Messages in this thread | | | Subject | Re: [PATCH] misc. kernel preemption bits | From | Robert Love <> | Date | 29 Aug 2002 14:39:25 -0400 |
| |
On Thu, 2002-08-29 at 14:38, Linus Torvalds wrote:
> I think we should return silently, and simply consider the case of > disabled local interrupts to be equivalent to having preemption disabled. > > So I would remove even the warning. > > Comments?
This is a tough question.
I originally did just that but Ingo said we should aim to find the problem areas, too. The issue is, for 99% of the cases, disabling interrupts really is equivalent to disabling preemption (e.g. preempt_schedule() is never called). For the remaining 1% of the cases, it is possible to fix up the problems by playing safely with interrupts off.
We _must_ return since we are seeing these in the wild. If we want to leave the debug checking to try to "fix" the remaining cases we can do so too.
How about this: add the return now (i.e. accept the patch as-is) and keep the debug check so we can continue to find areas that cause incorrect preemptions. Before 2.6, I will send a patch to remove the check and just return silently.
Sound good?
Robert Love
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |