HP-UX Floating-Point Guide

114 Chapter 4

HP-UX Math Libraries on HP 9000 Systems

Contents of the HP-UX Math Libraries

The C Math Library (libm)

The C math library, libm, provides all the math functions speciﬁed by

the ANSI C standard. It also provides

• float versions of many mathematical functions

• Degree-valued trigonometric functions

• A group of functions and macros recommended by the IEEE standard

(see Table 2-12 on page 67), including ﬂoating-point classiﬁcation

macros, and some additional classiﬁcation macros that the ISO/ANSI

C committee has included in the C9X draft standard, the proposed

new C standard

• A group of functions required by the COSE Common API

Speciﬁcation (Spec 1170) and by the XPG4.2 speciﬁcation

• The fenv(5) suite, a collection of functions that allow an application to

manipulate the ﬂoating-point status register. The ISO/ANSI C

committee has included these functions in the C9X draft standard.

• Additional functions and macros that the ISO/ANSI C committee has

included in the C9X draft standard

All of these functions are declared in math.h or fenv.h. However, many

are not speciﬁed by the current ANSI C standard. If you compile in strict

ANSI mode (with the -Aa option), only the declarations of functions

speciﬁed by the ANSI C standard are ordinarily visible to your program.

Therefore, to make the additional functions visible, do one of the

following:

• Compile with the extended ANSI option -Ae, which makes both ANSI

and non-ANSI function declarations visible. This option is the

default, so you need not specify it explicitly:

cc program_name.c -lm

• If your program does not use ANSI C features, compile in

compatibility mode (with the -Ac option), which also makes both

ANSI and non-ANSI function declarations visible:

cc -Ac program_name.c -lm

Do not use this option to compile programs that call float type math

functions, because they depend on ANSI C features. See “ﬂoat Type

Math Functions” on page 115.