Messages in this thread Patch in this message | | | From | "Uladzislau Rezki (Sony)" <> | Subject | [RFC][PATCH]: sched/fair: search a task from the tail of the queue | Date | Thu, 24 Aug 2017 12:44:45 +0200 |
| |
From: Uladzislau Rezki <urezki@gmail.com>
When a task is enqueued back from a physical CPU to the running list it is placed in the beginning of the queue. Thus, the cfs_tasks list is more or less sorted (except woken tasks) starting from recently given CPU time tasks toward tasks with max wait time in a run-queue.
As part of the load balance operation, this approach starts detach_tasks()/detach_one_task() from the tail of the queue instead of the head, giving some advantages:
- tends to pick a task with highest wait time; - tasks located in the tail are less likely cache-hot, therefore the can_migrate_task() decision is higher.
hackbench illustrates slightly better performance. For example doing 1000 samples and 40 groups on i5-3320M CPU, it shows below figures:
default: 0.672 avg / 0.674 median patched: 0.665 avg / 0.667 median
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com> --- kernel/sched/fair.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index c77e4b1d51c0..028f5e8bc830 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6806,11 +6806,12 @@ static void detach_task(struct task_struct *p, struct lb_env *env) */ static struct task_struct *detach_one_task(struct lb_env *env) { - struct task_struct *p, *n; + struct task_struct *p; lockdep_assert_held(&env->src_rq->lock); - list_for_each_entry_safe(p, n, &env->src_rq->cfs_tasks, se.group_node) { + list_for_each_entry_reverse(p, + &env->src_rq->cfs_tasks, se.group_node) { if (!can_migrate_task(p, env)) continue; @@ -6856,7 +6857,7 @@ static int detach_tasks(struct lb_env *env) if (env->idle != CPU_NOT_IDLE && env->src_rq->nr_running <= 1) break; - p = list_first_entry(tasks, struct task_struct, se.group_node); + p = list_last_entry(tasks, struct task_struct, se.group_node); env->loop++; /* We've more or less seen every task there is, call it quits */ @@ -6906,7 +6907,7 @@ static int detach_tasks(struct lb_env *env) continue; next: - list_move_tail(&p->se.group_node, tasks); + list_move(&p->se.group_node, tasks); } /* -- 2.11.0
| |