From f711f6090a81cbd396b63de90f415d33f563af9b Mon Sep 17 00:00:00 2001 From: Gautham R Shenoy Date: Tue, 14 Apr 2009 10:25:30 +0530 Subject: sched: Nominate idle load balancer from a semi-idle package. Currently the nomination of idle-load balancer is done by choosing the first idle cpu in the nohz.cpu_mask. This may not be power-efficient, since such an idle cpu could come from a completely idle core/package thereby preventing the whole core/package from being in a low-power state. For eg, consider a quad-core dual package system. The cpu numbering need not be sequential and can something like [0, 2, 4, 6] and [1, 3, 5, 7]. With sched_mc/smt_power_savings and the power-aware IRQ balance, we try to keep as fewer Packages/Cores active. But the current idle load balancer logic goes against this by choosing the first_cpu in the nohz.cpu_mask and not taking the system topology into consideration. Improve the algorithm to nominate the idle load balancer from a semi idle cores/packages thereby increasing the probability of the cores/packages being in deeper sleep states for longer duration. The algorithm is activated only when sched_mc/smt_power_savings != 0. Signed-off-by: Gautham R Shenoy Acked-by: Peter Zijlstra LKML-Reference: <20090414045530.7645.12175.stgit@sofia.in.ibm.com> Signed-off-by: Ingo Molnar --- kernel/sched.c | 127 +++++++++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 118 insertions(+), 9 deletions(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 5724508c3b66..b0fefa300b40 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4240,10 +4240,126 @@ static void active_load_balance(struct rq *busiest_rq, int busiest_cpu) static struct { atomic_t load_balancer; cpumask_var_t cpu_mask; + cpumask_var_t ilb_grp_nohz_mask; } nohz ____cacheline_aligned = { .load_balancer = ATOMIC_INIT(-1), }; +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) +/** + * lowest_flag_domain - Return lowest sched_domain containing flag. + * @cpu: The cpu whose lowest level of sched domain is to + * be returned. + * @flag: The flag to check for the lowest sched_domain + * for the given cpu. + * + * Returns the lowest sched_domain of a cpu which contains the given flag. + */ +static inline struct sched_domain *lowest_flag_domain(int cpu, int flag) +{ + struct sched_domain *sd; + + for_each_domain(cpu, sd) + if (sd && (sd->flags & flag)) + break; + + return sd; +} + +/** + * for_each_flag_domain - Iterates over sched_domains containing the flag. + * @cpu: The cpu whose domains we're iterating over. + * @sd: variable holding the value of the power_savings_sd + * for cpu. + * @flag: The flag to filter the sched_domains to be iterated. + * + * Iterates over all the scheduler domains for a given cpu that has the 'flag' + * set, starting from the lowest sched_domain to the highest. + */ +#define for_each_flag_domain(cpu, sd, flag) \ + for (sd = lowest_flag_domain(cpu, flag); \ + (sd && (sd->flags & flag)); sd = sd->parent) + +/** + * is_semi_idle_group - Checks if the given sched_group is semi-idle. + * @ilb_group: group to be checked for semi-idleness + * + * Returns: 1 if the group is semi-idle. 0 otherwise. + * + * We define a sched_group to be semi idle if it has atleast one idle-CPU + * and atleast one non-idle CPU. This helper function checks if the given + * sched_group is semi-idle or not. + */ +static inline int is_semi_idle_group(struct sched_group *ilb_group) +{ + cpumask_and(nohz.ilb_grp_nohz_mask, nohz.cpu_mask, + sched_group_cpus(ilb_group)); + + /* + * A sched_group is semi-idle when it has atleast one busy cpu + * and atleast one idle cpu. + */ + if (cpumask_empty(nohz.ilb_grp_nohz_mask)) + return 0; + + if (cpumask_equal(nohz.ilb_grp_nohz_mask, sched_group_cpus(ilb_group))) + return 0; + + return 1; +} +/** + * find_new_ilb - Finds the optimum idle load balancer for nomination. + * @cpu: The cpu which is nominating a new idle_load_balancer. + * + * Returns: Returns the id of the idle load balancer if it exists, + * Else, returns >= nr_cpu_ids. + * + * This algorithm picks the idle load balancer such that it belongs to a + * semi-idle powersavings sched_domain. The idea is to try and avoid + * completely idle packages/cores just for the purpose of idle load balancing + * when there are other idle cpu's which are better suited for that job. + */ +static int find_new_ilb(int cpu) +{ + struct sched_domain *sd; + struct sched_group *ilb_group; + + /* + * Have idle load balancer selection from semi-idle packages only + * when power-aware load balancing is enabled + */ + if (!