1 \input texinfo @c -*-texinfo-*-
3 @setfilename gfortran.info
4 @set copyrights-gfortran 1999-2005
6 @include gcc-common.texi
8 @settitle The GNU Fortran 95 Compiler
10 @c Create a separate index for command line options
12 @c Merge the standard indexes into a single one.
21 @c Use with @@smallbook.
23 @c %** start of document
25 @c Cause even numbered pages to be printed on the left hand side of
26 @c the page and odd numbered pages to be printed on the right hand
27 @c side of the page. Using this, you can print on both sides of a
28 @c sheet of paper and have the text on the same part of the sheet.
30 @c The text on right hand pages is pushed towards the right hand
31 @c margin and the text on left hand pages is pushed toward the left
33 @c (To provide the reverse effect, set bindingoffset to -0.75in.)
36 @c \global\bindingoffset=0.75in
37 @c \global\normaloffset =0.75in
41 Copyright @copyright{} @value{copyrights-gfortran} Free Software Foundation, Inc.
43 Permission is granted to copy, distribute and/or modify this document
44 under the terms of the GNU Free Documentation License, Version 1.1 or
45 any later version published by the Free Software Foundation; with the
46 Invariant Sections being ``GNU General Public License'' and ``Funding
47 Free Software'', the Front-Cover
48 texts being (a) (see below), and with the Back-Cover Texts being (b)
49 (see below). A copy of the license is included in the section entitled
50 ``GNU Free Documentation License''.
52 (a) The FSF's Front-Cover Text is:
56 (b) The FSF's Back-Cover Text is:
58 You have freedom to copy and modify this GNU Manual, like GNU
59 software. Copies published by the Free Software Foundation raise
60 funds for GNU development.
64 @dircategory Programming
66 * gfortran: (gfortran). The GNU Fortran 95 Compiler.
68 This file documents the use and the internals of
69 the GNU Fortran 95 compiler, (@command{gfortran}).
71 Published by the Free Software Foundation
72 51 Franklin Street, Fifth Floor
73 Boston, MA 02110-1301 USA
79 @setchapternewpage odd
81 @title Using GNU Fortran 95
83 @center The gfortran team
85 @vskip 0pt plus 1filll
86 For the @value{version-GCC} Version*
88 Published by the Free Software Foundation @*
89 51 Franklin Street, Fifth Floor@*
90 Boston, MA 02110-1301, USA@*
91 @c Last printed ??ber, 19??.@*
92 @c Printed copies are available for $? each.@*
101 @node Top, Copying,, (DIR)
105 This manual documents the use of @command{gfortran},
106 the GNU Fortran 95 compiler. You can find in this manual how to invoke
107 @command{gfortran}, as well as its features and incompatibilities.
110 @emph{Warning:} This document, and the compiler it describes, are still
111 under development. While efforts are made to keep it up-to-date, it might
112 not accurately reflect the status of the most recent @command{gfortran}.
116 @comment When you add a new menu item, please keep the right hand
117 @comment aligned to the same column. Do not use tabs. This provides
118 @comment better formatting.
121 * Copying:: GNU General Public License says
122 how you can copy and share GNU Fortran.
123 * GNU Free Documentation License::
124 How you can copy and share this manual.
125 * Funding:: How to help assure continued work for free software.
126 * Getting Started:: What you should know about @command{gfortran}.
127 * GFORTRAN and GCC:: You can compile Fortran, C, or other programs.
128 * GFORTRAN and G77:: Why we chose to start from scratch.
129 * Invoking GFORTRAN:: Command options supported by @command{gfortran}.
130 * Project Status:: Status of @command{gfortran}, roadmap, proposed extensions.
131 * Contributing:: How you can help.
132 * Standards:: Standards supported by @command{gfortran}
133 * Extensions:: Language extensions implemented by @command{gfortran}
134 * Intrinsic Procedures:: Intrinsic procedures supported by @command{gfortran}
135 * Index:: Index of this documentation.
