On Jun 24, 2005, at 6:28 PM, Hubert Feyrer wrote:

> On Mon, 30 May 2005, Christos Zoulas wrote:
>
>> Modified Files:
>>     src/sys/ufs/ffs: ffs_softdep.c
>>
>> Log Message:
>> rename delay because it is a function on sparc.
>>
>
> Wasn't it agreed that MI variables take precedence over MD  
> variables? I think there was something about ncpus on sparc64, and  
> sparc64 was changed.
> I think that should happen here, too.

No, delay() has been "maybe a real function" for a long time... it's  
an MI function name, but has an MD implementation.

-- thorpej