1 \input texinfo @c -*-texinfo-*-
3 @setfilename gfortran.info
4 @set copyrights-gfortran 1999-2007
6 @include gcc-common.texi
8 @settitle The GNU Fortran Compiler
10 @c Create a separate index for command line options
12 @c Merge the standard indexes into a single one.
19 @c TODO: The following "Part" definitions are included here temporarily
20 @c until they are incorporated into the official Texinfo distribution.
21 @c They borrow heavily from Texinfo's \unnchapentry definitions.
28 \vglue\titlepagetopglue
30 \leftline{Part #1:@* #2}
31 \vskip4pt \hrule height 4pt width \hsize \vskip4pt
33 \writetocentry{part}{#2}{#1}
36 \writetocentry{blankpart}{}{}
38 % Part TOC-entry definition for summary contents.
39 \gdef\dosmallpartentry#1#2#3#4{%
40 \vskip .5\baselineskip plus.2\baselineskip
43 \tocentry{Part #2: #1}{\doshortpageno\bgroup#4\egroup}
46 \gdef\dosmallblankpartentry#1#2#3#4{%
47 \vskip .5\baselineskip plus.2\baselineskip
49 % Part TOC-entry definition for regular contents. This has to be
50 % equated to an existing entry to not cause problems when the PDF
52 \gdef\dopartentry#1#2#3#4{%
53 \unnchapentry{Part #2: #1}{}{#3}{#4}
55 \gdef\doblankpartentry#1#2#3#4{}
60 @c Use with @@smallbook.
62 @c %** start of document
64 @c Cause even numbered pages to be printed on the left hand side of
65 @c the page and odd numbered pages to be printed on the right hand
66 @c side of the page. Using this, you can print on both sides of a
67 @c sheet of paper and have the text on the same part of the sheet.
69 @c The text on right hand pages is pushed towards the right hand
70 @c margin and the text on left hand pages is pushed toward the left
72 @c (To provide the reverse effect, set bindingoffset to -0.75in.)
75 @c \global\bindingoffset=0.75in
76 @c \global\normaloffset =0.75in
80 Copyright @copyright{} @value{copyrights-gfortran} Free Software Foundation, Inc.
82 Permission is granted to copy, distribute and/or modify this document
83 under the terms of the GNU Free Documentation License, Version 1.1 or
84 any later version published by the Free Software Foundation; with the
85 Invariant Sections being ``GNU General Public License'' and ``Funding
86 Free Software'', the Front-Cover
87 texts being (a) (see below), and with the Back-Cover Texts being (b)
88 (see below). A copy of the license is included in the section entitled
89 ``GNU Free Documentation License''.
91 (a) The FSF's Front-Cover Text is:
95 (b) The FSF's Back-Cover Text is:
97 You have freedom to copy and modify this GNU Manual, like GNU
98 software. Copies published by the Free Software Foundation raise
99 funds for GNU development.
103 @dircategory Software development
105 * gfortran: (gfortran). The GNU Fortran Compiler.
107 This file documents the use and the internals of
108 the GNU Fortran compiler, (@command{gfortran}).
110 Published by the Free Software Foundation
111 51 Franklin Street, Fifth Floor
112 Boston, MA 02110-1301 USA
118 @setchapternewpage odd
120 @title Using GNU Fortran
122 @author The @t{gfortran} team
124 @vskip 0pt plus 1filll
125 Published by the Free Software Foundation@*
126 51 Franklin Street, Fifth Floor@*
127 Boston, MA 02110-1301, USA@*
128 @c Last printed ??ber, 19??.@*
129 @c Printed copies are available for $? each.@*
135 @c TODO: The following "Part" definitions are included here temporarily
136 @c until they are incorporated into the official Texinfo distribution.
139 \global\let\partentry=\dosmallpartentry
140 \global\let\blankpartentry=\dosmallblankpartentry
145 \global\let\partentry=\dopartentry
146 \global\let\blankpartentry=\doblankpartentry
152 @c ---------------------------------------------------------------------
153 @c TexInfo table of contents.
154 @c ---------------------------------------------------------------------
161 This manual documents the use of @command{gfortran},
162 the GNU Fortran compiler. You can find in this manual how to invoke
163 @command{gfortran}, as well as its features and incompatibilities.
166 @emph{Warning:} This document, and the compiler it describes, are still
167 under development. While efforts are made to keep it up-to-date, it might
168 not accurately reflect the status of the most recent GNU Fortran compiler.
172 @comment When you add a new menu item, please keep the right hand
173 @comment aligned to the same column. Do not use tabs. This provides
174 @comment better formatting.
179 Part I: Invoking GNU Fortran
180 * Invoking GNU Fortran:: Command options supported by @command{gfortran}.
181 * Runtime:: Influencing runtime behavior with environment variables.
183 Part II: Language Reference
184 * Fortran 2003 status:: Fortran 2003 features supported by GNU Fortran.
185 * Extensions:: Language extensions implemented by GNU Fortran.
186 * Intrinsic Procedures:: Intrinsic procedures supported by GNU Fortran.
188 * Contributing:: How you can help.
189 * Copying:: GNU General Public License says
190 how you can copy and share GNU Fortran.
191 * GNU Free Documentation License::
192 How you can copy and share this manual.
193 * Funding:: How to help assure continued work for free software.
194 * Option Index:: Index of command line options
195 * Keyword Index:: Index of concepts
199 @c ---------------------------------------------------------------------
201 @c ---------------------------------------------------------------------
204 @chapter Introduction
206 @c The following duplicates the text on the TexInfo table of contents.
208 This manual documents the use of @command{gfortran}, the GNU Fortran
209 compiler. You can find in this manual how to invoke @command{gfortran},
210 as well as its features and incompatibilities.
213 @emph{Warning:} This document, and the compiler it describes, are still
214 under development. While efforts are made to keep it up-to-date, it
215 might not accurately reflect the status of the most recent GNU Fortran
220 The GNU Fortran compiler front end was
221 designed initially as a free replacement for,
222 or alternative to, the unix @command{f95} command;
223 @command{gfortran} is the command you'll use to invoke the compiler.
226 * About GNU Fortran:: What you should know about the GNU Fortran compiler.
227 * GNU Fortran and GCC:: You can compile Fortran, C, or other programs.
228 * GNU Fortran and G77:: Why we chose to start from scratch.