140 @c ---------------------------------------------------------------------
141 @c GNU General Public License
142 @c ---------------------------------------------------------------------
148 @c ---------------------------------------------------------------------
149 @c GNU Free Documentation License
150 @c ---------------------------------------------------------------------
156 @c ---------------------------------------------------------------------
157 @c Funding Free Software
158 @c ---------------------------------------------------------------------
160 @include funding.texi
164 @c ---------------------------------------------------------------------
166 @c ---------------------------------------------------------------------
168 @node Getting Started
169 @chapter Getting Started
171 Gfortran is the GNU Fortran 95 compiler front end,
172 designed initially as a free replacement for,
173 or alternative to, the unix @command{f95} command;
174 @command{gfortran} is the command you'll use to invoke the compiler.
176 Gfortran is still in an early state of development.
177 @command{gfortran} can generate code for most constructs and expressions,
178 but much work remains to be done.
180 When @command{gfortran} is finished,
181 it will do everything you expect from any decent compiler:
185 Read a user's program,
186 stored in a file and containing instructions written
187 in Fortran 77, Fortran 90 or Fortran 95.
188 This file contains @dfn{source code}.
191 Translate the user's program into instructions a computer
192 can carry out more quickly than it takes to translate the
193 instructions in the first
194 place. The result after compilation of a program is
196 code designed to be efficiently translated and processed
197 by a machine such as your computer.
198 Humans usually aren't as good writing machine code
199 as they are at writing Fortran (or C++, Ada, or Java),
200 because is easy to make tiny mistakes writing machine code.
203 Provide the user with information about the reasons why
204 the compiler is unable to create a binary from the source code.
205 Usually this will be the case if the source code is flawed.
206 When writing Fortran, it is easy to make big mistakes.
207 The Fortran 90 requires that the compiler can point out
208 mistakes to the user.
209 An incorrect usage of the language causes an @dfn{error message}.
211 The compiler will also attempt to diagnose cases where the
212 user's program contains a correct usage of the language,
213 but instructs the computer to do something questionable.
214 This kind of diagnostics message is called a @dfn{warning message}.
217 Provide optional information about the translation passes
218 from the source code to machine code.
219 This can help a user of the compiler to find the cause of
220 certain bugs which may not be obvious in the source code,
221 but may be more easily found at a lower level compiler output.
222 It also helps developers to find bugs in the compiler itself.
225 Provide information in the generated machine code that can
226 make it easier to find bugs in the program (using a debugging tool,
227 called a @dfn{debugger}, such as the GNU Debugger @command{gdb}).
230 Locate and gather machine code already generated to
231 perform actions requested by statements in the user's program.
232 This machine code is organized into @dfn{modules} and is located
233 and @dfn{linked} to the user program.
236 Gfortran consists of several components:
240 A version of the @command{gcc} command
241 (which also might be installed as the system's @command{cc} command)
242 that also understands and accepts Fortran source code.
243 The @command{gcc} command is the @dfn{driver} program for
244 all the languages in the GNU Compiler Collection (GCC);
246 you can compile the source code of any language for
247 which a front end is available in GCC.
250 The @command{gfortran} command itself,
251 which also might be installed as the
252 system's @command{f95} command.
253 @command{gfortran} is just another driver program,
254 but specifically for the Fortran 95 compiler only.
255 The difference with @command{gcc} is that @command{gfortran}
256 will automatically link the correct libraries to your program.
259 A collection of run-time libraries.
260 These libraries contain the machine code needed to support
261 capabilities of the Fortran language that are not directly
262 provided by the machine code generated by the
263 @command{gfortran} compilation phase,
264 such as intrinsic functions and subroutines,
265 and routines for interaction with files and the operating system.
266 @c and mechanisms to spawn,
267 @c unleash and pause threads in parallelized code.
270 The Fortran compiler itself, (@command{f951}).
271 This is the gfortran parser and code generator,
272 linked to and interfaced with the GCC backend library.
273 @command{f951} ``translates'' the source code to
274 assembler code. You would typically not use this
276 instead, the @command{gcc} or @command{gfortran} driver
277 programs will call it for you.
