Smartreflex disable timedout\n", - __func__); + __func__); /* Disable MCUDisableAcknowledge interrupt & clear pending interrupt */ sr_modify_reg(sr, ERRCONFIG_V1, ERRCONFIG_MCUDISACKINTEN, @@ -308,7 +307,7 @@ static void sr_v2_disable(struct omap_sr *sr) if (timeout >= SR_DISABLE_TIMEOUT) dev_warn(&sr->pdev->dev, "%s: Smartreflex disable timedout\n", - __func__); + __func__); /* Disable MCUDisableAcknowledge interrupt & clear pending interrupt */ sr_write_reg(sr, IRQENABLE_CLR, IRQENABLE_MCUDISABLEACKINT); @@ -322,7 +321,7 @@ static struct omap_sr_nvalue_table *sr_retrieve_nvalue_row( if (!sr->nvalue_table) { dev_warn(&sr->pdev->dev, "%s: Missing ntarget value table\n", - __func__); + __func__); return NULL; } @@ -356,8 +355,8 @@ int sr_configure_errgen(struct omap_sr *sr) u8 senp_shift, senn_shift; if (!sr) { - pr_warn("%s: NULL omap_sr from %pF\n", __func__, - (void *)_RET_IP_); + pr_warn("%s: NULL omap_sr from %pF\n", + __func__, (void *)_RET_IP_); return -EINVAL; } @@ -387,8 +386,8 @@ int sr_configure_errgen(struct omap_sr *sr) vpboundint_st = ERRCONFIG_VPBOUNDINTST_V2; break; default: - dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex" - "module without specifying the ip\n", __func__); + dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex module without specifying the ip\n", + __func__); return -EINVAL; } @@ -423,8 +422,8 @@ int sr_disable_errgen(struct omap_sr *sr) u32 vpboundint_en, vpboundint_st; if (!sr) { - pr_warn("%s: NULL omap_sr from %pF\n", __func__, - (void *)_RET_IP_); + pr_warn("%s: NULL omap_sr from %pF\n", + __func__, (void *)_RET_IP_); return -EINVAL; } @@ -440,8 +439,8 @@ int sr_disable_errgen(struct omap_sr *sr) vpboundint_st = ERRCONFIG_VPBOUNDINTST_V2; break; default: - dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex" - "module without specifying the ip\n", __func__); + dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex module without specifying the ip\n", + __func__); return -EINVAL; } @@ -478,8 +477,8 @@ int sr_configure_minmax(struct omap_sr *sr) u8 senp_shift, senn_shift; if (!sr) { - pr_warn("%s: NULL omap_sr from %pF\n", __func__, - (void *)_RET_IP_); + pr_warn("%s: NULL omap_sr from %pF\n", + __func__, (void *)_RET_IP_); return -EINVAL; } @@ -504,8 +503,8 @@ int sr_configure_minmax(struct omap_sr *sr) senp_shift = SRCONFIG_SENPENABLE_V2_SHIFT; break; default: - dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex" - "module without specifying the ip\n", __func__); + dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex module without specifying the ip\n", + __func__); return -EINVAL; } @@ -537,8 +536,8 @@ int sr_configure_minmax(struct omap_sr *sr) IRQENABLE_MCUBOUNDSINT | IRQENABLE_MCUDISABLEACKINT); break; default: - dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex" - "module without specifying the ip\n", __func__); + dev_err(&sr->pdev->dev, "%s: Trying to Configure smartreflex module without specifying the ip\n", + __func__); return -EINVAL; } @@ -563,16 +562,16 @@ int sr_enable(struct omap_sr *sr, unsigned long volt) int ret; if (!sr) { - pr_warn("%s: NULL omap_sr from %pF\n", __func__, - (void *)_RET_IP_); + pr_warn("%s: NULL omap_sr from %pF\n", + __func__, (void *)_RET_IP_); return -EINVAL; } volt_data = omap_voltage_get_voltdata(sr->voltdm, volt); if (IS_ERR(volt_data)) { - dev_warn(&sr->pdev->dev, "%s: Unable to get voltage table" - "for nominal voltage %ld\n", __func__, volt); + dev_warn(&sr->pdev->dev, "%s: Unable to get voltage table for nominal voltage %ld\n", + __func__, volt); return PTR_ERR(volt_data); } @@ -615,8 +614,8 @@ int sr_enable(struct omap_sr *sr, unsigned long volt) void sr_disable(struct omap_sr *sr) { if (!sr) { - pr_warn("%s: NULL omap_sr from %pF\n", __func__, - (void *)_RET_IP_); + pr_warn("%s: NULL omap_sr from %pF\n", + __func__, (void *)_RET_IP_); return; } @@ -658,13 +657,13 @@ int sr_register_class(struct omap_sr_class_data *class_data) struct omap_sr *sr_info; if (!class_data) { - pr_warning("%s:, Smartreflex class data passed is NULL\n", + pr_warn("%s:, Smartreflex class data passed is NULL\n", __func__); return -EINVAL; } if (sr_class) { - pr_warning("%s: Smartreflex class driver already registered\n", + pr_warn("%s: Smartreflex class driver already registered\n", __func__); return -EBUSY; } @@ -696,7 +695,7 @@ void omap_sr_enable(struct voltagedomain *voltdm) struct omap_sr *sr = _sr_lookup(voltdm); if (IS_ERR(sr)) { - pr_warning("%s: omap_sr struct for voltdm not found\n", __func__); + pr_warn("%s: omap_sr struct for voltdm not found\n", __func__); return; } @@ -704,8 +703,8 @@ void omap_sr_enable(struct voltagedomain *voltdm) return; if (!sr_class || !(sr_class->enable) || !(sr_class->configure)) { - dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not" - "registered\n", __func__); + dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not registered\n", + __func__); return; } @@ -728,7 +727,7 @@ void omap_sr_disable(struct voltagedomain *voltdm) struct omap_sr *sr = _sr_lookup(voltdm); if (IS_ERR(sr)) { - pr_warning("%s: omap_sr struct for voltdm not found\n", __func__); + pr_warn("%s: omap_sr struct for voltdm not found\n", __func__); return; } @@ -736,8 +735,8 @@ void omap_sr_disable(struct voltagedomain *voltdm) return; if (!sr_class || !(sr_class->disable)) { - dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not" - "registered\n", __func__); + dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not registered\n", + __func__); return; } @@ -760,7 +759,7 @@ void omap_sr_disable_reset_volt(struct voltagedomain *voltdm) struct omap_sr *sr = _sr_lookup(voltdm); if (IS_ERR(sr)) { - pr_warning("%s: omap_sr struct for voltdm not found\n", __func__); + pr_warn("%s: omap_sr struct for voltdm not found\n", __func__); return; } @@ -768,8 +767,8 @@ void omap_sr_disable_reset_volt(struct voltagedomain *voltdm) return; if (!sr_class || !(sr_class->disable)) { - dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not" - "registered\n", __func__); + dev_warn(&sr->pdev->dev, "%s: smartreflex class driver not registered\n", + __func__); return; } @@ -787,8 +786,8 @@ void omap_sr_disable_reset_volt(struct voltagedomain *voltdm) void omap_sr_register_pmic(struct omap_sr_pmic_data *pmic_data) { if (!pmic_data) { - pr_warning("%s: Trying to register NULL PMIC data structure" - "with smartreflex\n", __func__); + pr_warn("%s: Trying to register NULL PMIC data structure with smartreflex\n", + __func__); return; } @@ -801,7 +800,7 @@ static int omap_sr_autocomp_show(void *data, u64 *val) struct omap_sr *sr_info = data; if (!sr_info) { - pr_warning("%s: omap_sr struct not found\n", __func__); + pr_warn("%s: omap_sr struct not found\n", __func__); return -EINVAL; } @@ -815,13 +814,13 @@ static int omap_sr_autocomp_store(void *data, u64 val) struct omap_sr *sr_info = data; if (!sr_info) { - pr_warning("%s: omap_sr struct not found\n", __func__); + pr_warn("%s: omap_sr struct not found\n", __func__); return -EINVAL; } /* Sanity check */ if (val > 1) { - pr_warning("%s: Invalid argument %lld\n", __func__, val); + pr_warn("%s: Invalid argument %lld\n", __func__, val); return -EINVAL; } @@ -848,19 +847,13 @@ static int __init omap_sr_probe(struct platform_device *pdev) int i, ret = 0; sr_info = devm_kzalloc(&pdev->dev, sizeof(struct omap_sr), GFP_KERNEL); - if (!sr_info) { - dev_err(&pdev->dev, "%s: unable to allocate sr_info\n", - __func__); + if (!sr_info) return -ENOMEM; - } sr_info->name = devm_kzalloc(&pdev->dev, SMARTREFLEX_NAME_LEN, GFP_KERNEL); - if (!sr_info->name) { - dev_err(&pdev->dev, "%s: unable to allocate SR instance name\n", - __func__); + if (!sr_info->name) return -ENOMEM; - } platform_set_drvdata(pdev, sr_info); @@ -912,7 +905,7 @@ static int __init omap_sr_probe(struct platform_device *pdev) if (sr_class) { ret = sr_late_init(sr_info); if (ret) { - pr_warning("%s: Error in SR late init\n", __func__); + pr_warn("%s: Error in SR late init\n", __func__); goto err_list_del; } } @@ -923,7 +916,7 @@ static int __init omap_sr_probe(struct platform_device *pdev) if (IS_ERR_OR_NULL(sr_dbg_dir)) { ret = PTR_ERR(sr_dbg_dir); pr_err("%s:sr debugfs dir creation failed(%d)\n", - __func__, ret); + __func__, ret); goto err_list_del; } } @@ -945,8 +938,8 @@ static int __init omap_sr_probe(struct platform_device *pdev) nvalue_dir = debugfs_create_dir("nvalue", sr_info->dbg_dir); if (IS_ERR_OR_NULL(nvalue_dir)) { - dev_err(&pdev->dev, "%s: Unable to create debugfs directory" - "for n-values\n", __func__); + dev_err(&pdev->dev, "%s: Unable to create debugfs directory for n-values\n", + __func__); ret = PTR_ERR(nvalue_dir); goto err_debugfs; } @@ -1053,12 +1046,12 @@ static int __init sr_init(void) if (sr_pmic_data && sr_pmic_data->sr_pmic_init) sr_pmic_data->sr_pmic_init(); else - pr_warning("%s: No PMIC hook to init smartreflex\n", __func__); + pr_warn("%s: No PMIC hook to init smartreflex\n", __func__); ret = platform_driver_probe(&smartreflex_driver, omap_sr_probe); if (ret) { pr_err("%s: platform driver register failed for SR\n", - __func__); + __func__); return ret; } -- cgit v0.10.2 From ad38677df44b67e0f5b6c4d31e9c2734abde8ed9 Mon Sep 17 00:00:00 2001 From: Al Stone Date: Wed, 20 Jul 2016 15:10:04 -0600 Subject: cpufreq: CPPC: Force reporting values in KHz to fix user space interface When CPPC is being used by ACPI on arm64, user space tools such as cpupower report CPU frequency values from sysfs that are incorrect. What the driver was doing was reporting the values given by ACPI tables in whatever scale was used to provide them. However, the ACPI spec defines the CPPC values as unitless abstract numbers. Internal kernel structures such as struct perf_cap, in contrast, expect these values to be in KHz. When these struct values get reported via sysfs, the user space tools also assume they are in KHz, causing them to report incorrect values (for example, reporting a CPU frequency of 1MHz when it should be 1.8GHz). The downside is that this approach has some assumptions: (1) It relies on SMBIOS3 being used, *and* that the Max Frequency value for a processor is set to a non-zero value. (2) It assumes that all processors run at the same speed, or that the CPPC values have all been scaled to reflect relative speed. This patch retrieves the largest CPU Max Frequency from a type 4 DMI record that it can find. This may not be an issue, however, as a sampling of DMI data on x86 and arm64 indicates there is often only one such record regardless. Since CPPC is relatively new, it is unclear if the ACPI ASL will always be written to reflect any sort of relative performance of processors of differing speeds. (3) It assumes that performance and frequency both scale linearly. For arm64 servers, this may be sufficient, but it does rely on firmware values being set correctly. Hence, other approaches will be considered in the future. This has been tested on three arm64 servers, with and without DMI, with and without CPPC support. Signed-off-by: Al Stone Signed-off-by: Prashanth Prakash Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index 8882b8e..6debc18 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -19,10 +19,19 @@ #include #include #include +#include #include +#include + #include +/* Minimum struct length needed for the DMI processor entry