229 * Project Status:: Status of GNU Fortran, roadmap, proposed extensions.
230 * Standards:: Standards supported by GNU Fortran.
234 @c ---------------------------------------------------------------------
236 @c ---------------------------------------------------------------------
238 @node About GNU Fortran
239 @section About GNU Fortran
241 The GNU Fortran compiler is still in an early state of development.
242 It can generate code for most constructs and expressions,
243 but much work remains to be done.
245 When the GNU Fortran compiler is finished,
246 it will do everything you expect from any decent compiler:
250 Read a user's program,
251 stored in a file and containing instructions written
252 in Fortran 77, Fortran 90, Fortran 95 or Fortran 2003.
253 This file contains @dfn{source code}.
256 Translate the user's program into instructions a computer
257 can carry out more quickly than it takes to translate the
258 instructions in the first
259 place. The result after compilation of a program is
261 code designed to be efficiently translated and processed
262 by a machine such as your computer.
263 Humans usually aren't as good writing machine code
264 as they are at writing Fortran (or C++, Ada, or Java),
265 because is easy to make tiny mistakes writing machine code.
268 Provide the user with information about the reasons why
269 the compiler is unable to create a binary from the source code.
270 Usually this will be the case if the source code is flawed.
271 When writing Fortran, it is easy to make big mistakes.
272 The Fortran 90 requires that the compiler can point out
273 mistakes to the user.
274 An incorrect usage of the language causes an @dfn{error message}.
276 The compiler will also attempt to diagnose cases where the
277 user's program contains a correct usage of the language,
278 but instructs the computer to do something questionable.
279 This kind of diagnostics message is called a @dfn{warning message}.
282 Provide optional information about the translation passes
283 from the source code to machine code.
284 This can help a user of the compiler to find the cause of
285 certain bugs which may not be obvious in the source code,
286 but may be more easily found at a lower level compiler output.
287 It also helps developers to find bugs in the compiler itself.
290 Provide information in the generated machine code that can
291 make it easier to find bugs in the program (using a debugging tool,
292 called a @dfn{debugger}, such as the GNU Debugger @command{gdb}).
295 Locate and gather machine code already generated to
296 perform actions requested by statements in the user's program.
297 This machine code is organized into @dfn{modules} and is located
298 and @dfn{linked} to the user program.
301 The GNU Fortran compiler consists of several components:
305 A version of the @command{gcc} command
306 (which also might be installed as the system's @command{cc} command)
307 that also understands and accepts Fortran source code.
308 The @command{gcc} command is the @dfn{driver} program for
309 all the languages in the GNU Compiler Collection (GCC);
311 you can compile the source code of any language for
312 which a front end is available in GCC.
315 The @command{gfortran} command itself,
316 which also might be installed as the
317 system's @command{f95} command.
318 @command{gfortran} is just another driver program,
319 but specifically for the Fortran compiler only.
320 The difference with @command{gcc} is that @command{gfortran}
321 will automatically link the correct libraries to your program.
324 A collection of run-time libraries.
325 These libraries contain the machine code needed to support
326 capabilities of the Fortran language that are not directly
327 provided by the machine code generated by the
328 @command{gfortran} compilation phase,
329 such as intrinsic functions and subroutines,
330 and routines for interaction with files and the operating system.
331 @c and mechanisms to spawn,
332 @c unleash and pause threads in parallelized code.
335 The Fortran compiler itself, (@command{f951}).
336 This is the GNU Fortran parser and code generator,
337 linked to and interfaced with the GCC backend library.
338 @command{f951} ``translates'' the source code to
339 assembler code. You would typically not use this
341 instead, the @command{gcc} or @command{gfortran} driver
342 programs will call it for you.
346 @c ---------------------------------------------------------------------
347 @c GNU Fortran and GCC
348 @c ---------------------------------------------------------------------
350 @node GNU Fortran and GCC
351 @section GNU Fortran and GCC
352 @cindex GNU Compiler Collection
355 GNU Fortran is a part of GCC, the @dfn{GNU Compiler Collection}. GCC
356 consists of a collection of front ends for various languages, which
357 translate the source code into a language-independent form called
358 @dfn{GENERIC}. This is then processed by a common middle end which
359 provides optimization, and then passed to one of a collection of back
360 ends which generate code for different computer architectures and
363 Functionally, this is implemented with a driver program (@command{gcc})
364 which provides the command-line interface for the compiler. It calls
365 the relevant compiler front-end program (e.g., @command{f951} for
366 Fortran) for each file in the source code, and then calls the assembler
367 and linker as appropriate to produce the compiled output. In a copy of
368 GCC which has been compiled with Fortran language support enabled,
369 @command{gcc} will recognize files with @file{.f}, @file{.f90}, @file{.f95},
370 and @file{.f03} extensions as Fortran source code, and compile it
371 accordingly. A @command{gfortran} driver program is also provided,
372 which is identical to @command{gcc} except that it automatically links
373 the Fortran runtime libraries into the compiled program.
375 This manual specifically documents the Fortran front end, which handles
376 the programming language's syntax and semantics. The aspects of GCC
377 which relate to the optimization passes and the back-end code generation
378 are documented in the GCC manual; see
379 @ref{Top,,Introduction,gcc,Using the GNU Compiler Collection (GCC)}.
380 The two manuals together provide a complete reference for the GNU
384 @c ---------------------------------------------------------------------
385 @c GNU Fortran and G77
386 @c ---------------------------------------------------------------------
388 @node GNU Fortran and G77
389 @section GNU Fortran and G77
393 The GNU Fortran compiler is the successor to G77, the Fortran 77 front
394 end included in GCC prior to version 4. It is an entirely new program
395 that has been designed to provide Fortran 95 support and extensibility
396 for future Fortran language standards, as well as providing backwards
397 compatibility for Fortran 77 and nearly all of the GNU language
398 extensions supported by G77.
401 @c ---------------------------------------------------------------------
403 @c ---------------------------------------------------------------------
406 @section Project Status
409 As soon as @command{gfortran} can parse all of the statements correctly,
410 it will be in the ``larva'' state.
411 When we generate code, the ``puppa'' state.
412 When @command{gfortran} is done,
413 we'll see if it will be a beautiful butterfly,
414 or just a big bug....
