1 \input texinfo @c -*-texinfo-*-
4 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
6 @c GNAT DOCUMENTATION o
10 @c Copyright (C) 1992-2005 Ada Core Technologies, Inc. o
12 @c GNAT is free software; you can redistribute it and/or modify it under o
13 @c terms of the GNU General Public License as published by the Free Soft- o
14 @c ware Foundation; either version 2, or (at your option) any later ver- o
15 @c sion. GNAT is distributed in the hope that it will be useful, but WITH- o
16 @c OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY o
17 @c or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License o
18 @c for more details. You should have received a copy of the GNU General o
19 @c Public License distributed with GNAT; see file COPYING. If not, write o
20 @c to the Free Software Foundation, 59 Temple Place - Suite 330, Boston, o
21 @c MA 02111-1307, USA. o
23 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
25 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
27 @c GNAT_UGN Style Guide
29 @c 1. Always put a @noindent on the line before the first paragraph
30 @c after any of these commands:
42 @c 2. DO NOT use @example. Use @smallexample instead.
43 @c a) DO NOT use highlighting commands (@b{}, @i{}) inside an @smallexample
44 @c context. These can interfere with the readability of the texi
45 @c source file. Instead, use one of the following annotated
46 @c @smallexample commands, and preprocess the texi file with the
47 @c ada2texi tool (which generates appropriate highlighting):
48 @c @smallexample @c ada
49 @c @smallexample @c adanocomment
50 @c @smallexample @c projectfile
51 @c b) The "@c ada" markup will result in boldface for reserved words
52 @c and italics for comments
53 @c c) The "@c adanocomment" markup will result only in boldface for
54 @c reserved words (comments are left alone)
55 @c d) The "@c projectfile" markup is like "@c ada" except that the set
56 @c of reserved words include the new reserved words for project files
58 @c 3. Each @chapter, @section, @subsection, @subsubsection, etc.
59 @c command must be preceded by two empty lines
61 @c 4. The @item command should be on a line of its own if it is in an
62 @c @itemize or @enumerate command.
64 @c 5. When talking about ALI files use "ALI" (all uppercase), not "Ali"
67 @c 6. DO NOT put trailing spaces at the end of a line. Such spaces will
68 @c cause the document build to fail.
70 @c 7. DO NOT use @cartouche for examples that are longer than around 10 lines.
71 @c This command inhibits page breaks, so long examples in a @cartouche can
72 @c lead to large, ugly patches of empty space on a page.
74 @c NOTE: This file should be submitted to xgnatugn with either the vms flag
75 @c or the unw flag set. The unw flag covers topics for both Unix and
78 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
81 @setfilename gnat_ugn_vms.info
85 @setfilename gnat_ugn_unw.info
93 @set FILE gnat_ugn_unw
98 @set FILE gnat_ugn_vms
101 @settitle @value{EDITION} User's Guide @value{PLATFORM}
102 @dircategory GNU Ada tools
104 * @value{EDITION} User's Guide (@value{FILE}) @value{PLATFORM}
107 @include gcc-common.texi
109 @setchapternewpage odd
114 Copyright @copyright{} 1995-2005, Free Software Foundation
116 Permission is granted to copy, distribute and/or modify this document
117 under the terms of the GNU Free Documentation License, Version 1.2
118 or any later version published by the Free Software Foundation;
119 with the Invariant Sections being ``GNU Free Documentation License'', with the
120 Front-Cover Texts being
121 ``@value{EDITION} User's Guide'',
122 and with no Back-Cover Texts.
123 A copy of the license is included in the section entitled
124 ``GNU Free Documentation License''.
129 @title @value{EDITION} User's Guide
134 @titlefont{@i{@value{PLATFORM}}}
140 @subtitle GNAT, The GNU Ada 95 Compiler
141 @subtitle GCC version @value{version-GCC}
146 @vskip 0pt plus 1filll
153 @node Top, About This Guide, (dir), (dir)
154 @top @value{EDITION} User's Guide
157 @value{EDITION} User's Guide @value{PLATFORM}
160 GNAT, The GNU Ada 95 Compiler@*
161 GCC version @value{version-GCC}@*
168 * Getting Started with GNAT::
169 * The GNAT Compilation Model::
170 * Compiling Using gcc::
171 * Binding Using gnatbind::
172 * Linking Using gnatlink::
173 * The GNAT Make Program gnatmake::
174 * Improving Performance::
175 * Renaming Files Using gnatchop::
176 * Configuration Pragmas::
177 * Handling Arbitrary File Naming Conventions Using gnatname::
178 * GNAT Project Manager::
179 * The Cross-Referencing Tools gnatxref and gnatfind::
180 * The GNAT Pretty-Printer gnatpp::
181 * The GNAT Metric Tool gnatmetric::
182 * File Name Krunching Using gnatkr::
183 * Preprocessing Using gnatprep::
185 * The GNAT Run-Time Library Builder gnatlbr::
187 * The GNAT Library Browser gnatls::
188 * Cleaning Up Using gnatclean::
190 * GNAT and Libraries::
191 * Using the GNU make Utility::
193 * Memory Management Issues::
194 * Creating Sample Bodies Using gnatstub::
195 * Other Utility Programs::
196 * Running and Debugging Ada Programs::
198 * Compatibility with DEC Ada::
200 * Platform-Specific Information for the Run-Time Libraries::
201 * Example of Binder Output File::
202 * Elaboration Order Handling in GNAT::
204 * Compatibility and Porting Guide::
206 * Microsoft Windows Topics::
208 * GNU Free Documentation License::
211 --- The Detailed Node Listing ---
215 * What This Guide Contains::
216 * What You Should Know before Reading This Guide::
217 * Related Information::
220 Getting Started with GNAT
223 * Running a Simple Ada Program::
224 * Running a Program with Multiple Units::
225 * Using the gnatmake Utility::
227 * Editing with Emacs::
230 * Introduction to GPS::
231 * Introduction to Glide and GVD::
234 The GNAT Compilation Model
236 * Source Representation::
237 * Foreign Language Representation::
238 * File Naming Rules::
239 * Using Other File Names::
240 * Alternative File Naming Schemes::
241 * Generating Object Files::
242 * Source Dependencies::
243 * The Ada Library Information Files::
244 * Binding an Ada Program::
245 * Mixed Language Programming::
246 * Building Mixed Ada & C++ Programs::
247 * Comparison between GNAT and C/C++ Compilation Models::
248 * Comparison between GNAT and Conventional Ada Library Models::
250 * Placement of temporary files::
253 Foreign Language Representation
256 * Other 8-Bit Codes::
257 * Wide Character Encodings::
259 Compiling Ada Programs With gcc
261 * Compiling Programs::
263 * Search Paths and the Run-Time Library (RTL)::
264 * Order of Compilation Issues::
269 * Output and Error Message Control::
270 * Warning Message Control::
271 * Debugging and Assertion Control::
272 * Validity Checking::
275 * Stack Overflow Checking::
276 * Using gcc for Syntax Checking::
277 * Using gcc for Semantic Checking::
278 * Compiling Different Versions of Ada::
279 * Character Set Control::
280 * File Naming Control::
281 * Subprogram Inlining Control::
282 * Auxiliary Output Control::
283 * Debugging Control::
284 * Exception Handling Control::
285 * Units to Sources Mapping Files::
286 * Integrated Preprocessing::
291 Binding Ada Programs With gnatbind
294 * Switches for gnatbind::
295 * Command-Line Access::
296 * Search Paths for gnatbind::
297 * Examples of gnatbind Usage::
299 Switches for gnatbind
301 * Consistency-Checking Modes::
302 * Binder Error Message Control::
303 * Elaboration Control::
305 * Binding with Non-Ada Main Programs::
306 * Binding Programs with No Main Subprogram::
308 Linking Using gnatlink
311 * Switches for gnatlink::
312 * Setting Stack Size from gnatlink::
313 * Setting Heap Size from gnatlink::
315 The GNAT Make Program gnatmake
318 * Switches for gnatmake::
319 * Mode Switches for gnatmake::
320 * Notes on the Command Line::
321 * How gnatmake Works::
322 * Examples of gnatmake Usage::
324 Improving Performance
325 * Performance Considerations::
326 * Reducing the Size of Ada Executables with gnatelim::
328 Performance Considerations
329 * Controlling Run-Time Checks::
330 * Use of Restrictions::
331 * Optimization Levels::
332 * Debugging Optimized Code::
333 * Inlining of Subprograms::
334 * Optimization and Strict Aliasing::
336 * Coverage Analysis::
339 Reducing the Size of Ada Executables with gnatelim
342 * Correcting the List of Eliminate Pragmas::
343 * Making Your Executables Smaller::
344 * Summary of the gnatelim Usage Cycle::
346 Renaming Files Using gnatchop
348 * Handling Files with Multiple Units::
349 * Operating gnatchop in Compilation Mode::
350 * Command Line for gnatchop::
351 * Switches for gnatchop::
352 * Examples of gnatchop Usage::
354 Configuration Pragmas
356 * Handling of Configuration Pragmas::
357 * The Configuration Pragmas Files::
359 Handling Arbitrary File Naming Conventions Using gnatname
361 * Arbitrary File Naming Conventions::
363 * Switches for gnatname::
364 * Examples of gnatname Usage::
369 * Examples of Project Files::
370 * Project File Syntax::
371 * Objects and Sources in Project Files::
372 * Importing Projects::
373 * Project Extension::
374 * Project Hierarchy Extension::
375 * External References in Project Files::
376 * Packages in Project Files::
377 * Variables from Imported Projects::
380 * Stand-alone Library Projects::
381 * Switches Related to Project Files::
382 * Tools Supporting Project Files::
383 * An Extended Example::
384 * Project File Complete Syntax::
386 The Cross-Referencing Tools gnatxref and gnatfind
388 * gnatxref Switches::
389 * gnatfind Switches::
390 * Project Files for gnatxref and gnatfind::
391 * Regular Expressions in gnatfind and gnatxref::
392 * Examples of gnatxref Usage::
393 * Examples of gnatfind Usage::
395 The GNAT Pretty-Printer gnatpp
397 * Switches for gnatpp::
400 The GNAT Metrics Tool gnatmetric
402 * Switches for gnatmetric::
404 File Name Krunching Using gnatkr
409 * Examples of gnatkr Usage::
411 Preprocessing Using gnatprep
414 * Switches for gnatprep::
415 * Form of Definitions File::
416 * Form of Input Text for gnatprep::
419 The GNAT Run-Time Library Builder gnatlbr
422 * Switches for gnatlbr::
423 * Examples of gnatlbr Usage::
426 The GNAT Library Browser gnatls
429 * Switches for gnatls::
430 * Examples of gnatls Usage::
432 Cleaning Up Using gnatclean
434 * Running gnatclean::
435 * Switches for gnatclean::
436 @c * Examples of gnatclean Usage::
442 * Introduction to Libraries in GNAT::
443 * General Ada Libraries::
444 * Stand-alone Ada Libraries::
445 * Rebuilding the GNAT Run-Time Library::
447 Using the GNU make Utility
449 * Using gnatmake in a Makefile::
450 * Automatically Creating a List of Directories::
451 * Generating the Command Line Switches::
452 * Overcoming Command Line Length Limits::
455 Memory Management Issues
457 * Some Useful Memory Pools::
458 * The GNAT Debug Pool Facility::
463 Some Useful Memory Pools
465 The GNAT Debug Pool Facility
471 * Switches for gnatmem::
472 * Example of gnatmem Usage::
475 Sample Bodies Using gnatstub
478 * Switches for gnatstub::
480 Other Utility Programs
482 * Using Other Utility Programs with GNAT::
483 * The External Symbol Naming Scheme of GNAT::
485 * Ada Mode for Glide::
487 * Converting Ada Files to html with gnathtml::
489 Running and Debugging Ada Programs
491 * The GNAT Debugger GDB::
493 * Introduction to GDB Commands::
494 * Using Ada Expressions::
495 * Calling User-Defined Subprograms::
496 * Using the Next Command in a Function::
499 * Debugging Generic Units::
500 * GNAT Abnormal Termination or Failure to Terminate::
501 * Naming Conventions for GNAT Source Files::
502 * Getting Internal Debugging Information::
510 Compatibility with DEC Ada
512 * Ada 95 Compatibility::
513 * Differences in the Definition of Package System::
514 * Language-Related Features::
515 * The Package STANDARD::
516 * The Package SYSTEM::
517 * Tasking and Task-Related Features::
518 * Implementation of Tasks in DEC Ada for OpenVMS Alpha Systems::
519 * Pragmas and Pragma-Related Features::
520 * Library of Predefined Units::
522 * Main Program Definition::
523 * Implementation-Defined Attributes::
524 * Compiler and Run-Time Interfacing::
525 * Program Compilation and Library Management::
527 * Implementation Limits::
530 Language-Related Features
532 * Integer Types and Representations::
533 * Floating-Point Types and Representations::
534 * Pragmas Float_Representation and Long_Float::
535 * Fixed-Point Types and Representations::
536 * Record and Array Component Alignment::
538 * Other Representation Clauses::
540 Implementation of Tasks in DEC Ada for OpenVMS Alpha Systems
542 * Assigning Task IDs::
543 * Task IDs and Delays::
544 * Task-Related Pragmas::
545 * Scheduling and Task Priority::
547 * External Interrupts::
549 Pragmas and Pragma-Related Features
551 * Restrictions on the Pragma INLINE::
552 * Restrictions on the Pragma INTERFACE::
553 * Restrictions on the Pragma SYSTEM_NAME::
555 Library of Predefined Units
557 * Changes to DECLIB::
561 * Shared Libraries and Options Files::
565 Platform-Specific Information for the Run-Time Libraries
567 * Summary of Run-Time Configurations::
568 * Specifying a Run-Time Library::
569 * Choosing the Scheduling Policy::
570 * Solaris-Specific Considerations::
571 * IRIX-Specific Considerations::
572 * Linux-Specific Considerations::
573 * AIX-Specific Considerations::
575 Example of Binder Output File
577 Elaboration Order Handling in GNAT
579 * Elaboration Code in Ada 95::
580 * Checking the Elaboration Order in Ada 95::
581 * Controlling the Elaboration Order in Ada 95::
582 * Controlling Elaboration in GNAT - Internal Calls::
583 * Controlling Elaboration in GNAT - External Calls::
584 * Default Behavior in GNAT - Ensuring Safety::
585 * Treatment of Pragma Elaborate::
586 * Elaboration Issues for Library Tasks::
587 * Mixing Elaboration Models::
588 * What to Do If the Default Elaboration Behavior Fails::
589 * Elaboration for Access-to-Subprogram Values::
590 * Summary of Procedures for Elaboration Control::
591 * Other Elaboration Order Considerations::
595 * Basic Assembler Syntax::
596 * A Simple Example of Inline Assembler::
597 * Output Variables in Inline Assembler::
598 * Input Variables in Inline Assembler::
599 * Inlining Inline Assembler Code::
600 * Other Asm Functionality::
602 Compatibility and Porting Guide
604 * Compatibility with Ada 83::
605 * Implementation-dependent characteristics::
606 * Compatibility with DEC Ada 83::
607 * Compatibility with Other Ada 95 Systems::
608 * Representation Clauses::
610 * Transitioning from Alpha to Integrity OpenVMS::
614 Microsoft Windows Topics
616 * Using GNAT on Windows::
617 * CONSOLE and WINDOWS subsystems::
619 * Mixed-Language Programming on Windows::
620 * Windows Calling Conventions::
621 * Introduction to Dynamic Link Libraries (DLLs)::
622 * Using DLLs with GNAT::
623 * Building DLLs with GNAT::
624 * GNAT and Windows Resources::
626 * GNAT and COM/DCOM Objects::
633 @node About This Guide
634 @unnumbered About This Guide
638 This guide describes the use of @value{EDITION},
639 a full language compiler for the Ada
640 95 programming language, implemented on HP's Alpha and
641 Integrity (ia64) OpenVMS platforms.
644 This guide describes the use of @value{EDITION},
645 a compiler and software development
646 toolset for the full Ada 95 programming language.
648 It describes the features of the compiler and tools, and details
649 how to use them to build Ada 95 applications.
652 For ease of exposition, ``GNAT Pro'' will be referred to simply as
653 ``GNAT'' in the remainder of this document.
657 * What This Guide Contains::
658 * What You Should Know before Reading This Guide::
659 * Related Information::
663 @node What This Guide Contains
664 @unnumberedsec What This Guide Contains
667 This guide contains the following chapters:
671 @ref{Getting Started with GNAT}, describes how to get started compiling
672 and running Ada programs with the GNAT Ada programming environment.
674 @ref{The GNAT Compilation Model}, describes the compilation model used
678 @ref{Compiling Using gcc}, describes how to compile
679 Ada programs with @command{gcc}, the Ada compiler.
682 @ref{Binding Using gnatbind}, describes how to
683 perform binding of Ada programs with @code{gnatbind}, the GNAT binding
687 @ref{Linking Using gnatlink},
688 describes @command{gnatlink}, a
689 program that provides for linking using the GNAT run-time library to
690 construct a program. @command{gnatlink} can also incorporate foreign language
691 object units into the executable.
694 @ref{The GNAT Make Program gnatmake}, describes @command{gnatmake}, a
695 utility that automatically determines the set of sources
696 needed by an Ada compilation unit, and executes the necessary compilations
700 @ref{Improving Performance}, shows various techniques for making your
701 Ada program run faster or take less space.
702 It discusses the effect of the compiler's optimization switch and
703 also describes the @command{gnatelim} tool.
706 @ref{Renaming Files Using gnatchop}, describes
707 @code{gnatchop}, a utility that allows you to preprocess a file that
708 contains Ada source code, and split it into one or more new files, one
709 for each compilation unit.
712 @ref{Configuration Pragmas}, describes the configuration pragmas
716 @ref{Handling Arbitrary File Naming Conventions Using gnatname},
717 shows how to override the default GNAT file naming conventions,
718 either for an individual unit or globally.
721 @ref{GNAT Project Manager}, describes how to use project files
722 to organize large projects.
725 @ref{The Cross-Referencing Tools gnatxref and gnatfind}, discusses
726 @code{gnatxref} and @code{gnatfind}, two tools that provide an easy
727 way to navigate through sources.
730 @ref{The GNAT Pretty-Printer gnatpp}, shows how to produce a reformatted
731 version of an Ada source file with control over casing, indentation,
732 comment placement, and other elements of program presentation style.
735 @ref{The GNAT Metric Tool gnatmetric}, shows how to compute various
736 metrics for an Ada source file, such as the number of types and subprograms,
737 and assorted complexity measures.
740 @ref{File Name Krunching Using gnatkr}, describes the @code{gnatkr}
741 file name krunching utility, used to handle shortened
742 file names on operating systems with a limit on the length of names.
745 @ref{Preprocessing Using gnatprep}, describes @code{gnatprep}, a
746 preprocessor utility that allows a single source file to be used to
747 generate multiple or parameterized source files, by means of macro
752 @ref{The GNAT Run-Time Library Builder gnatlbr}, describes @command{gnatlbr},
753 a tool for rebuilding the GNAT run time with user-supplied
754 configuration pragmas.
758 @ref{The GNAT Library Browser gnatls}, describes @code{gnatls}, a
759 utility that displays information about compiled units, including dependences
760 on the corresponding sources files, and consistency of compilations.
763 @ref{Cleaning Up Using gnatclean}, describes @code{gnatclean}, a utility
764 to delete files that are produced by the compiler, binder and linker.
768 @ref{GNAT and Libraries}, describes the process of creating and using
769 Libraries with GNAT. It also describes how to recompile the GNAT run-time
773 @ref{Using the GNU make Utility}, describes some techniques for using
774 the GNAT toolset in Makefiles.
778 @ref{Memory Management Issues}, describes some useful predefined storage pools
779 and in particular the GNAT Debug Pool facility, which helps detect incorrect
782 It also describes @command{gnatmem}, a utility that monitors dynamic
783 allocation and deallocation and helps detect ``memory leaks''.
787 @ref{Creating Sample Bodies Using gnatstub}, discusses @code{gnatstub},
788 a utility that generates empty but compilable bodies for library units.
791 @ref{Other Utility Programs}, discusses several other GNAT utilities,
792 including @code{gnathtml}.
795 @ref{Running and Debugging Ada Programs}, describes how to run and debug
800 @ref{Compatibility with DEC Ada}, details the compatibility of GNAT with
801 DEC Ada 83 @footnote{``DEC Ada'' refers to the legacy product originally
802 developed by Digital Equipment Corporation and currently supported by HP.}
807 @ref{Platform-Specific Information for the Run-Time Libraries},
808 describes the various run-time
809 libraries supported by GNAT on various platforms and explains how to
810 choose a particular library.
813 @ref{Example of Binder Output File}, shows the source code for the binder
814 output file for a sample program.
817 @ref{Elaboration Order Handling in GNAT}, describes how GNAT helps
818 you deal with elaboration order issues.
821 @ref{Inline Assembler}, shows how to use the inline assembly facility
825 @ref{Compatibility and Porting Guide}, includes sections on compatibility
826 of GNAT with other Ada 83 and Ada 95 compilation systems, to assist
827 in porting code from other environments.
831 @ref{Microsoft Windows Topics}, presents information relevant to the
832 Microsoft Windows platform.
836 @c *************************************************
837 @node What You Should Know before Reading This Guide
838 @c *************************************************
839 @unnumberedsec What You Should Know before Reading This Guide
841 @cindex Ada 95 Language Reference Manual
843 This user's guide assumes that you are familiar with Ada 95 language, as
844 described in the International Standard ANSI/ISO/IEC-8652:1995, January
847 @node Related Information
848 @unnumberedsec Related Information
851 For further information about related tools, refer to the following
856 @cite{GNAT Reference Manual}, which contains all reference
857 material for the GNAT implementation of Ada 95.
861 @cite{Using the GNAT Programming System}, which describes the GPS
862 integrated development environment.
865 @cite{GNAT Programming System Tutorial}, which introduces the
866 main GPS features through examples.
870 @cite{Ada 95 Language Reference Manual}, which contains all reference
871 material for the Ada 95 programming language.
874 @cite{Debugging with GDB}
876 , located in the GNU:[DOCS] directory,
878 contains all details on the use of the GNU source-level debugger.
881 @cite{GNU Emacs Manual}
883 , located in the GNU:[DOCS] directory if the EMACS kit is installed,
885 contains full information on the extensible editor and programming
892 @unnumberedsec Conventions
894 @cindex Typographical conventions
897 Following are examples of the typographical and graphic conventions used
902 @code{Functions}, @code{utility program names}, @code{standard names},
909 @file{File Names}, @file{button names}, and @file{field names}.
918 [optional information or parameters]
921 Examples are described by text
923 and then shown this way.
928 Commands that are entered by the user are preceded in this manual by the
929 characters @w{``@code{$ }''} (dollar sign followed by space). If your system
930 uses this sequence as a prompt, then the commands will appear exactly as
931 you see them in the manual. If your system uses some other prompt, then
932 the command will appear with the @code{$} replaced by whatever prompt
933 character you are using.
936 Full file names are shown with the ``@code{/}'' character
937 as the directory separator; e.g., @file{parent-dir/subdir/myfile.adb}.
938 If you are using GNAT on a Windows platform, please note that
939 the ``@code{\}'' character should be used instead.
942 @c ****************************
943 @node Getting Started with GNAT
944 @chapter Getting Started with GNAT
947 This chapter describes some simple ways of using GNAT to build
948 executable Ada programs.
950 @ref{Running GNAT}, through @ref{Using the gnatmake Utility},
951 show how to use the command line environment.
952 @ref{Introduction to Glide and GVD}, provides a brief
953 introduction to the visually-oriented IDE for GNAT.
954 Supplementing Glide on some platforms is GPS, the
955 GNAT Programming System, which offers a richer graphical
956 ``look and feel'', enhanced configurability, support for
957 development in other programming language, comprehensive
958 browsing features, and many other capabilities.
959 For information on GPS please refer to
960 @cite{Using the GNAT Programming System}.
965 * Running a Simple Ada Program::
966 * Running a Program with Multiple Units::
967 * Using the gnatmake Utility::
969 * Editing with Emacs::
972 * Introduction to GPS::
973 * Introduction to Glide and GVD::
978 @section Running GNAT
981 Three steps are needed to create an executable file from an Ada source
986 The source file(s) must be compiled.
988 The file(s) must be bound using the GNAT binder.
990 All appropriate object files must be linked to produce an executable.
994 All three steps are most commonly handled by using the @command{gnatmake}
995 utility program that, given the name of the main program, automatically
996 performs the necessary compilation, binding and linking steps.
998 @node Running a Simple Ada Program
999 @section Running a Simple Ada Program
1002 Any text editor may be used to prepare an Ada program.
1005 used, the optional Ada mode may be helpful in laying out the program.
1008 program text is a normal text file. We will suppose in our initial
1009 example that you have used your editor to prepare the following
1010 standard format text file:
1012 @smallexample @c ada
1014 with Ada.Text_IO; use Ada.Text_IO;
1017 Put_Line ("Hello WORLD!");
1023 This file should be named @file{hello.adb}.
1024 With the normal default file naming conventions, GNAT requires
1026 contain a single compilation unit whose file name is the
1028 with periods replaced by hyphens; the
1029 extension is @file{ads} for a
1030 spec and @file{adb} for a body.
1031 You can override this default file naming convention by use of the
1032 special pragma @code{Source_File_Name} (@pxref{Using Other File Names}).
1033 Alternatively, if you want to rename your files according to this default
1034 convention, which is probably more convenient if you will be using GNAT
1035 for all your compilations, then the @code{gnatchop} utility
1036 can be used to generate correctly-named source files
1037 (@pxref{Renaming Files Using gnatchop}).
1039 You can compile the program using the following command (@code{$} is used
1040 as the command prompt in the examples in this document):
1047 @command{gcc} is the command used to run the compiler. This compiler is
1048 capable of compiling programs in several languages, including Ada 95 and
1049 C. It assumes that you have given it an Ada program if the file extension is
1050 either @file{.ads} or @file{.adb}, and it will then call
1051 the GNAT compiler to compile the specified file.
1054 The @option{-c} switch is required. It tells @command{gcc} to only do a
1055 compilation. (For C programs, @command{gcc} can also do linking, but this
1056 capability is not used directly for Ada programs, so the @option{-c}
1057 switch must always be present.)
1060 This compile command generates a file
1061 @file{hello.o}, which is the object
1062 file corresponding to your Ada program. It also generates
1063 an ``Ada Library Information'' file @file{hello.ali},
1064 which contains additional information used to check
1065 that an Ada program is consistent.
1066 To build an executable file,
1067 use @code{gnatbind} to bind the program
1068 and @command{gnatlink} to link it. The
1069 argument to both @code{gnatbind} and @command{gnatlink} is the name of the
1070 @file{ALI} file, but the default extension of @file{.ali} can
1071 be omitted. This means that in the most common case, the argument
1072 is simply the name of the main program:
1080 A simpler method of carrying out these steps is to use
1082 a master program that invokes all the required
1083 compilation, binding and linking tools in the correct order. In particular,
1084 @command{gnatmake} automatically recompiles any sources that have been
1085 modified since they were last compiled, or sources that depend
1086 on such modified sources, so that ``version skew'' is avoided.
1087 @cindex Version skew (avoided by @command{gnatmake})
1090 $ gnatmake hello.adb
1094 The result is an executable program called @file{hello}, which can be
1102 assuming that the current directory is on the search path
1103 for executable programs.
1106 and, if all has gone well, you will see
1113 appear in response to this command.
1115 @c ****************************************
1116 @node Running a Program with Multiple Units
1117 @section Running a Program with Multiple Units
1120 Consider a slightly more complicated example that has three files: a
1121 main program, and the spec and body of a package:
1123 @smallexample @c ada
1126 package Greetings is
1131 with Ada.Text_IO; use Ada.Text_IO;
1132 package body Greetings is
1135 Put_Line ("Hello WORLD!");
1138 procedure Goodbye is
1140 Put_Line ("Goodbye WORLD!");
1157 Following the one-unit-per-file rule, place this program in the
1158 following three separate files:
1162 spec of package @code{Greetings}
1165 body of package @code{Greetings}
1168 body of main program
1172 To build an executable version of
1173 this program, we could use four separate steps to compile, bind, and link
1174 the program, as follows:
1178 $ gcc -c greetings.adb
1184 Note that there is no required order of compilation when using GNAT.
1185 In particular it is perfectly fine to compile the main program first.
1186 Also, it is not necessary to compile package specs in the case where
1187 there is an accompanying body; you only need to compile the body. If you want
1188 to submit these files to the compiler for semantic checking and not code
1189 generation, then use the
1190 @option{-gnatc} switch:
1193 $ gcc -c greetings.ads -gnatc
1197 Although the compilation can be done in separate steps as in the
1198 above example, in practice it is almost always more convenient
1199 to use the @command{gnatmake} tool. All you need to know in this case
1200 is the name of the main program's source file. The effect of the above four
1201 commands can be achieved with a single one:
1204 $ gnatmake gmain.adb
1208 In the next section we discuss the advantages of using @command{gnatmake} in
1211 @c *****************************
1212 @node Using the gnatmake Utility
1213 @section Using the @command{gnatmake} Utility
1216 If you work on a program by compiling single components at a time using
1217 @command{gcc}, you typically keep track of the units you modify. In order to
1218 build a consistent system, you compile not only these units, but also any
1219 units that depend on the units you have modified.
1220 For example, in the preceding case,
1221 if you edit @file{gmain.adb}, you only need to recompile that file. But if
1222 you edit @file{greetings.ads}, you must recompile both
1223 @file{greetings.adb} and @file{gmain.adb}, because both files contain
1224 units that depend on @file{greetings.ads}.
1226 @code{gnatbind} will warn you if you forget one of these compilation
1227 steps, so that it is impossible to generate an inconsistent program as a
1228 result of forgetting to do a compilation. Nevertheless it is tedious and
1229 error-prone to keep track of dependencies among units.
1230 One approach to handle the dependency-bookkeeping is to use a
1231 makefile. However, makefiles present maintenance problems of their own:
1232 if the dependencies change as you change the program, you must make
1233 sure that the makefile is kept up-to-date manually, which is also an
1234 error-prone process.
1236 The @command{gnatmake} utility takes care of these details automatically.
1237 Invoke it using either one of the following forms:
1240 $ gnatmake gmain.adb
1241 $ gnatmake ^gmain^GMAIN^
1245 The argument is the name of the file containing the main program;
1246 you may omit the extension. @command{gnatmake}
1247 examines the environment, automatically recompiles any files that need
1248 recompiling, and binds and links the resulting set of object files,
1249 generating the executable file, @file{^gmain^GMAIN.EXE^}.
1250 In a large program, it
1251 can be extremely helpful to use @command{gnatmake}, because working out by hand
1252 what needs to be recompiled can be difficult.
1254 Note that @command{gnatmake}
1255 takes into account all the Ada 95 rules that
1256 establish dependencies among units. These include dependencies that result
1257 from inlining subprogram bodies, and from
1258 generic instantiation. Unlike some other
1259 Ada make tools, @command{gnatmake} does not rely on the dependencies that were
1260 found by the compiler on a previous compilation, which may possibly
1261 be wrong when sources change. @command{gnatmake} determines the exact set of
1262 dependencies from scratch each time it is run.
1265 @node Editing with Emacs
1266 @section Editing with Emacs
1270 Emacs is an extensible self-documenting text editor that is available in a
1271 separate VMSINSTAL kit.
1273 Invoke Emacs by typing @kbd{Emacs} at the command prompt. To get started,
1274 click on the Emacs Help menu and run the Emacs Tutorial.
1275 In a character cell terminal, Emacs help is invoked with @kbd{Ctrl-h} (also
1276 written as @kbd{C-h}), and the tutorial by @kbd{C-h t}.
1278 Documentation on Emacs and other tools is available in Emacs under the
1279 pull-down menu button: @code{Help - Info}. After selecting @code{Info},
1280 use the middle mouse button to select a topic (e.g. Emacs).
1282 In a character cell terminal, do @kbd{C-h i} to invoke info, and then @kbd{m}
1283 (stands for menu) followed by the menu item desired, as in @kbd{m Emacs}, to
1284 get to the Emacs manual.
1285 Help on Emacs is also available by typing @kbd{HELP EMACS} at the DCL command
1288 The tutorial is highly recommended in order to learn the intricacies of Emacs,
1289 which is sufficiently extensible to provide for a complete programming
1290 environment and shell for the sophisticated user.
1294 @node Introduction to GPS
1295 @section Introduction to GPS
1296 @cindex GPS (GNAT Programming System)
1297 @cindex GNAT Programming System (GPS)
1299 Although the command line interface (@command{gnatmake}, etc.) alone
1300 is sufficient, a graphical Interactive Development
1301 Environment can make it easier for you to compose, navigate, and debug
1302 programs. This section describes the main features of GPS
1303 (``GNAT Programming System''), the GNAT graphical IDE.
1304 You will see how to use GPS to build and debug an executable, and
1305 you will also learn some of the basics of the GNAT ``project'' facility.
1307 GPS enables you to do much more than is presented here;
1308 e.g., you can produce a call graph, interface to a third-party
1309 Version Control System, and inspect the generated assembly language
1311 Indeed, GPS also supports languages other than Ada.
1312 Such additional information, and an explanation of all of the GPS menu
1313 items. may be found in the on-line help, which includes
1314 a user's guide and a tutorial (these are also accessible from the GNAT
1318 * Building a New Program with GPS::
1319 * Simple Debugging with GPS::
1322 @node Building a New Program with GPS
1323 @subsection Building a New Program with GPS
1325 GPS invokes the GNAT compilation tools using information
1326 contained in a @emph{project} (also known as a @emph{project file}):
1327 a collection of properties such
1328 as source directories, identities of main subprograms, tool switches, etc.,
1329 and their associated values.
1330 See @ref{GNAT Project Manager} for details.
1331 In order to run GPS, you will need to either create a new project
1332 or else open an existing one.
1334 This section will explain how you can use GPS to create a project,
1335 to associate Ada source files with a project, and to build and run
1339 @item @emph{Creating a project}
1341 Invoke GPS, either from the command line or the platform's IDE.
1342 After it starts, GPS will display a ``Welcome'' screen with three
1347 @code{Start with default project in directory}
1350 @code{Create new project with wizard}
1353 @code{Open existing project}
1357 Select @code{Create new project with wizard} and press @code{OK}.
1358 A new window will appear. In the text box labeled with
1359 @code{Enter the name of the project to create}, type @file{sample}
1360 as the project name.
1361 In the next box, browse to choose the directory in which you
1362 would like to create the project file.
1363 After selecting an appropriate directory, press @code{Forward}.
1365 A window will appear with the title
1366 @code{Version Control System Configuration}.
1367 Simply press @code{Forward}.
1369 A window will appear with the title
1370 @code{Please select the source directories for this project}.
1371 The directory that you specified for the project file will be selected
1372 by default as the one to use for sources; simply press @code{Forward}.
1374 A window will appear with the title
1375 @code{Please select the build directory for this project}.
1376 The directory that you specified for the project file will be selected
1377 by default for object files and executables;
1378 simply press @code{Forward}.
1380 A window will appear with the title
1381 @code{Please select the main units for this project}.
1382 You will supply this information later, after creating the source file.
1383 Simply press @code{Forward} for now.
1385 A window will appear with the title
1386 @code{Please select the switches to build the project}.
1387 Press @code{Apply}. This will create a project file named
1388 @file{sample.prj} in the directory that you had specified.
1390 @item @emph{Creating and saving the source file}
1392 After you create the new project, a GPS window will appear, which is
1393 partitioned into two main sections:
1397 A @emph{Workspace area}, initially greyed out, which you will use for
1398 creating and editing source files
1401 Directly below, a @emph{Messages area}, which initially displays a
1402 ``Welcome'' message.
1403 (If the Messages area is not visible, drag its border upward to expand it.)
1407 Select @code{File} on the menu bar, and then the @code{New} command.
1408 The Workspace area will become white, and you can now
1409 enter the source program explicitly.
1410 Type the following text
1412 @smallexample @c ada
1414 with Ada.Text_IO; use Ada.Text_IO;
1417 Put_Line("Hello from GPS!");
1423 Select @code{File}, then @code{Save As}, and enter the source file name
1425 The file will be saved in the same directory you specified as the
1426 location of the default project file.
1428 @item @emph{Updating the project file}
1430 You need to add the new source file to the project.
1432 the @code{Project} menu and then @code{Edit project properties}.
1433 Click the @code{Main files} tab on the left, and then the
1435 Choose @file{hello.adb} from the list, and press @code{Open}.
1436 The project settings window will reflect this action.
1439 @item @emph{Building and running the program}
1441 In the main GPS window, now choose the @code{Build} menu, then @code{Make},
1442 and select @file{hello.adb}.
1443 The Messages window will display the resulting invocations of @command{gcc},
1444 @command{gnatbind}, and @command{gnatlink}
1445 (reflecting the default switch settings from the
1446 project file that you created) and then a ``successful compilation/build''
1449 To run the program, choose the @code{Build} menu, then @code{Run}, and
1450 select @command{hello}.
1451 An @emph{Arguments Selection} window will appear.
1452 There are no command line arguments, so just click @code{OK}.
1454 The Messages window will now display the program's output (the string
1455 @code{Hello from GPS}), and at the bottom of the GPS window a status
1456 update is displayed (@code{Run: hello}).
1457 Close the GPS window (or select @code{File}, then @code{Exit}) to
1458 terminate this GPS session.
1461 @node Simple Debugging with GPS
1462 @subsection Simple Debugging with GPS
1464 This section illustrates basic debugging techniques (setting breakpoints,
1465 examining/modifying variables, single stepping).
1468 @item @emph{Opening a project}
1470 Start GPS and select @code{Open existing project}; browse to
1471 specify the project file @file{sample.prj} that you had created in the
1474 @item @emph{Creating a source file}
1476 Select @code{File}, then @code{New}, and type in the following program:
1478 @smallexample @c ada
1480 with Ada.Text_IO; use Ada.Text_IO;
1481 procedure Example is
1482 Line : String (1..80);
1485 Put_Line("Type a line of text at each prompt; an empty line to exit");
1489 Put_Line (Line (1..N) );
1497 Select @code{File}, then @code{Save as}, and enter the file name
1500 @item @emph{Updating the project file}
1502 Add @code{Example} as a new main unit for the project:
1505 Select @code{Project}, then @code{Edit Project Properties}.
1508 Select the @code{Main files} tab, click @code{Add}, then
1509 select the file @file{example.adb} from the list, and
1511 You will see the file name appear in the list of main units
1517 @item @emph{Building/running the executable}
1519 To build the executable
1520 select @code{Build}, then @code{Make}, and then choose @file{example.adb}.
1522 Run the program to see its effect (in the Messages area).
1523 Each line that you enter is displayed; an empty line will
1524 cause the loop to exit and the program to terminate.
1526 @item @emph{Debugging the program}
1528 Note that the @option{-g} switches to @command{gcc} and @command{gnatlink},
1529 which are required for debugging, are on by default when you create
1531 Thus unless you intentionally remove these settings, you will be able
1532 to debug any program that you develop using GPS.
1535 @item @emph{Initializing}
1537 Select @code{Debug}, then @code{Initialize}, then @file{example}
1539 @item @emph{Setting a breakpoint}
1541 After performing the initialization step, you will observe a small
1542 icon to the right of each line number.
1543 This serves as a toggle for breakpoints; clicking the icon will
1544 set a breakpoint at the corresponding line (the icon will change to
1545 a red circle with an ``x''), and clicking it again
1546 will remove the breakpoint / reset the icon.
1548 For purposes of this example, set a breakpoint at line 10 (the
1549 statement @code{Put_Line@ (Line@ (1..N));}
1551 @item @emph{Starting program execution}
1553 Select @code{Debug}, then @code{Run}. When the
1554 @code{Program Arguments} window appears, click @code{OK}.
1555 A console window will appear; enter some line of text,
1556 e.g. @code{abcde}, at the prompt.
1557 The program will pause execution when it gets to the
1558 breakpoint, and the corresponding line is highlighted.
1560 @item @emph{Examining a variable}
1562 Move the mouse over one of the occurrences of the variable @code{N}.
1563 You will see the value (5) displayed, in ``tool tip'' fashion.
1564 Right click on @code{N}, select @code{Debug}, then select @code{Display N}.
1565 You will see information about @code{N} appear in the @code{Debugger Data}
1566 pane, showing the value as 5.
1568 @item @emph{Assigning a new value to a variable}
1570 Right click on the @code{N} in the @code{Debugger Data} pane, and
1571 select @code{Set value of N}.
1572 When the input window appears, enter the value @code{4} and click
1574 This value does not automatically appear in the @code{Debugger Data}
1575 pane; to see it, right click again on the @code{N} in the
1576 @code{Debugger Data} pane and select @code{Update value}.
1577 The new value, 4, will appear in red.
1579 @item @emph{Single stepping}
1581 Select @code{Debug}, then @code{Next}.
1582 This will cause the next statement to be executed, in this case the
1583 call of @code{Put_Line} with the string slice.
1584 Notice in the console window that the displayed string is simply
1585 @code{abcd} and not @code{abcde} which you had entered.
1586 This is because the upper bound of the slice is now 4 rather than 5.
1588 @item @emph{Removing a breakpoint}
1590 Toggle the breakpoint icon at line 10.
1592 @item @emph{Resuming execution from a breakpoint}
1594 Select @code{Debug}, then @code{Continue}.
1595 The program will reach the next iteration of the loop, and
1596 wait for input after displaying the prompt.
1597 This time, just hit the @kbd{Enter} key.
1598 The value of @code{N} will be 0, and the program will terminate.
1599 The console window will disappear.
1603 @node Introduction to Glide and GVD
1604 @section Introduction to Glide and GVD
1608 This section describes the main features of Glide,
1609 a GNAT graphical IDE, and also shows how to use the basic commands in GVD,
1610 the GNU Visual Debugger.
1611 These tools may be present in addition to, or in place of, GPS on some
1613 Additional information on Glide and GVD may be found
1614 in the on-line help for these tools.
1617 * Building a New Program with Glide::
1618 * Simple Debugging with GVD::
1619 * Other Glide Features::
1622 @node Building a New Program with Glide
1623 @subsection Building a New Program with Glide
1625 The simplest way to invoke Glide is to enter @command{glide}
1626 at the command prompt. It will generally be useful to issue this
1627 as a background command, thus allowing you to continue using
1628 your command window for other purposes while Glide is running:
1635 Glide will start up with an initial screen displaying the top-level menu items
1636 as well as some other information. The menu selections are as follows
1638 @item @code{Buffers}
1649 For this introductory example, you will need to create a new Ada source file.
1650 First, select the @code{Files} menu. This will pop open a menu with around
1651 a dozen or so items. To create a file, select the @code{Open file...} choice.
1652 Depending on the platform, you may see a pop-up window where you can browse
1653 to an appropriate directory and then enter the file name, or else simply
1654 see a line at the bottom of the Glide window where you can likewise enter
1655 the file name. Note that in Glide, when you attempt to open a non-existent
1656 file, the effect is to create a file with that name. For this example enter
1657 @file{hello.adb} as the name of the file.
1659 A new buffer will now appear, occupying the entire Glide window,
1660 with the file name at the top. The menu selections are slightly different
1661 from the ones you saw on the opening screen; there is an @code{Entities} item,
1662 and in place of @code{Glide} there is now an @code{Ada} item. Glide uses
1663 the file extension to identify the source language, so @file{adb} indicates
1666 You will enter some of the source program lines explicitly,
1667 and use the syntax-oriented template mechanism to enter other lines.
1668 First, type the following text:
1670 with Ada.Text_IO; use Ada.Text_IO;
1676 Observe that Glide uses different colors to distinguish reserved words from
1677 identifiers. Also, after the @code{procedure Hello is} line, the cursor is
1678 automatically indented in anticipation of declarations. When you enter
1679 @code{begin}, Glide recognizes that there are no declarations and thus places
1680 @code{begin} flush left. But after the @code{begin} line the cursor is again
1681 indented, where the statement(s) will be placed.
1683 The main part of the program will be a @code{for} loop. Instead of entering
1684 the text explicitly, however, use a statement template. Select the @code{Ada}
1685 item on the top menu bar, move the mouse to the @code{Statements} item,
1686 and you will see a large selection of alternatives. Choose @code{for loop}.
1687 You will be prompted (at the bottom of the buffer) for a loop name;
1688 simply press the @key{Enter} key since a loop name is not needed.
1689 You should see the beginning of a @code{for} loop appear in the source
1690 program window. You will now be prompted for the name of the loop variable;
1691 enter a line with the identifier @code{ind} (lower case). Note that,
1692 by default, Glide capitalizes the name (you can override such behavior
1693 if you wish, although this is outside the scope of this introduction).
1694 Next, Glide prompts you for the loop range; enter a line containing
1695 @code{1..5} and you will see this also appear in the source program,
1696 together with the remaining elements of the @code{for} loop syntax.
1698 Next enter the statement (with an intentional error, a missing semicolon)
1699 that will form the body of the loop:
1701 Put_Line("Hello, World" & Integer'Image(I))
1705 Finally, type @code{end Hello;} as the last line in the program.
1706 Now save the file: choose the @code{File} menu item, and then the
1707 @code{Save buffer} selection. You will see a message at the bottom
1708 of the buffer confirming that the file has been saved.
1710 You are now ready to attempt to build the program. Select the @code{Ada}
1711 item from the top menu bar. Although we could choose simply to compile
1712 the file, we will instead attempt to do a build (which invokes
1713 @command{gnatmake}) since, if the compile is successful, we want to build
1714 an executable. Thus select @code{Ada build}. This will fail because of the
1715 compilation error, and you will notice that the Glide window has been split:
1716 the top window contains the source file, and the bottom window contains the
1717 output from the GNAT tools. Glide allows you to navigate from a compilation
1718 error to the source file position corresponding to the error: click the
1719 middle mouse button (or simultaneously press the left and right buttons,
1720 on a two-button mouse) on the diagnostic line in the tool window. The
1721 focus will shift to the source window, and the cursor will be positioned
1722 on the character at which the error was detected.
1724 Correct the error: type in a semicolon to terminate the statement.
1725 Although you can again save the file explicitly, you can also simply invoke
1726 @code{Ada} @result{} @code{Build} and you will be prompted to save the file.
1727 This time the build will succeed; the tool output window shows you the
1728 options that are supplied by default. The GNAT tools' output (e.g.
1729 object and ALI files, executable) will go in the directory from which
1732 To execute the program, choose @code{Ada} and then @code{Run}.
1733 You should see the program's output displayed in the bottom window:
1743 @node Simple Debugging with GVD
1744 @subsection Simple Debugging with GVD
1747 This section describes how to set breakpoints, examine/modify variables,
1748 and step through execution.
1750 In order to enable debugging, you need to pass the @option{-g} switch
1751 to both the compiler and to @command{gnatlink}. If you are using
1752 the command line, passing @option{-g} to @command{gnatmake} will have
1753 this effect. You can then launch GVD, e.g. on the @code{hello} program,
1754 by issuing the command:
1761 If you are using Glide, then @option{-g} is passed to the relevant tools
1762 by default when you do a build. Start the debugger by selecting the
1763 @code{Ada} menu item, and then @code{Debug}.
1765 GVD comes up in a multi-part window. One pane shows the names of files
1766 comprising your executable; another pane shows the source code of the current
1767 unit (initially your main subprogram), another pane shows the debugger output
1768 and user interactions, and the fourth pane (the data canvas at the top
1769 of the window) displays data objects that you have selected.
