Re: tasklet_kill will always hang for recursive tasklets on a UP

[Juergen Quade]

> 	   Thanx for ur inputs. I think that I am missing something in ur 
> explanation. Can u please elaborate. In the meantime, the approach that I 

Maybe it is easier we make it the other way round.
If you look at the code of tasklet_kill:

	void tasklet_kill(struct tasklet_struct *t)
	{
		if (in_interrupt())
			printk("Attempt to kill tasklet from interrupt\n");

		while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) {
			do
				yield();
			while (test_bit(TASKLET_STATE_SCHED, &t->state));
		}
		tasklet_unlock_wait(t);
		clear_bit(TASKLET_STATE_SCHED, &t->state);
	}

Can you explain me, what the last statement 
	clear_bit(TASKLET_STATE_SCHED, &t->state);
is for?

> will like is to have another state TASKLET_STATE_KILLED so the code 
> changes that need to be done are
> 
> void tasklet_kill(struct tasklet_struct *t)
> {
> 
>      ...
>      ...
>      /*
>       * Mark the tasklet as killed, so the next time around
>       * tasklet_action does not call the handler for this tasklet
>       */
>      set_bit(TASKLET_STATE_KILLED, &t->state);  	<-- ADDED
> ...
> 

What I don't like on this approach is, to add another flag (=state) to
the tasklet, which might make the world more complicated as necessary.
I will take some time to think about it, but can't do that today :-(

Beside this, if you can't use a function without looking
at the code and without experimenting with it, that
must lead to bugs! IMHO, here is a call for action.

            Juergen.
-
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/