Very Computer

by **Steve Summ** » Sun, 28 Mar 1999 04:00:00

[I've crossposted this followup -- beware!]

> I'm learning C on a Linux system with gcc 2.7...
> When I try to calculate sin (x) the compiler says:
> undefined reference to 'sin'
> although I wrote
> #include <math.h>
> at the beginning.

Bjoern: you need to use the -lm flag at the end of the command
line when compiling. You might also want to get ahold of a copy
of the comp.lang.c FAQ list. It answers this question of yours
(question 14.3), so it just might answer your next question, too.

Linux and gcc folks: I know I've asked this before, but why can't
this be fixed in glibc? In 1978 on the PDP-11, there might have
been good reason to leave the math functions out of libc.a, but
surely we could afford to place them there today, especially
since the ANSI C standard says they're just as standard as the
rest of the standard library functions. As it is, gcc+glibc
(and therefore C under Linux) is unnecessarily noncompliant,
unless you argue that "gcc ... -lm" is the standard invocation.

Steve Summit

by **Brian McCaule** » Tue, 30 Mar 1999 04:00:00

> [I've crossposted this followup -- beware!]

> > I'm learning C on a Linux system with gcc 2.7...
> > When I try to calculate sin (x) the compiler says:
> > undefined reference to 'sin'
> > although I wrote
> > #include <math.h>
> > at the beginning.

> Bjoern: you need to use the -lm flag at the end of the command
> line when compiling. You might also want to get ahold of a copy
> of the comp.lang.c FAQ list. It answers this question of yours
> (question 14.3), so it just might answer your next question, too.

> Linux and gcc folks: I know I've asked this before, but why can't
> this be fixed in glibc? In 1978 on the PDP-11, there might have
> been good reason to leave the math functions out of libc.a, but
> surely we could afford to place them there today, especially
> since the ANSI C standard says they're just as standard as the
> rest of the standard library functions. As it is, gcc+glibc
> (and therefore C under Linux) is unnecessarily noncompliant,
> unless you argue that "gcc ... -lm" is the standard invocation.

I agree. I accept that "gcc ... -lm" *is* the standard but we've
rolled "nls", "socket", "resolv" and so on into "c" why not "m" too?

--

. _\\__[oo faeces from | Phones: +44 121 471 3789 (home)

. l___\\ /~~) /~~[ / [ | PGP-fp: D7 03 2A 4B D8 3A 05 37...
# ll l\\ ~~~~ ~ ~ ~ ~ | http://www.wcl.bham.ac.uk/~bam/
###LL LL\\ (Brian McCauley) |

Very Computer

#include <math.h> - problem

#include <math.h> - problem

#include <math.h> - problem