1771 To the left of the source file pane, you will notice green dots adjacent
1772 to some lines. These are lines for which object code exists and where
1773 breakpoints can thus be set. You set/reset a breakpoint by clicking
1774 the green dot. When a breakpoint is set, the dot is replaced by an @code{X}
1775 in a red circle. Clicking the circle toggles the breakpoint off,
1776 and the red circle is replaced by the green dot.
1778 For this example, set a breakpoint at the statement where @code{Put_Line}
1781 Start program execution by selecting the @code{Run} button on the top menu bar.
1782 (The @code{Start} button will also start your program, but it will
1783 cause program execution to break at the entry to your main subprogram.)
1784 Evidence of reaching the breakpoint will appear: the source file line will be
1785 highlighted, and the debugger interactions pane will display
1788 You can examine the values of variables in several ways. Move the mouse
1789 over an occurrence of @code{Ind} in the @code{for} loop, and you will see
1790 the value (now @code{1}) displayed. Alternatively, right-click on @code{Ind}
1791 and select @code{Display Ind}; a box showing the variable's name and value
1792 will appear in the data canvas.
1794 Although a loop index is a constant with respect to Ada semantics,
1795 you can change its value in the debugger. Right-click in the box
1796 for @code{Ind}, and select the @code{Set Value of Ind} item.
1797 Enter @code{2} as the new value, and press @command{OK}.
1798 The box for @code{Ind} shows the update.
1800 Press the @code{Step} button on the top menu bar; this will step through
1801 one line of program text (the invocation of @code{Put_Line}), and you can
1802 observe the effect of having modified @code{Ind} since the value displayed
1805 Remove the breakpoint, and resume execution by selecting the @code{Cont}
1806 button. You will see the remaining output lines displayed in the debugger
1807 interaction window, along with a message confirming normal program
1810 @node Other Glide Features
1811 @subsection Other Glide Features
1814 You may have observed that some of the menu selections contain abbreviations;
1815 e.g., @code{(C-x C-f)} for @code{Open file...} in the @code{Files} menu.
1816 These are @emph{shortcut keys} that you can use instead of selecting
1817 menu items. The @key{C} stands for @key{Ctrl}; thus @code{(C-x C-f)} means
1818 @key{Ctrl-x} followed by @key{Ctrl-f}, and this sequence can be used instead
1819 of selecting @code{Files} and then @code{Open file...}.
1821 To abort a Glide command, type @key{Ctrl-g}.
1823 If you want Glide to start with an existing source file, you can either
1824 launch Glide as above and then open the file via @code{Files} @result{}
1825 @code{Open file...}, or else simply pass the name of the source file
1826 on the command line:
1833 While you are using Glide, a number of @emph{buffers} exist.
1834 You create some explicitly; e.g., when you open/create a file.
1835 Others arise as an effect of the commands that you issue; e.g., the buffer
1836 containing the output of the tools invoked during a build. If a buffer
1837 is hidden, you can bring it into a visible window by first opening
1838 the @code{Buffers} menu and then selecting the desired entry.
1840 If a buffer occupies only part of the Glide screen and you want to expand it
1841 to fill the entire screen, then click in the buffer and then select
1842 @code{Files} @result{} @code{One Window}.
1844 If a window is occupied by one buffer and you want to split the window
1845 to bring up a second buffer, perform the following steps:
1847 @item Select @code{Files} @result{} @code{Split Window};
1848 this will produce two windows each of which holds the original buffer
1849 (these are not copies, but rather different views of the same buffer contents)
1851 @item With the focus in one of the windows,
1852 select the desired buffer from the @code{Buffers} menu
1856 To exit from Glide, choose @code{Files} @result{} @code{Exit}.
1859 @node The GNAT Compilation Model
1860 @chapter The GNAT Compilation Model
1861 @cindex GNAT compilation model
1862 @cindex Compilation model
1865 * Source Representation::
1866 * Foreign Language Representation::
1867 * File Naming Rules::
1868 * Using Other File Names::
1869 * Alternative File Naming Schemes::
1870 * Generating Object Files::
1871 * Source Dependencies::
1872 * The Ada Library Information Files::
1873 * Binding an Ada Program::
1874 * Mixed Language Programming::
1875 * Building Mixed Ada & C++ Programs::
1876 * Comparison between GNAT and C/C++ Compilation Models::
1877 * Comparison between GNAT and Conventional Ada Library Models::
1879 * Placement of temporary files::
1884 This chapter describes the compilation model used by GNAT. Although
1885 similar to that used by other languages, such as C and C++, this model
1886 is substantially different from the traditional Ada compilation models,
1887 which are based on a library. The model is initially described without
1888 reference to the library-based model. If you have not previously used an
1889 Ada compiler, you need only read the first part of this chapter. The
1890 last section describes and discusses the differences between the GNAT
1891 model and the traditional Ada compiler models. If you have used other
1892 Ada compilers, this section will help you to understand those
1893 differences, and the advantages of the GNAT model.
1895 @node Source Representation
1896 @section Source Representation
1900 Ada source programs are represented in standard text files, using
1901 Latin-1 coding. Latin-1 is an 8-bit code that includes the familiar
1902 7-bit ASCII set, plus additional characters used for
1903 representing foreign languages (@pxref{Foreign Language Representation}
1904 for support of non-USA character sets). The format effector characters
1905 are represented using their standard ASCII encodings, as follows:
1910 Vertical tab, @code{16#0B#}
1914 Horizontal tab, @code{16#09#}
1918 Carriage return, @code{16#0D#}
1922 Line feed, @code{16#0A#}
1926 Form feed, @code{16#0C#}
1930 Source files are in standard text file format. In addition, GNAT will
1931 recognize a wide variety of stream formats, in which the end of
1932 physical lines is marked by any of the following sequences:
1933 @code{LF}, @code{CR}, @code{CR-LF}, or @code{LF-CR}. This is useful
1934 in accommodating files that are imported from other operating systems.
1936 @cindex End of source file
1937 @cindex Source file, end
1939 The end of a source file is normally represented by the physical end of
1940 file. However, the control character @code{16#1A#} (@code{SUB}) is also
1941 recognized as signalling the end of the source file. Again, this is
1942 provided for compatibility with other operating systems where this
1943 code is used to represent the end of file.
1945 Each file contains a single Ada compilation unit, including any pragmas
1946 associated with the unit. For example, this means you must place a
1947 package declaration (a package @dfn{spec}) and the corresponding body in
1948 separate files. An Ada @dfn{compilation} (which is a sequence of
1949 compilation units) is represented using a sequence of files. Similarly,
1950 you will place each subunit or child unit in a separate file.
1952 @node Foreign Language Representation
1953 @section Foreign Language Representation
1956 GNAT supports the standard character sets defined in Ada 95 as well as
1957 several other non-standard character sets for use in localized versions
1958 of the compiler (@pxref{Character Set Control}).
1961 * Other 8-Bit Codes::
1962 * Wide Character Encodings::
1970 The basic character set is Latin-1. This character set is defined by ISO
1971 standard 8859, part 1. The lower half (character codes @code{16#00#}
1972 ... @code{16#7F#)} is identical to standard ASCII coding, but the upper half
1973 is used to represent additional characters. These include extended letters
1974 used by European languages, such as French accents, the vowels with umlauts
1975 used in German, and the extra letter A-ring used in Swedish.
1977 @findex Ada.Characters.Latin_1
1978 For a complete list of Latin-1 codes and their encodings, see the source
1979 file of library unit @code{Ada.Characters.Latin_1} in file
1980 @file{a-chlat1.ads}.
1981 You may use any of these extended characters freely in character or
1982 string literals. In addition, the extended characters that represent
1983 letters can be used in identifiers.
1985 @node Other 8-Bit Codes
1986 @subsection Other 8-Bit Codes
1989 GNAT also supports several other 8-bit coding schemes:
1992 @item ISO 8859-2 (Latin-2)
1995 Latin-2 letters allowed in identifiers, with uppercase and lowercase
1998 @item ISO 8859-3 (Latin-3)
2001 Latin-3 letters allowed in identifiers, with uppercase and lowercase
2004 @item ISO 8859-4 (Latin-4)
2007 Latin-4 letters allowed in identifiers, with uppercase and lowercase
2010 @item ISO 8859-5 (Cyrillic)
2013 ISO 8859-5 letters (Cyrillic) allowed in identifiers, with uppercase and
2014 lowercase equivalence.
2016 @item ISO 8859-15 (Latin-9)
2019 ISO 8859-15 (Latin-9) letters allowed in identifiers, with uppercase and
2020 lowercase equivalence
2022 @item IBM PC (code page 437)
2023 @cindex code page 437
2024 This code page is the normal default for PCs in the U.S. It corresponds
2025 to the original IBM PC character set. This set has some, but not all, of
2026 the extended Latin-1 letters, but these letters do not have the same
2027 encoding as Latin-1. In this mode, these letters are allowed in
2028 identifiers with uppercase and lowercase equivalence.
2030 @item IBM PC (code page 850)
2031 @cindex code page 850
2032 This code page is a modification of 437 extended to include all the
2033 Latin-1 letters, but still not with the usual Latin-1 encoding. In this
2034 mode, all these letters are allowed in identifiers with uppercase and
2035 lowercase equivalence.
2037 @item Full Upper 8-bit
2038 Any character in the range 80-FF allowed in identifiers, and all are
2039 considered distinct. In other words, there are no uppercase and lowercase
2040 equivalences in this range. This is useful in conjunction with
2041 certain encoding schemes used for some foreign character sets (e.g.
2042 the typical method of representing Chinese characters on the PC).
2045 No upper-half characters in the range 80-FF are allowed in identifiers.
2046 This gives Ada 83 compatibility for identifier names.
2050 For precise data on the encodings permitted, and the uppercase and lowercase
2051 equivalences that are recognized, see the file @file{csets.adb} in
2052 the GNAT compiler sources. You will need to obtain a full source release
2053 of GNAT to obtain this file.
2055 @node Wide Character Encodings
2056 @subsection Wide Character Encodings
2059 GNAT allows wide character codes to appear in character and string
2060 literals, and also optionally in identifiers, by means of the following
2061 possible encoding schemes:
2066 In this encoding, a wide character is represented by the following five
2074 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
2075 characters (using uppercase letters) of the wide character code. For
2076 example, ESC A345 is used to represent the wide character with code
2078 This scheme is compatible with use of the full Wide_Character set.
2080 @item Upper-Half Coding
2081 @cindex Upper-Half Coding
2082 The wide character with encoding @code{16#abcd#} where the upper bit is on
2083 (in other words, ``a'' is in the range 8-F) is represented as two bytes,
2084 @code{16#ab#} and @code{16#cd#}. The second byte cannot be a format control
2085 character, but is not required to be in the upper half. This method can
2086 be also used for shift-JIS or EUC, where the internal coding matches the
2089 @item Shift JIS Coding
2090 @cindex Shift JIS Coding
2091 A wide character is represented by a two-character sequence,
2093 @code{16#cd#}, with the restrictions described for upper-half encoding as
2094 described above. The internal character code is the corresponding JIS
2095 character according to the standard algorithm for Shift-JIS
2096 conversion. Only characters defined in the JIS code set table can be
2097 used with this encoding method.
2101 A wide character is represented by a two-character sequence
2103 @code{16#cd#}, with both characters being in the upper half. The internal
2104 character code is the corresponding JIS character according to the EUC
2105 encoding algorithm. Only characters defined in the JIS code set table
2106 can be used with this encoding method.
2109 A wide character is represented using
2110 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
2111 10646-1/Am.2. Depending on the character value, the representation
2112 is a one, two, or three byte sequence:
2117 16#0000#-16#007f#: 2#0xxxxxxx#
2118 16#0080#-16#07ff#: 2#110xxxxx# 2#10xxxxxx#
2119 16#0800#-16#ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
2124 where the xxx bits correspond to the left-padded bits of the
2125 16-bit character value. Note that all lower half ASCII characters
2126 are represented as ASCII bytes and all upper half characters and
2127 other wide characters are represented as sequences of upper-half
2128 (The full UTF-8 scheme allows for encoding 31-bit characters as
2129 6-byte sequences, but in this implementation, all UTF-8 sequences
2130 of four or more bytes length will be treated as illegal).
2131 @item Brackets Coding
2132 In this encoding, a wide character is represented by the following eight
2140 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
2141 characters (using uppercase letters) of the wide character code. For
2142 example, [``A345''] is used to represent the wide character with code
2143 @code{16#A345#}. It is also possible (though not required) to use the
2144 Brackets coding for upper half characters. For example, the code
2145 @code{16#A3#} can be represented as @code{[``A3'']}.
2147 This scheme is compatible with use of the full Wide_Character set,
2148 and is also the method used for wide character encoding in the standard
2149 ACVC (Ada Compiler Validation Capability) test suite distributions.
2154 Note: Some of these coding schemes do not permit the full use of the
2155 Ada 95 character set. For example, neither Shift JIS, nor EUC allow the
2156 use of the upper half of the Latin-1 set.
2158 @node File Naming Rules
2159 @section File Naming Rules
2162 The default file name is determined by the name of the unit that the
2163 file contains. The name is formed by taking the full expanded name of
2164 the unit and replacing the separating dots with hyphens and using
2165 ^lowercase^uppercase^ for all letters.
2167 An exception arises if the file name generated by the above rules starts
2168 with one of the characters
2175 and the second character is a
2176 minus. In this case, the character ^tilde^dollar sign^ is used in place
2177 of the minus. The reason for this special rule is to avoid clashes with
2178 the standard names for child units of the packages System, Ada,
2179 Interfaces, and GNAT, which use the prefixes
2188 The file extension is @file{.ads} for a spec and
2189 @file{.adb} for a body. The following list shows some
2190 examples of these rules.
2197 @item arith_functions.ads
2198 Arith_Functions (package spec)
2199 @item arith_functions.adb
2200 Arith_Functions (package body)
2202 Func.Spec (child package spec)
2204 Func.Spec (child package body)
2206 Sub (subunit of Main)
2207 @item ^a~bad.adb^A$BAD.ADB^
2208 A.Bad (child package body)
2212 Following these rules can result in excessively long
2213 file names if corresponding
2214 unit names are long (for example, if child units or subunits are
2215 heavily nested). An option is available to shorten such long file names
2216 (called file name ``krunching''). This may be particularly useful when
2217 programs being developed with GNAT are to be used on operating systems
2218 with limited file name lengths. @xref{Using gnatkr}.
2220 Of course, no file shortening algorithm can guarantee uniqueness over
2221 all possible unit names; if file name krunching is used, it is your
2222 responsibility to ensure no name clashes occur. Alternatively you
2223 can specify the exact file names that you want used, as described
2224 in the next section. Finally, if your Ada programs are migrating from a
2225 compiler with a different naming convention, you can use the gnatchop
2226 utility to produce source files that follow the GNAT naming conventions.
2227 (For details @pxref{Renaming Files Using gnatchop}.)
2229 Note: in the case of @code{Windows NT/XP} or @code{OpenVMS} operating
2230 systems, case is not significant. So for example on @code{Windows XP}
2231 if the canonical name is @code{main-sub.adb}, you can use the file name
2232 @code{Main-Sub.adb} instead. However, case is significant for other
2233 operating systems, so for example, if you want to use other than
2234 canonically cased file names on a Unix system, you need to follow
2235 the procedures described in the next section.
2237 @node Using Other File Names
2238 @section Using Other File Names
2242 In the previous section, we have described the default rules used by
2243 GNAT to determine the file name in which a given unit resides. It is
2244 often convenient to follow these default rules, and if you follow them,
2245 the compiler knows without being explicitly told where to find all
2248 However, in some cases, particularly when a program is imported from
2249 another Ada compiler environment, it may be more convenient for the
2250 programmer to specify which file names contain which units. GNAT allows
2251 arbitrary file names to be used by means of the Source_File_Name pragma.
2252 The form of this pragma is as shown in the following examples:
2253 @cindex Source_File_Name pragma
2255 @smallexample @c ada
2257 pragma Source_File_Name (My_Utilities.Stacks,
2258 Spec_File_Name => "myutilst_a.ada");
2259 pragma Source_File_name (My_Utilities.Stacks,
2260 Body_File_Name => "myutilst.ada");
2265 As shown in this example, the first argument for the pragma is the unit
2266 name (in this example a child unit). The second argument has the form
2267 of a named association. The identifier
2268 indicates whether the file name is for a spec or a body;
2269 the file name itself is given by a string literal.
2271 The source file name pragma is a configuration pragma, which means that
2272 normally it will be placed in the @file{gnat.adc}
2273 file used to hold configuration
2274 pragmas that apply to a complete compilation environment.
2275 For more details on how the @file{gnat.adc} file is created and used
2276 see @ref{Handling of Configuration Pragmas}.
2277 @cindex @file{gnat.adc}
2280 GNAT allows completely arbitrary file names to be specified using the
2281 source file name pragma. However, if the file name specified has an
2282 extension other than @file{.ads} or @file{.adb} it is necessary to use
2283 a special syntax when compiling the file. The name in this case must be
2284 preceded by the special sequence @code{-x} followed by a space and the name
2285 of the language, here @code{ada}, as in:
2288 $ gcc -c -x ada peculiar_file_name.sim
2293 @command{gnatmake} handles non-standard file names in the usual manner (the
2294 non-standard file name for the main program is simply used as the
2295 argument to gnatmake). Note that if the extension is also non-standard,
2296 then it must be included in the gnatmake command, it may not be omitted.
2298 @node Alternative File Naming Schemes
2299 @section Alternative File Naming Schemes
2300 @cindex File naming schemes, alternative
2303 In the previous section, we described the use of the @code{Source_File_Name}
2304 pragma to allow arbitrary names to be assigned to individual source files.
2305 However, this approach requires one pragma for each file, and especially in
2306 large systems can result in very long @file{gnat.adc} files, and also create
2307 a maintenance problem.
2309 GNAT also provides a facility for specifying systematic file naming schemes
2310 other than the standard default naming scheme previously described. An
2311 alternative scheme for naming is specified by the use of
2312 @code{Source_File_Name} pragmas having the following format:
2313 @cindex Source_File_Name pragma
2315 @smallexample @c ada
2316 pragma Source_File_Name (
2317 Spec_File_Name => FILE_NAME_PATTERN
2318 [,Casing => CASING_SPEC]
2319 [,Dot_Replacement => STRING_LITERAL]);
2321 pragma Source_File_Name (
2322 Body_File_Name => FILE_NAME_PATTERN
2323 [,Casing => CASING_SPEC]
2324 [,Dot_Replacement => STRING_LITERAL]);
2326 pragma Source_File_Name (
2327 Subunit_File_Name => FILE_NAME_PATTERN
2328 [,Casing => CASING_SPEC]
2329 [,Dot_Replacement => STRING_LITERAL]);
2331 FILE_NAME_PATTERN ::= STRING_LITERAL
2332 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
2336 The @code{FILE_NAME_PATTERN} string shows how the file name is constructed.
2337 It contains a single asterisk character, and the unit name is substituted
2338 systematically for this asterisk. The optional parameter
2339 @code{Casing} indicates
2340 whether the unit name is to be all upper-case letters, all lower-case letters,
2341 or mixed-case. If no
2342 @code{Casing} parameter is used, then the default is all
2343 ^lower-case^upper-case^.
2345 The optional @code{Dot_Replacement} string is used to replace any periods
2346 that occur in subunit or child unit names. If no @code{Dot_Replacement}
2347 argument is used then separating dots appear unchanged in the resulting
2349 Although the above syntax indicates that the
2350 @code{Casing} argument must appear
2351 before the @code{Dot_Replacement} argument, but it
2352 is also permissible to write these arguments in the opposite order.
2354 As indicated, it is possible to specify different naming schemes for
2355 bodies, specs, and subunits. Quite often the rule for subunits is the
2356 same as the rule for bodies, in which case, there is no need to give
2357 a separate @code{Subunit_File_Name} rule, and in this case the
2358 @code{Body_File_name} rule is used for subunits as well.
2360 The separate rule for subunits can also be used to implement the rather
2361 unusual case of a compilation environment (e.g. a single directory) which
2362 contains a subunit and a child unit with the same unit name. Although
2363 both units cannot appear in the same partition, the Ada Reference Manual
2364 allows (but does not require) the possibility of the two units coexisting
2365 in the same environment.
2367 The file name translation works in the following steps:
2372 If there is a specific @code{Source_File_Name} pragma for the given unit,
2373 then this is always used, and any general pattern rules are ignored.
2376 If there is a pattern type @code{Source_File_Name} pragma that applies to
2377 the unit, then the resulting file name will be used if the file exists. If
2378 more than one pattern matches, the latest one will be tried first, and the
2379 first attempt resulting in a reference to a file that exists will be used.
2382 If no pattern type @code{Source_File_Name} pragma that applies to the unit
2383 for which the corresponding file exists, then the standard GNAT default
2384 naming rules are used.
2389 As an example of the use of this mechanism, consider a commonly used scheme
2390 in which file names are all lower case, with separating periods copied
2391 unchanged to the resulting file name, and specs end with @file{.1.ada}, and
2392 bodies end with @file{.2.ada}. GNAT will follow this scheme if the following
2395 @smallexample @c ada
2396 pragma Source_File_Name
2397 (Spec_File_Name => "*.1.ada");
2398 pragma Source_File_Name
2399 (Body_File_Name => "*.2.ada");
2403 The default GNAT scheme is actually implemented by providing the following
2404 default pragmas internally:
2406 @smallexample @c ada
2407 pragma Source_File_Name
2408 (Spec_File_Name => "*.ads", Dot_Replacement => "-");
2409 pragma Source_File_Name
2410 (Body_File_Name => "*.adb", Dot_Replacement => "-");
2414 Our final example implements a scheme typically used with one of the
2415 Ada 83 compilers, where the separator character for subunits was ``__''
2416 (two underscores), specs were identified by adding @file{_.ADA}, bodies
2417 by adding @file{.ADA}, and subunits by
2418 adding @file{.SEP}. All file names were
2419 upper case. Child units were not present of course since this was an
2420 Ada 83 compiler, but it seems reasonable to extend this scheme to use
2421 the same double underscore separator for child units.
2423 @smallexample @c ada
2424 pragma Source_File_Name
2425 (Spec_File_Name => "*_.ADA",
2426 Dot_Replacement => "__",
2427 Casing = Uppercase);
2428 pragma Source_File_Name
2429 (Body_File_Name => "*.ADA",
2430 Dot_Replacement => "__",
2431 Casing = Uppercase);
2432 pragma Source_File_Name
2433 (Subunit_File_Name => "*.SEP",
2434 Dot_Replacement => "__",
2435 Casing = Uppercase);
2438 @node Generating Object Files
2439 @section Generating Object Files
2442 An Ada program consists of a set of source files, and the first step in
2443 compiling the program is to generate the corresponding object files.
2444 These are generated by compiling a subset of these source files.
2445 The files you need to compile are the following:
2449 If a package spec has no body, compile the package spec to produce the
2450 object file for the package.
2453 If a package has both a spec and a body, compile the body to produce the
2454 object file for the package. The source file for the package spec need
2455 not be compiled in this case because there is only one object file, which
2456 contains the code for both the spec and body of the package.
2459 For a subprogram, compile the subprogram body to produce the object file
2460 for the subprogram. The spec, if one is present, is as usual in a
2461 separate file, and need not be compiled.
2465 In the case of subunits, only compile the parent unit. A single object
2466 file is generated for the entire subunit tree, which includes all the
2470 Compile child units independently of their parent units
2471 (though, of course, the spec of all the ancestor unit must be present in order
2472 to compile a child unit).
2476 Compile generic units in the same manner as any other units. The object
2477 files in this case are small dummy files that contain at most the
2478 flag used for elaboration checking. This is because GNAT always handles generic
2479 instantiation by means of macro expansion. However, it is still necessary to
2480 compile generic units, for dependency checking and elaboration purposes.
2484 The preceding rules describe the set of files that must be compiled to
2485 generate the object files for a program. Each object file has the same
2486 name as the corresponding source file, except that the extension is
2489 You may wish to compile other files for the purpose of checking their
2490 syntactic and semantic correctness. For example, in the case where a
2491 package has a separate spec and body, you would not normally compile the
2492 spec. However, it is convenient in practice to compile the spec to make
2493 sure it is error-free before compiling clients of this spec, because such
2494 compilations will fail if there is an error in the spec.
2496 GNAT provides an option for compiling such files purely for the
2497 purposes of checking correctness; such compilations are not required as
2498 part of the process of building a program. To compile a file in this
2499 checking mode, use the @option{-gnatc} switch.
2501 @node Source Dependencies
2502 @section Source Dependencies
2505 A given object file clearly depends on the source file which is compiled
2506 to produce it. Here we are using @dfn{depends} in the sense of a typical
2507 @code{make} utility; in other words, an object file depends on a source
2508 file if changes to the source file require the object file to be
2510 In addition to this basic dependency, a given object may depend on
2511 additional source files as follows:
2515 If a file being compiled @code{with}'s a unit @var{X}, the object file
2516 depends on the file containing the spec of unit @var{X}. This includes
2517 files that are @code{with}'ed implicitly either because they are parents
2518 of @code{with}'ed child units or they are run-time units required by the
2519 language constructs used in a particular unit.
2522 If a file being compiled instantiates a library level generic unit, the
2523 object file depends on both the spec and body files for this generic
2527 If a file being compiled instantiates a generic unit defined within a
2528 package, the object file depends on the body file for the package as
2529 well as the spec file.
2533 @cindex @option{-gnatn} switch
2534 If a file being compiled contains a call to a subprogram for which
2535 pragma @code{Inline} applies and inlining is activated with the
2536 @option{-gnatn} switch, the object file depends on the file containing the
2537 body of this subprogram as well as on the file containing the spec. Note
2538 that for inlining to actually occur as a result of the use of this switch,
2539 it is necessary to compile in optimizing mode.
2541 @cindex @option{-gnatN} switch
2542 The use of @option{-gnatN} activates a more extensive inlining optimization
2543 that is performed by the front end of the compiler. This inlining does
2544 not require that the code generation be optimized. Like @option{-gnatn},
2545 the use of this switch generates additional dependencies.
2547 @option{-gnatN} automatically implies @option{-gnatn} so it is not necessary
2548 to specify both options.
2551 If an object file O depends on the proper body of a subunit through inlining
2552 or instantiation, it depends on the parent unit of the subunit. This means that
2553 any modification of the parent unit or one of its subunits affects the
2557 The object file for a parent unit depends on all its subunit body files.
2560 The previous two rules meant that for purposes of computing dependencies and
2561 recompilation, a body and all its subunits are treated as an indivisible whole.
2564 These rules are applied transitively: if unit @code{A} @code{with}'s
2565 unit @code{B}, whose elaboration calls an inlined procedure in package
2566 @code{C}, the object file for unit @code{A} will depend on the body of
2567 @code{C}, in file @file{c.adb}.
2569 The set of dependent files described by these rules includes all the
2570 files on which the unit is semantically dependent, as described in the
2571 Ada 95 Language Reference Manual. However, it is a superset of what the
2572 ARM describes, because it includes generic, inline, and subunit dependencies.
2574 An object file must be recreated by recompiling the corresponding source
2575 file if any of the source files on which it depends are modified. For
2576 example, if the @code{make} utility is used to control compilation,
2577 the rule for an Ada object file must mention all the source files on
2578 which the object file depends, according to the above definition.
2579 The determination of the necessary
2580 recompilations is done automatically when one uses @command{gnatmake}.
2583 @node The Ada Library Information Files
2584 @section The Ada Library Information Files
2585 @cindex Ada Library Information files
2586 @cindex @file{ALI} files
2589 Each compilation actually generates two output files. The first of these
2590 is the normal object file that has a @file{.o} extension. The second is a
2591 text file containing full dependency information. It has the same
2592 name as the source file, but an @file{.ali} extension.
2593 This file is known as the Ada Library Information (@file{ALI}) file.
2594 The following information is contained in the @file{ALI} file.
2598 Version information (indicates which version of GNAT was used to compile
2599 the unit(s) in question)
2602 Main program information (including priority and time slice settings,
2603 as well as the wide character encoding used during compilation).
2606 List of arguments used in the @command{gcc} command for the compilation
2609 Attributes of the unit, including configuration pragmas used, an indication
2610 of whether the compilation was successful, exception model used etc.
2613 A list of relevant restrictions applying to the unit (used for consistency)
2617 Categorization information (e.g. use of pragma @code{Pure}).
2620 Information on all @code{with}'ed units, including presence of
2621 @code{Elaborate} or @code{Elaborate_All} pragmas.
2624 Information from any @code{Linker_Options} pragmas used in the unit
2627 Information on the use of @code{Body_Version} or @code{Version}
2628 attributes in the unit.
2631 Dependency information. This is a list of files, together with
2632 time stamp and checksum information. These are files on which
2633 the unit depends in the sense that recompilation is required
2634 if any of these units are modified.
2637 Cross-reference data. Contains information on all entities referenced
2638 in the unit. Used by tools like @code{gnatxref} and @code{gnatfind} to
2639 provide cross-reference information.
2644 For a full detailed description of the format of the @file{ALI} file,
2645 see the source of the body of unit @code{Lib.Writ}, contained in file
2646 @file{lib-writ.adb} in the GNAT compiler sources.
2648 @node Binding an Ada Program
2649 @section Binding an Ada Program
2652 When using languages such as C and C++, once the source files have been
2653 compiled the only remaining step in building an executable program
2654 is linking the object modules together. This means that it is possible to
2655 link an inconsistent version of a program, in which two units have
2656 included different versions of the same header.
2658 The rules of Ada do not permit such an inconsistent program to be built.
2659 For example, if two clients have different versions of the same package,
2660 it is illegal to build a program containing these two clients.
2661 These rules are enforced by the GNAT binder, which also determines an
2662 elaboration order consistent with the Ada rules.
2664 The GNAT binder is run after all the object files for a program have
2665 been created. It is given the name of the main program unit, and from
2666 this it determines the set of units required by the program, by reading the
2667 corresponding ALI files. It generates error messages if the program is
2668 inconsistent or if no valid order of elaboration exists.
2670 If no errors are detected, the binder produces a main program, in Ada by
2671 default, that contains calls to the elaboration procedures of those
2672 compilation unit that require them, followed by
2673 a call to the main program. This Ada program is compiled to generate the
2674 object file for the main program. The name of
2675 the Ada file is @file{b~@var{xxx}.adb} (with the corresponding spec
2676 @file{b~@var{xxx}.ads}) where @var{xxx} is the name of the
2679 Finally, the linker is used to build the resulting executable program,
2680 using the object from the main program from the bind step as well as the
2681 object files for the Ada units of the program.
2683 @node Mixed Language Programming
2684 @section Mixed Language Programming
2685 @cindex Mixed Language Programming
2688 This section describes how to develop a mixed-language program,
2689 specifically one that comprises units in both Ada and C.
2692 * Interfacing to C::
2693 * Calling Conventions::
2696 @node Interfacing to C
2697 @subsection Interfacing to C
2699 Interfacing Ada with a foreign language such as C involves using
2700 compiler directives to import and/or export entity definitions in each
2701 language---using @code{extern} statements in C, for instance, and the
2702 @code{Import}, @code{Export}, and @code{Convention} pragmas in Ada. For
2703 a full treatment of these topics, read Appendix B, section 1 of the Ada
2704 95 Language Reference Manual.
2706 There are two ways to build a program using GNAT that contains some Ada
2707 sources and some foreign language sources, depending on whether or not
2708 the main subprogram is written in Ada. Here is a source example with
2709 the main subprogram in Ada:
2715 void print_num (int num)
2717 printf ("num is %d.\n", num);
2723 /* num_from_Ada is declared in my_main.adb */
2724 extern int num_from_Ada;
2728 return num_from_Ada;
2732 @smallexample @c ada
2734 procedure My_Main is
2736 -- Declare then export an Integer entity called num_from_Ada
2737 My_Num : Integer := 10;
2738 pragma Export (C, My_Num, "num_from_Ada");
2740 -- Declare an Ada function spec for Get_Num, then use
2741 -- C function get_num for the implementation.
2742 function Get_Num return Integer;
2743 pragma Import (C, Get_Num, "get_num");
2745 -- Declare an Ada procedure spec for Print_Num, then use
2746 -- C function print_num for the implementation.
2747 procedure Print_Num (Num : Integer);
2748 pragma Import (C, Print_Num, "print_num");
2751 Print_Num (Get_Num);
2757 To build this example, first compile the foreign language files to
2758 generate object files:
2765 Then, compile the Ada units to produce a set of object files and ALI
2768 gnatmake ^-c^/ACTIONS=COMPILE^ my_main.adb
2772 Run the Ada binder on the Ada main program:
2774 gnatbind my_main.ali
2778 Link the Ada main program, the Ada objects and the other language
2781 gnatlink my_main.ali file1.o file2.o
2785 The last three steps can be grouped in a single command:
2787 gnatmake my_main.adb -largs file1.o file2.o
2790 @cindex Binder output file
2792 If the main program is in a language other than Ada, then you may have
2793 more than one entry point into the Ada subsystem. You must use a special
2794 binder option to generate callable routines that initialize and
2795 finalize the Ada units (@pxref{Binding with Non-Ada Main Programs}).
2796 Calls to the initialization and finalization routines must be inserted
2797 in the main program, or some other appropriate point in the code. The
2798 call to initialize the Ada units must occur before the first Ada
2799 subprogram is called, and the call to finalize the Ada units must occur
2800 after the last Ada subprogram returns. The binder will place the
2801 initialization and finalization subprograms into the
2802 @file{b~@var{xxx}.adb} file where they can be accessed by your C
2803 sources. To illustrate, we have the following example:
2807 extern void adainit (void);
2808 extern void adafinal (void);
2809 extern int add (int, int);
2810 extern int sub (int, int);
2812 int main (int argc, char *argv[])
2818 /* Should print "21 + 7 = 28" */
2819 printf ("%d + %d = %d\n", a, b, add (a, b));
2820 /* Should print "21 - 7 = 14" */
2821 printf ("%d - %d = %d\n", a, b, sub (a, b));
2827 @smallexample @c ada
2830 function Add (A, B : Integer) return Integer;
2831 pragma Export (C, Add, "add");
2835 package body Unit1 is
2836 function Add (A, B : Integer) return Integer is
2844 function Sub (A, B : Integer) return Integer;
2845 pragma Export (C, Sub, "sub");
2849 package body Unit2 is
2850 function Sub (A, B : Integer) return Integer is
2859 The build procedure for this application is similar to the last
2860 example's. First, compile the foreign language files to generate object
2867 Next, compile the Ada units to produce a set of object files and ALI
2870 gnatmake ^-c^/ACTIONS=COMPILE^ unit1.adb
2871 gnatmake ^-c^/ACTIONS=COMPILE^ unit2.adb
2875 Run the Ada binder on every generated ALI file. Make sure to use the
2876 @option{-n} option to specify a foreign main program:
2878 gnatbind ^-n^/NOMAIN^ unit1.ali unit2.ali
2882 Link the Ada main program, the Ada objects and the foreign language
2883 objects. You need only list the last ALI file here:
2885 gnatlink unit2.ali main.o -o exec_file
2888 This procedure yields a binary executable called @file{exec_file}.
2891 @node Calling Conventions
2892 @subsection Calling Conventions
2893 @cindex Foreign Languages
2894 @cindex Calling Conventions
2895 GNAT follows standard calling sequence conventions and will thus interface
2896 to any other language that also follows these conventions. The following
2897 Convention identifiers are recognized by GNAT:
2900 @cindex Interfacing to Ada
2901 @cindex Other Ada compilers
2902 @cindex Convention Ada
2904 This indicates that the standard Ada calling sequence will be
2905 used and all Ada data items may be passed without any limitations in the
2906 case where GNAT is used to generate both the caller and callee. It is also
2907 possible to mix GNAT generated code and code generated by another Ada
2908 compiler. In this case, the data types should be restricted to simple
2909 cases, including primitive types. Whether complex data types can be passed
2910 depends on the situation. Probably it is safe to pass simple arrays, such
2911 as arrays of integers or floats. Records may or may not work, depending
2912 on whether both compilers lay them out identically. Complex structures
2913 involving variant records, access parameters, tasks, or protected types,
2914 are unlikely to be able to be passed.
2916 Note that in the case of GNAT running
2917 on a platform that supports DEC Ada 83, a higher degree of compatibility
2918 can be guaranteed, and in particular records are layed out in an identical
2919 manner in the two compilers. Note also that if output from two different
2920 compilers is mixed, the program is responsible for dealing with elaboration
2921 issues. Probably the safest approach is to write the main program in the
2922 version of Ada other than GNAT, so that it takes care of its own elaboration
2923 requirements, and then call the GNAT-generated adainit procedure to ensure
2924 elaboration of the GNAT components. Consult the documentation of the other
2925 Ada compiler for further details on elaboration.
2927 However, it is not possible to mix the tasking run time of GNAT and
2928 DEC Ada 83, All the tasking operations must either be entirely within
2929 GNAT compiled sections of the program, or entirely within DEC Ada 83
2930 compiled sections of the program.
2932 @cindex Interfacing to Assembly
2933 @cindex Convention Assembler
2935 Specifies assembler as the convention. In practice this has the
2936 same effect as convention Ada (but is not equivalent in the sense of being
2937 considered the same convention).
2939 @cindex Convention Asm
2942 Equivalent to Assembler.
2944 @cindex Interfacing to COBOL
2945 @cindex Convention COBOL
2948 Data will be passed according to the conventions described
2949 in section B.4 of the Ada 95 Reference Manual.
2952 @cindex Interfacing to C
2953 @cindex Convention C
2955 Data will be passed according to the conventions described
2956 in section B.3 of the Ada 95 Reference Manual.
2958 @findex C varargs function
2959 @cindex Intefacing to C varargs function
2960 @cindex varargs function interfaces
2961 @item C varargs function
2962 In C, @code{varargs} allows a function to take a variable number of
2963 arguments. There is no direct equivalent in this to Ada. One
2964 approach that can be used is to create a C wrapper for each
2965 different profile and then interface to this C wrapper. For
2966 example, to print an @code{int} value using @code{printf},
2967 create a C function @code{printfi} that takes two arguments, a
2968 pointer to a string and an int, and calls @code{printf}.
2969 Then in the Ada program, use pragma @code{Import} to
2970 interface to printfi.
2972 It may work on some platforms to directly interface to
2973 a @code{varargs} function by providing a specific Ada profile
2974 for a a particular call. However, this does not work on
2975 all platforms, since there is no guarantee that the
2976 calling sequence for a two argument normal C function
2977 is the same as for calling a @code{varargs} C function with
2978 the same two arguments.
2980 @cindex Convention Default
2985 @cindex Convention External
2991 @cindex Interfacing to C++
2992 @cindex Convention C++
2994 This stands for C++. For most purposes this is identical to C.
2995 See the separate description of the specialized GNAT pragmas relating to
2996 C++ interfacing for further details.
2999 @cindex Interfacing to Fortran
3000 @cindex Convention Fortran
3002 Data will be passed according to the conventions described
3003 in section B.5 of the Ada 95 Reference Manual.
3006 This applies to an intrinsic operation, as defined in the Ada 95
3007 Reference Manual. If a a pragma Import (Intrinsic) applies to a subprogram,
3008 this means that the body of the subprogram is provided by the compiler itself,
3009 usually by means of an efficient code sequence, and that the user does not
3010 supply an explicit body for it. In an application program, the pragma can
3011 only be applied to the following two sets of names, which the GNAT compiler
3016 Rotate_Left, Rotate_Right, Shift_Left, Shift_Right, Shift_Right_-
3017 Arithmetic. The corresponding subprogram declaration must have
3018 two formal parameters. The
3019 first one must be a signed integer type or a modular type with a binary
3020 modulus, and the second parameter must be of type Natural.
3021 The return type must be the same as the type of the first argument. The size
3022 of this type can only be 8, 16, 32, or 64.
3023 @item binary arithmetic operators: ``+'', ``-'', ``*'', ``/''
3024 The corresponding operator declaration must have parameters and result type
3025 that have the same root numeric type (for example, all three are long_float
3026 types). This simplifies the definition of operations that use type checking
3027 to perform dimensional checks:
3029 @smallexample @c ada
3030 type Distance is new Long_Float;
3031 type Time is new Long_Float;
3032 type Velocity is new Long_Float;
3033 function "/" (D : Distance; T : Time)
3035 pragma Import (Intrinsic, "/");
3039 This common idiom is often programmed with a generic definition and an
3040 explicit body. The pragma makes it simpler to introduce such declarations.
3041 It incurs no overhead in compilation time or code size, because it is
3042 implemented as a single machine instruction.
3048 @cindex Convention Stdcall
3050 This is relevant only to NT/Win95 implementations of GNAT,
3051 and specifies that the Stdcall calling sequence will be used, as defined
3052 by the NT API. Nevertheless, to ease building cross-platform bindings this
3053 convention will be handled as a C calling convention on non Windows
3057 @cindex Convention DLL
3059 This is equivalent to Stdcall.
3062 @cindex Convention Win32
3064 This is equivalent to Stdcall.
3068 @cindex Convention Stubbed
3070 This is a special convention that indicates that the compiler
3071 should provide a stub body that raises @code{Program_Error}.
3075 GNAT additionally provides a useful pragma @code{Convention_Identifier}
3076 that can be used to parametrize conventions and allow additional synonyms
3077 to be specified. For example if you have legacy code in which the convention
3078 identifier Fortran77 was used for Fortran, you can use the configuration
3081 @smallexample @c ada
3082 pragma Convention_Identifier (Fortran77, Fortran);
3086 And from now on the identifier Fortran77 may be used as a convention
3087 identifier (for example in an @code{Import} pragma) with the same
3090 @node Building Mixed Ada & C++ Programs
3091 @section Building Mixed Ada & C++ Programs
3094 A programmer inexperienced with mixed-language development may find that
3095 building an application containing both Ada and C++ code can be a
3096 challenge. As a matter of fact, interfacing with C++ has not been
3097 standardized in the Ada 95 Reference Manual due to the immaturity of --
3098 and lack of standards for -- C++ at the time. This section gives a few
3099 hints that should make this task easier. The first section addresses
3100 the differences regarding interfacing with C. The second section
3101 looks into the delicate problem of linking the complete application from
3102 its Ada and C++ parts. The last section gives some hints on how the GNAT
3103 run time can be adapted in order to allow inter-language dispatching
3104 with a new C++ compiler.
3107 * Interfacing to C++::
3108 * Linking a Mixed C++ & Ada Program::
3109 * A Simple Example::
3110 * Adapting the Run Time to a New C++ Compiler::
3113 @node Interfacing to C++
3114 @subsection Interfacing to C++
3117 GNAT supports interfacing with C++ compilers generating code that is
3118 compatible with the standard Application Binary Interface of the given
3122 Interfacing can be done at 3 levels: simple data, subprograms, and
3123 classes. In the first two cases, GNAT offers a specific @var{Convention
3124 CPP} that behaves exactly like @var{Convention C}. Usually, C++ mangles
3125 the names of subprograms, and currently, GNAT does not provide any help
3126 to solve the demangling problem. This problem can be addressed in two
3130 by modifying the C++ code in order to force a C convention using
3131 the @code{extern "C"} syntax.
3134 by figuring out the mangled name and use it as the Link_Name argument of
3139 Interfacing at the class level can be achieved by using the GNAT specific
3140 pragmas such as @code{CPP_Class} and @code{CPP_Virtual}. See the GNAT
3141 Reference Manual for additional information.
3143 @node Linking a Mixed C++ & Ada Program
3144 @subsection Linking a Mixed C++ & Ada Program
3147 Usually the linker of the C++ development system must be used to link
3148 mixed applications because most C++ systems will resolve elaboration
3149 issues (such as calling constructors on global class instances)
3150 transparently during the link phase. GNAT has been adapted to ease the
3151 use of a foreign linker for the last phase. Three cases can be
3156 Using GNAT and G++ (GNU C++ compiler) from the same GCC installation:
3157 The C++ linker can simply be called by using the C++ specific driver
3158 called @code{c++}. Note that this setup is not very common because it
3159 may involve recompiling the whole GCC tree from sources, which makes it
3160 harder to upgrade the compilation system for one language without
3161 destabilizing the other.
3166 $ gnatmake ada_unit -largs file1.o file2.o --LINK=c++
3170 Using GNAT and G++ from two different GCC installations: If both
3171 compilers are on the PATH, the previous method may be used. It is
3172 important to note that environment variables such as C_INCLUDE_PATH,
3173 GCC_EXEC_PREFIX, BINUTILS_ROOT, and GCC_ROOT will affect both compilers
3174 at the same time and may make one of the two compilers operate
3175 improperly if set during invocation of the wrong compiler. It is also
3176 very important that the linker uses the proper @file{libgcc.a} GCC
3177 library -- that is, the one from the C++ compiler installation. The
3178 implicit link command as suggested in the gnatmake command from the
3179 former example can be replaced by an explicit link command with the
3180 full-verbosity option in order to verify which library is used:
3183 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=c++
3185 If there is a problem due to interfering environment variables, it can
3186 be worked around by using an intermediate script. The following example
3187 shows the proper script to use when GNAT has not been installed at its
3188 default location and g++ has been installed at its default location:
3196 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=./my_script
3200 Using a non-GNU C++ compiler: The commands previously described can be
3201 used to insure that the C++ linker is used. Nonetheless, you need to add
3202 a few more parameters to the link command line, depending on the exception
3205 If the @code{setjmp/longjmp} exception mechanism is used, only the paths
3206 to the libgcc libraries are required:
3211 CC $* `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a`
3212 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3215 Where CC is the name of the non-GNU C++ compiler.
3217 If the @code{zero cost} exception mechanism is used, and the platform
3218 supports automatic registration of exception tables (e.g. Solaris or IRIX),
3219 paths to more objects are required:
3224 CC `gcc -print-file-name=crtbegin.o` $* \
3225 `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a` \
3226 `gcc -print-file-name=crtend.o`
3227 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3230 If the @code{zero cost} exception mechanism is used, and the platform
3231 doesn't support automatic registration of exception tables (e.g. HP-UX,
3232 Tru64 or AIX), the simple approach described above will not work and
3233 a pre-linking phase using GNAT will be necessary.
3237 @node A Simple Example
3238 @subsection A Simple Example
3240 The following example, provided as part of the GNAT examples, shows how
3241 to achieve procedural interfacing between Ada and C++ in both
3242 directions. The C++ class A has two methods. The first method is exported
3243 to Ada by the means of an extern C wrapper function. The second method
3244 calls an Ada subprogram. On the Ada side, The C++ calls are modelled by
3245 a limited record with a layout comparable to the C++ class. The Ada
3246 subprogram, in turn, calls the C++ method. So, starting from the C++
3247 main program, the process passes back and forth between the two
3251 Here are the compilation commands:
3253 $ gnatmake -c simple_cpp_interface
3256 $ gnatbind -n simple_cpp_interface
3257 $ gnatlink simple_cpp_interface -o cpp_main --LINK=$(CPLUSPLUS)
3258 -lstdc++ ex7.o cpp_main.o
3262 Here are the corresponding sources:
3270 void adainit (void);
3271 void adafinal (void);
3272 void method1 (A *t);
3294 class A : public Origin @{
3296 void method1 (void);
3297 void method2 (int v);
3307 extern "C" @{ void ada_method2 (A *t, int v);@}
3309 void A::method1 (void)
3312 printf ("in A::method1, a_value = %d \n",a_value);
3316 void A::method2 (int v)
3318 ada_method2 (this, v);
3319 printf ("in A::method2, a_value = %d \n",a_value);
3326 printf ("in A::A, a_value = %d \n",a_value);
3330 @b{package} @b{body} Simple_Cpp_Interface @b{is}
3332 @b{procedure} Ada_Method2 (This : @b{in} @b{out} A; V : Integer) @b{is}
3336 @b{end} Ada_Method2;
3338 @b{end} Simple_Cpp_Interface;
3340 @b{package} Simple_Cpp_Interface @b{is}
3341 @b{type} A @b{is} @b{limited}
3346 @b{pragma} Convention (C, A);
3348 @b{procedure} Method1 (This : @b{in} @b{out} A);
3349 @b{pragma} Import (C, Method1);
3351 @b{procedure} Ada_Method2 (This : @b{in} @b{out} A; V : Integer);
3352 @b{pragma} Export (C, Ada_Method2);
3354 @b{end} Simple_Cpp_Interface;
3357 @node Adapting the Run Time to a New C++ Compiler
3358 @subsection Adapting the Run Time to a New C++ Compiler
3360 GNAT offers the capability to derive Ada 95 tagged types directly from
3361 preexisting C++ classes and . See ``Interfacing with C++'' in the
3362 @cite{GNAT Reference Manual}. The mechanism used by GNAT for achieving
3364 has been made user configurable through a GNAT library unit
3365 @code{Interfaces.CPP}. The default version of this file is adapted to
3366 the GNU C++ compiler. Internal knowledge of the virtual
3367 table layout used by the new C++ compiler is needed to configure
3368 properly this unit. The Interface of this unit is known by the compiler
3369 and cannot be changed except for the value of the constants defining the
3370 characteristics of the virtual table: CPP_DT_Prologue_Size, CPP_DT_Entry_Size,
3371 CPP_TSD_Prologue_Size, CPP_TSD_Entry_Size. Read comments in the source
3372 of this unit for more details.
3374 @node Comparison between GNAT and C/C++ Compilation Models
3375 @section Comparison between GNAT and C/C++ Compilation Models
3378 The GNAT model of compilation is close to the C and C++ models. You can
3379 think of Ada specs as corresponding to header files in C. As in C, you
3380 don't need to compile specs; they are compiled when they are used. The
3381 Ada @code{with} is similar in effect to the @code{#include} of a C
3384 One notable difference is that, in Ada, you may compile specs separately
3385 to check them for semantic and syntactic accuracy. This is not always
3386 possible with C headers because they are fragments of programs that have
3387 less specific syntactic or semantic rules.
3389 The other major difference is the requirement for running the binder,
3390 which performs two important functions. First, it checks for
3391 consistency. In C or C++, the only defense against assembling
3392 inconsistent programs lies outside the compiler, in a makefile, for
3393 example. The binder satisfies the Ada requirement that it be impossible
3394 to construct an inconsistent program when the compiler is used in normal
3397 @cindex Elaboration order control
3398 The other important function of the binder is to deal with elaboration
3399 issues. There are also elaboration issues in C++ that are handled
3400 automatically. This automatic handling has the advantage of being
3401 simpler to use, but the C++ programmer has no control over elaboration.
3402 Where @code{gnatbind} might complain there was no valid order of
3403 elaboration, a C++ compiler would simply construct a program that
3404 malfunctioned at run time.
3406 @node Comparison between GNAT and Conventional Ada Library Models
3407 @section Comparison between GNAT and Conventional Ada Library Models
3410 This section is intended to be useful to Ada programmers who have
3411 previously used an Ada compiler implementing the traditional Ada library
3412 model, as described in the Ada 95 Language Reference Manual. If you
3413 have not used such a system, please go on to the next section.
3415 @cindex GNAT library
3416 In GNAT, there is no @dfn{library} in the normal sense. Instead, the set of
3417 source files themselves acts as the library. Compiling Ada programs does
3418 not generate any centralized information, but rather an object file and
3419 a ALI file, which are of interest only to the binder and linker.
3420 In a traditional system, the compiler reads information not only from
3421 the source file being compiled, but also from the centralized library.
3422 This means that the effect of a compilation depends on what has been
3423 previously compiled. In particular:
3427 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3428 to the version of the unit most recently compiled into the library.
3431 Inlining is effective only if the necessary body has already been
3432 compiled into the library.
3435 Compiling a unit may obsolete other units in the library.
3439 In GNAT, compiling one unit never affects the compilation of any other
3440 units because the compiler reads only source files. Only changes to source
3441 files can affect the results of a compilation. In particular:
3445 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3446 to the source version of the unit that is currently accessible to the
3451 Inlining requires the appropriate source files for the package or
3452 subprogram bodies to be available to the compiler. Inlining is always
3453 effective, independent of the order in which units are complied.
3456 Compiling a unit never affects any other compilations. The editing of
3457 sources may cause previous compilations to be out of date if they
3458 depended on the source file being modified.
3462 The most important result of these differences is that order of compilation
3463 is never significant in GNAT. There is no situation in which one is
3464 required to do one compilation before another. What shows up as order of
3465 compilation requirements in the traditional Ada library becomes, in
3466 GNAT, simple source dependencies; in other words, there is only a set
3467 of rules saying what source files must be present when a file is
3471 @node Placement of temporary files
3472 @section Placement of temporary files
3473 @cindex Temporary files (user control over placement)
3476 GNAT creates temporary files in the directory designated by the environment
3477 variable @env{TMPDIR}.
3478 (See the HP @emph{C RTL Reference Manual} on the function @code{getenv()}
3479 for detailed information on how environment variables are resolved.
3480 For most users the easiest way to make use of this feature is to simply
3481 define @env{TMPDIR} as a job level logical name).
3482 For example, if you wish to use a Ramdisk (assuming DECRAM is installed)
3483 for compiler temporary files, then you can include something like the
3484 following command in your @file{LOGIN.COM} file:
3487 $ define/job TMPDIR "/disk$scratchram/000000/temp/"
3491 If @env{TMPDIR} is not defined, then GNAT uses the directory designated by
3492 @env{TMP}; if @env{TMP} is not defined, then GNAT uses the directory
3493 designated by @env{TEMP}.
3494 If none of these environment variables are defined then GNAT uses the
3495 directory designated by the logical name @code{SYS$SCRATCH:}
3496 (by default the user's home directory). If all else fails
3497 GNAT uses the current directory for temporary files.
3500 @c *************************
3501 @node Compiling Using gcc
3502 @chapter Compiling Using @command{gcc}
3505 This chapter discusses how to compile Ada programs using the @command{gcc}
3506 command. It also describes the set of switches
3507 that can be used to control the behavior of the compiler.
3509 * Compiling Programs::
3510 * Switches for gcc::
3511 * Search Paths and the Run-Time Library (RTL)::
3512 * Order of Compilation Issues::
3516 @node Compiling Programs
3517 @section Compiling Programs
3520 The first step in creating an executable program is to compile the units
3521 of the program using the @command{gcc} command. You must compile the
3526 the body file (@file{.adb}) for a library level subprogram or generic
3530 the spec file (@file{.ads}) for a library level package or generic
3531 package that has no body
3534 the body file (@file{.adb}) for a library level package
3535 or generic package that has a body
3540 You need @emph{not} compile the following files
3545 the spec of a library unit which has a body
3552 because they are compiled as part of compiling related units. GNAT
3554 when the corresponding body is compiled, and subunits when the parent is
3557 @cindex cannot generate code
3558 If you attempt to compile any of these files, you will get one of the
3559 following error messages (where fff is the name of the file you compiled):
3562 cannot generate code for file @var{fff} (package spec)
3563 to check package spec, use -gnatc
3565 cannot generate code for file @var{fff} (missing subunits)
3566 to check parent unit, use -gnatc
3568 cannot generate code for file @var{fff} (subprogram spec)
3569 to check subprogram spec, use -gnatc
3571 cannot generate code for file @var{fff} (subunit)
3572 to check subunit, use -gnatc
3576 As indicated by the above error messages, if you want to submit
3577 one of these files to the compiler to check for correct semantics
3578 without generating code, then use the @option{-gnatc} switch.
3580 The basic command for compiling a file containing an Ada unit is
3583 $ gcc -c [@var{switches}] @file{file name}
3587 where @var{file name} is the name of the Ada file (usually
3589 @file{.ads} for a spec or @file{.adb} for a body).
3592 @option{-c} switch to tell @command{gcc} to compile, but not link, the file.
3594 The result of a successful compilation is an object file, which has the
3595 same name as the source file but an extension of @file{.o} and an Ada
3596 Library Information (ALI) file, which also has the same name as the
3597 source file, but with @file{.ali} as the extension. GNAT creates these
3598 two output files in the current directory, but you may specify a source
3599 file in any directory using an absolute or relative path specification
3600 containing the directory information.
3603 @command{gcc} is actually a driver program that looks at the extensions of
3604 the file arguments and loads the appropriate compiler. For example, the
3605 GNU C compiler is @file{cc1}, and the Ada compiler is @file{gnat1}.
3606 These programs are in directories known to the driver program (in some
3607 configurations via environment variables you set), but need not be in
3608 your path. The @command{gcc} driver also calls the assembler and any other
3609 utilities needed to complete the generation of the required object
3612 It is possible to supply several file names on the same @command{gcc}
3613 command. This causes @command{gcc} to call the appropriate compiler for
3614 each file. For example, the following command lists three separate
3615 files to be compiled:
3618 $ gcc -c x.adb y.adb z.c
3622 calls @code{gnat1} (the Ada compiler) twice to compile @file{x.adb} and
3623 @file{y.adb}, and @code{cc1} (the C compiler) once to compile @file{z.c}.
3624 The compiler generates three object files @file{x.o}, @file{y.o} and
3625 @file{z.o} and the two ALI files @file{x.ali} and @file{y.ali} from the
3626 Ada compilations. Any switches apply to all the files ^listed,^listed.^
3629 @option{-gnat@var{x}} switches, which apply only to Ada compilations.
3632 @node Switches for gcc
3633 @section Switches for @command{gcc}
3636 The @command{gcc} command accepts switches that control the
3637 compilation process. These switches are fully described in this section.
3638 First we briefly list all the switches, in alphabetical order, then we
3639 describe the switches in more detail in functionally grouped sections.
3641 More switches exist for GCC than those documented here, especially
3642 for specific targets. However, their use is not recommended as
3643 they may change code generation in ways that are incompatible with
3644 the Ada run-time library, or can cause inconsistencies between
3648 * Output and Error Message Control::
3649 * Warning Message Control::
3650 * Debugging and Assertion Control::
3651 * Validity Checking::
3654 * Stack Overflow Checking::
3655 * Using gcc for Syntax Checking::
3656 * Using gcc for Semantic Checking::
3657 * Compiling Different Versions of Ada::
3658 * Character Set Control::
3659 * File Naming Control::
3660 * Subprogram Inlining Control::
3661 * Auxiliary Output Control::
3662 * Debugging Control::
3663 * Exception Handling Control::
3664 * Units to Sources Mapping Files::
3665 * Integrated Preprocessing::
3666 * Code Generation Control::
3675 @cindex @option{-b} (@command{gcc})
3676 @item -b @var{target}
3677 Compile your program to run on @var{target}, which is the name of a
3678 system configuration. You must have a GNAT cross-compiler built if
3679 @var{target} is not the same as your host system.
3682 @cindex @option{-B} (@command{gcc})
3683 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
3684 from @var{dir} instead of the default location. Only use this switch
3685 when multiple versions of the GNAT compiler are available. See the
3686 @command{gcc} manual page for further details. You would normally use the
3687 @option{-b} or @option{-V} switch instead.
3690 @cindex @option{-c} (@command{gcc})
3691 Compile. Always use this switch when compiling Ada programs.
3693 Note: for some other languages when using @command{gcc}, notably in
3694 the case of C and C++, it is possible to use
3695 use @command{gcc} without a @option{-c} switch to
3696 compile and link in one step. In the case of GNAT, you
3697 cannot use this approach, because the binder must be run
3698 and @command{gcc} cannot be used to run the GNAT binder.
3702 @cindex @option{-fno-inline} (@command{gcc})
3703 Suppresses all back-end inlining, even if other optimization or inlining
3705 This includes suppression of inlining that results
3706 from the use of the pragma @code{Inline_Always}.
3707 See also @option{-gnatn} and @option{-gnatN}.
3709 @item -fno-strict-aliasing
3710 @cindex @option{-fno-strict-aliasing} (@command{gcc})
3711 Causes the compiler to avoid assumptions regarding non-aliasing
3712 of objects of different types. See
3713 @ref{Optimization and Strict Aliasing} for details.
3716 @cindex @option{-fstack-check} (@command{gcc})
3717 Activates stack checking.
3718 See @ref{Stack Overflow Checking} for details of the use of this option.
3721 @cindex @option{-fstack-usage} (@command{gcc})
3722 Makes the compiler output stack usage information for the program, on a
3723 per-function basis. The description of the format is to be found in
3724 the GCC documentation.
3726 @item -fcallgraph-info
3727 @cindex @option{-fcallgraph-info} (@command{gcc})
3728 Makes the compiler output callgraph information for the program, on a
3729 per-file basis. The information is generated in the VCG format. It can
3730 be decorated with additional, per-node information if other debugging
3731 options are enabled (only works with -fstack-usage as of this writing).
3734 @cindex @option{^-g^/DEBUG^} (@command{gcc})
3735 Generate debugging information. This information is stored in the object
3736 file and copied from there to the final executable file by the linker,
3737 where it can be read by the debugger. You must use the
3738 @option{^-g^/DEBUG^} switch if you plan on using the debugger.
3741 @cindex @option{-gnat83} (@command{gcc})
3742 Enforce Ada 83 restrictions.
3745 @cindex @option{-gnat95} (@command{gcc})
3746 Enforce Ada 95 restrictions.
3749 @cindex @option{-gnat05} (@command{gcc})
3750 Allow full Ada 2005 features.
3753 @cindex @option{-gnata} (@command{gcc})
3754 Assertions enabled. @code{Pragma Assert} and @code{pragma Debug} to be
3758 @cindex @option{-gnatA} (@command{gcc})
3759 Avoid processing @file{gnat.adc}. If a gnat.adc file is present,
3763 @cindex @option{-gnatb} (@command{gcc})
3764 Generate brief messages to @file{stderr} even if verbose mode set.
3767 @cindex @option{-gnatc} (@command{gcc})
3768 Check syntax and semantics only (no code generation attempted).
3771 @cindex @option{-gnatd} (@command{gcc})
3772 Specify debug options for the compiler. The string of characters after
3773 the @option{-gnatd} specify the specific debug options. The possible
3774 characters are 0-9, a-z, A-Z, optionally preceded by a dot. See
3775 compiler source file @file{debug.adb} for details of the implemented
3776 debug options. Certain debug options are relevant to applications
3777 programmers, and these are documented at appropriate points in this
3781 @cindex @option{-gnatD} (@command{gcc})
3782 Create expanded source files for source level debugging. This switch
3783 also suppress generation of cross-reference information
3784 (see @option{-gnatx}).
3786 @item -gnatec=@var{path}
3787 @cindex @option{-gnatec} (@command{gcc})
3788 Specify a configuration pragma file
3790 (the equal sign is optional)
3792 (@pxref{The Configuration Pragmas Files}).
3794 @item ^-gnateD^/DATA_PREPROCESSING=^symbol[=value]
3795 @cindex @option{-gnateD} (@command{gcc})
3796 Defines a symbol, associated with value, for preprocessing.
3797 (@pxref{Integrated Preprocessing}).
3800 @cindex @option{-gnatef} (@command{gcc})
3801 Display full source path name in brief error messages.
3803 @item -gnatem=@var{path}
3804 @cindex @option{-gnatem} (@command{gcc})
3805 Specify a mapping file
3807 (the equal sign is optional)
3809 (@pxref{Units to Sources Mapping Files}).
3811 @item -gnatep=@var{file}
3812 @cindex @option{-gnatep} (@command{gcc})
3813 Specify a preprocessing data file
3815 (the equal sign is optional)
3817 (@pxref{Integrated Preprocessing}).
3820 @cindex @option{-gnatE} (@command{gcc})
3821 Full dynamic elaboration checks.
3824 @cindex @option{-gnatf} (@command{gcc})
3825 Full errors. Multiple errors per line, all undefined references, do not
3826 attempt to suppress cascaded errors.
3829 @cindex @option{-gnatF} (@command{gcc})
3830 Externals names are folded to all uppercase.
3833 @cindex @option{-gnatg} (@command{gcc})
3834 Internal GNAT implementation mode. This should not be used for
3835 applications programs, it is intended only for use by the compiler
3836 and its run-time library. For documentation, see the GNAT sources.
3837 Note that @option{-gnatg} implies @option{-gnatwu} so that warnings
3838 are generated on unreferenced entities, and all warnings are treated
3842 @cindex @option{-gnatG} (@command{gcc})
3843 List generated expanded code in source form.
3845 @item ^-gnath^/HELP^
3846 @cindex @option{^-gnath^/HELP^} (@command{gcc})
3847 Output usage information. The output is written to @file{stdout}.
3849 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
3850 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
3851 Identifier character set
3853 (@var{c}=1/2/3/4/8/9/p/f/n/w).
3856 For details of the possible selections for @var{c},
3857 see @ref{Character Set Control}.
3860 @item -gnatk=@var{n}
3861 @cindex @option{-gnatk} (@command{gcc})
3862 Limit file names to @var{n} (1-999) characters ^(@code{k} = krunch)^^.
3865 @cindex @option{-gnatl} (@command{gcc})
3866 Output full source listing with embedded error messages.
3869 @cindex @option{-gnatL} (@command{gcc})
3870 This switch is deprecated. You can use @option{--RTS=sjlj} instead to enable
3871 @code{setjmp/longjmp} exception mechanism.
3873 @item -gnatm=@var{n}
3874 @cindex @option{-gnatm} (@command{gcc})
3875 Limit number of detected error or warning messages to @var{n}
3876 where @var{n} is in the range 1..999_999. The default setting if
3877 no switch is given is 9999. Compilation is terminated if this
3881 @cindex @option{-gnatn} (@command{gcc})
3882 Activate inlining for subprograms for which
3883 pragma @code{inline} is specified. This inlining is performed
3884 by the GCC back-end.
3887 @cindex @option{-gnatN} (@command{gcc})
3888 Activate front end inlining for subprograms for which
3889 pragma @code{Inline} is specified. This inlining is performed
3890 by the front end and will be visible in the
3891 @option{-gnatG} output.
3892 In some cases, this has proved more effective than the back end
3893 inlining resulting from the use of
3896 @option{-gnatN} automatically implies
3897 @option{-gnatn} so it is not necessary
3898 to specify both options. There are a few cases that the back-end inlining
3899 catches that cannot be dealt with in the front-end.
3902 @cindex @option{-gnato} (@command{gcc})
3903 Enable numeric overflow checking (which is not normally enabled by
3904 default). Not that division by zero is a separate check that is not
3905 controlled by this switch (division by zero checking is on by default).
3908 @cindex @option{-gnatp} (@command{gcc})
3909 Suppress all checks.
3912 @cindex @option{-gnatP} (@command{gcc})
3913 Enable polling. This is required on some systems (notably Windows NT) to
3914 obtain asynchronous abort and asynchronous transfer of control capability.
3915 See the description of pragma Polling in the GNAT Reference Manual for
3919 @cindex @option{-gnatq} (@command{gcc})
3920 Don't quit; try semantics, even if parse errors.
3923 @cindex @option{-gnatQ} (@command{gcc})
3924 Don't quit; generate @file{ALI} and tree files even if illegalities.
3926 @item ^-gnatR[0/1/2/3[s]]^/REPRESENTATION_INFO^
3927 @cindex @option{-gnatR} (@command{gcc})
3928 Output representation information for declared types and objects.
3931 @cindex @option{-gnats} (@command{gcc})
3935 @cindex @option{-gnatS} (@command{gcc})
3936 Print package Standard.
3939 @cindex @option{-gnatt} (@command{gcc})
3940 Generate tree output file.
3942 @item ^-gnatT^/TABLE_MULTIPLIER=^@var{nnn}
3943 @cindex @option{^-gnatT^/TABLE_MULTIPLIER^} (@command{gcc})
3944 All compiler tables start at @var{nnn} times usual starting size.
3947 @cindex @option{-gnatu} (@command{gcc})
3948 List units for this compilation.
3951 @cindex @option{-gnatU} (@command{gcc})
3952 Tag all error messages with the unique string ``error:''
3955 @cindex @option{-gnatv} (@command{gcc})
3956 Verbose mode. Full error output with source lines to @file{stdout}.
3959 @cindex @option{-gnatV} (@command{gcc})
3960 Control level of validity checking. See separate section describing
3963 @item ^-gnatw@var{xxx}^/WARNINGS=(@var{option}[,...])^
3964 @cindex @option{^-gnatw^/WARNINGS^} (@command{gcc})
3966 ^@var{xxx} is a string of option letters that^the list of options^ denotes
3967 the exact warnings that
3968 are enabled or disabled (@pxref{Warning Message Control}).
3970 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
3971 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
3972 Wide character encoding method
3974 (@var{e}=n/h/u/s/e/8).
3977 (@var{e}=@code{BRACKETS, NONE, HEX, UPPER, SHIFT_JIS, EUC, UTF8})
3981 @cindex @option{-gnatx} (@command{gcc})
3982 Suppress generation of cross-reference information.
3984 @item ^-gnaty^/STYLE_CHECKS=(option,option..)^
3985 @cindex @option{^-gnaty^/STYLE_CHECKS^} (@command{gcc})
3986 Enable built-in style checks (@pxref{Style Checking}).
3988 @item ^-gnatz^/DISTRIBUTION_STUBS=^@var{m}
3989 @cindex @option{^-gnatz^/DISTRIBUTION_STUBS^} (@command{gcc})
3990 Distribution stub generation and compilation
3992 (@var{m}=r/c for receiver/caller stubs).
3995 (@var{m}=@code{RECEIVER} or @code{CALLER} to specify the type of stubs
3996 to be generated and compiled).
4000 This switch is deprecated. When zero cost exception handling is not the
4001 default and this is supported, you can use @option{--RTS=zcx} instead.
4003 @item ^-I^/SEARCH=^@var{dir}
4004 @cindex @option{^-I^/SEARCH^} (@command{gcc})
4006 Direct GNAT to search the @var{dir} directory for source files needed by
4007 the current compilation
4008 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4010 @item ^-I-^/NOCURRENT_DIRECTORY^
4011 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gcc})
4013 Except for the source file named in the command line, do not look for source
4014 files in the directory containing the source file named in the command line
4015 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4019 @cindex @option{-mbig-switch} (@command{gcc})
4020 @cindex @code{case} statement (effect of @option{-mbig-switch} option)
4021 This standard gcc switch causes the compiler to use larger offsets in its
4022 jump table representation for @code{case} statements.
4023 This may result in less efficient code, but is sometimes necessary
4024 (for example on HP-UX targets)
4025 @cindex HP-UX and @option{-mbig-switch} option
4026 in order to compile large and/or nested @code{case} statements.
4029 @cindex @option{-o} (@command{gcc})
4030 This switch is used in @command{gcc} to redirect the generated object file
4031 and its associated ALI file. Beware of this switch with GNAT, because it may
4032 cause the object file and ALI file to have different names which in turn
4033 may confuse the binder and the linker.
4037 @cindex @option{-nostdinc} (@command{gcc})
4038 Inhibit the search of the default location for the GNAT Run Time
4039 Library (RTL) source files.
4042 @cindex @option{-nostdlib} (@command{gcc})
4043 Inhibit the search of the default location for the GNAT Run Time
4044 Library (RTL) ALI files.
4048 @cindex @option{-O} (@command{gcc})
4049 @var{n} controls the optimization level.
4053 No optimization, the default setting if no @option{-O} appears
4056 Normal optimization, the default if you specify @option{-O} without
4060 Extensive optimization
4063 Extensive optimization with automatic inlining of subprograms not
4064 specified by pragma @code{Inline}. This applies only to
4065 inlining within a unit. For details on control of inlining
4066 see @ref{Subprogram Inlining Control}.
4072 @cindex @option{/NOOPTIMIZE} (@code{GNAT COMPILE})
4073 Equivalent to @option{/OPTIMIZE=NONE}.
4074 This is the default behavior in the absence of an @option{/OPTMIZE}
4077 @item /OPTIMIZE[=(keyword[,...])]
4078 @cindex @option{/OPTIMIZE} (@code{GNAT COMPILE})
4079 Selects the level of optimization for your program. The supported
4080 keywords are as follows:
4083 Perform most optimizations, including those that
4085 This is the default if the @option{/OPTMIZE} qualifier is supplied
4086 without keyword options.
4089 Do not do any optimizations. Same as @code{/NOOPTIMIZE}.
4092 Perform some optimizations, but omit ones that are costly.
4095 Same as @code{SOME}.
4098 Full optimization, and also attempt automatic inlining of small
4099 subprograms within a unit even when pragma @code{Inline}
4100 is not specified (@pxref{Inlining of Subprograms}).
4103 Try to unroll loops. This keyword may be specified together with
4104 any keyword above other than @code{NONE}. Loop unrolling
4105 usually, but not always, improves the performance of programs.
4110 @item -pass-exit-codes
4111 @cindex @option{-pass-exit-codes} (@command{gcc})
4112 Catch exit codes from the compiler and use the most meaningful as
4116 @item --RTS=@var{rts-path}
4117 @cindex @option{--RTS} (@command{gcc})
4118 Specifies the default location of the runtime library. Same meaning as the
4119 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
4122 @cindex @option{^-S^/ASM^} (@command{gcc})
4123 ^Used in place of @option{-c} to^Used to^
4124 cause the assembler source file to be
4125 generated, using @file{^.s^.S^} as the extension,
4126 instead of the object file.
4127 This may be useful if you need to examine the generated assembly code.
4129 @item ^-fverbose-asm^/VERBOSE_ASM^
4130 @cindex @option{^-fverbose-asm^/VERBOSE_ASM^} (@command{gcc})
4131 ^Used in conjunction with @option{-S}^Used in place of @option{/ASM}^
4132 to cause the generated assembly code file to be annotated with variable
4133 names, making it significantly easier to follow.
4136 @cindex @option{^-v^/VERBOSE^} (@command{gcc})
4137 Show commands generated by the @command{gcc} driver. Normally used only for
4138 debugging purposes or if you need to be sure what version of the
4139 compiler you are executing.
4143 @cindex @option{-V} (@command{gcc})
4144 Execute @var{ver} version of the compiler. This is the @command{gcc}
4145 version, not the GNAT version.
4151 You may combine a sequence of GNAT switches into a single switch. For
4152 example, the combined switch
4154 @cindex Combining GNAT switches
4160 is equivalent to specifying the following sequence of switches:
4163 -gnato -gnatf -gnati3
4167 @c NEED TO CHECK THIS FOR VMS
4170 The following restrictions apply to the combination of switches
4175 The switch @option{-gnatc} if combined with other switches must come
4176 first in the string.
4179 The switch @option{-gnats} if combined with other switches must come
4180 first in the string.
4184 @option{^-gnatz^/DISTRIBUTION_STUBS^}, @option{-gnatzc}, and @option{-gnatzr}
4185 may not be combined with any other switches.
4189 Once a ``y'' appears in the string (that is a use of the @option{-gnaty}
4190 switch), then all further characters in the switch are interpreted
4191 as style modifiers (see description of @option{-gnaty}).
4194 Once a ``d'' appears in the string (that is a use of the @option{-gnatd}
4195 switch), then all further characters in the switch are interpreted
4196 as debug flags (see description of @option{-gnatd}).
4199 Once a ``w'' appears in the string (that is a use of the @option{-gnatw}
4200 switch), then all further characters in the switch are interpreted
4201 as warning mode modifiers (see description of @option{-gnatw}).
4204 Once a ``V'' appears in the string (that is a use of the @option{-gnatV}
4205 switch), then all further characters in the switch are interpreted
4206 as validity checking options (see description of @option{-gnatV}).
4210 @node Output and Error Message Control
4211 @subsection Output and Error Message Control
4215 The standard default format for error messages is called ``brief format''.
4216 Brief format messages are written to @file{stderr} (the standard error
4217 file) and have the following form:
4220 e.adb:3:04: Incorrect spelling of keyword "function"
4221 e.adb:4:20: ";" should be "is"
4225 The first integer after the file name is the line number in the file,
4226 and the second integer is the column number within the line.
4227 @code{glide} can parse the error messages
4228 and point to the referenced character.
4229 The following switches provide control over the error message
4235 @cindex @option{-gnatv} (@command{gcc})
4238 The v stands for verbose.
4240 The effect of this setting is to write long-format error
4241 messages to @file{stdout} (the standard output file.
4242 The same program compiled with the
4243 @option{-gnatv} switch would generate:
4247 3. funcion X (Q : Integer)
4249 >>> Incorrect spelling of keyword "function"
4252 >>> ";" should be "is"
4257 The vertical bar indicates the location of the error, and the @samp{>>>}
4258 prefix can be used to search for error messages. When this switch is
4259 used the only source lines output are those with errors.
4262 @cindex @option{-gnatl} (@command{gcc})
4264 The @code{l} stands for list.
4266 This switch causes a full listing of
4267 the file to be generated. The output might look as follows:
4273 3. funcion X (Q : Integer)
4275 >>> Incorrect spelling of keyword "function"
4278 >>> ";" should be "is"
4290 When you specify the @option{-gnatv} or @option{-gnatl} switches and
4291 standard output is redirected, a brief summary is written to
4292 @file{stderr} (standard error) giving the number of error messages and
4293 warning messages generated.
4296 @cindex @option{-gnatU} (@command{gcc})
4297 This switch forces all error messages to be preceded by the unique
4298 string ``error:''. This means that error messages take a few more
4299 characters in space, but allows easy searching for and identification
4303 @cindex @option{-gnatb} (@command{gcc})
4305 The @code{b} stands for brief.
4307 This switch causes GNAT to generate the
4308 brief format error messages to @file{stderr} (the standard error
4309 file) as well as the verbose
4310 format message or full listing (which as usual is written to
4311 @file{stdout} (the standard output file).
4313 @item -gnatm^^=^@var{n}
4314 @cindex @option{-gnatm} (@command{gcc})
4316 The @code{m} stands for maximum.
4318 @var{n} is a decimal integer in the
4319 range of 1 to 999 and limits the number of error messages to be
4320 generated. For example, using @option{-gnatm2} might yield
4323 e.adb:3:04: Incorrect spelling of keyword "function"
4324 e.adb:5:35: missing ".."
4325 fatal error: maximum errors reached
4326 compilation abandoned
4330 @cindex @option{-gnatf} (@command{gcc})
4331 @cindex Error messages, suppressing
4333 The @code{f} stands for full.
4335 Normally, the compiler suppresses error messages that are likely to be
4336 redundant. This switch causes all error
4337 messages to be generated. In particular, in the case of
4338 references to undefined variables. If a given variable is referenced
4339 several times, the normal format of messages is
4341 e.adb:7:07: "V" is undefined (more references follow)
4345 where the parenthetical comment warns that there are additional
4346 references to the variable @code{V}. Compiling the same program with the
4347 @option{-gnatf} switch yields
4350 e.adb:7:07: "V" is undefined
4351 e.adb:8:07: "V" is undefined
4352 e.adb:8:12: "V" is undefined
4353 e.adb:8:16: "V" is undefined
4354 e.adb:9:07: "V" is undefined
4355 e.adb:9:12: "V" is undefined
4359 The @option{-gnatf} switch also generates additional information for
4360 some error messages. Some examples are:
4364 Full details on entities not available in high integrity mode
4366 Details on possibly non-portable unchecked conversion
4368 List possible interpretations for ambiguous calls
4370 Additional details on incorrect parameters
4374 @cindex @option{-gnatq} (@command{gcc})
4376 The @code{q} stands for quit (really ``don't quit'').
4378 In normal operation mode, the compiler first parses the program and
4379 determines if there are any syntax errors. If there are, appropriate
4380 error messages are generated and compilation is immediately terminated.
4382 GNAT to continue with semantic analysis even if syntax errors have been
4383 found. This may enable the detection of more errors in a single run. On
4384 the other hand, the semantic analyzer is more likely to encounter some
4385 internal fatal error when given a syntactically invalid tree.
4388 @cindex @option{-gnatQ} (@command{gcc})
4389 In normal operation mode, the @file{ALI} file is not generated if any
4390 illegalities are detected in the program. The use of @option{-gnatQ} forces
4391 generation of the @file{ALI} file. This file is marked as being in
4392 error, so it cannot be used for binding purposes, but it does contain
4393 reasonably complete cross-reference information, and thus may be useful
4394 for use by tools (e.g. semantic browsing tools or integrated development
4395 environments) that are driven from the @file{ALI} file. This switch
4396 implies @option{-gnatq}, since the semantic phase must be run to get a
4397 meaningful ALI file.
4399 In addition, if @option{-gnatt} is also specified, then the tree file is
4400 generated even if there are illegalities. It may be useful in this case
4401 to also specify @option{-gnatq} to ensure that full semantic processing
4402 occurs. The resulting tree file can be processed by ASIS, for the purpose
4403 of providing partial information about illegal units, but if the error
4404 causes the tree to be badly malformed, then ASIS may crash during the
4407 When @option{-gnatQ} is used and the generated @file{ALI} file is marked as
4408 being in error, @command{gnatmake} will attempt to recompile the source when it
4409 finds such an @file{ALI} file, including with switch @option{-gnatc}.
4411 Note that @option{-gnatQ} has no effect if @option{-gnats} is specified,
4412 since ALI files are never generated if @option{-gnats} is set.
4416 @node Warning Message Control
4417 @subsection Warning Message Control
4418 @cindex Warning messages
4420 In addition to error messages, which correspond to illegalities as defined
4421 in the Ada 95 Reference Manual, the compiler detects two kinds of warning
4424 First, the compiler considers some constructs suspicious and generates a
4425 warning message to alert you to a possible error. Second, if the
4426 compiler detects a situation that is sure to raise an exception at
4427 run time, it generates a warning message. The following shows an example
4428 of warning messages:
4430 e.adb:4:24: warning: creation of object may raise Storage_Error
4431 e.adb:10:17: warning: static value out of range
4432 e.adb:10:17: warning: "Constraint_Error" will be raised at run time
4436 GNAT considers a large number of situations as appropriate
4437 for the generation of warning messages. As always, warnings are not
4438 definite indications of errors. For example, if you do an out-of-range
4439 assignment with the deliberate intention of raising a
4440 @code{Constraint_Error} exception, then the warning that may be
4441 issued does not indicate an error. Some of the situations for which GNAT
4442 issues warnings (at least some of the time) are given in the following
4443 list. This list is not complete, and new warnings are often added to
4444 subsequent versions of GNAT. The list is intended to give a general idea
4445 of the kinds of warnings that are generated.
4449 Possible infinitely recursive calls
4452 Out-of-range values being assigned
4455 Possible order of elaboration problems
4461 Fixed-point type declarations with a null range
4464 Direct_IO or Sequential_IO instantiated with a type that has access values
4467 Variables that are never assigned a value
4470 Variables that are referenced before being initialized
4473 Task entries with no corresponding @code{accept} statement
4476 Duplicate accepts for the same task entry in a @code{select}
4479 Objects that take too much storage
4482 Unchecked conversion between types of differing sizes
4485 Missing @code{return} statement along some execution path in a function
4488 Incorrect (unrecognized) pragmas
4491 Incorrect external names
4494 Allocation from empty storage pool
4497 Potentially blocking operation in protected type
4500 Suspicious parenthesization of expressions
4503 Mismatching bounds in an aggregate
4506 Attempt to return local value by reference
4509 Premature instantiation of a generic body
4512 Attempt to pack aliased components
4515 Out of bounds array subscripts
4518 Wrong length on string assignment
4521 Violations of style rules if style checking is enabled
4524 Unused @code{with} clauses
4527 @code{Bit_Order} usage that does not have any effect
4530 @code{Standard.Duration} used to resolve universal fixed expression
4533 Dereference of possibly null value
4536 Declaration that is likely to cause storage error
4539 Internal GNAT unit @code{with}'ed by application unit
4542 Values known to be out of range at compile time
4545 Unreferenced labels and variables
4548 Address overlays that could clobber memory
4551 Unexpected initialization when address clause present
4554 Bad alignment for address clause
4557 Useless type conversions
4560 Redundant assignment statements and other redundant constructs
4563 Useless exception handlers
4566 Accidental hiding of name by child unit
4569 Access before elaboration detected at compile time
4572 A range in a @code{for} loop that is known to be null or might be null
4577 The following switches are available to control the handling of
4583 @emph{Activate all optional errors.}
4584 @cindex @option{-gnatwa} (@command{gcc})
4585 This switch activates most optional warning messages, see remaining list
4586 in this section for details on optional warning messages that can be
4587 individually controlled. The warnings that are not turned on by this
4589 @option{-gnatwd} (implicit dereferencing),
4590 @option{-gnatwh} (hiding),
4591 and @option{-gnatwl} (elaboration warnings).
4592 All other optional warnings are turned on.
4595 @emph{Suppress all optional errors.}
4596 @cindex @option{-gnatwA} (@command{gcc})
4597 This switch suppresses all optional warning messages, see remaining list
4598 in this section for details on optional warning messages that can be
4599 individually controlled.
4602 @emph{Activate warnings on bad fixed values.}
4603 @cindex @option{-gnatwb} (@command{gcc})
4604 @cindex Bad fixed values
4605 @cindex Fixed-point Small value
4607 This switch activates warnings for static fixed-point expressions whose
4608 value is not an exact multiple of Small. Such values are implementation
4609 dependent, since an implementation is free to choose either of the multiples
4610 that surround the value. GNAT always chooses the closer one, but this is not
4611 required behavior, and it is better to specify a value that is an exact
4612 multiple, ensuring predictable execution. The default is that such warnings
4616 @emph{Suppress warnings on bad fixed values.}
4617 @cindex @option{-gnatwB} (@command{gcc})
4618 This switch suppresses warnings for static fixed-point expressions whose
4619 value is not an exact multiple of Small.
4622 @emph{Activate warnings on conditionals.}
4623 @cindex @option{-gnatwc} (@command{gcc})
4624 @cindex Conditionals, constant
4625 This switch activates warnings for conditional expressions used in
4626 tests that are known to be True or False at compile time. The default
4627 is that such warnings are not generated.
4628 Note that this warning does
4629 not get issued for the use of boolean variables or constants whose
4630 values are known at compile time, since this is a standard technique
4631 for conditional compilation in Ada, and this would generate too many
4632 ``false positive'' warnings.
4633 This warning can also be turned on using @option{-gnatwa}.
4636 @emph{Suppress warnings on conditionals.}
4637 @cindex @option{-gnatwC} (@command{gcc})
4638 This switch suppresses warnings for conditional expressions used in
4639 tests that are known to be True or False at compile time.
4642 @emph{Activate warnings on implicit dereferencing.}
4643 @cindex @option{-gnatwd} (@command{gcc})
4644 If this switch is set, then the use of a prefix of an access type
4645 in an indexed component, slice, or selected component without an
4646 explicit @code{.all} will generate a warning. With this warning
4647 enabled, access checks occur only at points where an explicit
4648 @code{.all} appears in the source code (assuming no warnings are
4649 generated as a result of this switch). The default is that such
4650 warnings are not generated.
4651 Note that @option{-gnatwa} does not affect the setting of
4652 this warning option.
4655 @emph{Suppress warnings on implicit dereferencing.}
4656 @cindex @option{-gnatwD} (@command{gcc})
4657 @cindex Implicit dereferencing
4658 @cindex Dereferencing, implicit
4659 This switch suppresses warnings for implicit dereferences in
4660 indexed components, slices, and selected components.
4663 @emph{Treat warnings as errors.}
4664 @cindex @option{-gnatwe} (@command{gcc})
4665 @cindex Warnings, treat as error
4666 This switch causes warning messages to be treated as errors.
4667 The warning string still appears, but the warning messages are counted
4668 as errors, and prevent the generation of an object file.
4671 @emph{Activate warnings on unreferenced formals.}
4672 @cindex @option{-gnatwf} (@command{gcc})
4673 @cindex Formals, unreferenced
4674 This switch causes a warning to be generated if a formal parameter
4675 is not referenced in the body of the subprogram. This warning can
4676 also be turned on using @option{-gnatwa} or @option{-gnatwu}.
4679 @emph{Suppress warnings on unreferenced formals.}
4680 @cindex @option{-gnatwF} (@command{gcc})
4681 This switch suppresses warnings for unreferenced formal
4682 parameters. Note that the
4683 combination @option{-gnatwu} followed by @option{-gnatwF} has the
4684 effect of warning on unreferenced entities other than subprogram
4688 @emph{Activate warnings on unrecognized pragmas.}
4689 @cindex @option{-gnatwg} (@command{gcc})
4690 @cindex Pragmas, unrecognized
4691 This switch causes a warning to be generated if an unrecognized
4692 pragma is encountered. Apart from issuing this warning, the
4693 pragma is ignored and has no effect. This warning can
4694 also be turned on using @option{-gnatwa}. The default
4695 is that such warnings are issued (satisfying the Ada Reference
4696 Manual requirement that such warnings appear).
4699 @emph{Suppress warnings on unrecognized pragmas.}
4700 @cindex @option{-gnatwG} (@command{gcc})
4701 This switch suppresses warnings for unrecognized pragmas.
4704 @emph{Activate warnings on hiding.}
4705 @cindex @option{-gnatwh} (@command{gcc})
4706 @cindex Hiding of Declarations
4707 This switch activates warnings on hiding declarations.
4708 A declaration is considered hiding
4709 if it is for a non-overloadable entity, and it declares an entity with the
4710 same name as some other entity that is directly or use-visible. The default
4711 is that such warnings are not generated.
4712 Note that @option{-gnatwa} does not affect the setting of this warning option.
4715 @emph{Suppress warnings on hiding.}
4716 @cindex @option{-gnatwH} (@command{gcc})
4717 This switch suppresses warnings on hiding declarations.
4720 @emph{Activate warnings on implementation units.}
4721 @cindex @option{-gnatwi} (@command{gcc})
4722 This switch activates warnings for a @code{with} of an internal GNAT
4723 implementation unit, defined as any unit from the @code{Ada},
4724 @code{Interfaces}, @code{GNAT},
4725 ^^@code{DEC},^ or @code{System}
4726 hierarchies that is not
4727 documented in either the Ada Reference Manual or the GNAT
4728 Programmer's Reference Manual. Such units are intended only
4729 for internal implementation purposes and should not be @code{with}'ed
4730 by user programs. The default is that such warnings are generated
4731 This warning can also be turned on using @option{-gnatwa}.
4734 @emph{Disable warnings on implementation units.}
4735 @cindex @option{-gnatwI} (@command{gcc})
4736 This switch disables warnings for a @code{with} of an internal GNAT
4737 implementation unit.
4740 @emph{Activate warnings on obsolescent features (Annex J).}
4741 @cindex @option{-gnatwj} (@command{gcc})
4742 @cindex Features, obsolescent
4743 @cindex Obsolescent features
4744 If this warning option is activated, then warnings are generated for
4745 calls to subprograms marked with @code{pragma Obsolescent} and
4746 for use of features in Annex J of the Ada Reference Manual. In the
4747 case of Annex J, not all features are flagged. In particular use
4748 of the renamed packages (like @code{Text_IO}) and use of package
4749 @code{ASCII} are not flagged, since these are very common and
4750 would generate many annoying positive warnings. The default is that
4751 such warnings are not generated.
4753 In addition to the above cases, warnings are also generated for
4754 GNAT features that have been provided in past versions but which
4755 have been superseded (typically by features in the new Ada standard).
4756 For example, @code{pragma Ravenscar} will be flagged since its
4757 function is replaced by @code{pragma Profile(Ravenscar)}.
4759 Note that this warning option functions differently from the
4760 restriction @code{No_Obsolescent_Features} in two respects.
4761 First, the restriction applies only to annex J features.
4762 Second, the restriction does flag uses of package @code{ASCII}.
4765 @emph{Suppress warnings on obsolescent features (Annex J).}
4766 @cindex @option{-gnatwJ} (@command{gcc})
4767 This switch disables warnings on use of obsolescent features.
4770 @emph{Activate warnings on variables that could be constants.}
4771 @cindex @option{-gnatwk} (@command{gcc})
4772 This switch activates warnings for variables that are initialized but
4773 never modified, and then could be declared constants.
4776 @emph{Suppress warnings on variables that could be constants.}
4777 @cindex @option{-gnatwK} (@command{gcc})
4778 This switch disables warnings on variables that could be declared constants.
4781 @emph{Activate warnings for missing elaboration pragmas.}
4782 @cindex @option{-gnatwl} (@command{gcc})
4783 @cindex Elaboration, warnings
4784 This switch activates warnings on missing
4785 @code{pragma Elaborate_All} statements.
4786 See the section in this guide on elaboration checking for details on
4787 when such pragma should be used. Warnings are also generated if you
4788 are using the static mode of elaboration, and a @code{pragma Elaborate}
4789 is encountered. The default is that such warnings
4791 This warning is not automatically turned on by the use of @option{-gnatwa}.
4794 @emph{Suppress warnings for missing elaboration pragmas.}
4795 @cindex @option{-gnatwL} (@command{gcc})
4796 This switch suppresses warnings on missing pragma Elaborate_All statements.
4797 See the section in this guide on elaboration checking for details on
4798 when such pragma should be used.
4801 @emph{Activate warnings on modified but unreferenced variables.}
4802 @cindex @option{-gnatwm} (@command{gcc})
4803 This switch activates warnings for variables that are assigned (using
4804 an initialization value or with one or more assignment statements) but
4805 whose value is never read. The warning is suppressed for volatile
4806 variables and also for variables that are renamings of other variables
4807 or for which an address clause is given.
4808 This warning can also be turned on using @option{-gnatwa}.
4811 @emph{Disable warnings on modified but unreferenced variables.}
4812 @cindex @option{-gnatwM} (@command{gcc})
4813 This switch disables warnings for variables that are assigned or
4814 initialized, but never read.
4817 @emph{Set normal warnings mode.}
4818 @cindex @option{-gnatwn} (@command{gcc})
4819 This switch sets normal warning mode, in which enabled warnings are
4820 issued and treated as warnings rather than errors. This is the default
4821 mode. the switch @option{-gnatwn} can be used to cancel the effect of
4822 an explicit @option{-gnatws} or
4823 @option{-gnatwe}. It also cancels the effect of the
4824 implicit @option{-gnatwe} that is activated by the
4825 use of @option{-gnatg}.
4828 @emph{Activate warnings on address clause overlays.}
4829 @cindex @option{-gnatwo} (@command{gcc})
4830 @cindex Address Clauses, warnings
4831 This switch activates warnings for possibly unintended initialization
4832 effects of defining address clauses that cause one variable to overlap
4833 another. The default is that such warnings are generated.
4834 This warning can also be turned on using @option{-gnatwa}.
4837 @emph{Suppress warnings on address clause overlays.}
4838 @cindex @option{-gnatwO} (@command{gcc})
4839 This switch suppresses warnings on possibly unintended initialization
4840 effects of defining address clauses that cause one variable to overlap
4844 @emph{Activate warnings on ineffective pragma Inlines.}
4845 @cindex @option{-gnatwp} (@command{gcc})
4846 @cindex Inlining, warnings
4847 This switch activates warnings for failure of front end inlining
4848 (activated by @option{-gnatN}) to inline a particular call. There are
4849 many reasons for not being able to inline a call, including most
4850 commonly that the call is too complex to inline.
4851 This warning can also be turned on using @option{-gnatwa}.
4854 @emph{Suppress warnings on ineffective pragma Inlines.}
4855 @cindex @option{-gnatwP} (@command{gcc})
4856 This switch suppresses warnings on ineffective pragma Inlines. If the
4857 inlining mechanism cannot inline a call, it will simply ignore the
4861 @emph{Activate warnings on redundant constructs.}
4862 @cindex @option{-gnatwr} (@command{gcc})
4863 This switch activates warnings for redundant constructs. The following
4864 is the current list of constructs regarded as redundant:
4865 This warning can also be turned on using @option{-gnatwa}.
4869 Assignment of an item to itself.
4871 Type conversion that converts an expression to its own type.
4873 Use of the attribute @code{Base} where @code{typ'Base} is the same
4876 Use of pragma @code{Pack} when all components are placed by a record
4877 representation clause.
4879 Exception handler containing only a reraise statement (raise with no
4880 operand) which has no effect.
4882 Use of the operator abs on an operand that is known at compile time
4885 Comparison of boolean expressions to an explicit True value.
4889 @emph{Suppress warnings on redundant constructs.}
4890 @cindex @option{-gnatwR} (@command{gcc})
4891 This switch suppresses warnings for redundant constructs.
4894 @emph{Suppress all warnings.}
4895 @cindex @option{-gnatws} (@command{gcc})
4896 This switch completely suppresses the
4897 output of all warning messages from the GNAT front end.
4898 Note that it does not suppress warnings from the @command{gcc} back end.
4899 To suppress these back end warnings as well, use the switch @option{-w}
4900 in addition to @option{-gnatws}.
4903 @emph{Activate warnings on unused entities.}
4904 @cindex @option{-gnatwu} (@command{gcc})
4905 This switch activates warnings to be generated for entities that
4906 are declared but not referenced, and for units that are @code{with}'ed
4908 referenced. In the case of packages, a warning is also generated if
4909 no entities in the package are referenced. This means that if the package
4910 is referenced but the only references are in @code{use}
4911 clauses or @code{renames}
4912 declarations, a warning is still generated. A warning is also generated
4913 for a generic package that is @code{with}'ed but never instantiated.
4914 In the case where a package or subprogram body is compiled, and there
4915 is a @code{with} on the corresponding spec
4916 that is only referenced in the body,
4917 a warning is also generated, noting that the
4918 @code{with} can be moved to the body. The default is that
4919 such warnings are not generated.
4920 This switch also activates warnings on unreferenced formals
4921 (it includes the effect of @option{-gnatwf}).
4922 This warning can also be turned on using @option{-gnatwa}.
4925 @emph{Suppress warnings on unused entities.}
4926 @cindex @option{-gnatwU} (@command{gcc})
4927 This switch suppresses warnings for unused entities and packages.
4928 It also turns off warnings on unreferenced formals (and thus includes
4929 the effect of @option{-gnatwF}).
4932 @emph{Activate warnings on unassigned variables.}
4933 @cindex @option{-gnatwv} (@command{gcc})
4934 @cindex Unassigned variable warnings
4935 This switch activates warnings for access to variables which
4936 may not be properly initialized. The default is that
4937 such warnings are generated.
4940 @emph{Suppress warnings on unassigned variables.}
4941 @cindex @option{-gnatwV} (@command{gcc})
4942 This switch suppresses warnings for access to variables which
4943 may not be properly initialized.
4946 @emph{Activate warnings for Ada 2005 compatibility issues.}
4947 @cindex @option{-gnatwy} (@command{gcc})
4948 @cindex Ada 2005 compatibility issues warnings
4949 For the most part Ada 2005 is upwards compatible with Ada 95,
4950 but there are some exceptions (for example the fact that
4951 @code{interface} is now a reserved word in Ada 2005. This
4952 switch activates several warnings to help in identifying
4953 and correcting such incompatibilities. The default is that
4954 these warnings are generated. Note that at one point Ada 2005
4955 was called Ada 0Y, hence the choice of character.
4958 @emph{Disab le warnings for Ada 2005 compatibility issues.}
4959 @cindex @option{-gnatwY} (@command{gcc})
4960 @cindex Ada 2005 compatibility issues warnings
4961 This switch suppresses several warnings intended to help in identifying
4962 incompatibilities between Ada 95 and Ada 2005.
4965 @emph{Activate warnings on Export/Import pragmas.}
4966 @cindex @option{-gnatwx} (@command{gcc})
4967 @cindex Export/Import pragma warnings
4968 This switch activates warnings on Export/Import pragmas when
4969 the compiler detects a possible conflict between the Ada and
4970 foreign language calling sequences. For example, the use of
4971 default parameters in a convention C procedure is dubious
4972 because the C compiler cannot supply the proper default, so
4973 a warning is issued. The default is that such warnings are
4977 @emph{Suppress warnings on Export/Import pragmas.}
4978 @cindex @option{-gnatwX} (@command{gcc})
4979 This switch suppresses warnings on Export/Import pragmas.
4980 The sense of this is that you are telling the compiler that
4981 you know what you are doing in writing the pragma, and it
4982 should not complain at you.
4985 @emph{Activate warnings on unchecked conversions.}
4986 @cindex @option{-gnatwz} (@command{gcc})
4987 @cindex Unchecked_Conversion warnings
4988 This switch activates warnings for unchecked conversions
4989 where the types are known at compile time to have different
4991 is that such warnings are generated.
4994 @emph{Suppress warnings on unchecked conversions.}
4995 @cindex @option{-gnatwZ} (@command{gcc})
4996 This switch suppresses warnings for unchecked conversions
4997 where the types are known at compile time to have different
5000 @item ^-Wuninitialized^WARNINGS=UNINITIALIZED^
5001 @cindex @option{-Wuninitialized}
5002 The warnings controlled by the @option{-gnatw} switch are generated by the
5003 front end of the compiler. In some cases, the @option{^gcc^GCC^} back end
5004 can provide additional warnings. One such useful warning is provided by
5005 @option{^-Wuninitialized^WARNINGS=UNINITIALIZED^}. This must be used in
5006 conjunction with tunrning on optimization mode. This causes the flow
5007 analysis circuits of the back end optimizer to output additional
5008 warnings about uninitialized variables.
5010 @item ^-w^/NO_BACK_END_WARNINGS^
5012 This switch suppresses warnings from the @option{^gcc^GCC^} back end. It may
5013 be used in conjunction with @option{-gnatws} to ensure that all warnings
5014 are suppressed during the entire compilation process.
5020 A string of warning parameters can be used in the same parameter. For example:
5027 will turn on all optional warnings except for elaboration pragma warnings,
5028 and also specify that warnings should be treated as errors.
5030 When no switch @option{^-gnatw^/WARNINGS^} is used, this is equivalent to:
5055 @node Debugging and Assertion Control
5056 @subsection Debugging and Assertion Control
5060 @cindex @option{-gnata} (@command{gcc})
5066 The pragmas @code{Assert} and @code{Debug} normally have no effect and
5067 are ignored. This switch, where @samp{a} stands for assert, causes
5068 @code{Assert} and @code{Debug} pragmas to be activated.
5070 The pragmas have the form:
5074 @b{pragma} Assert (@var{Boolean-expression} [,
5075 @var{static-string-expression}])
5076 @b{pragma} Debug (@var{procedure call})
5081 The @code{Assert} pragma causes @var{Boolean-expression} to be tested.
5082 If the result is @code{True}, the pragma has no effect (other than
5083 possible side effects from evaluating the expression). If the result is
5084 @code{False}, the exception @code{Assert_Failure} declared in the package
5085 @code{System.Assertions} is
5086 raised (passing @var{static-string-expression}, if present, as the
5087 message associated with the exception). If no string expression is
5088 given the default is a string giving the file name and line number
5091 The @code{Debug} pragma causes @var{procedure} to be called. Note that
5092 @code{pragma Debug} may appear within a declaration sequence, allowing
5093 debugging procedures to be called between declarations.
5096 @item /DEBUG[=debug-level]
5098 Specifies how much debugging information is to be included in
5099 the resulting object file where 'debug-level' is one of the following:
5102 Include both debugger symbol records and traceback
5104 This is the default setting.
5106 Include both debugger symbol records and traceback in
5109 Excludes both debugger symbol records and traceback
5110 the object file. Same as /NODEBUG.
5112 Includes only debugger symbol records in the object
5113 file. Note that this doesn't include traceback information.
5118 @node Validity Checking
5119 @subsection Validity Checking
5120 @findex Validity Checking
5123 The Ada 95 Reference Manual has specific requirements for checking
5124 for invalid values. In particular, RM 13.9.1 requires that the
5125 evaluation of invalid values (for example from unchecked conversions),
5126 not result in erroneous execution. In GNAT, the result of such an
5127 evaluation in normal default mode is to either use the value
5128 unmodified, or to raise Constraint_Error in those cases where use
5129 of the unmodified value would cause erroneous execution. The cases
5130 where unmodified values might lead to erroneous execution are case
5131 statements (where a wild jump might result from an invalid value),
5132 and subscripts on the left hand side (where memory corruption could
5133 occur as a result of an invalid value).
5135 The @option{-gnatV^@var{x}^^} switch allows more control over the validity
5138 The @code{x} argument is a string of letters that
5139 indicate validity checks that are performed or not performed in addition
5140 to the default checks described above.
5143 The options allowed for this qualifier
5144 indicate validity checks that are performed or not performed in addition
5145 to the default checks described above.
5151 @emph{All validity checks.}
5152 @cindex @option{-gnatVa} (@command{gcc})
5153 All validity checks are turned on.
5155 That is, @option{-gnatVa} is
5156 equivalent to @option{gnatVcdfimorst}.
5160 @emph{Validity checks for copies.}
5161 @cindex @option{-gnatVc} (@command{gcc})
5162 The right hand side of assignments, and the initializing values of
5163 object declarations are validity checked.
5166 @emph{Default (RM) validity checks.}
5167 @cindex @option{-gnatVd} (@command{gcc})
5168 Some validity checks are done by default following normal Ada semantics
5170 A check is done in case statements that the expression is within the range
5171 of the subtype. If it is not, Constraint_Error is raised.
5172 For assignments to array components, a check is done that the expression used
5173 as index is within the range. If it is not, Constraint_Error is raised.
5174 Both these validity checks may be turned off using switch @option{-gnatVD}.
5175 They are turned on by default. If @option{-gnatVD} is specified, a subsequent
5176 switch @option{-gnatVd} will leave the checks turned on.
5177 Switch @option{-gnatVD} should be used only if you are sure that all such
5178 expressions have valid values. If you use this switch and invalid values
5179 are present, then the program is erroneous, and wild jumps or memory
5180 overwriting may occur.
5183 @emph{Validity checks for floating-point values.}
5184 @cindex @option{-gnatVf} (@command{gcc})
5185 In the absence of this switch, validity checking occurs only for discrete
5186 values. If @option{-gnatVf} is specified, then validity checking also applies
5187 for floating-point values, and NaN's and infinities are considered invalid,
5188 as well as out of range values for constrained types. Note that this means
5189 that standard @code{IEEE} infinity mode is not allowed. The exact contexts
5190 in which floating-point values are checked depends on the setting of other
5191 options. For example,
5192 @option{^-gnatVif^VALIDITY_CHECKING=(IN_PARAMS,FLOATS)^} or
5193 @option{^-gnatVfi^VALIDITY_CHECKING=(FLOATS,IN_PARAMS)^}
5194 (the order does not matter) specifies that floating-point parameters of mode
5195 @code{in} should be validity checked.
5198 @emph{Validity checks for @code{in} mode parameters}
5199 @cindex @option{-gnatVi} (@command{gcc})
5200 Arguments for parameters of mode @code{in} are validity checked in function
5201 and procedure calls at the point of call.
5204 @emph{Validity checks for @code{in out} mode parameters.}
5205 @cindex @option{-gnatVm} (@command{gcc})
5206 Arguments for parameters of mode @code{in out} are validity checked in
5207 procedure calls at the point of call. The @code{'m'} here stands for
5208 modify, since this concerns parameters that can be modified by the call.
5209 Note that there is no specific option to test @code{out} parameters,
5210 but any reference within the subprogram will be tested in the usual
5211 manner, and if an invalid value is copied back, any reference to it
5212 will be subject to validity checking.
5215 @emph{No validity checks.}
5216 @cindex @option{-gnatVn} (@command{gcc})
5217 This switch turns off all validity checking, including the default checking
5218 for case statements and left hand side subscripts. Note that the use of
5219 the switch @option{-gnatp} suppresses all run-time checks, including
5220 validity checks, and thus implies @option{-gnatVn}. When this switch
5221 is used, it cancels any other @option{-gnatV} previously issued.
5224 @emph{Validity checks for operator and attribute operands.}
5225 @cindex @option{-gnatVo} (@command{gcc})
5226 Arguments for predefined operators and attributes are validity checked.
5227 This includes all operators in package @code{Standard},
5228 the shift operators defined as intrinsic in package @code{Interfaces}
5229 and operands for attributes such as @code{Pos}. Checks are also made
5230 on individual component values for composite comparisons.
5233 @emph{Validity checks for parameters.}
5234 @cindex @option{-gnatVp} (@command{gcc})
5235 This controls the treatment of parameters within a subprogram (as opposed
5236 to @option{-gnatVi} and @option{-gnatVm} which control validity testing
5237 of parameters on a call. If either of these call options is used, then
5238 normally an assumption is made within a subprogram that the input arguments
5239 have been validity checking at the point of call, and do not need checking
5240 again within a subprogram). If @option{-gnatVp} is set, then this assumption
5241 is not made, and parameters are not assumed to be valid, so their validity
5242 will be checked (or rechecked) within the subprogram.
5245 @emph{Validity checks for function returns.}
5246 @cindex @option{-gnatVr} (@command{gcc})
5247 The expression in @code{return} statements in functions is validity
5251 @emph{Validity checks for subscripts.}
5252 @cindex @option{-gnatVs} (@command{gcc})
5253 All subscripts expressions are checked for validity, whether they appear
5254 on the right side or left side (in default mode only left side subscripts
5255 are validity checked).
5258 @emph{Validity checks for tests.}
5259 @cindex @option{-gnatVt} (@command{gcc})
5260 Expressions used as conditions in @code{if}, @code{while} or @code{exit}
5261 statements are checked, as well as guard expressions in entry calls.
5266 The @option{-gnatV} switch may be followed by
5267 ^a string of letters^a list of options^
5268 to turn on a series of validity checking options.
5270 @option{^-gnatVcr^/VALIDITY_CHECKING=(COPIES, RETURNS)^}
5271 specifies that in addition to the default validity checking, copies and
5272 function return expressions are to be validity checked.
5273 In order to make it easier
5274 to specify the desired combination of effects,
5276 the upper case letters @code{CDFIMORST} may
5277 be used to turn off the corresponding lower case option.
5280 the prefix @code{NO} on an option turns off the corresponding validity
5283 @item @code{NOCOPIES}
5284 @item @code{NODEFAULT}
5285 @item @code{NOFLOATS}
5286 @item @code{NOIN_PARAMS}
5287 @item @code{NOMOD_PARAMS}
5288 @item @code{NOOPERANDS}
5289 @item @code{NORETURNS}
5290 @item @code{NOSUBSCRIPTS}
5291 @item @code{NOTESTS}
5295 @option{^-gnatVaM^/VALIDITY_CHECKING=(ALL, NOMOD_PARAMS)^}
5296 turns on all validity checking options except for
5297 checking of @code{@b{in out}} procedure arguments.
5299 The specification of additional validity checking generates extra code (and
5300 in the case of @option{-gnatVa} the code expansion can be substantial.
5301 However, these additional checks can be very useful in detecting
5302 uninitialized variables, incorrect use of unchecked conversion, and other
5303 errors leading to invalid values. The use of pragma @code{Initialize_Scalars}
5304 is useful in conjunction with the extra validity checking, since this
5305 ensures that wherever possible uninitialized variables have invalid values.
5307 See also the pragma @code{Validity_Checks} which allows modification of
5308 the validity checking mode at the program source level, and also allows for
5309 temporary disabling of validity checks.
5311 @node Style Checking
5312 @subsection Style Checking
5313 @findex Style checking
5316 The @option{-gnaty^x^(option,option,...)^} switch
5317 @cindex @option{-gnaty} (@command{gcc})
5318 causes the compiler to
5319 enforce specified style rules. A limited set of style rules has been used
5320 in writing the GNAT sources themselves. This switch allows user programs
5321 to activate all or some of these checks. If the source program fails a
5322 specified style check, an appropriate warning message is given, preceded by
5323 the character sequence ``(style)''.
5325 @code{(option,option,...)} is a sequence of keywords
5328 The string @var{x} is a sequence of letters or digits
5330 indicating the particular style
5331 checks to be performed. The following checks are defined:
5336 @emph{Specify indentation level.}
5337 If a digit from 1-9 appears
5338 ^in the string after @option{-gnaty}^as an option for /STYLE_CHECKS^
5339 then proper indentation is checked, with the digit indicating the
5340 indentation level required.
5341 The general style of required indentation is as specified by
5342 the examples in the Ada Reference Manual. Full line comments must be
5343 aligned with the @code{--} starting on a column that is a multiple of
5344 the alignment level.
5347 @emph{Check attribute casing.}
5348 If the ^letter a^word ATTRIBUTE^ appears in the string after @option{-gnaty}
5349 then attribute names, including the case of keywords such as @code{digits}
5350 used as attributes names, must be written in mixed case, that is, the
5351 initial letter and any letter following an underscore must be uppercase.
5352 All other letters must be lowercase.
5355 @emph{Blanks not allowed at statement end.}
5356 If the ^letter b^word BLANKS^ appears in the string after @option{-gnaty} then
5357 trailing blanks are not allowed at the end of statements. The purpose of this
5358 rule, together with h (no horizontal tabs), is to enforce a canonical format
5359 for the use of blanks to separate source tokens.
5362 @emph{Check comments.}
5363 If the ^letter c^word COMMENTS^ appears in the string after @option{-gnaty}
5364 then comments must meet the following set of rules:
5369 The ``@code{--}'' that starts the column must either start in column one,
5370 or else at least one blank must precede this sequence.
5373 Comments that follow other tokens on a line must have at least one blank
5374 following the ``@code{--}'' at the start of the comment.
5377 Full line comments must have two blanks following the ``@code{--}'' that
5378 starts the comment, with the following exceptions.
5381 A line consisting only of the ``@code{--}'' characters, possibly preceded
5382 by blanks is permitted.
5385 A comment starting with ``@code{--x}'' where @code{x} is a special character
5387 This allows proper processing of the output generated by specialized tools
5388 including @command{gnatprep} (where ``@code{--!}'' is used) and the SPARK
5390 language (where ``@code{--#}'' is used). For the purposes of this rule, a
5391 special character is defined as being in one of the ASCII ranges
5392 @code{16#21#..16#2F#} or @code{16#3A#..16#3F#}.
5393 Note that this usage is not permitted
5394 in GNAT implementation units (i.e. when @option{-gnatg} is used).
5397 A line consisting entirely of minus signs, possibly preceded by blanks, is
5398 permitted. This allows the construction of box comments where lines of minus
5399 signs are used to form the top and bottom of the box.
5402 A comment that starts and ends with ``@code{--}'' is permitted as long as at
5403 least one blank follows the initial ``@code{--}''. Together with the preceding
5404 rule, this allows the construction of box comments, as shown in the following
5407 ---------------------------
5408 -- This is a box comment --
5409 -- with two text lines. --
5410 ---------------------------
5414 @item ^d^DOS_LINE_ENDINGS^
5415 @emph{Check no DOS line terminators present.}
5416 If the ^letter d^word DOS_LINE_ENDINGS^ appears in the string after
5417 @option{-gnaty} then all lines must be terminated by a single ASCII.LF
5418 character (in particular the DOS line terminator sequence CR/LF is not
5422 @emph{Check end/exit labels.}
5423 If the ^letter e^word END^ appears in the string after @option{-gnaty} then
5424 optional labels on @code{end} statements ending subprograms and on
5425 @code{exit} statements exiting named loops, are required to be present.
5428 @emph{No form feeds or vertical tabs.}
5429 If the ^letter f^word VTABS^ appears in the string after @option{-gnaty} then
5430 neither form feeds nor vertical tab characters are permitted
5434 @emph{No horizontal tabs.}
5435 If the ^letter h^word HTABS^ appears in the string after @option{-gnaty} then
5436 horizontal tab characters are not permitted in the source text.
5437 Together with the b (no blanks at end of line) check, this
5438 enforces a canonical form for the use of blanks to separate
5442 @emph{Check if-then layout.}
5443 If the ^letter i^word IF_THEN^ appears in the string after @option{-gnaty},
5444 then the keyword @code{then} must appear either on the same
5445 line as corresponding @code{if}, or on a line on its own, lined
5446 up under the @code{if} with at least one non-blank line in between
5447 containing all or part of the condition to be tested.
5450 @emph{Check keyword casing.}
5451 If the ^letter k^word KEYWORD^ appears in the string after @option{-gnaty} then
5452 all keywords must be in lower case (with the exception of keywords
5453 such as @code{digits} used as attribute names to which this check
5457 @emph{Check layout.}
5458 If the ^letter l^word LAYOUT^ appears in the string after @option{-gnaty} then
5459 layout of statement and declaration constructs must follow the
5460 recommendations in the Ada Reference Manual, as indicated by the
5461 form of the syntax rules. For example an @code{else} keyword must
5462 be lined up with the corresponding @code{if} keyword.
5464 There are two respects in which the style rule enforced by this check
5465 option are more liberal than those in the Ada Reference Manual. First
5466 in the case of record declarations, it is permissible to put the
5467 @code{record} keyword on the same line as the @code{type} keyword, and
5468 then the @code{end} in @code{end record} must line up under @code{type}.
5469 For example, either of the following two layouts is acceptable:
5471 @smallexample @c ada
5487 Second, in the case of a block statement, a permitted alternative
5488 is to put the block label on the same line as the @code{declare} or
5489 @code{begin} keyword, and then line the @code{end} keyword up under
5490 the block label. For example both the following are permitted:
5492 @smallexample @c ada
5510 The same alternative format is allowed for loops. For example, both of
5511 the following are permitted:
5513 @smallexample @c ada
5515 Clear : while J < 10 loop
5526 @item ^Lnnn^MAX_NESTING=nnn^
5527 @emph{Set maximum nesting level}
5528 If the sequence ^Lnnn^MAX_NESTING=nnn^, where nnn is a decimal number in
5529 the range 0-999, appears in the string after @option{-gnaty} then the
5530 maximum level of nesting of constructs (including subprograms, loops,
5531 blocks, packages, and conditionals) may not exceed the given value. A
5532 value of zero disconnects this style check.
5534 @item ^m^LINE_LENGTH^
5535 @emph{Check maximum line length.}
5536 If the ^letter m^word LINE_LENGTH^ appears in the string after @option{-gnaty}
5537 then the length of source lines must not exceed 79 characters, including
5538 any trailing blanks. The value of 79 allows convenient display on an
5539 80 character wide device or window, allowing for possible special
5540 treatment of 80 character lines. Note that this count is of raw
5541 characters in the source text. This means that a tab character counts
5542 as one character in this count and a wide character sequence counts as
5543 several characters (however many are needed in the encoding).
5545 @item ^Mnnn^MAX_LENGTH=nnn^
5546 @emph{Set maximum line length.}
5547 If the sequence ^M^MAX_LENGTH=^nnn, where nnn is a decimal number, appears in
5548 the string after @option{-gnaty} then the length of lines must not exceed the
5551 @item ^n^STANDARD_CASING^
5552 @emph{Check casing of entities in Standard.}
5553 If the ^letter n^word STANDARD_CASING^ appears in the string
5554 after @option{-gnaty} then any identifier from Standard must be cased
5555 to match the presentation in the Ada Reference Manual (for example,
5556 @code{Integer} and @code{ASCII.NUL}).
5558 @item ^o^ORDERED_SUBPROGRAMS^
5559 @emph{Check order of subprogram bodies.}
5560 If the ^letter o^word ORDERED_SUBPROGRAMS^ appears in the string
5561 after @option{-gnaty} then all subprogram bodies in a given scope
5562 (e.g. a package body) must be in alphabetical order. The ordering
5563 rule uses normal Ada rules for comparing strings, ignoring casing
5564 of letters, except that if there is a trailing numeric suffix, then
5565 the value of this suffix is used in the ordering (e.g. Junk2 comes
5569 @emph{Check pragma casing.}
5570 If the ^letter p^word PRAGMA^ appears in the string after @option{-gnaty} then
5571 pragma names must be written in mixed case, that is, the
5572 initial letter and any letter following an underscore must be uppercase.
5573 All other letters must be lowercase.
5575 @item ^r^REFERENCES^
5576 @emph{Check references.}
5577 If the ^letter r^word REFERENCES^ appears in the string after @option{-gnaty}
5578 then all identifier references must be cased in the same way as the
5579 corresponding declaration. No specific casing style is imposed on
5580 identifiers. The only requirement is for consistency of references
5584 @emph{Check separate specs.}
5585 If the ^letter s^word SPECS^ appears in the string after @option{-gnaty} then
5586 separate declarations (``specs'') are required for subprograms (a
5587 body is not allowed to serve as its own declaration). The only
5588 exception is that parameterless library level procedures are
5589 not required to have a separate declaration. This exception covers
5590 the most frequent form of main program procedures.
5593 @emph{Check token spacing.}
5594 If the ^letter t^word TOKEN^ appears in the string after @option{-gnaty} then
5595 the following token spacing rules are enforced:
5600 The keywords @code{@b{abs}} and @code{@b{not}} must be followed by a space.
5603 The token @code{=>} must be surrounded by spaces.
5606 The token @code{<>} must be preceded by a space or a left parenthesis.
5609 Binary operators other than @code{**} must be surrounded by spaces.
5610 There is no restriction on the layout of the @code{**} binary operator.
5613 Colon must be surrounded by spaces.
5616 Colon-equal (assignment, initialization) must be surrounded by spaces.
5619 Comma must be the first non-blank character on the line, or be
5620 immediately preceded by a non-blank character, and must be followed
5624 If the token preceding a left parenthesis ends with a letter or digit, then
5625 a space must separate the two tokens.
5628 A right parenthesis must either be the first non-blank character on
5629 a line, or it must be preceded by a non-blank character.
5632 A semicolon must not be preceded by a space, and must not be followed by
5633 a non-blank character.
5636 A unary plus or minus may not be followed by a space.
5639 A vertical bar must be surrounded by spaces.
5642 @item ^u^UNNECESSARY_BLANK_LINES^
5643 @emph{Check unnecessary blank lines.}
5644 Check for unnecessary blank lines. A blank line is considered
5645 unnecessary if it appears at the end of the file, or if more than
5646 one blank line occurs in sequence.
5648 @item ^x^XTRA_PARENS^
5649 @emph{Check extra parentheses.}
5650 Check for the use of an unnecessary extra level of parentheses (C-style)
5651 around conditions in @code{if} statements, @code{while} statements and
5652 @code{exit} statements.
5657 In the above rules, appearing in column one is always permitted, that is,
5658 counts as meeting either a requirement for a required preceding space,
5659 or as meeting a requirement for no preceding space.
5661 Appearing at the end of a line is also always permitted, that is, counts
5662 as meeting either a requirement for a following space, or as meeting
5663 a requirement for no following space.
5666 If any of these style rules is violated, a message is generated giving
5667 details on the violation. The initial characters of such messages are
5668 always ``@code{(style)}''. Note that these messages are treated as warning
5669 messages, so they normally do not prevent the generation of an object
5670 file. The @option{-gnatwe} switch can be used to treat warning messages,
5671 including style messages, as fatal errors.
5675 @option{-gnaty} on its own (that is not
5676 followed by any letters or digits),
5677 is equivalent to @code{gnaty3abcefhiklmnprst}, that is all checking
5678 options enabled with the exception of @option{-gnatyo},
5679 @option{-gnatyd}, @option{-gnatyu}, and @option{-gnatyx}.
5682 /STYLE_CHECKS=ALL_BUILTIN enables all checking options with
5683 the exception of ORDERED_SUBPROGRAMS, UNNECESSARY_BLANK_LINES,
5684 XTRA_PARENS, and DOS_LINE_ENDINGS. In addition
5686 an indentation level of 3 is set. This is similar to the standard
5687 checking option that is used for the GNAT sources.
5696 clears any previously set style checks.
5698 @node Run-Time Checks
5699 @subsection Run-Time Checks
5700 @cindex Division by zero
5701 @cindex Access before elaboration
5702 @cindex Checks, division by zero
5703 @cindex Checks, access before elaboration
5706 If you compile with the default options, GNAT will insert many run-time
5707 checks into the compiled code, including code that performs range
5708 checking against constraints, but not arithmetic overflow checking for
5709 integer operations (including division by zero) or checks for access
5710 before elaboration on subprogram calls. All other run-time checks, as
5711 required by the Ada 95 Reference Manual, are generated by default.
5712 The following @command{gcc} switches refine this default behavior:
5717 @cindex @option{-gnatp} (@command{gcc})
5718 @cindex Suppressing checks
5719 @cindex Checks, suppressing
5721 Suppress all run-time checks as though @code{pragma Suppress (all_checks})
5722 had been present in the source. Validity checks are also suppressed (in
5723 other words @option{-gnatp} also implies @option{-gnatVn}.
5724 Use this switch to improve the performance
5725 of the code at the expense of safety in the presence of invalid data or
5729 @cindex @option{-gnato} (@command{gcc})
5730 @cindex Overflow checks
5731 @cindex Check, overflow
5732 Enables overflow checking for integer operations.
5733 This causes GNAT to generate slower and larger executable
5734 programs by adding code to check for overflow (resulting in raising
5735 @code{Constraint_Error} as required by standard Ada
5736 semantics). These overflow checks correspond to situations in which
5737 the true value of the result of an operation may be outside the base
5738 range of the result type. The following example shows the distinction:
5740 @smallexample @c ada
5741 X1 : Integer := Integer'Last;
5742 X2 : Integer range 1 .. 5 := 5;
5743 X3 : Integer := Integer'Last;
5744 X4 : Integer range 1 .. 5 := 5;
5745 F : Float := 2.0E+20;
5754 Here the first addition results in a value that is outside the base range
5755 of Integer, and hence requires an overflow check for detection of the
5756 constraint error. Thus the first assignment to @code{X1} raises a
5757 @code{Constraint_Error} exception only if @option{-gnato} is set.
5759 The second increment operation results in a violation
5760 of the explicit range constraint, and such range checks are always
5761 performed (unless specifically suppressed with a pragma @code{suppress}
5762 or the use of @option{-gnatp}).
5764 The two conversions of @code{F} both result in values that are outside
5765 the base range of type @code{Integer} and thus will raise
5766 @code{Constraint_Error} exceptions only if @option{-gnato} is used.
5767 The fact that the result of the second conversion is assigned to
5768 variable @code{X4} with a restricted range is irrelevant, since the problem
5769 is in the conversion, not the assignment.
5771 Basically the rule is that in the default mode (@option{-gnato} not
5772 used), the generated code assures that all integer variables stay
5773 within their declared ranges, or within the base range if there is
5774 no declared range. This prevents any serious problems like indexes
5775 out of range for array operations.
5777 What is not checked in default mode is an overflow that results in
5778 an in-range, but incorrect value. In the above example, the assignments
5779 to @code{X1}, @code{X2}, @code{X3} all give results that are within the
5780 range of the target variable, but the result is wrong in the sense that
5781 it is too large to be represented correctly. Typically the assignment
5782 to @code{X1} will result in wrap around to the largest negative number.
5783 The conversions of @code{F} will result in some @code{Integer} value
5784 and if that integer value is out of the @code{X4} range then the
5785 subsequent assignment would generate an exception.
5787 @findex Machine_Overflows
5788 Note that the @option{-gnato} switch does not affect the code generated
5789 for any floating-point operations; it applies only to integer
5791 For floating-point, GNAT has the @code{Machine_Overflows}
5792 attribute set to @code{False} and the normal mode of operation is to
5793 generate IEEE NaN and infinite values on overflow or invalid operations
5794 (such as dividing 0.0 by 0.0).
5796 The reason that we distinguish overflow checking from other kinds of
5797 range constraint checking is that a failure of an overflow check can
5798 generate an incorrect value, but cannot cause erroneous behavior. This
5799 is unlike the situation with a constraint check on an array subscript,
5800 where failure to perform the check can result in random memory description,
5801 or the range check on a case statement, where failure to perform the check
5802 can cause a wild jump.
5804 Note again that @option{-gnato} is off by default, so overflow checking is
5805 not performed in default mode. This means that out of the box, with the
5806 default settings, GNAT does not do all the checks expected from the
5807 language description in the Ada Reference Manual. If you want all constraint
5808 checks to be performed, as described in this Manual, then you must
5809 explicitly use the -gnato switch either on the @command{gnatmake} or
5810 @command{gcc} command.
5813 @cindex @option{-gnatE} (@command{gcc})
5814 @cindex Elaboration checks
5815 @cindex Check, elaboration
5816 Enables dynamic checks for access-before-elaboration
5817 on subprogram calls and generic instantiations.
5818 For full details of the effect and use of this switch,
5819 @xref{Compiling Using gcc}.
5824 The setting of these switches only controls the default setting of the
5825 checks. You may modify them using either @code{Suppress} (to remove
5826 checks) or @code{Unsuppress} (to add back suppressed checks) pragmas in
5829 @node Stack Overflow Checking
5830 @subsection Stack Overflow Checking
5831 @cindex Stack Overflow Checking
5832 @cindex -fstack-check
5835 For most operating systems, @command{gcc} does not perform stack overflow
5836 checking by default. This means that if the main environment task or
5837 some other task exceeds the available stack space, then unpredictable
5838 behavior will occur.
5840 To activate stack checking, compile all units with the gcc option
5841 @option{-fstack-check}. For example:
5844 gcc -c -fstack-check package1.adb
5848 Units compiled with this option will generate extra instructions to check
5849 that any use of the stack (for procedure calls or for declaring local
5850 variables in declare blocks) do not exceed the available stack space.
5851 If the space is exceeded, then a @code{Storage_Error} exception is raised.
5853 For declared tasks, the stack size is always controlled by the size
5854 given in an applicable @code{Storage_Size} pragma (or is set to
5855 the default size if no pragma is used.
5857 For the environment task, the stack size depends on
5858 system defaults and is unknown to the compiler. The stack
5859 may even dynamically grow on some systems, precluding the
5860 normal Ada semantics for stack overflow. In the worst case,
5861 unbounded stack usage, causes unbounded stack expansion
5862 resulting in the system running out of virtual memory.
5864 The stack checking may still work correctly if a fixed
5865 size stack is allocated, but this cannot be guaranteed.
5866 To ensure that a clean exception is signalled for stack
5867 overflow, set the environment variable
5868 @code{GNAT_STACK_LIMIT} to indicate the maximum
5869 stack area that can be used, as in:
5870 @cindex GNAT_STACK_LIMIT
5873 SET GNAT_STACK_LIMIT 1600
5877 The limit is given in kilobytes, so the above declaration would
5878 set the stack limit of the environment task to 1.6 megabytes.
5879 Note that the only purpose of this usage is to limit the amount
5880 of stack used by the environment task. If it is necessary to
5881 increase the amount of stack for the environment task, then this
5882 is an operating systems issue, and must be addressed with the
5883 appropriate operating systems commands.
5885 @node Using gcc for Syntax Checking
5886 @subsection Using @command{gcc} for Syntax Checking
5889 @cindex @option{-gnats} (@command{gcc})
5893 The @code{s} stands for ``syntax''.
5896 Run GNAT in syntax checking only mode. For
5897 example, the command
5900 $ gcc -c -gnats x.adb
5904 compiles file @file{x.adb} in syntax-check-only mode. You can check a
5905 series of files in a single command
5907 , and can use wild cards to specify such a group of files.
5908 Note that you must specify the @option{-c} (compile
5909 only) flag in addition to the @option{-gnats} flag.
5912 You may use other switches in conjunction with @option{-gnats}. In
5913 particular, @option{-gnatl} and @option{-gnatv} are useful to control the
5914 format of any generated error messages.
5916 When the source file is empty or contains only empty lines and/or comments,
5917 the output is a warning:
5920 $ gcc -c -gnats -x ada toto.txt
5921 toto.txt:1:01: warning: empty file, contains no compilation units
5925 Otherwise, the output is simply the error messages, if any. No object file or
5926 ALI file is generated by a syntax-only compilation. Also, no units other
5927 than the one specified are accessed. For example, if a unit @code{X}
5928 @code{with}'s a unit @code{Y}, compiling unit @code{X} in syntax
5929 check only mode does not access the source file containing unit
5932 @cindex Multiple units, syntax checking
5933 Normally, GNAT allows only a single unit in a source file. However, this
5934 restriction does not apply in syntax-check-only mode, and it is possible
5935 to check a file containing multiple compilation units concatenated
5936 together. This is primarily used by the @code{gnatchop} utility
5937 (@pxref{Renaming Files Using gnatchop}).
5940 @node Using gcc for Semantic Checking
5941 @subsection Using @command{gcc} for Semantic Checking
5944 @cindex @option{-gnatc} (@command{gcc})
5948 The @code{c} stands for ``check''.
5950 Causes the compiler to operate in semantic check mode,
5951 with full checking for all illegalities specified in the
5952 Ada 95 Reference Manual, but without generation of any object code
5953 (no object file is generated).
5955 Because dependent files must be accessed, you must follow the GNAT
5956 semantic restrictions on file structuring to operate in this mode:
5960 The needed source files must be accessible
5961 (@pxref{Search Paths and the Run-Time Library (RTL)}).
5964 Each file must contain only one compilation unit.
5967 The file name and unit name must match (@pxref{File Naming Rules}).
5970 The output consists of error messages as appropriate. No object file is
5971 generated. An @file{ALI} file is generated for use in the context of
5972 cross-reference tools, but this file is marked as not being suitable
5973 for binding (since no object file is generated).
5974 The checking corresponds exactly to the notion of
5975 legality in the Ada 95 Reference Manual.
5977 Any unit can be compiled in semantics-checking-only mode, including
5978 units that would not normally be compiled (subunits,
5979 and specifications where a separate body is present).
5982 @node Compiling Different Versions of Ada
5983 @subsection Compiling Different Versions of Ada
5985 @cindex Compatibility with Ada 83
5988 @cindex Ada 2005 mode
5990 GNAT is primarily an Ada 95 compiler, but the switches described in
5991 this section allow operation in Ada 83 compatibility mode, and also
5992 allow the use of a preliminary implementation of many of the expected
5993 new features in Ada 2005, the forthcoming new version of the standard.
5995 @item -gnat83 (Ada 83 Compatibility Mode)
5996 @cindex @option{-gnat83} (@command{gcc})
5997 @cindex ACVC, Ada 83 tests
6000 Although GNAT is primarily an Ada 95 compiler, it accepts this switch to
6001 specify that an Ada 83 program is to be compiled in Ada 83 mode. If you specify
6002 this switch, GNAT rejects most Ada 95 extensions and applies Ada 83 semantics
6003 where this can be done easily.
6004 It is not possible to guarantee this switch does a perfect
6005 job; for example, some subtle tests, such as are
6006 found in earlier ACVC tests (and that have been removed from the ACATS suite
6007 for Ada 95), might not compile correctly.
6008 Nevertheless, this switch may be useful in some circumstances, for example
6009 where, due to contractual reasons, legacy code needs to be maintained
6010 using only Ada 83 features.
6012 With few exceptions (most notably the need to use @code{<>} on
6013 @cindex Generic formal parameters
6014 unconstrained generic formal parameters, the use of the new Ada 95
6015 reserved words, and the use of packages
6016 with optional bodies), it is not necessary to use the
6017 @option{-gnat83} switch when compiling Ada 83 programs, because, with rare
6018 exceptions, Ada 95 is upwardly compatible with Ada 83. This
6019 means that a correct Ada 83 program is usually also a correct Ada 95
6021 For further information, please refer to @ref{Compatibility and Porting Guide}.
6023 @item -gnat95 (Ada 95 mode)
6024 @cindex @option{-gnat95} (@command{gcc})
6027 GNAT is primarily an Ada 95 compiler, and all current releases of GNAT Pro
6028 compile in Ada 95 mode by default. Typically, Ada 95 is sufficiently upwards
6029 compatible with Ada 83, that legacy Ada 83 programs may be compiled using
6030 this default Ada95 mode without problems (see section above describing the
6031 use of @option{-gnat83} to run in Ada 83 mode).
6033 In Ada 95 mode, the use of Ada 2005 features will in general cause error
6034 messages or warnings. Some specialized releases of GNAT (notably the GAP
6035 academic version) operate in Ada 2005 mode by default (see section below
6036 describing the use of @option{-gnat05} to run in Ada 2005 mode). For such
6037 versions the @option{-gnat95} switch may be used to enforce Ada 95 mode.
6038 This option also can be used to cancel the effect of a previous
6039 @option{-gnat83} or @option{-gnat05} switch earlier in the command line.
6042 @item -gnat05 (Ada 2005 mode)
6043 @cindex @option{-gnat05} (@command{gcc})
6046 Although GNAT is primarily an Ada 95 compiler, it can be set to operate
6047 in Ada 2005 mode using this option. Although the new standard has not
6048 yet been issued (as of early 2005), many features have been discussed and
6049 approved in ``Ada Issues'' (AI's). For the text of these AI's, see
6050 @url{www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs}. Included with GNAT
6051 releases is a file @file{features-ada0y} that describes the current set
6052 of implemented Ada 2005 features.
6054 If these features are used in Ada 95 mode (which is the normal default),
6055 then error messages or warnings may be
6056 generated, reflecting the fact that these new features are otherwise
6057 unauthorized extensions to Ada 95. The use of the @option{-gnat05}
6058 switch (or an equivalent pragma) causes these messages to be suppressed.
6060 Note that some specialized releases of GNAT (notably the GAP academic
6061 version) have Ada 2005 mode on by default, and in such environments,
6062 the Ada 2005 features can be used freely without the use of switches.
6066 @node Character Set Control
6067 @subsection Character Set Control
6069 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
6070 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
6073 Normally GNAT recognizes the Latin-1 character set in source program
6074 identifiers, as described in the Ada 95 Reference Manual.
6076 GNAT to recognize alternate character sets in identifiers. @var{c} is a
6077 single character ^^or word^ indicating the character set, as follows:
6081 ISO 8859-1 (Latin-1) identifiers
6084 ISO 8859-2 (Latin-2) letters allowed in identifiers
6087 ISO 8859-3 (Latin-3) letters allowed in identifiers
6090 ISO 8859-4 (Latin-4) letters allowed in identifiers
6093 ISO 8859-5 (Cyrillic) letters allowed in identifiers
6096 ISO 8859-15 (Latin-9) letters allowed in identifiers
6099 IBM PC letters (code page 437) allowed in identifiers
6102 IBM PC letters (code page 850) allowed in identifiers
6104 @item ^f^FULL_UPPER^
6105 Full upper-half codes allowed in identifiers
6108 No upper-half codes allowed in identifiers
6111 Wide-character codes (that is, codes greater than 255)
6112 allowed in identifiers
6115 @xref{Foreign Language Representation}, for full details on the
6116 implementation of these character sets.
6118 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
6119 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
6120 Specify the method of encoding for wide characters.
6121 @var{e} is one of the following:
6126 Hex encoding (brackets coding also recognized)
6129 Upper half encoding (brackets encoding also recognized)
6132 Shift/JIS encoding (brackets encoding also recognized)
6135 EUC encoding (brackets encoding also recognized)
6138 UTF-8 encoding (brackets encoding also recognized)
6141 Brackets encoding only (default value)
6143 For full details on the these encoding
6144 methods see @ref{Wide Character Encodings}.
6145 Note that brackets coding is always accepted, even if one of the other
6146 options is specified, so for example @option{-gnatW8} specifies that both
6147 brackets and @code{UTF-8} encodings will be recognized. The units that are
6148 with'ed directly or indirectly will be scanned using the specified
6149 representation scheme, and so if one of the non-brackets scheme is
6150 used, it must be used consistently throughout the program. However,
6151 since brackets encoding is always recognized, it may be conveniently
6152 used in standard libraries, allowing these libraries to be used with
6153 any of the available coding schemes.
6154 scheme. If no @option{-gnatW?} parameter is present, then the default
6155 representation is Brackets encoding only.
6157 Note that the wide character representation that is specified (explicitly
6158 or by default) for the main program also acts as the default encoding used
6159 for Wide_Text_IO files if not specifically overridden by a WCEM form
6163 @node File Naming Control
6164 @subsection File Naming Control
6167 @item ^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{n}
6168 @cindex @option{-gnatk} (@command{gcc})
6169 Activates file name ``krunching''. @var{n}, a decimal integer in the range
6170 1-999, indicates the maximum allowable length of a file name (not
6171 including the @file{.ads} or @file{.adb} extension). The default is not
6172 to enable file name krunching.
6174 For the source file naming rules, @xref{File Naming Rules}.
6177 @node Subprogram Inlining Control
6178 @subsection Subprogram Inlining Control
6183 @cindex @option{-gnatn} (@command{gcc})
6185 The @code{n} here is intended to suggest the first syllable of the
6188 GNAT recognizes and processes @code{Inline} pragmas. However, for the
6189 inlining to actually occur, optimization must be enabled. To enable
6190 inlining of subprograms specified by pragma @code{Inline},
6191 you must also specify this switch.
6192 In the absence of this switch, GNAT does not attempt
6193 inlining and does not need to access the bodies of
6194 subprograms for which @code{pragma Inline} is specified if they are not
6195 in the current unit.
6197 If you specify this switch the compiler will access these bodies,
6198 creating an extra source dependency for the resulting object file, and
6199 where possible, the call will be inlined.
6200 For further details on when inlining is possible
6201 see @ref{Inlining of Subprograms}.
6204 @cindex @option{-gnatN} (@command{gcc})
6205 The front end inlining activated by this switch is generally more extensive,
6206 and quite often more effective than the standard @option{-gnatn} inlining mode.
6207 It will also generate additional dependencies.
6209 @option{-gnatN} automatically implies @option{-gnatn} so it is not necessary
6210 to specify both options.
6213 @node Auxiliary Output Control
6214 @subsection Auxiliary Output Control
6218 @cindex @option{-gnatt} (@command{gcc})
6219 @cindex Writing internal trees
6220 @cindex Internal trees, writing to file
6221 Causes GNAT to write the internal tree for a unit to a file (with the
6222 extension @file{.adt}.
6223 This not normally required, but is used by separate analysis tools.
6225 these tools do the necessary compilations automatically, so you should
6226 not have to specify this switch in normal operation.
6229 @cindex @option{-gnatu} (@command{gcc})
6230 Print a list of units required by this compilation on @file{stdout}.
6231 The listing includes all units on which the unit being compiled depends
6232 either directly or indirectly.
6235 @item -pass-exit-codes
6236 @cindex @option{-pass-exit-codes} (@command{gcc})
6237 If this switch is not used, the exit code returned by @command{gcc} when
6238 compiling multiple files indicates whether all source files have
6239 been successfully used to generate object files or not.
6241 When @option{-pass-exit-codes} is used, @command{gcc} exits with an extended
6242 exit status and allows an integrated development environment to better
6243 react to a compilation failure. Those exit status are:
6247 There was an error in at least one source file.
6249 At least one source file did not generate an object file.
6251 The compiler died unexpectedly (internal error for example).
6253 An object file has been generated for every source file.
6258 @node Debugging Control
6259 @subsection Debugging Control
6263 @cindex Debugging options
6266 @cindex @option{-gnatd} (@command{gcc})
6267 Activate internal debugging switches. @var{x} is a letter or digit, or
6268 string of letters or digits, which specifies the type of debugging
6269 outputs desired. Normally these are used only for internal development
6270 or system debugging purposes. You can find full documentation for these
6271 switches in the body of the @code{Debug} unit in the compiler source
6272 file @file{debug.adb}.
