From ce96b5ac742801718ae86d2adf0500c5abef3782 Mon Sep 17 00:00:00 2001 From: Dmitry Adamushko Date: Thu, 15 Nov 2007 20:57:40 +0100 Subject: sched: fix __set_task_cpu() SMP race Grant Wilson has reported rare SCHED_FAIR_USER crashes on his quad-core system, which crashes can only be explained via runqueue corruption. there is a narrow SMP race in __set_task_cpu(): after ->cpu is set up to a new value, task_rq_lock(p, ...) can be successfuly executed on another CPU. We must ensure that updates of per-task data have been completed by this moment. this bug has been hiding in the Linux scheduler for an eternity (we never had any explicit barrier for task->cpu in set_task_cpu() - so the bug was introduced in 2.5.1), but only became visible via set_task_cfs_rq() being accidentally put after the task->cpu update. It also probably needs a sufficiently out-of-order CPU to trigger. Reported-by: Grant Wilson Signed-off-by: Dmitry Adamushko Signed-off-by: Ingo Molnar --- kernel/sched.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/sched.c b/kernel/sched.c index d62b759753f3..db1f71e31310 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -216,15 +216,15 @@ static inline struct task_group *task_group(struct task_struct *p) } /* Change a task's cfs_rq and parent entity if it moves across CPUs/groups */ -static inline void set_task_cfs_rq(struct task_struct *p) +static inline void set_task_cfs_rq(struct task_struct *p, unsigned int cpu) { - p->se.cfs_rq = task_group(p)->cfs_rq[task_cpu(p)]; - p->se.parent = task_group(p)->se[task_cpu(p)]; + p->se.cfs_rq = task_group(p)->cfs_rq[cpu]; + p->se.parent = task_group(p)->se[cpu]; } #else -static inline void set_task_cfs_rq(struct task_struct *p) { } +static inline void set_task_cfs_rq(struct task_struct *p, unsigned int cpu) { } #endif /* CONFIG_FAIR_GROUP_SCHED */ @@ -1022,10 +1022,16 @@ unsigned long weighted_cpuload(const int cpu) static inline void __set_task_cpu(struct task_struct *p, unsigned int cpu) { + set_task_cfs_rq(p, cpu); #ifdef CONFIG_SMP + /* + * After ->cpu is set up to a new value, task_rq_lock(p, ...) can be + * successfuly executed on another CPU. We must ensure that updates of + * per-task data have been completed by this moment. + */ + smp_wmb(); task_thread_info(p)->cpu = cpu; #endif - set_task_cfs_rq(p); } #ifdef CONFIG_SMP @@ -7088,7 +7094,7 @@ void sched_move_task(struct task_struct *tsk) rq = task_rq_lock(tsk, &flags); if (tsk->sched_class != &fair_sched_class) { - set_task_cfs_rq(tsk); + set_task_cfs_rq(tsk, task_cpu(tsk)); goto done; } @@ -7103,7 +7109,7 @@ void sched_move_task(struct task_struct *tsk) tsk->sched_class->put_prev_task(rq, tsk); } - set_task_cfs_rq(tsk); + set_task_cfs_rq(tsk, task_cpu(tsk)); if (on_rq) { if (unlikely(running)) -- cgit v1.2.3-59-g8ed1b