HP-UX Floating-Point Guide

Index
Index 239
L
-l linker option
, 25
ld link editor
+FP option
, 148
-a option
, 181
cache aliasing and
, 188
-l option
, 25
least significant word
, 37
libblas BLAS library
, 121
and performance
, 180, 185
libcl
Fortran and Pascal library
,
112, 215
libm
C math library
, 114, 193
libraries
archive
, 99, 181
introduction
, 99
math
, 97
shared
, 99, 181
libvec vector library (obsolete)
,
121
linker errors
, 224
linker options
+FP
, 148, 153
-a
, 181
-l
, 25
llrint function
, 120
llround function
, 120
log1p function
, 118
logb function
, 118
lrint function
, 120
lround function
, 120
M
mantissa. See fraction field
mask bits. See exception trap
enable bits
math libraries
, 97
BLAS library
, 121, 180, 185
C
, 104, 114, 123, 193
changes for HP-UX 11.0
, 16
contents
, 112
directory hierarchy,
illustration
, 31
error handling for C
, 104
error handling for Fortran
, 106
fenv(5) functions
, 127
Fortran
, 106, 112, 215
function calls
, 102
introduction
, 101
libblas
, 121, 185
libcl
, 112, 215
libm
, 112, 114, 193
libvec (obsolete)
, 121
locations
, 29
millicode versions of functions
,
112
overview
, 24
Pascal
, 112, 215
rounding mode and
, 55
scalar
, 112
selecting different versions
, 27
significance loss and
, 88
system architecture and
, 25
vector library (obsolete)
, 121
math.h header file
, 114, 195
matherr function
obsolete
, 16
matrix operations
and performance
, 185
maximum representable
floating-point values
table
, 56
millicode versions of math
functions
, 112, 175
mixed-precision expressions
effect on performance
, 184
modules
reordering to avoid cache alias
problems
, 188
modulo operation
and significance loss
, 88
most significant word
, 37
multiplication
operation
, 61
N
NaN
converting to integer format
,
58
in comparison operations
, 58,
63, 64
properties of, table
, 46
quiet
, 45
representation of
, 45
signaling
, 45
nan function
, 120
narrow mode
, 25
natural alignment
, 186
nearbyint function
, 120
negative infinity
round toward
, 55
negative value
square root of
, 58
negative zero
, 46
nextafter function
, 118
nextafterf function
, 120
non-aware version of comparison
assertions
, 63
normalized values
, 41
not-a-number. See NaN
numerically unstable
computations. See ill-
conditioned computations
O
obsolete library functions
, 16
ON statement
, 157
Fortran math library error
handling
, 108
operand errors. See invalid
operation conditions
operating system release
effect on application results
, 81