This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: cheaders=c: Where are overloaded cmath fuctions?
- From: jtc at acorntoolworks dot com (J.T. Conklin)
- To: Danny Smith <danny_r_smith_2001 at yahoo dot co dot nz>
- Cc: libstdc++ <libstdc++ at gcc dot gnu dot org>
- Date: 15 Oct 2002 11:35:11 -0700
- Subject: Re: cheaders=c: Where are overloaded cmath fuctions?
- References: <20020928043851.76153.qmail@web21406.mail.yahoo.com>
- Reply-to: jtc at acorntoolworks dot com
Danny Smith <danny_r_smith_2001@yahoo.co.nz> writes:
> Unlike cheaders=cstd model, cmath in cheaders=c model does not
> contain definitions of std C++98 math functions (except modf)
> overloaded for float and long double nor function versions of C99
> classification macros.
>
> Is there a reason for this? Are they the responsibility of
> C-runtime headers for cheaders=c model?
As a user of a system (QNX) which uses cheaders=c, this is my prefered
interpretation. I've had to comment out the existing abs() and modf()
overloads to avoid conflicts with the system headers. So rather than
add additional overloaded functions, I think the ones that exist should
be removed.
--jtc
--
J.T. Conklin