From: Vineeth Remanan Pillai <vpillai@digitalocean.com>
To: Nishanth Aravamudan <naravamudan@digitalocean.com>,
Julien Desfossez <jdesfossez@digitalocean.com>,
Peter Zijlstra <peterz@infradead.org>,
Tim Chen <tim.c.chen@linux.intel.com>,
mingo@kernel.org, tglx@linutronix.de, pjt@google.com,
torvalds@linux-foundation.org
Cc: linux-kernel@vger.kernel.org, Dario Faggioli <dfaggioli@suse.com>,
fweisbec@gmail.com, keescook@chromium.org, kerrnel@google.com,
Phil Auld <pauld@redhat.com>, Aaron Lu <aaron.lwe@gmail.com>,
Aubrey Li <aubrey.intel@gmail.com>,
Valentin Schneider <valentin.schneider@arm.com>,
Mel Gorman <mgorman@techsingularity.net>,
Pawan Gupta <pawan.kumar.gupta@linux.intel.com>,
Paolo Bonzini <pbonzini@redhat.com>
Subject: [RFC PATCH v4 15/19] sched: Trivial forced-newidle balancer
Date: Wed, 30 Oct 2019 18:33:28 +0000 [thread overview]
Message-ID: <89da8db0c28e165bef0a24bd970186b3ca0821bd.1572437285.git.vpillai@digitalocean.com> (raw)
In-Reply-To: <cover.1572437285.git.vpillai@digitalocean.com>
In-Reply-To: <cover.1572437285.git.vpillai@digitalocean.com>
From: Peter Zijlstra <peterz@infradead.org>
When a sibling is forced-idle to match the core-cookie; search for
matching tasks to fill the core.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
include/linux/sched.h | 1 +
kernel/sched/core.c | 131 +++++++++++++++++++++++++++++++++++++++++-
kernel/sched/idle.c | 1 +
kernel/sched/sched.h | 6 ++
4 files changed, 138 insertions(+), 1 deletion(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index d319219ea58f..896dd00e393f 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -693,6 +693,7 @@ struct task_struct {
#ifdef CONFIG_SCHED_CORE
struct rb_node core_node;
unsigned long core_cookie;
+ unsigned int core_occupation;
#endif
#ifdef CONFIG_CGROUP_SCHED
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 10b20de100e3..f0a63fe5f0a4 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -211,6 +211,21 @@ static struct task_struct *sched_core_find(struct rq *rq, unsigned long cookie)
return match;
}
+static struct task_struct *sched_core_next(struct task_struct *p, unsigned long cookie)
+{
+ struct rb_node *node = &p->core_node;
+
+ node = rb_next(node);
+ if (!node)
+ return NULL;
+
+ p = container_of(node, struct task_struct, core_node);
+ if (p->core_cookie != cookie)
+ return NULL;
+
+ return p;
+}
+
/*
* The static-key + stop-machine variable are needed such that:
*
@@ -4048,7 +4063,7 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf)
struct task_struct *next, *max = NULL;
const struct sched_class *class;
const struct cpumask *smt_mask;
- int i, j, cpu;
+ int i, j, cpu, occ = 0;
bool need_sync = false;
if (!sched_core_enabled(rq))
@@ -4150,6 +4165,9 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf)
goto done;
}
+ if (!is_idle_task(p))
+ occ++;
+
rq_i->core_pick = p;
/*
@@ -4175,6 +4193,7 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf)
cpu_rq(j)->core_pick = NULL;
}
+ occ = 1;
goto again;
} else {
/*
@@ -4214,6 +4233,8 @@ next_class:;
if (is_idle_task(rq_i->core_pick) && rq_i->nr_running)
rq_i->core_forceidle = true;
+ rq_i->core_pick->core_occupation = occ;
+
if (i == cpu)
continue;
@@ -4229,6 +4250,114 @@ next_class:;
return next;
}
+static bool try_steal_cookie(int this, int that)
+{
+ struct rq *dst = cpu_rq(this), *src = cpu_rq(that);
+ struct task_struct *p;
+ unsigned long cookie;
+ bool success = false;
+
+ local_irq_disable();
+ double_rq_lock(dst, src);
+
+ cookie = dst->core->core_cookie;
+ if (!cookie)
+ goto unlock;
+
+ if (dst->curr != dst->idle)
+ goto unlock;
+
+ p = sched_core_find(src, cookie);
+ if (p == src->idle)
+ goto unlock;
+
+ do {
+ if (p == src->core_pick || p == src->curr)
+ goto next;
+
+ if (!cpumask_test_cpu(this, &p->cpus_mask))
+ goto next;
+
+ if (p->core_occupation > dst->idle->core_occupation)