6276 @cindex @option{-gnatG} (@command{gcc})
6277 This switch causes the compiler to generate auxiliary output containing
6278 a pseudo-source listing of the generated expanded code. Like most Ada
6279 compilers, GNAT works by first transforming the high level Ada code into
6280 lower level constructs. For example, tasking operations are transformed
6281 into calls to the tasking run-time routines. A unique capability of GNAT
6282 is to list this expanded code in a form very close to normal Ada source.
6283 This is very useful in understanding the implications of various Ada
6284 usage on the efficiency of the generated code. There are many cases in
6285 Ada (e.g. the use of controlled types), where simple Ada statements can
6286 generate a lot of run-time code. By using @option{-gnatG} you can identify
6287 these cases, and consider whether it may be desirable to modify the coding
6288 approach to improve efficiency.
6290 The format of the output is very similar to standard Ada source, and is
6291 easily understood by an Ada programmer. The following special syntactic
6292 additions correspond to low level features used in the generated code that
6293 do not have any exact analogies in pure Ada source form. The following
6294 is a partial list of these special constructions. See the specification
6295 of package @code{Sprint} in file @file{sprint.ads} for a full list.
6298 @item new @var{xxx} [storage_pool = @var{yyy}]
6299 Shows the storage pool being used for an allocator.
6301 @item at end @var{procedure-name};
6302 Shows the finalization (cleanup) procedure for a scope.
6304 @item (if @var{expr} then @var{expr} else @var{expr})
6305 Conditional expression equivalent to the @code{x?y:z} construction in C.
6307 @item @var{target}^^^(@var{source})
6308 A conversion with floating-point truncation instead of rounding.
6310 @item @var{target}?(@var{source})
6311 A conversion that bypasses normal Ada semantic checking. In particular
6312 enumeration types and fixed-point types are treated simply as integers.
6314 @item @var{target}?^^^(@var{source})
6315 Combines the above two cases.
6317 @item @var{x} #/ @var{y}
6318 @itemx @var{x} #mod @var{y}
6319 @itemx @var{x} #* @var{y}
6320 @itemx @var{x} #rem @var{y}
6321 A division or multiplication of fixed-point values which are treated as
6322 integers without any kind of scaling.
6324 @item free @var{expr} [storage_pool = @var{xxx}]
6325 Shows the storage pool associated with a @code{free} statement.
6327 @item freeze @var{typename} [@var{actions}]
6328 Shows the point at which @var{typename} is frozen, with possible
6329 associated actions to be performed at the freeze point.
6331 @item reference @var{itype}
6332 Reference (and hence definition) to internal type @var{itype}.
6334 @item @var{function-name}! (@var{arg}, @var{arg}, @var{arg})
6335 Intrinsic function call.
6337 @item @var{labelname} : label
6338 Declaration of label @var{labelname}.
6340 @item @var{expr} && @var{expr} && @var{expr} ... && @var{expr}
6341 A multiple concatenation (same effect as @var{expr} & @var{expr} &
6342 @var{expr}, but handled more efficiently).
6344 @item [constraint_error]
6345 Raise the @code{Constraint_Error} exception.
6347 @item @var{expression}'reference
6348 A pointer to the result of evaluating @var{expression}.
6350 @item @var{target-type}!(@var{source-expression})
6351 An unchecked conversion of @var{source-expression} to @var{target-type}.
6353 @item [@var{numerator}/@var{denominator}]
6354 Used to represent internal real literals (that) have no exact
6355 representation in base 2-16 (for example, the result of compile time
6356 evaluation of the expression 1.0/27.0).
6360 @cindex @option{-gnatD} (@command{gcc})
6361 When used in conjunction with @option{-gnatG}, this switch causes
6362 the expanded source, as described above for
6363 @option{-gnatG} to be written to files with names
6364 @file{^xxx.dg^XXX_DG^}, where @file{xxx} is the normal file name,
6365 instead of to the standard ooutput file. For
6366 example, if the source file name is @file{hello.adb}, then a file
6367 @file{^hello.adb.dg^HELLO.ADB_DG^} will be written. The debugging
6368 information generated by the @command{gcc} @option{^-g^/DEBUG^} switch
6369 will refer to the generated @file{^xxx.dg^XXX_DG^} file. This allows
6370 you to do source level debugging using the generated code which is
6371 sometimes useful for complex code, for example to find out exactly
6372 which part of a complex construction raised an exception. This switch
6373 also suppress generation of cross-reference information (see
6374 @option{-gnatx}) since otherwise the cross-reference information
6375 would refer to the @file{^.dg^.DG^} file, which would cause
6376 confusion since this is not the original source file.
6378 Note that @option{-gnatD} actually implies @option{-gnatG}
6379 automatically, so it is not necessary to give both options.
6380 In other words @option{-gnatD} is equivalent to @option{-gnatDG}).
6383 @item -gnatR[0|1|2|3[s]]
6384 @cindex @option{-gnatR} (@command{gcc})
6385 This switch controls output from the compiler of a listing showing
6386 representation information for declared types and objects. For
6387 @option{-gnatR0}, no information is output (equivalent to omitting
6388 the @option{-gnatR} switch). For @option{-gnatR1} (which is the default,
6389 so @option{-gnatR} with no parameter has the same effect), size and alignment
6390 information is listed for declared array and record types. For
6391 @option{-gnatR2}, size and alignment information is listed for all
6392 expression information for values that are computed at run time for
6393 variant records. These symbolic expressions have a mostly obvious
6394 format with #n being used to represent the value of the n'th
6395 discriminant. See source files @file{repinfo.ads/adb} in the
6396 @code{GNAT} sources for full details on the format of @option{-gnatR3}
6397 output. If the switch is followed by an s (e.g. @option{-gnatR2s}), then
6398 the output is to a file with the name @file{^file.rep^file_REP^} where
6399 file is the name of the corresponding source file.
6402 @item /REPRESENTATION_INFO
6403 @cindex @option{/REPRESENTATION_INFO} (@command{gcc})
6404 This qualifier controls output from the compiler of a listing showing
6405 representation information for declared types and objects. For
6406 @option{/REPRESENTATION_INFO=NONE}, no information is output
6407 (equivalent to omitting the @option{/REPRESENTATION_INFO} qualifier).
6408 @option{/REPRESENTATION_INFO} without option is equivalent to
6409 @option{/REPRESENTATION_INFO=ARRAYS}.
6410 For @option{/REPRESENTATION_INFO=ARRAYS}, size and alignment
6411 information is listed for declared array and record types. For
6412 @option{/REPRESENTATION_INFO=OBJECTS}, size and alignment information
6413 is listed for all expression information for values that are computed
6414 at run time for variant records. These symbolic expressions have a mostly
6415 obvious format with #n being used to represent the value of the n'th
6416 discriminant. See source files @file{REPINFO.ADS/ADB} in the
6417 @code{GNAT} sources for full details on the format of
6418 @option{/REPRESENTATION_INFO=SYMBOLIC} output.
6419 If _FILE is added at the end of an option
6420 (e.g. @option{/REPRESENTATION_INFO=ARRAYS_FILE}),
6421 then the output is to a file with the name @file{file_REP} where
6422 file is the name of the corresponding source file.
6426 @cindex @option{-gnatS} (@command{gcc})
6427 The use of the switch @option{-gnatS} for an
6428 Ada compilation will cause the compiler to output a
6429 representation of package Standard in a form very
6430 close to standard Ada. It is not quite possible to
6431 do this entirely in standard Ada (since new
6432 numeric base types cannot be created in standard
6433 Ada), but the output is easily
6434 readable to any Ada programmer, and is useful to
6435 determine the characteristics of target dependent
6436 types in package Standard.
6439 @cindex @option{-gnatx} (@command{gcc})
6440 Normally the compiler generates full cross-referencing information in
6441 the @file{ALI} file. This information is used by a number of tools,
6442 including @code{gnatfind} and @code{gnatxref}. The @option{-gnatx} switch
6443 suppresses this information. This saves some space and may slightly
6444 speed up compilation, but means that these tools cannot be used.
6447 @node Exception Handling Control
6448 @subsection Exception Handling Control
6451 GNAT uses two methods for handling exceptions at run-time. The
6452 @code{setjmp/longjmp} method saves the context when entering
6453 a frame with an exception handler. Then when an exception is
6454 raised, the context can be restored immediately, without the
6455 need for tracing stack frames. This method provides very fast
6456 exception propagation, but introduces significant overhead for
6457 the use of exception handlers, even if no exception is raised.
6459 The other approach is called ``zero cost'' exception handling.
6460 With this method, the compiler builds static tables to describe
6461 the exception ranges. No dynamic code is required when entering
6462 a frame containing an exception handler. When an exception is
6463 raised, the tables are used to control a back trace of the
6464 subprogram invocation stack to locate the required exception
6465 handler. This method has considerably poorer performance for
6466 the propagation of exceptions, but there is no overhead for
6467 exception handlers if no exception is raised. Note that in this
6468 mode and in the context of mixed Ada and C/C++ programming,
6469 to propagate an exception through a C/C++ code, the C/C++ code
6470 must be compiled with the @option{-funwind-tables} GCC's
6473 The following switches can be used to control which of the
6474 two exception handling methods is used.
6480 @cindex @option{--RTS=sjlj} (@command{gnatmake})
6481 This switch causes the setjmp/longjmp run-time to be used
6482 for exception handling. If this is the default mechanism for the
6483 target (see below), then this has no effect. If the default
6484 mechanism for the target is zero cost exceptions, then
6485 this switch can be used to modify this default, and must be
6486 used for all units in the partition.
6487 This option is rarely used. One case in which it may be
6488 advantageous is if you have an application where exception
6489 raising is common and the overall performance of the
6490 application is improved by favoring exception propagation.
6493 @cindex @option{--RTS=zcx} (@command{gnatmake})
6494 @cindex Zero Cost Exceptions
6495 This switch causes the zero cost approach to be used
6496 for exception handling. If this is the default mechanism for the
6497 target (see below), then this has no effect. If the default
6498 mechanism for the target is setjmp/longjmp exceptions, then
6499 this switch can be used to modify this default, and must be
6500 used for all units in the partition.
6501 This option can only be used if the zero cost approach
6502 is available for the target in use (see below).
6506 The @code{setjmp/longjmp} approach is available on all targets, while
6507 the @code{zero cost} approach is available on selected targets.
6508 To determine whether zero cost exceptions can be used for a
6509 particular target, look at the private part of the file system.ads.
6510 Either @code{GCC_ZCX_Support} or @code{Front_End_ZCX_Support} must
6511 be True to use the zero cost approach. If both of these switches
6512 are set to False, this means that zero cost exception handling
6513 is not yet available for that target. The switch
6514 @code{ZCX_By_Default} indicates the default approach. If this
6515 switch is set to True, then the @code{zero cost} approach is
6518 @node Units to Sources Mapping Files
6519 @subsection Units to Sources Mapping Files
6523 @item -gnatem^^=^@var{path}
6524 @cindex @option{-gnatem} (@command{gcc})
6525 A mapping file is a way to communicate to the compiler two mappings:
6526 from unit names to file names (without any directory information) and from
6527 file names to path names (with full directory information). These mappings
6528 are used by the compiler to short-circuit the path search.
6530 The use of mapping files is not required for correct operation of the
6531 compiler, but mapping files can improve efficiency, particularly when
6532 sources are read over a slow network connection. In normal operation,
6533 you need not be concerned with the format or use of mapping files,
6534 and the @option{-gnatem} switch is not a switch that you would use
6535 explicitly. it is intended only for use by automatic tools such as
6536 @command{gnatmake} running under the project file facility. The
6537 description here of the format of mapping files is provided
6538 for completeness and for possible use by other tools.
6540 A mapping file is a sequence of sets of three lines. In each set,
6541 the first line is the unit name, in lower case, with ``@code{%s}''
6543 specifications and ``@code{%b}'' appended for bodies; the second line is the
6544 file name; and the third line is the path name.
6550 /gnat/project1/sources/main.2.ada
6553 When the switch @option{-gnatem} is specified, the compiler will create
6554 in memory the two mappings from the specified file. If there is any problem
6555 (non existent file, truncated file or duplicate entries), no mapping
6558 Several @option{-gnatem} switches may be specified; however, only the last
6559 one on the command line will be taken into account.
6561 When using a project file, @command{gnatmake} create a temporary mapping file
6562 and communicates it to the compiler using this switch.
6566 @node Integrated Preprocessing
6567 @subsection Integrated Preprocessing
6570 GNAT sources may be preprocessed immediately before compilation; the actual
6571 text of the source is not the text of the source file, but is derived from it
6572 through a process called preprocessing. Integrated preprocessing is specified
6573 through switches @option{-gnatep} and/or @option{-gnateD}. @option{-gnatep}
6574 indicates, through a text file, the preprocessing data to be used.
6575 @option{-gnateD} specifies or modifies the values of preprocessing symbol.
6578 It is recommended that @command{gnatmake} switch ^-s^/SWITCH_CHECK^ should be
6579 used when Integrated Preprocessing is used. The reason is that preprocessing
6580 with another Preprocessing Data file without changing the sources will
6581 not trigger recompilation without this switch.
6584 Note that @command{gnatmake} switch ^-m^/MINIMAL_RECOMPILATION^ will almost
6585 always trigger recompilation for sources that are preprocessed,
6586 because @command{gnatmake} cannot compute the checksum of the source after
6590 The actual preprocessing function is described in details in section
6591 @ref{Preprocessing Using gnatprep}. This section only describes how integrated
6592 preprocessing is triggered and parameterized.
6596 @item -gnatep=@var{file}
6597 @cindex @option{-gnatep} (@command{gcc})
6598 This switch indicates to the compiler the file name (without directory
6599 information) of the preprocessor data file to use. The preprocessor data file
6600 should be found in the source directories.
6603 A preprocessing data file is a text file with significant lines indicating
6604 how should be preprocessed either a specific source or all sources not
6605 mentioned in other lines. A significant line is a non empty, non comment line.
6606 Comments are similar to Ada comments.
6609 Each significant line starts with either a literal string or the character '*'.
6610 A literal string is the file name (without directory information) of the source
6611 to preprocess. A character '*' indicates the preprocessing for all the sources
6612 that are not specified explicitly on other lines (order of the lines is not
6613 significant). It is an error to have two lines with the same file name or two
6614 lines starting with the character '*'.
6617 After the file name or the character '*', another optional literal string
6618 indicating the file name of the definition file to be used for preprocessing
6619 (@pxref{Form of Definitions File}). The definition files are found by the
6620 compiler in one of the source directories. In some cases, when compiling
6621 a source in a directory other than the current directory, if the definition
6622 file is in the current directory, it may be necessary to add the current
6623 directory as a source directory through switch ^-I.^/SEARCH=[]^, otherwise
6624 the compiler would not find the definition file.
6627 Then, optionally, ^switches^switches^ similar to those of @code{gnatprep} may
6628 be found. Those ^switches^switches^ are:
6633 Causes both preprocessor lines and the lines deleted by
6634 preprocessing to be replaced by blank lines, preserving the line number.
6635 This ^switch^switch^ is always implied; however, if specified after @option{-c}
6636 it cancels the effect of @option{-c}.
6639 Causes both preprocessor lines and the lines deleted
6640 by preprocessing to be retained as comments marked
6641 with the special string ``@code{--! }''.
6643 @item -Dsymbol=value
6644 Define or redefine a symbol, associated with value. A symbol is an Ada
6645 identifier, or an Ada reserved word, with the exception of @code{if},
6646 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
6647 @code{value} is either a literal string, an Ada identifier or any Ada reserved
6648 word. A symbol declared with this ^switch^switch^ replaces a symbol with the
6649 same name defined in a definition file.
6652 Causes a sorted list of symbol names and values to be
6653 listed on the standard output file.
6656 Causes undefined symbols to be treated as having the value @code{FALSE}
6658 of a preprocessor test. In the absence of this option, an undefined symbol in
6659 a @code{#if} or @code{#elsif} test will be treated as an error.
6664 Examples of valid lines in a preprocessor data file:
6667 "toto.adb" "prep.def" -u
6668 -- preprocess "toto.adb", using definition file "prep.def",
6669 -- undefined symbol are False.
6672 -- preprocess all other sources without a definition file;
6673 -- suppressed lined are commented; symbol VERSION has the value V101.
6675 "titi.adb" "prep2.def" -s
6676 -- preprocess "titi.adb", using definition file "prep2.def";
6677 -- list all symbols with their values.
6680 @item ^-gnateD^/DATA_PREPROCESSING=^symbol[=value]
6681 @cindex @option{-gnateD} (@command{gcc})
6682 Define or redefine a preprocessing symbol, associated with value. If no value
6683 is given on the command line, then the value of the symbol is @code{True}.
6684 A symbol is an identifier, following normal Ada (case-insensitive)
6685 rules for its syntax, and value is any sequence (including an empty sequence)
6686 of characters from the set (letters, digits, period, underline).
6687 Ada reserved words may be used as symbols, with the exceptions of @code{if},
6688 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
6691 A symbol declared with this ^switch^switch^ on the command line replaces a
6692 symbol with the same name either in a definition file or specified with a
6693 ^switch^switch^ -D in the preprocessor data file.
6696 This switch is similar to switch @option{^-D^/ASSOCIATE^} of @code{gnatprep}.
6700 @node Code Generation Control
6701 @subsection Code Generation Control
6705 The GCC technology provides a wide range of target dependent
6706 @option{-m} switches for controlling
6707 details of code generation with respect to different versions of
6708 architectures. This includes variations in instruction sets (e.g.
6709 different members of the power pc family), and different requirements
6710 for optimal arrangement of instructions (e.g. different members of
6711 the x86 family). The list of available @option{-m} switches may be
6712 found in the GCC documentation.
6714 Use of the these @option{-m} switches may in some cases result in improved
6717 The GNAT Pro technology is tested and qualified without any
6718 @option{-m} switches,
6719 so generally the most reliable approach is to avoid the use of these
6720 switches. However, we generally expect most of these switches to work
6721 successfully with GNAT Pro, and many customers have reported successful
6722 use of these options.
6724 Our general advice is to avoid the use of @option{-m} switches unless
6725 special needs lead to requirements in this area. In particular,
6726 there is no point in using @option{-m} switches to improve performance
6727 unless you actually see a performance improvement.
6731 @subsection Return Codes
6732 @cindex Return Codes
6733 @cindex @option{/RETURN_CODES=VMS}
6736 On VMS, GNAT compiled programs return POSIX-style codes by default,
6737 e.g. @option{/RETURN_CODES=POSIX}.
6739 To enable VMS style return codes, use GNAT BIND and LINK with the option
6740 @option{/RETURN_CODES=VMS}. For example:
6743 GNAT BIND MYMAIN.ALI /RETURN_CODES=VMS
6744 GNAT LINK MYMAIN.ALI /RETURN_CODES=VMS
6748 Programs built with /RETURN_CODES=VMS are suitable to be called in
6749 VMS DCL scripts. Programs compiled with the default /RETURN_CODES=POSIX
6750 are suitable for spawning with appropriate GNAT RTL routines.
6754 @node Search Paths and the Run-Time Library (RTL)
6755 @section Search Paths and the Run-Time Library (RTL)
6758 With the GNAT source-based library system, the compiler must be able to
6759 find source files for units that are needed by the unit being compiled.
6760 Search paths are used to guide this process.
6762 The compiler compiles one source file whose name must be given
6763 explicitly on the command line. In other words, no searching is done
6764 for this file. To find all other source files that are needed (the most
6765 common being the specs of units), the compiler examines the following
6766 directories, in the following order:
6770 The directory containing the source file of the main unit being compiled
6771 (the file name on the command line).
6774 Each directory named by an @option{^-I^/SOURCE_SEARCH^} switch given on the
6775 @command{gcc} command line, in the order given.
6778 @findex ADA_PRJ_INCLUDE_FILE
6779 Each of the directories listed in the text file whose name is given
6780 by the @code{ADA_PRJ_INCLUDE_FILE} ^environment variable^logical name^.
6783 @code{ADA_PRJ_INCLUDE_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
6784 driver when project files are used. It should not normally be set
6788 @findex ADA_INCLUDE_PATH
6789 Each of the directories listed in the value of the
6790 @code{ADA_INCLUDE_PATH} ^environment variable^logical name^.
6792 Construct this value
6793 exactly as the @code{PATH} environment variable: a list of directory
6794 names separated by colons (semicolons when working with the NT version).
6797 Normally, define this value as a logical name containing a comma separated
6798 list of directory names.
6800 This variable can also be defined by means of an environment string
6801 (an argument to the DEC C exec* set of functions).
6805 DEFINE ANOTHER_PATH FOO:[BAG]
6806 DEFINE ADA_INCLUDE_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
6809 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
6810 first, followed by the standard Ada 95
6811 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADAINCLUDE].
6812 If this is not redefined, the user will obtain the DEC Ada 83 IO packages
6813 (Text_IO, Sequential_IO, etc)
6814 instead of the Ada95 packages. Thus, in order to get the Ada 95
6815 packages by default, ADA_INCLUDE_PATH must be redefined.
6819 The content of the @file{ada_source_path} file which is part of the GNAT
6820 installation tree and is used to store standard libraries such as the
6821 GNAT Run Time Library (RTL) source files.
6823 @ref{Installing a library}
6828 Specifying the switch @option{^-I-^/NOCURRENT_DIRECTORY^}
6829 inhibits the use of the directory
6830 containing the source file named in the command line. You can still
6831 have this directory on your search path, but in this case it must be
6832 explicitly requested with a @option{^-I^/SOURCE_SEARCH^} switch.
6834 Specifying the switch @option{-nostdinc}
6835 inhibits the search of the default location for the GNAT Run Time
6836 Library (RTL) source files.
6838 The compiler outputs its object files and ALI files in the current
6841 Caution: The object file can be redirected with the @option{-o} switch;
6842 however, @command{gcc} and @code{gnat1} have not been coordinated on this
6843 so the @file{ALI} file will not go to the right place. Therefore, you should
6844 avoid using the @option{-o} switch.
6848 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
6849 children make up the GNAT RTL, together with the simple @code{System.IO}
6850 package used in the @code{"Hello World"} example. The sources for these units
6851 are needed by the compiler and are kept together in one directory. Not
6852 all of the bodies are needed, but all of the sources are kept together
6853 anyway. In a normal installation, you need not specify these directory
6854 names when compiling or binding. Either the environment variables or
6855 the built-in defaults cause these files to be found.
6857 In addition to the language-defined hierarchies (@code{System}, @code{Ada} and
6858 @code{Interfaces}), the GNAT distribution provides a fourth hierarchy,
6859 consisting of child units of @code{GNAT}. This is a collection of generally
6860 useful types, subprograms, etc. See the @cite{GNAT Reference Manual} for
6863 Besides simplifying access to the RTL, a major use of search paths is
6864 in compiling sources from multiple directories. This can make
6865 development environments much more flexible.
6867 @node Order of Compilation Issues
6868 @section Order of Compilation Issues
6871 If, in our earlier example, there was a spec for the @code{hello}
6872 procedure, it would be contained in the file @file{hello.ads}; yet this
6873 file would not have to be explicitly compiled. This is the result of the
6874 model we chose to implement library management. Some of the consequences
6875 of this model are as follows:
6879 There is no point in compiling specs (except for package
6880 specs with no bodies) because these are compiled as needed by clients. If
6881 you attempt a useless compilation, you will receive an error message.
6882 It is also useless to compile subunits because they are compiled as needed
6886 There are no order of compilation requirements: performing a
6887 compilation never obsoletes anything. The only way you can obsolete
6888 something and require recompilations is to modify one of the
6889 source files on which it depends.
6892 There is no library as such, apart from the ALI files
6893 (@pxref{The Ada Library Information Files}, for information on the format
6894 of these files). For now we find it convenient to create separate ALI files,
6895 but eventually the information therein may be incorporated into the object
6899 When you compile a unit, the source files for the specs of all units
6900 that it @code{with}'s, all its subunits, and the bodies of any generics it
6901 instantiates must be available (reachable by the search-paths mechanism
6902 described above), or you will receive a fatal error message.
6909 The following are some typical Ada compilation command line examples:
6912 @item $ gcc -c xyz.adb
6913 Compile body in file @file{xyz.adb} with all default options.
6916 @item $ gcc -c -O2 -gnata xyz-def.adb
6919 @item $ GNAT COMPILE /OPTIMIZE=ALL -gnata xyz-def.adb
6922 Compile the child unit package in file @file{xyz-def.adb} with extensive
6923 optimizations, and pragma @code{Assert}/@code{Debug} statements
6926 @item $ gcc -c -gnatc abc-def.adb
6927 Compile the subunit in file @file{abc-def.adb} in semantic-checking-only
6931 @node Binding Using gnatbind
6932 @chapter Binding Using @code{gnatbind}
6936 * Running gnatbind::
6937 * Switches for gnatbind::
6938 * Command-Line Access::
6939 * Search Paths for gnatbind::
6940 * Examples of gnatbind Usage::
6944 This chapter describes the GNAT binder, @code{gnatbind}, which is used
6945 to bind compiled GNAT objects. The @code{gnatbind} program performs
6946 four separate functions:
6950 Checks that a program is consistent, in accordance with the rules in
6951 Chapter 10 of the Ada 95 Reference Manual. In particular, error
6952 messages are generated if a program uses inconsistent versions of a
6956 Checks that an acceptable order of elaboration exists for the program
6957 and issues an error message if it cannot find an order of elaboration
6958 that satisfies the rules in Chapter 10 of the Ada 95 Language Manual.
6961 Generates a main program incorporating the given elaboration order.
6962 This program is a small Ada package (body and spec) that
6963 must be subsequently compiled
6964 using the GNAT compiler. The necessary compilation step is usually
6965 performed automatically by @command{gnatlink}. The two most important
6966 functions of this program
6967 are to call the elaboration routines of units in an appropriate order
6968 and to call the main program.
6971 Determines the set of object files required by the given main program.
6972 This information is output in the forms of comments in the generated program,
6973 to be read by the @command{gnatlink} utility used to link the Ada application.
6976 @node Running gnatbind
6977 @section Running @code{gnatbind}
6980 The form of the @code{gnatbind} command is
6983 $ gnatbind [@i{switches}] @i{mainprog}[.ali] [@i{switches}]
6987 where @file{@i{mainprog}.adb} is the Ada file containing the main program
6988 unit body. If no switches are specified, @code{gnatbind} constructs an Ada
6989 package in two files whose names are
6990 @file{b~@i{mainprog}.ads}, and @file{b~@i{mainprog}.adb}.
6991 For example, if given the
6992 parameter @file{hello.ali}, for a main program contained in file
6993 @file{hello.adb}, the binder output files would be @file{b~hello.ads}
6994 and @file{b~hello.adb}.
6996 When doing consistency checking, the binder takes into consideration
6997 any source files it can locate. For example, if the binder determines
6998 that the given main program requires the package @code{Pack}, whose
7000 file is @file{pack.ali} and whose corresponding source spec file is
7001 @file{pack.ads}, it attempts to locate the source file @file{pack.ads}
7002 (using the same search path conventions as previously described for the
7003 @command{gcc} command). If it can locate this source file, it checks that
7005 or source checksums of the source and its references to in @file{ALI} files
7006 match. In other words, any @file{ALI} files that mentions this spec must have
7007 resulted from compiling this version of the source file (or in the case
7008 where the source checksums match, a version close enough that the
7009 difference does not matter).
7011 @cindex Source files, use by binder
7012 The effect of this consistency checking, which includes source files, is
7013 that the binder ensures that the program is consistent with the latest
7014 version of the source files that can be located at bind time. Editing a
7015 source file without compiling files that depend on the source file cause
7016 error messages to be generated by the binder.
7018 For example, suppose you have a main program @file{hello.adb} and a
7019 package @code{P}, from file @file{p.ads} and you perform the following
7024 Enter @code{gcc -c hello.adb} to compile the main program.
7027 Enter @code{gcc -c p.ads} to compile package @code{P}.
7030 Edit file @file{p.ads}.
7033 Enter @code{gnatbind hello}.
7037 At this point, the file @file{p.ali} contains an out-of-date time stamp
7038 because the file @file{p.ads} has been edited. The attempt at binding
7039 fails, and the binder generates the following error messages:
7042 error: "hello.adb" must be recompiled ("p.ads" has been modified)
7043 error: "p.ads" has been modified and must be recompiled
7047 Now both files must be recompiled as indicated, and then the bind can
7048 succeed, generating a main program. You need not normally be concerned
7049 with the contents of this file, but for reference purposes a sample
7050 binder output file is given in @ref{Example of Binder Output File}.
7052 In most normal usage, the default mode of @command{gnatbind} which is to
7053 generate the main package in Ada, as described in the previous section.
7054 In particular, this means that any Ada programmer can read and understand
7055 the generated main program. It can also be debugged just like any other
7056 Ada code provided the @option{^-g^/DEBUG^} switch is used for
7057 @command{gnatbind} and @command{gnatlink}.
7059 However for some purposes it may be convenient to generate the main
7060 program in C rather than Ada. This may for example be helpful when you
7061 are generating a mixed language program with the main program in C. The
7062 GNAT compiler itself is an example.
7063 The use of the @option{^-C^/BIND_FILE=C^} switch
7064 for both @code{gnatbind} and @command{gnatlink} will cause the program to
7065 be generated in C (and compiled using the gnu C compiler).
7067 @node Switches for gnatbind
7068 @section Switches for @command{gnatbind}
7071 The following switches are available with @code{gnatbind}; details will
7072 be presented in subsequent sections.
7075 * Consistency-Checking Modes::
7076 * Binder Error Message Control::
7077 * Elaboration Control::
7079 * Binding with Non-Ada Main Programs::
7080 * Binding Programs with No Main Subprogram::
7085 @item ^-aO^/OBJECT_SEARCH^
7086 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatbind})
7087 Specify directory to be searched for ALI files.
7089 @item ^-aI^/SOURCE_SEARCH^
7090 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
7091 Specify directory to be searched for source file.
7093 @item ^-A^/BIND_FILE=ADA^
7094 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatbind})
7095 Generate binder program in Ada (default)
7097 @item ^-b^/REPORT_ERRORS=BRIEF^
7098 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@command{gnatbind})
7099 Generate brief messages to @file{stderr} even if verbose mode set.
7101 @item ^-c^/NOOUTPUT^
7102 @cindex @option{^-c^/NOOUTPUT^} (@command{gnatbind})
7103 Check only, no generation of binder output file.
7105 @item ^-C^/BIND_FILE=C^
7106 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatbind})
7107 Generate binder program in C
7109 @item ^-e^/ELABORATION_DEPENDENCIES^
7110 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@command{gnatbind})
7111 Output complete list of elaboration-order dependencies.
7113 @item ^-E^/STORE_TRACEBACKS^
7114 @cindex @option{^-E^/STORE_TRACEBACKS^} (@command{gnatbind})
7115 Store tracebacks in exception occurrences when the target supports it.
7116 This is the default with the zero cost exception mechanism.
7118 @c The following may get moved to an appendix
7119 This option is currently supported on the following targets:
7120 all x86 ports, Solaris, Windows, HP-UX, AIX, PowerPC VxWorks and Alpha VxWorks.
7122 See also the packages @code{GNAT.Traceback} and
7123 @code{GNAT.Traceback.Symbolic} for more information.
7125 Note that on x86 ports, you must not use @option{-fomit-frame-pointer}
7126 @command{gcc} option.
7129 @item ^-F^/FORCE_ELABS_FLAGS^
7130 @cindex @option{^-F^/FORCE_ELABS_FLAGS^} (@command{gnatbind})
7131 Force the checks of elaboration flags. @command{gnatbind} does not normally
7132 generate checks of elaboration flags for the main executable, except when
7133 a Stand-Alone Library is used. However, there are cases when this cannot be
7134 detected by gnatbind. An example is importing an interface of a Stand-Alone
7135 Library through a pragma Import and only specifying through a linker switch
7136 this Stand-Alone Library. This switch is used to guarantee that elaboration
7137 flag checks are generated.
7140 @cindex @option{^-h^/HELP^} (@command{gnatbind})
7141 Output usage (help) information
7144 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
7145 Specify directory to be searched for source and ALI files.
7147 @item ^-I-^/NOCURRENT_DIRECTORY^
7148 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatbind})
7149 Do not look for sources in the current directory where @code{gnatbind} was
7150 invoked, and do not look for ALI files in the directory containing the
7151 ALI file named in the @code{gnatbind} command line.
7153 @item ^-l^/ORDER_OF_ELABORATION^
7154 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@command{gnatbind})
7155 Output chosen elaboration order.
7157 @item ^-Lxxx^/BUILD_LIBRARY=xxx^
7158 @cindex @option{^-L^/BUILD_LIBRARY^} (@command{gnatbind})
7159 Bind the units for library building. In this case the adainit and
7160 adafinal procedures (@pxref{Binding with Non-Ada Main Programs})
7161 are renamed to ^xxxinit^XXXINIT^ and
7162 ^xxxfinal^XXXFINAL^.
7163 Implies ^-n^/NOCOMPILE^.
7165 (@xref{GNAT and Libraries}, for more details.)
7168 On OpenVMS, these init and final procedures are exported in uppercase
7169 letters. For example if /BUILD_LIBRARY=toto is used, the exported name of
7170 the init procedure will be "TOTOINIT" and the exported name of the final
7171 procedure will be "TOTOFINAL".
7174 @item ^-Mxyz^/RENAME_MAIN=xyz^
7175 @cindex @option{^-M^/RENAME_MAIN^} (@command{gnatbind})
7176 Rename generated main program from main to xyz
7178 @item ^-m^/ERROR_LIMIT=^@var{n}
7179 @cindex @option{^-m^/ERROR_LIMIT^} (@command{gnatbind})
7180 Limit number of detected errors to @var{n}, where @var{n} is
7181 in the range 1..999_999. The default value if no switch is
7182 given is 9999. Binding is terminated if the limit is exceeded.
7184 Furthermore, under Windows, the sources pointed to by the libraries path
7185 set in the registry are not searched for.
7189 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
7193 @cindex @option{-nostdinc} (@command{gnatbind})
7194 Do not look for sources in the system default directory.
7197 @cindex @option{-nostdlib} (@command{gnatbind})
7198 Do not look for library files in the system default directory.
7200 @item --RTS=@var{rts-path}
7201 @cindex @option{--RTS} (@code{gnatbind})
7202 Specifies the default location of the runtime library. Same meaning as the
7203 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
7205 @item ^-o ^/OUTPUT=^@var{file}
7206 @cindex @option{^-o ^/OUTPUT^} (@command{gnatbind})
7207 Name the output file @var{file} (default is @file{b~@var{xxx}.adb}).
7208 Note that if this option is used, then linking must be done manually,
7209 gnatlink cannot be used.
7211 @item ^-O^/OBJECT_LIST^
7212 @cindex @option{^-O^/OBJECT_LIST^} (@command{gnatbind})
7215 @item ^-p^/PESSIMISTIC_ELABORATION^
7216 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@command{gnatbind})
7217 Pessimistic (worst-case) elaboration order
7219 @item ^-s^/READ_SOURCES=ALL^
7220 @cindex @option{^-s^/READ_SOURCES=ALL^} (@command{gnatbind})
7221 Require all source files to be present.
7223 @item ^-S@var{xxx}^/INITIALIZE_SCALARS=@var{xxx}^
7224 @cindex @option{^-S^/INITIALIZE_SCALARS^} (@command{gnatbind})
7225 Specifies the value to be used when detecting uninitialized scalar
7226 objects with pragma Initialize_Scalars.
7227 The @var{xxx} ^string specified with the switch^option^ may be either
7229 @item ``@option{^in^INVALID^}'' requesting an invalid value where possible
7230 @item ``@option{^lo^LOW^}'' for the lowest possible value
7231 @item ``@option{^hi^HIGH^}'' for the highest possible value
7232 @item ``@option{xx}'' for a value consisting of repeated bytes with the
7233 value 16#xx# (i.e. xx is a string of two hexadecimal digits).
7236 In addition, you can specify @option{-Sev} to indicate that the value is
7237 to be set at run time. In this case, the program will look for an environment
7238 @cindex GNAT_INIT_SCALARS
7239 variable of the form @code{GNAT_INIT_SCALARS=xx}, where xx is one
7240 of @option{in/lo/hi/xx} with the same meanings as above.
7241 If no environment variable is found, or if it does not have a valid value,
7242 then the default is @option{in} (invalid values).
7246 @cindex @option{-static} (@code{gnatbind})
7247 Link against a static GNAT run time.
7250 @cindex @option{-shared} (@code{gnatbind})
7251 Link against a shared GNAT run time when available.
7254 @item ^-t^/NOTIME_STAMP_CHECK^
7255 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
7256 Tolerate time stamp and other consistency errors
7258 @item ^-T@var{n}^/TIME_SLICE=@var{n}^
7259 @cindex @option{^-T^/TIME_SLICE^} (@code{gnatbind})
7260 Set the time slice value to @var{n} milliseconds. If the system supports
7261 the specification of a specific time slice value, then the indicated value
7262 is used. If the system does not support specific time slice values, but
7263 does support some general notion of round-robin scheduling, then any
7264 non-zero value will activate round-robin scheduling.
7266 A value of zero is treated specially. It turns off time
7267 slicing, and in addition, indicates to the tasking run time that the
7268 semantics should match as closely as possible the Annex D
7269 requirements of the Ada RM, and in particular sets the default
7270 scheduling policy to @code{FIFO_Within_Priorities}.
7272 @item ^-v^/REPORT_ERRORS=VERBOSE^
7273 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
7274 Verbose mode. Write error messages, header, summary output to
7279 @cindex @option{-w} (@code{gnatbind})
7280 Warning mode (@var{x}=s/e for suppress/treat as error)
7284 @item /WARNINGS=NORMAL
7285 @cindex @option{/WARNINGS} (@code{gnatbind})
7286 Normal warnings mode. Warnings are issued but ignored
7288 @item /WARNINGS=SUPPRESS
7289 @cindex @option{/WARNINGS} (@code{gnatbind})
7290 All warning messages are suppressed
7292 @item /WARNINGS=ERROR
7293 @cindex @option{/WARNINGS} (@code{gnatbind})
7294 Warning messages are treated as fatal errors
7297 @item ^-x^/READ_SOURCES=NONE^
7298 @cindex @option{^-x^/READ_SOURCES^} (@code{gnatbind})
7299 Exclude source files (check object consistency only).
7302 @item /READ_SOURCES=AVAILABLE
7303 @cindex @option{/READ_SOURCES} (@code{gnatbind})
7304 Default mode, in which sources are checked for consistency only if
7308 @item ^-z^/ZERO_MAIN^
7309 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
7315 You may obtain this listing of switches by running @code{gnatbind} with
7319 @node Consistency-Checking Modes
7320 @subsection Consistency-Checking Modes
7323 As described earlier, by default @code{gnatbind} checks
7324 that object files are consistent with one another and are consistent
7325 with any source files it can locate. The following switches control binder
7330 @item ^-s^/READ_SOURCES=ALL^
7331 @cindex @option{^-s^/READ_SOURCES=ALL^} (@code{gnatbind})
7332 Require source files to be present. In this mode, the binder must be
7333 able to locate all source files that are referenced, in order to check
7334 their consistency. In normal mode, if a source file cannot be located it
7335 is simply ignored. If you specify this switch, a missing source
7338 @item ^-x^/READ_SOURCES=NONE^
7339 @cindex @option{^-x^/READ_SOURCES=NONE^} (@code{gnatbind})
7340 Exclude source files. In this mode, the binder only checks that ALI
7341 files are consistent with one another. Source files are not accessed.
7342 The binder runs faster in this mode, and there is still a guarantee that
7343 the resulting program is self-consistent.
7344 If a source file has been edited since it was last compiled, and you
7345 specify this switch, the binder will not detect that the object
7346 file is out of date with respect to the source file. Note that this is the
7347 mode that is automatically used by @command{gnatmake} because in this
7348 case the checking against sources has already been performed by
7349 @command{gnatmake} in the course of compilation (i.e. before binding).
7352 @item /READ_SOURCES=AVAILABLE
7353 @cindex @code{/READ_SOURCES=AVAILABLE} (@code{gnatbind})
7354 This is the default mode in which source files are checked if they are
7355 available, and ignored if they are not available.
7359 @node Binder Error Message Control
7360 @subsection Binder Error Message Control
7363 The following switches provide control over the generation of error
7364 messages from the binder:
7368 @item ^-v^/REPORT_ERRORS=VERBOSE^
7369 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
7370 Verbose mode. In the normal mode, brief error messages are generated to
7371 @file{stderr}. If this switch is present, a header is written
7372 to @file{stdout} and any error messages are directed to @file{stdout}.
7373 All that is written to @file{stderr} is a brief summary message.
7375 @item ^-b^/REPORT_ERRORS=BRIEF^
7376 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@code{gnatbind})
7377 Generate brief error messages to @file{stderr} even if verbose mode is
7378 specified. This is relevant only when used with the
7379 @option{^-v^/REPORT_ERRORS=VERBOSE^} switch.
7383 @cindex @option{-m} (@code{gnatbind})
7384 Limits the number of error messages to @var{n}, a decimal integer in the
7385 range 1-999. The binder terminates immediately if this limit is reached.
7388 @cindex @option{-M} (@code{gnatbind})
7389 Renames the generated main program from @code{main} to @code{xxx}.
7390 This is useful in the case of some cross-building environments, where
7391 the actual main program is separate from the one generated
7395 @item ^-ws^/WARNINGS=SUPPRESS^
7396 @cindex @option{^-ws^/WARNINGS=SUPPRESS^} (@code{gnatbind})
7398 Suppress all warning messages.
7400 @item ^-we^/WARNINGS=ERROR^
7401 @cindex @option{^-we^/WARNINGS=ERROR^} (@code{gnatbind})
7402 Treat any warning messages as fatal errors.
7405 @item /WARNINGS=NORMAL
7406 Standard mode with warnings generated, but warnings do not get treated
7410 @item ^-t^/NOTIME_STAMP_CHECK^
7411 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
7412 @cindex Time stamp checks, in binder
7413 @cindex Binder consistency checks
7414 @cindex Consistency checks, in binder
7415 The binder performs a number of consistency checks including:
7419 Check that time stamps of a given source unit are consistent
7421 Check that checksums of a given source unit are consistent
7423 Check that consistent versions of @code{GNAT} were used for compilation
7425 Check consistency of configuration pragmas as required
7429 Normally failure of such checks, in accordance with the consistency
7430 requirements of the Ada Reference Manual, causes error messages to be
7431 generated which abort the binder and prevent the output of a binder
7432 file and subsequent link to obtain an executable.
7434 The @option{^-t^/NOTIME_STAMP_CHECK^} switch converts these error messages
7435 into warnings, so that
7436 binding and linking can continue to completion even in the presence of such
7437 errors. The result may be a failed link (due to missing symbols), or a
7438 non-functional executable which has undefined semantics.
7439 @emph{This means that
7440 @option{^-t^/NOTIME_STAMP_CHECK^} should be used only in unusual situations,
7444 @node Elaboration Control
7445 @subsection Elaboration Control
7448 The following switches provide additional control over the elaboration
7449 order. For full details see @ref{Elaboration Order Handling in GNAT}.
7452 @item ^-p^/PESSIMISTIC_ELABORATION^
7453 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@code{gnatbind})
7454 Normally the binder attempts to choose an elaboration order that is
7455 likely to minimize the likelihood of an elaboration order error resulting
7456 in raising a @code{Program_Error} exception. This switch reverses the
7457 action of the binder, and requests that it deliberately choose an order
7458 that is likely to maximize the likelihood of an elaboration error.
7459 This is useful in ensuring portability and avoiding dependence on
7460 accidental fortuitous elaboration ordering.
7462 Normally it only makes sense to use the @option{^-p^/PESSIMISTIC_ELABORATION^}
7464 elaboration checking is used (@option{-gnatE} switch used for compilation).
7465 This is because in the default static elaboration mode, all necessary
7466 @code{Elaborate_All} pragmas are implicitly inserted.
7467 These implicit pragmas are still respected by the binder in
7468 @option{^-p^/PESSIMISTIC_ELABORATION^} mode, so a
7469 safe elaboration order is assured.
7472 @node Output Control
7473 @subsection Output Control
7476 The following switches allow additional control over the output
7477 generated by the binder.
7482 @item ^-A^/BIND_FILE=ADA^
7483 @cindex @option{^-A^/BIND_FILE=ADA^} (@code{gnatbind})
7484 Generate binder program in Ada (default). The binder program is named
7485 @file{b~@var{mainprog}.adb} by default. This can be changed with
7486 @option{^-o^/OUTPUT^} @code{gnatbind} option.
7488 @item ^-c^/NOOUTPUT^
7489 @cindex @option{^-c^/NOOUTPUT^} (@code{gnatbind})
7490 Check only. Do not generate the binder output file. In this mode the
7491 binder performs all error checks but does not generate an output file.
7493 @item ^-C^/BIND_FILE=C^
7494 @cindex @option{^-C^/BIND_FILE=C^} (@code{gnatbind})
7495 Generate binder program in C. The binder program is named
7496 @file{b_@var{mainprog}.c}.
7497 This can be changed with @option{^-o^/OUTPUT^} @code{gnatbind}
7500 @item ^-e^/ELABORATION_DEPENDENCIES^
7501 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@code{gnatbind})
7502 Output complete list of elaboration-order dependencies, showing the
7503 reason for each dependency. This output can be rather extensive but may
7504 be useful in diagnosing problems with elaboration order. The output is
7505 written to @file{stdout}.
7508 @cindex @option{^-h^/HELP^} (@code{gnatbind})
7509 Output usage information. The output is written to @file{stdout}.
7511 @item ^-K^/LINKER_OPTION_LIST^
7512 @cindex @option{^-K^/LINKER_OPTION_LIST^} (@code{gnatbind})
7513 Output linker options to @file{stdout}. Includes library search paths,
7514 contents of pragmas Ident and Linker_Options, and libraries added
7517 @item ^-l^/ORDER_OF_ELABORATION^
7518 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@code{gnatbind})
7519 Output chosen elaboration order. The output is written to @file{stdout}.
7521 @item ^-O^/OBJECT_LIST^
7522 @cindex @option{^-O^/OBJECT_LIST^} (@code{gnatbind})
7523 Output full names of all the object files that must be linked to provide
7524 the Ada component of the program. The output is written to @file{stdout}.
7525 This list includes the files explicitly supplied and referenced by the user
7526 as well as implicitly referenced run-time unit files. The latter are
7527 omitted if the corresponding units reside in shared libraries. The
7528 directory names for the run-time units depend on the system configuration.
7530 @item ^-o ^/OUTPUT=^@var{file}
7531 @cindex @option{^-o^/OUTPUT^} (@code{gnatbind})
7532 Set name of output file to @var{file} instead of the normal
7533 @file{b~@var{mainprog}.adb} default. Note that @var{file} denote the Ada
7534 binder generated body filename. In C mode you would normally give
7535 @var{file} an extension of @file{.c} because it will be a C source program.
7536 Note that if this option is used, then linking must be done manually.
7537 It is not possible to use gnatlink in this case, since it cannot locate
7540 @item ^-r^/RESTRICTION_LIST^
7541 @cindex @option{^-r^/RESTRICTION_LIST^} (@code{gnatbind})
7542 Generate list of @code{pragma Restrictions} that could be applied to
7543 the current unit. This is useful for code audit purposes, and also may
7544 be used to improve code generation in some cases.
