A version of the gcccommand (which can also be installed as the system's cccommand) that also understands and accepts the Fortran source code. The middle and back-end code is shared by all the compiler front-ends contained in the GNU Compiler Collection. The code generator translates the semantics of the source code into a language independent form called GENERIC.
There is a good Fortran 77 compiler in the GNU Compiler Collection that accepts some standard Fortran 90 features as extensions.
Option Summary
For information on the non-Fortran-specific aspects of the gcc command (and therefore the gfortran command), see the "GCC Command Options" section in Using the GNU Compiler Collection (GCC). All gcc and gfortran options are accepted by both gfortran and gcc (as well as all other concurrently built drivers such as g++), since adding gfortran to the gcc distribution allows gfortran options to be accepted by all relevant drivers.
Options Controlling Fortran Dialect
Options to Request or Suppress Warnings
This manual documents only one of these two forms, whichever is non-standard. If the '-fd-lines-as-comments' option is given, they are treated as comment lines. Each of these specific warning options also has a negative form beginning with '-Wno-' to turn off warnings; e.g,.
This may change in future versions of gfortran See the "Options to check for or suppress warnings" section in Using the GNU Compiler Collection (GCC), for information on more options provided by the GBE separate from gfortran, gcc and other GNU compilers.
Options for Debugging Your Program or GNU Fortran
Options for Directory Search
Set the column by which characters are ignored in typical fixed-format lines in the source file and through which spaces are assumed (as if padded to that length) at the ends of short fixed-format lines. Set the column by which characters are ignored in typical free-form lines in the source file. Specifically, it warns if the same actual argument is associated with a virtual argument with intent(in) and a virtual argument within intent(out) in a call to an explicit interface.
Warn if the user tries to use an intrinsic that is not in the default that the user chose through the -std option.
Influencing runtime behavior
Options for Code Generation Conventions
If this option is used, gfortran adds two underscores to names with underscores and one underscore to external names without underscores. Code compiled with this option is likely to be incompatible with code compiled without this option and may execute more slowly. This option is provided for interoperability with C code that was compiled with the -fshort-enums option.
See section "Options for Code Generation Conventions" in Using the GNU Compiler Collection (GCC), for information about more options offered by the GBE-shared gfortran gcc and other GNU compilers.
Environment Variables Affecting GNU Fortran
As soon as gfortran can parse all the statements correctly, it will be in. When gfortran is done, we will see if it will be a beautiful butterfly or just a big bug. The start of the GNU Fortran 95 project was announced on the GCC website on March 18, 2000 (although of course Andy had already been working on it for a while).
Gfortran is currently reaching the stage where it is able to compile real-world programs.
Compiler Status
Library Status
Proposed Extensions
Compiler extensions
Compile code that saves memory by dynamically allocating shared and module storage either on the stack or heap. Compile setting for the maximum size of the stack frame size before wasting parts to static or heaps. Generate code to check for null pointer dereferences - prints locus of dereference instead of segfaulting.
This includes several new features, including floating-point exceptions, extended use of allocable arrays, C interoperability, Parameterizer data types, and function pointers.
Environment Options
GFORTRAN STDIN UNIT – Unit number for standard input
GFORTRAN STDOUT UNIT – Unit number for standard output
GFORTRAN STDERR UNIT – Unit number for standard error
GFORTRAN USE STDERR:: Send library output to standard
GFORTRAN TMPDIR – Directory for scratch files
GFORTRAN UNBUFFERED ALL – Don’t buffer output
GFORTRAN SHOW LOCUS – Show location for runtime errors
GFORTRAN OPTIONAL PLUS – Print leading + where
GFORTRAN LIST SEPARATOR – Separator for list output
GFORTRAN CONVERT UNIT – Set endianness for
Setting environment variables must be done on the command line or via the export command for sh-compatible shells and viasetenvforcsh-compatible shells. See Section 7.13 [CONVERT Specifier], page 31, for an alternative way to specify data representation for raw files. By default, '-std=gnu' allows the compiler to accept both types of extensions, but to warn about using the latter.
Specifying "-std=f95" or "-std=f2003" disables both types of extensions, and "-std=legacy" allows both without warning.
