DIGITAL Fortran 90
User Manual for
DIGITAL UNIX Systems

1
-1 option
6
-66 option
A
a.out file #1
a.out file #2
specifying different name
abort library routine
abstract_to_physical library routine
ACCEPT statement
See also LRM
access library routine
Access modes
direct
limitations by file organization and record type
OPEN statement specifiers
requirement for direct access to sequential files
sequential
ACCESS specifier #1
ACCESS specifier #2
See also LRM
Accuracy
and dummy aliases
and numerical data I/O
-assume noaccuracy_sensitive option
floating-point constants
-fpconstant option
-fprm keyword options
-fp_reorder option
hoisting divide operations
-intconstant option
integer constants
-math_library fast option
rounding during calculations
when converting OpenVMS Fortran unformatted data
ACTION specifier
See also LRM
Actual arguments
Adjusting the run-time parallel environment
ADVANCE specifier #1
ADVANCE specifier #2
See also LRM
Advancing I/O
AIMAG intrinsic function
options controlling size returned
See also LRM
alarm library routine
-align commons option #1
-align commons option #2
effect of -fast option
-align dcommons option #1
-align dcommons option #2
-align nocommons option
-align norecords option #1
-align norecords option #2
-align recnbyte option
-align records option
effect of -vms option #1
effect of -vms option #2
Alignment
argument passing
DIGITAL Fortran 90 and C structures
array elements
cDEC$ directives and -align options
checking for unaligned data #1
checking for unaligned data #2
common block data
derived-type data
effect of declaration statements
locating unaligned data (debugger)
of data types
options controlling
common block data #1
common block data #2
common block data #3
common block data #4
common block data #5
derived-type data #1
derived-type data #2
effect of -vms option
record structures #1
record structures #2
record structures #3
warnings for unaligned data
record structures
SEQUENCE statement
effect on derived-type data
unaligned data
causes
checking for
effect on performance #1
effect on performance #2
locating in debugger
options #1
options #2
ordering data declarations
warning messages for #1
warning messages for #2
Allocating and freeing virtual memory (library routine)
Alpha processor generation
specifying to compiler #1
specifying to compiler #2
specifying to compiler #3
Alternate entry points (Fortran)
-altparam option
and function
APPEND specifier
See also LRM
ar command
-arch option
Archive library
creating and maintaining
linker searching options
list searched by f90 command
nonshared optimizations
obtaining information about
recognized file name suffix
specifying with f90
Argument passing
alignment of data passed
C and DIGITAL Fortran 90 #1
C and DIGITAL Fortran 90 #2
C and DIGITAL Fortran 90 #3
alignment of structures
arrays
cDEC$ ATTRIBUTES C directive
changing default mechanisms #1
changing default mechanisms #2
changing default mechanisms #3
character data
character data example
character null terminator
common block values
complex data example
data types
examples #1
examples #2
examples #3
examples #4
integer data example
passing arrays
pointer data #1
pointer data #2
structures
using C conventions
checking for mismatches at compile-time
DIGITAL Fortran 77 and DIGITAL Fortran 90
alignment options
common block values
data type sizes
data types to avoid
differences
example
function values #1
function values #2
I/O compatibility
mechanisms
passing target or pointer data
pointer data
similarities
using DIGITAL Fortran 77 features
DIGITAL Fortran 90
array temporary creation
arrays
by address (%LOC)
by reference (%REF)
by value (%VAL)
changing default mechanisms #1
changing default mechanisms #2
characters
default passing mechanism
descriptor format
explicit interface
hidden length #1
hidden length #2
hidden length #3
hidden length #4
implicit interface #1
implicit interface #2
omitted arguments (extra comma)
pointers
rules #1
rules #2
efficient array use
from USEROPEN function and DIGITAL Fortran 90 RTL
when temporary array copy is created
Arguments
actual
differences between DIGITAL Fortran 77 and DIGITAL Fortran 90
dummy
aliasing and accuracy
and implied-DO loop collapsing
option for aliasing
maximum allowed on CALL statement
See also Argument passing; LRM
Arithmetic exception handling
controlling reporting
-check nopower option
-check underflow option
-fpe options
-speculate option
-synchronous_exceptions option
debugger handling
defaults and applicable options
effect of speculative execution optimization
effect of using fast math library routines
example program
exceptional values
floating-point data #1
floating-point data #2
-check nopower option
-check underflow option
-fpe options
-synchronous_exceptions option
floating-point exceptional values
floating-point underflow #1
floating-point underflow #2
floating-point underflow #3
floating-point underflow #4
forcing core dump at run time
FP_CLASS intrinsic
IEEE NaN values (quiet and signaling)
infinity values
integer overflow
ISNAN intrinsic
options for
See also Data types, numeric ranges
signals
caught by DIGITAL Fortran 90 RTL
definition
handling in debugger #1
handling in debugger #2
handling in debugger #3
Arithmetic operators
for efficient run-time performance
Arrays
alignment
allocatable
arguments #1
arguments #2
arguments #3
DIGITAL Fortran 90 and DIGITAL Fortran 77 similarities
example #1
example #2
example #3
example #4
example #5
See also LRM
array sections
viewing in debugger
array temporary creation
assumed-shape #1
assumed-shape #2
assumed-size
bounds checking
character
arguments passed with hidden length
column-major order
conformable
declaring
See LRM
differences between DIGITAL Fortran 90 and C
dimension limits
efficient array syntax
efficient array syntax in I/O list
efficient combinations of input and output arguments
element sequence association
explicit-shape
explicit-shape arguments
example (C and DIGITAL Fortran 90)
example (DIGITAL Fortran 77 and DIGITAL Fortran 90)
expression syntax for debugger
HPF_LOCAL_LIBRARY routines
natural storage order
optimizations #1
optimizations #2
passed by descriptor
passing as arguments
example
pointers to
row-major order
See also Argument passing; LRM
sorting
library routines for
syntax for run-time efficiency
using efficiently
when temporary copy is created for argument passing
writing for efficiency
zero-sized
Assembler file
creating
passed to cc
ASSOCIATEVARIABLE specifier
See also LRM
Association
host
procedure interface
use
procedure interface block #1
procedure interface block #2
-assume accuracy_sensitive option
-assume bigarrays option
-assume byterecl option
-assume dummy_aliases option #1
-assume dummy_aliases option #2
-assume dummy_aliases option #3
-assume gfullpath option
-assume minus0 option
-assume noaccuracy_sensitive option
effect of -fast option
-assume nogfullpath option
-assume nominus0 option
-assume nopthread_lock option
-assume nosource_include option
-assume nounderscore option
-assume nozsize option
effect of -fast option
-assume pthread_lock option
-assume source_include option
Atom toolkit
ATOMIC directive #1
ATOMIC directive #2
ATOMIC directive #3
-automatic option #1
-automatic option #2
effect on -recursive