Re: Priority Inheritance Test (Real-Time Preemption)

[Ingo Molnar]

* Esben Nielsen <simlo@xxxxxxxxxx> wrote:

> > 	task-A			task-B			task-RT
> > 
> > 	spin_lock(&lock2);
> > 	[ gets lock2 ]
> > 				spin_lock(&lock1);
> > 				[ gets lock1 ]
> > 							spin_lock(&lock2);
> > 							[ boosts task-A ]
> > 							[ waits ]
> > 	[ gets RT prio ]				.
> > 	spin_lock(&lock1);				.
> > 	[ boosts task-B ]				.
> > 	[ waits ]					.
> > 	.			[ gets RT prio ]	.
> > 	.			[ 1 msec loop ]		.
> > 	.			spin_unlock(&lock1);	.
> > 	[ gets lock 1 ]					.
> > 				spin_lock(&lock1);	.
> 
> point of disagreement          ----^   

> No :-)

> Why should task B get lock1 the 2. time before the rt-task? That would
> be an error! 

then make it task-C, which tried to take the lock before the RT task
came into the picture. Btw., the above scenario can still happen on SMP.

when task-A unlocks lock1, it can very well give it to task-C - there's
no reason not to do it, task-RT has not expressed any interest in lock1 
yet.

so my example and analysis still stands.

> I can't see how it can produce a flow like the one you describe above!

it can produce such a flow on SMP, or if you add in a third non-RT task
(task-C). Agreed?

In the test where you got 3 msecs you had more than 2 non-RT tasks,
correct?

	Ingo
-
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/