From 5efbe4279f959a3f5ed26adf5f05cb78dd1ffa7e Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Tue, 23 Oct 2012 01:07:46 +0200 Subject: PM / QoS: Introduce request and constraint data types for PM QoS flags Introduce struct pm_qos_flags_request and struct pm_qos_flags representing PM QoS flags request type and PM QoS flags constraint type, respectively. With these definitions the data structures will be arranged so that the list member of a struct pm_qos_flags object will contain the head of a list of struct pm_qos_flags_request objects representing all of the "flags" requests present for the given device. Then, the effective_flags member of a struct pm_qos_flags object will contain the bitwise OR of the flags members of all the struct pm_qos_flags_request objects in the list. Additionally, introduce helper function pm_qos_update_flags() allowing the caller to manage the list of struct pm_qos_flags_request pointed to by the list member of struct pm_qos_flags. The flags are of type s32 so that the request's "value" field is always of the same type regardless of what kind of request it is (latency requests already have value fields of type s32). Signed-off-by: Rafael J. Wysocki Reviewed-by: Jean Pihet Acked-by: mark gross --- kernel/power/qos.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 63 insertions(+) (limited to 'kernel') diff --git a/kernel/power/qos.c b/kernel/power/qos.c index 846bd42c7ed..2ab2819aee6 100644 --- a/kernel/power/qos.c +++ b/kernel/power/qos.c @@ -212,6 +212,69 @@ int pm_qos_update_target(struct pm_qos_constraints *c, struct plist_node *node, } } +/** + * pm_qos_flags_remove_req - Remove device PM QoS flags request. + * @pqf: Device PM QoS flags set to remove the request from. + * @req: Request to remove from the set. + */ +static void pm_qos_flags_remove_req(struct pm_qos_flags *pqf, + struct pm_qos_flags_request *req) +{ + s32 val = 0; + + list_del(&req->node); + list_for_each_entry(req, &pqf->list, node) + val |= req->flags; + + pqf->effective_flags = val; +} + +/** + * pm_qos_update_flags - Update a set of PM QoS flags. + * @pqf: Set of flags to update. + * @req: Request to add to the set, to modify, or to remove from the set. + * @action: Action to take on the set. + * @val: Value of the request to add or modify. + * + * Update the given set of PM QoS flags and call notifiers if the aggregate + * value has changed. Returns 1 if the aggregate constraint value has changed, + * 0 otherwise. + */ +bool pm_qos_update_flags(struct pm_qos_flags *pqf, + struct pm_qos_flags_request *req, + enum pm_qos_req_action action, s32 val) +{ + unsigned long irqflags; + s32 prev_value, curr_value; + + spin_lock_irqsave(&pm_qos_lock, irqflags); + + prev_value = list_empty(&pqf->list) ? 0 : pqf->effective_flags; + + switch (action) { + case PM_QOS_REMOVE_REQ: + pm_qos_flags_remove_req(pqf, req); + break; + case PM_QOS_UPDATE_REQ: + pm_qos_flags_remove_req(pqf, req); + case PM_QOS_ADD_REQ: + req->flags = val; + INIT_LIST_HEAD(&req->node); + list_add_tail(&req->node, &pqf->list); + pqf->effective_flags |= val; + break; + default: + /* no action */ + ; + } + + curr_value = list_empty(&pqf->list) ? 0 : pqf->effective_flags; + + spin_unlock_irqrestore(&pm_qos_lock, irqflags); + + return prev_value != curr_value; +} + /** * pm_qos_request - returns current system wide qos expectation * @pm_qos_class: identification of which qos value is requested -- cgit v1.2.3-70-g09d2 From 5e5041f3527b36b58e864886ba34c179ad40ff92 Mon Sep 17 00:00:00 2001 From: Yasuaki Ishimatsu Date: Tue, 23 Oct 2012 01:30:54 +0200 Subject: ACPI / processor: prevent cpu from becoming online Even if acpi_processor_handle_eject() offlines cpu, there is a chance to online the cpu after that. So the patch closes the window