(sched_smt_power_savings || sched_mc_power_savings)) + goto out_done; + + /* + * Optimize for the case when we have no idle CPUs or only one + * idle CPU. Don't walk the sched_domain hierarchy in such cases + */ + if (cpumask_weight(nohz.cpu_mask) < 2) + goto out_done; + + for_each_flag_domain(cpu, sd, SD_POWERSAVINGS_BALANCE) { + ilb_group = sd->groups; + + do { + if (is_semi_idle_group(ilb_group)) + return cpumask_first(nohz.ilb_grp_nohz_mask); + + ilb_group = ilb_group->next; + + } while (ilb_group != sd->groups); + } + +out_done: + return cpumask_first(nohz.cpu_mask); +} +#else /* (CONFIG_SCHED_MC || CONFIG_SCHED_SMT) */ +static inline int find_new_ilb(int call_cpu) +{ + return first_cpu(nohz.cpu_mask); +} +#endif + /* * This routine will try to nominate the ilb (idle load balancing) * owner among the cpus whose ticks are stopped. ilb owner will do the idle @@ -4468,15 +4584,7 @@ static inline void trigger_load_balance(struct rq *rq, int cpu) } if (atomic_read(&nohz.load_balancer) == -1) { - /* - * simple selection for now: Nominate the - * first cpu in the nohz list to be the next - * ilb owner. - * - * TBD: Traverse the sched domains and nominate - * the nearest cpu in the nohz.cpu_mask. - */ - int ilb = cpumask_first(nohz.cpu_mask); + int ilb = find_new_ilb(cpu); if (ilb < nr_cpu_ids) resched_cpu(ilb); @@ -9051,6 +9159,7 @@ void __init sched_init(void) #ifdef CONFIG_SMP #ifdef CONFIG_NO_HZ alloc_bootmem_cpumask_var(&nohz.cpu_mask); + alloc_bootmem_cpumask_var(&nohz.ilb_grp_nohz_mask); #endif alloc_bootmem_cpumask_var(&cpu_isolated_map); #endif /* SMP */ -- cgit v1.2.3 From e790fb0ba64bfec158e1219d899cb588275d12ab Mon Sep 17 00:00:00 2001 From: Gautham R Shenoy Date: Tue, 14 Apr 2009 10:25:35 +0530 Subject: sched: Nominate a power-efficient ilb in select_nohz_balancer() The CPU that first goes idle becomes the idle-load-balancer and remains that until either it picks up a task or till all the CPUs of the system goes idle. Optimize this further to allow it to relinquish it's post once all it's siblings in the power-aware sched_domain go idle, thereby allowing the whole package-core to go idle. While relinquising the post, nominate another an idle-load balancer from a semi-idle core/package. Signed-off-by: Gautham R Shenoy Acked-by: Peter Zijlstra LKML-Reference: <20090414045535.7645.31641.stgit@sofia.in.ibm.com> Signed-off-by: Ingo Molnar --- kernel/sched.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index b0fefa300b40..36d213bca473 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4414,8 +4414,24 @@ int select_nohz_load_balancer(int stop_tick) /* make me the ilb owner */ if (atomic_cmpxchg(&nohz.load_balancer, -1, cpu) == -1) return 1; - } else if (atomic_read(&nohz.load_balancer) == cpu) + } else if (atomic_read(&nohz.load_balancer) == cpu) { + int new_ilb; + + if (!(sched_smt_power_savings || + sched_mc_power_savings)) + return 1; + /* + * Check to see if there is a more power-efficient + * ilb. + */ + new_ilb = find_new_ilb(cpu); + if (new_ilb < nr_cpu_ids && new_ilb != cpu) { + atomic_set(&nohz.load_balancer, -1); + resched_cpu(new_ilb); + return 0; + } return 1; + } } else { if (!cpumask_test_cpu(cpu, nohz.cpu_mask)) return 0; -- cgit v1.2.3 From 78ddb08feb7d4fbe3c0a9931804c51ee58be4023 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Tue, 14 Apr 2009 16:53:05 +0200 Subject: wait: don't use __wake_up_common() '777c6c5 wait: prevent exclusive waiter starvation' made __wake_up_common() global to be used from abort_exclusive_wait(). It was needed to do