[bug #68547] Provide fused multiply-add (FMA) function

<[email protected]> Thu, 23 Jul 2026 16:11:11 -0400 (EDT)
Newsgroups gmane.comp.gnu.octave.bugs
Message-ID <[email protected]>
--===============5633842023790191998==
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: base64
Content-Disposition: inline

UGxlYXNlIHVzZSB0aGUgYnVnIHRyYWNrZXIgdG8gcG9zdCB1cGRhdGVzIHRvIGEgYnVnIHJlcG9y
dC4gIFRoZSBtYWlsaW5nIGxpc3QgaXMgaW50ZW5kZWQgYXMgYSByZWFkLW9ubHkgbm90aWZpY2F0
aW9uIHN0cmVhbS4gIEluZm8gcG9zdGVkIHRvIHRoaXMgbWFpbGluZyBsaXN0IGFkZHJlc3Mgd29u
J3QgYXBwZWFyIGluIHRoZSB0cmFja2VyIGRhdGFiYXNlIHdoZXJlIGl0IGlzIG1vc3QgdXNlZnVs
LgoK

--===============5633842023790191998==
Content-Type: MULTIPART/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; BOUNDARY="8323329-424238335-1784837471=:1196930"
Content-Disposition: inline

--8323329-424238335-1784837471=:1196930
Content-Type: TEXT/plain; CHARSET=utf-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Content-Disposition: inline

Follow-up Comment #2, bug #68547 (group octave):=0A=0AI think you're probab=
ly correct that it's not in core MATLAB.  I didn't see=0Aanything on the pa=
ge itself indicating that, but fma is not listed on=0A[https://www.mathwork=
s.com/help/matlab/referencelist.html this page] which I'm=0Aguessing is int=
ended to be a complete list of core functions.  All but ten of=0Afunctions =
linked from there seem to have URL paths starting with=0A"/help/matlab/".=
=0A=0AI don't know how Octave decides whether a function belongs in core vs=
. a=0Aseparate package.  Some points that might argue in favor of having it=
 in=0Acore:=0A* FMA is a basic low-level building block.=0A* FMA is useful =
in a large fraction of numerical computations.=0A* FMA would likely be usef=
ul in some of Octave's own m-file functions.=0A* FMA is a standard part of =
the floating point environment in many other=0Alanguages.=0A* FMA is a requ=
ired operation in IEEE 754.=0A=09* I haven't seen a claim that Octave is ai=
ming for IEEE 754 conformance, but=0Agetting closer still seems valuable.=
=0A=0A=0A=0A    _______________________________________________________=0A=
=0AReply to this item at:=0A=0A  <https://savannah.gnu.org/bugs/?68547>=0A=
=0A_______________________________________________=0AMessage sent via Savan=
nah=0Ahttps://savannah.gnu.org/=0A
--8323329-424238335-1784837471=:1196930
Content-Type: APPLICATION/pgp-signature; name=signature.asc

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQQk97aszIMMAvLLwm6qLAuaBUf3TgUCamJ1XwAKCRCqLAuaBUf3
TlQuAP0SAioJed1rbJet3TcAqeloXIBAU+Re9zUki0ihCvrj9gEAyFtSul07mDYc
YsLmbA4kf1fjyt5+7BpcQgmfmUtjgwY=
=xa68
-----END PGP SIGNATURE-----

--8323329-424238335-1784837471=:1196930--



--===============5633842023790191998==--