sched: fix hotplug cpus on ia64

Cliff Wickman wrote:

> I built an ia64 kernel from Andrew's tree (2.6.26-rc2-mm1)
> and get a very predictable hotplug cpu problem.
> billberry1:/tmp/cpw # ./dis
> disabled cpu 17
> enabled cpu 17
> billberry1:/tmp/cpw # ./dis
> disabled cpu 17
> enabled cpu 17
> billberry1:/tmp/cpw # ./dis
>
> The script that disables the cpu always hangs (unkillable)
> on the 3rd attempt.
>
> And a bit further:
> The kstopmachine thread always sits on the run queue (real time) for about
> 30 minutes before running.

this fix solves some (but not all) issues between CPU hotplug and
RT bandwidth throttling.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This commit is contained in:
Peter Zijlstra 2008-06-05 14:49:58 +02:00 коммит произвёл Ingo Molnar
Родитель e958b36004
Коммит 7def2be1dc
2 изменённых файлов: 115 добавлений и 9 удалений

Просмотреть файл

@ -7513,21 +7513,28 @@ int sched_create_sysfs_power_savings_entries(struct sysdev_class *cls)
static int update_sched_domains(struct notifier_block *nfb, static int update_sched_domains(struct notifier_block *nfb,
unsigned long action, void *hcpu) unsigned long action, void *hcpu)
{ {
int cpu = (int)(long)hcpu;
switch (action) { switch (action) {
case CPU_UP_PREPARE:
case CPU_UP_PREPARE_FROZEN:
case CPU_DOWN_PREPARE: case CPU_DOWN_PREPARE:
case CPU_DOWN_PREPARE_FROZEN: case CPU_DOWN_PREPARE_FROZEN:
disable_runtime(cpu_rq(cpu));
/* fall-through */
case CPU_UP_PREPARE:
case CPU_UP_PREPARE_FROZEN:
detach_destroy_domains(&cpu_online_map); detach_destroy_domains(&cpu_online_map);
free_sched_domains(); free_sched_domains();
return NOTIFY_OK; return NOTIFY_OK;
case CPU_UP_CANCELED:
case CPU_UP_CANCELED_FROZEN:
case CPU_DOWN_FAILED: case CPU_DOWN_FAILED:
case CPU_DOWN_FAILED_FROZEN: case CPU_DOWN_FAILED_FROZEN:
case CPU_ONLINE: case CPU_ONLINE:
case CPU_ONLINE_FROZEN: case CPU_ONLINE_FROZEN:
enable_runtime(cpu_rq(cpu));
/* fall-through */
case CPU_UP_CANCELED:
case CPU_UP_CANCELED_FROZEN:
case CPU_DEAD: case CPU_DEAD:
case CPU_DEAD_FROZEN: case CPU_DEAD_FROZEN:
/* /*