we want */ +#define DMI_ENTRY_PROCESSOR_MIN_LENGTH 48 + +/* Offest in the DMI processor structure for the max frequency */ +#define DMI_PROCESSOR_MAX_SPEED 0x14 + /* * These structs contain information parsed from per CPU * ACPI _CPC structures. @@ -32,6 +41,39 @@ */ static struct cpudata **all_cpu_data; +/* Capture the max KHz from DMI */ +static u64 cppc_dmi_max_khz; + +/* Callback function used to retrieve the max frequency from DMI */ +static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) +{ + const u8 *dmi_data = (const u8 *)dm; + u16 *mhz = (u16 *)private; + + if (dm->type == DMI_ENTRY_PROCESSOR && + dm->length >= DMI_ENTRY_PROCESSOR_MIN_LENGTH) { + u16 val = (u16)get_unaligned((const u16 *) + (dmi_data + DMI_PROCESSOR_MAX_SPEED)); + *mhz = val > *mhz ? val : *mhz; + } +} + +/* Look up the max frequency in DMI */ +static u64 cppc_get_dmi_max_khz(void) +{ + u16 mhz = 0; + + dmi_walk(cppc_find_dmi_mhz, &mhz); + + /* + * Real stupid fallback value, just in case there is no + * actual value set. + */ + mhz = mhz ? mhz : 1; + + return (1000 * mhz); +} + static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, unsigned int target_freq, unsigned int relation) @@ -42,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, cpu = all_cpu_data[policy->cpu]; - cpu->perf_ctrls.desired_perf = target_freq; + cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; freqs.old = policy->cur; freqs.new = target_freq; @@ -94,8 +136,10 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) return ret; } - policy->min = cpu->perf_caps.lowest_perf; - policy->max = cpu->perf_caps.highest_perf; + cppc_dmi_max_khz = cppc_get_dmi_max_khz(); + + policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu->perf_caps.highest_perf; + policy->max = cppc_dmi_max_khz; policy->cpuinfo.min_freq = policy->min; policy->cpuinfo.max_freq = policy->max; policy->shared_type = cpu->shared_type; @@ -112,7 +156,8 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) cpu->cur_policy = policy; /* Set policy->cur to max now. The governors will adjust later. */ - policy->cur = cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; + policy->cur = cppc_dmi_max_khz; + cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; ret = cppc_set_perf(cpu_num, &cpu->perf_ctrls); if (ret) -- cgit v0.10.2 From ec69572b3f0daf1f26b3e563d56e487d515a3cd1 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:05 +0100 Subject: PM / Domains: Add new helper functions for device-tree Ideally, if we are returning a reference to a PM domain via a call to of_genpd_get_from_provider(), then we should keep track of such references via a reference count. The reference count could then be used to determine if a PM domain can be safely removed. Alternatively, it is possible to avoid such external references by providing APIs to access the PM domain and hence, eliminate any calls to of_genpd_get_from_provider(). Add new helper functions for adding a device and a subdomain to a PM domain when using device-tree, so that external calls to of_genpd_get_from_provider() can be removed. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 058c8b6..e12d3a8 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1502,6 +1502,52 @@ struct generic_pm_domain *of_genpd_get_from_provider( EXPORT_SYMBOL_GPL(of_genpd_get_from_provider); /** + * of_genpd_add_device() - Add a device to an I/O PM domain + * @genpdspec: OF phandle args to use for look-up PM domain + * @dev: Device to be added. + * + * Looks-up an I/O PM domain based upon phandle args provided and adds + * the device to the PM domain. Returns a negative error code on failure. + */ +int of_genpd_add_device(struct of_phandle_args *genpdspec, struct device *dev) +{ + struct generic_pm_domain *genpd; + + genpd = of_genpd_get_from_provider(genpdspec); + if (IS_ERR(genpd)) + return PTR_ERR(genpd); + + return pm_genpd_add_device(genpd, dev); +} +EXPORT_SYMBOL_GPL(of_genpd_add_device); + +/** + * of_genpd_add_subdomain - Add a subdomain to an I/O PM domain. + * @parent_spec: OF phandle args to use for parent PM domain look-up + * @subdomain_spec: OF phandle args to use for subdomain look-up + * + * Looks-up a parent PM domain and subdomain based upon phandle args + * provided and adds the subdomain to the parent PM domain. Returns a + * negative error code on failure. + */ +int of_genpd_add_subdomain(struct of_phandle_args *parent_spec, + struct of_phandle_args *subdomain_spec) +{ + struct generic_pm_domain *parent, *subdomain; + + parent = of_genpd_get_from_provider(parent_spec); + if (IS_ERR(parent)) + return PTR_ERR(parent); + + subdomain = of_genpd_get_from_provider(subdomain_spec); + if (IS_ERR(subdomain)) + return PTR_ERR(subdomain); + + return pm_genpd_add_subdomain(parent, subdomain); +} +EXPORT_SYMBOL_GPL(of_genpd_add_subdomain); + +/** * genpd_dev_pm_detach - Detach a device from its PM domain. * @dev: Device to detach. * @power_off: Currently not used diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index 31fec85..e1964a2 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -208,6 +208,10 @@ struct generic_pm_domain *__of_genpd_xlate_simple( struct generic_pm_domain *__of_genpd_xlate_onecell( struct of_phandle_args *genpdspec, void *data); +extern int of_genpd_add_device(struct of_phandle_args *args, + struct device *dev); +extern int of_genpd_add_subdomain(struct of_phandle_args *parent, + struct of_phandle_args *new_subdomain); int genpd_dev_pm_attach(struct device *dev); #else /* !CONFIG_PM_GENERIC_DOMAINS_OF */ @@ -227,6 +231,18 @@ static inline struct generic_pm_domain *of_genpd_get_from_provider( #define __of_genpd_xlate_simple NULL #define __of_genpd_xlate_onecell NULL +static inline int of_genpd_add_device(struct of_phandle_args *args, + struct device *dev) +{ + return -ENODEV; +} + +static inline int of_genpd_add_subdomain(struct of_phandle_args *parent, + struct of_phandle_args *new_subdomain) +{ + return -ENODEV; +} + static inline int genpd_dev_pm_attach(struct device *dev) { return -ENODEV; -- cgit v0.10.2 From 8d68c3713aa6dcf4f65f7fa8f0b6f692ace18d07 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:06 +0100 Subject: ARM: EXYNOS: Remove calls to of_genpd_get_from_provider() Update the EXYNOS PM domain code to use the of_genpd_add_subdomain() and remove any calls to of_genpd_get_from_provider(). Signed-off-by: Jon Hunter Reviewed-by: Krzysztof Kozlowski Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/soc/samsung/pm_domains.c b/drivers/soc/samsung/pm_domains.c index 4822346..7112004 100644 --- a/drivers/soc/samsung/pm_domains.c +++ b/drivers/soc/samsung/pm_domains.c @@ -215,29 +215,22 @@ no_clk: /* Assign the child power domains to their parents */ for_each_matching_node(np, exynos_pm_domain_of_match) { - struct generic_pm_domain *child_domain, *parent_domain; - struct of_phandle_args args; + struct of_phandle_args child, parent; - args.np = np; - args.args_count = 0; - child_domain = of_genpd_get_from_provider(&args); - if (IS_ERR(child_domain)) - continue; + child.np = np; + child.args_count = 0; if (of_parse_phandle_with_args(np, "power-domains", - "#power-domain-cells", 0, &args) != 0) - continue; - - parent_domain = of_genpd_get_from_provider(&args); - if (IS_ERR(parent_domain)) + "#power-domain-cells", 0, + &parent) != 0) continue; - if (pm_genpd_add_subdomain(parent_domain, child_domain)) + if (of_genpd_add_subdomain(&parent, &child)) pr_warn("%s failed to add subdomain: %s\n", - parent_domain->name, child_domain->name); + parent.np->name, child.np->name); else pr_info("%s has as child subdomain: %s.\n", - parent_domain->name, child_domain->name); + parent.np->name, child.np->name); } return 0; -- cgit v0.10.2 From 3468c9d9ef5cc0f5375b7bc0b2ab360b559fc1a9 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:07 +0100 Subject: staging: board: Remove calls to of_genpd_get_from_provider() Update the staging/board PM domain code to use the of_genpd_add_subdomain() and remove any calls to of_genpd_get_from_provider(). Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/staging/board/board.c b/drivers/staging/board/board.c index 45807d8..86dc411 100644 --- a/drivers/staging/board/board.c +++ b/drivers/staging/board/board.c @@ -140,7 +140,6 @@ static int board_staging_add_dev_domain(struct platform_device *pdev, const char *domain) { struct of_phandle_args pd_args; - struct generic_pm_domain *pd; struct device_node *np; np = of_find_node_by_path(domain); @@ -151,14 +150,8 @@ static int board_staging_add_dev_domain(struct platform_device *pdev, pd_args.np = np; pd_args.args_count = 0; - pd = of_genpd_get_from_provider(&pd_args); - if (IS_ERR(pd)) { - pr_err("Cannot find genpd %s (%ld)\n", domain, PTR_ERR(pd)); - return PTR_ERR(pd); - } - pr_debug("Found genpd %s for device %s\n", pd->name, pdev->name); - return pm_genpd_add_device(pd, &pdev->dev); + return of_genpd_add_device(&pd_args, &pdev->dev); } #else static inline int board_staging_add_dev_domain(struct platform_device *pdev, -- cgit v0.10.2 From f58d4e5ab0ca3453f091eab514474e9fdbfc539f Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:08 +0100 Subject: PM / Domains: Don't expose generic_pm_domain structure to clients There should be no need to expose the generic_pm_domain structure to clients and this eliminates the need to implement reference counting for any external reference to a PM domain. Therefore, make the functions pm_genpd_lookup_dev() and of_genpd_get_from_provider() private to the PM domain core. The functions are renamed in accordance with the naming conventions for genpd static functions. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index e12d3a8..d8c1eab 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -45,7 +45,7 @@ static DEFINE_MUTEX(gpd_list_lock); * and checks that the PM domain pointer is a real generic PM domain. * Any failure results in NULL being returned. */ -struct generic_pm_domain *pm_genpd_lookup_dev(struct device *dev) +static struct generic_pm_domain *genpd_lookup_dev(struct device *dev) { struct generic_pm_domain *genpd = NULL, *gpd; @@ -1119,7 +1119,7 @@ int pm_genpd_remove_device(struct generic_pm_domain *genpd, dev_dbg(dev, "%s()\n", __func__); - if (!genpd || genpd != pm_genpd_lookup_dev(dev)) + if (!genpd || genpd != genpd_lookup_dev(dev)) return -EINVAL; /* The above validation also means we have existing domain_data. */ @@ -1466,7 +1466,7 @@ void of_genpd_del_provider(struct device_node *np) EXPORT_SYMBOL_GPL(of_genpd_del_provider); /** - * of_genpd_get_from_provider() - Look-up PM domain + * genpd_get_from_provider() - Look-up PM domain * @genpdspec: OF phandle args to use for look-up * * Looks for a PM domain provider under the node specified by @genpdspec and if @@ -1476,7 +1476,7 @@ EXPORT_SYMBOL_GPL(of_genpd_del_provider); * Returns a valid pointer to struct generic_pm_domain on success or ERR_PTR() * on failure. */ -struct generic_pm_domain *of_genpd_get_from_provider( +static struct generic_pm_domain *genpd_get_from_provider( struct of_phandle_args *genpdspec) { struct generic_pm_domain *genpd = ERR_PTR(-ENOENT); @@ -1499,7 +1499,6 @@ struct generic_pm_domain *of_genpd_get_from_provider( return genpd; } -EXPORT_SYMBOL_GPL(of_genpd_get_from_provider); /** * of_genpd_add_device() - Add a device to an I/O PM domain @@ -1513,7 +1512,7 @@ int of_genpd_add_device(struct of_phandle_args *genpdspec, struct device *dev) { struct generic_pm_domain *genpd; - genpd = of_genpd_get_from_provider(genpdspec); + genpd = genpd_get_from_provider(genpdspec); if (IS_ERR(genpd)) return PTR_ERR(genpd); @@ -1535,11 +1534,11 @@ int of_genpd_add_subdomain(struct of_phandle_args *parent_spec, { struct generic_pm_domain *parent, *subdomain; - parent = of_genpd_get_from_provider(parent_spec); + parent = genpd_get_from_provider(parent_spec); if (IS_ERR(parent)) return PTR_ERR(parent); - subdomain = of_genpd_get_from_provider(subdomain_spec); + subdomain = genpd_get_from_provider(subdomain_spec); if (IS_ERR(subdomain)) return PTR_ERR(subdomain); @@ -1561,7 +1560,7 @@ static void genpd_dev_pm_detach(struct device *dev, bool power_off) unsigned int i; int ret = 0; - pd = pm_genpd_lookup_dev(dev); + pd = genpd_lookup_dev(dev); if (!pd) return; @@ -1642,7 +1641,7 @@ int genpd_dev_pm_attach(struct device *dev) return -ENOENT; } - pd = of_genpd_get_from_provider(&pd_args); + pd = genpd_get_from_provider(&pd_args); of_node_put(pd_args.np); if (IS_ERR(pd)) { dev_dbg(dev, "%s() failed to find PM domain: %ld\n", diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index e1964a2..bd411e7 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -116,7 +116,6 @@ static inline struct generic_pm_domain_data *dev_gpd_data(struct device *dev) return to_gpd_data(dev->power.subsys_data->domain_data); } -extern struct generic_pm_domain *pm_genpd_lookup_dev(struct device *dev); extern int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, struct gpd_timing_data *td); @@ -138,10 +137,6 @@ static inline struct generic_pm_domain_data *dev_gpd_data(struct device *dev) { return ERR_PTR(-ENOSYS); } -static inline struct generic_pm_domain *pm_genpd_lookup_dev(struct device *dev) -{ - return NULL; -} static inline int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, struct gpd_timing_data *td) @@ -199,9 +194,6 @@ typedef struct generic_pm_domain *(*genpd_xlate_t)(struct of_phandle_args *args, int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, void *data); void of_genpd_del_provider(struct device_node *np); -struct generic_pm_domain *of_genpd_get_from_provider( - struct of_phandle_args *genpdspec); - struct generic_pm_domain *__of_genpd_xlate_simple( struct of_phandle_args *genpdspec, void *data); @@ -222,12 +214,6 @@ static inline int __of_genpd_add_provider(struct device_node *np, } static inline void of_genpd_del_provider(struct device_node *np) {} -static inline struct generic_pm_domain *of_genpd_get_from_provider( - struct of_phandle_args *genpdspec) -{ - return NULL; -} - #define __of_genpd_xlate_simple NULL #define __of_genpd_xlate_onecell NULL -- cgit v0.10.2 From 892ebdcccb4e11c5b43036a50e89e3869df5b429 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:09 +0100 Subject: PM / Domains: Don't expose xlate and provider helper functions Functions __of_genpd_xlate_simple(), __of_genpd_xlate_onecell() and __of_genpd_add_provider() are not used outside of the core generic PM domain code. Therefore, reduce the number of APIs exposed by making these static. At the same time don't expose the typedef for genpd_xlate_t either and make this a local definition as well. The functions are renamed to follow the naming conventions for static functions in the generic PM domain core. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index d8c1eab..c93aa3a 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1329,6 +1329,10 @@ int pm_genpd_init(struct generic_pm_domain *genpd, EXPORT_SYMBOL_GPL(pm_genpd_init); #ifdef CONFIG_PM_GENERIC_DOMAINS_OF + +typedef struct generic_pm_domain *(*genpd_xlate_t)(struct of_phandle_args *args, + void *data); + /* * Device Tree based PM domain providers. * @@ -1340,8 +1344,8 @@ EXPORT_SYMBOL_GPL(pm_genpd_init); * maps a PM domain specifier retrieved from the device tree to a PM domain. * * Two simple mapping functions have been provided for convenience: - * - __of_genpd_xlate_simple() for 1:1 device tree node to PM domain mapping. - * - __of_genpd_xlate_onecell() for mapping of multiple PM domains per node by + * - genpd_xlate_simple() for 1:1 device tree node to PM domain mapping. + * - genpd_xlate_onecell() for mapping of multiple PM domains per node by * index. */ @@ -1366,7 +1370,7 @@ static LIST_HEAD(of_genpd_providers); static DEFINE_MUTEX(of_genpd_mutex); /** - * __of_genpd_xlate_simple() - Xlate function for direct node-domain mapping + * genpd_xlate_simple() - Xlate function for direct node-domain mapping * @genpdspec: OF phandle args to map into a PM domain * @data: xlate function private data - pointer to struct generic_pm_domain * @@ -1374,7 +1378,7 @@ static DEFINE_MUTEX(of_genpd_mutex); * have their own device tree nodes. The private data of xlate function needs * to be a valid pointer to struct generic_pm_domain. */ -struct generic_pm_domain *__of_genpd_xlate_simple( +static struct generic_pm_domain *genpd_xlate_simple( struct of_phandle_args *genpdspec, void *data) { @@ -1382,10 +1386,9 @@ struct generic_pm_domain *__of_genpd_xlate_simple( return ERR_PTR(-EINVAL); return data; } -EXPORT_SYMBOL_GPL(__of_genpd_xlate_simple); /** - * __of_genpd_xlate_onecell() - Xlate function using a single index. + * genpd_xlate_onecell() - Xlate function using a single index. * @genpdspec: OF phandle args to map into a PM domain * @data: xlate function private data - pointer to struct genpd_onecell_data * @@ -1394,7 +1397,7 @@ EXPORT_SYMBOL_GPL(__of_genpd_xlate_simple); * A single cell is used as an index into an array of PM domains specified in * the genpd_onecell_data struct when registering the provider. */ -struct generic_pm_domain *__of_genpd_xlate_onecell( +static struct generic_pm_domain *genpd_xlate_onecell( struct of_phandle_args *genpdspec, void *data) { @@ -1414,16 +1417,15 @@ struct generic_pm_domain *__of_genpd_xlate_onecell( return genpd_data->domains[idx]; } -EXPORT_SYMBOL_GPL(__of_genpd_xlate_onecell); /** - * __of_genpd_add_provider() - Register a PM domain provider for a node + * genpd_add_provider() - Register a PM domain provider for a node * @np: Device node pointer associated with the PM domain provider. * @xlate: Callback for decoding PM domain from phandle arguments. * @data: Context pointer for @xlate callback. */ -int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, - void *data) +static int genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, + void *data) { struct of_genpd_provider *cp; @@ -1442,7 +1444,30 @@ int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, return 0; } -EXPORT_SYMBOL_GPL(__of_genpd_add_provider); + +/** + * of_genpd_add_provider_simple() - Register a simple PM domain provider + * @np: Device node pointer associated with the PM domain provider. + * @genpd: Pointer to PM domain associated with the PM domain provider. + */ +int of_genpd_add_provider_simple(struct device_node *np, + struct generic_pm_domain *genpd) +{ + return genpd_add_provider(np, genpd_xlate_simple, genpd); +} +EXPORT_SYMBOL_GPL(of_genpd_add_provider_simple); + +/** + * of_genpd_add_provider_onecell() - Register a onecell PM domain provider + * @np: Device node pointer associated with the PM domain provider. + * @data: Pointer to the data associated with the PM domain provider. + */ +int of_genpd_add_provider_onecell(struct device_node *np, + struct genpd_onecell_data *data) +{ + return genpd_add_provider(np, genpd_xlate_onecell, data); +} +EXPORT_SYMBOL_GPL(of_genpd_add_provider_onecell); /** * of_genpd_del_provider() - Remove a previously registered PM domain provider diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index bd411e7..f103869 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -187,19 +187,12 @@ struct genpd_onecell_data { unsigned int num_domains; }; -typedef struct generic_pm_domain *(*genpd_xlate_t)(struct of_phandle_args *args, - void *data); - #ifdef CONFIG_PM_GENERIC_DOMAINS_OF -int __of_genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, - void *data); +int of_genpd_add_provider_simple(struct device_node *np, + struct generic_pm_domain *genpd); +int of_genpd_add_provider_onecell(struct device_node *np, + struct genpd_onecell_data *data); void of_genpd_del_provider(struct device_node *np); -struct generic_pm_domain *__of_genpd_xlate_simple( - struct of_phandle_args *genpdspec, - void *data); -struct generic_pm_domain *__of_genpd_xlate_onecell( - struct of_phandle_args *genpdspec, - void *data); extern int of_genpd_add_device(struct of_phandle_args *args, struct device *dev); extern int of_genpd_add_subdomain(struct of_phandle_args *parent, @@ -207,15 +200,19 @@ extern int of_genpd_add_subdomain(struct of_phandle_args *parent, int genpd_dev_pm_attach(struct device *dev); #else /* !CONFIG_PM_GENERIC_DOMAINS_OF */ -static inline int __of_genpd_add_provider(struct device_node *np, - genpd_xlate_t xlate, void *data) +static inline int of_genpd_add_provider_simple(struct device_node *np, + struct generic_pm_domain *genpd) { - return 0; + return -ENOTSUPP; +} + +static inline int of_genpd_add_provider_onecell(struct device_node *np, + struct genpd_onecell_data *data) +{ + return -ENOTSUPP; } -static inline void of_genpd_del_provider(struct device_node *np) {} -#define __of_genpd_xlate_simple NULL -#define __of_genpd_xlate_onecell NULL +static inline void of_genpd_del_provider(struct device_node *np) {} static inline int of_genpd_add_device(struct of_phandle_args *args, struct device *dev) @@ -235,17 +232,6 @@ static inline int genpd_dev_pm_attach(struct device *dev) } #endif /* CONFIG_PM_GENERIC_DOMAINS_OF */ -static inline int of_genpd_add_provider_simple(struct device_node *np, - struct generic_pm_domain *genpd) -{ - return __of_genpd_add_provider(np, __of_genpd_xlate_simple, genpd); -} -static inline int of_genpd_add_provider_onecell(struct device_node *np, - struct genpd_onecell_data *data) -{ - return __of_genpd_add_provider(np, __of_genpd_xlate_onecell, data); -} - #ifdef CONFIG_PM extern int dev_pm_domain_attach(struct device *dev, bool power_on); extern void dev_pm_domain_detach(struct device *dev, bool power_off); -- cgit v0.10.2 From 0159ec670763dde7c8518034eda8ab9ec267fd23 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:10 +0100 Subject: PM / Domains: Verify the PM domain is present when adding a provider When a PM domain provider is added, there is currently no way to tell if any PM domains associated with the provider are present. Naturally, the PM domain provider should not be registered if the PM domains have not been added. Nonetheless, verify that the PM domain(s) associated with a provider are present when registering the PM domain provider. This change adds a dependency on the function pm_genpd_present() when CONFIG_PM_GENERIC_DOMAINS_OF is enabled and so ensure this function is available when CONFIG_PM_GENERIC_DOMAINS_OF selected. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index c93aa3a..fc9f11c 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -586,7 +586,7 @@ static int __init genpd_poweroff_unused(void) } late_initcall(genpd_poweroff_unused); -#ifdef CONFIG_PM_SLEEP +#if defined(CONFIG_PM_SLEEP) || defined(CONFIG_PM_GENERIC_DOMAINS_OF) /** * pm_genpd_present - Check if the given PM domain has been initialized. @@ -606,6 +606,10 @@ static bool pm_genpd_present(const struct generic_pm_domain *genpd) return false; } +#endif + +#ifdef CONFIG_PM_SLEEP + static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd, struct device *dev) { @@ -1453,7 +1457,19 @@ static int genpd_add_provider(struct device_node *np, genpd_xlate_t xlate, int of_genpd_add_provider_simple(struct device_node *np, struct generic_pm_domain *genpd) { - return genpd_add_provider(np, genpd_xlate_simple, genpd); + int ret = -EINVAL; + + if (!np || !genpd) + return -EINVAL; + + mutex_lock(&gpd_list_lock); + + if (pm_genpd_present(genpd)) + ret = genpd_add_provider(np, genpd_xlate_simple, genpd); + + mutex_unlock(&gpd_list_lock); + + return ret; } EXPORT_SYMBOL_GPL(of_genpd_add_provider_simple); @@ -1465,7 +1481,26 @@ EXPORT_SYMBOL_GPL(of_genpd_add_provider_simple); int of_genpd_add_provider_onecell(struct device_node *np, struct genpd_onecell_data *data) { - return genpd_add_provider(np, genpd_xlate_onecell, data); + unsigned int i; + int ret; + + if (!np || !data) + return -EINVAL; + + mutex_lock(&gpd_list_lock); + + for (i = 0; i < data->num_domains; i++) { + if (!pm_genpd_present(data->domains[i])) { + mutex_unlock(&gpd_list_lock); + return -EINVAL; + } + } + + ret = genpd_add_provider(np, genpd_xlate_onecell, data); + + mutex_unlock(&gpd_list_lock); + + return ret; } EXPORT_SYMBOL_GPL(of_genpd_add_provider_onecell); -- cgit v0.10.2 From 19efa5ff63dc5ed41ac85f5a8518aa77cece3776 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:11 +0100 Subject: PM / Domains: Prepare for adding support to remove PM domains In order to remove PM domains safely from the list of PM domains, it is necessary to adding locking for the PM domain list around any places where devices or subdomains are added to a PM domain. There are places where a reference to a PM domain is obtained via calling of_genpd_get_from_provider() before adding the device or the subdomain. In these cases a lock for the PM domain list needs to be held around the call to of_genpd_get_from_provider() and the call to add the device/subdomain. To avoid deadlocks by multiple attempts to obtain the PM domain list lock, add functions genpd_add_device() and genpd_add_subdomain() which require the user to hold the PM domain list lock when calling. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index fc9f11c..1bd8d41 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1060,14 +1060,8 @@ static void genpd_free_dev_data(struct device *dev, dev_pm_put_subsys_data(dev); } -/** - * __pm_genpd_add_device - Add a device to an I/O PM domain. - * @genpd: PM domain to add the device to. - * @dev: Device to be added. - * @td: Set of PM QoS timing parameters to attach to the device. - */ -int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, - struct gpd_timing_data *td) +static int genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, + struct gpd_timing_data *td) { struct generic_pm_domain_data *gpd_data; int ret = 0; @@ -1107,6 +1101,24 @@ int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, return ret; } + +/** + * __pm_genpd_add_device - Add a device to an I/O PM domain. + * @genpd: PM domain to add the device to. + * @dev: Device to be added. + * @td: Set of PM QoS timing parameters to attach to the device. + */ +int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, + struct gpd_timing_data *td) +{ + int ret; + + mutex_lock(&gpd_list_lock); + ret = genpd_add_device(genpd, dev, td); + mutex_unlock(&gpd_list_lock); + + return ret; +} EXPORT_SYMBOL_GPL(__pm_genpd_add_device); /** @@ -1160,13 +1172,8 @@ int pm_genpd_remove_device(struct generic_pm_domain *genpd, } EXPORT_SYMBOL_GPL(pm_genpd_remove_device); -/** - * pm_genpd_add_subdomain - Add a subdomain to an I/O PM domain. - * @genpd: Master PM domain to add the subdomain to. - * @subdomain: Subdomain to be added. - */ -int pm_genpd_add_subdomain(struct generic_pm_domain *genpd, - struct generic_pm_domain *subdomain) +static int genpd_add_subdomain(struct generic_pm_domain *genpd, + struct generic_pm_domain *subdomain) { struct gpd_link *link, *itr; int ret = 0; @@ -1209,6 +1216,23 @@ int pm_genpd_add_subdomain(struct generic_pm_domain *genpd, kfree(link); return ret; } + +/** + * pm_genpd_add_subdomain - Add a subdomain to an I/O PM domain. + * @genpd: Master PM domain to add the subdomain to. + * @subdomain: Subdomain to be added. + */ +int pm_genpd_add_subdomain(struct generic_pm_domain *genpd, + struct generic_pm_domain *subdomain) +{ + int ret; + + mutex_lock(&gpd_list_lock); + ret = genpd_add_subdomain(genpd, subdomain); + mutex_unlock(&gpd_list_lock); + + return ret; +} EXPORT_SYMBOL_GPL(pm_genpd_add_subdomain); /** @@ -1571,12 +1595,22 @@ static struct generic_pm_domain *genpd_get_from_provider( int of_genpd_add_device(struct of_phandle_args *genpdspec, struct device *dev) { struct generic_pm_domain *genpd; + int ret; + + mutex_lock(&gpd_list_lock); genpd = genpd_get_from_provider(genpdspec); - if (IS_ERR(genpd)) - return PTR_ERR(genpd); + if (IS_ERR(genpd)) { + ret = PTR_ERR(genpd); + goto out; + } + + ret = genpd_add_device(genpd, dev, NULL); - return pm_genpd_add_device(genpd, dev); +out: + mutex_unlock(&gpd_list_lock); + + return ret; } EXPORT_SYMBOL_GPL(of_genpd_add_device); @@ -1593,16 +1627,28 @@ int of_genpd_add_subdomain(struct of_phandle_args *parent_spec, struct of_phandle_args *subdomain_spec) { struct generic_pm_domain *parent, *subdomain; + int ret; + + mutex_lock(&gpd_list_lock); parent = genpd_get_from_provider(parent_spec); - if (IS_ERR(parent)) - return PTR_ERR(parent); + if (IS_ERR(parent)) { + ret = PTR_ERR(parent); + goto out; + } subdomain = genpd_get_from_provider(subdomain_spec); - if (IS_ERR(subdomain)) - return PTR_ERR(subdomain); + if (IS_ERR(subdomain)) { + ret = PTR_ERR(subdomain); + goto out; + } + + ret = genpd_add_subdomain(parent, subdomain); - return pm_genpd_add_subdomain(parent, subdomain); +out: + mutex_unlock(&gpd_list_lock); + + return ret; } EXPORT_SYMBOL_GPL(of_genpd_add_subdomain); @@ -1701,9 +1747,11 @@ int genpd_dev_pm_attach(struct device *dev) return -ENOENT; } + mutex_lock(&gpd_list_lock); pd = genpd_get_from_provider(&pd_args); of_node_put(pd_args.np); if (IS_ERR(pd)) { + mutex_unlock(&gpd_list_lock); dev_dbg(dev, "%s() failed to find PM domain: %ld\n", __func__, PTR_ERR(pd)); return -EPROBE_DEFER; @@ -1712,13 +1760,14 @@ int genpd_dev_pm_attach(struct device *dev) dev_dbg(dev, "adding to PM domain %s\n", pd->name); for (i = 1; i < GENPD_RETRY_MAX_MS; i <<= 1) { - ret = pm_genpd_add_device(pd, dev); + ret = genpd_add_device(pd, dev, NULL); if (ret != -EAGAIN) break; mdelay(i); cond_resched(); } + mutex_unlock(&gpd_list_lock); if (ret < 0) { dev_err(dev, "failed to add to PM domain %s: %d", -- cgit v0.10.2 From de0aa06d8b1c39df1071bfe169b3b97ca6bc01ac Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:12 +0100 Subject: PM / Domains: Store the provider in the PM domain structure It is possible that a device has more than one provider of PM domains and to support the removal of a PM domain by provider, it is necessary to store a reference to the provider in the PM domain structure. Therefore, store a reference to the firmware node handle in the PM domain structure and populate it when providers (only device-tree based providers are currently supported by PM domains) are registered. Please note that when removing PM domains, it is necessary to verify that the PM domain provider has been removed from the list of providers before the PM domain can be removed. To do this add another member to the PM domain structure that indicates if the provider is present and set this member accordingly when providers are added and removed. Initialise the 'provider' and 'has_provider' members of the generic_pm_domain structure when a PM domains is added by calling pm_genpd_init(). Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 1bd8d41..d5135ca 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1306,6 +1306,8 @@ int pm_genpd_init(struct generic_pm_domain *genpd, genpd->device_count = 0; genpd->max_off_time_ns = -1; genpd->max_off_time_changed = true; + genpd->provider = NULL; + genpd->has_provider = false; genpd->domain.ops.runtime_suspend = genpd_runtime_suspend; genpd->domain.ops.runtime_resume = genpd_runtime_resume; genpd->domain.ops.prepare = pm_genpd_prepare; @@ -1491,6 +1493,11 @@ int of_genpd_add_provider_simple(struct device_node *np, if (pm_genpd_present(genpd)) ret = genpd_add_provider(np, genpd_xlate_simple, genpd); + if (!ret) { + genpd->provider = &np->fwnode; + genpd->has_provider = true; + } + mutex_unlock(&gpd_list_lock); return ret; @@ -1506,7 +1513,7 @@ int of_genpd_add_provider_onecell(struct device_node *np, struct genpd_onecell_data *data) { unsigned int i; - int ret; + int ret = -EINVAL; if (!np || !data) return -EINVAL; @@ -1514,13 +1521,26 @@ int of_genpd_add_provider_onecell(struct device_node *np, mutex_lock(&gpd_list_lock); for (i = 0; i < data->num_domains; i++) { - if (!pm_genpd_present(data->domains[i])) { - mutex_unlock(&gpd_list_lock); - return -EINVAL; - } + if (!pm_genpd_present(data->domains[i])) + goto error; + + data->domains[i]->provider = &np->fwnode; + data->domains[i]->has_provider = true; } ret = genpd_add_provider(np, genpd_xlate_onecell, data); + if (ret < 0) + goto error; + + mutex_unlock(&gpd_list_lock); + + return 0; + +error: + while (i--) { + data->domains[i]->provider = NULL; + data->domains[i]->has_provider = false; + } mutex_unlock(&gpd_list_lock); @@ -1535,10 +1555,21 @@ EXPORT_SYMBOL_GPL(of_genpd_add_provider_onecell); void of_genpd_del_provider(struct device_node *np) { struct of_genpd_provider *cp; + struct generic_pm_domain *gpd; + mutex_lock(&gpd_list_lock); mutex_lock(&of_genpd_mutex); list_for_each_entry(cp, &of_genpd_providers, link) { if (cp->node == np) { + /* + * For each PM domain associated with the + * provider, set the 'has_provider' to false + * so that the PM domain can be safely removed. + */ + list_for_each_entry(gpd, &gpd_list, gpd_list_node) + if (gpd->provider == &np->fwnode) + gpd->has_provider = false; + list_del(&cp->link); of_node_put(cp->node); kfree(cp); @@ -1546,6 +1577,7 @@ void of_genpd_del_provider(struct device_node *np) } } mutex_unlock(&of_genpd_mutex); + mutex_unlock(&gpd_list_lock); } EXPORT_SYMBOL_GPL(of_genpd_del_provider); diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index f103869..554f891 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -51,6 +51,8 @@ struct generic_pm_domain { struct mutex lock; struct dev_power_governor *gov; struct work_struct power_off_work; + struct fwnode_handle *provider; /* Identity of the domain provider */ + bool has_provider; const char *name; atomic_t sd_count; /* Number of subdomains with power "on" */ enum gpd_status status; /* Current state of the domain */ -- cgit v0.10.2 From 3fe577107ccf1974958701df710e0b07ef16db75 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:13 +0100 Subject: PM / Domains: Add support for removing PM domains The genpd framework allows users to add PM domains via the pm_genpd_init() function, however, there is no corresponding function to remove a PM domain. For most devices this may be fine as the PM domains are never removed, however, for devices that wish to populate the PM domains from within a driver, having the ability to remove a PM domain if the probing of the device fails or the driver is unloaded is necessary. Add the function pm_genpd_remove() to remove a PM domain by referencing it's generic_pm_domain structure. Note that the bulk of the code that removes the PM domain is placed in a separate local function genpd_remove() (which is called by pm_genpd_remove()). The code is structured in this way to prepare for adding another function to remove a PM domain by provider that will also call genpd_remove(). Note that users of genpd_remove() must call this function with the mutex, gpd_list_lock, held. PM domains can only be removed if the associated provider has been removed, they are not a parent domain to another PM domain and have no devices associated with them. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index d5135ca..7481e3e 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1358,6 +1358,66 @@ int pm_genpd_init(struct generic_pm_domain *genpd, } EXPORT_SYMBOL_GPL(pm_genpd_init); +static int genpd_remove(struct generic_pm_domain *genpd) +{ + struct gpd_link *l, *link; + + if (IS_ERR_OR_NULL(genpd)) + return -EINVAL; + + mutex_lock(&genpd->lock); + + if (genpd->has_provider) { + mutex_unlock(&genpd->lock); + pr_err("Provider present, unable to remove %s\n", genpd->name); + return -EBUSY; + } + + if (!list_empty(&genpd->master_links) || genpd->device_count) { + mutex_unlock(&genpd->lock); + pr_err("%s: unable to remove %s\n", __func__, genpd->name); + return -EBUSY; + } + + list_for_each_entry_safe(link, l, &genpd->slave_links, slave_node) { + list_del(&link->master_node); + list_del(&link->slave_node); + kfree(link); + } + + list_del(&genpd->gpd_list_node); + mutex_unlock(&genpd->lock); + cancel_work_sync(&genpd->power_off_work); + pr_debug("%s: removed %s\n", __func__, genpd->name); + + return 0; +} + +/** + * pm_genpd_remove - Remove a generic I/O PM domain + * @genpd: Pointer to PM domain that is to be removed. + * + * To remove the PM domain, this function: + * - Removes the PM domain as a subdomain to any parent domains, + * if it was added. + * - Removes the PM domain from the list of registered PM domains. + * + * The PM domain will only be removed, if the associated provider has + * been removed, it is not a parent to any other PM domain and has no + * devices associated with it. + */ +int pm_genpd_remove(struct generic_pm_domain *genpd) +{ + int ret; + + mutex_lock(&gpd_list_lock); + ret = genpd_remove(genpd); + mutex_unlock(&gpd_list_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(pm_genpd_remove); + #ifdef CONFIG_PM_GENERIC_DOMAINS_OF typedef struct generic_pm_domain *(*genpd_xlate_t)(struct of_phandle_args *args, diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index 554f891..85f7d53 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -130,6 +130,7 @@ extern int pm_genpd_remove_subdomain(struct generic_pm_domain *genpd, struct generic_pm_domain *target); extern int pm_genpd_init(struct generic_pm_domain *genpd, struct dev_power_governor *gov, bool is_off); +extern int pm_genpd_remove(struct generic_pm_domain *genpd); extern struct dev_power_governor simple_qos_governor; extern struct dev_power_governor pm_domain_always_on_gov; @@ -165,6 +166,10 @@ static inline int pm_genpd_init(struct generic_pm_domain *genpd, { return -ENOSYS; } +static inline int pm_genpd_remove(struct generic_pm_domain *genpd) +{ + return -ENOTSUPP; +} #endif static inline int pm_genpd_add_device(struct generic_pm_domain *genpd, -- cgit v0.10.2 From 17926551c98a4ff5d7fa3a574c60534fedb3f2c6 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Mon, 12 Sep 2016 12:01:14 +0100 Subject: PM / Domains: Add support for removing nested PM domains by provider If a device supports PM domains that are subdomains of another PM domain, then the PM domains should be removed in reverse order to ensure that the subdomains are removed first. Furthermore, if there is more than one provider, then there needs to be a way to remove the domains in reverse order for a specific provider. Add the function of_genpd_remove_last() to remove the last PM domain added by a given PM domain provider and return the generic_pm_domain structure for the PM domain that was removed. Signed-off-by: Jon Hunter Acked-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 7481e3e..b0cf46d 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1745,6 +1745,41 @@ out: EXPORT_SYMBOL_GPL(of_genpd_add_subdomain); /** + * of_genpd_remove_last - Remove the last PM domain registered for a provider + * @provider: Pointer to device structure associated with provider + * + * Find the last PM domain that was added by a particular provider and + * remove this PM domain from the list of PM domains. The provider is + * identified by the 'provider' device structure that is passed. The PM + * domain will only be removed, if the provider associated with domain + * has been removed. + * + * Returns a valid pointer to struct generic_pm_domain on success or + * ERR_PTR() on failure. + */ +struct generic_pm_domain *of_genpd_remove_last(struct device_node *np) +{ + struct generic_pm_domain *gpd, *genpd = ERR_PTR(-ENOENT); + int ret; + + if (IS_ERR_OR_NULL(np)) + return ERR_PTR(-EINVAL); + + mutex_lock(&gpd_list_lock); + list_for_each_entry(gpd, &gpd_list, gpd_list_node) { + if (gpd->provider == &np->fwnode) { + ret = genpd_remove(gpd); + genpd = ret ? ERR_PTR(ret) : gpd; + break; + } + } + mutex_unlock(&gpd_list_lock); + + return genpd; +} +EXPORT_SYMBOL_GPL(of_genpd_remove_last); + +/** * genpd_dev_pm_detach - Detach a device from its PM domain. * @dev: Device to detach. * @power_off: Currently not used diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index 85f7d53..a09fe5c 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -204,6 +204,7 @@ extern int of_genpd_add_device(struct of_phandle_args *args, struct device *dev); extern int of_genpd_add_subdomain(struct of_phandle_args *parent, struct of_phandle_args *new_subdomain); +extern struct generic_pm_domain *of_genpd_remove_last(struct device_node *np); int genpd_dev_pm_attach(struct device *dev); #else /* !CONFIG_PM_GENERIC_DOMAINS_OF */ @@ -237,6 +238,12 @@ static inline int genpd_dev_pm_attach(struct device *dev) { return -ENODEV; } + +static inline +struct generic_pm_domain *of_genpd_remove_last(struct device_node *np) +{ + return ERR_PTR(-ENOTSUPP); +} #endif /* CONFIG_PM_GENERIC_DOMAINS_OF */ #ifdef CONFIG_PM -- cgit v0.10.2 From 8c34ab1910a79319731107ec8ecd2e80893ea30c Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 9 Sep 2016 23:59:33 +0200 Subject: cpufreq / sched: SCHED_CPUFREQ_IOWAIT flag to indicate iowait condition Testing indicates that it is possible to improve performace significantly without increasing energy consumption too much by teaching cpufreq governors to bump up the CPU performance level if the in_iowait flag is set for the task in enqueue_task_fair(). For this purpose, define a new cpufreq_update_util() flag SCHED_CPUFREQ_IOWAIT and modify enqueue_task_fair() to pass that flag to cpufreq_update_util() in the in_iowait case. That generally requires cpufreq_update_util() to be called directly from there, because update_load_avg() may not be invoked in that case. Signed-off-by: Rafael J. Wysocki Looks-good-to: Steve Muckle Acked-by: Peter Zijlstra (Intel) diff --git a/include/linux/sched.h b/include/linux/sched.h index b0fa726..98fe95f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -3471,6 +3471,7 @@ static inline unsigned long rlimit_max(unsigned int limit) #define SCHED_CPUFREQ_RT (1U << 0) #define SCHED_CPUFREQ_DL (1U << 1) +#define SCHED_CPUFREQ_IOWAIT (1U << 2) #define SCHED_CPUFREQ_RT_DL (SCHED_CPUFREQ_RT | SCHED_CPUFREQ_DL) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 5d558cc..a5cd07b 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -4500,6 +4500,14 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) struct cfs_rq *cfs_rq; struct sched_entity *se = &p->se; + /* + * If in_iowait is set, the code below may not trigger any cpufreq + * utilization updates, so do it here explicitly with the IOWAIT flag + * passed. + */ + if (p->in_iowait) + cpufreq_update_this_cpu(rq, SCHED_CPUFREQ_IOWAIT); + for_each_sched_entity(se) { if (se->on_rq) break; -- cgit v0.10.2 From 21ca6d2c52f8ca8638129c1dfc489d0b0ae68532 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sat, 10 Sep 2016 00:00:31 +0200 Subject: cpufreq: schedutil: Add iowait boosting Modify the schedutil cpufreq governor to boost the CPU frequency if the SCHED_CPUFREQ_IOWAIT flag is passed to it via cpufreq_update_util(). If that happens, the frequency is set to the maximum during the first update after receiving the SCHED_CPUFREQ_IOWAIT flag and then the boost is reduced by half during each following update. Signed-off-by: Rafael J. Wysocki Looks-good-to: Steve Muckle Acked-by: Peter Zijlstra (Intel) diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index cb8a77b..69e0689 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -47,11 +47,13 @@ struct sugov_cpu { struct sugov_policy *sg_policy; unsigned int cached_raw_freq; + unsigned long iowait_boost; + unsigned long iowait_boost_max; + u64 last_update; /* The fields below are only needed when sharing a policy. */ unsigned long util; unsigned long max; - u64 last_update; unsigned int flags; }; @@ -155,6 +157,36 @@ static void sugov_get_util(unsigned long *util, unsigned long *max) *max = cfs_max; } +static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time, + unsigned int flags) +{ + if (flags & SCHED_CPUFREQ_IOWAIT) { + sg_cpu->iowait_boost = sg_cpu->iowait_boost_max; + } else if (sg_cpu->iowait_boost) { + s64 delta_ns = time - sg_cpu->last_update; + + /* Clear iowait_boost if the CPU apprears to have been idle. */ + if (delta_ns > TICK_NSEC) + sg_cpu->iowait_boost = 0; + } +} + +static void sugov_iowait_boost(struct sugov_cpu *sg_cpu, unsigned long *util, + unsigned long *max) +{ + unsigned long boost_util = sg_cpu->iowait_boost; + unsigned long boost_max = sg_cpu->iowait_boost_max; + + if (!boost_util) + return; + + if (*util * boost_max < *max * boost_util) { + *util = boost_util; + *max = boost_max; + } + sg_cpu->iowait_boost >>= 1; +} + static void sugov_update_single(struct update_util_data *hook, u64 time, unsigned int flags) { @@ -164,6 +196,9 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, unsigned long util, max; unsigned int next_f; + sugov_set_iowait_boost(sg_cpu, time, flags); + sg_cpu->last_update = time; + if (!sugov_should_update_freq(sg_policy, time)) return; @@ -171,6 +206,7 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, next_f = policy->cpuinfo.max_freq; } else { sugov_get_util(&util, &max); + sugov_iowait_boost(sg_cpu, &util, &max); next_f = get_next_freq(sg_cpu, util, max); } sugov_update_commit(sg_policy, time, next_f); @@ -189,6 +225,8 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, if (flags & SCHED_CPUFREQ_RT_DL) return max_f; + sugov_iowait_boost(sg_cpu, &util, &max); + for_each_cpu(j, policy->cpus) { struct sugov_cpu *j_sg_cpu; unsigned long j_util, j_max; @@ -203,12 +241,13 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, * frequency update and the time elapsed between the last update * of the CPU utilization and the last frequency update is long * enough, don't take the CPU into account as it probably is - * idle now. + * idle now (and clear iowait_boost for it). */ delta_ns = last_freq_update_time - j_sg_cpu->last_update; - if (delta_ns > TICK_NSEC) + if (delta_ns > TICK_NSEC) { + j_sg_cpu->iowait_boost = 0; continue; - + } if (j_sg_cpu->flags & SCHED_CPUFREQ_RT_DL) return max_f; @@ -218,6 +257,8 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, util = j_util; max = j_max; } + + sugov_iowait_boost(j_sg_cpu, &util, &max); } return get_next_freq(sg_cpu, util, max); @@ -238,6 +279,8 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time, sg_cpu->util = util; sg_cpu->max = max; sg_cpu->flags = flags; + + sugov_set_iowait_boost(sg_cpu, time, flags); sg_cpu->last_update = time; if (sugov_should_update_freq(sg_policy, time)) { @@ -470,6 +513,8 @@ static int sugov_start(struct cpufreq_policy *policy) sg_cpu->flags = SCHED_CPUFREQ_RT; sg_cpu->last_update = 0; sg_cpu->cached_raw_freq = 0; + sg_cpu->iowait_boost = 0; + sg_cpu->iowait_boost_max = policy->cpuinfo.max_freq; cpufreq_add_update_util_hook(cpu, &sg_cpu->update_util, sugov_update_shared); } else { -- cgit v0.10.2 From 09c448d3c61f31322c097cb4c1484778d50da399 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 14 Sep 2016 02:28:13 +0200 Subject: cpufreq: intel_pstate: Use IOWAIT flag in Atom algorithm Modify the P-state selection algorithm for Atom processors to use the new SCHED_CPUFREQ_IOWAIT flag instead of the questionable get_cpu_iowait_time_us() function. Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index bdbe936..7c457cc 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -181,6 +181,8 @@ struct _pid { * @cpu: CPU number for this instance data * @update_util: CPUFreq utility callback information * @update_util_set: CPUFreq utility callback is set + * @iowait_boost: iowait-related boost fraction + * @last_update: Time of the last update. * @pstate: Stores P state limits for this CPU * @vid: Stores VID limits for this CPU * @pid: Stores PID parameters for this CPU @@ -206,6 +208,7 @@ struct cpudata { struct vid_data vid; struct _pid pid; + u64 last_update; u64 last_sample_time; u64 prev_aperf; u64 prev_mperf; @@ -216,6 +219,7 @@ struct cpudata { struct acpi_processor_performance acpi_perf_data; bool valid_pss_table; #endif + unsigned int iowait_boost; }; static struct cpudata **all_cpu_data; @@ -229,6 +233,7 @@ static struct cpudata **all_cpu_data; * @p_gain_pct: PID proportional gain * @i_gain_pct: PID integral gain * @d_gain_pct: PID derivative gain + * @boost_iowait: Whether or not to use iowait boosting. * * Stores per CPU model static PID configuration data. */ @@ -240,6 +245,7 @@ struct pstate_adjust_policy { int p_gain_pct; int d_gain_pct; int i_gain_pct; + bool boost_iowait; }; /** @@ -1037,6 +1043,7 @@ static struct cpu_defaults silvermont_params = { .p_gain_pct = 14, .d_gain_pct = 0, .i_gain_pct = 4, + .boost_iowait = true, }, .funcs = { .get_max = atom_get_max_pstate, @@ -1058,6 +1065,7 @@ static struct cpu_defaults airmont_params = { .p_gain_pct = 14, .d_gain_pct = 0, .i_gain_pct = 4, + .boost_iowait = true, }, .funcs = { .get_max = atom_get_max_pstate, @@ -1099,6 +1107,7 @@ static struct cpu_defaults bxt_params = { .p_gain_pct = 14, .d_gain_pct = 0, .i_gain_pct = 4, + .boost_iowait = true, }, .funcs = { .get_max = core_get_max_pstate, @@ -1222,36 +1231,18 @@ static inline int32_t get_avg_pstate(struct cpudata *cpu) static inline int32_t get_target_pstate_use_cpu_load(struct cpudata *cpu) { struct sample *sample = &cpu->sample; - u64 cummulative_iowait, delta_iowait_us; - u64 delta_iowait_mperf; - u64 mperf, now; - int32_t cpu_load; + int32_t busy_frac, boost; - cummulative_iowait = get_cpu_iowait_time_us(cpu->cpu, &now); + busy_frac = div_fp(sample->mperf, sample->tsc); - /* - * Convert iowait time into number of IO cycles spent at max_freq. - * IO is considered as busy only for the cpu_load algorithm. For - * performance this is not needed since we always try to reach the - * maximum P-State, so we are already boosting the IOs. - */ - delta_iowait_us = cummulative_iowait - cpu->prev_cummulative_iowait; - delta_iowait_mperf = div64_u64(delta_iowait_us * cpu->pstate.scaling * - cpu->pstate.max_pstate, MSEC_PER_SEC); + boost = cpu->iowait_boost; + cpu->iowait_boost >>= 1; - mperf = cpu->sample.mperf + delta_iowait_mperf; - cpu->prev_cummulative_iowait = cummulative_iowait; + if (busy_frac < boost) + busy_frac = boost; - /* - * The load can be estimated as the ratio of the mperf counter - * running at a constant frequency during active periods - * (C0) and the time stamp counter running at the same frequency - * also during C-states. - */ - cpu_load = div64_u64(int_tofp(100) * mperf, sample->tsc); - cpu->sample.busy_scaled = cpu_load; - - return get_avg_pstate(cpu) - pid_calc(&cpu->pid, cpu_load); + sample->busy_scaled = busy_frac * 100; + return get_avg_pstate(cpu) - pid_calc(&cpu->pid, sample->busy_scaled); } static inline int32_t get_target_pstate_use_performance(struct cpudata *cpu) @@ -1332,8 +1323,21 @@ static void intel_pstate_update_util(struct update_util_data *data, u64 time, unsigned int flags) { struct cpudata *cpu = container_of(data, struct cpudata, update_util); - u64 delta_ns = time - cpu->sample.time; + u64 delta_ns; + + if (pid_params.boost_iowait) { + if (flags & SCHED_CPUFREQ_IOWAIT) { + cpu->iowait_boost = int_tofp(1); + } else if (cpu->iowait_boost) { + /* Clear iowait_boost if the CPU may have been idle. */ + delta_ns = time - cpu->last_update; + if (delta_ns > TICK_NSEC) + cpu->iowait_boost = 0; + } + cpu->last_update = time; + } + delta_ns = time - cpu->sample.time; if ((s64)delta_ns >= pid_params.sample_rate_ns) { bool sample_taken = intel_pstate_sample(cpu, time); -- cgit v0.10.2 From 290128ac139310e6d1fe47c023978907f2324f14 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 14 Sep 2016 14:22:25 +1000 Subject: partial revert of "PM / devfreq: Add COMPILE_TEST for build coverage" This reverts the Tegra part of commit 797da5598f3a (PM / devfreq: Add COMPILE_TEST for build coverage) that introduced a build failute in in linux-next. [ rjw: Changelog ] Signed-off-by: Stephen Rothwell Signed-off-by: Rafael J. Wysocki diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig index cadd56e..93b6ada 100644 --- a/drivers/devfreq/Kconfig +++ b/drivers/devfreq/Kconfig @@ -92,7 +92,7 @@ config ARM_EXYNOS_BUS_DEVFREQ config ARM_TEGRA_DEVFREQ tristate "Tegra DEVFREQ Driver" - depends on ARCH_TEGRA_124_SOC || COMPILE_TEST + depends on ARCH_TEGRA_124_SOC select DEVFREQ_GOV_SIMPLE_ONDEMAND select PM_OPP help -- cgit v0.10.2 From 3ba7bcaa3657f5fe32295ebd17fbdaaf16608e2f Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Tue, 13 Sep 2016 17:41:33 -0700 Subject: cpufreq: intel_pstate: Add io_boost trace Add io_boost percent to current pstate_sample tracepoint. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 7c457cc..86c29af 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1316,7 +1316,8 @@ static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu) sample->mperf, sample->aperf, sample->tsc, - get_avg_frequency(cpu)); + get_avg_frequency(cpu), + fp_toint(cpu->iowait_boost * 100)); } static void intel_pstate_update_util(struct update_util_data *data, u64 time, diff --git a/include/trace/events/power.h b/include/trace/events/power.h index 19e5030..54e3aad 100644 --- a/include/trace/events/power.h +++ b/include/trace/events/power.h @@ -69,7 +69,8 @@ TRACE_EVENT(pstate_sample, u64 mperf, u64 aperf, u64 tsc, - u32 freq + u32 freq, + u32 io_boost ), TP_ARGS(core_busy, @@ -79,7 +80,8 @@ TRACE_EVENT(pstate_sample, mperf, aperf, tsc, - freq + freq, + io_boost ), TP_STRUCT__entry( @@ -91,6 +93,7 @@ TRACE_EVENT(pstate_sample, __field(u64, aperf) __field(u64, tsc) __field(u32, freq) + __field(u32, io_boost) ), TP_fast_assign( @@ -102,9 +105,10 @@ TRACE_EVENT(pstate_sample, __entry->aperf = aperf; __entry->tsc = tsc; __entry->freq = freq; + __entry->io_boost = io_boost; ), - TP_printk("core_busy=%lu scaled=%lu from=%lu to=%lu mperf=%llu aperf=%llu tsc=%llu freq=%lu ", + TP_printk("core_busy=%lu scaled=%lu from=%lu to=%lu mperf=%llu aperf=%llu tsc=%llu freq=%lu io_boost=%lu", (unsigned long)__entry->core_busy, (unsigned long)__entry->scaled_busy, (unsigned long)__entry->from, @@ -112,7 +116,8 @@ TRACE_EVENT(pstate_sample, (unsigned long long)__entry->mperf, (unsigned long long)__entry->aperf, (unsigned long long)__entry->tsc, - (unsigned long)__entry->freq + (unsigned long)__entry->freq, + (unsigned long)__entry->io_boost ) ); -- cgit v0.10.2 From e01072d22d4e7f9ca966f848def22fe41eaef4de Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Wed, 14 Sep 2016 15:41:37 -0500 Subject: cpufreq: ti: Use generic platdev driver Now that the cpufreq-dt-platdev is used to create the cpufreq-dt platform device for all OMAP platforms and the platform code that did it before has been removed, add ti,am33xx and ti,dra7xx to the machine list in cpufreq-dt-platdev which had relied on the removed platform code to do this previously. Fixes: 7694ca6e1d6f (cpufreq: omap: Use generic platdev driver) Signed-off-by: Dave Gerlach Acked-by: Viresh Kumar Cc: 4.7+ # 4.7+ Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cpufreq-dt-platdev.c b/drivers/cpufreq/cpufreq-dt-platdev.c index d2637e1..7126762 100644 --- a/drivers/cpufreq/cpufreq-dt-platdev.c +++ b/drivers/cpufreq/cpufreq-dt-platdev.c @@ -72,6 +72,8 @@ static const struct of_device_id machines[] __initconst = { { .compatible = "sigma,tango4" }, + { .compatible = "ti,am33xx", }, + { .compatible = "ti,dra7", }, { .compatible = "ti,omap2", }, { .compatible = "ti,omap3", }, { .compatible = "ti,omap4", }, -- cgit v0.10.2 From f89f4147f76f91bfff6f32890fc34148178f144d Mon Sep 17 00:00:00 2001 From: Hoan Tran Date: Wed, 14 Sep 2016 16:08:28 -0700 Subject: cpufreq: CPPC: Avoid overflow when calculating desired_perf This patch fixes overflow issue when calculating the desired_perf. Fixes: ad38677df44b (cpufreq: CPPC: Force reporting values in KHz to fix user space interface) Signed-off-by: Hoan Tran Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index 6debc18..99db422 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, cpu = all_cpu_data[policy->cpu]; - cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; + cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz; freqs.old = policy->cur; freqs.new = target_freq; -- cgit v0.10.2 From 609bed67bd8f72981faba75d80b3b48d746b7206 Mon Sep 17 00:00:00 2001 From: Tomeu Vizoso Date: Thu, 15 Sep 2016 14:05:23 +0200 Subject: PM / Domains: Allow holes in genpd_data.domains array In platforms such as Rockchip's, the array of domains isn't always filled without holes, as which domains are present depend on the particular SoC revision. By allowing holes to be in the array, such SoCs can still use a single set of constants to index the array of power domains. Fixes: 0159ec670763 (PM / Domains: Verify the PM domain is present when adding a provider) Signed-off-by: Tomeu Vizoso Acked-by: Jon Hunter Acked-by: Ulf Hansson Reviewed-by: Heiko Stuebner Tested-by: Heiko Stuebner Acked-by: Kevin Hilman Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index