by using get/put_online_cpus(). Why does the patch change _cpu_up() logic? The patch cares the race of hot-remove cpu and _cpu_up(). If the patch does not change it, there is the following race. hot-remove cpu | _cpu_up() ------------------------------------- ------------------------------------ call acpi_processor_handle_eject() | call cpu_down() | call get_online_cpus() | | call cpu_hotplug_begin() and stop here call arch_unregister_cpu() | call acpi_unmap_lsapic() | call put_online_cpus() | | start and continue _cpu_up() return acpi_processor_remove() | continue hot-remove the cpu | So _cpu_up() can continue to itself. And hot-remove cpu can also continue itself. If the patch changes _cpu_up() logic, the race disappears as below: hot-remove cpu | _cpu_up() ----------------------------------------------------------------------- call acpi_processor_handle_eject() | call cpu_down() | call get_online_cpus() | | call cpu_hotplug_begin() and stop here call arch_unregister_cpu() | call acpi_unmap_lsapic() | cpu's cpu_present is set | to false by set_cpu_present()| call put_online_cpus() | | start _cpu_up() | check cpu_present() and return -EINVAL return acpi_processor_remove() | continue hot-remove the cpu | Signed-off-by: Yasuaki Ishimatsu Reviewed-by: Srivatsa S. Bhat Reviewed-by: Toshi Kani Signed-off-by: Rafael J. Wysocki --- drivers/acpi/processor_driver.c | 14 ++++++++++++++ kernel/cpu.c | 8 +++++--- 2 files changed, 19 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index bd4e5dca3ff..a4352b88a33 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -852,8 +852,22 @@ static int acpi_processor_handle_eject(struct acpi_processor *pr) if (cpu_online(pr->id)) cpu_down(pr->id); + get_online_cpus(); + /* + * The cpu might become online again at this point. So we check whether + * the cpu has been onlined or not. If the cpu became online, it means + * that someone wants to use the cpu. So acpi_processor_handle_eject() + * returns -EAGAIN. + */ + if (unlikely(cpu_online(pr->id))) { + put_online_cpus(); + pr_warn("Failed to remove CPU %d, because other task " + "brought the CPU back online\n", pr->id); + return -EAGAIN; + } arch_unregister_cpu(pr->id); acpi_unmap_lsapic(pr->id); + put_online_cpus(); return (0); } #else diff --git a/kernel/cpu.c b/kernel/cpu.c index 42bd331ee0a..f45657f1eb8 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -348,11 +348,13 @@ static int __cpuinit _cpu_up(unsigned int cpu, int tasks_frozen) unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0; struct task_struct *idle; - if (cpu_online(cpu) || !cpu_present(cpu)) - return -EINVAL; - cpu_hotplug_begin(); + if (cpu_online(cpu) || !cpu_present(cpu)) { + ret = -EINVAL; + goto out; + } + idle = idle_thread_get(cpu); if (IS_ERR(idle)) { ret = PTR_ERR(idle); -- cgit v1.2.3-70-g09d2 From 69a37beabf1f0a6705c08e879bdd5d82ff6486c4 Mon Sep 17 00:00:00 2001 From: Youquan Song Date: Fri, 26 Oct 2012 12:26:41 +0200 Subject: cpuidle: Quickly notice prediction failure for repeat mode The prediction for future is difficult and when the cpuidle governor prediction fails and govenor possibly choose the shallower C-state than it should. How to quickly notice and find the failure becomes important for power saving. cpuidle menu governor has a method to predict the repeat pattern if there are 8 C-states residency which are continuous and the same or very close, so it will predict the next C-states residency will keep same residency time. There is a real case that turbostat utility (tools/power/x86/turbostat) at kernel 3.3 or early. turbostat utility will read 10 registers one by one at Sandybridge, so it will generate 10 IPIs to wake up idle CPUs. So cpuidle menu governor will predict it is repeat mode and there is another IPI wake up idle CPU soon, so it keeps idle CPU stay at C1 state even though CPU is totally idle. However, in the turbostat, following 10 registers reading is sleep 5 seconds by default, so the idle CPU will keep at C1 for a long time though it is idle until break event occurs. In a idle Sandybridge system, run "./turbostat -v", we will notice that deep C-state dangles between "70% ~ 99%". After patched the kernel, we will notice deep C-state stays at >99.98%. In the patch, a timer is added when menu governor detects a repeat mode and choose a shallow C-state. The timer is set to a time out value that greater than predicted time, and we conclude repeat mode prediction failure if timer is triggered. When repeat mode happens as expected, the timer is not triggered and CPU waken up from C-states and it will cancel the timer initiatively. When repeat mode does not happen, the timer will be time out and menu governor will quickly notice that the repeat mode prediction fails and then re-evaluates deeper C-states possibility. Below is another case which will clearly show the patch much benefit: #include #include #include #include #include #include #include volatile int * shutdown; volatile long * count; int delay = 20; int loop = 8; void usage(void) { fprintf(stderr, "Usage: idle_predict [options]\n" " --help -h Print this help\n" " --thread -n Thread number\n" " --loop -l Loop times in shallow Cstate\n" " --delay -t Sleep time (uS)in shallow Cstate\n"); } void *simple_loop() { int idle_num = 1; while (!(*shutdown)) { *count = *count + 1; if (idle_num % loop) usleep(delay); else { /* sleep 1 second */ usleep(1000000); idle_num = 0; } idle_num++; } } static void sighand(int sig) { *shutdown = 1; } int main(int argc, char *argv[]) { sigset_t sigset; int signum = SIGALRM; int i, c, er = 0, thread_num = 8; pthread_t pt[1024]; static char optstr[] = "n:l:t:h:"; while ((c = getopt(argc, argv, optstr)) != EOF) switch (c) { case 'n': thread_num = atoi(optarg); break; case 'l': loop = atoi(optarg); break; case 't': delay = atoi(optarg); break; case 'h': default: usage(); exit(1); } printf("thread=%d,loop=%d,delay=%d\n",thread_num,loop,delay); count = malloc(sizeof(long)); shutdown = malloc(sizeof(int)); *count = 0; *shutdown = 0; sigemptyset(&sigset); sigaddset(&sigset, signum); sigprocmask (SIG_BLOCK, &sigset, NULL); signal(SIGINT, sighand); signal(SIGTERM, sighand); for(i = 0; i < thread_num ; i++) pthread_create(&pt[i], NULL, simple_loop, NULL); for (i = 0; i < thread_num; i++) pthread_join(pt[i], NULL); exit(0); } Get powertop V2 from git://github.com/fenrus75/powertop, build powertop. After build the above test application, then run it. Test plaform can be Intel Sandybridge or other recent platforms. #./idle_predict -l 10 & #./powertop We will find that deep C-state will dangle between 40%~100% and much time spent on C1 state. It is because menu governor wrongly predict that repeat mode is kept, so it will choose the C1 shallow C-state even though it has chance to sleep 1 second in deep C-state. While after patched the kernel, we find that deep C-state will keep >99.6%. Signed-off-by: Rik van Riel Signed-off-by: Youquan Song Signed-off-by: Rafael J. Wysocki --- drivers/cpuidle/governors/menu.c | 75 +++++++++++++++++++++++++++++++++++++--- include/linux/tick.h | 6 ++++ kernel/time/tick-sched.c | 4 +++ 3 files changed, 80 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c index 5b1f2c372c1..37c0ff6c805 100644 --- a/drivers/cpuidle/governors/menu.c +++ b/drivers/cpuidle/governors/menu.c @@ -28,6 +28,13 @@ #define MAX_INTERESTING 50000 #define STDDEV_THRESH 400 +/* 60 * 60 > STDDEV_THRESH * INTERVALS = 400 * 8 */ +#define MAX_DEVIATION 60 + +static DEFINE_PER_CPU(struct hrtimer, menu_hrtimer); +static DEFINE_PER_CPU(int, hrtimer_status); +/* menu hrtimer mode */ +enum {MENU_HRTIMER_STOP, MENU_HRTIMER_REPEAT}; /* * Concepts and ideas behind the menu governor @@ -191,17 +198,42 @@ static u64 div_round64(u64 dividend, u32 divisor) return div_u64(dividend + (divisor / 2), divisor); } +/* Cancel the hrtimer if it is not triggered yet */ +void menu_hrtimer_cancel(void) +{ + int cpu = smp_processor_id(); + struct hrtimer *hrtmr = &per_cpu(menu_hrtimer, cpu); + + /* The timer is still not time out*/ + if (per_cpu(hrtimer_status, cpu)) { + hrtimer_cancel(hrtmr); + per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_STOP; + } +} +EXPORT_SYMBOL_GPL(menu_hrtimer_cancel); + +/* Call back for hrtimer is triggered */ +static enum hrtimer_restart menu_hrtimer_notify(struct hrtimer *hrtimer) +{ + int cpu = smp_processor_id(); + + per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_STOP; + + return HRTIMER_NORESTART; +} + /* * Try detecting repeating patterns by keeping track of the last 8 * intervals, and checking if the standard deviation of that set * of points is below a threshold. If it is... then use the * average of these 8 points as the estimated value. */ -static void detect_repeating_patterns(struct menu_device *data) +static int detect_repeating_patterns(struct menu_device *data) { int i; uint64_t avg = 0; uint64_t stddev = 0; /* contains the square of the std deviation */ + int ret = 0; /* first calculate average and standard deviation of the past */ for (i = 0; i < INTERVALS; i++) @@ -210,7 +242,7 @@ static void detect_repeating_patterns(struct menu_device *data) /* if the avg is beyond the known next tick, it's worthless */ if (avg > data->expected_us) - return; + return 0; for (i = 0; i < INTERVALS; i++) stddev += (data->intervals[i] - avg) * @@ -223,8 +255,12 @@ static void detect_repeating_patterns(struct menu_device *data) * repeating pattern and predict we keep doing this. */ - if (avg && stddev < STDDEV_THRESH) + if (avg && stddev < STDDEV_THRESH) { data->predicted_us = avg; + ret = 1; + } + + return ret; } /** @@ -240,6 +276,9 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev) int i; int multiplier; struct timespec t; + int repeat = 0, low_predicted = 0; + int cpu = smp_processor_id(); + struct hrtimer *hrtmr = &per_cpu(menu_hrtimer, cpu); if (data->needs_update) { menu_update(drv, dev); @@ -274,7 +313,7 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev) data->predicted_us = div_round64(data->expected_us * data->correction_factor[data->bucket], RESOLUTION * DECAY); - detect_repeating_patterns(data); + repeat = detect_repeating_patterns(data); /* * We want to default to C1 (hlt), not to busy polling @@ -295,8 +334,10 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev) if (s->disabled || su->disable) continue; - if (s->target_residency > data->predicted_us) + if (s->target_residency > data->predicted_us) { + low_predicted = 1; continue; + } if (s->exit_latency > latency_req) continue; if (s->exit_latency * multiplier > data->predicted_us) @@ -309,6 +350,27 @@ static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev) } } + /* not deepest C-state chosen for low predicted residency */ + if (low_predicted) { + unsigned int timer_us = 0; + + /* + * Set a timer to detect whether this sleep is much + * longer than repeat mode predicted. If the timer + * triggers, the code will evaluate whether to put + * the CPU into a deeper C-state. + * The timer is cancelled on CPU wakeup. + */ + timer_us = 2 * (data->predicted_us + MAX_DEVIATION); + + if (repeat && (4 * timer_us < data->expected_us)) { + hrtimer_start(hrtmr, ns_to_ktime(1000 * timer_us), + HRTIMER_MODE_REL_PINNED); + /* In repeat case, menu hrtimer is started */ + per_cpu(hrtimer_status, cpu) = MENU_HRTIMER_REPEAT; + } + } + return