7548 @node Binding with Non-Ada Main Programs
7549 @subsection Binding with Non-Ada Main Programs
7552 In our description so far we have assumed that the main
7553 program is in Ada, and that the task of the binder is to generate a
7554 corresponding function @code{main} that invokes this Ada main
7555 program. GNAT also supports the building of executable programs where
7556 the main program is not in Ada, but some of the called routines are
7557 written in Ada and compiled using GNAT (@pxref{Mixed Language Programming}).
7558 The following switch is used in this situation:
7562 @cindex @option{^-n^/NOMAIN^} (@code{gnatbind})
7563 No main program. The main program is not in Ada.
7567 In this case, most of the functions of the binder are still required,
7568 but instead of generating a main program, the binder generates a file
7569 containing the following callable routines:
7574 You must call this routine to initialize the Ada part of the program by
7575 calling the necessary elaboration routines. A call to @code{adainit} is
7576 required before the first call to an Ada subprogram.
7578 Note that it is assumed that the basic execution environment must be setup
7579 to be appropriate for Ada execution at the point where the first Ada
7580 subprogram is called. In particular, if the Ada code will do any
7581 floating-point operations, then the FPU must be setup in an appropriate
7582 manner. For the case of the x86, for example, full precision mode is
7583 required. The procedure GNAT.Float_Control.Reset may be used to ensure
7584 that the FPU is in the right state.
7588 You must call this routine to perform any library-level finalization
7589 required by the Ada subprograms. A call to @code{adafinal} is required
7590 after the last call to an Ada subprogram, and before the program
7595 If the @option{^-n^/NOMAIN^} switch
7596 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
7597 @cindex Binder, multiple input files
7598 is given, more than one ALI file may appear on
7599 the command line for @code{gnatbind}. The normal @dfn{closure}
7600 calculation is performed for each of the specified units. Calculating
7601 the closure means finding out the set of units involved by tracing
7602 @code{with} references. The reason it is necessary to be able to
7603 specify more than one ALI file is that a given program may invoke two or
7604 more quite separate groups of Ada units.
7606 The binder takes the name of its output file from the last specified ALI
7607 file, unless overridden by the use of the @option{^-o file^/OUTPUT=file^}.
7608 @cindex @option{^-o^/OUTPUT^} (@command{gnatbind})
7609 The output is an Ada unit in source form that can
7610 be compiled with GNAT unless the -C switch is used in which case the
7611 output is a C source file, which must be compiled using the C compiler.
7612 This compilation occurs automatically as part of the @command{gnatlink}
7615 Currently the GNAT run time requires a FPU using 80 bits mode
7616 precision. Under targets where this is not the default it is required to
7617 call GNAT.Float_Control.Reset before using floating point numbers (this
7618 include float computation, float input and output) in the Ada code. A
7619 side effect is that this could be the wrong mode for the foreign code
7620 where floating point computation could be broken after this call.
7622 @node Binding Programs with No Main Subprogram
7623 @subsection Binding Programs with No Main Subprogram
7626 It is possible to have an Ada program which does not have a main
7627 subprogram. This program will call the elaboration routines of all the
7628 packages, then the finalization routines.
7630 The following switch is used to bind programs organized in this manner:
7633 @item ^-z^/ZERO_MAIN^
7634 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
7635 Normally the binder checks that the unit name given on the command line
7636 corresponds to a suitable main subprogram. When this switch is used,
7637 a list of ALI files can be given, and the execution of the program
7638 consists of elaboration of these units in an appropriate order.
7641 @node Command-Line Access
7642 @section Command-Line Access
7645 The package @code{Ada.Command_Line} provides access to the command-line
7646 arguments and program name. In order for this interface to operate
7647 correctly, the two variables
7659 are declared in one of the GNAT library routines. These variables must
7660 be set from the actual @code{argc} and @code{argv} values passed to the
7661 main program. With no @option{^n^/NOMAIN^} present, @code{gnatbind}
7662 generates the C main program to automatically set these variables.
7663 If the @option{^n^/NOMAIN^} switch is used, there is no automatic way to
7664 set these variables. If they are not set, the procedures in
7665 @code{Ada.Command_Line} will not be available, and any attempt to use
7666 them will raise @code{Constraint_Error}. If command line access is
7667 required, your main program must set @code{gnat_argc} and
7668 @code{gnat_argv} from the @code{argc} and @code{argv} values passed to
7671 @node Search Paths for gnatbind
7672 @section Search Paths for @code{gnatbind}
7675 The binder takes the name of an ALI file as its argument and needs to
7676 locate source files as well as other ALI files to verify object consistency.
7678 For source files, it follows exactly the same search rules as @command{gcc}
7679 (@pxref{Search Paths and the Run-Time Library (RTL)}). For ALI files the
7680 directories searched are:
7684 The directory containing the ALI file named in the command line, unless
7685 the switch @option{^-I-^/NOCURRENT_DIRECTORY^} is specified.
7688 All directories specified by @option{^-I^/SEARCH^}
7689 switches on the @code{gnatbind}
7690 command line, in the order given.
7693 @findex ADA_PRJ_OBJECTS_FILE
7694 Each of the directories listed in the text file whose name is given
7695 by the @code{ADA_PRJ_OBJECTS_FILE} ^environment variable^logical name^.
7698 @code{ADA_PRJ_OBJECTS_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
7699 driver when project files are used. It should not normally be set
7703 @findex ADA_OBJECTS_PATH
7704 Each of the directories listed in the value of the
7705 @code{ADA_OBJECTS_PATH} ^environment variable^logical name^.
7707 Construct this value
7708 exactly as the @code{PATH} environment variable: a list of directory
7709 names separated by colons (semicolons when working with the NT version
7713 Normally, define this value as a logical name containing a comma separated
7714 list of directory names.
7716 This variable can also be defined by means of an environment string
7717 (an argument to the DEC C exec* set of functions).
7721 DEFINE ANOTHER_PATH FOO:[BAG]
7722 DEFINE ADA_OBJECTS_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
7725 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
7726 first, followed by the standard Ada 95
7727 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADALIB].
7728 If this is not redefined, the user will obtain the DEC Ada 83 IO packages
7729 (Text_IO, Sequential_IO, etc)
7730 instead of the Ada95 packages. Thus, in order to get the Ada 95
7731 packages by default, ADA_OBJECTS_PATH must be redefined.
7735 The content of the @file{ada_object_path} file which is part of the GNAT
7736 installation tree and is used to store standard libraries such as the
7737 GNAT Run Time Library (RTL) unless the switch @option{-nostdlib} is
7740 @ref{Installing a library}
7745 In the binder the switch @option{^-I^/SEARCH^}
7746 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
7747 is used to specify both source and
7748 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
7749 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
7750 instead if you want to specify
7751 source paths only, and @option{^-aO^/LIBRARY_SEARCH^}
7752 @cindex @option{^-aO^/LIBRARY_SEARCH^} (@command{gnatbind})
7753 if you want to specify library paths
7754 only. This means that for the binder
7755 @option{^-I^/SEARCH=^}@var{dir} is equivalent to
7756 @option{^-aI^/SOURCE_SEARCH=^}@var{dir}
7757 @option{^-aO^/OBJECT_SEARCH=^}@var{dir}.
7758 The binder generates the bind file (a C language source file) in the
7759 current working directory.
7765 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
7766 children make up the GNAT Run-Time Library, together with the package
7767 GNAT and its children, which contain a set of useful additional
7768 library functions provided by GNAT. The sources for these units are
7769 needed by the compiler and are kept together in one directory. The ALI
7770 files and object files generated by compiling the RTL are needed by the
7771 binder and the linker and are kept together in one directory, typically
7772 different from the directory containing the sources. In a normal
7773 installation, you need not specify these directory names when compiling
7774 or binding. Either the environment variables or the built-in defaults
7775 cause these files to be found.
7777 Besides simplifying access to the RTL, a major use of search paths is
7778 in compiling sources from multiple directories. This can make
7779 development environments much more flexible.
7781 @node Examples of gnatbind Usage
7782 @section Examples of @code{gnatbind} Usage
7785 This section contains a number of examples of using the GNAT binding
7786 utility @code{gnatbind}.
7789 @item gnatbind hello
7790 The main program @code{Hello} (source program in @file{hello.adb}) is
7791 bound using the standard switch settings. The generated main program is
7792 @file{b~hello.adb}. This is the normal, default use of the binder.
7795 @item gnatbind hello -o mainprog.adb
7798 @item gnatbind HELLO.ALI /OUTPUT=Mainprog.ADB
7800 The main program @code{Hello} (source program in @file{hello.adb}) is
7801 bound using the standard switch settings. The generated main program is
7802 @file{mainprog.adb} with the associated spec in
7803 @file{mainprog.ads}. Note that you must specify the body here not the
7804 spec, in the case where the output is in Ada. Note that if this option
7805 is used, then linking must be done manually, since gnatlink will not
7806 be able to find the generated file.
7809 @item gnatbind main -C -o mainprog.c -x
7812 @item gnatbind MAIN.ALI /BIND_FILE=C /OUTPUT=Mainprog.C /READ_SOURCES=NONE
7814 The main program @code{Main} (source program in
7815 @file{main.adb}) is bound, excluding source files from the
7816 consistency checking, generating
7817 the file @file{mainprog.c}.
7820 @item gnatbind -x main_program -C -o mainprog.c
7821 This command is exactly the same as the previous example. Switches may
7822 appear anywhere in the command line, and single letter switches may be
7823 combined into a single switch.
7827 @item gnatbind -n math dbase -C -o ada-control.c
7830 @item gnatbind /NOMAIN math dbase /BIND_FILE=C /OUTPUT=ada-control.c
7832 The main program is in a language other than Ada, but calls to
7833 subprograms in packages @code{Math} and @code{Dbase} appear. This call
7834 to @code{gnatbind} generates the file @file{ada-control.c} containing
7835 the @code{adainit} and @code{adafinal} routines to be called before and
7836 after accessing the Ada units.
7839 @c ------------------------------------
7840 @node Linking Using gnatlink
7841 @chapter Linking Using @command{gnatlink}
7842 @c ------------------------------------
7846 This chapter discusses @command{gnatlink}, a tool that links
7847 an Ada program and builds an executable file. This utility
7848 invokes the system linker ^(via the @command{gcc} command)^^
7849 with a correct list of object files and library references.
7850 @command{gnatlink} automatically determines the list of files and
7851 references for the Ada part of a program. It uses the binder file
7852 generated by the @command{gnatbind} to determine this list.
7855 * Running gnatlink::
7856 * Switches for gnatlink::
7857 * Setting Stack Size from gnatlink::
7858 * Setting Heap Size from gnatlink::
7861 @node Running gnatlink
7862 @section Running @command{gnatlink}
7865 The form of the @command{gnatlink} command is
7868 $ gnatlink [@var{switches}] @var{mainprog}[.ali]
7869 [@var{non-Ada objects}] [@var{linker options}]
7873 The arguments of @command{gnatlink} (switches, main @file{ALI} file,
7875 or linker options) may be in any order, provided that no non-Ada object may
7876 be mistaken for a main @file{ALI} file.
7877 Any file name @file{F} without the @file{.ali}
7878 extension will be taken as the main @file{ALI} file if a file exists
7879 whose name is the concatenation of @file{F} and @file{.ali}.
7882 @file{@var{mainprog}.ali} references the ALI file of the main program.
7883 The @file{.ali} extension of this file can be omitted. From this
7884 reference, @command{gnatlink} locates the corresponding binder file
7885 @file{b~@var{mainprog}.adb} and, using the information in this file along
7886 with the list of non-Ada objects and linker options, constructs a
7887 linker command file to create the executable.
7889 The arguments other than the @command{gnatlink} switches and the main
7890 @file{ALI} file are passed to the linker uninterpreted.
7891 They typically include the names of
7892 object files for units written in other languages than Ada and any library
7893 references required to resolve references in any of these foreign language
7894 units, or in @code{Import} pragmas in any Ada units.
7896 @var{linker options} is an optional list of linker specific
7898 The default linker called by gnatlink is @var{gcc} which in
7899 turn calls the appropriate system linker.
7900 Standard options for the linker such as @option{-lmy_lib} or
7901 @option{-Ldir} can be added as is.
7902 For options that are not recognized by
7903 @var{gcc} as linker options, use the @var{gcc} switches @option{-Xlinker} or
7905 Refer to the GCC documentation for
7906 details. Here is an example showing how to generate a linker map:
7910 $ gnatlink my_prog -Wl,-Map,MAPFILE
7915 <<Need example for VMS>>
7918 Using @var{linker options} it is possible to set the program stack and
7919 heap size. See @ref{Setting Stack Size from gnatlink} and
7920 @ref{Setting Heap Size from gnatlink}.
7922 @command{gnatlink} determines the list of objects required by the Ada
7923 program and prepends them to the list of objects passed to the linker.
7924 @command{gnatlink} also gathers any arguments set by the use of
7925 @code{pragma Linker_Options} and adds them to the list of arguments
7926 presented to the linker.
7929 @command{gnatlink} accepts the following types of extra files on the command
7930 line: objects (.OBJ), libraries (.OLB), sharable images (.EXE), and
7931 options files (.OPT). These are recognized and handled according to their
7935 @node Switches for gnatlink
7936 @section Switches for @command{gnatlink}
7939 The following switches are available with the @command{gnatlink} utility:
7944 @item ^-A^/BIND_FILE=ADA^
7945 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatlink})
7946 The binder has generated code in Ada. This is the default.
7948 @item ^-C^/BIND_FILE=C^
7949 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatlink})
7950 If instead of generating a file in Ada, the binder has generated one in
7951 C, then the linker needs to know about it. Use this switch to signal
7952 to @command{gnatlink} that the binder has generated C code rather than
7955 @item ^-f^/FORCE_OBJECT_FILE_LIST^
7956 @cindex Command line length
7957 @cindex @option{^-f^/FORCE_OBJECT_FILE_LIST^} (@command{gnatlink})
7958 On some targets, the command line length is limited, and @command{gnatlink}
7959 will generate a separate file for the linker if the list of object files
7961 The @option{^-f^/FORCE_OBJECT_FILE_LIST^} switch forces this file
7962 to be generated even if
7963 the limit is not exceeded. This is useful in some cases to deal with
7964 special situations where the command line length is exceeded.
7967 @cindex Debugging information, including
7968 @cindex @option{^-g^/DEBUG^} (@command{gnatlink})
7969 The option to include debugging information causes the Ada bind file (in
7970 other words, @file{b~@var{mainprog}.adb}) to be compiled with
7971 @option{^-g^/DEBUG^}.
7972 In addition, the binder does not delete the @file{b~@var{mainprog}.adb},
7973 @file{b~@var{mainprog}.o} and @file{b~@var{mainprog}.ali} files.
7974 Without @option{^-g^/DEBUG^}, the binder removes these files by
7975 default. The same procedure apply if a C bind file was generated using
7976 @option{^-C^/BIND_FILE=C^} @code{gnatbind} option, in this case the filenames
7977 are @file{b_@var{mainprog}.c} and @file{b_@var{mainprog}.o}.
7979 @item ^-n^/NOCOMPILE^
7980 @cindex @option{^-n^/NOCOMPILE^} (@command{gnatlink})
7981 Do not compile the file generated by the binder. This may be used when
7982 a link is rerun with different options, but there is no need to recompile
7986 @cindex @option{^-v^/VERBOSE^} (@command{gnatlink})
7987 Causes additional information to be output, including a full list of the
7988 included object files. This switch option is most useful when you want
7989 to see what set of object files are being used in the link step.
7991 @item ^-v -v^/VERBOSE/VERBOSE^
7992 @cindex @option{^-v -v^/VERBOSE/VERBOSE^} (@command{gnatlink})
7993 Very verbose mode. Requests that the compiler operate in verbose mode when
7994 it compiles the binder file, and that the system linker run in verbose mode.
7996 @item ^-o ^/EXECUTABLE=^@var{exec-name}
7997 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatlink})
7998 @var{exec-name} specifies an alternate name for the generated
7999 executable program. If this switch is omitted, the executable has the same
8000 name as the main unit. For example, @code{gnatlink try.ali} creates
8001 an executable called @file{^try^TRY.EXE^}.
8004 @item -b @var{target}
8005 @cindex @option{-b} (@command{gnatlink})
8006 Compile your program to run on @var{target}, which is the name of a
8007 system configuration. You must have a GNAT cross-compiler built if
8008 @var{target} is not the same as your host system.
8011 @cindex @option{-B} (@command{gnatlink})
8012 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
8013 from @var{dir} instead of the default location. Only use this switch
8014 when multiple versions of the GNAT compiler are available. See the
8015 @command{gcc} manual page for further details. You would normally use the
8016 @option{-b} or @option{-V} switch instead.
8018 @item --GCC=@var{compiler_name}
8019 @cindex @option{--GCC=compiler_name} (@command{gnatlink})
8020 Program used for compiling the binder file. The default is
8021 @command{gcc}. You need to use quotes around @var{compiler_name} if
8022 @code{compiler_name} contains spaces or other separator characters. As
8023 an example @option{--GCC="foo -x -y"} will instruct @command{gnatlink} to use
8024 @code{foo -x -y} as your compiler. Note that switch @option{-c} is always
8025 inserted after your command name. Thus in the above example the compiler
8026 command that will be used by @command{gnatlink} will be @code{foo -c -x -y}.
8027 If several @option{--GCC=compiler_name} are used, only the last
8028 @var{compiler_name} is taken into account. However, all the additional
8029 switches are also taken into account. Thus,
8030 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
8031 @option{--GCC="bar -x -y -z -t"}.
8033 @item --LINK=@var{name}
8034 @cindex @option{--LINK=} (@command{gnatlink})
8035 @var{name} is the name of the linker to be invoked. This is especially
8036 useful in mixed language programs since languages such as C++ require
8037 their own linker to be used. When this switch is omitted, the default
8038 name for the linker is @command{gcc}. When this switch is used, the
8039 specified linker is called instead of @command{gcc} with exactly the same
8040 parameters that would have been passed to @command{gcc} so if the desired
8041 linker requires different parameters it is necessary to use a wrapper
8042 script that massages the parameters before invoking the real linker. It
8043 may be useful to control the exact invocation by using the verbose
8049 @item /DEBUG=TRACEBACK
8050 @cindex @code{/DEBUG=TRACEBACK} (@command{gnatlink})
8051 This qualifier causes sufficient information to be included in the
8052 executable file to allow a traceback, but does not include the full
8053 symbol information needed by the debugger.
8055 @item /IDENTIFICATION="<string>"
8056 @code{"<string>"} specifies the string to be stored in the image file
8057 identification field in the image header.
8058 It overrides any pragma @code{Ident} specified string.
8060 @item /NOINHIBIT-EXEC
8061 Generate the executable file even if there are linker warnings.
8063 @item /NOSTART_FILES
8064 Don't link in the object file containing the ``main'' transfer address.
8065 Used when linking with a foreign language main program compiled with a
8069 Prefer linking with object libraries over sharable images, even without
8075 @node Setting Stack Size from gnatlink
8076 @section Setting Stack Size from @command{gnatlink}
8079 Under Windows systems, it is possible to specify the program stack size from
8080 @command{gnatlink} using either:
8084 @item using @option{-Xlinker} linker option
8087 $ gnatlink hello -Xlinker --stack=0x10000,0x1000
8090 This sets the stack reserve size to 0x10000 bytes and the stack commit
8091 size to 0x1000 bytes.
8093 @item using @option{-Wl} linker option
8096 $ gnatlink hello -Wl,--stack=0x1000000
8099 This sets the stack reserve size to 0x1000000 bytes. Note that with
8100 @option{-Wl} option it is not possible to set the stack commit size
8101 because the coma is a separator for this option.
8105 @node Setting Heap Size from gnatlink
8106 @section Setting Heap Size from @command{gnatlink}
8109 Under Windows systems, it is possible to specify the program heap size from
8110 @command{gnatlink} using either:
8114 @item using @option{-Xlinker} linker option
8117 $ gnatlink hello -Xlinker --heap=0x10000,0x1000
8120 This sets the heap reserve size to 0x10000 bytes and the heap commit
8121 size to 0x1000 bytes.
8123 @item using @option{-Wl} linker option
8126 $ gnatlink hello -Wl,--heap=0x1000000
8129 This sets the heap reserve size to 0x1000000 bytes. Note that with
8130 @option{-Wl} option it is not possible to set the heap commit size
8131 because the coma is a separator for this option.
8135 @node The GNAT Make Program gnatmake
8136 @chapter The GNAT Make Program @command{gnatmake}
8140 * Running gnatmake::
8141 * Switches for gnatmake::
8142 * Mode Switches for gnatmake::
8143 * Notes on the Command Line::
8144 * How gnatmake Works::
8145 * Examples of gnatmake Usage::
8148 A typical development cycle when working on an Ada program consists of
8149 the following steps:
8153 Edit some sources to fix bugs.
8159 Compile all sources affected.
8169 The third step can be tricky, because not only do the modified files
8170 @cindex Dependency rules
8171 have to be compiled, but any files depending on these files must also be
8172 recompiled. The dependency rules in Ada can be quite complex, especially
8173 in the presence of overloading, @code{use} clauses, generics and inlined
8176 @command{gnatmake} automatically takes care of the third and fourth steps
8177 of this process. It determines which sources need to be compiled,
8178 compiles them, and binds and links the resulting object files.
8180 Unlike some other Ada make programs, the dependencies are always
8181 accurately recomputed from the new sources. The source based approach of
8182 the GNAT compilation model makes this possible. This means that if
8183 changes to the source program cause corresponding changes in
8184 dependencies, they will always be tracked exactly correctly by
8187 @node Running gnatmake
8188 @section Running @command{gnatmake}
8191 The usual form of the @command{gnatmake} command is
8194 $ gnatmake [@var{switches}] @var{file_name}
8195 [@var{file_names}] [@var{mode_switches}]
8199 The only required argument is one @var{file_name}, which specifies
8200 a compilation unit that is a main program. Several @var{file_names} can be
8201 specified: this will result in several executables being built.
8202 If @code{switches} are present, they can be placed before the first
8203 @var{file_name}, between @var{file_names} or after the last @var{file_name}.
8204 If @var{mode_switches} are present, they must always be placed after
8205 the last @var{file_name} and all @code{switches}.
8207 If you are using standard file extensions (.adb and .ads), then the
8208 extension may be omitted from the @var{file_name} arguments. However, if
8209 you are using non-standard extensions, then it is required that the
8210 extension be given. A relative or absolute directory path can be
8211 specified in a @var{file_name}, in which case, the input source file will
8212 be searched for in the specified directory only. Otherwise, the input
8213 source file will first be searched in the directory where
8214 @command{gnatmake} was invoked and if it is not found, it will be search on
8215 the source path of the compiler as described in
8216 @ref{Search Paths and the Run-Time Library (RTL)}.
8218 All @command{gnatmake} output (except when you specify
8219 @option{^-M^/DEPENDENCIES_LIST^}) is to
8220 @file{stderr}. The output produced by the
8221 @option{^-M^/DEPENDENCIES_LIST^} switch is send to
8224 @node Switches for gnatmake
8225 @section Switches for @command{gnatmake}
8228 You may specify any of the following switches to @command{gnatmake}:
8233 @item --GCC=@var{compiler_name}
8234 @cindex @option{--GCC=compiler_name} (@command{gnatmake})
8235 Program used for compiling. The default is `@command{gcc}'. You need to use
8236 quotes around @var{compiler_name} if @code{compiler_name} contains
8237 spaces or other separator characters. As an example @option{--GCC="foo -x
8238 -y"} will instruct @command{gnatmake} to use @code{foo -x -y} as your
8239 compiler. Note that switch @option{-c} is always inserted after your
8240 command name. Thus in the above example the compiler command that will
8241 be used by @command{gnatmake} will be @code{foo -c -x -y}.
8242 If several @option{--GCC=compiler_name} are used, only the last
8243 @var{compiler_name} is taken into account. However, all the additional
8244 switches are also taken into account. Thus,
8245 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
8246 @option{--GCC="bar -x -y -z -t"}.
8248 @item --GNATBIND=@var{binder_name}
8249 @cindex @option{--GNATBIND=binder_name} (@command{gnatmake})
8250 Program used for binding. The default is `@code{gnatbind}'. You need to
8251 use quotes around @var{binder_name} if @var{binder_name} contains spaces
8252 or other separator characters. As an example @option{--GNATBIND="bar -x
8253 -y"} will instruct @command{gnatmake} to use @code{bar -x -y} as your
8254 binder. Binder switches that are normally appended by @command{gnatmake} to
8255 `@code{gnatbind}' are now appended to the end of @code{bar -x -y}.
8257 @item --GNATLINK=@var{linker_name}
8258 @cindex @option{--GNATLINK=linker_name} (@command{gnatmake})
8259 Program used for linking. The default is `@command{gnatlink}'. You need to
8260 use quotes around @var{linker_name} if @var{linker_name} contains spaces
8261 or other separator characters. As an example @option{--GNATLINK="lan -x
8262 -y"} will instruct @command{gnatmake} to use @code{lan -x -y} as your
8263 linker. Linker switches that are normally appended by @command{gnatmake} to
8264 `@command{gnatlink}' are now appended to the end of @code{lan -x -y}.
8268 @item ^-a^/ALL_FILES^
8269 @cindex @option{^-a^/ALL_FILES^} (@command{gnatmake})
8270 Consider all files in the make process, even the GNAT internal system
8271 files (for example, the predefined Ada library files), as well as any
8272 locked files. Locked files are files whose ALI file is write-protected.
8274 @command{gnatmake} does not check these files,
8275 because the assumption is that the GNAT internal files are properly up
8276 to date, and also that any write protected ALI files have been properly
8277 installed. Note that if there is an installation problem, such that one
8278 of these files is not up to date, it will be properly caught by the
8280 You may have to specify this switch if you are working on GNAT
8281 itself. The switch @option{^-a^/ALL_FILES^} is also useful
8282 in conjunction with @option{^-f^/FORCE_COMPILE^}
8283 if you need to recompile an entire application,
8284 including run-time files, using special configuration pragmas,
8285 such as a @code{Normalize_Scalars} pragma.
8288 @code{gnatmake ^-a^/ALL_FILES^} compiles all GNAT
8291 @code{gcc -c -gnatpg} rather than @code{gcc -c}.
8294 the @code{/CHECKS=SUPPRESS_ALL /STYLE_CHECKS=GNAT} switch.
8297 @item ^-b^/ACTIONS=BIND^
8298 @cindex @option{^-b^/ACTIONS=BIND^} (@command{gnatmake})
8299 Bind only. Can be combined with @option{^-c^/ACTIONS=COMPILE^} to do
8300 compilation and binding, but no link.
8301 Can be combined with @option{^-l^/ACTIONS=LINK^}
8302 to do binding and linking. When not combined with
8303 @option{^-c^/ACTIONS=COMPILE^}
8304 all the units in the closure of the main program must have been previously
8305 compiled and must be up to date. The root unit specified by @var{file_name}
8306 may be given without extension, with the source extension or, if no GNAT
8307 Project File is specified, with the ALI file extension.
8309 @item ^-c^/ACTIONS=COMPILE^
8310 @cindex @option{^-c^/ACTIONS=COMPILE^} (@command{gnatmake})
8311 Compile only. Do not perform binding, except when @option{^-b^/ACTIONS=BIND^}
8312 is also specified. Do not perform linking, except if both
8313 @option{^-b^/ACTIONS=BIND^} and
8314 @option{^-l^/ACTIONS=LINK^} are also specified.
8315 If the root unit specified by @var{file_name} is not a main unit, this is the
8316 default. Otherwise @command{gnatmake} will attempt binding and linking
8317 unless all objects are up to date and the executable is more recent than
8321 @cindex @option{^-C^/MAPPING^} (@command{gnatmake})
8322 Use a temporary mapping file. A mapping file is a way to communicate to the
8323 compiler two mappings: from unit names to file names (without any directory
8324 information) and from file names to path names (with full directory
8325 information). These mappings are used by the compiler to short-circuit the path
8326 search. When @command{gnatmake} is invoked with this switch, it will create
8327 a temporary mapping file, initially populated by the project manager,
8328 if @option{^-P^/PROJECT_FILE^} is used, otherwise initially empty.
8329 Each invocation of the compiler will add the newly accessed sources to the
8330 mapping file. This will improve the source search during the next invocation
8333 @item ^-C=^/USE_MAPPING_FILE=^@var{file}
8334 @cindex @option{^-C=^/USE_MAPPING^} (@command{gnatmake})
8335 Use a specific mapping file. The file, specified as a path name (absolute or
8336 relative) by this switch, should already exist, otherwise the switch is
8337 ineffective. The specified mapping file will be communicated to the compiler.
8338 This switch is not compatible with a project file
8339 (^-P^/PROJECT_FILE=^@var{file}) or with multiple compiling processes
8340 (^-j^/PROCESSES=^nnn, when nnn is greater than 1).
8342 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
8343 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@command{gnatmake})
8344 Put all object files and ALI file in directory @var{dir}.
8345 If the @option{^-D^/DIRECTORY_OBJECTS^} switch is not used, all object files
8346 and ALI files go in the current working directory.
8348 This switch cannot be used when using a project file.
8352 @cindex @option{-eL} (@command{gnatmake})
8353 Follow all symbolic links when processing project files.
8356 @item ^-f^/FORCE_COMPILE^
8357 @cindex @option{^-f^/FORCE_COMPILE^} (@command{gnatmake})
8358 Force recompilations. Recompile all sources, even though some object
8359 files may be up to date, but don't recompile predefined or GNAT internal
8360 files or locked files (files with a write-protected ALI file),
8361 unless the @option{^-a^/ALL_FILES^} switch is also specified.
8363 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
8364 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@command{gnatmake})
8365 When using project files, if some errors or warnings are detected during
8366 parsing and verbose mode is not in effect (no use of switch
8367 ^-v^/VERBOSE^), then error lines start with the full path name of the project
8368 file, rather than its simple file name.
8370 @item ^-i^/IN_PLACE^
8371 @cindex @option{^-i^/IN_PLACE^} (@command{gnatmake})
8372 In normal mode, @command{gnatmake} compiles all object files and ALI files
8373 into the current directory. If the @option{^-i^/IN_PLACE^} switch is used,
8374 then instead object files and ALI files that already exist are overwritten
8375 in place. This means that once a large project is organized into separate
8376 directories in the desired manner, then @command{gnatmake} will automatically
8377 maintain and update this organization. If no ALI files are found on the
8378 Ada object path (@ref{Search Paths and the Run-Time Library (RTL)}),
8379 the new object and ALI files are created in the
8380 directory containing the source being compiled. If another organization
8381 is desired, where objects and sources are kept in different directories,
8382 a useful technique is to create dummy ALI files in the desired directories.
8383 When detecting such a dummy file, @command{gnatmake} will be forced to
8384 recompile the corresponding source file, and it will be put the resulting
8385 object and ALI files in the directory where it found the dummy file.
8387 @item ^-j^/PROCESSES=^@var{n}
8388 @cindex @option{^-j^/PROCESSES^} (@command{gnatmake})
8389 @cindex Parallel make
8390 Use @var{n} processes to carry out the (re)compilations. On a
8391 multiprocessor machine compilations will occur in parallel. In the
8392 event of compilation errors, messages from various compilations might
8393 get interspersed (but @command{gnatmake} will give you the full ordered
8394 list of failing compiles at the end). If this is problematic, rerun
8395 the make process with n set to 1 to get a clean list of messages.
8397 @item ^-k^/CONTINUE_ON_ERROR^
8398 @cindex @option{^-k^/CONTINUE_ON_ERROR^} (@command{gnatmake})
8399 Keep going. Continue as much as possible after a compilation error. To
8400 ease the programmer's task in case of compilation errors, the list of
8401 sources for which the compile fails is given when @command{gnatmake}
8404 If @command{gnatmake} is invoked with several @file{file_names} and with this
8405 switch, if there are compilation errors when building an executable,
8406 @command{gnatmake} will not attempt to build the following executables.
8408 @item ^-l^/ACTIONS=LINK^
8409 @cindex @option{^-l^/ACTIONS=LINK^} (@command{gnatmake})
8410 Link only. Can be combined with @option{^-b^/ACTIONS=BIND^} to binding
8411 and linking. Linking will not be performed if combined with
8412 @option{^-c^/ACTIONS=COMPILE^}
8413 but not with @option{^-b^/ACTIONS=BIND^}.
8414 When not combined with @option{^-b^/ACTIONS=BIND^}
8415 all the units in the closure of the main program must have been previously
8416 compiled and must be up to date, and the main program needs to have been bound.
8417 The root unit specified by @var{file_name}
8418 may be given without extension, with the source extension or, if no GNAT
8419 Project File is specified, with the ALI file extension.
8421 @item ^-m^/MINIMAL_RECOMPILATION^
8422 @cindex @option{^-m^/MINIMAL_RECOMPILATION^} (@command{gnatmake})
8423 Specify that the minimum necessary amount of recompilations
8424 be performed. In this mode @command{gnatmake} ignores time
8425 stamp differences when the only
8426 modifications to a source file consist in adding/removing comments,
8427 empty lines, spaces or tabs. This means that if you have changed the
8428 comments in a source file or have simply reformatted it, using this
8429 switch will tell gnatmake not to recompile files that depend on it
8430 (provided other sources on which these files depend have undergone no
8431 semantic modifications). Note that the debugging information may be
8432 out of date with respect to the sources if the @option{-m} switch causes
8433 a compilation to be switched, so the use of this switch represents a
8434 trade-off between compilation time and accurate debugging information.
8436 @item ^-M^/DEPENDENCIES_LIST^
8437 @cindex Dependencies, producing list
8438 @cindex @option{^-M^/DEPENDENCIES_LIST^} (@command{gnatmake})
8439 Check if all objects are up to date. If they are, output the object
8440 dependences to @file{stdout} in a form that can be directly exploited in
8441 a @file{Makefile}. By default, each source file is prefixed with its
8442 (relative or absolute) directory name. This name is whatever you
8443 specified in the various @option{^-aI^/SOURCE_SEARCH^}
8444 and @option{^-I^/SEARCH^} switches. If you use
8445 @code{gnatmake ^-M^/DEPENDENCIES_LIST^}
8446 @option{^-q^/QUIET^}
8447 (see below), only the source file names,
8448 without relative paths, are output. If you just specify the
8449 @option{^-M^/DEPENDENCIES_LIST^}
8450 switch, dependencies of the GNAT internal system files are omitted. This
8451 is typically what you want. If you also specify
8452 the @option{^-a^/ALL_FILES^} switch,
8453 dependencies of the GNAT internal files are also listed. Note that
8454 dependencies of the objects in external Ada libraries (see switch
8455 @option{^-aL^/SKIP_MISSING=^}@var{dir} in the following list)
8458 @item ^-n^/DO_OBJECT_CHECK^
8459 @cindex @option{^-n^/DO_OBJECT_CHECK^} (@command{gnatmake})
8460 Don't compile, bind, or link. Checks if all objects are up to date.
8461 If they are not, the full name of the first file that needs to be
8462 recompiled is printed.
8463 Repeated use of this option, followed by compiling the indicated source
8464 file, will eventually result in recompiling all required units.
8466 @item ^-o ^/EXECUTABLE=^@var{exec_name}
8467 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatmake})
8468 Output executable name. The name of the final executable program will be
8469 @var{exec_name}. If the @option{^-o^/EXECUTABLE^} switch is omitted the default
8470 name for the executable will be the name of the input file in appropriate form
8471 for an executable file on the host system.
8473 This switch cannot be used when invoking @command{gnatmake} with several
8476 @item ^-P^/PROJECT_FILE=^@var{project}
8477 @cindex @option{^-P^/PROJECT_FILE^} (@command{gnatmake})
8478 Use project file @var{project}. Only one such switch can be used.
8479 @xref{gnatmake and Project Files}.
8482 @cindex @option{^-q^/QUIET^} (@command{gnatmake})
8483 Quiet. When this flag is not set, the commands carried out by
8484 @command{gnatmake} are displayed.
8486 @item ^-s^/SWITCH_CHECK/^
8487 @cindex @option{^-s^/SWITCH_CHECK^} (@command{gnatmake})
8488 Recompile if compiler switches have changed since last compilation.
8489 All compiler switches but -I and -o are taken into account in the
8491 orders between different ``first letter'' switches are ignored, but
8492 orders between same switches are taken into account. For example,
8493 @option{-O -O2} is different than @option{-O2 -O}, but @option{-g -O}
8494 is equivalent to @option{-O -g}.
8496 This switch is recommended when Integrated Preprocessing is used.
8499 @cindex @option{^-u^/UNIQUE^} (@command{gnatmake})
8500 Unique. Recompile at most the main files. It implies -c. Combined with
8501 -f, it is equivalent to calling the compiler directly. Note that using
8502 ^-u^/UNIQUE^ with a project file and no main has a special meaning
8503 (@pxref{Project Files and Main Subprograms}).
8505 @item ^-U^/ALL_PROJECTS^
8506 @cindex @option{^-U^/ALL_PROJECTS^} (@command{gnatmake})
8507 When used without a project file or with one or several mains on the command
8508 line, is equivalent to ^-u^/UNIQUE^. When used with a project file and no main
8509 on the command line, all sources of all project files are checked and compiled
8510 if not up to date, and libraries are rebuilt, if necessary.
8513 @cindex @option{^-v^/REASONS^} (@command{gnatmake})
8514 Verbose. Display the reason for all recompilations @command{gnatmake}
8515 decides are necessary.
8517 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
8518 Indicate the verbosity of the parsing of GNAT project files.
8519 @xref{Switches Related to Project Files}.
8521 @item ^-x^/NON_PROJECT_UNIT_COMPILATION^
8522 @cindex @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} (@command{gnatmake})
8523 Indicate that sources that are not part of any Project File may be compiled.
8524 Normally, when using Project Files, only sources that are part of a Project
8525 File may be compile. When this switch is used, a source outside of all Project
8526 Files may be compiled. The ALI file and the object file will be put in the
8527 object directory of the main Project. The compilation switches used will only
8528 be those specified on the command line.
8530 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
8531 Indicate that external variable @var{name} has the value @var{value}.
8532 The Project Manager will use this value for occurrences of
8533 @code{external(name)} when parsing the project file.
8534 @xref{Switches Related to Project Files}.
8537 @cindex @option{^-z^/NOMAIN^} (@command{gnatmake})
8538 No main subprogram. Bind and link the program even if the unit name
8539 given on the command line is a package name. The resulting executable
8540 will execute the elaboration routines of the package and its closure,
8541 then the finalization routines.
8544 @cindex @option{^-g^/DEBUG^} (@command{gnatmake})
8545 Enable debugging. This switch is simply passed to the compiler and to the
8551 @item @command{gcc} @asis{switches}
8553 Any uppercase or multi-character switch that is not a @command{gnatmake} switch
8554 is passed to @command{gcc} (e.g. @option{-O}, @option{-gnato,} etc.)
8557 Any qualifier that cannot be recognized as a qualifier for @code{GNAT MAKE}
8558 but is recognizable as a valid qualifier for @code{GNAT COMPILE} is
8559 automatically treated as a compiler switch, and passed on to all
8560 compilations that are carried out.
8565 Source and library search path switches:
8569 @item ^-aI^/SOURCE_SEARCH=^@var{dir}
8570 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatmake})
8571 When looking for source files also look in directory @var{dir}.
8572 The order in which source files search is undertaken is
8573 described in @ref{Search Paths and the Run-Time Library (RTL)}.
8575 @item ^-aL^/SKIP_MISSING=^@var{dir}
8576 @cindex @option{^-aL^/SKIP_MISSING^} (@command{gnatmake})
8577 Consider @var{dir} as being an externally provided Ada library.
8578 Instructs @command{gnatmake} to skip compilation units whose @file{.ALI}
8579 files have been located in directory @var{dir}. This allows you to have
8580 missing bodies for the units in @var{dir} and to ignore out of date bodies
8581 for the same units. You still need to specify
8582 the location of the specs for these units by using the switches
8583 @option{^-aI^/SOURCE_SEARCH=^@var{dir}}
8584 or @option{^-I^/SEARCH=^@var{dir}}.
8585 Note: this switch is provided for compatibility with previous versions
8586 of @command{gnatmake}. The easier method of causing standard libraries
8587 to be excluded from consideration is to write-protect the corresponding
8590 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
8591 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatmake})
8592 When searching for library and object files, look in directory
8593 @var{dir}. The order in which library files are searched is described in
8594 @ref{Search Paths for gnatbind}.
8596 @item ^-A^/CONDITIONAL_SOURCE_SEARCH=^@var{dir}
8597 @cindex Search paths, for @command{gnatmake}
8598 @cindex @option{^-A^/CONDITIONAL_SOURCE_SEARCH^} (@command{gnatmake})
8599 Equivalent to @option{^-aL^/SKIP_MISSING=^@var{dir}
8600 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
8602 @item ^-I^/SEARCH=^@var{dir}
8603 @cindex @option{^-I^/SEARCH^} (@command{gnatmake})
8604 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}
8605 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
8607 @item ^-I-^/NOCURRENT_DIRECTORY^
8608 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatmake})
8609 @cindex Source files, suppressing search
8610 Do not look for source files in the directory containing the source
8611 file named in the command line.
8612 Do not look for ALI or object files in the directory
8613 where @command{gnatmake} was invoked.
8615 @item ^-L^/LIBRARY_SEARCH=^@var{dir}
8616 @cindex @option{^-L^/LIBRARY_SEARCH^} (@command{gnatmake})
8617 @cindex Linker libraries
8618 Add directory @var{dir} to the list of directories in which the linker
8619 will search for libraries. This is equivalent to
8620 @option{-largs ^-L^/LIBRARY_SEARCH=^}@var{dir}.
8622 Furthermore, under Windows, the sources pointed to by the libraries path
8623 set in the registry are not searched for.
8627 @cindex @option{-nostdinc} (@command{gnatmake})
8628 Do not look for source files in the system default directory.
8631 @cindex @option{-nostdlib} (@command{gnatmake})
8632 Do not look for library files in the system default directory.
8634 @item --RTS=@var{rts-path}
8635 @cindex @option{--RTS} (@command{gnatmake})
8636 Specifies the default location of the runtime library. GNAT looks for the
8638 in the following directories, and stops as soon as a valid runtime is found
8639 (@file{adainclude} or @file{ada_source_path}, and @file{adalib} or
8640 @file{ada_object_path} present):
8643 @item <current directory>/$rts_path
8645 @item <default-search-dir>/$rts_path
8647 @item <default-search-dir>/rts-$rts_path
8651 The selected path is handled like a normal RTS path.
8655 @node Mode Switches for gnatmake
8656 @section Mode Switches for @command{gnatmake}
8659 The mode switches (referred to as @code{mode_switches}) allow the
8660 inclusion of switches that are to be passed to the compiler itself, the
8661 binder or the linker. The effect of a mode switch is to cause all
8662 subsequent switches up to the end of the switch list, or up to the next
8663 mode switch, to be interpreted as switches to be passed on to the
8664 designated component of GNAT.
8668 @item -cargs @var{switches}
8669 @cindex @option{-cargs} (@command{gnatmake})
8670 Compiler switches. Here @var{switches} is a list of switches
8671 that are valid switches for @command{gcc}. They will be passed on to
8672 all compile steps performed by @command{gnatmake}.
8674 @item -bargs @var{switches}
8675 @cindex @option{-bargs} (@command{gnatmake})
8676 Binder switches. Here @var{switches} is a list of switches
8677 that are valid switches for @code{gnatbind}. They will be passed on to
8678 all bind steps performed by @command{gnatmake}.
8680 @item -largs @var{switches}
8681 @cindex @option{-largs} (@command{gnatmake})
8682 Linker switches. Here @var{switches} is a list of switches
8683 that are valid switches for @command{gnatlink}. They will be passed on to
8684 all link steps performed by @command{gnatmake}.
8686 @item -margs @var{switches}
8687 @cindex @option{-margs} (@command{gnatmake})
8688 Make switches. The switches are directly interpreted by @command{gnatmake},
8689 regardless of any previous occurrence of @option{-cargs}, @option{-bargs}
8693 @node Notes on the Command Line
8694 @section Notes on the Command Line
8697 This section contains some additional useful notes on the operation
8698 of the @command{gnatmake} command.
8702 @cindex Recompilation, by @command{gnatmake}
8703 If @command{gnatmake} finds no ALI files, it recompiles the main program
8704 and all other units required by the main program.
8705 This means that @command{gnatmake}
8706 can be used for the initial compile, as well as during subsequent steps of
8707 the development cycle.
8710 If you enter @code{gnatmake @var{file}.adb}, where @file{@var{file}.adb}
8711 is a subunit or body of a generic unit, @command{gnatmake} recompiles
8712 @file{@var{file}.adb} (because it finds no ALI) and stops, issuing a
8716 In @command{gnatmake} the switch @option{^-I^/SEARCH^}
8717 is used to specify both source and
8718 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
8719 instead if you just want to specify
8720 source paths only and @option{^-aO^/OBJECT_SEARCH^}
8721 if you want to specify library paths
8725 @command{gnatmake} will ignore any files whose ALI file is write-protected.
8726 This may conveniently be used to exclude standard libraries from
8727 consideration and in particular it means that the use of the
8728 @option{^-f^/FORCE_COMPILE^} switch will not recompile these files
8729 unless @option{^-a^/ALL_FILES^} is also specified.
8732 @command{gnatmake} has been designed to make the use of Ada libraries
8733 particularly convenient. Assume you have an Ada library organized
8734 as follows: @i{^obj-dir^[OBJ_DIR]^} contains the objects and ALI files for
8735 of your Ada compilation units,
8736 whereas @i{^include-dir^[INCLUDE_DIR]^} contains the
8737 specs of these units, but no bodies. Then to compile a unit
8738 stored in @code{main.adb}, which uses this Ada library you would just type
8742 $ gnatmake -aI@var{include-dir} -aL@var{obj-dir} main
8745 $ gnatmake /SOURCE_SEARCH=@i{[INCLUDE_DIR]}
8746 /SKIP_MISSING=@i{[OBJ_DIR]} main
8751 Using @command{gnatmake} along with the
8752 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}
8753 switch provides a mechanism for avoiding unnecessary rcompilations. Using
8755 you can update the comments/format of your
8756 source files without having to recompile everything. Note, however, that
8757 adding or deleting lines in a source files may render its debugging
8758 info obsolete. If the file in question is a spec, the impact is rather
8759 limited, as that debugging info will only be useful during the
8760 elaboration phase of your program. For bodies the impact can be more
8761 significant. In all events, your debugger will warn you if a source file
8762 is more recent than the corresponding object, and alert you to the fact
8763 that the debugging information may be out of date.
8766 @node How gnatmake Works
8767 @section How @command{gnatmake} Works
8770 Generally @command{gnatmake} automatically performs all necessary
8771 recompilations and you don't need to worry about how it works. However,
8772 it may be useful to have some basic understanding of the @command{gnatmake}
8773 approach and in particular to understand how it uses the results of
8774 previous compilations without incorrectly depending on them.
8776 First a definition: an object file is considered @dfn{up to date} if the
8777 corresponding ALI file exists and if all the source files listed in the
8778 dependency section of this ALI file have time stamps matching those in
8779 the ALI file. This means that neither the source file itself nor any
8780 files that it depends on have been modified, and hence there is no need
8781 to recompile this file.
8783 @command{gnatmake} works by first checking if the specified main unit is up
8784 to date. If so, no compilations are required for the main unit. If not,
8785 @command{gnatmake} compiles the main program to build a new ALI file that
8786 reflects the latest sources. Then the ALI file of the main unit is
8787 examined to find all the source files on which the main program depends,
8788 and @command{gnatmake} recursively applies the above procedure on all these
8791 This process ensures that @command{gnatmake} only trusts the dependencies
8792 in an existing ALI file if they are known to be correct. Otherwise it
8793 always recompiles to determine a new, guaranteed accurate set of
8794 dependencies. As a result the program is compiled ``upside down'' from what may
8795 be more familiar as the required order of compilation in some other Ada
8796 systems. In particular, clients are compiled before the units on which
8797 they depend. The ability of GNAT to compile in any order is critical in
8798 allowing an order of compilation to be chosen that guarantees that
8799 @command{gnatmake} will recompute a correct set of new dependencies if
8802 When invoking @command{gnatmake} with several @var{file_names}, if a unit is
8803 imported by several of the executables, it will be recompiled at most once.
8805 Note: when using non-standard naming conventions
8806 (@pxref{Using Other File Names}), changing through a configuration pragmas
8807 file the version of a source and invoking @command{gnatmake} to recompile may
8808 have no effect, if the previous version of the source is still accessible
8809 by @command{gnatmake}. It may be necessary to use the switch
8810 ^-f^/FORCE_COMPILE^.
8812 @node Examples of gnatmake Usage
8813 @section Examples of @command{gnatmake} Usage
8816 @item gnatmake hello.adb
8817 Compile all files necessary to bind and link the main program
8818 @file{hello.adb} (containing unit @code{Hello}) and bind and link the
8819 resulting object files to generate an executable file @file{^hello^HELLO.EXE^}.
8821 @item gnatmake main1 main2 main3
8822 Compile all files necessary to bind and link the main programs
8823 @file{main1.adb} (containing unit @code{Main1}), @file{main2.adb}
8824 (containing unit @code{Main2}) and @file{main3.adb}
8825 (containing unit @code{Main3}) and bind and link the resulting object files
8826 to generate three executable files @file{^main1^MAIN1.EXE^},
8827 @file{^main2^MAIN2.EXE^}
8828 and @file{^main3^MAIN3.EXE^}.
8831 @item gnatmake -q Main_Unit -cargs -O2 -bargs -l
8835 @item gnatmake Main_Unit /QUIET
8836 /COMPILER_QUALIFIERS /OPTIMIZE=ALL
8837 /BINDER_QUALIFIERS /ORDER_OF_ELABORATION
8839 Compile all files necessary to bind and link the main program unit
8840 @code{Main_Unit} (from file @file{main_unit.adb}). All compilations will
8841 be done with optimization level 2 and the order of elaboration will be
8842 listed by the binder. @command{gnatmake} will operate in quiet mode, not
8843 displaying commands it is executing.
8846 @c *************************
8847 @node Improving Performance
8848 @chapter Improving Performance
8849 @cindex Improving performance
8852 This chapter presents several topics related to program performance.
8853 It first describes some of the tradeoffs that need to be considered
8854 and some of the techniques for making your program run faster.
8855 It then documents the @command{gnatelim} tool, which can reduce
8856 the size of program executables.
8860 * Performance Considerations::
8861 * Reducing the Size of Ada Executables with gnatelim::
8865 @c *****************************
8866 @node Performance Considerations
8867 @section Performance Considerations
8870 The GNAT system provides a number of options that allow a trade-off
8875 performance of the generated code
8878 speed of compilation
8881 minimization of dependences and recompilation
8884 the degree of run-time checking.
8888 The defaults (if no options are selected) aim at improving the speed
8889 of compilation and minimizing dependences, at the expense of performance
8890 of the generated code:
8897 no inlining of subprogram calls
8900 all run-time checks enabled except overflow and elaboration checks
8904 These options are suitable for most program development purposes. This
8905 chapter describes how you can modify these choices, and also provides
8906 some guidelines on debugging optimized code.
8909 * Controlling Run-Time Checks::
8910 * Use of Restrictions::
8911 * Optimization Levels::
8912 * Debugging Optimized Code::
8913 * Inlining of Subprograms::
8914 * Optimization and Strict Aliasing::
8916 * Coverage Analysis::
8920 @node Controlling Run-Time Checks
8921 @subsection Controlling Run-Time Checks
8924 By default, GNAT generates all run-time checks, except arithmetic overflow
8925 checking for integer operations and checks for access before elaboration on
8926 subprogram calls. The latter are not required in default mode, because all
8927 necessary checking is done at compile time.
8928 @cindex @option{-gnatp} (@command{gcc})
8929 @cindex @option{-gnato} (@command{gcc})
8930 Two gnat switches, @option{-gnatp} and @option{-gnato} allow this default to
8931 be modified. @xref{Run-Time Checks}.
8933 Our experience is that the default is suitable for most development
8936 We treat integer overflow specially because these
8937 are quite expensive and in our experience are not as important as other
8938 run-time checks in the development process. Note that division by zero
8939 is not considered an overflow check, and divide by zero checks are
8940 generated where required by default.
8942 Elaboration checks are off by default, and also not needed by default, since
8943 GNAT uses a static elaboration analysis approach that avoids the need for
8944 run-time checking. This manual contains a full chapter discussing the issue
8945 of elaboration checks, and if the default is not satisfactory for your use,
8946 you should read this chapter.
8948 For validity checks, the minimal checks required by the Ada Reference
8949 Manual (for case statements and assignments to array elements) are on
8950 by default. These can be suppressed by use of the @option{-gnatVn} switch.
8951 Note that in Ada 83, there were no validity checks, so if the Ada 83 mode
8952 is acceptable (or when comparing GNAT performance with an Ada 83 compiler),
8953 it may be reasonable to routinely use @option{-gnatVn}. Validity checks
8954 are also suppressed entirely if @option{-gnatp} is used.
8956 @cindex Overflow checks
8957 @cindex Checks, overflow
8960 @cindex pragma Suppress
8961 @cindex pragma Unsuppress
8962 Note that the setting of the switches controls the default setting of
8963 the checks. They may be modified using either @code{pragma Suppress} (to
8964 remove checks) or @code{pragma Unsuppress} (to add back suppressed
8965 checks) in the program source.
8967 @node Use of Restrictions
8968 @subsection Use of Restrictions
8971 The use of pragma Restrictions allows you to control which features are
8972 permitted in your program. Apart from the obvious point that if you avoid
8973 relatively expensive features like finalization (enforceable by the use
8974 of pragma Restrictions (No_Finalization), the use of this pragma does not
8975 affect the generated code in most cases.
8977 One notable exception to this rule is that the possibility of task abort
8978 results in some distributed overhead, particularly if finalization or
8979 exception handlers are used. The reason is that certain sections of code
8980 have to be marked as non-abortable.
8982 If you use neither the @code{abort} statement, nor asynchronous transfer
8983 of control (@code{select .. then abort}), then this distributed overhead
8984 is removed, which may have a general positive effect in improving
8985 overall performance. Especially code involving frequent use of tasking
8986 constructs and controlled types will show much improved performance.
8987 The relevant restrictions pragmas are
8990 pragma Restrictions (No_Abort_Statements);
8991 pragma Restrictions (Max_Asynchronous_Select_Nesting => 0);
8995 It is recommended that these restriction pragmas be used if possible. Note
8996 that this also means that you can write code without worrying about the
8997 possibility of an immediate abort at any point.
8999 @node Optimization Levels
9000 @subsection Optimization Levels
9001 @cindex @option{^-O^/OPTIMIZE^} (@command{gcc})
9004 The default is optimization off. This results in the fastest compile
9005 times, but GNAT makes absolutely no attempt to optimize, and the
9006 generated programs are considerably larger and slower than when
9007 optimization is enabled. You can use the
9009 @option{-O@var{n}} switch, where @var{n} is an integer from 0 to 3,
9012 @code{OPTIMIZE} qualifier
9014 to @command{gcc} to control the optimization level:
9017 @item ^-O0^/OPTIMIZE=NONE^
9018 No optimization (the default);
9019 generates unoptimized code but has
9020 the fastest compilation time.
9022 @item ^-O1^/OPTIMIZE=SOME^
9023 Medium level optimization;
9024 optimizes reasonably well but does not
9025 degrade compilation time significantly.
9027 @item ^-O2^/OPTIMIZE=ALL^
9029 @itemx /OPTIMIZE=DEVELOPMENT
9032 generates highly optimized code and has
9033 the slowest compilation time.
9035 @item ^-O3^/OPTIMIZE=INLINING^
9036 Full optimization as in @option{-O2},
9037 and also attempts automatic inlining of small
9038 subprograms within a unit (@pxref{Inlining of Subprograms}).
9042 Higher optimization levels perform more global transformations on the
9043 program and apply more expensive analysis algorithms in order to generate
9044 faster and more compact code. The price in compilation time, and the
9045 resulting improvement in execution time,
9046 both depend on the particular application and the hardware environment.
9047 You should experiment to find the best level for your application.
9049 Since the precise set of optimizations done at each level will vary from
9050 release to release (and sometime from target to target), it is best to think
9051 of the optimization settings in general terms.
9052 The @cite{Using GNU GCC} manual contains details about
9053 ^the @option{-O} settings and a number of @option{-f} options that^how to^
9054 individually enable or disable specific optimizations.
9056 Unlike some other compilation systems, ^@command{gcc}^GNAT^ has
9057 been tested extensively at all optimization levels. There are some bugs
9058 which appear only with optimization turned on, but there have also been
9059 bugs which show up only in @emph{unoptimized} code. Selecting a lower
9060 level of optimization does not improve the reliability of the code
9061 generator, which in practice is highly reliable at all optimization
9064 Note regarding the use of @option{-O3}: The use of this optimization level
9065 is generally discouraged with GNAT, since it often results in larger
9066 executables which run more slowly. See further discussion of this point
9067 in @ref{Inlining of Subprograms}.
9069 @node Debugging Optimized Code
9070 @subsection Debugging Optimized Code
9071 @cindex Debugging optimized code
9072 @cindex Optimization and debugging
9075 Although it is possible to do a reasonable amount of debugging at
9077 non-zero optimization levels,
9078 the higher the level the more likely that
9081 @option{/OPTIMIZE} settings other than @code{NONE},
9082 such settings will make it more likely that
9084 source-level constructs will have been eliminated by optimization.
9085 For example, if a loop is strength-reduced, the loop
9086 control variable may be completely eliminated and thus cannot be
9087 displayed in the debugger.
9088 This can only happen at @option{-O2} or @option{-O3}.
9089 Explicit temporary variables that you code might be eliminated at
9090 ^level^setting^ @option{-O1} or higher.
9092 The use of the @option{^-g^/DEBUG^} switch,
9093 @cindex @option{^-g^/DEBUG^} (@command{gcc})
9094 which is needed for source-level debugging,
9095 affects the size of the program executable on disk,
9096 and indeed the debugging information can be quite large.
9097 However, it has no effect on the generated code (and thus does not
9098 degrade performance)
9100 Since the compiler generates debugging tables for a compilation unit before
9101 it performs optimizations, the optimizing transformations may invalidate some
9102 of the debugging data. You therefore need to anticipate certain
9103 anomalous situations that may arise while debugging optimized code.
9104 These are the most common cases:
9108 @i{The ``hopping Program Counter'':} Repeated @code{step} or @code{next}
9110 the PC bouncing back and forth in the code. This may result from any of
9111 the following optimizations:
9115 @i{Common subexpression elimination:} using a single instance of code for a
9116 quantity that the source computes several times. As a result you
9117 may not be able to stop on what looks like a statement.
9120 @i{Invariant code motion:} moving an expression that does not change within a
9121 loop, to the beginning of the loop.
9124 @i{Instruction scheduling:} moving instructions so as to
9125 overlap loads and stores (typically) with other code, or in
9126 general to move computations of values closer to their uses. Often
9127 this causes you to pass an assignment statement without the assignment
9128 happening and then later bounce back to the statement when the
9129 value is actually needed. Placing a breakpoint on a line of code
9130 and then stepping over it may, therefore, not always cause all the
9131 expected side-effects.
9135 @i{The ``big leap'':} More commonly known as @emph{cross-jumping}, in which
9136 two identical pieces of code are merged and the program counter suddenly
9137 jumps to a statement that is not supposed to be executed, simply because
9138 it (and the code following) translates to the same thing as the code
9139 that @emph{was} supposed to be executed. This effect is typically seen in
9140 sequences that end in a jump, such as a @code{goto}, a @code{return}, or
9141 a @code{break} in a C @code{^switch^switch^} statement.
9144 @i{The ``roving variable'':} The symptom is an unexpected value in a variable.
9145 There are various reasons for this effect:
9149 In a subprogram prologue, a parameter may not yet have been moved to its
9153 A variable may be dead, and its register re-used. This is
9154 probably the most common cause.
9157 As mentioned above, the assignment of a value to a variable may
9161 A variable may be eliminated entirely by value propagation or
9162 other means. In this case, GCC may incorrectly generate debugging
9163 information for the variable
9167 In general, when an unexpected value appears for a local variable or parameter
9168 you should first ascertain if that value was actually computed by
9169 your program, as opposed to being incorrectly reported by the debugger.
9171 array elements in an object designated by an access value
9172 are generally less of a problem, once you have ascertained that the access
9174 Typically, this means checking variables in the preceding code and in the
9175 calling subprogram to verify that the value observed is explainable from other
9176 values (one must apply the procedure recursively to those
9177 other values); or re-running the code and stopping a little earlier
9178 (perhaps before the call) and stepping to better see how the variable obtained
9179 the value in question; or continuing to step @emph{from} the point of the
9180 strange value to see if code motion had simply moved the variable's
9185 In light of such anomalies, a recommended technique is to use @option{-O0}
9186 early in the software development cycle, when extensive debugging capabilities
9187 are most needed, and then move to @option{-O1} and later @option{-O2} as
9188 the debugger becomes less critical.
9189 Whether to use the @option{^-g^/DEBUG^} switch in the release version is
9190 a release management issue.
9192 Note that if you use @option{-g} you can then use the @command{strip} program
9193 on the resulting executable,
9194 which removes both debugging information and global symbols.
9197 @node Inlining of Subprograms
9198 @subsection Inlining of Subprograms
9201 A call to a subprogram in the current unit is inlined if all the
9202 following conditions are met:
9206 The optimization level is at least @option{-O1}.
9209 The called subprogram is suitable for inlining: It must be small enough
9210 and not contain nested subprograms or anything else that @command{gcc}
9211 cannot support in inlined subprograms.
9214 The call occurs after the definition of the body of the subprogram.
9217 @cindex pragma Inline
9219 Either @code{pragma Inline} applies to the subprogram or it is
9220 small and automatic inlining (optimization level @option{-O3}) is
9225 Calls to subprograms in @code{with}'ed units are normally not inlined.
9226 To achieve this level of inlining, the following conditions must all be
9231 The optimization level is at least @option{-O1}.
9234 The called subprogram is suitable for inlining: It must be small enough
9235 and not contain nested subprograms or anything else @command{gcc} cannot
9236 support in inlined subprograms.
9239 The call appears in a body (not in a package spec).
9242 There is a @code{pragma Inline} for the subprogram.
9245 @cindex @option{-gnatn} (@command{gcc})
9246 The @option{^-gnatn^/INLINE^} switch
9247 is used in the @command{gcc} command line
9250 Note that specifying the @option{-gnatn} switch causes additional
9251 compilation dependencies. Consider the following:
9253 @smallexample @c ada
9273 With the default behavior (no @option{-gnatn} switch specified), the
9274 compilation of the @code{Main} procedure depends only on its own source,
9275 @file{main.adb}, and the spec of the package in file @file{r.ads}. This
9276 means that editing the body of @code{R} does not require recompiling
9279 On the other hand, the call @code{R.Q} is not inlined under these
9280 circumstances. If the @option{-gnatn} switch is present when @code{Main}
9281 is compiled, the call will be inlined if the body of @code{Q} is small
9282 enough, but now @code{Main} depends on the body of @code{R} in
9283 @file{r.adb} as well as on the spec. This means that if this body is edited,
9284 the main program must be recompiled. Note that this extra dependency
9285 occurs whether or not the call is in fact inlined by @command{gcc}.
9287 The use of front end inlining with @option{-gnatN} generates similar
9288 additional dependencies.
9290 @cindex @option{^-fno-inline^/INLINE=SUPPRESS^} (@command{gcc})
9291 Note: The @option{^-fno-inline^/INLINE=SUPPRESS^} switch
9292 can be used to prevent
9293 all inlining. This switch overrides all other conditions and ensures
9294 that no inlining occurs. The extra dependences resulting from
9295 @option{-gnatn} will still be active, even if
9296 this switch is used to suppress the resulting inlining actions.
9298 Note regarding the use of @option{-O3}: There is no difference in inlining
9299 behavior between @option{-O2} and @option{-O3} for subprograms with an explicit
9300 pragma @code{Inline} assuming the use of @option{-gnatn}
9301 or @option{-gnatN} (the switches that activate inlining). If you have used
9302 pragma @code{Inline} in appropriate cases, then it is usually much better
9303 to use @option{-O2} and @option{-gnatn} and avoid the use of @option{-O3} which
9304 in this case only has the effect of inlining subprograms you did not
9305 think should be inlined. We often find that the use of @option{-O3} slows
9306 down code by performing excessive inlining, leading to increased instruction
9307 cache pressure from the increased code size. So the bottom line here is
9308 that you should not automatically assume that @option{-O3} is better than
9309 @option{-O2}, and indeed you should use @option{-O3} only if tests show that
9310 it actually improves performance.
9312 @node Optimization and Strict Aliasing
9313 @subsection Optimization and Strict Aliasing
9315 @cindex Strict Aliasing
9316 @cindex No_Strict_Aliasing
9319 The strong typing capabilities of Ada allow an optimizer to generate
9320 efficient code in situations where other languages would be forced to
9321 make worst case assumptions preventing such optimizations. Consider
9322 the following example:
9324 @smallexample @c ada
9327 type Int1 is new Integer;
9328 type Int2 is new Integer;
9329 type Int1A is access Int1;
9330 type Int2A is access Int2;
9337 for J in Data'Range loop
9338 if Data (J) = Int1V.all then
9339 Int2V.all := Int2V.all + 1;
9348 In this example, since the variable @code{Int1V} can only access objects
9349 of type @code{Int1}, and @code{Int2V} can only access objects of type
9350 @code{Int2}, there is no possibility that the assignment to
9351 @code{Int2V.all} affects the value of @code{Int1V.all}. This means that
9352 the compiler optimizer can "know" that the value @code{Int1V.all} is constant
9353 for all iterations of the loop and avoid the extra memory reference
9354 required to dereference it each time through the loop.
9356 This kind of optimization, called strict aliasing analysis, is
9357 triggered by specifying an optimization level of @option{-O2} or
9358 higher and allows @code{GNAT} to generate more efficient code
9359 when access values are involved.
9361 However, although this optimization is always correct in terms of
9362 the formal semantics of the Ada Reference Manual, difficulties can
9363 arise if features like @code{Unchecked_Conversion} are used to break
9364 the typing system. Consider the following complete program example:
9366 @smallexample @c ada
9369 type int1 is new integer;
9370 type int2 is new integer;
9371 type a1 is access int1;
9372 type a2 is access int2;
9377 function to_a2 (Input : a1) return a2;
9380 with Unchecked_Conversion;
9382 function to_a2 (Input : a1) return a2 is
9384 new Unchecked_Conversion (a1, a2);
9386 return to_a2u (Input);
9392 with Text_IO; use Text_IO;
9394 v1 : a1 := new int1;
9395 v2 : a2 := to_a2 (v1);
9399 put_line (int1'image (v1.all));
9405 This program prints out 0 in @code{-O0} or @code{-O1}
9406 mode, but it prints out 1 in @code{-O2} mode. That's
9407 because in strict aliasing mode, the compiler can and
9408 does assume that the assignment to @code{v2.all} could not
9409 affect the value of @code{v1.all}, since different types
9412 This behavior is not a case of non-conformance with the standard, since
9413 the Ada RM specifies that an unchecked conversion where the resulting
9414 bit pattern is not a correct value of the target type can result in an
9415 abnormal value and attempting to reference an abnormal value makes the
9416 execution of a program erroneous. That's the case here since the result
9417 does not point to an object of type @code{int2}. This means that the
9418 effect is entirely unpredictable.
9420 However, although that explanation may satisfy a language
9421 lawyer, in practice an applications programmer expects an
9422 unchecked conversion involving pointers to create true
9423 aliases and the behavior of printing 1 seems plain wrong.
9424 In this case, the strict aliasing optimization is unwelcome.
9426 Indeed the compiler recognizes this possibility, and the
9427 unchecked conversion generates a warning:
9430 p2.adb:5:07: warning: possible aliasing problem with type "a2"
9431 p2.adb:5:07: warning: use -fno-strict-aliasing switch for references
9432 p2.adb:5:07: warning: or use "pragma No_Strict_Aliasing (a2);"
9436 Unfortunately the problem is recognized when compiling the body of
9437 package @code{p2}, but the actual "bad" code is generated while
9438 compiling the body of @code{m} and this latter compilation does not see
9439 the suspicious @code{Unchecked_Conversion}.
9441 As implied by the warning message, there are approaches you can use to
9442 avoid the unwanted strict aliasing optimization in a case like this.
9444 One possibility is to simply avoid the use of @code{-O2}, but
9445 that is a bit drastic, since it throws away a number of useful
9446 optimizations that do not involve strict aliasing assumptions.
9448 A less drastic approach is to compile the program using the
9449 option @code{-fno-strict-aliasing}. Actually it is only the
9450 unit containing the dereferencing of the suspicious pointer
9451 that needs to be compiled. So in this case, if we compile
9452 unit @code{m} with this switch, then we get the expected
9453 value of zero printed. Analyzing which units might need
9454 the switch can be painful, so a more reasonable approach
9455 is to compile the entire program with options @code{-O2}
9456 and @code{-fno-strict-aliasing}. If the performance is
9457 satisfactory with this combination of options, then the
9458 advantage is that the entire issue of possible "wrong"
9459 optimization due to strict aliasing is avoided.
9461 To avoid the use of compiler switches, the configuration
9462 pragma @code{No_Strict_Aliasing} with no parameters may be
9463 used to specify that for all access types, the strict
9464 aliasing optimization should be suppressed.
9466 However, these approaches are still overkill, in that they causes
9467 all manipulations of all access values to be deoptimized. A more
9468 refined approach is to concentrate attention on the specific
9469 access type identified as problematic.
9471 First, if a careful analysis of uses of the pointer shows
9472 that there are no possible problematic references, then
9473 the warning can be suppressed by bracketing the
9474 instantiation of @code{Unchecked_Conversion} to turn
9477 @smallexample @c ada
9478 pragma Warnings (Off);
9480 new Unchecked_Conversion (a1, a2);
9481 pragma Warnings (On);
9485 Of course that approach is not appropriate for this particular
9486 example, since indeed there is a problematic reference. In this
9487 case we can take one of two other approaches.
9489 The first possibility is to move the instantiation of unchecked
9490 conversion to the unit in which the type is declared. In
9491 this example, we would move the instantiation of
9492 @code{Unchecked_Conversion} from the body of package
9493 @code{p2} to the spec of package @code{p1}. Now the
9494 warning disappears. That's because any use of the
9495 access type knows there is a suspicious unchecked
9496 conversion, and the strict aliasing optimization
9497 is automatically suppressed for the type.
9499 If it is not practical to move the unchecked conversion to the same unit
9500 in which the destination access type is declared (perhaps because the
9501 source type is not visible in that unit), you may use pragma
9502 @code{No_Strict_Aliasing} for the type. This pragma must occur in the
9503 same declarative sequence as the declaration of the access type:
9505 @smallexample @c ada
9506 type a2 is access int2;
9507 pragma No_Strict_Aliasing (a2);
9511 Here again, the compiler now knows that the strict aliasing optimization
9512 should be suppressed for any reference to type @code{a2} and the
9513 expected behavior is obtained.
9515 Finally, note that although the compiler can generate warnings for
9516 simple cases of unchecked conversions, there are tricker and more
9517 indirect ways of creating type incorrect aliases which the compiler
9518 cannot detect. Examples are the use of address overlays and unchecked
9519 conversions involving composite types containing access types as
9520 components. In such cases, no warnings are generated, but there can
9521 still be aliasing problems. One safe coding practice is to forbid the
9522 use of address clauses for type overlaying, and to allow unchecked
9523 conversion only for primitive types. This is not really a significant
9524 restriction since any possible desired effect can be achieved by
9525 unchecked conversion of access values.
9528 @node Coverage Analysis
9529 @subsection Coverage Analysis
9532 GNAT supports the Digital Performance Coverage Analyzer (PCA), which allows
9533 the user to determine the distribution of execution time across a program,
9534 @pxref{Profiling} for details of usage.
9537 @node Reducing the Size of Ada Executables with gnatelim
9538 @section Reducing the Size of Ada Executables with @code{gnatelim}
9542 This section describes @command{gnatelim}, a tool which detects unused
9543 subprograms and helps the compiler to create a smaller executable for your
9548 * Running gnatelim::
9549 * Correcting the List of Eliminate Pragmas::
9550 * Making Your Executables Smaller::
9551 * Summary of the gnatelim Usage Cycle::
9554 @node About gnatelim
9555 @subsection About @code{gnatelim}
9558 When a program shares a set of Ada
9559 packages with other programs, it may happen that this program uses
9560 only a fraction of the subprograms defined in these packages. The code
9561 created for these unused subprograms increases the size of the executable.
9563 @code{gnatelim} tracks unused subprograms in an Ada program and
9564 outputs a list of GNAT-specific pragmas @code{Eliminate} marking all the
9565 subprograms that are declared but never called. By placing the list of
9566 @code{Eliminate} pragmas in the GNAT configuration file @file{gnat.adc} and
9567 recompiling your program, you may decrease the size of its executable,
9568 because the compiler will not generate the code for 'eliminated' subprograms.
9569 See GNAT Reference Manual for more information about this pragma.
9571 @code{gnatelim} needs as its input data the name of the main subprogram
9572 and a bind file for a main subprogram.
9574 To create a bind file for @code{gnatelim}, run @code{gnatbind} for
9575 the main subprogram. @code{gnatelim} can work with both Ada and C
9576 bind files; when both are present, it uses the Ada bind file.
9577 The following commands will build the program and create the bind file:
9580 $ gnatmake ^-c Main_Prog^/ACTIONS=COMPILE MAIN_PROG^
9581 $ gnatbind main_prog
9584 Note that @code{gnatelim} needs neither object nor ALI files.
9586 @node Running gnatelim
9587 @subsection Running @code{gnatelim}
9590 @code{gnatelim} has the following command-line interface:
9593 $ gnatelim [options] name
9597 @code{name} should be a name of a source file that contains the main subprogram
9598 of a program (partition).
9600 @code{gnatelim} has the following switches:
9605 @cindex @option{^-q^/QUIET^} (@command{gnatelim})
9606 Quiet mode: by default @code{gnatelim} outputs to the standard error
9607 stream the number of program units left to be processed. This option turns
9611 @cindex @option{^-v^/VERBOSE^} (@command{gnatelim})
9612 Verbose mode: @code{gnatelim} version information is printed as Ada
9613 comments to the standard output stream. Also, in addition to the number of
9614 program units left @code{gnatelim} will output the name of the current unit
9618 @cindex @option{^-a^/ALL^} (@command{gnatelim})
9619 Also look for subprograms from the GNAT run time that can be eliminated. Note
9620 that when @file{gnat.adc} is produced using this switch, the entire program
9621 must be recompiled with switch @option{^-a^/ALL_FILES^} to @command{gnatmake}.
9623 @item ^-I^/INCLUDE_DIRS=^@var{dir}
9624 @cindex @option{^-I^/INCLUDE_DIRS^} (@command{gnatelim})
9625 When looking for source files also look in directory @var{dir}. Specifying
9626 @option{^-I-^/INCLUDE_DIRS=-^} instructs @code{gnatelim} not to look for
9627 sources in the current directory.
9629 @item ^-b^/BIND_FILE=^@var{bind_file}
9630 @cindex @option{^-b^/BIND_FILE^} (@command{gnatelim})
9631 Specifies @var{bind_file} as the bind file to process. If not set, the name
9632 of the bind file is computed from the full expanded Ada name
9633 of a main subprogram.
9635 @item ^-C^/CONFIG_FILE=^@var{config_file}
9636 @cindex @option{^-C^/CONFIG_FILE^} (@command{gnatelim})
9637 Specifies a file @var{config_file} that contains configuration pragmas. The
9638 file must be specified with full path.
9640 @item ^--GCC^/COMPILER^=@var{compiler_name}
9641 @cindex @option{^-GCC^/COMPILER^} (@command{gnatelim})
9642 Instructs @code{gnatelim} to use specific @command{gcc} compiler instead of one
9643 available on the path.
9645 @item ^--GNATMAKE^/GNATMAKE^=@var{gnatmake_name}
9646 @cindex @option{^--GNATMAKE^/GNATMAKE^} (@command{gnatelim})
9647 Instructs @code{gnatelim} to use specific @command{gnatmake} instead of one
9648 available on the path.
9652 @code{gnatelim} sends its output to the standard output stream, and all the
9653 tracing and debug information is sent to the standard error stream.
9654 In order to produce a proper GNAT configuration file
9655 @file{gnat.adc}, redirection must be used:
9659 $ PIPE GNAT ELIM MAIN_PROG.ADB > GNAT.ADC
9662 $ gnatelim main_prog.adb > gnat.adc
9671 $ gnatelim main_prog.adb >> gnat.adc
9675 in order to append the @code{gnatelim} output to the existing contents of
9679 @node Correcting the List of Eliminate Pragmas
9680 @subsection Correcting the List of Eliminate Pragmas
9683 In some rare cases @code{gnatelim} may try to eliminate
9684 subprograms that are actually called in the program. In this case, the
9685 compiler will generate an error message of the form:
9688 file.adb:106:07: cannot call eliminated subprogram "My_Prog"
9692 You will need to manually remove the wrong @code{Eliminate} pragmas from
9693 the @file{gnat.adc} file. You should recompile your program
9694 from scratch after that, because you need a consistent @file{gnat.adc} file
9695 during the entire compilation.
9697 @node Making Your Executables Smaller
9698 @subsection Making Your Executables Smaller
9701 In order to get a smaller executable for your program you now have to
9702 recompile the program completely with the new @file{gnat.adc} file
9703 created by @code{gnatelim} in your current directory:
9706 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
9710 (Use the @option{^-f^/FORCE_COMPILE^} option for @command{gnatmake} to
9711 recompile everything
9712 with the set of pragmas @code{Eliminate} that you have obtained with
9713 @command{gnatelim}).
9715 Be aware that the set of @code{Eliminate} pragmas is specific to each
9716 program. It is not recommended to merge sets of @code{Eliminate}
9717 pragmas created for different programs in one @file{gnat.adc} file.
9719 @node Summary of the gnatelim Usage Cycle
9720 @subsection Summary of the gnatelim Usage Cycle
9723 Here is a quick summary of the steps to be taken in order to reduce
9724 the size of your executables with @code{gnatelim}. You may use
9725 other GNAT options to control the optimization level,
9726 to produce the debugging information, to set search path, etc.
9733 $ gnatmake ^-c main_prog^/ACTIONS=COMPILE MAIN_PROG^
9734 $ gnatbind main_prog
9738 Generate a list of @code{Eliminate} pragmas
9741 $ PIPE GNAT ELIM MAIN_PROG > GNAT.ADC
9744 $ gnatelim main_prog >[>] gnat.adc
9749 Recompile the application
9752 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
9757 @c ********************************
9758 @node Renaming Files Using gnatchop
9759 @chapter Renaming Files Using @code{gnatchop}
9763 This chapter discusses how to handle files with multiple units by using
9764 the @code{gnatchop} utility. This utility is also useful in renaming
9765 files to meet the standard GNAT default file naming conventions.
9768 * Handling Files with Multiple Units::
9769 * Operating gnatchop in Compilation Mode::
9770 * Command Line for gnatchop::
9771 * Switches for gnatchop::
9772 * Examples of gnatchop Usage::
9775 @node Handling Files with Multiple Units
9776 @section Handling Files with Multiple Units
9779 The basic compilation model of GNAT requires that a file submitted to the
9780 compiler have only one unit and there be a strict correspondence
9781 between the file name and the unit name.
9783 The @code{gnatchop} utility allows both of these rules to be relaxed,
9784 allowing GNAT to process files which contain multiple compilation units
9785 and files with arbitrary file names. @code{gnatchop}
9786 reads the specified file and generates one or more output files,
9787 containing one unit per file. The unit and the file name correspond,
9788 as required by GNAT.
9790 If you want to permanently restructure a set of ``foreign'' files so that
9791 they match the GNAT rules, and do the remaining development using the
9792 GNAT structure, you can simply use @command{gnatchop} once, generate the
9793 new set of files and work with them from that point on.
9795 Alternatively, if you want to keep your files in the ``foreign'' format,
9796 perhaps to maintain compatibility with some other Ada compilation
9797 system, you can set up a procedure where you use @command{gnatchop} each
9798 time you compile, regarding the source files that it writes as temporary
9799 files that you throw away.
9801 @node Operating gnatchop in Compilation Mode
9802 @section Operating gnatchop in Compilation Mode
9805 The basic function of @code{gnatchop} is to take a file with multiple units
9806 and split it into separate files. The boundary between files is reasonably
9807 clear, except for the issue of comments and pragmas. In default mode, the
9808 rule is that any pragmas between units belong to the previous unit, except
9809 that configuration pragmas always belong to the following unit. Any comments
9810 belong to the following unit. These rules
9811 almost always result in the right choice of
9812 the split point without needing to mark it explicitly and most users will
9813 find this default to be what they want. In this default mode it is incorrect to
9814 submit a file containing only configuration pragmas, or one that ends in
9815 configuration pragmas, to @code{gnatchop}.
9817 However, using a special option to activate ``compilation mode'',
9819 can perform another function, which is to provide exactly the semantics
9820 required by the RM for handling of configuration pragmas in a compilation.
9821 In the absence of configuration pragmas (at the main file level), this
9822 option has no effect, but it causes such configuration pragmas to be handled
9823 in a quite different manner.
9825 First, in compilation mode, if @code{gnatchop} is given a file that consists of
9826 only configuration pragmas, then this file is appended to the
9827 @file{gnat.adc} file in the current directory. This behavior provides
9828 the required behavior described in the RM for the actions to be taken
9829 on submitting such a file to the compiler, namely that these pragmas
9830 should apply to all subsequent compilations in the same compilation
9831 environment. Using GNAT, the current directory, possibly containing a
9832 @file{gnat.adc} file is the representation
9833 of a compilation environment. For more information on the
9834 @file{gnat.adc} file, see @ref{Handling of Configuration Pragmas}.
9836 Second, in compilation mode, if @code{gnatchop}
9837 is given a file that starts with
9838 configuration pragmas, and contains one or more units, then these
9839 configuration pragmas are prepended to each of the chopped files. This
9840 behavior provides the required behavior described in the RM for the
9841 actions to be taken on compiling such a file, namely that the pragmas
9842 apply to all units in the compilation, but not to subsequently compiled
9845 Finally, if configuration pragmas appear between units, they are appended
9846 to the previous unit. This results in the previous unit being illegal,
9847 since the compiler does not accept configuration pragmas that follow
9848 a unit. This provides the required RM behavior that forbids configuration
9849 pragmas other than those preceding the first compilation unit of a
9852 For most purposes, @code{gnatchop} will be used in default mode. The
9853 compilation mode described above is used only if you need exactly
9854 accurate behavior with respect to compilations, and you have files
9855 that contain multiple units and configuration pragmas. In this
9856 circumstance the use of @code{gnatchop} with the compilation mode
9857 switch provides the required behavior, and is for example the mode
9858 in which GNAT processes the ACVC tests.
9860 @node Command Line for gnatchop
9861 @section Command Line for @code{gnatchop}
9864 The @code{gnatchop} command has the form:
9867 $ gnatchop switches @var{file name} [@var{file name} @var{file name} ...]
9872 The only required argument is the file name of the file to be chopped.
9873 There are no restrictions on the form of this file name. The file itself
9874 contains one or more Ada units, in normal GNAT format, concatenated
9875 together. As shown, more than one file may be presented to be chopped.
9877 When run in default mode, @code{gnatchop} generates one output file in
9878 the current directory for each unit in each of the files.
9880 @var{directory}, if specified, gives the name of the directory to which
9881 the output files will be written. If it is not specified, all files are
9882 written to the current directory.
9884 For example, given a
9885 file called @file{hellofiles} containing
9887 @smallexample @c ada
9892 with Text_IO; use Text_IO;
9905 $ gnatchop ^hellofiles^HELLOFILES.^
9909 generates two files in the current directory, one called
9910 @file{hello.ads} containing the single line that is the procedure spec,
9911 and the other called @file{hello.adb} containing the remaining text. The
9912 original file is not affected. The generated files can be compiled in
9916 When gnatchop is invoked on a file that is empty or that contains only empty
9917 lines and/or comments, gnatchop will not fail, but will not produce any
9920 For example, given a
9921 file called @file{toto.txt} containing
9923 @smallexample @c ada
9935 $ gnatchop ^toto.txt^TOT.TXT^
9939 will not produce any new file and will result in the following warnings:
9942 toto.txt:1:01: warning: empty file, contains no compilation units
9943 no compilation units found
9944 no source files written
9947 @node Switches for gnatchop
9948 @section Switches for @code{gnatchop}
9951 @command{gnatchop} recognizes the following switches:
9956 @item ^-c^/COMPILATION^
9957 @cindex @option{^-c^/COMPILATION^} (@code{gnatchop})
9958 Causes @code{gnatchop} to operate in compilation mode, in which
9959 configuration pragmas are handled according to strict RM rules. See
9960 previous section for a full description of this mode.
9964 This passes the given @option{-gnatxxx} switch to @code{gnat} which is
9965 used to parse the given file. Not all @code{xxx} options make sense,
9966 but for example, the use of @option{-gnati2} allows @code{gnatchop} to
9967 process a source file that uses Latin-2 coding for identifiers.
9971 Causes @code{gnatchop} to generate a brief help summary to the standard
9972 output file showing usage information.
9974 @item ^-k@var{mm}^/FILE_NAME_MAX_LENGTH=@var{mm}^
9975 @cindex @option{^-k^/FILE_NAME_MAX_LENGTH^} (@code{gnatchop})
9976 Limit generated file names to the specified number @code{mm}
9978 This is useful if the
9979 resulting set of files is required to be interoperable with systems
9980 which limit the length of file names.
9982 If no value is given, or
9983 if no @code{/FILE_NAME_MAX_LENGTH} qualifier is given,
9984 a default of 39, suitable for OpenVMS Alpha
9988 No space is allowed between the @option{-k} and the numeric value. The numeric
9989 value may be omitted in which case a default of @option{-k8},
9991 with DOS-like file systems, is used. If no @option{-k} switch
9993 there is no limit on the length of file names.
9996 @item ^-p^/PRESERVE^
9997 @cindex @option{^-p^/PRESERVE^} (@code{gnatchop})
9998 Causes the file ^modification^creation^ time stamp of the input file to be
9999 preserved and used for the time stamp of the output file(s). This may be
10000 useful for preserving coherency of time stamps in an environment where
10001 @code{gnatchop} is used as part of a standard build process.
10004 @cindex @option{^-q^/QUIET^} (@code{gnatchop})
10005 Causes output of informational messages indicating the set of generated
10006 files to be suppressed. Warnings and error messages are unaffected.
10008 @item ^-r^/REFERENCE^
10009 @cindex @option{^-r^/REFERENCE^} (@code{gnatchop})
10010 @findex Source_Reference
10011 Generate @code{Source_Reference} pragmas. Use this switch if the output
10012 files are regarded as temporary and development is to be done in terms
10013 of the original unchopped file. This switch causes
10014 @code{Source_Reference} pragmas to be inserted into each of the
10015 generated files to refers back to the original file name and line number.
10016 The result is that all error messages refer back to the original
10018 In addition, the debugging information placed into the object file (when
10019 the @option{^-g^/DEBUG^} switch of @command{gcc} or @command{gnatmake} is
10021 also refers back to this original file so that tools like profilers and
10022 debuggers will give information in terms of the original unchopped file.
10024 If the original file to be chopped itself contains
10025 a @code{Source_Reference}
10026 pragma referencing a third file, then gnatchop respects
10027 this pragma, and the generated @code{Source_Reference} pragmas
10028 in the chopped file refer to the original file, with appropriate
10029 line numbers. This is particularly useful when @code{gnatchop}
10030 is used in conjunction with @code{gnatprep} to compile files that
10031 contain preprocessing statements and multiple units.
10033 @item ^-v^/VERBOSE^
10034 @cindex @option{^-v^/VERBOSE^} (@code{gnatchop})
10035 Causes @code{gnatchop} to operate in verbose mode. The version
10036 number and copyright notice are output, as well as exact copies of
10037 the gnat1 commands spawned to obtain the chop control information.
10039 @item ^-w^/OVERWRITE^
10040 @cindex @option{^-w^/OVERWRITE^} (@code{gnatchop})
10041 Overwrite existing file names. Normally @code{gnatchop} regards it as a
10042 fatal error if there is already a file with the same name as a
10043 file it would otherwise output, in other words if the files to be
10044 chopped contain duplicated units. This switch bypasses this
10045 check, and causes all but the last instance of such duplicated
10046 units to be skipped.
10050 @cindex @option{--GCC=} (@code{gnatchop})
10051 Specify the path of the GNAT parser to be used. When this switch is used,
10052 no attempt is made to add the prefix to the GNAT parser executable.
10056 @node Examples of gnatchop Usage
10057 @section Examples of @code{gnatchop} Usage
10061 @item gnatchop /OVERWRITE HELLO_S.ADA [PRERELEASE.FILES]
10064 @item gnatchop -w hello_s.ada prerelease/files
10067 Chops the source file @file{hello_s.ada}. The output files will be
10068 placed in the directory @file{^prerelease/files^[PRERELEASE.FILES]^},
10070 files with matching names in that directory (no files in the current
10071 directory are modified).
10073 @item gnatchop ^archive^ARCHIVE.^
10074 Chops the source file @file{^archive^ARCHIVE.^}
10075 into the current directory. One
10076 useful application of @code{gnatchop} is in sending sets of sources
10077 around, for example in email messages. The required sources are simply
10078 concatenated (for example, using a ^Unix @code{cat}^VMS @code{APPEND/NEW}^
10080 @code{gnatchop} is used at the other end to reconstitute the original
10083 @item gnatchop file1 file2 file3 direc
10084 Chops all units in files @file{file1}, @file{file2}, @file{file3}, placing
10085 the resulting files in the directory @file{direc}. Note that if any units
10086 occur more than once anywhere within this set of files, an error message
10087 is generated, and no files are written. To override this check, use the
10088 @option{^-w^/OVERWRITE^} switch,
10089 in which case the last occurrence in the last file will
10090 be the one that is output, and earlier duplicate occurrences for a given
10091 unit will be skipped.
10094 @node Configuration Pragmas
10095 @chapter Configuration Pragmas
10096 @cindex Configuration pragmas
10097 @cindex Pragmas, configuration
10100 In Ada 95, configuration pragmas include those pragmas described as
10101 such in the Ada 95 Reference Manual, as well as
10102 implementation-dependent pragmas that are configuration pragmas. See the
10103 individual descriptions of pragmas in the GNAT Reference Manual for
10104 details on these additional GNAT-specific configuration pragmas. Most
10105 notably, the pragma @code{Source_File_Name}, which allows
10106 specifying non-default names for source files, is a configuration
10107 pragma. The following is a complete list of configuration pragmas
10108 recognized by @code{GNAT}:
10115 Component_Alignment
10121 External_Name_Casing
10122 Float_Representation
10131 Propagate_Exceptions
10134 Restricted_Run_Time
10136 Restrictions_Warnings
10141 Task_Dispatching_Policy
10150 * Handling of Configuration Pragmas::
10151 * The Configuration Pragmas Files::
10154 @node Handling of Configuration Pragmas
10155 @section Handling of Configuration Pragmas
10157 Configuration pragmas may either appear at the start of a compilation
10158 unit, in which case they apply only to that unit, or they may apply to
10159 all compilations performed in a given compilation environment.
10161 GNAT also provides the @code{gnatchop} utility to provide an automatic
10162 way to handle configuration pragmas following the semantics for
10163 compilations (that is, files with multiple units), described in the RM.
10164 See @ref{Operating gnatchop in Compilation Mode} for details.
10165 However, for most purposes, it will be more convenient to edit the
10166 @file{gnat.adc} file that contains configuration pragmas directly,
10167 as described in the following section.
10169 @node The Configuration Pragmas Files
10170 @section The Configuration Pragmas Files
10171 @cindex @file{gnat.adc}
10174 In GNAT a compilation environment is defined by the current
10175 directory at the time that a compile command is given. This current
10176 directory is searched for a file whose name is @file{gnat.adc}. If
10177 this file is present, it is expected to contain one or more
10178 configuration pragmas that will be applied to the current compilation.
10179 However, if the switch @option{-gnatA} is used, @file{gnat.adc} is not
10182 Configuration pragmas may be entered into the @file{gnat.adc} file
10183 either by running @code{gnatchop} on a source file that consists only of
10184 configuration pragmas, or more conveniently by
10185 direct editing of the @file{gnat.adc} file, which is a standard format
10188 In addition to @file{gnat.adc}, one additional file containing configuration
10189 pragmas may be applied to the current compilation using the switch
10190 @option{-gnatec}@var{path}. @var{path} must designate an existing file that
10191 contains only configuration pragmas. These configuration pragmas are
10192 in addition to those found in @file{gnat.adc} (provided @file{gnat.adc}
10193 is present and switch @option{-gnatA} is not used).
10195 It is allowed to specify several switches @option{-gnatec}, however only
10196 the last one on the command line will be taken into account.
10198 If you are using project file, a separate mechanism is provided using
10199 project attributes, see @ref{Specifying Configuration Pragmas} for more
10203 Of special interest to GNAT OpenVMS Alpha is the following
10204 configuration pragma:
10206 @smallexample @c ada
10208 pragma Extend_System (Aux_DEC);
10213 In the presence of this pragma, GNAT adds to the definition of the
10214 predefined package SYSTEM all the additional types and subprograms that are
10215 defined in DEC Ada. See @ref{Compatibility with DEC Ada} for details.
10218 @node Handling Arbitrary File Naming Conventions Using gnatname
10219 @chapter Handling Arbitrary File Naming Conventions Using @code{gnatname}
10220 @cindex Arbitrary File Naming Conventions
10223 * Arbitrary File Naming Conventions::
10224 * Running gnatname::
10225 * Switches for gnatname::
10226 * Examples of gnatname Usage::
10229 @node Arbitrary File Naming Conventions
10230 @section Arbitrary File Naming Conventions
10233 The GNAT compiler must be able to know the source file name of a compilation
10234 unit. When using the standard GNAT default file naming conventions
10235 (@code{.ads} for specs, @code{.adb} for bodies), the GNAT compiler
10236 does not need additional information.
10239 When the source file names do not follow the standard GNAT default file naming
10240 conventions, the GNAT compiler must be given additional information through
10241 a configuration pragmas file (@pxref{Configuration Pragmas})
10243 When the non standard file naming conventions are well-defined,
10244 a small number of pragmas @code{Source_File_Name} specifying a naming pattern
10245 (@pxref{Alternative File Naming Schemes}) may be sufficient. However,
10246 if the file naming conventions are irregular or arbitrary, a number
10247 of pragma @code{Source_File_Name} for individual compilation units
10249 To help maintain the correspondence between compilation unit names and
10250 source file names within the compiler,
10251 GNAT provides a tool @code{gnatname} to generate the required pragmas for a
10254 @node Running gnatname
10255 @section Running @code{gnatname}
10258 The usual form of the @code{gnatname} command is
10261 $ gnatname [@var{switches}] @var{naming_pattern} [@var{naming_patterns}]
10265 All of the arguments are optional. If invoked without any argument,
10266 @code{gnatname} will display its usage.