Old-style kind specifications
There are currently two categories of gfortran extensions, those that provide functionality beyond that provided by any standard, and those that are supported by gfortran simply for backward compatibility with legacy compilers.
Old-style variable initialization
Extensions to namelist
X format descriptor
Commas in FORMAT specifications
I/O item lists
Hexadecimal constants
Real array indices
Unary operators
Implicitly interconvert LOGICAL and INTEGER
Hollerith constants support
Cray pointers
CONVERT specifier
The type of the pointee can be declared before or after the pointer statement, and the array specification (if any) can be declared before, during, or after the pointer statement. Any expression involving the pointee is translated to use the value stored in the pointer as the base address. There is no malloc implemented intrinsically as part of the Cray pointer extension, but it could be a useful future addition to gfortran.
Subsequent changes to the pointer will not change the base address of the array that was passed.
Introduction to intrinsic procedures
ABORT — Abort the program
ABS — Absolute value
ACHAR — Character in ASCII collating sequence
ACOS — Arc cosine function
ADJUSTL — Left adjust a string
The return value is of type CHARACTER where leading spaces are removed and the same number of spaces are inserted at the end of STR.
ADJUSTR — Right adjust a string
AIMAG — Imaginary part of complex number
AINT — Imaginary part of complex number
ALARM — Execute a routine after a given delay
If STATUS is provided, it will be returned with the number of seconds remaining until any previously scheduled alarms were sent, or zero if there were no previously scheduled alarms.
ALL — All values in MASK along DIM are true
If the rank is greater than one, ALL(MASK,DIM) is determined by applying ALL to the array sections. b, 2) end subroutine section end program test_all.
ALLOCATED — Status of an allocatable entity
ANINT — Nearest whole number
The return value is of type real with the argument's type type parameter if the optional KIND is absent; otherwise the childtype parameter is given by KIND.
ANY — Any value in MASK along DIM is true
ASIN — Arcsine function
ASSOCIATED — Status of a pointer or pointer/target pair
If the optional TGT is not present, then ASSOCIATED(PTR). is true if PTRis is associated with a target; otherwise it returns false. The TGT is not a 0-size storage sequence, and the target associated with the PTR occupies the same storage units. TGT and PTR have the same form, are not arrays of size 0, arrays whose elements are not 0 are large storage sequences, and TGT and PTR occupy the same storage units in array element order.
ATAN — Arctangent function
ATAN2 — Arctangent function
BESJ0 — Bessel function of the first kind of order 0
BESJ1 — Bessel function of the first kind of order 1
BESJN — Bessel function of the first kind
BESY0 — Bessel function of the second kind of order 0
BESY1 — Bessel function of the second kind of order 1
BESYN — Bessel function of the second kind
BIT_SIZE — Bit size inquiry function
BTEST — Bit test function
CEILING — Integer ceiling function
CHAR — Character conversion function
CMPLX — Complex conversion function
COMMAND_ARGUMENT_COUNT — Argument count function
CONJG — Complex conjugate function
COS — Cosine function
COSH — Hyperbolic cosine function
COUNT — Count function
CPU_TIME — CPU elapsed time in seconds
CSHIFT — Circular shift function
CTIME — Convert a time into a string
DATE_AND_TIME — Date and time subroutine
DBLE — Double conversion function
DCMPLX — Double complex conversion function
DFLOAT — Double conversion function
DIGITS — Significant digits function
DIM — Dim function
DOT_PRODUCT — Dot product function
DPROD — Double product function
DREAL — Double real part function
DTIME — Execution time subroutine (or function)
If DTIME is called as a function, it cannot be called as a subroutine, and vice versa.
EOSHIFT — End-off shift function
Therefore, the values returned by this intrinsic value may be or become negative or numerically less than previous values during a single run of the compiled program. If BOUNDARY is present, the corresponding value is copied from BOUNDARY back at the other end. If BOUNDARY is not present, the following is copied depending on the type of ARRAY.
