[PATCH] hypotl, cabsl, and code removal in cabs

Bruce Evans brde at optusnet.com.au
Sun Dec 9 21:29:50 PST 2007

On Sun, 9 Dec 2007, David Schultz wrote:

> On Fri, Dec 07, 2007, Bruce Evans wrote:
> [...]
>> Anyway, the software version of sqrtl is irrelevant on
>> athlon-xp, since athlon-xp has sqrtl in hardware (takes 35 cycles).
>> Similarly for amd64, ia64 and possibly sparc64 (sparc64 has sqrt in
>> hardware so it hopefully has sqrtl in hardware).  arm and powerpc
>> apparently have long double == double, so the software version of sqrtl
>> is apparently only needed on ia64.
> In general, even if we don't use it on any architectures that
> FreeBSD currently supports, I'd like to have a working MI
> implementation checked in before we add a bunch of MD
> versions. Otherwise we risk overburdening people trying to port
> FreeBSD to new architectures. Having an MI version makes
> cross-testing easier, too.

I agree.  I could write amd64 and i386 versions in asm the most important
functions in about 5 minutes each, but have refrained from doing so since 
I think this would get in the way of general support.

> For ia64, we can use Intel's BSD-licensed math lib.

Sure.  I use it a hacked up version of the glibc version of it locally
as a gold standard, but don't want to maintain it.  Is is fully


More information about the freebsd-standards mailing list