linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: luca abeni <luca.abeni@santannapisa.it>
To: linux-kernel@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>, Juri Lelli <juri.lelli@arm.com>,
	Claudio Scordino <claudio@evidence.eu.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	Tommaso Cucinotta <tommaso.cucinotta@sssup.it>,
	Daniel Bristot de Oliveira <bristot@redhat.com>,
	Joel Fernandes <joelaf@google.com>,
	Mathieu Poirier <mathieu.poirier@linaro.org>,
	Luca Abeni <luca.abeni@santannapisa.it>
Subject: [PATCH 04/10] sched/deadline: implement GRUB accounting
Date: Thu, 18 May 2017 22:13:31 +0200	[thread overview]
Message-ID: <1495138417-6203-5-git-send-email-luca.abeni@santannapisa.it> (raw)
In-Reply-To: <1495138417-6203-1-git-send-email-luca.abeni@santannapisa.it>

From: Luca Abeni <luca.abeni@santannapisa.it>

According to the GRUB (Greedy Reclaimation of Unused Bandwidth)
reclaiming algorithm, the runtime is not decreased as "dq = -dt",
but as "dq = -Uact dt" (where Uact is the per-runqueue active
utilization).
Hence, this commit modifies the runtime accounting rule in
update_curr_dl() to implement the GRUB rule.

Signed-off-by: Luca Abeni <luca.abeni@santannapisa.it>
Tested-by: Daniel Bristot de Oliveira <bristot@redhat.com>
---
 kernel/sched/core.c     |  4 ++--
 kernel/sched/deadline.c | 17 +++++++++++++++++
 kernel/sched/sched.h    |  2 ++
 3 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 57856b2..14dd97e 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2435,7 +2435,7 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p)
 unsigned long to_ratio(u64 period, u64 runtime)
 {
 	if (runtime == RUNTIME_INF)
-		return 1ULL << 20;
+		return BW_UNIT;
 
 	/*
 	 * Doing this here saves a lot of checks in all
@@ -2445,7 +2445,7 @@ unsigned long to_ratio(u64 period, u64 runtime)
 	if (period == 0)
 		return 0;
 
-	return div64_u64(runtime << 20, period);
+	return div64_u64(runtime << BW_SHIFT, period);
 }
 
 #ifdef CONFIG_SMP
diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
index bb71d65..4e0c02c 100644
--- a/kernel/sched/deadline.c
+++ b/kernel/sched/deadline.c
@@ -918,6 +918,22 @@ int dl_runtime_exceeded(struct sched_dl_entity *dl_se)
 extern bool sched_rt_bandwidth_account(struct rt_rq *rt_rq);
 
 /*
+ * This function implements the GRUB accounting rule:
+ * according to the GRUB reclaiming algorithm, the runtime is
+ * not decreased as "dq = -dt", but as "dq = -Uact dt", where
+ * Uact is the (per-runqueue) active utilization.
+ * Since rq->dl.running_bw contains Uact * 2^BW_SHIFT, the result
+ * has to be shifted right by BW_SHIFT.
+ */
+u64 grub_reclaim(u64 delta, struct rq *rq)
+{
+	delta *= rq->dl.running_bw;
+	delta >>= BW_SHIFT;
+
+	return delta;
+}
+
+/*
  * Update the current task's runtime statistics (provided it is still
  * a -deadline task and has not been removed from the dl_rq).
  */
@@ -959,6 +975,7 @@ static void update_curr_dl(struct rq *rq)
 
 	sched_rt_avg_update(rq, delta_exec);
 
+	delta_exec = grub_reclaim(delta_exec, rq);
 	dl_se->runtime -= delta_exec;
 
 throttle:
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index 3069e45..c0001ab 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1496,6 +1496,8 @@ extern void init_dl_bandwidth(struct dl_bandwidth *dl_b, u64 period, u64 runtime
 extern void init_dl_task_timer(struct sched_dl_entity *dl_se);
 extern void init_dl_inactive_task_timer(struct sched_dl_entity *dl_se);
 
+#define BW_SHIFT	20
+#define BW_UNIT		(1 << BW_SHIFT)
 unsigned long to_ratio(u64 period, u64 runtime);
 
 extern void init_entity_runnable_average(struct sched_entity *se);
-- 
2.7.4

  parent reply	other threads:[~2017-05-18 20:16 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-05-18 20:13 [PATCH 00/10] CPU reclaiming for SCHED_DEADLINE luca abeni
2017-05-18 20:13 ` [PATCH 01/10] sched/deadline: track the active utilization luca abeni
2017-06-08  8:31   ` Ingo Molnar
2017-06-08  8:43     ` Luca Abeni
2017-06-08  9:05       ` Juri Lelli
2017-06-08 13:36         ` Steven Rostedt
2017-06-08 13:47           ` Juri Lelli
2017-06-08  9:23   ` [tip:sched/core] sched/deadline: Track " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 02/10] sched/deadline: improve the tracking of " luca abeni
2017-06-08  9:24   ` [tip:sched/core] sched/deadline: Improve " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 03/10] sched/deadline: fix the update of the total -deadline utilization luca abeni
2017-06-08  9:24   ` [tip:sched/core] sched/deadline: Fix " tip-bot for Luca Abeni
2017-05-18 20:13 ` luca abeni [this message]
2017-06-08  9:25   ` [tip:sched/core] sched/deadline: Implement GRUB accounting tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 05/10] sched/deadline: do not reclaim the whole CPU bandwidth luca abeni
2017-06-08  9:25   ` [tip:sched/core] sched/deadline: Do " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 06/10] sched/deadline: make GRUB a task's flag luca abeni
2017-06-08  9:26   ` [tip:sched/core] sched/deadline: Make " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 07/10] sched/deadline: track the "total rq utilization" too luca abeni
2017-06-08  9:26   ` [tip:sched/core] sched/deadline: Track " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 08/10] sched/deadline: base GRUB reclaiming on the inactive utilization luca abeni
2017-06-08  9:27   ` [tip:sched/core] sched/deadline: Base " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 09/10] sched/deadline: also reclaim bandwidth not used by dl tasks luca abeni
2017-06-08  9:28   ` [tip:sched/core] sched/deadline: Reclaim " tip-bot for Luca Abeni
2017-05-18 20:13 ` [PATCH 10/10] sched/deadline: documentation about GRUB reclaiming luca abeni
2017-06-08  9:28   ` [tip:sched/core] sched/deadline: Add " tip-bot for Claudio Scordino

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=1495138417-6203-5-git-send-email-luca.abeni@santannapisa.it \
    --to=luca.abeni@santannapisa.it \
    --cc=bristot@redhat.com \
    --cc=claudio@evidence.eu.com \
    --cc=joelaf@google.com \
    --cc=juri.lelli@arm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mathieu.poirier@linaro.org \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=tommaso.cucinotta@sssup.it \
    /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).