workqueues and percpu (was: [PATCH] dm: remake of the verity target)

[Mikulas Patocka]

On Tue, 6 Mar 2012, Mandeep Singh Baines wrote:

> You are
> allocated a complete shash_desc per I/O. We only allocate one per CPU.

I looked at it --- and using percpu variables in workqueues isn't safe 
because the workqueue can change CPU if the CPU is hot-unplugged.

dm-crypt has the same bug --- it also uses workqueue with per-cpu 
variables and assumes that the CPU doesn't change for a single work item.

This program shows that work executed in a workqueue can be switched to a 
different CPU.

I'm wondering how much other kernel code assumes that workqueues are bound 
to a specific CPU, which isn't true if we unplug that CPU.

Mikulas

---

/*
 * A proof of concept that a work item executed on a workqueue may change CPU
 * when CPU hot-unplugging is used.
 * Compile this as a module and run:
 * insmod test.ko; sleep 1; echo 0 >/sys/devices/system/cpu/cpu1/online
 * You see that the work item starts executing on CPU 1 and ends up executing
 * on different CPU, usually 0.
 */

#include <linux/module.h>
#include <linux/delay.h>

static struct workqueue_struct *wq;
static struct work_struct work;

static void do_work(struct work_struct *w)
{
	printk("starting work on cpu %d\n", smp_processor_id());
	msleep(10000);
	printk("finishing work on cpu %d\n", smp_processor_id());
}

static int __init test_init(void)
{
	printk("module init\n");
	wq = alloc_workqueue("testd", WQ_MEM_RECLAIM | WQ_CPU_INTENSIVE, 1);
	if (!wq) {
		printk("alloc_workqueue failed\n");
		return -ENOMEM;
	}
	INIT_WORK(&work, do_work);
	queue_work_on(1, wq, &work);
	return 0;
}

static void __exit test_exit(void)
{
	destroy_workqueue(wq);
	printk("module exit\n");
}

module_init(test_init)
module_exit(test_exit)
MODULE_LICENSE("GPL");
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/