sched: Fix 32bit race

PATCH In review Peter Zijlstra 3 participants 3 messages Jul 12, 2011

lore

Share & tools

lore

Peter Zijlstra

author

2011-07-12 12:07 UTC #0 lore raw

PATCH 1/1 sched: Fix 32bit race

Ingo, can we get the below patch into sched/urgent and towards Linus before he cuts a release?

I'm surprised there weren't more 32bit-smp doesn't boot for me complaints, either Bruno is a very 'lucky' man, or nobody is using 32bit kernels any more since he was hitting this quite reliably.

---
Subject: sched: Fix 32bit race
From: Peter Zijlstra <[email protected]>
Date: Mon Jul 11 16:28:50 CEST 2011

Commit 3fe1698b7fe0 ("sched: Deal with non-atomic min_vruntime reads on 32bit") forgot to initialize min_vruntime_copy which could lead to an infinite while loop in task_waking_fair() under some circumstances (early boot, lucky timing).

Reported-and-tested-by Bruno Wolff III <[email protected]>

Signed-off-by Peter Zijlstra <[email protected]>

---

kernel/sched.c +3 -0

Index: linux-2.6/kernel/sched.c
===================================================================
@@ -7750,6 +7750,9 @@ static void init_cfs_rq(struct cfs_rq *c
 #endif
 #endif
 	cfs_rq->min_vruntime = (u64)(-(1LL << 20));
+#ifndef CONFIG_64BIT
+	cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime;
+#endif
 }
 
 static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq)

Bruno Wolff III

2011-07-12 12:56 UTC #1 lore raw

Quoted 6 lines · Peter Zijlstra

Ingo, can we get the below patch into sched/urgent and towards Linus before he cuts a release?
I'm surprised there weren't more 32bit-smp doesn't boot for me complaints, either Bruno is a very 'lucky' man, or nobody is using 32bit kernels any more since he was hitting this quite reliably.

On two different machines. Both an Athlon MP and an old Xeon were hitting this problem on every boot.

Paul E. McKenney

Reviewed-by on v1 1/1

2011-07-12 20:09 UTC #2 lore raw

In reply to Peter Zijlstra

Quoted 6 lines · Peter Zijlstra

Ingo, can we get the below patch into sched/urgent and towards Linus before he cuts a release?
I'm surprised there weren't more 32bit-smp doesn't boot for me complaints, either Bruno is a very 'lucky' man, or nobody is using 32bit kernels any more since he was hitting this quite reliably.

It appears that there have been several, but they were blaming RCU instead. ;-)

Quoted 9 lines

---
Subject: sched: Fix 32bit race
From: Peter Zijlstra <[email protected]>
Date: Mon Jul 11 16:28:50 CEST 2011
Commit 3fe1698b7fe0 ("sched: Deal with non-atomic min_vruntime reads on 32bit") forgot to initialize min_vruntime_copy which could lead to an infinite while loop in task_waking_fair() under some circumstances (early boot, lucky timing).

Reviewed-by Paul E. McKenney <[email protected]>

Quoted 24 lines

Reported-and-tested-by Bruno Wolff III <[email protected]>
Signed-off-by Peter Zijlstra <[email protected]>
---
Index: linux-2.6/kernel/sched.c
===================================================================
kernel/sched.c +3 -1
@@ -7750,6 +7750,9 @@ static void init_cfs_rq(struct cfs_rq *c
 #endif
 #endif
 	cfs_rq->min_vruntime = (u64)(-(1LL << 20));
+#ifndef CONFIG_64BIT
+	cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime;
+#endif
 }
 
 static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq)
 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/