+ goto next;
+
+ p->on_rq = TASK_ON_RQ_MIGRATING;
+ deactivate_task(src, p, 0);
+ set_task_cpu(p, this);
+ activate_task(dst, p, 0);
+ p->on_rq = TASK_ON_RQ_QUEUED;
+
+ resched_curr(dst);
+
+ success = true;
+ break;
+
+next:
+ p = sched_core_next(p, cookie);
+ } while (p);
+
+unlock:
+ double_rq_unlock(dst, src);
+ local_irq_enable();
+
+ return success;
+}
+
+static bool steal_cookie_task(int cpu, struct sched_domain *sd)
+{
+ int i;
+
+ for_each_cpu_wrap(i, sched_domain_span(sd), cpu) {
+ if (i == cpu)
+ continue;
+
+ if (need_resched())
+ break;
+
+ if (try_steal_cookie(cpu, i))
+ return true;
+ }
+
+ return false;
+}
+
+static void sched_core_balance(struct rq *rq)
+{
+ struct sched_domain *sd;
+ int cpu = cpu_of(rq);
+
+ rcu_read_lock();
+ raw_spin_unlock_irq(rq_lockp(rq));
+ for_each_domain(cpu, sd) {
+ if (!(sd->flags & SD_LOAD_BALANCE))
+ break;
+
+ if (need_resched())
+ break;
+
+ if (steal_cookie_task(cpu, sd))
+ break;
+ }
+ raw_spin_lock_irq(rq_lockp(rq));
+ rcu_read_unlock();
+}
+
+static DEFINE_PER_CPU(struct callback_head, core_balance_head);
+
+void queue_core_balance(struct rq *rq)
+{
+ if (!sched_core_enabled(rq))
+ return;
+
+ if (!rq->core->core_cookie)
+ return;
+
+ if (!rq->nr_running) /* not forced idle */
+ return;
+
+ queue_balance_callback(rq, &per_cpu(core_balance_head, rq->cpu), sched_core_balance);
+}
+
#else /* !CONFIG_SCHED_CORE */
static struct task_struct *
diff --git a/kernel/sched/idle.c b/kernel/sched/idle.c
index 87c55e3b988d..bd21d9922a05 100644
--- a/kernel/sched/idle.c
+++ b/kernel/sched/idle.c
@@ -388,6 +388,7 @@ static void set_next_task_idle(struct rq *rq, struct task_struct *next)
{
update_idle_core(rq);
schedstat_inc(rq->sched_goidle);
+ queue_core_balance(rq);
}
static struct task_struct *
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index fa92f6d3c73b..311ab1e2a00e 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1053,6 +1053,8 @@ static inline raw_spinlock_t *rq_lockp(struct rq *rq)
return &rq->__lock;
}
+extern void queue_core_balance(struct rq *rq);
+
#else /* !CONFIG_SCHED_CORE */
static inline bool sched_core_enabled(struct rq *rq)
@@ -1065,6 +1067,10 @@ static inline raw_spinlock_t *rq_lockp(struct rq *rq)
return &rq->__lock;
}
+static inline void queue_core_balance(struct rq *rq)
+{
+}
+
#endif /* CONFIG_SCHED_CORE */
#ifdef CONFIG_SCHED_SMT
--
2.17.1
next prev parent reply other threads:[~2019-10-30 18:34 UTC|newest]
Thread overview: 84+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-30 18:33 [RFC PATCH v4 00/19] Core scheduling v4 Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 01/19] stop_machine: Fix stop_cpus_in_progress ordering Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 02/19] sched: Fix kerneldoc comment for ia64_set_curr_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 03/19] sched: Wrap rq::lock access Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 04/19] sched/{rt,deadline}: Fix set_next_task vs pick_next_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 05/19] sched: Add task_struct pointer to sched_class::set_curr_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 06/19] sched/fair: Export newidle_balance() Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 07/19] sched: Allow put_prev_task() to drop rq->lock Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 08/19] sched: Rework pick_next_task() slow-path Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 09/19] sched: Introduce sched_class::pick_task() Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 10/19] sched: Core-wide rq->lock Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 11/19] sched: Basic tracking of matching tasks Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 12/19] sched: A quick and dirty cgroup tagging interface Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 13/19] sched: Add core wide task selection and scheduling Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 14/19] sched/fair: Add a few assertions Vineeth Remanan Pillai
