lkml.org 
[lkml]   [2006]   [Nov]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[PATCH 6/8] Use softirq for load balancing
Use softirq to balance sched domains.

Call rebalance_tick (renamed to run_rebalance_domains) from a newly introduced
softirq.

We calculate the earliest time for each layer of sched domains to be
rescanned (this is the rescan time for idle) and use the earliest
of those to schedule the softirq via a new field "next_balance"
added to struct rq.

Signed-off-by: Christoph Lameter <clameter@sgi.com>

Index: linux-2.6.19-rc5-mm1/kernel/sched.c
===================================================================
--- linux-2.6.19-rc5-mm1.orig/kernel/sched.c 2006-11-14 14:02:13.045549034 -0600
+++ linux-2.6.19-rc5-mm1/kernel/sched.c 2006-11-14 14:02:54.254194569 -0600
@@ -228,6 +228,7 @@ struct rq {
unsigned long expired_timestamp;
unsigned long long timestamp_last_tick;
struct task_struct *curr, *idle;
+ unsigned long next_balance;
struct mm_struct *prev_mm;
struct prio_array *active, *expired, arrays[2];
int best_expired_prio;
@@ -2861,7 +2862,7 @@ static void update_load(struct rq *this_
}

/*
- * rebalance_tick will get called every timer tick, on every CPU.
+ * run_rebalance_domains is triggered when needed from the scheduler tick.
*
* It checks each scheduling domain to see if it is due to be balanced,
* and initiates a balancing operation if so.
@@ -2869,9 +2870,10 @@ static void update_load(struct rq *this_
* Balancing parameters are set up in arch_init_sched_domains.
*/

-static void
-rebalance_tick(int this_cpu, struct rq *this_rq)
+static void run_rebalance_domains(struct softirq_action *h)
{
+ int this_cpu = smp_processor_id();
+ struct rq *this_rq = cpu_rq(this_cpu);
unsigned long interval;
struct sched_domain *sd;
/*
@@ -2880,6 +2882,8 @@ rebalance_tick(int this_cpu, struct rq *
*/
enum idle_type idle = !this_rq->nr_running ?
SCHED_IDLE : NOT_IDLE;
+ /* Earliest time when we have to call run_rebalance_domains again */
+ unsigned long next_balance = jiffies + 60*HZ;

for_each_domain(this_cpu, sd) {
if (!(sd->flags & SD_LOAD_BALANCE))
@@ -2894,7 +2898,7 @@ rebalance_tick(int this_cpu, struct rq *
if (unlikely(!interval))
interval = 1;

- if (jiffies - sd->last_balance >= interval) {
+ if (time_after_eq(jiffies, sd->last_balance + interval)) {
if (load_balance(this_cpu, this_rq, sd, idle)) {
/*
* We've pulled tasks over so either we're no
@@ -2905,7 +2909,10 @@ rebalance_tick(int this_cpu, struct rq *
}
sd->last_balance += interval;
}
+ if (time_after(next_balance, sd->last_balance + interval))
+ next_balance = sd->last_balance + interval;
}
+ this_rq->next_balance = next_balance;
}
#else
/*
@@ -3158,7 +3165,8 @@ void scheduler_tick(void)
task_running_tick(rq, p);
#ifdef CONFIG_SMP
update_load(rq);
- rebalance_tick(cpu, rq);
+ if (time_after_eq(jiffies, rq->next_balance))
+ raise_softirq(SCHED_SOFTIRQ);
#endif
}

@@ -6959,6 +6966,10 @@ void __init sched_init(void)

set_load_weight(&init_task);

+#ifdef CONFIG_SMP
+ open_softirq(SCHED_SOFTIRQ, run_rebalance_domains, NULL);
+#endif
+
#ifdef CONFIG_RT_MUTEXES
plist_head_init(&init_task.pi_waiters, &init_task.pi_lock);
#endif
Index: linux-2.6.19-rc5-mm1/include/linux/interrupt.h
===================================================================
--- linux-2.6.19-rc5-mm1.orig/include/linux/interrupt.h 2006-11-14 14:02:13.059221756 -0600
+++ linux-2.6.19-rc5-mm1/include/linux/interrupt.h 2006-11-14 14:02:15.109153475 -0600
@@ -237,6 +237,7 @@ enum
NET_RX_SOFTIRQ,
BLOCK_SOFTIRQ,
TASKLET_SOFTIRQ,
+ SCHED_SOFTIRQ,
#ifdef CONFIG_HIGH_RES_TIMERS
HRTIMER_SOFTIRQ,
#endif
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2006-11-14 21:39    [W:0.777 / U:0.232 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site