282 @c ---------------------------------------------------------------------
284 @c ---------------------------------------------------------------------
286 @node GFORTRAN and GCC
287 @chapter GFORTRAN and GCC
288 @cindex GNU Compiler Collection
290 GCC used to be the GNU ``C'' Compiler,
291 but is now known as the @dfn{GNU Compiler Collection}.
292 GCC provides the GNU system with a very versatile
293 compiler middle end (shared optimization passes),
294 and back ends (code generators) for many different
295 computer architectures and operating systems.
296 The code of the middle end and back end are shared by all
297 compiler front ends that are in the GNU Compiler Collection.
299 A GCC front end is essentially a source code parser
300 and an intermediate code generator. The code generator translates the
301 semantics of the source code into a language independent form called
304 The parser takes a source file written in a
305 particular computer language, reads and parses it,
306 and tries to make sure that the source code conforms to
308 Once the correctness of a program has been established,
309 the compiler will build a data structure known as the
310 @dfn{Abstract Syntax tree},
311 or just @dfn{AST} or ``tree'' for short.
312 This data structure represents the whole program
313 or a subroutine or a function.
314 The ``tree'' is passed to the GCC middle end,
315 which will perform optimization passes on it. The optimized AST is then
316 handed off too the back end which assembles the program unit.
318 Different phases in this translation process can be,
319 and in fact @emph{are} merged in many compiler front ends.
320 GNU Fortran 95 has a strict separation between the
321 parser and code generator.
323 The goal of the gfortran project is to build a new front end for GCC.
324 Specifically, a Fortran 95 front end.
325 In a non-gfortran installation,
326 @command{gcc} will not be able to compile Fortran 95 source code
327 (only the ``C'' front end has to be compiled if you want to build GCC,
328 all other languages are optional).
329 If you build GCC with gfortran, @command{gcc} will recognize
330 @file{.f/.f90/.f95} source files and accepts Fortran 95 specific
331 command line options.
335 @c ---------------------------------------------------------------------
337 @c ---------------------------------------------------------------------
339 @node GFORTRAN and G77
340 @chapter GFORTRAN and G77
344 Why do we write a compiler front end from scratch?
345 There's a fine Fortran 77 compiler in the
346 GNU Compiler Collection that accepts some features
347 of the Fortran 90 standard as extensions.
348 Why not start from there and revamp it?
350 One of the reasons is that Craig Burley, the author of G77,
351 has decided to stop working on the G77 front end.
352 On @uref{http://world.std.com/~burley/g77-why.html,
353 Craig explains the reasons for his decision to stop working on G77}
354 in one of the pages in his homepage.
355 Among the reasons is a lack of interest in improvements to
357 Users appear to be quite satisfied with @command{g77} as it is.
358 While @command{g77} is still being maintained (by Toon Moene),
359 it is unlikely that sufficient people will be willing
360 to completely rewrite the existing code.
362 But there are other reasons to start from scratch.
363 Many people, including Craig Burley,
364 no longer agreed with certain design decisions in the G77 front end.
365 Also, the interface of @command{g77} to the back end is written in
366 a style which is confusing and not up to date on recommended practice.
367 In fact, a full rewrite had already been planned for GCC 3.0.
369 When Craig decided to stop,
370 it just seemed to be a better idea to start a new project from scratch,
371 because it was expected to be easier to maintain code we
372 develop ourselves than to do a major overhaul of @command{g77} first,
373 and then build a Fortran 95 compiler out of it.
377 @c ---------------------------------------------------------------------
379 @c ---------------------------------------------------------------------
382 @chapter Project Status
385 As soon as gfortran can parse all of the statements correctly,
386 it will be in the ``larva'' state.
387 When we generate code, the ``puppa'' state.
388 When gfortran is done,
389 we'll see if it will be a beautiful butterfly,
390 or just a big bug....