a wake-up with the waitqueue lock held while passing down a key to the wake-up function. Since '4ede816 epoll keyed wakeups: add __wake_up_locked_key() and __wake_up_sync_key()' there is an appropriate wrapper for this case: __wake_up_locked_key(). Use it here and make __wake_up_common() private to the scheduler again. Signed-off-by: Johannes Weiner Cc: Andrew Morton Cc: Peter Zijlstra LKML-Reference: <1239720785-19661-1-git-send-email-hannes@cmpxchg.org> Signed-off-by: Ingo Molnar --- include/linux/wait.h | 2 -- kernel/sched.c | 2 +- kernel/wait.c | 2 +- 3 files changed, 2 insertions(+), 4 deletions(-) (limited to 'kernel/sched.c') diff --git a/include/linux/wait.h b/include/linux/wait.h index 5d631c17eaee..67d4e89b62d6 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -132,8 +132,6 @@ static inline void __remove_wait_queue(wait_queue_head_t *head, list_del(&old->task_list); } -void __wake_up_common(wait_queue_head_t *q, unsigned int mode, - int nr_exclusive, int sync, void *key); void __wake_up(wait_queue_head_t *q, unsigned int mode, int nr, void *key); void __wake_up_locked_key(wait_queue_head_t *q, unsigned int mode, void *key); void __wake_up_sync_key(wait_queue_head_t *q, unsigned int mode, int nr, diff --git a/kernel/sched.c b/kernel/sched.c index 36d213bca473..92b4b56ad093 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5345,7 +5345,7 @@ EXPORT_SYMBOL(default_wake_function); * started to run but is not in state TASK_RUNNING. try_to_wake_up() returns * zero in this (rare) case, and we handle it by continuing to scan the queue. */ -void __wake_up_common(wait_queue_head_t *q, unsigned int mode, +static void __wake_up_common(wait_queue_head_t *q, unsigned int mode, int nr_exclusive, int sync, void *key) { wait_queue_t *curr, *next; diff --git a/kernel/wait.c b/kernel/wait.c index 42a2dbc181c8..ea7c3b4275cf 100644 --- a/kernel/wait.c +++ b/kernel/wait.c @@ -154,7 +154,7 @@ void abort_exclusive_wait(wait_queue_head_t *q, wait_queue_t *wait, if (!list_empty(&wait->task_list)) list_del_init(&wait->task_list); else if (waitqueue_active(q)) - __wake_up_common(q, mode, 1, 0, key); + __wake_up_locked_key(q, mode, key); spin_unlock_irqrestore(&q->lock, flags); } EXPORT_SYMBOL(abort_exclusive_wait); -- cgit v1.2.3 From 13318a7186d8e0ae08c996ea4111a945e7789772 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 15 Apr 2009 09:59:10 +0800 Subject: sched: use group_first_cpu() instead of cpumask_first(sched_group_cpus()) Impact: cleanup This patch changes cpumask_first(sched_group_cpus()) to group_first_cpu() for maintainability. Signed-off-by: Miao Xie Acked-by: Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/sched.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 92b4b56ad093..7601ceebf7ce 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -7995,7 +7995,7 @@ static void init_numa_sched_groups_power(struct sched_group *group_head) struct sched_domain *sd; sd = &per_cpu(phys_domains, j).sd; - if (j != cpumask_first(sched_group_cpus(sd->groups))) { + if (j != group_first_cpu(sd->groups)) { /* * Only add "power" once for each * physical package. @@ -8073,7 +8073,7 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd) WARN_ON(!sd || !sd->groups); - if (cpu != cpumask_first(sched_group_cpus(sd->groups))) + if (cpu != group_first_cpu(sd->groups)) return; child = sd->child; -- cgit v1.2.3 From ff743345bf7685a207868048a70e23164c4785e5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 13 Mar 2009 12:21:26 +0100 Subject: sched: remove extra call overhead for schedule() Lai Jiangshan's patch reminded me that I promised Nick to remove that extra call overhead in schedule(). Signed-off-by: Peter Zijlstra LKML-Reference: <20090313112300.927414207@chello.nl> Signed-off-by: Ingo Molnar --- kernel/mutex.c | 4 +++- kernel/sched.c | 12 ++++-------- 2 files changed, 7 insertions(+), 9 deletions(-) (limited to 'kernel/sched.c') diff --git a/kernel/mutex.c b/kernel/mutex.c index 5d79781394a3..e1fb73510409 100644 --- a/kernel/mutex.c +++ b/kernel/mutex.c @@ -248,7 +248,9 