Re: [RFC PATCH] workqueue: introduce queue_delayed_work_on_offline_safe.

[Haakon Bugge <haakon.bugge@oracle.com>]

On 31 Jan 2025, at 12:16, Imran Khan <imran.f.khan@oracle.com> wrote:
> Hello again Tejun,
> sorry, just found one mistake in earlier shared patch, it
> missed an unlock done below:
> 
> Thanks,
> Imran
> On 31/1/2025 9:37 pm, imran.f.khan@oracle.com wrote:
> Hello Tejun,
> 
> [...]
> 
> Could you kindly let me know, if it would be acceptable, to have
> queue_delayed_work_on_offline_safe, as a wrapper around
> queue_delayed_work_on, such that it can check and ensure CPU's
> availability. If it can't, then it can simply return false and let
> caller decide which cpu to use next. Something like below:
> 
> 
> diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
> index b0dc957c3e560..57f39807f3bf1 100644
> --- a/include/linux/workqueue.h
> +++ b/include/linux/workqueue.h
> @@ -589,6 +589,9 @@ extern bool queue_work_node(int node, struct workqueue_struct *wq,
>                            struct work_struct *work);
> extern bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq,
>                        struct delayed_work *work, unsigned long delay);
> +extern bool queue_delayed_work_on_offline_safe(int cpu,
> +                       struct workqueue_struct *wq, struct delayed_work *work,
> +                       unsigned long delay);

Hi Imran,


I am not quite sure this signature will be OK. See below.

> extern bool mod_delayed_work_on(int cpu, struct workqueue_struct *wq,
>                        struct delayed_work *dwork, unsigned long delay);
> extern bool queue_rcu_work(struct workqueue_struct *wq, struct rcu_work *rwork);
> 
> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> index 9362484a653c4..7d3b8050422e4 100644
> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c
> @@ -2565,6 +2565,37 @@ bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq,
> }
> EXPORT_SYMBOL(queue_delayed_work_on);
> 
> +/**
> + * queue_delayed_work_on_offline_safe - queue work on specific online CPU after
> + *                                     delay,
> + *
> + * @cpu: CPU number to execute work on
> + * @wq: workqueue to use
> + * @dwork: work to queue
> + * @delay: number of jiffies to wait before queueing
> + *
> + * a wrapper, around queue_delayed_work_on, that checks and ensures that
> + * specified @cpu is online. If @cpu is found to be offline or if its online
> + * status can't be reliably determined, return false and leave the decision,
> + * of selecting new cpu for delayed_work, to caller.

The return value here is ambiguous.

> + *
> + */
> +bool queue_delayed_work_on_offline_safe(int cpu, struct workqueue_struct *wq,
> +                          struct delayed_work *dwork, unsigned long delay)
> +{
> +       bool ret = false;
> +       int locked = 0;
> +
> +       if ((locked = cpus_read_trylock()) && cpu_online(cpu)) {
> +               ret = queue_delayed_work_on(cpu, wq, dwork, delay);

Now, ret will be false if the work was already queued.

>                  cpus_read_unlock();
> +       } else if (locked)
> +               cpus_read_unlock();
> +
> +       return ret;

If false is returned, it is a) because the designated @cpu was reliable detected online but the work was already queued or b) because the designated @cpu could not reliable be detected online.

Hence, I think you need to distinguish these cases. I suggest to keep the bool return value to mean what it does for queue_delayed_work_on() and add a bool pointer as argument which can be used to determine reliable online detection of @cpu or not.

A nit is that you should have braces on both the if/else clauses is any of them have it. But even simpler, do not use "else if" but an ordinary "if (locked)", and this way, you can remove the first cpus_read_unlock().


Thxs, Håkon


> +}
> +EXPORT_SYMBOL(queue_delayed_work_on_offline_safe);
> +
> +
> 
> If this looks acceptable to you, I can send a v2 of earlier patch.
> 
> Thanks,
> Imran