Re: [PATCH] swap migration: Fix lru drain

[Nick Piggin <piggin@cyberone.com.au>]

Christoph Lameter wrote:
> isolate_page() currently uses an IPI to notify other processors that the lru
> caches need to be drained if the page cannot be found on the LRU. The IPI
> interrupt may interrupt a processor that is just processing lru requests
> and cause a race condition.
> 
> This patch introduces a new function run_on_each_cpu() that uses the keventd()
> to run the LRU draining on each processor. Processors disable preemption
> when dealing the LRU caches (these are per processor) and thus executing
> LRU draining from another process is safe.
> 

Couple of comments:

> ===================================================================
> --- linux-2.6.15-rc5-mm1.orig/kernel/workqueue.c	2005-12-05 11:15:24.000000000 -0800
> +++ linux-2.6.15-rc5-mm1/kernel/workqueue.c	2005-12-06 17:50:44.000000000 -0800
> @@ -424,6 +424,19 @@ int schedule_delayed_work_on(int cpu,
>  	return ret;
>  }
>  
> +void schedule_on_each_cpu(void (*func) (void *info), void *info)
> +{
> +	int cpu;
> +	struct work_struct * work = kmalloc(NR_CPUS * sizeof(struct work_struct), GFP_KERNEL);
> +

Do we need a lock_cpu_hotplug() around here?

> +	for_each_online_cpu(cpu) {
> +		INIT_WORK(work + cpu, func, info);
> +		__queue_work(per_cpu_ptr(keventd_wq->cpu_wq, cpu), work + cpu);
> +	}
> +	flush_workqueue(keventd_wq);
> +	kfree(work);
> +}
> +

Can't this deadlock if 2 CPUs each send work to the other?

-- 
SUSE Labs, Novell Inc.