On Fri, 27 Oct 2000, Ignatios Souvatzis wrote:

> On Thu, Oct 26, 2000 at 03:48:52PM -0400, Todd Vierling wrote:
> > On Thu, 26 Oct 2000, David Brownlee wrote:
> >
> > : 	Presumably -march=strongarm will work similarly to armv4, but
> > : 	is less correct (armv4 includes FL_CO_PROC)?
> > :
> > : 	SHARK already had -march=strongarm, which should probably also
> > : 	be switched.
> >
> > I believe the shark can handle halfword instructions properly...?
> >
> > That much aside, the Shark has no coprocessor.  It should be using
> > softfloat, as well.
>
	Does FL_CO_PROC make any effective difference apart from
	the floating point case?

> I'm pretty sure (but check the docs yourself) that no SA-110 _can_ have
> any coprocessor (besides the internal system control coprocessor) because
> it doesn't have an external coprocessor bus.
>
> Don't know about SA-1100.

	IIRC arm32 uses a softfloat library already, but I didn't think
	that precluded the use of floating point hardware (with the
	addition of an alternative libm).

	Given the increasing availability of arm hardwith with floating
	point does it make sense to go the i386 route of a default libm
	with softfloat and an alternative hardfloat library used if
	sysctl machdep.fpu_present is true?

                David/absolute
			       -- www.netbsd.org: A pmap for every occasion --