On Fri, 18 Sep 1998 17:52:14 -0700 (PDT), David Brownlee wrote:
> 
> On Sat, 19 Sep 1998, Ken Nakata wrote:
> > 
> > Even if so, it does not scale *at all*.  So, in Linux or even FreeBSD,
> > maybe, but I wouldn't want an ugly hack like this in NetBSD!
> > 
> 	If its in an LKM for a DSP chip then I dont see a real problem
> 	a) It doesn't need to scale - just work for that DSP chip.

That's not what I meant.  If we are going to allow this particular
device driver to use FPU in kernel, why not another?  I mean, what if
someone else wants to use FPU in his XXX driver for whatever reason?
That's what I meant by "not scale".  IOW, the proposed "solution" does
not scale to multiple instances of such FPU-requiring drivers.

> 	b) It doesn't impact any other part of the kernel source.

AS LONG AS there's no other driver which wants to use FPU.

I think the idea of letting a device driver use FPU since there's
currently no other such thing just stinks.

Ken