Index

A

abort on exception

C example, 1

accessible documentation, 1

accuracy, 1

floating-point operations, 1

significant digits (number of), 1

threshold, 1

adb, 1

addrans

random number utilities, 1

argument reduction

trigonometric functions, 1

B

base conversion

base 10 to base 2, 1

base 2 to base 10, 1

formatted I/O, 1

Bessel functions, 1

C

C driver

example, call FORTRAN subroutines from C, 1

clock speed, 1

compilers, accessing, 1

conversion between number sets, 1

conversions between decimal strings and binary floating-point numbers, 1

convert_external

binary floating-point, 1

data conversion, 1

D

data types

relation to IEEE formats, 1

dbx, 1

decimal representation

maximum positive normal number, 1

minimum positive normal number, 1

precision, 1

ranges, 1

documentation index, 1

double-precision representation

C example, 1

FORTRAN example, 1

E

errno.h

define values for errno, 1

examine the accrued exception bits

C example, 1

examine the accrued exception flags

C example, 1

F

-fast, 1

floating-point

exceptions list, 1

rounding direction, 1

rounding precision, 1

tutorial, 1

floating-point accuracy

decimal strings and binary floating-point numbers, 1

floating-point exceptions, 1

abort on exceptions, 1

accrued exception bits, 1

common exceptions, 1

default result, 1

definition, 1

flags, 1

accrued, 1

current, 1

ieee_functions, 1

ieee_retrospective, 1

list of exceptions, 1

priority, 1

trap precedence, 1

floating-point options, 1

floating-point queue (FQ), 1

floating-point status register (FSR), 1, 2

floatingpoint.h

define handler types

C and C++, 1

flush to zero (see Store 0), 1

fmod, 1

-fnonstd, 1

G

generate an array of numbers

FORTRAN example, 1

Goldberg paper, 1

abstract, 1

acknowledgments, 1

details, 1

IEEE standard, 1

IEEE standards, 1

introduction, 1

references, 1

rounding error, 1

summary, 1

systems aspects, 1

gradual underflow

error properties, 1

H

HUGE

compatibility with IEEE standard, 1

HUGE_VAL

compatibility with IEEE standard, 1

I

IEEE double extended format

biased exponent

x86 architecture, 1

bit-field assignment

x86 architecture, 1

fraction

x86 architecture, 1

Inf

SPARC architecture, 1

x86 architecture, 1

NaN

x86 architecture, 1

normal number

SPARC architecture, 1

x86 architecture, 1

quadruple precision

SPARC architecture, 1

sign bit

x86 architecture, 1

significand

explicit leading bit (x86 architecture), 1

subnormal number

SPARC architecture, 1

x86 architecture, 1

IEEE double format

biased exponent, 1

bit patterns and equivalent values, 1

bit-field assignment, 1

denormalized number, 1

fraction, 1

storage on SPARC, 1

storage on x86, 1

implicit bit, 1

Inf, infinity, 1

NaN, not a number, 1

normal number, 1

precision, 1

sign bit, 1

significand, 1

subnormal number, 1

IEEE formats

relation to language data types, 1

IEEE single format

biased exponent, 1

biased exponent,implicit bit, 1

bit assignments, 1

bit patterns and equivalent values, 1

bit-field assignment, 1

denormalized number, 1

fraction, 1

Inf,positive infinity, 1

mixed number, significand, 1

NaN, not a number, 1

normal number

maximum positive, 1

normal number bit pattern, 1

precision, normal number, 1

sign bit, 1

subnormal number bit pattern, 1

IEEE Standard 754

double extended format, 1

double format, 1

single format, 1

ieee_flags

accrued exception flag, 1

examine accrued exception bits-C example, 1

rounding direction, 1

rounding precision, 1, 2

set exception flags-C example, 1

truncate rounding, 1

ieee_functions

bit mask operations, 1

floating-point exceptions, 1

ieee_handler

abort on exception

FORTRAN example, 1

trap on exception

C example, 1

ieee_handler, 1

example, calling sequence, 1

trap on common exceptions, 1

ieee_retrospective

check underflow exception flag, 1

floating-point exceptions, 1

floating-point status register (FSR), 1

getting information about nonstandard IEEE modes, 1

getting information about outstanding exceptions, 1

nonstandard_arithmetic in effect, 1

precision, 1

rounding, 1

suppress exception messages, 1

ieee_sun

IEEE classification functions, 1

ieee_values

quadruple-precision values, 1

representing floating-point values, 1

representing Inf, 1

representing NaN, 1

representing normal number, 1

single-precision values, 1

ieee_values functions

C example, 1

Inf, 1, 2

default result of divide by zero, 1

L

lcrans

random number utilities, 1

libm

default directories-executables, 1

default directories-header files, 1

list of functions, 1

standard installation, 1

SVID compliance, 1

libm functions

double precision, 1

quadruple precision, 1

single precision, 1

libmil (see also in-line templates), 1

libsunmath

default directories-executables, 1

default directories-header files, 1

list of functions, 1

standard installation, 1

M

man pages, accessing, 1

MANPATH environment variable, setting, 1

N

NaN, 1, 2, 3, 4

nonstandard_arithmetic

gradual underflow, 1

turn off IEEE gradual underflow, 1

underflow, 1

normal number

maximum positive, 1

minimum positive, 1, 2

number line

binary representation, 1

decimal representation, 1

powers of 2, 1

O

operating system math library

libm.a, 1

libm.so, 1

P

PATH environment variable, setting, 1

pi

infinitely precise value, 1

Q

quiet NaN

default result of invalid operation, 1

R

random number generators, 1

random number utilities

shufrans, 1

represent double-precision value

C example, 1

FORTRAN example, 1

represent single-precision value

C example, 1

rounding direction, 1

C example, 1

rounding precision, 1

roundoff error

accuracy

loss of, 1

S

set exception flags

C example, 1

shell prompts, 1

shufrans

shuffle pseudo-random numbers, 1

single format, 1

single precision representation

C example, 1

square root instruction, 1, 2

standard_arithmetic

turn on IEEE behavior, 1

Store 0, 1

flush underflow results, 1, 2

subnormal number, 1

floating-point calculations, 1

SVID behavior of libm

-Xt compiler option, 1

SVID exceptions

errno set to EDOM

improper operands, 1

errno set to ERANGE

overflow or underflow, 1

matherr, 1

PLOSS, 1

TLOSS, 1

System V Interface Definition (SVID), 1

T

trap

abort on exception, 1

ieee_retrospective, 1

trap on exception

C example, 1, 2

trap on floating-point exceptions

C example, 1

trigonometric functions

argument reduction, 1

tutorial, floating-point, 1

typographic conventions, 1

U

underflow

floating-point calculations, 1

gradual, 1

nonstandard_arithmetic, 1

threshold, 1

underflow thresholds

double extended precision, 1

double precision, 1

single precision, 1

unit in last place (ulp), 1

unordered comparison

floating-point values, 1

NaN, 1

V

values.h

define error messages, 1

X

X/Open behavior of libm

-Xa compiler option, 1

X_TLOSS, 1

-Xa, 1

-Xc, 1

-Xs, 1

-Xt, 1