Bill Studenmund wrote:

> On Wed, Mar 10, 2004 at 03:53:33PM +0100, Aymeric Vincent wrote:
>>
>> The other reason is because it looks bad to me to force a user of the
>> locking or (worse) pool API to know about SPLs.
>
> Unfortunately for pools and locks that are used in interrupt context, 
> the
> user _does_ need to know about SPL levels. Otherwise bad things happen 
> in
> SMP systems.

Yes, definitely. But, only for those ones. That's why I suggested two 
different init functions, so that callers which need not be aware of 
anything spl-related are not required to.

But the point of Darrin that his goal is to force people to think about 
spl-problems whenever they use pools is fine, too.

So I won't argue much on the subject.

Regards,
  Aymeric