392 --Andy Vaught, April 2000
395 The start of the GNU Fortran 95 project was announced on
396 the GCC homepage in March 18, 2000
397 (even though Andy had already been working on it for a while,
400 Gfortran is currently reaching the stage where is is able to compile real
401 world programs. However it is still under development and has many rough
407 * Proposed Extensions::
410 @node Compiler Status
411 @section Compiler Status
415 This is the part of gfortran which parses a source file, verifies that it
416 is valid Fortran 95, performs compile time replacement of constants
417 (PARAMETER variables) and reads and generate module files. This is
418 almost complete. Every Fortran 95 source should be accepted, and most
419 none-Fortran 95 source should be rejected. If you find a source file where
420 this is not true, please tell us. You can use the -fsyntax-only switch to
421 make gfortran quit after running the front end, effectively reducing it to
424 @item Middle end interface
425 These are the parts of gfortran that take the parse tree generated by the
426 front end and translate it to the GENERIC form required by the GCC back
427 end. Work is ongoing in these parts of gfortran, but a large part has
428 already been completed.
432 @section Library Status
434 Some intrinsic functions map directly to library functions, and in most
435 cases the name of the library function used depends on the type of the
436 arguments. For some intrinsics we generate inline code, and for others,
437 such as sin, cos and sqrt, we rely on the backend to use special
438 instructions in the floating point unit of the CPU if available, or to
439 fall back to a call to libm if these are not available.
441 Implementation of some non-elemental intrinsic functions (eg. DOT_PRODUCT,
442 AVERAGE) is not yet optimal. This is hard because we have to make decisions
443 whether to use inline code (good for small arrays as no function call
444 overhead occurs) or generate function calls (good for large arrays as it
445 allows use of hand-optimized assembly routines, SIMD instructions, etc.)
447 The IO library is still under development. The following features should be
448 usable for real programs:
452 @item Unformatted sequential
458 @item Formatted sequential ('T' edit descriptor, and others)
464 @item Unformatted direct access
465 @item Formatted direct access
468 Many Fortran programs only use a small subset of the available IO
469 capabilities, so your mileage may vary.
471 @node Proposed Extensions
472 @section Proposed Extensions
474 Here's a list of proposed extensions for @command{gfortran}, in no particular
475 order. Most of these are necessary to be fully compatible with
476 existing Fortran compilers, but they are not part of the official
477 J3 Fortran 95 standard.
479 @subsection Compiler extensions:
482 Flag for defining the kind number for default logicals.
485 User-specified alignment rules for structures.
487 Flag to generate @code{Makefile} info.
490 Automatically extend single precision constants to double.
493 Compile code that conserves memory by dynamically allocating common and
494 module storage either on stack or heap.
497 Flag to cause the compiler to distinguish between upper and lower case
498 names. The Fortran 95 standard does not distinguish them.
501 Compile flag to generate code for array conformance checking (suggest -CC).
504 User control of symbol names (underscores, etc).
507 Compile setting for maximum size of stack frame size before spilling
508 parts to static or heap.
511 Flag to force local variables into static space.
514 Flag to force local variables onto stack.
517 Flag to compile lines beginning with ``D''.
520 Flag to ignore lines beginning with ``D''.
523 Flag for maximum errors before ending compile.
526 Generate code to check for null pointer dereferences -- prints locus of
527 dereference instead of segfaulting. There was some discussion about this
528 option in the g95 development mailing list.
531 Allow setting the default unit number.
534 Option to initialize otherwise uninitialized integer and floating
538 Support for OpenMP directives. This also requires support from the runtime
539 library and the rest of the compiler.
542 Support for Fortran 200x. This includes several new features including
543 floating point exceptions, extended use of allocatable arrays, C
544 interoperability, Parameterizer data types and function pointers.
548 @subsection Environment Options
551 Pluggable library modules for random numbers, linear algebra.
552 LA should use BLAS calling conventions.
555 Environment variables controlling actions on arithmetic exceptions like
556 overflow, underflow, precision loss -- Generate NaN, abort, default.
