Re: [PATCH v5 UPDATEDv2 3/3] CPU hotplug, smp: Flush any pending IPI callbacks before CPU offline

[Oleg Nesterov <oleg@redhat.com>]

On 05/19, Srivatsa S. Bhat wrote:
>
> However, an IPI sent much earlier might arrive late on the target CPU
> (possibly _after_ the CPU has gone offline) due to hardware latencies,
> and due to this, the smp-call-function callbacks queued on the outgoing
> CPU might not get noticed (and hence not executed) at all.

OK, but

> +void flush_smp_call_function_queue(void)
> +{
> +	struct llist_head *head;
> +	struct llist_node *entry;
> +	struct call_single_data *csd, *csd_next;
> +
> +	WARN_ON(!irqs_disabled());
> +
> +	head = &__get_cpu_var(call_single_queue);
> +
> +	if (likely(llist_empty(head)))
> +		return;
> +
> +	entry = llist_del_all(head);
> +	entry = llist_reverse_order(entry);
> +
> +	llist_for_each_entry_safe(csd, csd_next, entry, llist) {
> +		csd->func(csd->info);
> +		csd_unlock(csd);
> +	}
> +}

why do we need it? Can't multi_cpu_stop() just call
generic_smp_call_function_single_interrupt() ? This cpu is still online,
we should not worry about WARN_ON(!cpu_online()) ?

Oleg.