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass, /* didnt get the lock, go to sleep: */ spin_unlock_mutex(&lock->wait_lock, flags); - __schedule(); + preempt_enable_no_resched(); + schedule(); + preempt_disable(); spin_lock_mutex(&lock->wait_lock, flags); } diff --git a/kernel/sched.c b/kernel/sched.c index 7601ceebf7ce..797f6fdabadf 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5131,13 +5131,15 @@ pick_next_task(struct rq *rq) /* * schedule() is the main scheduler function. */ -asmlinkage void __sched __schedule(void) +asmlinkage void __sched schedule(void) { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; +need_resched: + preempt_disable(); cpu = smp_processor_id(); rq = cpu_rq(cpu); rcu_qsctr_inc(cpu); @@ -5194,15 +5196,9 @@ need_resched_nonpreemptible: if (unlikely(reacquire_kernel_lock(current) < 0)) goto need_resched_nonpreemptible; -} -asmlinkage void __sched schedule(void) -{ -need_resched: - preempt_disable(); - __schedule(); preempt_enable_no_resched(); - if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) + if (need_resched()) goto need_resched; } EXPORT_SYMBOL(schedule); -- cgit v1.2.3 From 6e29ec5701e9d44fa02b96c1c5c45f7516182b65 Mon Sep 17 00:00:00 2001 From: Gautham R Shenoy Date: Tue, 21 Apr 2009 08:40:49 +0530 Subject: sched: Replace first_cpu() with cpumask_first() in ILB nomination code Stephen Rothwell reported this build warning: > kernel/sched.c: In function 'find_new_ilb': > kernel/sched.c:4355: warning: passing argument 1 of '__first_cpu' from incompatible pointer type > > Possibly caused by commit f711f6090a81cbd396b63de90f415d33f563af9b > ("sched: Nominate idle load balancer from a semi-idle package") from > the sched tree. Should this call to first_cpu be cpumask_first? For !(CONFIG_SCHED_MC || CONFIG_SCHED_SMT), find_new_ilb() nominates the Idle load balancer as the first cpu from the nohz.cpu_mask. This code uses the older API first_cpu(). Replace it with cpumask_first(), which is the correct API here. [ Impact: cleanup, address build warning ] Reported-by: Stephen Rothwell Signed-off-by: Gautham R Shenoy Cc: Rusty Russell LKML-Reference: <20090421031049.GA4140@in.ibm.com> Signed-off-by: Ingo Molnar --- kernel/sched.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 797f6fdabadf..54d67b94f1a9 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4356,7 +4356,7 @@ out_done: #else /* (CONFIG_SCHED_MC || CONFIG_SCHED_SMT) */ static inline int find_new_ilb(int call_cpu) { - return first_cpu(nohz.cpu_mask); + return cpumask_first(nohz.cpu_mask); } #endif -- cgit v1.2.3 From 60aa605dfce2976e54fa76e805ab0f221372d4d9 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 5 May 2009 17:50:21 +0200 Subject: sched: rt: document the risk of small values in the bandwidth settings Thomas noted that we should disallow sysctl_sched_rt_runtime == 0 for (!RT_GROUP) since the root group always has some RT tasks in it. Further, update the documentation to inspire clue. [ Impact: exclude corner-case sysctl_sched_rt_runtime value ] Reported-by: Thomas Gleixner Signed-off-by: Peter Zijlstra LKML-Reference: <20090505155436.863098054@chello.nl> Signed-off-by: Ingo Molnar --- Documentation/scheduler/sched-rt-group.txt | 18 ++++++++++++++++++ kernel/sched.c | 7 +++++++ 2 files changed, 25 insertions(+) (limited to 'kernel/sched.c') diff --git a/Documentation/scheduler/sched-rt-group.txt b/Documentation/scheduler/sched-rt-group.txt index 5ba4d3fc625a..eb74b014a3f8 100644 --- a/Documentation/scheduler/sched-rt-group.txt +++ b/Documentation/scheduler/sched-rt-group.txt @@ -4,6 +4,7 @@ CONTENTS ======== +0. WARNING 1. Overview 1.1 The problem 1.2 The solution @@ -14,6 +15,23 @@ CONTENTS 3. Future plans +0. WARNING +========== + + Fiddling with these settings can result in an unstable system, the knobs are + root only and assumes root knows what he is doing. + +Most notable: + + * very small values in sched_rt_period_us can result in an unstable + system when the period is smaller than either the available hrtimer + resolution, or the time it takes to handle