BH implementation question

[Anders Peter Fugmann <afu@fugmann.dhs.org>]

Hi I have a couple of questions to the kernel code.

I have been trying to fully inderstand (and doccument) the changes in 
2.4 wrt. Tasklets and softirq's, BH's and task queues.

In my try to understand how it all works, I came across the code:

(linux/kernel/softirq.c: 246)

static void bh_action(unsigned long nr)
{
	int cpu = smp_processor_id();

	if (!spin_trylock(&global_bh_lock))
		goto resched;

	if (!hardirq_trylock(cpu))
		goto resched_unlock;

	if (bh_base[nr])
		bh_base[nr]();

	hardirq_endlock(cpu);
	spin_unlock(&global_bh_lock);
	return;

resched_unlock:
	spin_unlock(&global_bh_lock);
resched:
	mark_bh(nr);
}

Now all of this but the hardirq_trylock(cpu) and hardirq_endlock(cpu) 
makes perfectly sence.

Anyone care to explain the what theese lines do.

Secondly.

Is there a reason why to implement a queue (TASKLET_HI) for the old 
BH's, instead of just using a single tasklet for all BH administraton. 
Would'ent this guarentee that no BH is executed at the same time, and at 
the same time reduce code complexity, and remove the global_bh_lock?


TIA
Anders Fugmann



-- 
Hi. I'm a .signature virus.
Please copy me into your .signature file and help me spread.