b0cf46d..83ae3d7 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1581,6 +1581,8 @@ int of_genpd_add_provider_onecell(struct device_node *np, mutex_lock(&gpd_list_lock); for (i = 0; i < data->num_domains; i++) { + if (!data->domains[i]) + continue; if (!pm_genpd_present(data->domains[i])) goto error; @@ -1598,6 +1600,8 @@ int of_genpd_add_provider_onecell(struct device_node *np, error: while (i--) { + if (!data->domains[i]) + continue; data->domains[i]->provider = NULL; data->domains[i]->has_provider = false; } -- cgit v0.10.2 From 4df27c91893fd13eaa30e9b0bca74f317816f428 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 15 Sep 2016 17:38:38 +0200 Subject: PM / OPP: avoid maybe-uninitialized warning When CONFIG_OPTIMIZE_INLINING is set and we are building with -Wmaybe-uninitialized enabled, we can get a warning for the opp core driver: drivers/base/power/opp/core.c: In function 'dev_pm_opp_set_rate': drivers/base/power/opp/core.c:560:8: warning: 'ou_volt_min' may be used uninitialized in this function [-Wmaybe-uninitialized] This has only now appeared as a result of commit 797da5598f3a ("PM / devfreq: Add COMPILE_TEST for build coverage"), which makes the driver visible in some configurations that didn't have it before. The warning is a false positive that I got with gcc-6.1.1, but there is a simple workaround in removing the local variables that we get warnings for (all three are affected depending on the configuration). This also makes the code easier to read. Signed-off-by: Arnd Bergmann Reviewed-by: Stephen Boyd Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/opp/core.c b/drivers/base/power/opp/core.c index df0c709..4c7c6da 100644 --- a/drivers/base/power/opp/core.c +++ b/drivers/base/power/opp/core.c @@ -584,7 +584,6 @@ int dev_pm_opp_set_rate(struct device *dev, unsigned long target_freq) struct clk *clk; unsigned long freq, old_freq; unsigned long u_volt, u_volt_min, u_volt_max; - unsigned long ou_volt, ou_volt_min, ou_volt_max; int ret; if (unlikely(!target_freq)) { @@ -620,11 +619,7 @@ int dev_pm_opp_set_rate(struct device *dev, unsigned long target_freq) } old_opp = _find_freq_ceil(opp_table, &old_freq); - if (!IS_ERR(old_opp)) { - ou_volt = old_opp->u_volt; - ou_volt_min = old_opp->u_volt_min; - ou_volt_max = old_opp->u_volt_max; - } else { + if (IS_ERR(old_opp)) { dev_err(dev, "%s: failed to find current OPP for freq %lu (%ld)\n", __func__, old_freq, PTR_ERR(old_opp)); } @@ -683,7 +678,8 @@ restore_freq: restore_voltage: /* This shouldn't harm even if the voltages weren't updated earlier */ if (!IS_ERR(old_opp)) - _set_opp_voltage(dev, reg, ou_volt, ou_volt_min, ou_volt_max); + _set_opp_voltage(dev, reg, old_opp->u_volt, + old_opp->u_volt_min, old_opp->u_volt_max); return ret; } -- cgit v0.10.2 From 54dec69bb975e71db83baf82bf4043df545b204c Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 15 Sep 2016 17:44:58 +0200 Subject: PM / devfreq: rockchip: add PM_DEVFREQ_EVENT dependency The newly added ARM_RK3399_DMC_DEVFREQ driver requires the DEVFREQ_EVENT_ROCKCHIP_DFI driver and tries to turn that on through a 'select' statement, and that in turn has a dependency on PM_DEVFREQ_EVENT, which may be disabled here: warning: (ARM_RK3399_DMC_DEVFREQ) selects DEVFREQ_EVENT_ROCKCHIP_DFI which has unmet direct dependencies (PM_DEVFREQ && PM_DEVFREQ_EVENT && ARCH_ROCKCHIP) We probably want a 'depends on' here, but other drivers use 'select' too, so for consistency I'm doing the same. Signed-off-by: Arnd Bergmann Fixes: 5a893e31a636 (PM / devfreq: rockchip: add devfreq driver for rk3399 dmc) Acked-by: Chanwoo Choi Acked-by: MyungJoo Ham Signed-off-by: Rafael J. Wysocki diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig index 93b6ada..41254e7 100644 --- a/drivers/devfreq/Kconfig +++ b/drivers/devfreq/Kconfig @@ -105,6 +105,7 @@ config ARM_RK3399_DMC_DEVFREQ depends on ARCH_ROCKCHIP select DEVFREQ_EVENT_ROCKCHIP_DFI select DEVFREQ_GOV_SIMPLE_ONDEMAND + select PM_DEVFREQ_EVENT select PM_OPP help This adds the DEVFREQ driver for the RK3399 DMC(Dynamic Memory Controller). -- cgit v0.10.2 From da4a64481b3504a6f4f07c08ff03be8f08d9638d Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 18 Sep 2016 10:41:05 +0800 Subject: PM / devfreq: rk3399_dmc: Remove explictly regulator_put call in .remove Current code uses devm_regulator_get() in .probe so a regulator_put() will be automatically called when unload the module. Remove the explictly regulator_put() call and then we can also remove rk3399_dmcfreq_remove(). Signed-off-by: Axel Lin Reviewed-by: Chanwoo Choi Acked-by: MyungJoo Ham Signed-off-by: Rafael J. Wysocki diff --git a/drivers/devfreq/rk3399_dmc.c b/drivers/devfreq/rk3399_dmc.c index 54d65f2..e24b73d 100644 --- a/drivers/devfreq/rk3399_dmc.c +++ b/drivers/devfreq/rk3399_dmc.c @@ -450,15 +450,6 @@ static int rk3399_dmcfreq_probe(struct platform_device *pdev) return 0; } -static int rk3399_dmcfreq_remove(struct platform_device *pdev) -{ - struct rk3399_dmcfreq *dmcfreq = platform_get_drvdata(pdev); - - regulator_put(dmcfreq->vdd_center); - - return 0; -} - static const struct of_device_id rk3399dmc_devfreq_of_match[] = { { .compatible = "rockchip,rk3399-dmc" }, { }, @@ -466,7 +457,6 @@ static const struct of_device_id rk3399dmc_devfreq_of_match[] = { static struct platform_driver rk3399_dmcfreq_driver = { .probe = rk3399_dmcfreq_probe, - .remove = rk3399_dmcfreq_remove, .driver = { .name = "rk3399-dmc-freq", .pm = &rk3399_dmcfreq_pm, -- cgit v0.10.2 From 85168d56cc8603963cfdda1e2b9e8f9b68aef867 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 21 Sep 2016 15:38:50 +0200 Subject: PM / Domains: Simplify detaching a device from its genpd There's no need to validate the PM domain by using genpd_lookup_dev() when removing the device via genpd's genpd_dev_pm_detach() function. That's because this function can't be called, unless there is a valid PM domain for the device. To simplify the behaviour, let's move code from pm_genpd_remove_device() into a new internal function, genpd_remove_device(), which is called from pm_genpd_remove_device() and genpd_dev_pm_detach(). Signed-off-by: Ulf Hansson Reviewed-by: Lina Iyer Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 83ae3d7..63f0f8b 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1121,13 +1121,8 @@ int __pm_genpd_add_device(struct generic_pm_domain *genpd, struct device *dev, } EXPORT_SYMBOL_GPL(__pm_genpd_add_device); -/** - * pm_genpd_remove_device - Remove a device from an I/O PM domain. - * @genpd: PM domain to remove the device from. - * @dev: Device to be removed. - */ -int pm_genpd_remove_device(struct generic_pm_domain *genpd, - struct device *dev) +static int genpd_remove_device(struct generic_pm_domain *genpd, + struct device *dev) { struct generic_pm_domain_data *gpd_data; struct pm_domain_data *pdd; @@ -1135,10 +1130,6 @@ int pm_genpd_remove_device(struct generic_pm_domain *genpd, dev_dbg(dev, "%s()\n", __func__); - if (!genpd || genpd != genpd_lookup_dev(dev)) - return -EINVAL; - - /* The above validation also means we have existing domain_data. */ pdd = dev->power.subsys_data->domain_data; gpd_data = to_gpd_data(pdd); dev_pm_qos_remove_notifier(dev, &gpd_data->nb); @@ -1170,6 +1161,20 @@ int pm_genpd_remove_device(struct generic_pm_domain *genpd, return ret; } + +/** + * pm_genpd_remove_device - Remove a device from an I/O PM domain. + * @genpd: PM domain to remove the device from. + * @dev: Device to be removed. + */ +int pm_genpd_remove_device(struct generic_pm_domain *genpd, + struct device *dev) +{ + if (!genpd || genpd != genpd_lookup_dev(dev)) + return -EINVAL; + + return genpd_remove_device(genpd, dev); +} EXPORT_SYMBOL_GPL(pm_genpd_remove_device); static int genpd_add_subdomain(struct generic_pm_domain *genpd, @@ -1797,14 +1802,14 @@ static void genpd_dev_pm_detach(struct device *dev, bool power_off) unsigned int i; int ret = 0; - pd = genpd_lookup_dev(dev); - if (!pd) + pd = dev_to_genpd(dev); + if (IS_ERR(pd)) return; dev_dbg(dev, "removing from PM domain %s\n", pd->name); for (i = 1; i < GENPD_RETRY_MAX_MS; i <<= 1) { - ret = pm_genpd_remove_device(pd, dev); + ret = genpd_remove_device(pd, dev); if (ret != -EAGAIN) break; -- cgit v0.10.2 From 294f47ffd55c3710c6f8ca1122e0424f045a099c Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 21 Sep 2016 15:38:51 +0200 Subject: PM / Domains: Remove redundant system PM callbacks In cases when the PM domain haven't assigned a system PM callback, the PM core fall-backs to check for the callback at the driver level instead. This makes it redundant to assign a pm_generic_* helper function to a corresponding system PM callback at a PM domain level. Therefore, let's remove these assignments in pm_genpd_init(). Signed-off-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 63f0f8b..661ac3a 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -1316,24 +1316,12 @@ int pm_genpd_init(struct generic_pm_domain *genpd, genpd->domain.ops.runtime_suspend = genpd_runtime_suspend; genpd->domain.ops.runtime_resume = genpd_runtime_resume; genpd->domain.ops.prepare = pm_genpd_prepare; - genpd->domain.ops.suspend = pm_generic_suspend; - genpd->domain.ops.suspend_late = pm_generic_suspend_late; genpd->domain.ops.suspend_noirq = pm_genpd_suspend_noirq; genpd->domain.ops.resume_noirq = pm_genpd_resume_noirq; - genpd->domain.ops.resume_early = pm_generic_resume_early; - genpd->domain.ops.resume = pm_generic_resume; - genpd->domain.ops.freeze = pm_generic_freeze; - genpd->domain.ops.freeze_late = pm_generic_freeze_late; genpd->domain.ops.freeze_noirq = pm_genpd_freeze_noirq; genpd->domain.ops.thaw_noirq = pm_genpd_thaw_noirq; - genpd->domain.ops.thaw_early = pm_generic_thaw_early; - genpd->domain.ops.thaw = pm_generic_thaw; - genpd->domain.ops.poweroff = pm_generic_poweroff; - genpd->domain.ops.poweroff_late = pm_generic_poweroff_late; genpd->domain.ops.poweroff_noirq = pm_genpd_suspend_noirq; genpd->domain.ops.restore_noirq = pm_genpd_restore_noirq; - genpd->domain.ops.restore_early = pm_generic_restore_early; - genpd->domain.ops.restore = pm_generic_restore; genpd->domain.ops.complete = pm_genpd_complete; if (genpd->flags & GENPD_FLAG_PM_CLK) { -- cgit v0.10.2 From adb560b3ee395ac9688a93582d805edec4c3d3e7 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 21 Sep 2016 15:38:52 +0200 Subject: PM / Domains: Don't measure latency of ->power_on|off() during system PM Measure latency does by itself contribute to an increased latency, thus we should avoid it when it isn't needed. Currently genpd measures latencies in the system PM phase for the ->power_on|off() callbacks, except in the syscore case when it's not allowed to use ktime_get() as timekeeping may be suspended. Since there should be plenty of occasions during runtime PM to perform these measurements, let's rely on that and drop them from system PM. This will also make it consistent for how measurements are done of the runtime PM callbacks (as those may be invoked during system PM). Signed-off-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index 661ac3a..d52709e 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -619,7 +619,6 @@ static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd, /** * pm_genpd_sync_poweroff - Synchronously power off a PM domain and its masters. * @genpd: PM domain to power off, if possible. - * @timed: True if latency measurements are allowed. * * Check if the given PM domain can be powered off (during system suspend or * hibernation) and do that if so. Also, in that case propagate to its masters. @@ -629,8 +628,7 @@ static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd, * executed sequentially, so it is guaranteed that it will never run twice in * parallel). */ -static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd, - bool timed) +static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd) { struct gpd_link *link; @@ -643,28 +641,26 @@ static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd, /* Choose the deepest state when suspending */ genpd->state_idx = genpd->state_count - 1; - genpd_power_off(genpd, timed); + genpd_power_off(genpd, false); genpd->status = GPD_STATE_POWER_OFF; list_for_each_entry(link, &genpd->slave_links, slave_node) { genpd_sd_counter_dec(link->master); - pm_genpd_sync_poweroff(link->master, timed); + pm_genpd_sync_poweroff(link->master); } } /** * pm_genpd_sync_poweron - Synchronously power on a PM domain and its masters. * @genpd: PM domain to power on. - * @timed: True if latency measurements are allowed. * * This function is only called in "noirq" and "syscore" stages of system power * transitions, so it need not acquire locks (all of the "noirq" callbacks are * executed sequentially, so it is guaranteed that it will never run twice in * parallel). */ -static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd, - bool timed) +static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd) { struct gpd_link *link; @@ -672,11 +668,11 @@ static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd, return; list_for_each_entry(link, &genpd->slave_links, slave_node) { - pm_genpd_sync_poweron(link->master, timed); + pm_genpd_sync_poweron(link->master); genpd_sd_counter_inc(link->master); } - genpd_power_on(genpd, timed); + genpd_power_on(genpd, false); genpd->status = GPD_STATE_ACTIVE; } @@ -788,7 +784,7 @@ static int pm_genpd_suspend_noirq(struct device *dev) * the same PM domain, so it is not necessary to use locking here. */ genpd->suspended_count++; - pm_genpd_sync_poweroff(genpd, true); + pm_genpd_sync_poweroff(genpd); return 0; } @@ -818,7 +814,7 @@ static int pm_genpd_resume_noirq(struct device *dev) * guaranteed that this function will never run twice in parallel for * the same PM domain, so it is not necessary to use locking here. */ - pm_genpd_sync_poweron(genpd, true); + pm_genpd_sync_poweron(genpd); genpd->suspended_count--; if (genpd->dev_ops.stop && genpd->dev_ops.start) @@ -911,7 +907,7 @@ static int pm_genpd_restore_noirq(struct device *dev) */ genpd->status = GPD_STATE_POWER_OFF; - pm_genpd_sync_poweron(genpd, true); + pm_genpd_sync_poweron(genpd); if (genpd->dev_ops.stop && genpd->dev_ops.start) ret = pm_runtime_force_resume(dev); @@ -966,9 +962,9 @@ static void genpd_syscore_switch(struct device *dev, bool suspend) if (suspend) { genpd->suspended_count++; - pm_genpd_sync_poweroff(genpd, false); + pm_genpd_sync_poweroff(genpd); } else { - pm_genpd_sync_poweron(genpd, false); + pm_genpd_sync_poweron(genpd); genpd->suspended_count--; } } -- cgit v0.10.2 From eefdee07074db61764408583a23e96cf60d5a0a7 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Wed, 21 Sep 2016 15:38:53 +0200 Subject: PM / Domains: Rename pm_genpd_sync_poweron|poweroff() These are internal static functions to genpd. Let's conform to the naming rules, by dropping the "pm_" prefix from these. Signed-off-by: Ulf Hansson Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index d52709e..e023066 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -617,7 +617,7 @@ static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd, } /** - * pm_genpd_sync_poweroff - Synchronously power off a PM domain and its masters. + * genpd_sync_poweroff - Synchronously power off a PM domain and its masters. * @genpd: PM domain to power off, if possible. * * Check if the given PM domain can be powered off (during system suspend or @@ -628,7 +628,7 @@ static bool genpd_dev_active_wakeup(struct generic_pm_domain *genpd, * executed sequentially, so it is guaranteed that it will never run twice in * parallel). */ -static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd) +static void genpd_sync_poweroff(struct generic_pm_domain *genpd) { struct gpd_link *link; @@ -647,12 +647,12 @@ static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd) list_for_each_entry(link, &genpd->slave_links, slave_node) { genpd_sd_counter_dec(link->master); - pm_genpd_sync_poweroff(link->master); + genpd_sync_poweroff(link->master); } } /** - * pm_genpd_sync_poweron - Synchronously power on a PM domain and its masters. + * genpd_sync_poweron - Synchronously power on a PM domain and its masters. * @genpd: PM domain to power on. * * This function is only called in "noirq" and "syscore" stages of system power @@ -660,7 +660,7 @@ static void pm_genpd_sync_poweroff(struct generic_pm_domain *genpd) * executed sequentially, so it is guaranteed that it will never run twice in * parallel). */ -static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd) +static void genpd_sync_poweron(struct generic_pm_domain *genpd) { struct gpd_link *link; @@ -668,7 +668,7 @@ static void pm_genpd_sync_poweron(struct generic_pm_domain *genpd) return; list_for_each_entry(link, &genpd->slave_links, slave_node) { - pm_genpd_sync_poweron(link->master); + genpd_sync_poweron(link->master); genpd_sd_counter_inc(link->master); } @@ -784,7 +784,7 @@ static int pm_genpd_suspend_noirq(struct device *dev) * the same PM domain, so it is not necessary to use locking here. */ genpd->suspended_count++; - pm_genpd_sync_poweroff(genpd); + genpd_sync_poweroff(genpd); return 0; } @@ -814,7 +814,7 @@ static int pm_genpd_resume_noirq(struct device *dev) * guaranteed that this function will never run twice in parallel for * the same PM domain, so it is not necessary to use locking here. */ - pm_genpd_sync_poweron(genpd); + genpd_sync_poweron(genpd); genpd->suspended_count--; if (genpd->dev_ops.stop && genpd->dev_ops.start) @@ -902,12 +902,12 @@ static int pm_genpd_restore_noirq(struct device *dev) if (genpd->suspended_count++ == 0) /* * The boot kernel might put the domain into arbitrary state, - * so make it appear as powered off to pm_genpd_sync_poweron(), + * so make it appear as powered off to genpd_sync_poweron(), * so that it tries to power it on in case it was really off. */ genpd->status = GPD_STATE_POWER_OFF; - pm_genpd_sync_poweron(genpd); + genpd_sync_poweron(genpd); if (genpd->dev_ops.stop && genpd->dev_ops.start) ret = pm_runtime_force_resume(dev); @@ -962,9 +962,9 @@ static void genpd_syscore_switch(struct device *dev, bool suspend) if (suspend) { genpd->suspended_count++; - pm_genpd_sync_poweroff(genpd); + genpd_sync_poweroff(genpd); } else { - pm_genpd_sync_poweron(genpd); + genpd_sync_poweron(genpd); genpd->suspended_count--; } } -- cgit v0.10.2 From 4c232f94693a1d9cde4867b938830c361313eda7 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Sun, 25 Sep 2016 14:35:43 -0700 Subject: cpufreq: kirkwood: add missing \n to end of dev_err messages Trival fix, dev_err messages are missing a \n, so add it. Signed-off-by: Colin Ian King Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/kirkwood-cpufreq.c b/drivers/cpufreq/kirkwood-cpufreq.c index be42f10..1b9bcd7 100644 --- a/drivers/cpufreq/kirkwood-cpufreq.c +++ b/drivers/cpufreq/kirkwood-cpufreq.c @@ -123,7 +123,7 @@ static int kirkwood_cpufreq_probe(struct platform_device *pdev) priv.cpu_clk = of_clk_get_by_name(np, "cpu_clk"); if (IS_ERR(priv.cpu_clk)) { - dev_err(priv.dev, "Unable to get cpuclk"); + dev_err(priv.dev, "Unable to get cpuclk\n"); return PTR_ERR(priv.cpu_clk); } @@ -132,7 +132,7 @@ static int kirkwood_cpufreq_probe(struct platform_device *pdev) priv.ddr_clk = of_clk_get_by_name(np, "ddrclk"); if (IS_ERR(priv.ddr_clk)) { - dev_err(priv.dev, "Unable to get ddrclk"); + dev_err(priv.dev, "Unable to get ddrclk\n"); err = PTR_ERR(priv.ddr_clk); goto out_cpu; } @@ -142,7 +142,7 @@ static int kirkwood_cpufreq_probe(struct platform_device *pdev) priv.powersave_clk = of_clk_get_by_name(np, "powersave"); if (IS_ERR(priv.powersave_clk)) { - dev_err(priv.dev, "Unable to get powersave"); + dev_err(priv.dev, "Unable to get powersave\n"); err = PTR_ERR(priv.powersave_clk); goto out_ddr; } @@ -155,7 +155,7 @@ static int kirkwood_cpufreq_probe(struct platform_device *pdev) if (!err) return 0; - dev_err(priv.dev, "Failed to register cpufreq driver"); + dev_err(priv.dev, "Failed to register cpufreq driver\n"); clk_disable_unprepare(priv.powersave_clk); out_ddr: -- cgit v0.10.2 From 9ad0a1b6a21fbfa0b308e704fd924836f7ae4458 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Sun, 25 Sep 2016 14:40:13 -0700 Subject: cpufreq: st: add missing \n to end of dev_err message Trival fix, dev_err message is missing a \n, so add it. Signed-off-by: Colin Ian King Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/cpufreq/sti-cpufreq.c b/drivers/cpufreq/sti-cpufreq.c index 0404203..b366e6d 100644 --- a/drivers/cpufreq/sti-cpufreq.c +++ b/drivers/cpufreq/sti-cpufreq.c @@ -163,7 +163,7 @@ static int sti_cpufreq_set_opp_info(void) reg_fields = sti_cpufreq_match(); if (!reg_fields) { - dev_err(dev, "This SoC doesn't support voltage scaling"); + dev_err(dev, "This SoC doesn't support voltage scaling\n"); return -ENODEV; } -- cgit v0.10.2 From a4ee4545932d4b26ec0c190f2ce265de79990c7a Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Fri, 23 Sep 2016 15:07:47 -0500 Subject: PM / OPP: Don't support OPP if it provides supported-hw but platform does not The OPP framework allows each OPP to set a opp-supported-hw property which provides values that are matched against supported_hw values provided by the platform to limit support for certain OPPs on specific hardware. Currently, if the platform does not set supported_hw values, all OPPs are interpreted as supported, even if they have provided their own opp-supported-hw values. If an OPP has provided opp-supported-hw, it is indicating that there is some specific hardware configuration it is supported by. These constraints should be honored, and if no supported_hw has been provided by the platform, there is no way to determine if that OPP is actually supported, so it should be marked as not supported. Signed-off-by: Dave Gerlach Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki diff --git a/drivers/base/power/opp/of.c b/drivers/base/power/opp/of.c index 1dfd3dd..5552211 100644 --- a/drivers/base/power/opp/of.c +++ b/drivers/base/power/opp/of.c @@ -71,8 +71,18 @@ static bool _opp_is_supported(struct device *dev, struct opp_table *opp_table, u32 version; int ret; - if (!opp_table->supported_hw) - return true; + if (!opp_table->supported_hw) { + /* + * In the case that no supported_hw has been set by the + * platform but there is an opp-supported-hw value set for + * an OPP then the OPP should not be enabled as there is + * no way to see if the hardware supports it. + */ + if (of_find_property(np, "opp-supported-hw", NULL)) + return false; + else + return true; + } while (count--) { ret = of_property_read_u32_index(np, "opp-supported-hw", count, -- cgit v0.10.2