560 Set precision for fp units that support it (i387).
563 Variable for setting fp rounding mode.
566 Variable to fill uninitialized variables with a user-defined bit
570 Environment variable controlling filename that is opened for that unit
574 Environment variable to clear/trash memory being freed.
577 Environment variable to control tracing of allocations and frees.
580 Environment variable to display allocated memory at normal program end.
583 Environment variable for filename for * IO-unit.
586 Environment variable for temporary file directory.
589 Environment variable forcing standard output to be line buffered (unix).
592 Variable for swapping endianness during unformatted read.
595 Variable for swapping Endianness during unformatted write.
598 @c ---------------------------------------------------------------------
600 @c ---------------------------------------------------------------------
602 @c Maybe this chapter should be merged with the 'Standards' section,
603 @c whenever that is written :-)
609 @command{gfortran} implements a number of extensions over standard
610 Fortran. This chapter contains information on their syntax and
611 meaning. There are currently two categories of @command{gfortran}
612 extensions, those that provide functionality beyond that provided
613 by any standard, and those that are supported by @command{gfortran}
614 purely for backward compatibility with legacy compilers. By default,
615 @option{-std=gnu} allows the compiler to accept both types of
616 extensions, but to warn about the use of the latter. Specifying
617 either @option{-std=f95} or @option{-std=f2003} disables both types
618 of extensions, and @option{-std=legacy} allows both without warning.
621 * Old-style kind specifications::
622 * Old-style variable initialization::
623 * Extensions to namelist::
624 * X format descriptor::
625 * Commas in FORMAT specifications::
627 * Hexadecimal constants::
628 * Real array indices::
630 * Implicitly interconvert LOGICAL and INTEGER::
631 * Hollerith constants support::
635 @node Old-style kind specifications
636 @section Old-style kind specifications
637 @cindex Kind specifications
639 @command{gfortran} allows old-style kind specifications in
640 declarations. These look like:
644 where @code{TYPESPEC} is a basic type, and where @code{k} is a valid kind
645 number for that type. The statement then declares @code{x}, @code{y}
646 and @code{z} to be of type @code{TYPESPEC} with kind @code{k}. In
647 other words, it is equivalent to the standard conforming declaration
652 @node Old-style variable initialization
653 @section Old-style variable initialization
654 @cindex Initialization
656 @command{gfortran} allows old-style initialization of variables of the
660 REAL*8 x(2,2) /3*0.,1./
662 These are only allowed in declarations without double colons
663 (@code{::}), as these were introduced in Fortran 90 which also
664 introduced a new syntax for variable initializations. The syntax for
665 the individual initializers is as for the @code{DATA} statement, but
666 unlike in a @code{DATA} statement, an initializer only applies to the
667 variable immediately preceding. In other words, something like
668 @code{INTEGER I,J/2,3/} is not valid.
670 Examples of standard conforming code equivalent to the above example, are:
673 INTEGER(4) :: i = 1, j = 2
674 REAL(8) :: x(2,2) = RESHAPE((/0.,0.,0.,1./),SHAPE(x))
677 DOUBLE PRECISION x(2,2)
678 DATA i,j,x /1,2,3*0.,1./
681 @node Extensions to namelist
682 @section Extensions to namelist
685 @command{gfortran} fully supports the Fortran 95 standard for namelist I/O
686 including array qualifiers, substrings and fully qualified derived types.
687 The output from a namelist write is compatible with namelist read. The
688 output has all names in upper case and indentation to column 1 after the
689 namelist name. Two extensions are permitted:
691 Old-style use of $ instead of &
694 X(:)%Y(2) = 1.0 2.0 3.0
699 It should be noticed that the default terminator is / rather than &END.
701 Querying of the namelist when inputting from stdin. After at least
702 one space, entering ? sends to stdout the namelist name and the names of
703 the variables in the namelist:
714 Entering =? outputs the namelist to stdout, as if WRITE (*,NML = mynml)
720 X(1)%Y= 0.000000 , 1.000000 , 0.000000 ,
721 X(2)%Y= 0.000000 , 2.000000 , 0.000000 ,
722 X(3)%Y= 0.000000 , 3.000000 , 0.000000 ,
726 To aid this dialog, when input is from stdin, errors send their
727 messages to stderr and execution continues, even if IOSTAT is set.