the budget refresh itself. + + * very small values in sched_rt_runtime_us can result in an unstable + system when the runtime is so small the system has difficulty making + forward progress (NOTE: the migration thread and kstopmachine both + are real-time processes). + 1. Overview =========== diff --git a/kernel/sched.c b/kernel/sched.c index 54d67b94f1a9..2a43a581ead3 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -9917,6 +9917,13 @@ static int sched_rt_global_constraints(void) if (sysctl_sched_rt_period <= 0) return -EINVAL; + /* + * There's always some RT tasks in the root group + * -- migration, kstopmachine etc.. + */ + if (sysctl_sched_rt_runtime == 0) + return -EBUSY; + spin_lock_irqsave(&def_rt_bandwidth.rt_runtime_lock, flags); for_each_possible_cpu(i) { struct rt_rq *rt_rq = &cpu_rq(i)->rt; -- cgit v1.2.3 From aa47b7e0f89b9998dad4d1667447e8cb7703ff4e Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Mon, 4 May 2009 01:38:05 -0700 Subject: sched: emit thread info flags with stack trace When a thread is oom killed and fails to exit, it's helpful to know which threads have access to memory reserves if the machine livelocks. This is done by testing for the TIF_MEMDIE thread info flag and should be displayed alongside stack traces to identify tasks that have access to such reserves but are still stuck allocating pages, for instance. It would probably be helpful in other cases as well, so all thread info flags are emitted when showing a task. ( v2: fix warning reported by Stephen Rothwell ) [ Impact: extend debug printout info ] Signed-off-by: David Rientjes Cc: Peter Zijlstra Cc: Stephen Rothwell LKML-Reference: Signed-off-by: Ingo Molnar --- kernel/sched.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel/sched.c') diff --git a/kernel/sched.c b/kernel/sched.c index 2a43a581ead3..5aa63f50c696 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -6610,8 +6610,9 @@ void sched_show_task(struct task_struct *p) #ifdef CONFIG_DEBUG_STACK_USAGE free = stack_not_used(p); #endif - printk(KERN_CONT "%5lu %5d %6d\n", free, - task_pid_nr(p), task_pid_nr(p->real_parent)); + printk(KERN_CONT "%5lu %5d %6d 0x%08lx\n", free, + task_pid_nr(p), task_pid_nr(p->real_parent), + (unsigned long)task_thread_info(p)->flags); show_stack(p, NULL); } -- cgit v1.2.3 From dce48a84adf1806676319f6f480e30a6daa012f9 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 11 Apr 2009 10:43:41 +0200 Subject: sched, timers: move calc_load() to scheduler Dimitri Sivanich noticed that xtime_lock is held write locked across calc_load() which iterates over all online CPUs. That can cause long latencies for xtime_lock readers on large SMP systems. The load average calculation is an rough estimate anyway so there is no real need to protect the readers vs. the update. It's not a problem when the avenrun array is updated while a reader copies the values. Instead of iterating over all online CPUs let the scheduler_tick code update the number of active tasks shortly before the avenrun update happens. The avenrun update itself is handled by the CPU which calls do_timer(). [ Impact: reduce xtime_lock write locked section ] Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra --- include/linux/sched.h | 2 +- kernel/sched.c | 84 +++++++++++++++++++++++++++++++++++++++++------ kernel/sched_idletask.c | 3 +- kernel/time/timekeeping.c | 2 +- kernel/timer.c | 54 ++---------------------------- 5 files changed, 80 insertions(+), 65 deletions(-) (limited to 'kernel/sched.c') diff --git a/include/linux/sched.h b/include/linux/sched.h index b4c38bc8049c..6eb4892efe45 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -135,8 +135,8 @@ DECLARE_PER_CPU(unsigned long, process_counts); extern int nr_processes(void); extern unsigned long nr_running(void); extern unsigned long nr_uninterruptible(void); -extern unsigned long nr_active(void); extern unsigned long nr_iowait(void); +extern void calc_global_load(void); extern unsigned long get_parent_ip(unsigned long addr); diff --git a/kernel/sched.c b/kernel/sched.c index 