summaryrefslogtreecommitdiff
path: root/kernel/sched.c
diff options
context:
space:
mode:
authorPeter Zijlstra <a.p.zijlstra@chello.nl>2010-03-24 18:34:10 +0100
committerIngo Molnar <mingo@elte.hu>2010-04-02 20:12:03 +0200
commit0017d735092844118bef006696a750a0e4ef6ebd (patch)
tree8ed1540aaeb63da726f93da12950a8eaa0e0a3e0 /kernel/sched.c
parent9084bb8246ea935b98320554229e2f371f7f52fa (diff)
sched: Fix TASK_WAKING vs fork deadlock
Oleg noticed a few races with the TASK_WAKING usage on fork. - since TASK_WAKING is basically a spinlock, it should be IRQ safe - since we set TASK_WAKING (*) without holding rq->lock it could be there still is a rq->lock holder, thereby not actually providing full serialization. (*) in fact we clear PF_STARTING, which in effect enables TASK_WAKING. Cure the second issue by not setting TASK_WAKING in sched_fork(), but only temporarily in wake_up_new_task() while calling select_task_rq(). Cure the first by holding rq->lock around the select_task_rq() call, this will disable IRQs, this however requires that we push down the rq->lock release into select_task_rq_fair()'s cgroup stuff. Because select_task_rq_fair() still needs to drop the rq->lock we cannot fully get rid of TASK_WAKING. Reported-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'kernel/sched.c')
-rw-r--r--kernel/sched.c65
1 files changed, 24 insertions, 41 deletions
diff --git a/kernel/sched.c b/kernel/sched.c
index 9a38c7a24ed7..dcd17736dae1 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -916,14 +916,10 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev)
/*
* Check whether the task is waking, we use this to synchronize against
* ttwu() so that task_cpu() reports a stable number.
- *
- * We need to make an exception for PF_STARTING tasks because the fork
- * path might require task_rq_lock() to work, eg. it can call
- * set_cpus_allowed_ptr() from the cpuset clone_ns code.
*/
static inline int task_is_waking(struct task_struct *p)
{
- return unlikely((p->state == TASK_WAKING) && !(p->flags & PF_STARTING));
+ return unlikely(p->state == TASK_WAKING);
}
/*
@@ -2320,9 +2316,9 @@ static int select_fallback_rq(int cpu, struct task_struct *p)
* The caller (fork, wakeup) owns TASK_WAKING, ->cpus_allowed is stable.
*/
static inline
-int select_task_rq(struct task_struct *p, int sd_flags, int wake_flags)
+int select_task_rq(struct rq *rq, struct task_struct *p, int sd_flags, int wake_flags)
{
- int cpu = p->sched_class->select_task_rq(p, sd_flags, wake_flags);
+ int cpu = p->sched_class->select_task_rq(rq, p, sd_flags, wake_flags);
/*
* In order not to call set_task_cpu() on a blocking task we need
@@ -2393,17 +2389,10 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state,
if (p->sched_class->task_waking)
p->sched_class->task_waking(rq, p);
- __task_rq_unlock(rq);
-
- cpu = select_task_rq(p, SD_BALANCE_WAKE, wake_flags);
- if (cpu != orig_cpu) {
- /*
- * Since we migrate the task without holding any rq->lock,
- * we need to be careful with task_rq_lock(), since that
- * might end up locking an invalid rq.
- */
+ cpu = select_task_rq(rq, p, SD_BALANCE_WAKE, wake_flags);
+ if (cpu != orig_cpu)
set_task_cpu(p, cpu);
- }
+ __task_rq_unlock(rq);
rq = cpu_rq(cpu);
raw_spin_lock(&rq->lock);
@@ -2530,11 +2519,11 @@ void sched_fork(struct task_struct *p, int clone_flags)
__sched_fork(p);
/*
- * We mark the process as waking here. This guarantees that
+ * We mark the process as running here. This guarantees that
* nobody will actually run it, and a signal or other external
* event cannot wake it up and insert it on the runqueue either.
*/
- p->state = TASK_WAKING;
+ p->state = TASK_RUNNING;
/*
* Revert to default priority/policy on fork if requested.
@@ -2601,28 +2590,25 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags)
int cpu __maybe_unused = get_cpu();
#ifdef CONFIG_SMP
+ rq = task_rq_lock(p, &flags);
+ p->state = TASK_WAKING;
+
/*
* Fork balancing, do it here and not earlier because:
* - cpus_allowed can change in the fork path
* - any previously selected cpu might disappear through hotplug
*
- * We still have TASK_WAKING but PF_STARTING is gone now, meaning
- * ->cpus_allowed is stable, we have preemption disabled, meaning
- * cpu_online_mask is stable.
+ * We set TASK_WAKING so that select_task_rq() can drop rq->lock
+ * without people poking at ->cpus_allowed.
*/
- cpu = select_task_rq(p, SD_BALANCE_FORK, 0);
+ cpu = select_task_rq(rq, p, SD_BALANCE_FORK, 0);
set_task_cpu(p, cpu);
-#endif
-
- /*
- * Since the task is not on the rq and we still have TASK_WAKING set
- * nobody else will migrate this task.
- */
- rq = cpu_rq(cpu);
- raw_spin_lock_irqsave(&rq->lock, flags);
- BUG_ON(p->state != TASK_WAKING);
p->state = TASK_RUNNING;
+ task_rq_unlock(rq, &flags);
+#endif
+
+ rq = task_rq_lock(p, &flags);
activate_task(rq, p, 0);
trace_sched_wakeup_new(rq, p, 1);
check_preempt_curr(rq, p, WF_FORK);
@@ -3068,19 +3054,15 @@ void sched_exec(void)
{
struct task_struct *p = current;
struct migration_req req;
- int dest_cpu, this_cpu;
unsigned long flags;
struct rq *rq;
-
- this_cpu = get_cpu();
- dest_cpu = p->sched_class->select_task_rq(p, SD_BALANCE_EXEC, 0);
- if (dest_cpu == this_cpu) {
- put_cpu();
- return;
- }
+ int dest_cpu;
rq = task_rq_lock(p, &flags);
- put_cpu();
+ dest_cpu = p->sched_class->select_task_rq(rq, p, SD_BALANCE_EXEC, 0);
+ if (dest_cpu == smp_processor_id())
+ goto unlock;
+
/*
* select_task_rq() can race against ->cpus_allowed
*/
@@ -3098,6 +3080,7 @@ void sched_exec(void)
return;
}
+unlock:
task_rq_unlock(rq, &flags);
}