Re: RT patch acceptance

[Ingo Molnar]

* Andrea Arcangeli <andrea@xxxxxxx> wrote:

> > you are wrong. This codepath is not running with interrupts disabled on 
> > PREEMPT_RT. irqs-off spinlocks dont turn off interrupts on PREEMPT_RT.  
> 
> Then I'm afraid preempt-RT infringe on the patent [...]

i'd have expected you to say "oops, i was wrong, thanks for the 
explanation", but now you come up with a completely nontechnical topic 
instead?

> > (there are still some ways to introduce latencies into PREEMPT_RT, but 
> > they are not common and we are working on ways to cover them all.)
> 
> How can you schedule a task while a spinlock is held? Ok irqs will 
> keep going, but how can you reschedule risking to deadlock? As long as 
> there are regular spinlocks in any driver out there (i.e. all of them) 
> then you can still introduce latencies. It doesn't seem too uncommon 
> to me to take a spinlock. [...]

yes, and everything works just fine, if it's deadlock-free under normal 
SMP Linux. The key to that is that the whole execution flow is 'flat'.  
I.e. _all_ driver code (except a few notable exceptions that dont count) 
runs in a thread context. So spinlocks are _never_ recursively taken by 
the same thread under PREEMPT_RT - and the scheduler sorts out all the 
inter-thread blocking.

This is the basic concept of PREEMPT_RT. [ I have to say that I'm happy 
that while you have already written 12 emails into this thread with a 
seemingly firm and competent opinion - arguing against various aspects 
of PREEMPT_RT - today you finally seem to have gotten closer to 
understanding its basics ;-) ]

	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/