8908d190a348..f4eb88153bd1 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -630,6 +630,10 @@ struct rq { struct list_head migration_queue; #endif + /* calc_load related fields */ + unsigned long calc_load_update; + long calc_load_active; + #ifdef CONFIG_SCHED_HRTICK #ifdef CONFIG_SMP int hrtick_csd_pending; @@ -1728,6 +1732,8 @@ static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares) } #endif +static void calc_load_account_active(struct rq *this_rq); + #include "sched_stats.h" #include "sched_idletask.c" #include "sched_fair.c" @@ -2856,19 +2862,57 @@ unsigned long nr_iowait(void) return sum; } -unsigned long nr_active(void) +/* Variables and functions for calc_load */ +static atomic_long_t calc_load_tasks; +static unsigned long calc_load_update; +unsigned long avenrun[3]; +EXPORT_SYMBOL(avenrun); + +static unsigned long +calc_load(unsigned long load, unsigned long exp, unsigned long active) { - unsigned long i, running = 0, uninterruptible = 0; + load *= exp; + load += active * (FIXED_1 - exp); + return load >> FSHIFT; +} - for_each_online_cpu(i) { - running += cpu_rq(i)->nr_running; - uninterruptible += cpu_rq(i)->nr_uninterruptible; - } +/* + * calc_load - update the avenrun load estimates 10 ticks after the + * CPUs have updated calc_load_tasks. + */ +void calc_global_load(void) +{ + unsigned long upd = calc_load_update + 10; + long active; + + if (time_before(jiffies, upd)) + return; - if (unlikely((long)uninterruptible < 0)) - uninterruptible = 0; + active = atomic_long_read(&calc_load_tasks); + active = active > 0 ? active * FIXED_1 : 0; - return running + uninterruptible; + avenrun[0] = calc_load(avenrun[0], EXP_1, active); + avenrun[1] = calc_load(avenrun[1], EXP_5, active); + avenrun[2] = calc_load(avenrun[2], EXP_15, active); + + calc_load_update += LOAD_FREQ; +} + +/* + * Either called from update_cpu_load() or from a cpu going idle + */ +static void calc_load_account_active(struct rq *this_rq) +{ + long nr_active, delta; + + nr_active = this_rq->nr_running; + nr_active += (long) this_rq->nr_uninterruptible; + + if (nr_active != this_rq->calc_load_active) { + delta = nr_active - this_rq->calc_load_active; + this_rq->calc_load_active = nr_active; + atomic_long_add(delta, &calc_load_tasks); + } } /* @@ -2899,6 +2943,11 @@ static void update_cpu_load(struct rq *this_rq) new_load += scale-1; this_rq->cpu_load[i] = (old_load*(scale-1) + new_load) >> i; } + + if (time_after_eq(jiffies, this_rq->calc_load_update)) { + this_rq->calc_load_update += LOAD_FREQ; + calc_load_account_active(this_rq); + } } #ifdef CONFIG_SMP @@ -7091,6 +7140,14 @@ static void migrate_dead_tasks(unsigned int dead_cpu) } } + +/* + * remove the tasks which were accounted by rq from calc_load_tasks. + */ +static void calc_global_load_remove(struct rq *rq) +{ + atomic_long_sub(rq->calc_load_active, &calc_load_tasks); +} #endif /* CONFIG_HOTPLUG_CPU */ #if defined(CONFIG_SCHED_DEBUG) && defined(CONFIG_SYSCTL) @@ -7325,6 +7382,8 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) /* Update our root-domain */ rq = cpu_rq(cpu); spin_lock_irqsave(&rq->lock, flags); + rq->calc_load_update = calc_load_update; + rq->calc_load_active = 0; if (rq->rd) { BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); @@ -7364,7 +7423,7 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) cpuset_unlock(); migrate_nr_uninterruptible(rq); BUG_ON(rq->nr_running != 0); - + calc_global_load_remove(rq); /* * No need to migrate the tasks: it was best-effort if * they didn't take sched_hotcpu_mutex. Just wake up @@ -9059,6 +9118,8 @@ void __init sched_init(void) rq = cpu_rq(i); spin_lock_init(&rq->lock); rq->nr_running = 0; + rq->calc_load_active = 0; + rq->calc_load_update = jiffies + LOAD_FREQ; init_cfs_rq(&rq->cfs, rq); init_rt_rq(&rq->rt, rq); #ifdef CONFIG_FAIR_GROUP_SCHED @@ -9166,6 +9227,9 @@ void __init sched_init(void) * when this runqueue becomes "idle". */ init_idle(current, smp_processor_id()); + + calc_load_update = jiffies + LOAD_FREQ; + /* * During early bootup we pretend to be a normal task: */ diff --git a/kernel/sched_idletask.c b/kernel/sched_idletask.c index 8a21a2e28c13..499672c10cbd 100644 --- a/kernel/sched_idletask.c +++ b/kernel/sched_idletask.c @@ -22,7 +22,8 @@ static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p, int sy static struct task_struct *pick_next_task_idle(struct rq *rq) { schedstat_inc(rq, sched_goidle); - + /* adjust the active tasks as we might go into a long sleep */ + calc_load_account_active(rq); return rq->idle; } diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 687dff49f6e7..52a8bf8931f3 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -22,7 +22,7 @@ /* * This read-write spinlock protects us from races in SMP while - * playing with xtime and avenrun. + * playing with xtime. */ __cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); diff --git a/kernel/timer.c b/kernel/timer.c index cffffad01c31..6a21d7af9620 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -1122,47 +1122,6 @@ void update_process_times(int user_tick) run_posix_cpu_timers(p); } -/* - * Nr of active tasks - counted in fixed-point numbers - */ -static unsigned long count_active_tasks(void) -{ - return nr_active() * FIXED_1; -} - -/* - * Hmm.. Changed this, as the GNU make sources (load.c) seems to - * imply that avenrun[] is the standard name for this kind of thing. - * Nothing else seems to be standardized: the fractional size etc - * all seem to differ on different machines. - * - * Requires xtime_lock to access. - */ -unsigned long avenrun[3]; - -EXPORT_SYMBOL(avenrun); - -/* - * calc_load - given tick count, update the avenrun load estimates. - * This is called while holding a write_lock on xtime_lock. - */ -static inline void calc_load(unsigned long ticks) -{ - unsigned long active_tasks; /* fixed-point */ - static int count = LOAD_FREQ; - - count -= ticks; - if (unlikely(count < 0)) { - active_tasks = count_active_tasks(); - do { - CALC_LOAD(avenrun[0], EXP_1, active_tasks); - CALC_LOAD(avenrun[1], EXP_5, active_tasks); - CALC_LOAD(avenrun[2], EXP_15, active_tasks); - count += LOAD_FREQ; - } while (count < 0); - } -} - /* * This function runs timers and the timer-tq in bottom half context. */ @@ -1186,16 +1145,6 @@ void run_local_timers(void) softlockup_tick(); } -/* - * Called by the timer interrupt. xtime_lock must already be taken - * by the timer IRQ! - */ -static inline void update_times(unsigned long ticks) -{ - update_wall_time(); - calc_load(ticks); -} - /* * The 64-bit jiffies value is not atomic - you MUST NOT read it * without sampling the sequence number in xtime_lock. @@ -1205,7 +1154,8 @@ static inline void update_times(unsigned long ticks) void do_timer(unsigned long ticks) { jiffies_64 += ticks; - update_times(ticks); + update_wall_time(); + calc_global_load(); } #ifdef __ARCH_WANT_SYS_ALARM -- cgit v1.2.3 From 2d02494f5a90f2e4b3c4c6acc85ec94674cdc431 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 2 May 2009 20:08:52 +0200 Subject: sched, timers: cleanup avenrun users avenrun is an rough estimate so we don't have to worry about consistency of the three avenrun values. Remove the xtime lock dependency and provide a function to scale the values. Cleanup the users. [ Impact: cleanup ] Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra --- fs/proc/loadavg.c | 18 ++++++------------ include/linux/sched.h | 1 + kernel/sched.c | 15 +++++++++++++++ kernel/timer.c | 32 ++++++-------------------------- 4 files changed, 28 insertions(+), 38 deletions(-) (limited to 'kernel/sched.c') diff --git a/fs/proc/loadavg.c b/fs/proc/loadavg.c index 9bca39cf99ee..1afa4dd4cae2 100644 --- a/fs/proc/loadavg.c +++ b/fs/proc/loadavg.c @@ -12,20 +12,14 @@ static int loadavg_proc_show(struct seq_file *m, void *v) { - int a, b, c; - unsigned long seq; + unsigned long avnrun[3]; - do { - seq = read_seqbegin(&xtime_lock); - a = avenrun[0] + (FIXED_1/200); - b = avenrun[1] + (FIXED_1/200); - c = avenrun[2] + (FIXED_1/200); - } while (read_seqretry(&xtime_lock, seq)); + get_avenrun(avnrun, FIXED_1/200, 0); - seq_printf(m, "%d.%02d %d.%02d %d.