2019-10-30 18:33 ` Vineeth Remanan Pillai [this message]
2019-10-30 18:33 ` [RFC PATCH v4 16/19] sched: Debug bits Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 17/19] sched/fair: wrapper for cfs_rq->min_vruntime Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 18/19] sched/fair: core wide vruntime comparison Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 19/19] sched/fair : Wake up forced idle siblings if needed Vineeth Remanan Pillai
2019-10-31 11:42 ` [RFC PATCH v4 00/19] Core scheduling v4 Li, Aubrey
2019-11-01 11:33 ` Li, Aubrey
2019-11-08 3:20 ` Li, Aubrey
2019-10-31 18:42 ` Phil Auld
2019-11-01 14:03 ` Vineeth Remanan Pillai
2019-11-01 16:35 ` Greg Kerr
2019-11-01 18:07 ` Dario Faggioli
2019-11-12 1:45 ` Dario Faggioli
2019-11-13 17:16 ` Tim Chen
2020-01-02 2:28 ` Aubrey Li
2020-01-10 23:19 ` Tim Chen
2019-11-11 19:10 ` Tim Chen
2020-01-14 1:12 ` Tim Chen
2020-01-14 15:40 ` Vineeth Remanan Pillai
2020-01-15 3:43 ` Li, Aubrey
2020-01-15 19:33 ` Tim Chen
2020-01-16 1:45 ` Aubrey Li
2020-01-17 16:00 ` Vineeth Remanan Pillai
2020-01-22 18:04 ` Gruza, Agata
2020-01-28 2:40 ` Dario Faggioli
[not found] ` <CANaguZDDpzrzdTmvjXvCmV2c+wBt6mXWSz4Vn-LJ-onc_Oj=yw@mail.gmail.com>
2020-02-01 15:31 ` Dario Faggioli
2020-02-06 0:28 ` Tim Chen
2020-02-06 22:37 ` Julien Desfossez
2020-02-12 23:07 ` Julien Desfossez
2020-02-13 18:37 ` Tim Chen
2020-02-14 6:10 ` Aubrey Li
[not found] ` <CANaguZC40mDHfL1H_9AA7H8cyd028t9PQVRqQ3kB4ha8R7hhqg@mail.gmail.com>
2020-02-15 6:01 ` Aubrey Li
[not found] ` <CANaguZBj_x_2+9KwbHCQScsmraC_mHdQB6uRqMTYMmvhBYfv2Q@mail.gmail.com>
2020-02-21 23:20 ` Julien Desfossez
2020-03-17 0:55 ` Joel Fernandes
2020-03-17 19:07 ` Tim Chen
2020-03-17 20:18 ` Tim Chen
2020-03-18 1:10 ` Joel Fernandes
2020-03-17 21:17 ` Thomas Gleixner
2020-03-17 21:58 ` Tim Chen
2020-03-18 1:03 ` Joel Fernandes
2020-03-18 2:30 ` Joel Fernandes
2020-03-18 0:52 ` Joel Fernandes
2020-03-18 11:53 ` Thomas Gleixner
2020-03-19 1:54 ` Joel Fernandes
2020-02-25 3:44 ` Aaron Lu
2020-02-25 5:32 ` Aubrey Li
2020-02-25 7:34 ` Aaron Lu
2020-02-25 10:40 ` Aubrey Li
2020-02-25 11:21 ` Aaron Lu
2020-02-25 13:41 ` Aubrey Li
[not found] ` <CANaguZD205ccu1V_2W-QuMRrJA9SjJ5ng1do4NCdLy8NDKKrbA@mail.gmail.com>
2020-02-26 3:13 ` Aaron Lu
2020-02-26 7:21 ` Aubrey Li
[not found] ` <CANaguZDQZg-Z6aNpeLcjQ-cGm3X8CQOkZ_hnJNUyqDRM=yVDFQ@mail.gmail.com>
2020-02-27 4:45 ` Aubrey Li
2020-02-28 23:55 ` Tim Chen
2020-03-03 14:59 ` Li, Aubrey
2020-03-03 23:54 ` Li, Aubrey
2020-03-05 4:33 ` Aaron Lu
2020-03-05 6:10 ` Li, Aubrey
2020-03-05 8:52 ` Aaron Lu
2020-02-27 2:04 ` Aaron Lu
2020-02-27 14:10 ` Phil Auld
2020-02-27 14:37 ` Aubrey Li
2020-02-28 2:54 ` Aaron Lu
2020-03-05 13:45 ` Aubrey Li
2020-03-06 2:41 ` Aaron Lu
2020-03-06 18:06 ` Tim Chen
2020-03-06 18:33 ` Phil Auld
2020-03-06 21:44 ` Tim Chen
2020-03-07 3:13 ` Aaron Lu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=89da8db0c28e165bef0a24bd970186b3ca0821bd.1572437285.git.vpillai@digitalocean.com \
--to=vpillai@digitalocean.com \
--cc=aaron.lwe@gmail.com \
--cc=aubrey.intel@gmail.com \
--cc=dfaggioli@suse.com \
--cc=fweisbec@gmail.com \
--cc=jdesfossez@digitalocean.com \
--cc=keescook@chromium.org \
--cc=kerrnel@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@techsingularity.net \
--cc=mingo@kernel.org \
--cc=naravamudan@digitalocean.com \
--cc=pauld@redhat.com \
--cc=pawan.kumar.gupta@linux.intel.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=pjt@google.com \
--cc=tglx@linutronix.de \
--cc=tim.c.chen@linux.intel.com \
--cc=torvalds@linux-foundation.org \
--cc=valentin.schneider@arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).