sched: remove prio preference from balance decisions

Priority looses much of its meaning in a hierarchical context. So don't
use it in balance decisions.

Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Cc: Mike Galbraith <efault@gmx.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This commit is contained in:
Peter Zijlstra 2008-06-27 13:41:31 +02:00 committed by Ingo Molnar
parent 4be9daaa1b
commit 051c67640e

View file

@ -2896,7 +2896,7 @@ balance_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest,
enum cpu_idle_type idle, int *all_pinned, enum cpu_idle_type idle, int *all_pinned,
int *this_best_prio, struct rq_iterator *iterator) int *this_best_prio, struct rq_iterator *iterator)
{ {
int loops = 0, pulled = 0, pinned = 0, skip_for_load; int loops = 0, pulled = 0, pinned = 0;
struct task_struct *p; struct task_struct *p;
long rem_load_move = max_load_move; long rem_load_move = max_load_move;
@ -2912,14 +2912,8 @@ balance_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest,
next: next:
if (!p || loops++ > sysctl_sched_nr_migrate) if (!p || loops++ > sysctl_sched_nr_migrate)
goto out; goto out;
/*
* To help distribute high priority tasks across CPUs we don't if ((p->se.load.weight >> 1) > rem_load_move ||
* skip a task if it will be the highest priority task (i.e. smallest
* prio value) on its new queue regardless of its load weight
*/
skip_for_load = (p->se.load.weight >> 1) > rem_load_move +
SCHED_LOAD_SCALE_FUZZ;
if ((skip_for_load && p->prio >= *this_best_prio) ||
!can_migrate_task(p, busiest, this_cpu, sd, idle, &pinned)) { !can_migrate_task(p, busiest, this_cpu, sd, idle, &pinned)) {
p = iterator->next(iterator->arg); p = iterator->next(iterator->arg);
goto next; goto next;