Messages in this thread | | | Date | Sun, 12 Oct 2008 19:12:11 -0400 | From | Arjan van de Ven <> | Subject | Re: when spin_lock_irq (as opposed to spin_lock_irqsave) is appropriate? |
| |
On Sun, 12 Oct 2008 15:48:00 +0400 Andrey Borzenkov <arvidjaar@mail.ru> wrote:
> On Saturday 11 October 2008, Arjan van de Ven wrote: > [... very useful explanation omitted ...] > > Does this answer your question? > > > > As Oliver pointed out, part of confusion wa my asumption that _irqsave > verion saves actual interrupt mask. It actually does not. > > This leaves me with a question - how can I know whether interrupts may > (not) be disabled at particular point?
the _irq versions mask the interrupts in the *cpu*! Not in the hw. All CPUs have a flag that says "don't give me interrupts right now please", and the spin_lock_irq(save) functions work on that flag. And they block all interrupts (except NMI's, which are very special)
> In particular, is it safe to > assume that any place marked at "code may sleep" has interrupts > enabled?
yes. That's a good rule of thumb ;-) Anything else is a lot of "depends"
-- Arjan van de Ven Intel Open Source Technology Centre For development, discussion and tips for power savings, visit http://www.lesswatts.org
| |