Re: [PATCH v4] sched/wake_q: Reduce reference counting for special users

[Davidlohr Bueso <dave@stgolabs.net>]

Hi - considering that the wake_q patches were picked up for tip/urgent, can
this one make it in as well?

Thanks,
Davidlohr

On Tue, 18 Dec 2018, Waiman Long wrote:

>On 12/18/2018 02:53 PM, Davidlohr Bueso wrote:
>> 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>
>> ---
>>
>> - Changes from v3: fixed wake_q_add_safe. While previous version
>>  had been tested with a bootup, the failed cmpxchg path obviously
>>  hadn't been exercised.  Sorry about the noise.
>>
>> include/linux/sched/wake_q.h |  4 +--
>> kernel/futex.c               |  3 +--
>> kernel/locking/rwsem-xadd.c  |  4 +--
>> kernel/sched/core.c          | 60
>> ++++++++++++++++++++++++++++++++------------
>> 4 files changed, 48 insertions(+), 23 deletions(-)
>>
>> diff --git a/include/linux/sched/wake_q.h b/include/linux/sched/wake_q.h
>> index 545f37138057..ad826d2a4557 100644
>> --- a/include/linux/sched/wake_q.h
>> +++ b/include/linux/sched/wake_q.h
>> @@ -51,8 +51,8 @@ static inline void wake_q_init(struct wake_q_head
>> *head)
>>     head->lastp = &head->first;
>> }
>>
>> -extern void wake_q_add(struct wake_q_head *head,
>> -               struct task_struct *task);
>> +extern void wake_q_add(struct wake_q_head *head, 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);
>>
>> #endif /* _LINUX_SCHED_WAKE_Q_H */
>> diff --git a/kernel/futex.c b/kernel/futex.c
>> index d14971f6ed3d..6218d98f649b 100644
>> --- a/kernel/futex.c
>> +++ b/kernel/futex.c
>> @@ -1402,8 +1402,7 @@ 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
>>      * the hb->lock. wake_q_add() grabs reference to p.
>>      */
>> -    wake_q_add(wake_q, p);
>> -    put_task_struct(p);
>> +    wake_q_add_safe(wake_q, p);
>> }
>>
>> /*
>> diff --git a/kernel/locking/rwsem-xadd.c b/kernel/locking/rwsem-xadd.c
>> index 50d9af615dc4..fbe96341beee 100644
>> --- a/kernel/locking/rwsem-xadd.c
>> +++ b/kernel/locking/rwsem-xadd.c
>> @@ -211,9 +211,7 @@ static void __rwsem_mark_wake(struct rw_semaphore
>> *sem,
>>          * Ensure issuing the wakeup (either by us or someone else)
>>          * after setting the reader waiter to nil.
>>          */
>> -        wake_q_add(wake_q, tsk);
>> -        /* wake_q_add() already take the task ref */
>> -        put_task_struct(tsk);
>> +        wake_q_add_safe(wake_q, tsk);
>>     }
>>
>>     adjustment = woken * RWSEM_ACTIVE_READ_BIAS - adjustment;
>> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
>> index d740d7a3608d..be977df66a21 100644
>> --- a/kernel/sched/core.c
>> +++ b/kernel/sched/core.c
>> @@ -396,19 +396,7 @@ static bool set_nr_if_polling(struct task_struct *p)
>> #endif
>> #endif
>>
>> -/**
>> - * wake_q_add() - queue a wakeup for 'later' waking.
>> - * @head: the wake_q_head to add @task to
>> - * @task: the task to queue for 'later' wakeup
>> - *
>> - * Queue a task for later wakeup, most likely by the wake_up_q() call
>> in the
>> - * same context, _HOWEVER_ this is not guaranteed, the wakeup can come
>> - * instantly.
>> - *
>> - * This function must be used as-if it were wake_up_process(); IOW
>> the task
>> - * must be ready to be woken at this location.
>> - */
>> -void wake_q_add(struct wake_q_head *head, struct task_struct *task)
>> +static bool __wake_q_add(struct wake_q_head *head, struct task_struct
>> *task)
>> {
>>     struct wake_q_node *node = &task->wake_q;
>>
>> @@ -422,15 +410,55 @@ void wake_q_add(struct wake_q_head *head, struct
>> task_struct *task)
>>      */
>>     smp_mb__before_atomic();
>>     if (unlikely(cmpxchg_relaxed(&node->next, NULL, WAKE_Q_TAIL)))
>> -        return;
>> -
>> -    get_task_struct(task);
>> +        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.
>> + * @head: the wake_q_head to add @task to
>> + * @task: the task to queue for 'later' wakeup
>> + *
>> + * Queue a task for later wakeup, most likely by the wake_up_q() call
>> in the
>> + * same context, _HOWEVER_ this is not guaranteed, the wakeup can come
>> + * instantly.
>> + *
>> + * This function must be used as-if it were wake_up_process(); IOW
>> the task
>> + * must be ready to be woken at this location.
>> + */
>> +void wake_q_add(struct wake_q_head *head, struct task_struct *task)
>> +{
>> +    if (__wake_q_add(head, task))
>> +        get_task_struct(task);
>> +}
>> +
>> +/**
>> + * wake_q_add_safe() - safely queue a wakeup for 'later' waking.
>> + * @head: the wake_q_head to add @task to
>> + * @task: the task to queue for 'later' wakeup
>> + *
>> + * Queue a task for later wakeup, most likely by the wake_up_q() call
>> in the
>> + * same context, _HOWEVER_ this is not guaranteed, the wakeup can come
>> + * instantly.
>> + *
>> + * This function must be used as-if it were wake_up_process(); IOW
>> the task
>> + * must be ready to be woken at this location.
>> + *
>> + * 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
>> + * queued for wakeup.
>> + */
>> +void wake_q_add_safe(struct wake_q_head *head, struct task_struct *task)
>> +{
>> +    if (!__wake_q_add(head, task))
>> +        put_task_struct(task);
>> }
>>
>> void wake_up_q(struct wake_q_head *head)
>
>Acked-by: Waiman Long <longman@redhat.com>
>