HP-UX Floating-Point Guide

238 Index
Index
formats
floating-point
, 40
operand, converting between
,
64
Fortran
!$HP$ ALIAS and $ALIAS
directives
, 123
BLAS library
, 121, 180
calling C math library
functions from
, 123
static data, effect on
performance
, 178
vector library (obsolete)
, 121
Fortran math library
, 112, 215
error handling
, 106
fpclassify macro
, 48, 116
example
, 117
table of values
, 116
fpgetround functions
obsolete
, 16
FPU (floating-point unit)
, 126
fraction field
, 36, 38
fraction implicit bit
, 38
functions
C math library
, 114, 193
floating-point mode control
,
127
HP-UX math libraries
, 112
library, effect on application
results
, 73
new at HP-UX 11.0
, 16
recommended by IEEE
standard
, 67
G
global variables
and performance
, 190
gprof(1)
profiling tool
, 169
gradual underflow
disabling
, 145, 183
H
handling traps
, 157
ON statement
, 157
sigaction(2) function
, 161
header files
errno.h
, 105
fenv.h
, 127, 195
math.h
, 114, 195
hexadecimal
displaying floating-point
values in
, 77
hidden bit
, 38
I
IEEE formats
, 36
converting to and from decimal
formats
, 54
denormalized
, 41
double-precision
, 36
examples
, 40
infinity
, 43
limits of
, 40
NaN
, 45
normalized
, 41
quad-precision
, 36
single-precision
, 36
summary
, 48
zero
, 46
IEEE rounding modes
, 54
IEEE standard
, 16
conversions
, 64
introduction
, 35
recommended functions
, 67
ill-conditioned computations
and significance loss
, 94
ilogb function
, 118
implicit bit
fraction
, 38
inefficient code
, 171
inexact result condition
, 53
infinity
converting to integer format,
58
dividing by infinity
, 58
multiplying by zero
, 58
properties of, table
, 44
representation of
, 43
round toward negative
, 55
round toward positive
, 55
subtracting
, 58
instability, numeric. See ill-
conditioned computations
instruction cache aliasing
and performance
, 187
table
, 188
integer exceptions
handling
, 165
integers
converting infinity to
, 58
converting NaN to
, 58
converting to floating-point
, 64
handling exceptions
, 165
rounding floating-point to
, 62
truncating floating-point to
,
64, 92
invalid operation condition
, 58
isfinite macro
, 48, 118
isgreater macro
, 119
isgreaterequal macro
, 119
isinf macro
, 48, 118
isless macro
, 119
islessequal macro
, 119
islessgreater macro
, 120
isnan macro
, 48, 118
ISO/ANSI C standard. See C9X
draft standard
isunordered macro
, 120
K
-K compiler option
, 178