工作队列(workqueue)

转载于：http://blog.csdn.net/angle_birds/article/details/8448070

项目需要，在驱动模块里用内核计时器timer_list实现了一个状态机。
郁闷的是，运行时总报错“Scheduling while atomic”，网上搜了一下：

"Scheduling while atomic" indicates that you've tried to sleep somewhere that you shouldn't - like within a spinlock-protected critical section or an interrupt handler.

改进程序，在计时器里使用了workqueue，搞定问题。顺便把workqueue的实现代码总结了一下


一、workqueue简介

workqueue与tasklet类似，都是允许内核代码请求某个函数在将来的时间被调用（抄《ldd3》上的），每个workqueue就是一个内核进程。

workqueue与tasklet的区别：
   1.tasklet是通过软中断实现的，在软中断上下文中运行，tasklet代码必须是原子的，workqueue是通过内核进程实现的，就没有上述限制的，最爽的是，工作队列函数可以休眠 
         
     PS: 我的驱动模块就是印在计时器中调用了可休眠函数，所以出现了cheduling while atomic告警
         内核计时器也是通过软中断实现的

   2.tasklet始终运行在被初始提交的同一处理器上，workqueue不一定

   3.tasklet不能确定延时时间（即使很短），workqueue可以设定延迟时间


二、workqueue的API

 
workqueue的API自2.6.20后发生了变化

#include <linux/workqueue.h>
struct workqueue_struct;
struct work_struct;

struct workqueue_struct *create_workqueue(const char *name);
void destroy_workqueue(struct workqueue_struct *queue);

INIT_WORK(_work, _func);
INIT_DELAYED_WORK(_work, _func);

int queue_work(struct workqueue_struct *wq, struct work_struct *work);
int queue_delayed_work(struct workqueue_struct *wq,struct delayed_work *dwork, unsigned long delay);

int queue_delayed_work_on(int cpu, struct workqueue_struct *wq,
            struct delayed_work *dwork, unsigned long delay);

int cancel_work_sync(struct work_struct *work);
int cancel_delayed_work_sync(struct delayed_work *dwork);

void flush_workqueue(struct workqueue_struct *wq);

(代码基于2.6.24)

struct my_work_stuct{
    int test;
    struct work_stuct save;
};

struct my_work_stuct test_work; 
struct workqueue_struct *test_workqueue;

void do_save(struct work_struct *p_work)
{
    struct my_work_struct *p_test_work = container_of(p_work, struct my_work_stuct, save);
    printk("%d\n",p_test_work->test);
}
  
void test_init()
{
    test_workqueue = create_workqueue("test_workqueue");
    if (!test_workqueue)
        panic("Failed to create test_workqueue\n");

    INIT_WORK(&(test_work.save), do_save);

    queue_work(test_workqueue, &(test_work.save));
}

void test_destory(void)
{
    if(test_workqueue)
        destroy_workqueue(test_workqueue);
}

三、workqueue的实现


工作队列workqueue不是通过软中断实现的，它是通过内核进程实现的

首先，创建一个workqueue，实际上就是建立一个内核进程

create_workqueue("tap_workqueue")
--> __create_workqueue(“tap_workqueue”, 0, 0)
--> __create_workqueue_key((name), (singlethread), (freezeable), NULL, NULL){
         wq = kzalloc(sizeof(*wq), GFP_KERNEL);
         wq->cpu_wq = alloc_percpu(struct cpu_workqueue_struct);
         wq->name = name;
         wq->singlethread = singlethread;
         wq->freezeable = freezeable;
         INIT_LIST_HEAD(&wq->list);

         for_each_possible_cpu(cpu) {
             cwq = init_cpu_workqueue(wq, cpu);
             err = create_workqueue_thread(cwq, cpu);
             start_workqueue_thread(cwq, cpu);
         }
    }

create_workqueue_thread 建立了一个内核进程 worker_thread（linux_2_6_24/kernel/workqueue.c）

create_workqueue_thread(struct cpu_workqueue_struct *cwq, int cpu)
{
    struct workqueue_struct *wq = cwq->wq;

    const char *fmt = is_single_threaded(wq) ? "%s" : "%s/%d";
    struct task_struct *p;

    p = kthread_create(worker_thread, cwq, fmt, wq->name, cpu);

    if (IS_ERR(p))
        return PTR_ERR(p);

    cwq->thread = p;

    return 0;
}

（linux_2_6_24/kernel/workqueue.c）

int worker_thread (void *__cwq)
{
    struct cpu_workqueue_struct *cwq = __cwq;
    /*下面定义等待队列项*/
    DEFINE_WAIT(wait);

    /*下面freezeable一般为0*/
    if (cwq->wq->freezeable)
        set_freezable();