Просмотреть файл

@ -286,6 +286,9 @@ static int balance_runtime(struct rt_rq *rt_rq)
continue; continue;
spin_lock(&iter->rt_runtime_lock); spin_lock(&iter->rt_runtime_lock);
if (iter->rt_runtime == RUNTIME_INF)
goto next;
diff = iter->rt_runtime - iter->rt_time; diff = iter->rt_runtime - iter->rt_time;
if (diff > 0) { if (diff > 0) {
do_div(diff, weight); do_div(diff, weight);
@ -299,12 +302,105 @@ static int balance_runtime(struct rt_rq *rt_rq)
break; break;
} }
} }
next:
spin_unlock(&iter->rt_runtime_lock); spin_unlock(&iter->rt_runtime_lock);
} }
spin_unlock(&rt_b->rt_runtime_lock); spin_unlock(&rt_b->rt_runtime_lock);
return more; return more;
} }
static void __disable_runtime(struct rq *rq)
{
struct root_domain *rd = rq->rd;
struct rt_rq *rt_rq;
if (unlikely(!scheduler_running))
return;
for_each_leaf_rt_rq(rt_rq, rq) {
struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
s64 want;
int i;
spin_lock(&rt_b->rt_runtime_lock);
spin_lock(&rt_rq->rt_runtime_lock);
if (rt_rq->rt_runtime == RUNTIME_INF ||
rt_rq->rt_runtime == rt_b->rt_runtime)
goto balanced;
spin_unlock(&rt_rq->rt_runtime_lock);
want = rt_b->rt_runtime - rt_rq->rt_runtime;
for_each_cpu_mask(i, rd->span) {
struct rt_rq *iter = sched_rt_period_rt_rq(rt_b, i);
s64 diff;
if (iter == rt_rq)
continue;
spin_lock(&iter->rt_runtime_lock);
if (want > 0) {
diff = min_t(s64, iter->rt_runtime, want);
iter->rt_runtime -= diff;
want -= diff;
} else {
iter->rt_runtime -= want;
want -= want;
}
spin_unlock(&iter->rt_runtime_lock);
if (!want)
break;
}
spin_lock(&rt_rq->rt_runtime_lock);
BUG_ON(want);
balanced:
rt_rq->rt_runtime = RUNTIME_INF;
spin_unlock(&rt_rq->rt_runtime_lock);
spin_unlock(&rt_b->rt_runtime_lock);
}
}
static void disable_runtime(struct rq *rq)
{
unsigned long flags;
spin_lock_irqsave(&rq->lock, flags);
__disable_runtime(rq);
spin_unlock_irqrestore(&rq->lock, flags);
}
static void __enable_runtime(struct rq *rq)
{
struct root_domain *rd = rq->rd;
struct rt_rq *rt_rq;
if (unlikely(!scheduler_running))
return;
for_each_leaf_rt_rq(rt_rq, rq) {
struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
spin_lock(&rt_b->rt_runtime_lock);
spin_lock(&rt_rq->rt_runtime_lock);
rt_rq->rt_runtime = rt_b->rt_runtime;
rt_rq->rt_time = 0;
spin_unlock(&rt_rq->rt_runtime_lock);
spin_unlock(&rt_b->rt_runtime_lock);
}
}
static void enable_runtime(struct rq *rq)
{
unsigned long flags;
spin_lock_irqsave(&rq->lock, flags);
__enable_runtime(rq);
spin_unlock_irqrestore(&rq->lock, flags);
}
#endif #endif
static inline int rt_se_prio(struct sched_rt_entity *rt_se) static inline int rt_se_prio(struct sched_rt_entity *rt_se)
@ -334,14 +430,13 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
#ifdef CONFIG_SMP #ifdef CONFIG_SMP
if (rt_rq->rt_time > runtime) { if (rt_rq->rt_time > runtime) {
int more;
spin_unlock(&rt_rq->rt_runtime_lock); spin_unlock(&rt_rq->rt_runtime_lock);
more = balance_runtime(rt_rq); balance_runtime(rt_rq);
spin_lock(&rt_rq->rt_runtime_lock); spin_lock(&rt_rq->rt_runtime_lock);
if (more) runtime = sched_rt_runtime(rt_rq);
runtime = sched_rt_runtime(rt_rq); if (runtime == RUNTIME_INF)
return 0;
} }
#endif #endif
@ -1174,6 +1269,8 @@ static void rq_online_rt(struct rq *rq)
if (rq->rt.overloaded) if (rq->rt.overloaded)
rt_set_overload(rq); rt_set_overload(rq);
__enable_runtime(rq);
cpupri_set(&rq->rd->cpupri, rq->cpu, rq->rt.highest_prio); cpupri_set(&rq->rd->cpupri, rq->cpu, rq->rt.highest_prio);
} }
@ -1183,6 +1280,8 @@ static void rq_offline_rt(struct rq *rq)
if (rq->rt.overloaded) if (rq->rt.overloaded)
rt_clear_overload(rq); rt_clear_overload(rq);
__disable_runtime(rq);
cpupri_set(&rq->rd->cpupri, rq->cpu, CPUPRI_INVALID); cpupri_set(&rq->rd->cpupri, rq->cpu, CPUPRI_INVALID);
} }