729 PRINT namelist is permitted. This causes an error if -std=f95 is used.
732 REAL, dimension (4) :: x = (/1.0, 2.0, 3.0, 4.0/)
735 END PROGRAM test_print
738 @node X format descriptor
739 @section X format descriptor
740 @cindex X format descriptor
742 To support legacy codes, @command{gfortran} permits the count field
743 of the X edit descriptor in FORMAT statements to be omitted. When
744 omitted, the count is implicitly assumed to be one.
748 10 FORMAT (I1, X, I1)
751 @node Commas in FORMAT specifications
752 @section Commas in FORMAT specifications
753 @cindex Commas in FORMAT specifications
755 To support legacy codes, @command{gfortran} allows the comma separator
756 to be omitted immediately before and after character string edit
757 descriptors in FORMAT statements.
761 10 FORMAT ('FOO='I1' BAR='I2)
765 @section I/O item lists
766 @cindex I/O item lists
768 To support legacy codes, @command{gfortran} allows the input item list
769 of the READ statement, and the output item lists of the WRITE and PRINT
770 statements to start with a comma.
772 @node Hexadecimal constants
773 @section Hexadecimal constants
774 @cindex Hexadecimal constants
776 As a GNU extension, @command{gfortran} allows hexadecimal constants to
777 be specified using the X prefix, in addition to the standard Z prefix.
779 @node Real array indices
780 @section Real array indices
781 @cindex Real array indices
783 As a GNU extension, @command{gfortran} allows arrays to be indexed using
784 real types, whose values are implicitly converted to integers.
786 @node Unary operators
787 @section Unary operators
788 @cindex Unary operators
790 As a GNU extension, @command{gfortran} allows unary plus and unary
791 minus operators to appear as the second operand of binary arithmetic
792 operators without the need for parenthesis.
798 @node Implicitly interconvert LOGICAL and INTEGER
799 @section Implicitly interconvert LOGICAL and INTEGER
800 @cindex Implicitly interconvert LOGICAL and INTEGER
802 As a GNU extension for backwards compatibility with other compilers,
803 @command{gfortran} allows the implicit conversion of LOGICALs to INTEGERs
804 and vice versa. When converting from a LOGICAL to an INTEGER, the numeric
805 value of @code{.FALSE.} is zero, and that of @code{.TRUE.} is one. When
806 converting from INTEGER to LOGICAL, the value zero is interpreted as
807 @code{.FALSE.} and any nonzero value is interpreted as @code{.TRUE.}.
814 @node Hollerith constants support
815 @section Hollerith constants support
816 @cindex Hollerith constants
818 A Hollerith constant is a string of characters preceded by the letter @samp{H}
819 or @samp{h}, and there must be an literal, unsigned, nonzero default integer
820 constant indicating the number of characters in the string. Hollerith constants
821 are stored as byte strings, one character per byte.
823 @command{gfortran} supports Hollerith constants. They can be used as the right
824 hands in the @code{DATA} statement and @code{ASSIGN} statement, also as the
825 arguments. The left hands can be of Integer, Real, Complex and Logical type.
826 The constant will be padded or trancated to fit the size of left hand.
828 Valid Hollerith constants examples:
831 data x /16Habcdefghijklmnop, 16Hqrstuvwxyz012345/
833 x(1) = 16Habcdefghijklmnop
836 Invalid Hollerith constants examples:
839 a = 8H12345678 ! The Hollerith constant is too long. It will be truncated.
840 a = 0H ! At least one character needed.
844 @section Cray pointers
845 @cindex Cray pointers
847 Cray pointers are part of a non-standard extension that provides a
848 C-like pointer in Fortran. This is accomplished through a pair of
849 variables: an integer "pointer" that holds a memory address, and a
850 "pointee" that is used to dereference the pointer.
852 Pointer/pointee pairs are declared in statements of the form:
854 pointer ( <pointer> , <pointee> )
858 pointer ( <pointer1> , <pointee1> ), ( <pointer2> , <pointee2> ), ...
860 The pointer is an integer that is intended to hold a memory address.
861 The pointee may be an array or scalar. A pointee can be an assumed
862 size array -- that is, the last dimension may be left unspecified by
863 using a '*' in place of a value -- but a pointee cannot be an assumed
864 shape array. No space is allocated for the pointee.
866 The pointee may have its type declared before or after the pointer
867 statement, and its array specification (if any) may be declared
868 before, during, or after the pointer statement. The pointer may be
869 declared as an integer prior to the pointer statement. However, some
870 machines have default integer sizes that are different than the size
871 of a pointer, and so the following code is not portable:
876 If a pointer is declared with a kind that is too small, the compiler
877 will issue a warning; the resulting binary will probably not work
878 correctly, because the memory addresses stored in the pointers may be
879 truncated. It is safer to omit the first line of the above example;
880 if explicit declaration of ipt's type is omitted, then the compiler
881 will ensure that ipt is an integer variable large enough to hold a
884 Pointer arithmetic is valid with Cray pointers, but it is not the same
885 as C pointer arithmetic. Cray pointers are just ordinary integers, so
886 the user is responsible for determining how many bytes to add to a
887 pointer in order to increment it. Consider the following example:
891 pointer (ipt, pointee)
895 The last statement does not set ipt to the address of
896 @code{target(1)}, as one familiar with C pointer arithmetic might
897 expect. Adding 1 to ipt just adds one byte to the address stored in
900 Any expression involving the pointee will be translated to use the
901 value stored in the pointer as the base address.
903 To get the address of elements, this extension provides an intrinsic
904 function loc(), loc() is essentially the C '&' operator, except the
905 address is cast to an integer type:
908 pointer(ipt, arpte(10))
910 ipt = loc(ar) ! Makes arpte is an alias for ar
911 arpte(1) = 1.0 ! Sets ar(1) to 1.0
913 The pointer can also be set by a call to a malloc-type
914 function. There is no malloc intrinsic implemented as part of the
915 Cray pointer extension, but it might be a useful future addition to
916 @command{gfortran}. Even without an intrinsic malloc function,
917 dynamic memory allocation can be combined with Cray pointers by
918 calling a short C function:
922 void mymalloc_(void **ptr, int *nbytes)
924 *ptr = malloc(*nbytes);
933 pointer (ipdata, data(1024))
934 call mymalloc(ipdata,4*1024)
937 Cray pointees often are used to alias an existing variable. For
945 As long as ipt remains unchanged, iarr is now an alias for target.
946 The optimizer, however, will not detect this aliasing, so it is unsafe
947 to use iarr and target simultaneously. Using a pointee in any way
948 that violates the Fortran aliasing rules or assumptions is illegal.
949 It is the user's responsibility to avoid doing this; the compiler
950 works under the assumption that no such aliasing occurs.
952 Cray pointers will work correctly when there is no aliasing (i.e.,
953 when they're used to access a dynamically allocated block of memory),
954 and also in any routine where a pointee is used, but any variable with
955 which it shares storage is not used. Code that violates these rules
956 may not run as the user intends. This is not a bug in the optimizer;
957 any code that violates the aliasing rules is illegal. (Note that this
958 is not unique to gfortran; any Fortran compiler that supports Cray
959 pointers will ``incorrectly'' optimize code with illegal aliasing.)
