Message-Id:	<1.5.4.16.19960731074038.1e4fe9fe@vub.vub.ac.be>
Mime-Version:	1.0
Date:	Wed, 31 Jul 1996 09:40:38 +0200
To:	djgpp-workers AT delorie DOT com
From:	Pieter Vankeerberghen <pietervk AT minf DOT vub DOT ac DOT be>
Subject:	Re: long double

OK, last night, I recompiled libc from DJLSR200.zip (going the hard way):
Moshiers ioldoubl.o is 16232 bytes when compiled as -O2.
libc goes from 536958 to 553582 kbytes. Moshier's ieetst (IEEE test) program
goes from 62976 to 74240 bytes. BC3.1 (-386/387, 16 bit, no optim, no debug)
is 42236 bytes but fails on long double support.

Reading the mails on long double support, I think it is a good idea
to include all double support in libc, and nothing more. Thus no "Le%"
which will avoid confusion.

On the other hand, the full IEEE 754 compliant routines for double and
long double (also "Le%") go in libm. Basically, non-numerical users
don't link libm. This is how I would arrange my code, but since 
I'm no contributor, I can only speak for myself.

kind regards !

Pieter Vankeerberghen

----------------------------------------------------------------------------
Pieter Vankeerberghen                       tel: +00 32 2 477 43 29
Fabi/ChemoAc, Vrije Universiteit Brussel    fax: +00 32 2 477 47 35
Laarbeeklaan 103, 1090 Brussels, Belgium    email: pietervk AT vub DOT vub DOT ac DOT be
----------------------------------------------------------------------------

- Raw text -