bugfix for cgroup steal task
Cheng Yu (2): sched/fair: Fix the condition in overload_clear sched/debug: Fix h_nr_running/steal_h_nr_running in sched_debug
kernel/sched/debug.c | 10 +++++----- kernel/sched/fair.c | 7 +++---- 2 files changed, 8 insertions(+), 9 deletions(-)
hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IAQWPQ
-----------------------------------------
In the overload_clear function, there is a bug when setting the overload_cpus, that is, when the steal_task feature is globally enabled and rq->cfs.h_nr_running is less than 2, the need_clear should be set to true.
Fixes: 433c0b725642 ("sched/fair: Count the number of tasks marked as steal_task on cfs_rq") Signed-off-by: Cheng Yu serein.chengyu@huawei.com --- kernel/sched/fair.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 38841d1d640a..a518d636f07f 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -4492,11 +4492,10 @@ static void overload_clear(struct rq *rq) if (!steal_enabled()) return;
- if (!group_steal_used() && rq->cfs.h_nr_running >= 2) - return; + if (rq->cfs.h_nr_running < 2) + need_clear = true;
- if (group_steal_used() && - (rq->cfs.h_nr_running < 2 || rq->cfs.steal_h_nr_running == 0)) + if (group_steal_used() && rq->cfs.steal_h_nr_running < 1) need_clear = true;
if (!need_clear)
hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IAQWPQ
-----------------------------------------
It would be more practical to put the statistics of h_nr_running and steal_h_nr_running in the per-cpu column.
Fixes: 1f9456c873b7 ("sched/debug: Add h_nr_running/steal_h_nr_running in sched_debug") Signed-off-by: Cheng Yu serein.chengyu@huawei.com --- kernel/sched/debug.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index ff2d6dc59c14..4275398bc713 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -594,11 +594,6 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq) SEQ_printf(m, " .%-30s: %d\n", "nr_spread_over", cfs_rq->nr_spread_over); SEQ_printf(m, " .%-30s: %d\n", "nr_running", cfs_rq->nr_running); -#ifdef CONFIG_SCHED_STEAL - SEQ_printf(m, " .%-30s: %d\n", "h_nr_running", cfs_rq->h_nr_running); - SEQ_printf(m, " .%-30s: %ld\n", "steal_h_nr_running", - cfs_rq->steal_h_nr_running); -#endif SEQ_printf(m, " .%-30s: %ld\n", "load", cfs_rq->load.weight); #ifdef CONFIG_SMP SEQ_printf(m, " .%-30s: %lu\n", "load_avg", @@ -714,6 +709,11 @@ do { \ SEQ_printf(m, " .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rq->x))
P(nr_running); +#ifdef CONFIG_SCHED_STEAL + SEQ_printf(m, " .%-30s: %d\n", "h_nr_running", rq->cfs.h_nr_running); + SEQ_printf(m, " .%-30s: %ld\n", "steal_h_nr_running", + rq->cfs.steal_h_nr_running); +#endif P(nr_switches); P(nr_uninterruptible); PN(next_balance);
反馈： 您发送到kernel@openeuler.org的补丁/补丁集，已成功转换为PR！ PR链接地址： https://gitee.com/openeuler/kernel/pulls/12101 邮件列表地址：https://mailweb.openeuler.org/hyperkitty/list/kernel@openeuler.org/message/Y...
FeedBack: The patch(es) which you have sent to kernel@openeuler.org mailing list has been converted to a pull request successfully! Pull request link: https://gitee.com/openeuler/kernel/pulls/12101 Mailing list address: https://mailweb.openeuler.org/hyperkitty/list/kernel@openeuler.org/message/Y...
-
Cheng Yu -
patchwork bot