961 There are a number of restrictions on the attributes that can be
962 applied to Cray pointers and pointees. Pointees may not have the
963 attributes ALLOCATABLE, INTENT, OPTIONAL, DUMMY, TARGET, EXTERNAL,
964 INTRINSIC, or POINTER. Pointers may not have the attributes
965 DIMENSION, POINTER, TARGET, ALLOCATABLE, EXTERNAL, or INTRINSIC.
966 Pointees may not occur in more than one pointer statement. A pointee
967 cannot be a pointer. Pointees cannot occur in equivalence, common, or
970 A pointer may be modified during the course of a program, and this
971 will change the location to which the pointee refers. However, when
972 pointees are passed as arguments, they are treated as ordinary
973 variables in the invoked function. Subsequent changes to the pointer
974 will not change the base address of the array that was passed.
976 @include intrinsic.texi
977 @c ---------------------------------------------------------------------
979 @c ---------------------------------------------------------------------
982 @chapter Contributing
985 Free software is only possible if people contribute to efforts
987 We're always in need of more people helping out with ideas
988 and comments, writing documentation and contributing code.
990 If you want to contribute to GNU Fortran 95,
991 have a look at the long lists of projects you can take on.
992 Some of these projects are small,
993 some of them are large;
994 some are completely orthogonal to the rest of what is
995 happening on @command{gfortran},
996 but others are ``mainstream'' projects in need of enthusiastic hackers.
997 All of these projects are important!
998 We'll eventually get around to the things here,
999 but they are also things doable by someone who is willing and able.
1008 @section Contributors to GNU Fortran 95
1009 @cindex Contributors
1013 Most of the parser was hand-crafted by @emph{Andy Vaught}, who is
1014 also the initiator of the whole project. Thanks Andy!
1015 Most of the interface with GCC was written by @emph{Paul Brook}.
1017 The following individuals have contributed code and/or
1018 ideas and significant help to the gfortran project
1019 (in no particular order):
1023 @item Katherine Holcomb
1024 @item Tobias Schlüter
1025 @item Steven Bosscher
1028 @item Niels Kristian Bech Jensen
1029 @item Steven Johnson
1034 @item François-Xavier Coudert
1042 The following people have contributed bug reports,
1043 smaller or larger patches,
1044 and much needed feedback and encouragement for the
1045 @command{gfortran} project:
1048 @item Erik Schnetter
1053 Many other individuals have helped debug,
1054 test and improve @command{gfortran} over the past few years,
1055 and we welcome you to do the same!
1056 If you already have done so,
1057 and you would like to see your name listed in the
1058 list above, please contact us.
1066 @item Help build the test suite
1067 Solicit more code for donation to the test suite.
1068 We can keep code private on request.
1070 @item Bug hunting/squishing
1071 Find bugs and write more test cases!
1072 Test cases are especially very welcome,
1073 because it allows us to concentrate on fixing bugs
1074 instead of isolating them.
1076 @item Smaller projects (``bug'' fixes):
1078 @item Allow init exprs to be numbers raised to integer powers.
1079 @item Implement correct rounding.
1080 @item Implement F restrictions on Fortran 95 syntax.
1081 @item See about making Emacs-parsable error messages.
1085 If you wish to work on the runtime libraries,
1086 please contact a project maintainer.
1090 @c ---------------------------------------------------------------------
1092 @c ---------------------------------------------------------------------
1098 The GNU Fortran 95 Compiler aims to be a conforming implementation of
1099 ISO/IEC 1539:1997 (Fortran 95).
1101 In the future it may also support other variants of and extensions to
1102 the Fortran language. These include ANSI Fortran 77, ISO Fortran 90,
1103 ISO Fortran 2003 and OpenMP.
1106 * Fortran 2003 status::
1109 @node Fortran 2003 status
1110 @section Fortran 2003 status
1112 Although @command{gfortran} focuses on implementing the Fortran 95
1113 standard for the time being, a few Fortran 2003 features are currently
1118 Intrinsics @code{command_argument_count}, @code{get_command},
1119 @code{get_command_argument}, and @code{get_environment_variable}.
1122 Array constructors using square brackets. That is, @code{[...]} rather
1123 than @code{(/.../)}.
1126 @code{FLUSH} statement.
1129 @code{IOMSG=} specifier for I/O statements.
1133 @c ---------------------------------------------------------------------
1135 @c ---------------------------------------------------------------------