416 --Andy Vaught, April 2000
419 The start of the GNU Fortran 95 project was announced on
420 the GCC homepage in March 18, 2000
421 (even though Andy had already been working on it for a while,
424 The GNU Fortran compiler is able to compile nearly all
425 standard-compliant Fortran 95, Fortran 90, and Fortran 77 programs,
426 including a number of standard and non-standard extensions, and can be
427 used on real-world programs. In particular, the supported extensions
428 include OpenMP, Cray-style pointers, and several Fortran 2003 features
429 such as enumeration, stream I/O, and some of the enhancements to
430 allocatable array support from TR 15581. However, it is still under
431 development and has a few remaining rough edges.
433 At present, the GNU Fortran compiler passes the
434 @uref{http://www.fortran-2000.com/ArnaudRecipes/fcvs21_f95.html,
435 NIST Fortran 77 Test Suite}, and produces acceptable results on the
436 @uref{http://www.netlib.org/lapack/faq.html#1.21, LAPACK Test Suite}.
437 It also provides respectable performance on
438 the @uref{http://www.polyhedron.com/pb05.html, Polyhedron Fortran
439 compiler benchmarks} and the
440 @uref{http://www.llnl.gov/asci_benchmarks/asci/limited/lfk/README.html,
441 Livermore Fortran Kernels test}. It has been used to compile a number of
442 large real-world programs, including
443 @uref{http://mysite.verizon.net/serveall/moene.pdf, the HIRLAM
444 weather-forecasting code} and
445 @uref{http://www.theochem.uwa.edu.au/tonto/, the Tonto quantum
446 chemistry package}; see @url{http://gcc.gnu.org/wiki/GfortranApps} for an
449 Among other things, the GNU Fortran compiler is intended as a replacement
450 for G77. At this point, nearly all programs that could be compiled with
451 G77 can be compiled with GNU Fortran, although there are a few minor known
454 The primary work remaining to be done on GNU Fortran falls into three
455 categories: bug fixing (primarily regarding the treatment of invalid code
456 and providing useful error messages), improving the compiler optimizations
457 and the performance of compiled code, and extending the compiler to support
458 future standards---in particular, Fortran 2003.
461 @c ---------------------------------------------------------------------
463 @c ---------------------------------------------------------------------
469 The GNU Fortran compiler implements
470 ISO/IEC 1539:1997 (Fortran 95). As such, it can also compile essentially all
471 standard-compliant Fortran 90 and Fortran 77 programs. It also supports
472 the ISO/IEC TR-15581 enhancements to allocatable arrays, and
473 the @uref{http://www.openmp.org/drupal/mp-documents/spec25.pdf,
474 OpenMP Application Program Interface v2.5} specification.
476 In the future, the GNU Fortran compiler may also support other standard
477 variants of and extensions to the Fortran language. These include
478 ISO/IEC 1539-1:2004 (Fortran 2003).
481 @c =====================================================================
482 @c PART I: INVOCATION REFERENCE
483 @c =====================================================================
486 \part{I}{Invoking GNU Fortran}
489 @c ---------------------------------------------------------------------
491 @c ---------------------------------------------------------------------
496 @c ---------------------------------------------------------------------
498 @c ---------------------------------------------------------------------
501 @chapter Runtime: Influencing runtime behavior with environment variables
504 The behavior of the @command{gfortran} can be influenced by
505 environment variables.
507 Malformed environment variables are silently ignored.
510 * GFORTRAN_STDIN_UNIT:: Unit number for standard input
511 * GFORTRAN_STDOUT_UNIT:: Unit number for standard output
512 * GFORTRAN_STDERR_UNIT:: Unit number for standard error
513 * GFORTRAN_USE_STDERR:: Send library output to standard error
514 * GFORTRAN_TMPDIR:: Directory for scratch files
515 * GFORTRAN_UNBUFFERED_ALL:: Don't buffer output
516 * GFORTRAN_SHOW_LOCUS:: Show location for runtime errors
517 * GFORTRAN_OPTIONAL_PLUS:: Print leading + where permitted
518 * GFORTRAN_DEFAULT_RECL:: Default record length for new files
519 * GFORTRAN_LIST_SEPARATOR:: Separator for list output
520 * GFORTRAN_CONVERT_UNIT:: Set endianness for unformatted I/O
523 @node GFORTRAN_STDIN_UNIT
524 @section @env{GFORTRAN_STDIN_UNIT}---Unit number for standard input
526 This environment variable can be used to select the unit number
527 preconnected to standard input. This must be a positive integer.
528 The default value is 5.
530 @node GFORTRAN_STDOUT_UNIT
531 @section @env{GFORTRAN_STDOUT_UNIT}---Unit number for standard output
533 This environment variable can be used to select the unit number
534 preconnected to standard output. This must be a positive integer.
535 The default value is 6.
537 @node GFORTRAN_STDERR_UNIT
538 @section @env{GFORTRAN_STDERR_UNIT}---Unit number for standard error
540 This environment variable can be used to select the unit number
541 preconnected to standard error. This must be a positive integer.
542 The default value is 0.
544 @node GFORTRAN_USE_STDERR
545 @section @env{GFORTRAN_USE_STDERR}---Send library output to standard error
547 This environment variable controls where library output is sent.
548 If the first letter is @samp{y}, @samp{Y} or @samp{1}, standard
549 error is used. If the first letter is @samp{n}, @samp{N} or
550 @samp{0}, standard output is used.
552 @node GFORTRAN_TMPDIR
553 @section @env{GFORTRAN_TMPDIR}---Directory for scratch files
555 This environment variable controls where scratch files are
556 created. If this environment variable is missing,
557 GNU Fortran searches for the environment variable @env{TMP}. If
558 this is also missing, the default is @file{/tmp}.
560 @node GFORTRAN_UNBUFFERED_ALL
561 @section @env{GFORTRAN_UNBUFFERED_ALL}---Don't buffer output
563 This environment variable controls whether all output is unbuffered.
564 If the first letter is @samp{y}, @samp{Y} or @samp{1}, all output is
565 unbuffered. This will slow down large writes. If the first letter is
566 @samp{n}, @samp{N} or @samp{0}, output is buffered. This is the
569 @node GFORTRAN_SHOW_LOCUS
570 @section @env{GFORTRAN_SHOW_LOCUS}---Show location for runtime errors
572 If the first letter is @samp{y}, @samp{Y} or @samp{1}, filename and
573 line numbers for runtime errors are printed. If the first letter is
574 @samp{n}, @samp{N} or @samp{0}, don't print filename and line numbers
575 for runtime errors. The default is to print the location.
