On Sat, Nov 05, 2005 at 11:20:47PM -0600, Peter Seebach wrote:
> In message <mtuacgiuzle.fsf@contents-vnder-pressvre.mit.edu>, "Nathan J. Willia
> ms" writes:
> >seebs@plethora.net (Peter Seebach) writes:
> >> On some BSD somewhere, I was once told to use floating point emulation in
> >> the kernel so that the kernel wouldn't have to save and restore floating
> >> point registers as often when processing system calls.
> 
> >Right. By convention, the kernel isn't supposed to use hardware FP
> >resources.
> 
> Okay.
> 
...
> 
> So the kernel already DTRT anyway?  Or is this a separate option, such
> as -msoft-float?

You missed the point.
You aren't allowed to do FP calculations in the kernel.
(or use the FP registers for anything else)
Unless that piece of code explicitly saves them.

	David

-- 
David Laight: david@l8s.co.uk