sched/wake_q: Reduce reference counting for special users

Some users, specifically futexes and rwsems, required fixes
that allowed the callers to be safe when wakeups occur before
they are expected by wake_up_q(). Such scenarios also play
games and rely on reference counting, and until now were
pivoting on wake_q doing it. With the wake_q_add() call being
moved down, this can no longer be the case. As such we end up
with a a double task refcounting overhead; and these callers
care enough about this (being rather core-ish).

This patch introduces a wake_q_add_safe() call that serves
for callers that have already done refcounting and therefore the
task is 'safe' from wake_q point of view (int that it requires
reference throughout the entire queue/>wakeup cycle). In the one
case it has internal reference counting, in the other case it
consumes the reference counting.

Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Waiman Long <longman@redhat.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Xie Yongji <xieyongji@baidu.com>
Cc: Yongji Xie <elohimes@gmail.com>
Cc: andrea.parri@amarulasolutions.com
Cc: lilin24@baidu.com
Cc: liuqi16@baidu.com
Cc: nixun@baidu.com
Cc: yuanlinsi01@baidu.com
Cc: zhangyu31@baidu.com
Link: https://lkml.kernel.org/r/20181218195352.7orq3upiwfdbrdne@linux-r8p5
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This commit is contained in:
Davidlohr Bueso 2018-12-18 11:53:52 -08:00 committed by Ingo Molnar
parent 513e1073d5
commit 07879c6a37
4 changed files with 52 additions and 27 deletions

View File

@ -51,8 +51,8 @@ static inline void wake_q_init(struct wake_q_head *head)
head->lastp = &head->first; head->lastp = &head->first;
} }
extern void wake_q_add(struct wake_q_head *head, extern void wake_q_add(struct wake_q_head *head, struct task_struct *task);
struct task_struct *task); extern void wake_q_add_safe(struct wake_q_head *head, struct task_struct *task);
extern void wake_up_q(struct wake_q_head *head); extern void wake_up_q(struct wake_q_head *head);
#endif /* _LINUX_SCHED_WAKE_Q_H */ #endif /* _LINUX_SCHED_WAKE_Q_H */

View File

@ -1463,8 +1463,7 @@ static void mark_wake_futex(struct wake_q_head *wake_q, struct futex_q *q)
* Queue the task for later wakeup for after we've released * Queue the task for later wakeup for after we've released
* the hb->lock. wake_q_add() grabs reference to p. * the hb->lock. wake_q_add() grabs reference to p.
*/ */
wake_q_add(wake_q, p); wake_q_add_safe(wake_q, p);
put_task_struct(p);
} }
/* /*

View File

@ -211,9 +211,7 @@ static void __rwsem_mark_wake(struct rw_semaphore *sem,
* Ensure issuing the wakeup (either by us or someone else) * Ensure issuing the wakeup (either by us or someone else)
* after setting the reader waiter to nil. * after setting the reader waiter to nil.
*/ */
wake_q_add(wake_q, tsk); wake_q_add_safe(wake_q, tsk);
/* wake_q_add() already take the task ref */
put_task_struct(tsk);
} }
adjustment = woken * RWSEM_ACTIVE_READ_BIAS - adjustment; adjustment = woken * RWSEM_ACTIVE_READ_BIAS - adjustment;

View File

@ -396,6 +396,30 @@ static bool set_nr_if_polling(struct task_struct *p)
#endif #endif
#endif #endif
static bool __wake_q_add(struct wake_q_head *head, struct task_struct *task)
{
struct wake_q_node *node = &task->wake_q;
/*
* Atomically grab the task, if ->wake_q is !nil already it means
* its already queued (either by us or someone else) and will get the
* wakeup due to that.
*
* In order to ensure that a pending wakeup will observe our pending
* state, even in the failed case, an explicit smp_mb() must be used.
*/
smp_mb__before_atomic();
if (unlikely(cmpxchg_relaxed(&node->next, NULL, WAKE_Q_TAIL)))
return false;
/*
* The head is context local, there can be no concurrency.
*/
*head->lastp = node;
head->lastp = &node->next;
return true;
}
/** /**
* wake_q_add() - queue a wakeup for 'later' waking. * wake_q_add() - queue a wakeup for 'later' waking.
* @head: the wake_q_head to add @task to * @head: the wake_q_head to add @task to
@ -410,27 +434,31 @@ static bool set_nr_if_polling(struct task_struct *p)
*/ */
void wake_q_add(struct wake_q_head *head, struct task_struct *task) void wake_q_add(struct wake_q_head *head, struct task_struct *task)
{ {
struct wake_q_node *node = &task->wake_q; if (__wake_q_add(head, task))
get_task_struct(task);
}
/* /**
* Atomically grab the task, if ->wake_q is !nil already it means * wake_q_add_safe() - safely queue a wakeup for 'later' waking.
* its already queued (either by us or someone else) and will get the * @head: the wake_q_head to add @task to
* wakeup due to that. * @task: the task to queue for 'later' wakeup
* *
* In order to ensure that a pending wakeup will observe our pending * Queue a task for later wakeup, most likely by the wake_up_q() call in the
* state, even in the failed case, an explicit smp_mb() must be used. * same context, _HOWEVER_ this is not guaranteed, the wakeup can come
*/ * instantly.
smp_mb__before_atomic(); *
if (unlikely(cmpxchg_relaxed(&node->next, NULL, WAKE_Q_TAIL))) * This function must be used as-if it were wake_up_process(); IOW the task
return; * must be ready to be woken at this location.
*
get_task_struct(task); * This function is essentially a task-safe equivalent to wake_q_add(). Callers
* that already hold reference to @task can call the 'safe' version and trust
/* * wake_q to do the right thing depending whether or not the @task is already
* The head is context local, there can be no concurrency. * queued for wakeup.
*/ */
*head->lastp = node; void wake_q_add_safe(struct wake_q_head *head, struct task_struct *task)
head->lastp = &node->next; {
if (!__wake_q_add(head, task))
put_task_struct(task);
} }
void wake_up_q(struct wake_q_head *head) void wake_up_q(struct wake_q_head *head)