data->last_state_idx; } @@ -399,6 +461,9 @@ static int menu_enable_device(struct cpuidle_driver *drv, struct cpuidle_device *dev) { struct menu_device *data = &per_cpu(menu_devices, dev->cpu); + struct hrtimer *t = &per_cpu(menu_hrtimer, dev->cpu); + hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + t->function = menu_hrtimer_notify; memset(data, 0, sizeof(struct menu_device)); diff --git a/include/linux/tick.h b/include/linux/tick.h index f37fceb69b7..1a6567b4849 100644 --- a/include/linux/tick.h +++ b/include/linux/tick.h @@ -142,4 +142,10 @@ static inline u64 get_cpu_idle_time_us(int cpu, u64 *unused) { return -1; } static inline u64 get_cpu_iowait_time_us(int cpu, u64 *unused) { return -1; } # endif /* !NO_HZ */ +# ifdef CONFIG_CPU_IDLE_GOV_MENU +extern void menu_hrtimer_cancel(void); +# else +static inline void menu_hrtimer_cancel(void) {} +# endif /* CONFIG_CPU_IDLE_GOV_MENU */ + #endif diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index a4026088526..6f337068dc4 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -526,6 +526,8 @@ void tick_nohz_irq_exit(void) if (!ts->inidle) return; + /* Cancel the timer because CPU already waken up from the C-states*/ + menu_hrtimer_cancel(); __tick_nohz_idle_enter(ts); } @@ -621,6 +623,8 @@ void tick_nohz_idle_exit(void) ts->inidle = 0; + /* Cancel the timer because CPU already waken up from the C-states*/ + menu_hrtimer_cancel(); if (ts->idle_active || ts->tick_stopped) now = ktime_get(); -- cgit v1.2.3-70-g09d2 From 883ee4f79d636d13f24c2479c66d42cb652b0239 Mon Sep 17 00:00:00 2001 From: Daniel Walter Date: Tue, 23 Oct 2012 01:20:35 +0200 Subject: PM / sysfs: replace strict_str* with kstrto* Replace strict_strtoul() with kstrtoul() in pm_async_store() and pm_qos_power_write(). [rjw: Modified subject and changelog.] Signed-off-by: Daniel Walter Acked-by: Pavel Machek Signed-off-by: Rafael J. Wysocki --- kernel/power/main.c | 2 +- kernel/power/qos.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/power/main.c b/kernel/power/main.c index f458238109c..1c16f9167de 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -59,7 +59,7 @@ static ssize_t pm_async_store(struct kobject *kobj, struct kobj_attribute *attr, { unsigned long val; - if (strict_strtoul(buf, 10, &val)) + if (kstrtoul(buf, 10, &val)) return -EINVAL; if (val > 1) diff --git a/kernel/power/qos.c b/kernel/power/qos.c index 846bd42c7ed..4da05cee81b 100644 --- a/kernel/power/qos.c +++ b/kernel/power/qos.c @@ -500,7 +500,7 @@ static ssize_t pm_qos_power_write(struct file *filp, const char __user *buf, } else { ascii_value[count] = '\0'; } - ret = strict_strtoul(ascii_value, 16, &ulval); + ret = kstrtoul(ascii_value, 16, &ulval); if (ret) { pr_debug("%s, 0x%lx, 0x%x\n", ascii_value, ulval, ret); return -EINVAL; -- cgit v1.2.3-70-g09d2 From 8316bd72c0248adbb9572abf2dd045a95f682bcd Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Tue, 23 Oct 2012 01:21:09 +0200 Subject: PM / Hibernate: use rb_entry Since the software suspend extents are organized in an rbtree, use rb_entry instead of container_of, as it is semantically more appropriate in order to get a node as it is iterated. Signed-off-by: Davidlohr Bueso Signed-off-by: Rafael J. Wysocki --- kernel/power/swap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/power/swap.c b/kernel/power/swap.c index 3c9d764eb0d..7c33ed20041 100644 --- a/kernel/power/swap.c +++ b/kernel/power/swap.c @@ -126,7 +126,7 @@ static int swsusp_extents_insert(unsigned long swap_offset) /* Figure out where to put the new node */ while (*new) { - ext = container_of(*new, struct swsusp_extent, node); + ext = rb_entry(*new, struct swsusp_extent, node); parent = *new; if (swap_offset < ext->start) { /* Try to merge */ -- cgit v1.2.3-70-g09d2