577 @node GFORTRAN_OPTIONAL_PLUS
578 @section @env{GFORTRAN_OPTIONAL_PLUS}---Print leading + where permitted
580 If the first letter is @samp{y}, @samp{Y} or @samp{1},
581 a plus sign is printed
582 where permitted by the Fortran standard. If the first letter
583 is @samp{n}, @samp{N} or @samp{0}, a plus sign is not printed
584 in most cases. Default is not to print plus signs.
586 @node GFORTRAN_DEFAULT_RECL
587 @section @env{GFORTRAN_DEFAULT_RECL}---Default record length for new files
589 This environment variable specifies the default record length, in
590 bytes, for files which are opened without a @code{RECL} tag in the
591 @code{OPEN} statement. This must be a positive integer. The
592 default value is 1073741824 bytes (1 GB).
594 @node GFORTRAN_LIST_SEPARATOR
595 @section @env{GFORTRAN_LIST_SEPARATOR}---Separator for list output
597 This environment variable specifies the separator when writing
598 list-directed output. It may contain any number of spaces and
599 at most one comma. If you specify this on the command line,
600 be sure to quote spaces, as in
602 $ GFORTRAN_LIST_SEPARATOR=' , ' ./a.out
604 when @command{a.out} is the compiled Fortran program that you want to run.
605 Default is a single space.
607 @node GFORTRAN_CONVERT_UNIT
608 @section @env{GFORTRAN_CONVERT_UNIT}---Set endianness for unformatted I/O
610 By setting the @env{GFORTRAN_CONVERT_UNIT} variable, it is possible
611 to change the representation of data for unformatted files.
612 The syntax for the @env{GFORTRAN_CONVERT_UNIT} variable is:
614 GFORTRAN_CONVERT_UNIT: mode | mode ';' exception ;
615 mode: 'native' | 'swap' | 'big_endian' | 'little_endian' ;
616 exception: mode ':' unit_list | unit_list ;
617 unit_list: unit_spec | unit_list unit_spec ;
618 unit_spec: INTEGER | INTEGER '-' INTEGER ;
620 The variable consists of an optional default mode, followed by
621 a list of optional exceptions, which are separated by semicolons
622 from the preceding default and each other. Each exception consists
623 of a format and a comma-separated list of units. Valid values for
624 the modes are the same as for the @code{CONVERT} specifier:
627 @item @code{NATIVE} Use the native format. This is the default.
628 @item @code{SWAP} Swap between little- and big-endian.
629 @item @code{LITTLE_ENDIAN} Use the little-endian format
630 for unformatted files.
631 @item @code{BIG_ENDIAN} Use the big-endian format for unformatted files.
633 A missing mode for an exception is taken to mean @code{BIG_ENDIAN}.
634 Examples of values for @env{GFORTRAN_CONVERT_UNIT} are:
636 @item @code{'big_endian'} Do all unformatted I/O in big_endian mode.
637 @item @code{'little_endian;native:10-20,25'} Do all unformatted I/O
638 in little_endian mode, except for units 10 to 20 and 25, which are in
640 @item @code{'10-20'} Units 10 to 20 are big-endian, the rest is native.
643 Setting the environment variables should be done on the command
644 line or via the @command{export}
645 command for @command{sh}-compatible shells and via @command{setenv}
646 for @command{csh}-compatible shells.
648 Example for @command{sh}:
651 $ GFORTRAN_CONVERT_UNIT='big_endian;native:10-20' ./a.out
654 Example code for @command{csh}:
657 % setenv GFORTRAN_CONVERT_UNIT 'big_endian;native:10-20'
661 Using anything but the native representation for unformatted data
662 carries a significant speed overhead. If speed in this area matters
663 to you, it is best if you use this only for data that needs to be
666 @xref{CONVERT specifier}, for an alternative way to specify the
667 data representation for unformatted files. @xref{Runtime Options}, for
668 setting a default data representation for the whole program. The
669 @code{CONVERT} specifier overrides the @option{-fconvert} compile options.
672 @c =====================================================================
673 @c PART II: LANGUAGE REFERENCE
674 @c =====================================================================
677 \part{II}{Language Reference}
680 @c ---------------------------------------------------------------------
681 @c Fortran 2003 Status
682 @c ---------------------------------------------------------------------
684 @node Fortran 2003 status
685 @chapter Fortran 2003 Status
687 Although GNU Fortran focuses on implementing the Fortran 95
688 standard for the time being, a few Fortran 2003 features are currently
693 Intrinsics @code{command_argument_count}, @code{get_command},
694 @code{get_command_argument}, @code{get_environment_variable}, and
698 @cindex Array constructors
700 Array constructors using square brackets. That is, @code{[...]} rather
704 @cindex @code{FLUSH} statement
705 @code{FLUSH} statement.
708 @cindex @code{IOMSG=} specifier
709 @code{IOMSG=} specifier for I/O statements.
712 @cindex @code{ENUM} statement
713 @cindex @code{ENUMERATOR} statement
714 @opindex @code{fshort-enums}
715 Support for the declaration of enumeration constants via the
716 @code{ENUM} and @code{ENUMERATOR} statements. Interoperability with
717 @command{gcc} is guaranteed also for the case where the
718 @command{-fshort-enums} command line option is given.
725 @cindex @code{ALLOCATABLE} dummy arguments
726 @code{ALLOCATABLE} dummy arguments.
728 @cindex @code{ALLOCATABLE} function results
729 @code{ALLOCATABLE} function results
731 @cindex @code{ALLOCATABLE} components of derived types
732 @code{ALLOCATABLE} components of derived types
736 @cindex @code{STREAM} I/O
737 @cindex @code{ACCESS='STREAM'} I/O
738 The @code{OPEN} statement supports the @code{ACCESS='STREAM'} specifier,
739 allowing I/O without any record structure.
742 Namelist input/output for internal files.
745 @cindex @code{PROTECTED}
746 The @code{PROTECTED} statement and attribute.
750 The @code{VALUE} statement and attribute.
753 @cindex @code{VOLATILE}
754 The @code{VOLATILE} statement and attribute.
757 @cindex @code{IMPORT}
758 The @code{IMPORT} statement, allowing to import
759 host-associated derived types.
762 @cindex @code{USE, INTRINSIC}
763 @cindex @code{ISO_FORTRAN_ENV}
764 @code{USE} statement with @code{INTRINSIC} and @code{NON_INTRINSIC}
765 attribute; supported intrinsic modules: @code{ISO_FORTRAN_ENV},
766 @code{OMP_LIB} and @code{OMP_LIB_KINDS}.
769 Renaming of operators in the @code{USE} statement.
774 @c ---------------------------------------------------------------------
776 @c ---------------------------------------------------------------------
778 @c Maybe this chapter should be merged with the 'Standards' section,
779 @c whenever that is written :-)
785 GNU Fortran implements a number of extensions over standard
786 Fortran. This chapter contains information on their syntax and
787 meaning. There are currently two categories of GNU Fortran
788 extensions, those that provide functionality beyond that provided
789 by any standard, and those that are supported by GNU Fortran
790 purely for backward compatibility with legacy compilers. By default,
791 @option{-std=gnu} allows the compiler to accept both types of
792 extensions, but to warn about the use of the latter. Specifying
793 either @option{-std=f95} or @option{-std=f2003} disables both types
794 of extensions, and @option{-std=legacy} allows both without warning.
797 * Old-style kind specifications::
798 * Old-style variable initialization::
799 * Extensions to namelist::
800 * X format descriptor without count field::
801 * Commas in FORMAT specifications::
802 * Missing period in FORMAT specifications::
804 * BOZ literal constants::
805 * Real array indices::
807 * Implicitly convert LOGICAL and INTEGER values::
808 * Hollerith constants support::
810 * CONVERT specifier::
812 * Argument list functions::
815 @node Old-style kind specifications
816 @section Old-style kind specifications
817 @cindex Kind specifications
819 GNU Fortran allows old-style kind specifications in declarations. These
825 where @code{TYPESPEC} is a basic type (@code{INTEGER}, @code{REAL},
826 etc.), and where @code{size} is a byte count corresponding to the
827 storage size of a valid kind for that type. (For @code{COMPLEX}
828 variables, @code{size} is the total size of the real and imaginary
829 parts.) The statement then declares @code{x}, @code{y} and @code{z} to
830 be of type @code{TYPESPEC} with the appropriate kind. This is
831 equivalent to the standard-conforming declaration
836 where @code{k} is equal to @code{size} for most types, but is equal to
837 @code{size/2} for the @code{COMPLEX} type.
839 @node Old-style variable initialization
840 @section Old-style variable initialization
841 @cindex Initialization
843 GNU Fortran allows old-style initialization of variables of the
847 REAL x(2,2) /3*0.,1./
849 The syntax for the initializers is as for the @code{DATA} statement, but
850 unlike in a @code{DATA} statement, an initializer only applies to the
851 variable immediately preceding the initialization. In other words,
852 something like @code{INTEGER I,J/2,3/} is not valid. This style of
853 initialization is only allowed in declarations without double colons
854 (@code{::}); the double colons were introduced in Fortran 90, which also
855 introduced a standard syntax for initializing variables in type
858 Examples of standard-conforming code equivalent to the above example
862 INTEGER :: i = 1, j = 2
863 REAL :: x(2,2) = RESHAPE((/0.,0.,0.,1./),SHAPE(x))
867 DATA i/1/, j/2/, x/3*0.,1./
870 Note that variables which are explicitly initialized in declarations
871 or in @code{DATA} statements automatically acquire the @code{SAVE}
874 @node Extensions to namelist
875 @section Extensions to namelist
878 GNU Fortran fully supports the Fortran 95 standard for namelist I/O
879 including array qualifiers, substrings and fully qualified derived types.
880 The output from a namelist write is compatible with namelist read. The
881 output has all names in upper case and indentation to column 1 after the
882 namelist name. Two extensions are permitted:
884 Old-style use of @samp{$} instead of @samp{&}
887 X(:)%Y(2) = 1.0 2.0 3.0
892 It should be noted that the default terminator is @samp{/} rather than
895 Querying of the namelist when inputting from stdin. After at least
896 one space, entering @samp{?} sends to stdout the namelist name and the names of
897 the variables in the namelist:
908 Entering @samp{=?} outputs the namelist to stdout, as if
909 @code{WRITE(*,NML = mynml)} had been called:
914 X(1)%Y= 0.000000 , 1.000000 , 0.000000 ,
915 X(2)%Y= 0.000000 , 2.000000 , 0.000000 ,
916 X(3)%Y= 0.000000 , 3.000000 , 0.000000 ,
920 To aid this dialog, when input is from stdin, errors send their
921 messages to stderr and execution continues, even if @code{IOSTAT} is set.
923 @code{PRINT} namelist is permitted. This causes an error if
924 @option{-std=f95} is used.
927 REAL, dimension (4) :: x = (/1.0, 2.0, 3.0, 4.0/)
930 END PROGRAM test_print
933 Expanded namelist reads are permitted. This causes an error if
934 @option{-std=f95} is used. In the following example, the first element
935 of the array will be given the value 0.00 and the two succeeding
936 elements will be given the values 1.00 and 2.00.
939 X(1,1) = 0.00 , 1.00 , 2.00
943 @node X format descriptor without count field
944 @section @code{X} format descriptor without count field
945 @cindex @code{X} format descriptor without count field
947 To support legacy codes, GNU Fortran permits the count field of the
948 @code{X} edit descriptor in @code{FORMAT} statements to be omitted.
949 When omitted, the count is implicitly assumed to be one.
953 10 FORMAT (I1, X, I1)
956 @node Commas in FORMAT specifications
957 @section Commas in @code{FORMAT} specifications
958 @cindex Commas in @code{FORMAT} specifications
960 To support legacy codes, GNU Fortran allows the comma separator
961 to be omitted immediately before and after character string edit
962 descriptors in @code{FORMAT} statements.
966 10 FORMAT ('FOO='I1' BAR='I2)
970 @node Missing period in FORMAT specifications
971 @section Missing period in @code{FORMAT} specifications
972 @cindex Missing period in @code{FORMAT} specifications
974 To support legacy codes, GNU Fortran allows missing periods in format
975 specifications if and only if @option{-std=legacy} is given on the
976 command line. This is considered non-conforming code and is
986 @section I/O item lists
987 @cindex I/O item lists
989 To support legacy codes, GNU Fortran allows the input item list
990 of the @code{READ} statement, and the output item lists of the
991 @code{WRITE} and @code{PRINT} statements, to start with a comma.
993 @node BOZ literal constants
994 @section BOZ literal constants
995 @cindex BOZ literal constants
997 As an extension, GNU Fortran allows hexadecimal BOZ literal constants to
998 be specified using the X prefix, in addition to the standard Z prefix.
999 BOZ literal constants can also be specified by adding a suffix to the
1000 string. For example, @code{Z'ABC'} and @code{'ABC'Z} are equivalent.
1002 The Fortran standard restricts the appearance of a BOZ literal constant
1003 to the @code{DATA} statement, and it is expected to be assigned to an
1004 @code{INTEGER} variable. GNU Fortran permits a BOZ literal to appear in
1005 any initialization expression as well as assignment statements.
1007 Attempts to use a BOZ literal constant to do a bitwise initialization of
1008 a variable can lead to confusion. A BOZ literal constant is converted
1009 to an @code{INTEGER} value with the kind type with the largest decimal
1010 representation, and this value is then converted numerically to the type
1011 and kind of the variable in question. Thus, one should not expect a
1012 bitwise copy of the BOZ literal constant to be assigned to a @code{REAL}
1015 Similarly, initializing an @code{INTEGER} variable with a statement such
1016 as @code{DATA i/Z'FFFFFFFF'/} will produce an integer overflow rather
1017 than the desired result of @math{-1} when @code{i} is a 32-bit integer
1018 on a system that supports 64-bit integers. The @samp{-fno-range-check}
1019 option can be used as a workaround for legacy code that initializes
1020 integers in this manner.
1022 @node Real array indices
1023 @section Real array indices
1024 @cindex Real array indices
1026 As an extension, GNU Fortran allows the use of @code{REAL} expressions
1027 or variables as array indices.
1029 @node Unary operators
1030 @section Unary operators
1031 @cindex Unary operators
1033 As an extension, GNU Fortran allows unary plus and unary minus operators
1034 to appear as the second operand of binary arithmetic operators without
1035 the need for parenthesis.
1041 @node Implicitly convert LOGICAL and INTEGER values
1042 @section Implicitly convert @code{LOGICAL} and @code{INTEGER} values
1043 @cindex Implicitly convert @code{LOGICAL} and @code{INTEGER} values
1045 As an extension for backwards compatibility with other compilers, GNU
1046 Fortran allows the implicit conversion of @code{LOGICAL} values to
1047 @code{INTEGER} values and vice versa. When converting from a
1048 @code{LOGICAL} to an @code{INTEGER}, @code{.FALSE.} is interpreted as
1049 zero, and @code{.TRUE.} is interpreted as one. When converting from
1050 @code{INTEGER} to @code{LOGICAL}, the value zero is interpreted as
1051 @code{.FALSE.} and any nonzero value is interpreted as @code{.TRUE.}.
1055 IF (i) PRINT *, 'True'
1058 @node Hollerith constants support
1059 @section Hollerith constants support
1060 @cindex Hollerith constants
1062 GNU Fortran supports Hollerith constants in assignments, function
1063 arguments, and @code{DATA} and @code{ASSIGN} statements. A Hollerith
1064 constant is written as a string of characters preceded by an integer
1065 constant indicating the character count, and the letter @code{H} or
1066 @code{h}, and stored in bytewise fashion in a numeric (@code{INTEGER},
1067 @code{REAL}, or @code{complex}) or @code{LOGICAL} variable. The
1068 constant will be padded or truncated to fit the size of the variable in
1071 Examples of valid uses of Hollerith constants:
1074 data x /16Habcdefghijklmnop, 16Hqrstuvwxyz012345/
1075 x(1) = 16HABCDEFGHIJKLMNOP
1079 Invalid Hollerith constants examples:
1082 a = 8H12345678 ! Valid, but the Hollerith constant will be truncated.
1083 a = 0H ! At least one character is needed.
1086 In general, Hollerith constants were used to provide a rudimentary
1087 facility for handling character strings in early Fortran compilers,
1088 prior to the introduction of @code{CHARACTER} variables in Fortran 77;
1089 in those cases, the standard-compliant equivalent is to convert the
1090 program to use proper character strings. On occasion, there may be a
1091 case where the intent is specifically to initialize a numeric variable
1092 with a given byte sequence. In these cases, the same result can be
1093 obtained by using the @code{TRANSFER} statement, as in this example.
1095 INTEGER(KIND=4) :: a
1096 a = TRANSFER ("abcd", a) ! equivalent to: a = 4Habcd
1101 @section Cray pointers
1102 @cindex Cray pointers
1104 Cray pointers are part of a non-standard extension that provides a
1105 C-like pointer in Fortran. This is accomplished through a pair of
1106 variables: an integer "pointer" that holds a memory address, and a
1107 "pointee" that is used to dereference the pointer.
1109 Pointer/pointee pairs are declared in statements of the form:
1111 pointer ( <pointer> , <pointee> )
1115 pointer ( <pointer1> , <pointee1> ), ( <pointer2> , <pointee2> ), ...
1117 The pointer is an integer that is intended to hold a memory address.
1118 The pointee may be an array or scalar. A pointee can be an assumed
1119 size array---that is, the last dimension may be left unspecified by
1120 using a @code{*} in place of a value---but a pointee cannot be an
1121 assumed shape array. No space is allocated for the pointee.
1123 The pointee may have its type declared before or after the pointer
1124 statement, and its array specification (if any) may be declared
1125 before, during, or after the pointer statement. The pointer may be
1126 declared as an integer prior to the pointer statement. However, some
1127 machines have default integer sizes that are different than the size
1128 of a pointer, and so the following code is not portable:
1133 If a pointer is declared with a kind that is too small, the compiler
1134 will issue a warning; the resulting binary will probably not work
1135 correctly, because the memory addresses stored in the pointers may be
1136 truncated. It is safer to omit the first line of the above example;
1137 if explicit declaration of ipt's type is omitted, then the compiler
1138 will ensure that ipt is an integer variable large enough to hold a
1141 Pointer arithmetic is valid with Cray pointers, but it is not the same
1142 as C pointer arithmetic. Cray pointers are just ordinary integers, so
1143 the user is responsible for determining how many bytes to add to a
1144 pointer in order to increment it. Consider the following example:
1148 pointer (ipt, pointee)
1152 The last statement does not set @code{ipt} to the address of
1153 @code{target(1)}, as it would in C pointer arithmetic. Adding @code{1}
1154 to @code{ipt} just adds one byte to the address stored in @code{ipt}.
