linux-rt-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Gregory Haskins <ghaskins@novell.com>
To: Chirag Jog <chirag@linux.vnet.ibm.com>
Cc: linux-rt-users@vger.kernel.org, linux-kernel@vger.kernel.org,
	rostedt@goodmis.org, dvhltc@us.ibm.com, dino@in.ibm.com,
	Gilles.Carry@bull.net
Subject: [RT PATCH v2 2/2] RT: remove "paranoid" limit in push_rt_task
Date: Fri, 03 Oct 2008 13:26:18 -0400	[thread overview]
Message-ID: <20081003172618.23714.55546.stgit@dev.haskins.net> (raw)
In-Reply-To: <20081003172221.23714.71575.stgit@dev.haskins.net>

A panic was discovered by Chirag Jog and investigated by Gilles Carry
to be originating in the fact that a task being pushed away
may get migrated away during a double_lock_balance.  The result was
that the pushable_tasks list may become corrupted.

The root cause is that the "paranoid" retry limit could cause us to
bail out of a retry, but still try to remove the item from the (now
potentially incorrect) list.  There are numerous ways to correct the
condition, but the paranoid feature is no longer relevant with the new
pushable logic (since pushable naturally limits the loop anyway), so
lets just remove it.

Reported By: Chirag Jog <chirag@linux.vnet.ibm.com>
Found-by: Gilles Carry <gilles.carry@bull.net>
Signed-off-by: Gregory Haskins <ghaskins@novell.com>
---

 kernel/sched_rt.c |   34 ++++++++++++++++++++++------------
 1 files changed, 22 insertions(+), 12 deletions(-)

diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 59ead84..201bd97 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -1056,7 +1056,6 @@ static int push_rt_task(struct rq *rq)
 {
 	struct task_struct *next_task;
 	struct rq *lowest_rq;
-	int paranoid = RT_MAX_TRIES;
 
 	if (!rq->rt.overloaded)
 		return 0;
@@ -1090,23 +1089,34 @@ static int push_rt_task(struct rq *rq)
 		struct task_struct *task;
 		/*
 		 * find lock_lowest_rq releases rq->lock
-		 * so it is possible that next_task has changed.
-		 * If it has, then try again.
+		 * so it is possible that next_task has migrated.
+		 *
+		 * We need to make sure that the task is still on the same
+		 * run-queue and is also still the next task eligible for
+		 * pushing.
 		 */
 		task = pick_next_pushable_task(rq);
-		if (unlikely(task != next_task) && task && paranoid--) {
-			put_task_struct(next_task);
-			next_task = task;
-			goto retry;
+		if (task_cpu(next_task) == rq->cpu && task == next_task) {
+			/*
+			 * If we get here, the task hasnt moved it all, but
+			 * it has failed to push.  We will not try again,
+			 * since the other cpus will pull from us when they
+			 * are ready.
+			 */
+			dequeue_pushable_task(rq, next_task);
+			goto out;
 		}
+		
+		if (!task)
+			/* No more tasks, just exit */
+			goto out;
 
 		/*
-		 * Once we have failed to push this task, we will not
-		 * try again, since the other cpus will pull from us
-		 * when they are ready
+		 * Something has shifted, try again.
 		 */
-		dequeue_pushable_task(rq, next_task);
-		goto out;
+		put_task_struct(next_task);
+		next_task = task;
+		goto retry;
 	}
 
 	deactivate_task(rq, next_task, 0);


  parent reply	other threads:[~2008-10-03 17:21 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-09-25 12:32 [BUG][PPC64] BUG in 2.6.26.5-rt9 causing Hang Chirag Jog
2008-09-29 18:13 ` Gregory Haskins
2008-09-29 21:18 ` Gregory Haskins
2008-09-29 21:34   ` Gregory Haskins
2008-09-29 22:00     ` Gregory Haskins
2008-09-30  4:43       ` Chirag Jog
2008-09-30  6:47         ` Gilles Carry
2008-10-01 14:22         ` [PATCH] sched: add a stacktrace on enqueue_pushable error Gregory Haskins
2008-10-02  9:42           ` Gilles Carry
2008-10-02 11:18   ` [BUG][PPC64] BUG in 2.6.26.5-rt9 causing Hang Gilles Carry
2008-10-03 12:42 ` [RT PATCH 0/2] fix for BUG_ON crash in 26.5-rt9 Gregory Haskins
2008-10-03 12:43   ` [PATCH 1/2] RT: Remove comment that is no longer true Gregory Haskins
2008-10-03 12:43   ` [PATCH 2/2] RT: remove "paranoid" limit in push_rt_task Gregory Haskins
2008-10-03 13:46     ` Gilles Carry
2008-10-03 15:45       ` Chirag Jog
2008-10-03 17:27         ` Gregory Haskins
2008-10-03 17:26       ` [RT PATCH v2 0/2] Series short description Gregory Haskins
2008-10-03 17:26         ` [RT PATCH v2 1/2] RT: Remove comment that is no longer true Gregory Haskins
2008-10-03 17:26         ` Gregory Haskins [this message]
2008-10-03 12:54   ` [RT PATCH 0/2] fix for BUG_ON crash in 26.5-rt9 Gregory Haskins
2008-10-06 15:14 ` [RT PATCH v3 0/2] Fix for "[BUG][PPC64] BUG in 2.6.26.5-rt9 causing Hang" Gregory Haskins
2008-10-06 15:14   ` [RT PATCH v3 1/2] RT: Remove comment that is no longer true Gregory Haskins
2008-10-06 15:14   ` [RT PATCH v3 2/2] RT: fix push_rt_task() to handle dequeue_pushable properly Gregory Haskins
2008-10-07  6:04   ` [RT PATCH v3 0/2] Fix for "[BUG][PPC64] BUG in 2.6.26.5-rt9 causing Hang" Gilles Carry

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=20081003172618.23714.55546.stgit@dev.haskins.net \
    --to=ghaskins@novell.com \
    --cc=Gilles.Carry@bull.net \
    --cc=chirag@linux.vnet.ibm.com \
    --cc=dino@in.ibm.com \
    --cc=dvhltc@us.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rt-users@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    /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).