EPSILON — Epsilon function
ERF — Error function
ERFC — Error function
ETIME — Execution time subroutine (or function)
EXIT — Exit the program with status
EXP — Exponential function
EXPONENT — Exponent function
FDATE — Get the current time as a string
FLOAT — Convert integer to default real
FLOOR — Integer floor function
FLUSH — Flush I/O unit(s)
FNUM — File number function
FRACTION — Fractional part of the model representation
FREE — Frees memory
GETGID — Group ID function
GETPID — Process ID function
GETUID — User ID function
HUGE — Largest number of a kind
IACHAR — Code in ASCII collating sequence
ICHAR — Character-to-integer conversion function
IDATE — Get current local time subroutine (day/month/year)
IRAND — Integer pseudo-random number
ITIME — Get current local time subroutine
KIND — Kind of an entity
LOC — Returns the address of a variable
LOG — Logarithm function
LOG10 — Base 10 logarithm function
MALLOC — Allocate dynamic memory
This example is intended to run on 32-bit systems, where the default integer type is suitable for storing pointers; on 64-bit systems ptr x will need to be declared as integer (kind=8).
MAXEXPONENT — Maximum exponent of a real kind
MINEXPONENT — Minimum exponent of a real kind
The return value is of type INTEGER(K), with K such that variables of type INTEGER(K) have the same size as C pointers (sizeof(void.
MOD — Remainder function
MODULO — Modulo function
NEAREST — Nearest representable number
If S is positive, NEAREST returns the number represented by the processor greater than X and closest to it. If S is negative, NEAREST returns the number represented by the processor that is less than and closest to X.
NINT — Nearest whole number
PRECISION — Decimal precision of a real kind
RADIX — Base of a model number
RAND — Real pseudo-random number
RANGE — Decimal exponent range of a real kind
REAL — Convert to real type
RRSPACING — Reciprocal of the relative spacing
SCALE — Scale a real value
SELECTED_INT_KIND — Choose integer kind
SELECTED_REAL_KIND — Choose real kind
If more than one real data type meets the criteria, the data type with the smallest decimal precision is returned. 1 if the processor does not support a real data type with accuracy greater than or equal to P. 2 if the processor does not support a real type with an exponent range greater than or equal to R.
SECNDS — Time subroutine
SET_EXPONENT — Set the exponent of the model
SIGN — Sign copying function
SIGNAL — Signal handling subroutine (or function)
SIN — Sine function
SINH — Hyperbolic sine function
SNGL — Convert double precision real to default real
SQRT — Square-root function
Notes: The Fortran 2003 standard specifies the intrinsic RANDOM_SEED to initialize the pseudo-random number generator and RANDOM_NUMBER to generate pseudo-random numbers. Please note that in gfortran, these two internal arrays (RAND, IRAND, and SRAND on the one hand, RANDOM_NUMBER and RANDOM_SEED on the other) access two independent pseudo-random number generators. TINY(X) returns the smallest positive (non-zero) number in the pattern of type X.
The return value is of the same type and kind as X Example: See BIG for an example. We always need more people helping with ideas and comments, writing documentation and contributing code. If you want to contribute to GNU Fortran 95, check out the long lists of projects you can take on.
Some of these projects are small, some of them are large; some are completely orthogonal to the rest of what happens on gfortran, but others are "mainstream". We will eventually get around to the things here, but they are also things that can be done by someone who is willing and able.
Contributors to GNU Fortran 95
If you have already done so and would like your name to appear on the list above, please contact us.
Projects
In the future it may also support other variants and extensions to the Fortran language.
Fortran 2003 status
You may copy and distribute the Program (or a work based on it, subject to Section 2) in object code or executable form under the terms of Sections 1 and 2 above, provided that you also do one of the following: This Section intended to fully clarify what is believed to be the consequence of the rest of this license. The Free Software Foundation may from time to time publish revised and/or new versions of the General Public License.
THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. You should have received a copy of the GNU General Public License with this program; if not, write to the Free Software. This license is a kind of "copyleft", which means that derivative works of the document itself must be free in the same sense.
On the Title Page, list the name of the publisher of the Custom Version as the publisher. Make the same adjustment to the section titles in the list of Invariant Sections in the licensing notice of the combined work. When the Document is included in an aggregate, this License does not apply to the other works in the aggregate that are not themselves derivative works of the Document.
Translation is considered a type of modification, so you may distribute translations of the Document under the terms of Section 4. The Free Software Foundation may from time to time publish new, revised versions of the GNU Free Documentation License.