1156 Any expression involving the pointee will be translated to use the
1157 value stored in the pointer as the base address.
1159 To get the address of elements, this extension provides an intrinsic
1160 function @code{LOC()}. The @code{LOC()} function is equivalent to the
1161 @code{&} operator in C, except the address is cast to an integer type:
1164 pointer(ipt, arpte(10))
1166 ipt = loc(ar) ! Makes arpte is an alias for ar
1167 arpte(1) = 1.0 ! Sets ar(1) to 1.0
1169 The pointer can also be set by a call to the @code{MALLOC} intrinsic
1172 Cray pointees often are used to alias an existing variable. For
1180 As long as @code{ipt} remains unchanged, @code{iarr} is now an alias for
1181 @code{target}. The optimizer, however, will not detect this aliasing, so
1182 it is unsafe to use @code{iarr} and @code{target} simultaneously. Using
1183 a pointee in any way that violates the Fortran aliasing rules or
1184 assumptions is illegal. It is the user's responsibility to avoid doing
1185 this; the compiler works under the assumption that no such aliasing
1188 Cray pointers will work correctly when there is no aliasing (i.e., when
1189 they are used to access a dynamically allocated block of memory), and
1190 also in any routine where a pointee is used, but any variable with which
1191 it shares storage is not used. Code that violates these rules may not
1192 run as the user intends. This is not a bug in the optimizer; any code
1193 that violates the aliasing rules is illegal. (Note that this is not
1194 unique to GNU Fortran; any Fortran compiler that supports Cray pointers
1195 will ``incorrectly'' optimize code with illegal aliasing.)
1197 There are a number of restrictions on the attributes that can be applied
1198 to Cray pointers and pointees. Pointees may not have the
1199 @code{ALLOCATABLE}, @code{INTENT}, @code{OPTIONAL}, @code{DUMMY},
1200 @code{TARGET}, @code{INTRINSIC}, or @code{POINTER} attributes. Pointers
1201 may not have the @code{DIMENSION}, @code{POINTER}, @code{TARGET},
1202 @code{ALLOCATABLE}, @code{EXTERNAL}, or @code{INTRINSIC} attributes.
1203 Pointees may not occur in more than one pointer statement. A pointee
1204 cannot be a pointer. Pointees cannot occur in equivalence, common, or
1207 A Cray pointer may also point to a function or a subroutine. For
1208 example, the following excerpt is valid:
1212 pointer (subptr,subpte)
1222 A pointer may be modified during the course of a program, and this
1223 will change the location to which the pointee refers. However, when
1224 pointees are passed as arguments, they are treated as ordinary
1225 variables in the invoked function. Subsequent changes to the pointer
1226 will not change the base address of the array that was passed.
1228 @node CONVERT specifier
1229 @section CONVERT specifier
1230 @cindex CONVERT specifier
1232 GNU Fortran allows the conversion of unformatted data between little-
1233 and big-endian representation to facilitate moving of data
1234 between different systems. The conversion can be indicated with
1235 the @code{CONVERT} specifier on the @code{OPEN} statement.
1236 @xref{GFORTRAN_CONVERT_UNIT}, for an alternative way of specifying
1237 the data format via an environment variable.
1239 Valid values for @code{CONVERT} are:
1241 @item @code{CONVERT='NATIVE'} Use the native format. This is the default.
1242 @item @code{CONVERT='SWAP'} Swap between little- and big-endian.
1243 @item @code{CONVERT='LITTLE_ENDIAN'} Use the little-endian representation
1244 for unformatted files.
1245 @item @code{CONVERT='BIG_ENDIAN'} Use the big-endian representation for
1249 Using the option could look like this:
1251 open(file='big.dat',form='unformatted',access='sequential', &
1252 convert='big_endian')
1255 The value of the conversion can be queried by using
1256 @code{INQUIRE(CONVERT=ch)}. The values returned are
1257 @code{'BIG_ENDIAN'} and @code{'LITTLE_ENDIAN'}.
1259 @code{CONVERT} works between big- and little-endian for
1260 @code{INTEGER} values of all supported kinds and for @code{REAL}
1261 on IEEE systems of kinds 4 and 8. Conversion between different
1262 ``extended double'' types on different architectures such as
1263 m68k and x86_64, which GNU Fortran
1264 supports as @code{REAL(KIND=10)} and @code{REAL(KIND=16)}, will
1267 @emph{Note that the values specified via the GFORTRAN_CONVERT_UNIT
1268 environment variable will override the CONVERT specifier in the
1269 open statement}. This is to give control over data formats to
1270 users who do not have the source code of their program available.
1272 Using anything but the native representation for unformatted data
1273 carries a significant speed overhead. If speed in this area matters
1274 to you, it is best if you use this only for data that needs to be
1281 GNU Fortran attempts to be OpenMP Application Program Interface v2.5
1282 compatible when invoked with the @option{-fopenmp} option. GNU Fortran
1283 then generates parallelized code according to the OpenMP directives
1284 used in the source. The OpenMP Fortran runtime library
1285 routines are provided both in a form of a Fortran 90 module named
1286 @code{omp_lib} and in a form of a Fortran @code{include} file named
1289 For details refer to the actual
1290 @uref{http://www.openmp.org/drupal/mp-documents/spec25.pdf,
1291 OpenMP Application Program Interface v2.5} specification.
1293 @node Argument list functions
1294 @section Argument list functions %VAL, %REF and %LOC
1295 @cindex Argument list functions %VAL, %REF and %LOC
1297 GNU Fortran supports argument list functions @code{%VAL}, @code{%REF}
1298 and @code{%LOC} statements, for backward compatibility with g77.
1299 It is recommended that these should be used only for code that is
1300 accessing facilities outside of GNU Fortran, such as operating system
1301 or windowing facilities. It is best to constrain such uses to isolated
1302 portions of a program--portions that deal specifically and exclusively
1303 with low-level, system-dependent facilities. Such portions might well
1304 provide a portable interface for use by the program as a whole, but are
1305 themselves not portable, and should be thoroughly tested each time they
1306 are rebuilt using a new compiler or version of a compiler.
