All of lore.kernel.org
 help / color / mirror / Atom feed
From: Xunlei Pang <xlpang@redhat.com>
To: linux-kernel@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Juri Lelli <juri.lelli@arm.com>, Ingo Molnar <mingo@redhat.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	Xunlei Pang <xlpang@redhat.com>
Subject: [PATCH v3 1/6] rtmutex: Deboost before waking up the top waiter
Date: Thu, 14 Apr 2016 19:37:02 +0800	[thread overview]
Message-ID: <1460633827-345-2-git-send-email-xlpang@redhat.com> (raw)
In-Reply-To: <1460633827-345-1-git-send-email-xlpang@redhat.com>

We should deboost before waking the high-prio task such that
we don't run two tasks with the 'same' priority.

The patch fixed the logic, and introduced rt_mutex_postunlock()
to do some code refactor.

Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Xunlei Pang <xlpang@redhat.com>
---
 kernel/futex.c                  |  5 ++---
 kernel/locking/rtmutex.c        | 28 ++++++++++++++++++++++++----
 kernel/locking/rtmutex_common.h |  1 +
 3 files changed, 27 insertions(+), 7 deletions(-)

diff --git a/kernel/futex.c b/kernel/futex.c
index 4e1a53e..4ae3523 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -1524,9 +1524,8 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this,
 	 * scheduled away before the wake up can take place.
 	 */
 	spin_unlock(&hb->lock);
-	wake_up_q(&wake_q);
-	if (deboost)
-		rt_mutex_adjust_prio(current);
+
+	rt_mutex_postunlock(&wake_q, deboost);
 
 	return 0;
 }
diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c
index 3e74660..42bc59b 100644
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -1390,12 +1390,32 @@ rt_mutex_fastunlock(struct rt_mutex *lock,
 	} else {
 		bool deboost = slowfn(lock, &wake_q);
 
-		wake_up_q(&wake_q);
+		rt_mutex_postunlock(&wake_q, deboost);
+	}
+}
+
 
-		/* Undo pi boosting if necessary: */
-		if (deboost)
-			rt_mutex_adjust_prio(current);
+/*
+ * Undo pi boosting (if necessary) and wake top waiter.
+ */
+void rt_mutex_postunlock(struct wake_q_head *wake_q, bool deboost)
+{
+	/*
+	 * We should deboost before waking the high-prio task such that
+	 * we don't run two tasks with the 'same' priority. This however
+	 * can lead to prio-inversion if we would get preempted after
+	 * the deboost but before waking our high-prio task, hence the
+	 * preempt_disable.
+	 */
+	if (deboost) {
+		preempt_disable();
+		rt_mutex_adjust_prio(current);
 	}
+
+	wake_up_q(wake_q);
+
+	if (deboost)
+		preempt_enable();
 }
 
 /**
diff --git a/kernel/locking/rtmutex_common.h b/kernel/locking/rtmutex_common.h
index 4f5f83c..93b0924 100644
--- a/kernel/locking/rtmutex_common.h
+++ b/kernel/locking/rtmutex_common.h
@@ -111,6 +111,7 @@ extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
 extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to);
 extern bool rt_mutex_futex_unlock(struct rt_mutex *lock,
 				  struct wake_q_head *wqh);
+extern void rt_mutex_postunlock(struct wake_q_head *wake_q, bool deboost);
 extern void rt_mutex_adjust_prio(struct task_struct *task);
 
 #ifdef CONFIG_DEBUG_RT_MUTEXES
-- 
1.8.3.1

  reply	other threads:[~2016-04-14 11:37 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-14 11:37 [PATCH v3 0/6] sched/deadline/rtmutex: Fix two deadline PI issues Xunlei Pang
2016-04-14 11:37 ` Xunlei Pang [this message]
2016-04-18  8:23   ` [PATCH v3 1/6] rtmutex: Deboost before waking up the top waiter Thomas Gleixner
2016-04-18  8:44     ` Xunlei Pang
2016-04-18  9:02       ` Thomas Gleixner
2016-04-18  9:41         ` Xunlei Pang
2016-04-20 12:20         ` Peter Zijlstra
2016-04-20 12:43           ` Thomas Gleixner
2016-04-20 13:10             ` Peter Zijlstra
2016-04-14 11:37 ` [PATCH v3 2/6] sched/rtmutex/deadline: Fix a PI crash for deadline tasks Xunlei Pang
2016-04-20 13:19   ` Peter Zijlstra
2016-04-20 13:49     ` Xunlei Pang
2016-04-22  3:26       ` Xunlei Pang
2016-04-14 11:37 ` [PATCH v3 3/6] rtmutex: Move "rt_mutex_waiter" definition to "include/linux/rtmutex.h" Xunlei Pang
2016-04-14 11:37 ` [PATCH v3 4/6] sched: Move dl_policy() to "include/linux/sched.h" Xunlei Pang
2016-04-14 11:37 ` [PATCH v3 5/6] sched/deadline/rtmutex: Fix unprotected PI access in enqueue_task_dl() Xunlei Pang
2016-04-14 15:31   ` Peter Zijlstra
2016-04-15  1:58     ` Xunlei Pang
2016-04-15  2:19       ` Xunlei Pang
2016-04-20 12:25         ` Peter Zijlstra
2016-04-20 13:00           ` Xunlei Pang
2016-04-20 13:17             ` Peter Zijlstra
2016-04-20 13:45               ` Xunlei Pang
2016-04-14 11:37 ` [PATCH v3 6/6] sched/deadline/rtmutex: Don't miss the dl_runtime/dl_period update Xunlei Pang
2017-03-23 14:56 [PATCH -v3 0/8] PI vs SCHED_DEADLINE fixes Peter Zijlstra
2017-03-23 14:56 ` [PATCH -v3 1/8] rtmutex: Deboost before waking up the top waiter Peter Zijlstra
2017-04-04  9:48   ` [tip:locking/core] " tip-bot for Xunlei Pang
2017-04-05  8:08     ` Mike Galbraith
2017-04-05 14:55       ` [tip:locking/core] Retiplockingcore_rtmutex_Deboost_before_waking_up_the_top_waiter tip-bot for Mike Galbraith
2017-04-05 15:03       ` [tip:locking/core] rtmutex: Plug preempt count leak in rt_mutex_futex_unlock() tip-bot for Mike Galbraith
2017-04-06  6:16       ` [tip:locking/core] rtmutex: Deboost before waking up the top waiter Xunlei Pang
2017-03-23 14:56 ` [PATCH -v3 2/8] sched/rtmutex/deadline: Fix a PI crash for deadline tasks Peter Zijlstra
2017-04-04  9:49   ` [tip:locking/core] " tip-bot for Xunlei Pang
2017-03-23 14:56 ` [PATCH -v3 3/8] sched/deadline/rtmutex: Dont miss the dl_runtime/dl_period update Peter Zijlstra
2017-04-04  9:50   ` [tip:locking/core] " tip-bot for Xunlei Pang
2017-03-23 14:56 ` [PATCH -v3 4/8] rtmutex: Clean up Peter Zijlstra
2017-03-23 18:21   ` Steven Rostedt
2017-04-04  9:50   ` [tip:locking/core] " tip-bot for Peter Zijlstra
2017-03-23 14:56 ` [PATCH -v3 5/8] sched/rtmutex: Refactor rt_mutex_setprio() Peter Zijlstra
2017-04-04  9:51   ` [tip:locking/core] " tip-bot for Peter Zijlstra
2017-03-23 14:56 ` [PATCH -v3 6/8] sched,tracing: Update trace_sched_pi_setprio() Peter Zijlstra
2017-04-04  9:51   ` [tip:locking/core] " tip-bot for Peter Zijlstra
2017-03-23 14:56 ` [PATCH -v3 7/8] rtmutex: Fix PI chain order integrity Peter Zijlstra
2017-04-04  9:52   ` [tip:locking/core] " tip-bot for Peter Zijlstra
2017-03-23 14:56 ` [PATCH -v3 8/8] rtmutex: Fix more prio comparisons Peter Zijlstra
2017-04-04  9:52   ` [tip:locking/core] " tip-bot for Peter Zijlstra

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=1460633827-345-2-git-send-email-xlpang@redhat.com \
    --to=xlpang@redhat.com \
    --cc=juri.lelli@arm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=tglx@linutronix.de \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.