%02d %ld/%d %d\n", - LOAD_INT(a), LOAD_FRAC(a), - LOAD_INT(b), LOAD_FRAC(b), - LOAD_INT(c), LOAD_FRAC(c), + seq_printf(m, "%lu.%02lu %lu.%02lu %lu.%02lu %ld/%d %d\n", + LOAD_INT(avnrun[0]), LOAD_FRAC(avnrun[0]), + LOAD_INT(avnrun[1]), LOAD_FRAC(avnrun[1]), + LOAD_INT(avnrun[2]), LOAD_FRAC(avnrun[2]), nr_running(), nr_threads, task_active_pid_ns(current)->last_pid); return 0; diff --git a/include/linux/sched.h b/include/linux/sched.h index 6eb4892efe45..de7b3b217772 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -116,6 +116,7 @@ struct fs_struct; * 11 bit fractions. */ extern unsigned long avenrun[]; /* Load averages */ +extern void get_avenrun(unsigned long *loads, unsigned long offset, int shift); #define FSHIFT 11 /* nr of bits of precision */ #define FIXED_1 (1<= 0) { - tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC; - tp.tv_sec++; - } - info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0); + ktime_get_ts(&tp); + monotonic_to_bootbased(&tp); + info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0); - info->loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT); - info->loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT); - info->loads[2] = avenrun[2] << (SI_LOAD_SHIFT - FSHIFT); + get_avenrun(info->loads, 0, SI_LOAD_SHIFT - FSHIFT); - info->procs = nr_threads; - } while (read_seqretry(&xtime_lock, seq)); + info->procs = nr_threads; si_meminfo(info); si_swapinfo(info); -- cgit v1.2.3 From 4200efd9acda4accf24640f1e77d24fdcdb524df Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Tue, 19 May 2009 09:22:19 +0200 Subject: sched: properly define the sched_group::cpumask and sched_domain::span fields Properly document the variable-size structure tricks we are doing wrt. struct sched_group and sched_domain, and use the field[0] GCC extension instead of defining a vla array. Dont use unions for this, as pointed out by Linus. [ Impact: cleanup, un-confuse Sparse and LLVM ] Reported-by: Jeff Garzik Acked-by: Linus Torvalds LKML-Reference: Signed-off-by: Ingo Molnar --- include/linux/sched.h | 25 ++++++++++++++++++++++--- kernel/sched.c | 5 +++-- 2 files changed, 25 insertions(+), 5 deletions(-) (limited to 'kernel/sched.c') diff --git a/include/linux/sched.h b/include/linux/sched.h index de7b3b217772..dbb1043e8656 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -839,7 +839,17 @@ struct sched_group { */ u32 reciprocal_cpu_power; - unsigned long cpumask[]; + /* + * The CPUs this group covers. + * + * NOTE: this field is variable length. (Allocated dynamically + * by attaching extra space to the end of the structure, + * depending on how many CPUs the kernel has booted up with) + * + * It is also be embedded into static data structures at build + * time. (See 'struct static_sched_group' in kernel/sched.c) + */ + unsigned long cpumask[0]; }; static inline struct cpumask *sched_group_cpus(struct sched_group *sg) @@ -925,8 +935,17 @@ struct sched_domain { char *name; #endif - /* span of all CPUs in this domain */ - unsigned long span[]; + /* + * Span of all CPUs in this domain. + * + * NOTE: this field is variable length. (Allocated dynamically + * by attaching extra space to the end of the structure, + * depending on how many CPUs the kernel has booted up with) + * + * It is also be embedded into static data structures at build + * time. (See 'struct static_sched_domain' in kernel/sched.c) + */ + unsigned long span[0]; }; static inline struct cpumask *sched_domain_span(struct sched_domain *sd) diff --git a/kernel/sched.c b/kernel/sched.c index 497c09ba61e7..228acae8821f 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -7948,8 +7948,9 @@ int sched_smt_power_savings = 0, sched_mc_power_savings = 0; /* * The cpus mask in sched_group and sched_domain hangs off the end. - * FIXME: use cpumask_var_t or dynamic percpu alloc to avoid wasting space - * for nr_cpu_ids < CONFIG_NR_CPUS. + * + * ( See the the comments in include/linux/sched.h:struct sched_group + * and struct sched_domain. ) */ struct static_sched_group { struct sched_group sg; -- cgit v1.2.3