前面的(de)博文中, 大(dà)年夜緻分散介紹了(le／liǎo)一些中斷相幹的(de)器械, 然則對軟中斷的(de)部分沒有細心介紹, 在(zài)這(zhè)裏同一總結一下.
中斷上(shàng)半部的(de)處理,彙編到(dào)request_irq的(de)handle之(zhī)間的(de)過程.
http://blog.csdn.net/jackjones_008/article/details/42387241
MIPS平台的(de)一點記錄
http://blog.csdn.net/jackjones_008/article/details/41945909
tasklet/workqueue的(de)介紹鄙人(rén)面這(zhè)篇博文中有比較具體的(de)介紹.
http://blog.csdn.net/jackjones_008/article/details/42295411
中斷的(de)處理, 起首當然是(shì)硬件中斷産生後彙編部分的(de)處理, 然後在(zài)hardirq處理完之(zhī)後, 在(zài)函數irq_exit中, 如不(bù)雅發明當前沒有hardirq并且softirq沒有被禁止,
也(yě)沒有pending的(de)本地(dì / de)softirq, 則會顯示的(de)調用do_softirq 去處理softirq. 在(zài)處理softirq的(de)過程中, 是(shì)會樊籬硬件中斷的(de).

asmlinkage void do_softirq(void)
{
	__u32 pending;
	unsigned long flags;

	if (in_interrupt())
		return;

	local_irq_save(flags);

	pending = local_softirq_pending();

	if (pending)
		__do_softirq();

	local_irq_restore(flags);
}

在(zài)__do_softirq 琅绫擎, 在(zài)獲得pending的(de)softirq後, 又會打開硬件中斷. 然後會去依次處理pending 的(de)softirq.
這(zhè)個(gè)依次, 其實就(jiù)是(shì)根據enum中的(de)定義大(dà)年夜 HI_SOFTIRQ 履行到(dào) RCU_SOFTIRQ.
再多煩瑣一下, 爲(wéi / wèi)什麽會有pending的(de)softirq呢, 因爲(wéi / wèi)代碼有些處所去調用了(le／liǎo) __raise_softirq_irqoff .
我們還可以(yǐ)留意到(dào), __do_softirq 琅绫擎定義了(le／liǎo)一個(gè)max_restart, 這(zhè)個(gè)值是(shì)10, 是(shì)體系的(de)一個(gè)折中, 處理軟中斷, 然則不(bù)克不(bù)及過度的(de)影響其他(tā)過程的(de)履行.
完成不(bù)了(le／liǎo)的(de), 可以(yǐ)交給 ksoftirqd 這(zhè)個(gè)内核thread去做.

asmlinkage void __do_softirq(void)
{
	struct softirq_action *h;
	__u32 pending;
	int max_restart = MAX_SOFTIRQ_RESTART;
	int cpu;

	pending = local_softirq_pending();
	account_system_vtime(current);

	__local_bh_disable((unsigned long)__builtin_return_address(0));
	trace_softirq_enter();

	cpu = smp_processor_id();
restart:
	/* Reset the pending bitmask before enabling irqs */
	set_softirq_pending(0);

	local_irq_enable();

	h = softirq_vec;

	do {
		if (pending & 1) {
			int prev_count = preempt_count();

			h->action(h);

			if (unlikely(prev_count != preempt_count())) {
				printk(KERN_ERR "huh, entered softirq %td %p"
				       "with preempt_count %08x,"
				       " exited with %08x?\n", h - softirq_vec,
				       h->action, prev_count, preempt_count());
				preempt_count() = prev_count;
			}

			rcu_bh_qsctr_inc(cpu);
		}
		h++;
		pending >>= 1;
	} while (pending);

	local_irq_disable();

	pending = local_softirq_pending();
	if (pending && --max_restart)
		goto restart;

	if (pending)
		wakeup_softirqd();

	trace_softirq_exit();

	account_system_vtime(current);
	_local_bh_enable();
}

enum
{
	HI_SOFTIRQ=0,
	TIMER_SOFTIRQ,
	NET_TX_SOFTIRQ,
	NET_RX_SOFTIRQ,
	BLOCK_SOFTIRQ,
	TASKLET_SOFTIRQ,
	SCHED_SOFTIRQ,
	HRTIMER_SOFTIRQ,
	RCU_SOFTIRQ,	/* Preferable RCU should always be the last softirq */

	NR_SOFTIRQS
};

接上(shàng), restart 中處理完後, 還有pending的(de)softirq的(de)話, 就(jiù)交給内核過程 ksoftirqd 去做. 這(zhè)個(gè)内核過程也(yě)揮菰式的(de)調用 do_softirq 去處理.

static int ksoftirqd(void * __bind_cpu)
{
	set_current_state(TASK_INTERRUPTIBLE);

	while (!kthread_should_stop()) {
		preempt_disable();
		if (!local_softirq_pending()) {
			preempt_enable_no_resched();
			schedule();
			preempt_disable();
		}

		__set_current_state(TASK_RUNNING);

		while (local_softirq_pending()) {
			/* Preempt disable stops cpu going offline.
			   If already offline, we'll be on wrong CPU:
			   don't process */
			if (cpu_is_offline((long)__bind_cpu))
				goto wait_to_die;
			do_softirq();
			preempt_enable_no_resched();
			cond_resched();
			preempt_disable();
			rcu_qsctr_inc((long)__bind_cpu);
		}
		preempt_enable();
		set_current_state(TASK_INTERRUPTIBLE);
	}
	__set_current_state(TASK_RUNNING);
	return 0;

wait_to_die:
	preempt_enable();
	/* Wait for kthread_stop */
	set_current_state(TASK_INTERRUPTIBLE);
	while (!kthread_should_stop()) {
		schedule();
		set_current_state(TASK_INTERRUPTIBLE);
	}
	__set_current_state(TASK_RUNNING);
	return 0;
}

這(zhè)時(shí), 又弗成避免的(de)回到(dào)了(le／liǎo)中斷上(shàng)半部和(hé / huò)下半部的(de)話題, 上(shàng)半部我們可以(yǐ)懂得爲(wéi / wèi)一向履行到(dào) request_irq 中注冊的(de)那個(gè) handle 爲(wéi / wèi)止. 而(ér)softirq 開端的(de)處理都是(shì)屬于(yú)下半部的(de)處理.
鄙人(rén)半部的(de)處理中, 根據睡眠與否可以(yǐ)應用tasklet和(hé / huò)workqueue的(de)機制.
其實tasklet也(yě)是(shì)屬于(yú)softirq的(de)一種, 并且弗成睡眠. workqueue的(de)機制則是(shì)交給了(le／liǎo)内核thread去履行, 可以(yǐ)許可睡眠.