    /*提高优先级别*/
    set_user_nice(current, -5);

    for (;;) {
        /*在cwq->more_work上等待, 若有人调用queue_work,该函数将调用wake_up(&cwq->more_work) 激活本进程*/
        prepare_to_wait(&cwq->more_work, &wait, TASK_INTERRUPTIBLE);

        /*work队列空则切换出去*/
        if (!freezing(current) && !kthread_should_stop() && list_empty(&cwq->worklist))
            schedule();


        /*切换回来则结束等待 说明有人唤醒cwq->more_work上的等待 有work需要处理*/
        finish_wait(&cwq->more_work, &wait);

        /*下面空，因为没有定义电源管理*/
        try_to_freeze();

        if (kthread_should_stop())
            break;

        /*run_workqueue依次处理工作队列上所有的work*/
        run_workqueue(cwq);
    }
    return 0;
}


/*run_workqueue依次处理工作队列上所有的work*/
static void run_workqueue(struct cpu_workqueue_struct *cwq)
{
    spin_lock_irq(&cwq->lock);
    cwq->run_depth++;
    if (cwq->run_depth > 3) {
        /* morton gets to eat his hat */
        printk("%s: recursion depth exceeded: %d\n",
            __FUNCTION__, cwq->run_depth);
        dump_stack();
    }

    while (!list_empty(&cwq->worklist)) {
        struct work_struct *work = list_entry(cwq->worklist.next,
                        struct work_struct, entry);
        work_func_t f = work->func;
#ifdef CONFIG_LOCKDEP
        /*
         * It is permissible to free the struct work_struct
         * from inside the function that is called from it,
         * this we need to take into account for lockdep too.
         * To avoid bogus "held lock freed" warnings as well
         * as problems when looking into work->lockdep_map,
         * make a copy and use that here.
         */
        struct lockdep_map lockdep_map = work->lockdep_map;
#endif

        cwq->current_work = work;
        list_del_init(cwq->worklist.next);
        spin_unlock_irq(&cwq->lock);

        BUG_ON(get_wq_data(work) != cwq);
        work_clear_pending(work);
        lock_acquire(&cwq->wq->lockdep_map, 0, 0, 0, 2, _THIS_IP_);
        lock_acquire(&lockdep_map, 0, 0, 0, 2, _THIS_IP_);

        f(work); /*执行work项中的func*/
       
        lock_release(&lockdep_map, 1, _THIS_IP_);
        lock_release(&cwq->wq->lockdep_map, 1, _THIS_IP_);

        if (unlikely(in_atomic() || lockdep_depth(current) > 0)) {
            printk(KERN_ERR "BUG: workqueue leaked lock or atomic: "
                    "%s/0x%08x/%d\n",
                    current->comm, preempt_count(),
                        task_pid_nr(current));
            printk(KERN_ERR " last function: ");
            print_symbol("%s\n", (unsigned long)f);
            debug_show_held_locks(current);
            dump_stack();
        }

        spin_lock_irq(&cwq->lock);
        cwq->current_work = NULL;
    }
    cwq->run_depth--;
    spin_unlock_irq(&cwq->lock);
}

  int fastcall queue_work(struct workqueue_struct *wq, struct work_struct *work)
{
    int ret = 0;

    if (!test_and_set_bit(WORK_STRUCT_PENDING, work_data_bits(work))) {
        BUG_ON(!list_empty(&work->entry));
        __queue_work(wq_per_cpu(wq, get_cpu()), work);
        put_cpu();
        ret = 1;
    }
    return ret;
} 


/* Preempt must be disabled. */
static void __queue_work(struct cpu_workqueue_struct *cwq, struct work_struct *work)
{
    unsigned long flags;

    spin_lock_irqsave(&cwq->lock, flags);
    insert_work(cwq, work, 1);
    spin_unlock_irqrestore(&cwq->lock, flags);
}

static void insert_work(struct cpu_workqueue_struct *cwq,
                struct work_struct *work, int tail)
{
    set_wq_data(work, cwq);
    /*
     * Ensure that we get the right work->data if we see the
     * result of list_add() below, see try_to_grab_pending().
     */
    smp_wmb();
    if (tail)
        list_add_tail(&work->entry, &cwq->worklist);
    else
        list_add(&work->entry, &cwq->worklist);
    wake_up(&cwq->more_work);
}

使用共享队列，有这样一套API

int schedule_work(struct work_struct *work)
{
    queue_work(keventd_wq, work);
}

int schedule_delayed_work(struct delayed_work *dwork,unsigned long delay)
{
    timer_stats_timer_set_start_info(&dwork->timer);
    return queue_delayed_work(keventd_wq, dwork, delay);
}

void flush_scheduled_work(void)
{
    flush_workqueue(keventd_wq);
}