1308 @code{%VAL} passes a scalar argument by value, @code{%REF} passes it by
1309 reference and @code{%LOC} passes its memory location. Since gfortran
1310 already passes scalar arguments by reference, @code{%REF} is in effect
1311 a do-nothing. @code{%LOC} has the same effect as a fortran pointer.
1313 An example of passing an argument by value to a C subroutine foo.:
1316 C prototype void foo_ (float x);
1325 For details refer to the g77 manual
1326 @uref{http://gcc.gnu.org/onlinedocs/gcc-3.4.6/g77/index.html#Top}.
1328 Also, the gfortran testsuite c_by_val.f and its partner c_by_val.c are
1331 @c ---------------------------------------------------------------------
1332 @c Intrinsic Procedures
1333 @c ---------------------------------------------------------------------
1335 @include intrinsic.texi
1342 @c ---------------------------------------------------------------------
1344 @c ---------------------------------------------------------------------
1347 @unnumbered Contributing
1348 @cindex Contributing
1350 Free software is only possible if people contribute to efforts
1352 We're always in need of more people helping out with ideas
1353 and comments, writing documentation and contributing code.
1355 If you want to contribute to GNU Fortran,
1356 have a look at the long lists of projects you can take on.
1357 Some of these projects are small,
1358 some of them are large;
1359 some are completely orthogonal to the rest of what is
1360 happening on GNU Fortran,
1361 but others are ``mainstream'' projects in need of enthusiastic hackers.
1362 All of these projects are important!
1363 We'll eventually get around to the things here,
1364 but they are also things doable by someone who is willing and able.
1369 * Proposed Extensions::
1374 @section Contributors to GNU Fortran
1375 @cindex Contributors
1379 Most of the parser was hand-crafted by @emph{Andy Vaught}, who is
1380 also the initiator of the whole project. Thanks Andy!
1381 Most of the interface with GCC was written by @emph{Paul Brook}.
1383 The following individuals have contributed code and/or
1384 ideas and significant help to the GNU Fortran project
1385 (in no particular order):
1389 @item Katherine Holcomb
1390 @item Tobias Schl@"uter
1391 @item Steven Bosscher
1394 @item Niels Kristian Bech Jensen
1395 @item Steven Johnson
1400 @item Fran@,{c}ois-Xavier Coudert
1401 @item Steven G. Kargl
1403 @item Janne Blomqvist
1410 @item Richard Henderson
1411 @item Richard Sandiford
1412 @item Richard Guenther
1413 @item Bernhard Fischer
1416 The following people have contributed bug reports,
1417 smaller or larger patches,
1418 and much needed feedback and encouragement for the
1419 GNU Fortran project:
1422 @item Erik Schnetter
1427 Many other individuals have helped debug,
1428 test and improve the GNU Fortran compiler over the past few years,
1429 and we welcome you to do the same!
1430 If you already have done so,
1431 and you would like to see your name listed in the
1432 list above, please contact us.
1440 @item Help build the test suite
1441 Solicit more code for donation to the test suite.
1442 We can keep code private on request.
1444 @item Bug hunting/squishing
1445 Find bugs and write more test cases!
1446 Test cases are especially very welcome,
1447 because it allows us to concentrate on fixing bugs
1448 instead of isolating them.
1450 @item Smaller projects (``bug'' fixes):
1452 @item Allow init exprs to be numbers raised to integer powers.
1453 @item Implement correct rounding.
1454 @item Implement F restrictions on Fortran 95 syntax.
1455 @item See about making Emacs-parsable error messages.
1459 If you wish to work on the runtime libraries,
1460 please contact a project maintainer.
1464 @node Proposed Extensions
1465 @section Proposed Extensions
1467 Here's a list of proposed extensions for the GNU Fortran compiler, in no particular
1468 order. Most of these are necessary to be fully compatible with
1469 existing Fortran compilers, but they are not part of the official
1470 J3 Fortran 95 standard.
1472 @subsection Compiler extensions:
1475 User-specified alignment rules for structures.
1478 Flag to generate @code{Makefile} info.
1481 Automatically extend single precision constants to double.
1484 Compile code that conserves memory by dynamically allocating common and
1485 module storage either on stack or heap.
1488 Compile flag to generate code for array conformance checking (suggest -CC).
1491 User control of symbol names (underscores, etc).
1494 Compile setting for maximum size of stack frame size before spilling
1495 parts to static or heap.
1498 Flag to force local variables into static space.
1501 Flag to force local variables onto stack.
1504 Flag for maximum errors before ending compile.
1507 Option to initialize otherwise uninitialized integer and floating
1512 @subsection Environment Options
1515 Pluggable library modules for random numbers, linear algebra.
1516 LA should use BLAS calling conventions.
1519 Environment variables controlling actions on arithmetic exceptions like
1520 overflow, underflow, precision loss---Generate NaN, abort, default.
1524 Set precision for fp units that support it (i387).
1527 Variable for setting fp rounding mode.
1530 Variable to fill uninitialized variables with a user-defined bit
1534 Environment variable controlling filename that is opened for that unit
1538 Environment variable to clear/trash memory being freed.
1541 Environment variable to control tracing of allocations and frees.
1544 Environment variable to display allocated memory at normal program end.
1547 Environment variable for filename for * IO-unit.
1550 Environment variable for temporary file directory.
1553 Environment variable forcing standard output to be line buffered (unix).
1558 @c ---------------------------------------------------------------------
1559 @c GNU General Public License
1560 @c ---------------------------------------------------------------------
1566 @c ---------------------------------------------------------------------
1567 @c GNU Free Documentation License
1568 @c ---------------------------------------------------------------------
1574 @c ---------------------------------------------------------------------
1575 @c Funding Free Software
1576 @c ---------------------------------------------------------------------
1578 @include funding.texi
1580 @c ---------------------------------------------------------------------
1582 @c ---------------------------------------------------------------------
1585 @unnumbered Index of command line options
1586 @command{gfortran}'s command line options are indexed here without any
1587 initial `-' or `--'. Where an option has both positive and negative forms
1588 (such as -foption and -fno-option), relevant entries in the manual are
1589 indexed under the most appropriate form; it may sometimes be useful to
1594 @unnumbered Index of concepts