10269 When used with at least one naming pattern, @code{gnatname} will attempt to
10270 find all the compilation units in files that follow at least one of the
10271 naming patterns. To find these compilation units,
10272 @code{gnatname} will use the GNAT compiler in syntax-check-only mode on all
10276 One or several Naming Patterns may be given as arguments to @code{gnatname}.
10277 Each Naming Pattern is enclosed between double quotes.
10278 A Naming Pattern is a regular expression similar to the wildcard patterns
10279 used in file names by the Unix shells or the DOS prompt.
10282 Examples of Naming Patterns are
10291 For a more complete description of the syntax of Naming Patterns,
10292 see the second kind of regular expressions described in @file{g-regexp.ads}
10293 (the ``Glob'' regular expressions).
10296 When invoked with no switches, @code{gnatname} will create a configuration
10297 pragmas file @file{gnat.adc} in the current working directory, with pragmas
10298 @code{Source_File_Name} for each file that contains a valid Ada unit.
10300 @node Switches for gnatname
10301 @section Switches for @code{gnatname}
10304 Switches for @code{gnatname} must precede any specified Naming Pattern.
10307 You may specify any of the following switches to @code{gnatname}:
10312 @item ^-c^/CONFIG_FILE=^@file{file}
10313 @cindex @option{^-c^/CONFIG_FILE^} (@code{gnatname})
10314 Create a configuration pragmas file @file{file} (instead of the default
10317 There may be zero, one or more space between @option{-c} and
10320 @file{file} may include directory information. @file{file} must be
10321 writable. There may be only one switch @option{^-c^/CONFIG_FILE^}.
10322 When a switch @option{^-c^/CONFIG_FILE^} is
10323 specified, no switch @option{^-P^/PROJECT_FILE^} may be specified (see below).
10325 @item ^-d^/SOURCE_DIRS=^@file{dir}
10326 @cindex @option{^-d^/SOURCE_DIRS^} (@code{gnatname})
10327 Look for source files in directory @file{dir}. There may be zero, one or more
10328 spaces between @option{^-d^/SOURCE_DIRS=^} and @file{dir}.
10329 When a switch @option{^-d^/SOURCE_DIRS^}
10330 is specified, the current working directory will not be searched for source
10331 files, unless it is explicitly specified with a @option{^-d^/SOURCE_DIRS^}
10332 or @option{^-D^/DIR_FILES^} switch.
10333 Several switches @option{^-d^/SOURCE_DIRS^} may be specified.
10334 If @file{dir} is a relative path, it is relative to the directory of
10335 the configuration pragmas file specified with switch
10336 @option{^-c^/CONFIG_FILE^},
10337 or to the directory of the project file specified with switch
10338 @option{^-P^/PROJECT_FILE^} or,
10339 if neither switch @option{^-c^/CONFIG_FILE^}
10340 nor switch @option{^-P^/PROJECT_FILE^} are specified, it is relative to the
10341 current working directory. The directory
10342 specified with switch @option{^-d^/SOURCE_DIRS^} must exist and be readable.
10344 @item ^-D^/DIRS_FILE=^@file{file}
10345 @cindex @option{^-D^/DIRS_FILE^} (@code{gnatname})
10346 Look for source files in all directories listed in text file @file{file}.
10347 There may be zero, one or more spaces between @option{^-D^/DIRS_FILE=^}
10349 @file{file} must be an existing, readable text file.
10350 Each non empty line in @file{file} must be a directory.
10351 Specifying switch @option{^-D^/DIRS_FILE^} is equivalent to specifying as many
10352 switches @option{^-d^/SOURCE_DIRS^} as there are non empty lines in
10355 @item ^-f^/FOREIGN_PATTERN=^@file{pattern}
10356 @cindex @option{^-f^/FOREIGN_PATTERN^} (@code{gnatname})
10357 Foreign patterns. Using this switch, it is possible to add sources of languages
10358 other than Ada to the list of sources of a project file.
10359 It is only useful if a ^-P^/PROJECT_FILE^ switch is used.
10362 gnatname ^-Pprj -f"*.c"^/PROJECT_FILE=PRJ /FOREIGN_PATTERN=*.C^ "*.ada"
10365 will look for Ada units in all files with the @file{.ada} extension,
10366 and will add to the list of file for project @file{prj.gpr} the C files
10367 with extension ".^c^C^".
10370 @cindex @option{^-h^/HELP^} (@code{gnatname})
10371 Output usage (help) information. The output is written to @file{stdout}.
10373 @item ^-P^/PROJECT_FILE=^@file{proj}
10374 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatname})
10375 Create or update project file @file{proj}. There may be zero, one or more space
10376 between @option{-P} and @file{proj}. @file{proj} may include directory
10377 information. @file{proj} must be writable.
10378 There may be only one switch @option{^-P^/PROJECT_FILE^}.
10379 When a switch @option{^-P^/PROJECT_FILE^} is specified,
10380 no switch @option{^-c^/CONFIG_FILE^} may be specified.
10382 @item ^-v^/VERBOSE^
10383 @cindex @option{^-v^/VERBOSE^} (@code{gnatname})
10384 Verbose mode. Output detailed explanation of behavior to @file{stdout}.
10385 This includes name of the file written, the name of the directories to search
10386 and, for each file in those directories whose name matches at least one of
10387 the Naming Patterns, an indication of whether the file contains a unit,
10388 and if so the name of the unit.
10390 @item ^-v -v^/VERBOSE /VERBOSE^
10391 @cindex @option{^-v -v^/VERBOSE /VERBOSE^} (@code{gnatname})
10392 Very Verbose mode. In addition to the output produced in verbose mode,
10393 for each file in the searched directories whose name matches none of
10394 the Naming Patterns, an indication is given that there is no match.
10396 @item ^-x^/EXCLUDED_PATTERN=^@file{pattern}
10397 @cindex @option{^-x^/EXCLUDED_PATTERN^} (@code{gnatname})
10398 Excluded patterns. Using this switch, it is possible to exclude some files
10399 that would match the name patterns. For example,
10401 gnatname ^-x "*_nt.ada"^/EXCLUDED_PATTERN=*_nt.ada^ "*.ada"
10404 will look for Ada units in all files with the @file{.ada} extension,
10405 except those whose names end with @file{_nt.ada}.
10409 @node Examples of gnatname Usage
10410 @section Examples of @code{gnatname} Usage
10414 $ gnatname /CONFIG_FILE=[HOME.ME]NAMES.ADC /SOURCE_DIRS=SOURCES "[a-z]*.ada*"
10420 $ gnatname -c /home/me/names.adc -d sources "[a-z]*.ada*"
10425 In this example, the directory @file{^/home/me^[HOME.ME]^} must already exist
10426 and be writable. In addition, the directory
10427 @file{^/home/me/sources^[HOME.ME.SOURCES]^} (specified by
10428 @option{^-d sources^/SOURCE_DIRS=SOURCES^}) must exist and be readable.
10431 Note the optional spaces after @option{-c} and @option{-d}.
10436 $ gnatname -P/home/me/proj -x "*_nt_body.ada"
10437 -dsources -dsources/plus -Dcommon_dirs.txt "body_*" "spec_*"
10440 $ gnatname /PROJECT_FILE=[HOME.ME]PROJ
10441 /EXCLUDED_PATTERN=*_nt_body.ada
10442 /SOURCE_DIRS=(SOURCES,[SOURCES.PLUS])
10443 /DIRS_FILE=COMMON_DIRS.TXT "body_*" "spec_*"
10447 Note that several switches @option{^-d^/SOURCE_DIRS^} may be used,
10448 even in conjunction with one or several switches
10449 @option{^-D^/DIRS_FILE^}. Several Naming Patterns and one excluded pattern
10450 are used in this example.
10452 @c *****************************************
10453 @c * G N A T P r o j e c t M a n a g e r *
10454 @c *****************************************
10455 @node GNAT Project Manager
10456 @chapter GNAT Project Manager
10460 * Examples of Project Files::
10461 * Project File Syntax::
10462 * Objects and Sources in Project Files::
10463 * Importing Projects::
10464 * Project Extension::
10465 * Project Hierarchy Extension::
10466 * External References in Project Files::
10467 * Packages in Project Files::
10468 * Variables from Imported Projects::
10470 * Library Projects::
10471 * Stand-alone Library Projects::
10472 * Switches Related to Project Files::
10473 * Tools Supporting Project Files::
10474 * An Extended Example::
10475 * Project File Complete Syntax::
10478 @c ****************
10479 @c * Introduction *
10480 @c ****************
10483 @section Introduction
10486 This chapter describes GNAT's @emph{Project Manager}, a facility that allows
10487 you to manage complex builds involving a number of source files, directories,
10488 and compilation options for different system configurations. In particular,
10489 project files allow you to specify:
10492 The directory or set of directories containing the source files, and/or the
10493 names of the specific source files themselves
10495 The directory in which the compiler's output
10496 (@file{ALI} files, object files, tree files) is to be placed
10498 The directory in which the executable programs is to be placed
10500 ^Switch^Switch^ settings for any of the project-enabled tools
10501 (@command{gnatmake}, compiler, binder, linker, @code{gnatls}, @code{gnatxref},
10502 @code{gnatfind}); you can apply these settings either globally or to individual
10505 The source files containing the main subprogram(s) to be built
10507 The source programming language(s) (currently Ada and/or C)
10509 Source file naming conventions; you can specify these either globally or for
10510 individual compilation units
10517 @node Project Files
10518 @subsection Project Files
10521 Project files are written in a syntax close to that of Ada, using familiar
10522 notions such as packages, context clauses, declarations, default values,
10523 assignments, and inheritance. Finally, project files can be built
10524 hierarchically from other project files, simplifying complex system
10525 integration and project reuse.
10527 A @dfn{project} is a specific set of values for various compilation properties.
10528 The settings for a given project are described by means of
10529 a @dfn{project file}, which is a text file written in an Ada-like syntax.
10530 Property values in project files are either strings or lists of strings.
10531 Properties that are not explicitly set receive default values. A project
10532 file may interrogate the values of @dfn{external variables} (user-defined
10533 command-line switches or environment variables), and it may specify property
10534 settings conditionally, based on the value of such variables.
10536 In simple cases, a project's source files depend only on other source files
10537 in the same project, or on the predefined libraries. (@emph{Dependence} is
10539 the Ada technical sense; as in one Ada unit @code{with}ing another.) However,
10540 the Project Manager also allows more sophisticated arrangements,
10541 where the source files in one project depend on source files in other
10545 One project can @emph{import} other projects containing needed source files.
10547 You can organize GNAT projects in a hierarchy: a @emph{child} project
10548 can extend a @emph{parent} project, inheriting the parent's source files and
10549 optionally overriding any of them with alternative versions
10553 More generally, the Project Manager lets you structure large development
10554 efforts into hierarchical subsystems, where build decisions are delegated
10555 to the subsystem level, and thus different compilation environments
10556 (^switch^switch^ settings) used for different subsystems.
10558 The Project Manager is invoked through the
10559 @option{^-P^/PROJECT_FILE=^@emph{projectfile}}
10560 switch to @command{gnatmake} or to the @command{^gnat^GNAT^} front driver.
10562 There may be zero, one or more spaces between @option{-P} and
10563 @option{@emph{projectfile}}.
10565 If you want to define (on the command line) an external variable that is
10566 queried by the project file, you must use the
10567 @option{^-X^/EXTERNAT_REFERENCE=^@emph{vbl}=@emph{value}} switch.
10568 The Project Manager parses and interprets the project file, and drives the
10569 invoked tool based on the project settings.
10571 The Project Manager supports a wide range of development strategies,
10572 for systems of all sizes. Here are some typical practices that are
10576 Using a common set of source files, but generating object files in different
10577 directories via different ^switch^switch^ settings
10579 Using a mostly-shared set of source files, but with different versions of
10584 The destination of an executable can be controlled inside a project file
10585 using the @option{^-o^-o^}
10587 In the absence of such a ^switch^switch^ either inside
10588 the project file or on the command line, any executable files generated by
10589 @command{gnatmake} are placed in the directory @code{Exec_Dir} specified
10590 in the project file. If no @code{Exec_Dir} is specified, they will be placed
10591 in the object directory of the project.
10593 You can use project files to achieve some of the effects of a source
10594 versioning system (for example, defining separate projects for
10595 the different sets of sources that comprise different releases) but the
10596 Project Manager is independent of any source configuration management tools
10597 that might be used by the developers.
10599 The next section introduces the main features of GNAT's project facility
10600 through a sequence of examples; subsequent sections will present the syntax
10601 and semantics in more detail. A more formal description of the project
10602 facility appears in the GNAT Reference Manual.
10604 @c *****************************
10605 @c * Examples of Project Files *
10606 @c *****************************
10608 @node Examples of Project Files
10609 @section Examples of Project Files
10611 This section illustrates some of the typical uses of project files and
10612 explains their basic structure and behavior.
10615 * Common Sources with Different ^Switches^Switches^ and Directories::
10616 * Using External Variables::
10617 * Importing Other Projects::
10618 * Extending a Project::
10621 @node Common Sources with Different ^Switches^Switches^ and Directories
10622 @subsection Common Sources with Different ^Switches^Switches^ and Directories
10626 * Specifying the Object Directory::
10627 * Specifying the Exec Directory::
10628 * Project File Packages::
10629 * Specifying ^Switch^Switch^ Settings::
10630 * Main Subprograms::
10631 * Executable File Names::
10632 * Source File Naming Conventions::
10633 * Source Language(s)::
10637 Suppose that the Ada source files @file{pack.ads}, @file{pack.adb}, and
10638 @file{proc.adb} are in the @file{/common} directory. The file
10639 @file{proc.adb} contains an Ada main subprogram @code{Proc} that @code{with}s
10640 package @code{Pack}. We want to compile these source files under two sets
10641 of ^switches^switches^:
10644 When debugging, we want to pass the @option{-g} switch to @command{gnatmake},
10645 and the @option{^-gnata^-gnata^},
10646 @option{^-gnato^-gnato^},
10647 and @option{^-gnatE^-gnatE^} switches to the
10648 compiler; the compiler's output is to appear in @file{/common/debug}
10650 When preparing a release version, we want to pass the @option{^-O2^O2^} switch
10651 to the compiler; the compiler's output is to appear in @file{/common/release}
10655 The GNAT project files shown below, respectively @file{debug.gpr} and
10656 @file{release.gpr} in the @file{/common} directory, achieve these effects.
10669 ^/common/debug^[COMMON.DEBUG]^
10674 ^/common/release^[COMMON.RELEASE]^
10679 Here are the corresponding project files:
10681 @smallexample @c projectfile
10684 for Object_Dir use "debug";
10685 for Main use ("proc");
10688 for ^Default_Switches^Default_Switches^ ("Ada")
10690 for Executable ("proc.adb") use "proc1";
10695 package Compiler is
10696 for ^Default_Switches^Default_Switches^ ("Ada")
10697 use ("-fstack-check",
10700 "^-gnatE^-gnatE^");
10706 @smallexample @c projectfile
10709 for Object_Dir use "release";
10710 for Exec_Dir use ".";
10711 for Main use ("proc");
10713 package Compiler is
10714 for ^Default_Switches^Default_Switches^ ("Ada")
10722 The name of the project defined by @file{debug.gpr} is @code{"Debug"} (case
10723 insensitive), and analogously the project defined by @file{release.gpr} is
10724 @code{"Release"}. For consistency the file should have the same name as the
10725 project, and the project file's extension should be @code{"gpr"}. These
10726 conventions are not required, but a warning is issued if they are not followed.
10728 If the current directory is @file{^/temp^[TEMP]^}, then the command
10730 gnatmake ^-P/common/debug.gpr^/PROJECT_FILE=[COMMON]DEBUG^
10734 generates object and ALI files in @file{^/common/debug^[COMMON.DEBUG]^},
10735 as well as the @code{^proc1^PROC1.EXE^} executable,
10736 using the ^switch^switch^ settings defined in the project file.
10738 Likewise, the command
10740 gnatmake ^-P/common/release.gpr^/PROJECT_FILE=[COMMON]RELEASE^
10744 generates object and ALI files in @file{^/common/release^[COMMON.RELEASE]^},
10745 and the @code{^proc^PROC.EXE^}
10746 executable in @file{^/common^[COMMON]^},
10747 using the ^switch^switch^ settings from the project file.
10750 @unnumberedsubsubsec Source Files
10753 If a project file does not explicitly specify a set of source directories or
10754 a set of source files, then by default the project's source files are the
10755 Ada source files in the project file directory. Thus @file{pack.ads},
10756 @file{pack.adb}, and @file{proc.adb} are the source files for both projects.
10758 @node Specifying the Object Directory
10759 @unnumberedsubsubsec Specifying the Object Directory
10762 Several project properties are modeled by Ada-style @emph{attributes};
10763 a property is defined by supplying the equivalent of an Ada attribute
10764 definition clause in the project file.
10765 A project's object directory is another such a property; the corresponding
10766 attribute is @code{Object_Dir}, and its value is also a string expression,
10767 specified either as absolute or relative. In the later case,
10768 it is relative to the project file directory. Thus the compiler's
10769 output is directed to @file{^/common/debug^[COMMON.DEBUG]^}
10770 (for the @code{Debug} project)
10771 and to @file{^/common/release^[COMMON.RELEASE]^}
10772 (for the @code{Release} project).
10773 If @code{Object_Dir} is not specified, then the default is the project file
10776 @node Specifying the Exec Directory
10777 @unnumberedsubsubsec Specifying the Exec Directory
10780 A project's exec directory is another property; the corresponding
10781 attribute is @code{Exec_Dir}, and its value is also a string expression,
10782 either specified as relative or absolute. If @code{Exec_Dir} is not specified,
10783 then the default is the object directory (which may also be the project file
10784 directory if attribute @code{Object_Dir} is not specified). Thus the executable
10785 is placed in @file{^/common/debug^[COMMON.DEBUG]^}
10786 for the @code{Debug} project (attribute @code{Exec_Dir} not specified)
10787 and in @file{^/common^[COMMON]^} for the @code{Release} project.
10789 @node Project File Packages
10790 @unnumberedsubsubsec Project File Packages
10793 A GNAT tool that is integrated with the Project Manager is modeled by a
10794 corresponding package in the project file. In the example above,
10795 The @code{Debug} project defines the packages @code{Builder}
10796 (for @command{gnatmake}) and @code{Compiler};
10797 the @code{Release} project defines only the @code{Compiler} package.
10799 The Ada-like package syntax is not to be taken literally. Although packages in
10800 project files bear a surface resemblance to packages in Ada source code, the
10801 notation is simply a way to convey a grouping of properties for a named
10802 entity. Indeed, the package names permitted in project files are restricted
10803 to a predefined set, corresponding to the project-aware tools, and the contents
10804 of packages are limited to a small set of constructs.
10805 The packages in the example above contain attribute definitions.
10807 @node Specifying ^Switch^Switch^ Settings
10808 @unnumberedsubsubsec Specifying ^Switch^Switch^ Settings
10811 ^Switch^Switch^ settings for a project-aware tool can be specified through
10812 attributes in the package that corresponds to the tool.
10813 The example above illustrates one of the relevant attributes,
10814 @code{^Default_Switches^Default_Switches^}, which is defined in packages
10815 in both project files.
10816 Unlike simple attributes like @code{Source_Dirs},
10817 @code{^Default_Switches^Default_Switches^} is
10818 known as an @emph{associative array}. When you define this attribute, you must
10819 supply an ``index'' (a literal string), and the effect of the attribute
10820 definition is to set the value of the array at the specified index.
10821 For the @code{^Default_Switches^Default_Switches^} attribute,
10822 the index is a programming language (in our case, Ada),
10823 and the value specified (after @code{use}) must be a list
10824 of string expressions.
10826 The attributes permitted in project files are restricted to a predefined set.
10827 Some may appear at project level, others in packages.
10828 For any attribute that is an associative array, the index must always be a
10829 literal string, but the restrictions on this string (e.g., a file name or a
10830 language name) depend on the individual attribute.
10831 Also depending on the attribute, its specified value will need to be either a
10832 string or a string list.
10834 In the @code{Debug} project, we set the switches for two tools,
10835 @command{gnatmake} and the compiler, and thus we include the two corresponding
10836 packages; each package defines the @code{^Default_Switches^Default_Switches^}
10837 attribute with index @code{"Ada"}.
10838 Note that the package corresponding to
10839 @command{gnatmake} is named @code{Builder}. The @code{Release} project is
10840 similar, but only includes the @code{Compiler} package.
10842 In project @code{Debug} above, the ^switches^switches^ starting with
10843 @option{-gnat} that are specified in package @code{Compiler}
10844 could have been placed in package @code{Builder}, since @command{gnatmake}
10845 transmits all such ^switches^switches^ to the compiler.
10847 @node Main Subprograms
10848 @unnumberedsubsubsec Main Subprograms
10851 One of the specifiable properties of a project is a list of files that contain
10852 main subprograms. This property is captured in the @code{Main} attribute,
10853 whose value is a list of strings. If a project defines the @code{Main}
10854 attribute, it is not necessary to identify the main subprogram(s) when
10855 invoking @command{gnatmake} (@pxref{gnatmake and Project Files}).
10857 @node Executable File Names
10858 @unnumberedsubsubsec Executable File Names
10861 By default, the executable file name corresponding to a main source is
10862 deduced from the main source file name. Through the attributes
10863 @code{Executable} and @code{Executable_Suffix} of package @code{Builder},
10864 it is possible to change this default.
10865 In project @code{Debug} above, the executable file name
10866 for main source @file{^proc.adb^PROC.ADB^} is
10867 @file{^proc1^PROC1.EXE^}.
10868 Attribute @code{Executable_Suffix}, when specified, may change the suffix
10869 of the the executable files, when no attribute @code{Executable} applies:
10870 its value replace the platform-specific executable suffix.
10871 Attributes @code{Executable} and @code{Executable_Suffix} are the only ways to
10872 specify a non default executable file name when several mains are built at once
10873 in a single @command{gnatmake} command.
10875 @node Source File Naming Conventions
10876 @unnumberedsubsubsec Source File Naming Conventions
10879 Since the project files above do not specify any source file naming
10880 conventions, the GNAT defaults are used. The mechanism for defining source
10881 file naming conventions -- a package named @code{Naming} --
10882 is described below (@pxref{Naming Schemes}).
10884 @node Source Language(s)
10885 @unnumberedsubsubsec Source Language(s)
10888 Since the project files do not specify a @code{Languages} attribute, by
10889 default the GNAT tools assume that the language of the project file is Ada.
10890 More generally, a project can comprise source files
10891 in Ada, C, and/or other languages.
10893 @node Using External Variables
10894 @subsection Using External Variables
10897 Instead of supplying different project files for debug and release, we can
10898 define a single project file that queries an external variable (set either
10899 on the command line or via an ^environment variable^logical name^) in order to
10900 conditionally define the appropriate settings. Again, assume that the
10901 source files @file{pack.ads}, @file{pack.adb}, and @file{proc.adb} are
10902 located in directory @file{^/common^[COMMON]^}. The following project file,
10903 @file{build.gpr}, queries the external variable named @code{STYLE} and
10904 defines an object directory and ^switch^switch^ settings based on whether
10905 the value is @code{"deb"} (debug) or @code{"rel"} (release), and where
10906 the default is @code{"deb"}.
10908 @smallexample @c projectfile
10911 for Main use ("proc");
10913 type Style_Type is ("deb", "rel");
10914 Style : Style_Type := external ("STYLE", "deb");
10918 for Object_Dir use "debug";
10921 for Object_Dir use "release";
10922 for Exec_Dir use ".";
10931 for ^Default_Switches^Default_Switches^ ("Ada")
10933 for Executable ("proc") use "proc1";
10942 package Compiler is
10946 for ^Default_Switches^Default_Switches^ ("Ada")
10947 use ("^-gnata^-gnata^",
10949 "^-gnatE^-gnatE^");
10952 for ^Default_Switches^Default_Switches^ ("Ada")
10963 @code{Style_Type} is an example of a @emph{string type}, which is the project
10964 file analog of an Ada enumeration type but whose components are string literals
10965 rather than identifiers. @code{Style} is declared as a variable of this type.
10967 The form @code{external("STYLE", "deb")} is known as an
10968 @emph{external reference}; its first argument is the name of an
10969 @emph{external variable}, and the second argument is a default value to be
10970 used if the external variable doesn't exist. You can define an external
10971 variable on the command line via the @option{^-X^/EXTERNAL_REFERENCE^} switch,
10972 or you can use ^an environment variable^a logical name^
10973 as an external variable.
10975 Each @code{case} construct is expanded by the Project Manager based on the
10976 value of @code{Style}. Thus the command
10979 gnatmake -P/common/build.gpr -XSTYLE=deb
10985 gnatmake /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=deb
10990 is equivalent to the @command{gnatmake} invocation using the project file
10991 @file{debug.gpr} in the earlier example. So is the command
10993 gnatmake ^-P/common/build.gpr^/PROJECT_FILE=[COMMON]BUILD.GPR^
10997 since @code{"deb"} is the default for @code{STYLE}.
11003 gnatmake -P/common/build.gpr -XSTYLE=rel
11009 GNAT MAKE /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=rel
11014 is equivalent to the @command{gnatmake} invocation using the project file
11015 @file{release.gpr} in the earlier example.
11017 @node Importing Other Projects
11018 @subsection Importing Other Projects
11019 @cindex @code{ADA_PROJECT_PATH}
11022 A compilation unit in a source file in one project may depend on compilation
11023 units in source files in other projects. To compile this unit under
11024 control of a project file, the
11025 dependent project must @emph{import} the projects containing the needed source
11027 This effect is obtained using syntax similar to an Ada @code{with} clause,
11028 but where @code{with}ed entities are strings that denote project files.
11030 As an example, suppose that the two projects @code{GUI_Proj} and
11031 @code{Comm_Proj} are defined in the project files @file{gui_proj.gpr} and
11032 @file{comm_proj.gpr} in directories @file{^/gui^[GUI]^}
11033 and @file{^/comm^[COMM]^}, respectively.
11034 Suppose that the source files for @code{GUI_Proj} are
11035 @file{gui.ads} and @file{gui.adb}, and that the source files for
11036 @code{Comm_Proj} are @file{comm.ads} and @file{comm.adb}, where each set of
11037 files is located in its respective project file directory. Schematically:
11056 We want to develop an application in directory @file{^/app^[APP]^} that
11057 @code{with} the packages @code{GUI} and @code{Comm}, using the properties of
11058 the corresponding project files (e.g. the ^switch^switch^ settings
11059 and object directory).
11060 Skeletal code for a main procedure might be something like the following:
11062 @smallexample @c ada
11065 procedure App_Main is
11074 Here is a project file, @file{app_proj.gpr}, that achieves the desired
11077 @smallexample @c projectfile
11079 with "/gui/gui_proj", "/comm/comm_proj";
11080 project App_Proj is
11081 for Main use ("app_main");
11087 Building an executable is achieved through the command:
11089 gnatmake ^-P/app/app_proj^/PROJECT_FILE=[APP]APP_PROJ^
11092 which will generate the @code{^app_main^APP_MAIN.EXE^} executable
11093 in the directory where @file{app_proj.gpr} resides.
11095 If an imported project file uses the standard extension (@code{^gpr^GPR^}) then
11096 (as illustrated above) the @code{with} clause can omit the extension.
11098 Our example specified an absolute path for each imported project file.
11099 Alternatively, the directory name of an imported object can be omitted
11103 The imported project file is in the same directory as the importing project
11106 You have defined ^an environment variable^a logical name^
11107 that includes the directory containing
11108 the needed project file. The syntax of @code{ADA_PROJECT_PATH} is the same as
11109 the syntax of @code{ADA_INCLUDE_PATH} and @code{ADA_OBJECTS_PATH}: a list of
11110 directory names separated by colons (semicolons on Windows).
11114 Thus, if we define @code{ADA_PROJECT_PATH} to include @file{^/gui^[GUI]^} and
11115 @file{^/comm^[COMM]^}, then our project file @file{app_proj.gpr} can be written
11118 @smallexample @c projectfile
11120 with "gui_proj", "comm_proj";
11121 project App_Proj is
11122 for Main use ("app_main");
11128 Importing other projects can create ambiguities.
11129 For example, the same unit might be present in different imported projects, or
11130 it might be present in both the importing project and in an imported project.
11131 Both of these conditions are errors. Note that in the current version of
11132 the Project Manager, it is illegal to have an ambiguous unit even if the
11133 unit is never referenced by the importing project. This restriction may be
11134 relaxed in a future release.
11136 @node Extending a Project
11137 @subsection Extending a Project
11140 In large software systems it is common to have multiple
11141 implementations of a common interface; in Ada terms, multiple versions of a
11142 package body for the same specification. For example, one implementation
11143 might be safe for use in tasking programs, while another might only be used
11144 in sequential applications. This can be modeled in GNAT using the concept
11145 of @emph{project extension}. If one project (the ``child'') @emph{extends}
11146 another project (the ``parent'') then by default all source files of the
11147 parent project are inherited by the child, but the child project can
11148 override any of the parent's source files with new versions, and can also
11149 add new files. This facility is the project analog of a type extension in
11150 Object-Oriented Programming. Project hierarchies are permitted (a child
11151 project may be the parent of yet another project), and a project that
11152 inherits one project can also import other projects.
11154 As an example, suppose that directory @file{^/seq^[SEQ]^} contains the project
11155 file @file{seq_proj.gpr} as well as the source files @file{pack.ads},
11156 @file{pack.adb}, and @file{proc.adb}:
11169 Note that the project file can simply be empty (that is, no attribute or
11170 package is defined):
11172 @smallexample @c projectfile
11174 project Seq_Proj is
11180 implying that its source files are all the Ada source files in the project
11183 Suppose we want to supply an alternate version of @file{pack.adb}, in
11184 directory @file{^/tasking^[TASKING]^}, but use the existing versions of
11185 @file{pack.ads} and @file{proc.adb}. We can define a project
11186 @code{Tasking_Proj} that inherits @code{Seq_Proj}:
11190 ^/tasking^[TASKING]^
11196 project Tasking_Proj extends "/seq/seq_proj" is
11202 The version of @file{pack.adb} used in a build depends on which project file
11205 Note that we could have obtained the desired behavior using project import
11206 rather than project inheritance; a @code{base} project would contain the
11207 sources for @file{pack.ads} and @file{proc.adb}, a sequential project would
11208 import @code{base} and add @file{pack.adb}, and likewise a tasking project
11209 would import @code{base} and add a different version of @file{pack.adb}. The
11210 choice depends on whether other sources in the original project need to be
11211 overridden. If they do, then project extension is necessary, otherwise,
11212 importing is sufficient.
11215 In a project file that extends another project file, it is possible to
11216 indicate that an inherited source is not part of the sources of the extending
11217 project. This is necessary sometimes when a package spec has been overloaded
11218 and no longer requires a body: in this case, it is necessary to indicate that
11219 the inherited body is not part of the sources of the project, otherwise there
11220 will be a compilation error when compiling the spec.
11222 For that purpose, the attribute @code{Locally_Removed_Files} is used.
11223 Its value is a string list: a list of file names.
11225 @smallexample @c @projectfile
11226 project B extends "a" is
11227 for Source_Files use ("pkg.ads");
11228 -- New spec of Pkg does not need a completion
11229 for Locally_Removed_Files use ("pkg.adb");
11233 Attribute @code{Locally_Removed_Files} may also be used to check if a source
11234 is still needed: if it is possible to build using @command{gnatmake} when such
11235 a source is put in attribute @code{Locally_Removed_Files} of a project P, then
11236 it is possible to remove the source completely from a system that includes
11239 @c ***********************
11240 @c * Project File Syntax *
11241 @c ***********************
11243 @node Project File Syntax
11244 @section Project File Syntax
11253 * Associative Array Attributes::
11254 * case Constructions::
11258 This section describes the structure of project files.
11260 A project may be an @emph{independent project}, entirely defined by a single
11261 project file. Any Ada source file in an independent project depends only
11262 on the predefined library and other Ada source files in the same project.
11265 A project may also @dfn{depend on} other projects, in either or both of
11266 the following ways:
11268 @item It may import any number of projects
11269 @item It may extend at most one other project
11273 The dependence relation is a directed acyclic graph (the subgraph reflecting
11274 the ``extends'' relation is a tree).
11276 A project's @dfn{immediate sources} are the source files directly defined by
11277 that project, either implicitly by residing in the project file's directory,
11278 or explicitly through any of the source-related attributes described below.
11279 More generally, a project @var{proj}'s @dfn{sources} are the immediate sources
11280 of @var{proj} together with the immediate sources (unless overridden) of any
11281 project on which @var{proj} depends (either directly or indirectly).
11284 @subsection Basic Syntax
11287 As seen in the earlier examples, project files have an Ada-like syntax.
11288 The minimal project file is:
11289 @smallexample @c projectfile
11298 The identifier @code{Empty} is the name of the project.
11299 This project name must be present after the reserved
11300 word @code{end} at the end of the project file, followed by a semi-colon.
11302 Any name in a project file, such as the project name or a variable name,
11303 has the same syntax as an Ada identifier.
11305 The reserved words of project files are the Ada reserved words plus
11306 @code{extends}, @code{external}, and @code{project}. Note that the only Ada
11307 reserved words currently used in project file syntax are:
11335 Comments in project files have the same syntax as in Ada, two consecutives
11336 hyphens through the end of the line.
11339 @subsection Packages
11342 A project file may contain @emph{packages}. The name of a package must be one
11343 of the identifiers from the following list. A package
11344 with a given name may only appear once in a project file. Package names are
11345 case insensitive. The following package names are legal:
11361 @code{Cross_Reference}
11365 @code{Pretty_Printer}
11375 @code{Language_Processing}
11379 In its simplest form, a package may be empty:
11381 @smallexample @c projectfile
11391 A package may contain @emph{attribute declarations},
11392 @emph{variable declarations} and @emph{case constructions}, as will be
11395 When there is ambiguity between a project name and a package name,
11396 the name always designates the project. To avoid possible confusion, it is
11397 always a good idea to avoid naming a project with one of the
11398 names allowed for packages or any name that starts with @code{gnat}.
11401 @subsection Expressions
11404 An @emph{expression} is either a @emph{string expression} or a
11405 @emph{string list expression}.
11407 A @emph{string expression} is either a @emph{simple string expression} or a
11408 @emph{compound string expression}.
11410 A @emph{simple string expression} is one of the following:
11412 @item A literal string; e.g.@code{"comm/my_proj.gpr"}
11413 @item A string-valued variable reference (@pxref{Variables})
11414 @item A string-valued attribute reference (@pxref{Attributes})
11415 @item An external reference (@pxref{External References in Project Files})
11419 A @emph{compound string expression} is a concatenation of string expressions,
11420 using the operator @code{"&"}
11422 Path & "/" & File_Name & ".ads"
11426 A @emph{string list expression} is either a
11427 @emph{simple string list expression} or a
11428 @emph{compound string list expression}.
11430 A @emph{simple string list expression} is one of the following:
11432 @item A parenthesized list of zero or more string expressions,
11433 separated by commas
11435 File_Names := (File_Name, "gnat.adc", File_Name & ".orig");
11438 @item A string list-valued variable reference
11439 @item A string list-valued attribute reference
11443 A @emph{compound string list expression} is the concatenation (using
11444 @code{"&"}) of a simple string list expression and an expression. Note that
11445 each term in a compound string list expression, except the first, may be
11446 either a string expression or a string list expression.
11448 @smallexample @c projectfile
11450 File_Name_List := () & File_Name; -- One string in this list
11451 Extended_File_Name_List := File_Name_List & (File_Name & ".orig");
11453 Big_List := File_Name_List & Extended_File_Name_List;
11454 -- Concatenation of two string lists: three strings
11455 Illegal_List := "gnat.adc" & Extended_File_Name_List;
11456 -- Illegal: must start with a string list
11461 @subsection String Types
11464 A @emph{string type declaration} introduces a discrete set of string literals.
11465 If a string variable is declared to have this type, its value
11466 is restricted to the given set of literals.
11468 Here is an example of a string type declaration:
11470 @smallexample @c projectfile
11471 type OS is ("NT", "nt", "Unix", "GNU/Linux", "other OS");
11475 Variables of a string type are called @emph{typed variables}; all other
11476 variables are called @emph{untyped variables}. Typed variables are
11477 particularly useful in @code{case} constructions, to support conditional
11478 attribute declarations.
11479 (@pxref{case Constructions}).
11481 The string literals in the list are case sensitive and must all be different.
11482 They may include any graphic characters allowed in Ada, including spaces.
11484 A string type may only be declared at the project level, not inside a package.
11486 A string type may be referenced by its name if it has been declared in the same
11487 project file, or by an expanded name whose prefix is the name of the project
11488 in which it is declared.
11491 @subsection Variables
11494 A variable may be declared at the project file level, or within a package.
11495 Here are some examples of variable declarations:
11497 @smallexample @c projectfile
11499 This_OS : OS := external ("OS"); -- a typed variable declaration
11500 That_OS := "GNU/Linux"; -- an untyped variable declaration
11505 The syntax of a @emph{typed variable declaration} is identical to the Ada
11506 syntax for an object declaration. By contrast, the syntax of an untyped
11507 variable declaration is identical to an Ada assignment statement. In fact,
11508 variable declarations in project files have some of the characteristics of
11509 an assignment, in that successive declarations for the same variable are
11510 allowed. Untyped variable declarations do establish the expected kind of the
11511 variable (string or string list), and successive declarations for it must
11512 respect the initial kind.
11515 A string variable declaration (typed or untyped) declares a variable
11516 whose value is a string. This variable may be used as a string expression.
11517 @smallexample @c projectfile
11518 File_Name := "readme.txt";
11519 Saved_File_Name := File_Name & ".saved";
11523 A string list variable declaration declares a variable whose value is a list
11524 of strings. The list may contain any number (zero or more) of strings.
11526 @smallexample @c projectfile
11528 List_With_One_Element := ("^-gnaty^-gnaty^");
11529 List_With_Two_Elements := List_With_One_Element & "^-gnatg^-gnatg^";
11530 Long_List := ("main.ada", "pack1_.ada", "pack1.ada", "pack2_.ada"
11531 "pack2.ada", "util_.ada", "util.ada");
11535 The same typed variable may not be declared more than once at project level,
11536 and it may not be declared more than once in any package; it is in effect
11539 The same untyped variable may be declared several times. Declarations are
11540 elaborated in the order in which they appear, so the new value replaces
11541 the old one, and any subsequent reference to the variable uses the new value.
11542 However, as noted above, if a variable has been declared as a string, all
11544 declarations must give it a string value. Similarly, if a variable has
11545 been declared as a string list, all subsequent declarations
11546 must give it a string list value.
11548 A @emph{variable reference} may take several forms:
11551 @item The simple variable name, for a variable in the current package (if any)
11552 or in the current project
11553 @item An expanded name, whose prefix is a context name.
11557 A @emph{context} may be one of the following:
11560 @item The name of an existing package in the current project
11561 @item The name of an imported project of the current project
11562 @item The name of an ancestor project (i.e., a project extended by the current
11563 project, either directly or indirectly)
11564 @item An expanded name whose prefix is an imported/parent project name, and
11565 whose selector is a package name in that project.
11569 A variable reference may be used in an expression.
11572 @subsection Attributes
11575 A project (and its packages) may have @emph{attributes} that define
11576 the project's properties. Some attributes have values that are strings;
11577 others have values that are string lists.
11579 There are two categories of attributes: @emph{simple attributes}
11580 and @emph{associative arrays} (@pxref{Associative Array Attributes}).
11582 Legal project attribute names, and attribute names for each legal package are
11583 listed below. Attributes names are case-insensitive.
11585 The following attributes are defined on projects (all are simple attributes):
11587 @multitable @columnfractions .4 .3
11588 @item @emph{Attribute Name}
11590 @item @code{Source_Files}
11592 @item @code{Source_Dirs}
11594 @item @code{Source_List_File}
11596 @item @code{Object_Dir}
11598 @item @code{Exec_Dir}
11600 @item @code{Locally_Removed_Files}
11604 @item @code{Languages}
11606 @item @code{Main_Language}
11608 @item @code{Library_Dir}
11610 @item @code{Library_Name}
11612 @item @code{Library_Kind}
11614 @item @code{Library_Version}
11616 @item @code{Library_Interface}
11618 @item @code{Library_Auto_Init}
11620 @item @code{Library_Options}
11622 @item @code{Library_GCC}
11627 The following attributes are defined for package @code{Naming}
11628 (@pxref{Naming Schemes}):
11630 @multitable @columnfractions .4 .2 .2 .2
11631 @item Attribute Name @tab Category @tab Index @tab Value
11632 @item @code{Spec_Suffix}
11633 @tab associative array
11636 @item @code{Body_Suffix}
11637 @tab associative array
11640 @item @code{Separate_Suffix}
11641 @tab simple attribute
11644 @item @code{Casing}
11645 @tab simple attribute
11648 @item @code{Dot_Replacement}
11649 @tab simple attribute
11653 @tab associative array
11657 @tab associative array
11660 @item @code{Specification_Exceptions}
11661 @tab associative array
11664 @item @code{Implementation_Exceptions}
11665 @tab associative array
11671 The following attributes are defined for packages @code{Builder},
11672 @code{Compiler}, @code{Binder},
11673 @code{Linker}, @code{Cross_Reference}, and @code{Finder}
11674 (@pxref{^Switches^Switches^ and Project Files}).
11676 @multitable @columnfractions .4 .2 .2 .2
11677 @item Attribute Name @tab Category @tab Index @tab Value
11678 @item @code{^Default_Switches^Default_Switches^}
11679 @tab associative array
11682 @item @code{^Switches^Switches^}
11683 @tab associative array
11689 In addition, package @code{Compiler} has a single string attribute
11690 @code{Local_Configuration_Pragmas} and package @code{Builder} has a single
11691 string attribute @code{Global_Configuration_Pragmas}.
11694 Each simple attribute has a default value: the empty string (for string-valued
11695 attributes) and the empty list (for string list-valued attributes).
11697 An attribute declaration defines a new value for an attribute.
11699 Examples of simple attribute declarations:
11701 @smallexample @c projectfile
11702 for Object_Dir use "objects";
11703 for Source_Dirs use ("units", "test/drivers");
11707 The syntax of a @dfn{simple attribute declaration} is similar to that of an
11708 attribute definition clause in Ada.
11710 Attributes references may be appear in expressions.
11711 The general form for such a reference is @code{<entity>'<attribute>}:
11712 Associative array attributes are functions. Associative
11713 array attribute references must have an argument that is a string literal.
11717 @smallexample @c projectfile
11719 Naming'Dot_Replacement
11720 Imported_Project'Source_Dirs
11721 Imported_Project.Naming'Casing
11722 Builder'^Default_Switches^Default_Switches^("Ada")
11726 The prefix of an attribute may be:
11728 @item @code{project} for an attribute of the current project
11729 @item The name of an existing package of the current project
11730 @item The name of an imported project
11731 @item The name of a parent project that is extended by the current project
11732 @item An expanded name whose prefix is imported/parent project name,
11733 and whose selector is a package name
11738 @smallexample @c projectfile
11741 for Source_Dirs use project'Source_Dirs & "units";
11742 for Source_Dirs use project'Source_Dirs & "test/drivers"
11748 In the first attribute declaration, initially the attribute @code{Source_Dirs}
11749 has the default value: an empty string list. After this declaration,
11750 @code{Source_Dirs} is a string list of one element: @code{"units"}.
11751 After the second attribute declaration @code{Source_Dirs} is a string list of
11752 two elements: @code{"units"} and @code{"test/drivers"}.
11754 Note: this example is for illustration only. In practice,
11755 the project file would contain only one attribute declaration:
11757 @smallexample @c projectfile
11758 for Source_Dirs use ("units", "test/drivers");
11761 @node Associative Array Attributes
11762 @subsection Associative Array Attributes
11765 Some attributes are defined as @emph{associative arrays}. An associative
11766 array may be regarded as a function that takes a string as a parameter
11767 and delivers a string or string list value as its result.
11769 Here are some examples of single associative array attribute associations:
11771 @smallexample @c projectfile
11772 for Body ("main") use "Main.ada";
11773 for ^Switches^Switches^ ("main.ada")
11775 "^-gnatv^-gnatv^");
11776 for ^Switches^Switches^ ("main.ada")
11777 use Builder'^Switches^Switches^ ("main.ada")
11782 Like untyped variables and simple attributes, associative array attributes
11783 may be declared several times. Each declaration supplies a new value for the
11784 attribute, and replaces the previous setting.
11787 An associative array attribute may be declared as a full associative array
11788 declaration, with the value of the same attribute in an imported or extended
11791 @smallexample @c projectfile
11793 for Default_Switches use Default.Builder'Default_Switches;
11798 In this example, @code{Default} must be either an project imported by the
11799 current project, or the project that the current project extends. If the
11800 attribute is in a package (in this case, in package @code{Builder}), the same
11801 package needs to be specified.
11804 A full associative array declaration replaces any other declaration for the
11805 attribute, including other full associative array declaration. Single
11806 associative array associations may be declare after a full associative
11807 declaration, modifying the value for a single association of the attribute.
11809 @node case Constructions
11810 @subsection @code{case} Constructions
11813 A @code{case} construction is used in a project file to effect conditional
11815 Here is a typical example:
11817 @smallexample @c projectfile
11820 type OS_Type is ("GNU/Linux", "Unix", "NT", "VMS");
11822 OS : OS_Type := external ("OS", "GNU/Linux");
11826 package Compiler is
11828 when "GNU/Linux" | "Unix" =>
11829 for ^Default_Switches^Default_Switches^ ("Ada")
11830 use ("^-gnath^-gnath^");
11832 for ^Default_Switches^Default_Switches^ ("Ada")
11833 use ("^-gnatP^-gnatP^");
11842 The syntax of a @code{case} construction is based on the Ada case statement
11843 (although there is no @code{null} construction for empty alternatives).
11845 The case expression must a typed string variable.
11846 Each alternative comprises the reserved word @code{when}, either a list of
11847 literal strings separated by the @code{"|"} character or the reserved word
11848 @code{others}, and the @code{"=>"} token.
11849 Each literal string must belong to the string type that is the type of the
11851 An @code{others} alternative, if present, must occur last.
11853 After each @code{=>}, there are zero or more constructions. The only
11854 constructions allowed in a case construction are other case constructions and
11855 attribute declarations. String type declarations, variable declarations and
11856 package declarations are not allowed.
11858 The value of the case variable is often given by an external reference
11859 (@pxref{External References in Project Files}).
11861 @c ****************************************
11862 @c * Objects and Sources in Project Files *
11863 @c ****************************************
11865 @node Objects and Sources in Project Files
11866 @section Objects and Sources in Project Files
11869 * Object Directory::
11871 * Source Directories::
11872 * Source File Names::
11876 Each project has exactly one object directory and one or more source
11877 directories. The source directories must contain at least one source file,
11878 unless the project file explicitly specifies that no source files are present
11879 (@pxref{Source File Names}).
11881 @node Object Directory
11882 @subsection Object Directory
11885 The object directory for a project is the directory containing the compiler's
11886 output (such as @file{ALI} files and object files) for the project's immediate
11889 The object directory is given by the value of the attribute @code{Object_Dir}
11890 in the project file.
11892 @smallexample @c projectfile
11893 for Object_Dir use "objects";
11897 The attribute @var{Object_Dir} has a string value, the path name of the object
11898 directory. The path name may be absolute or relative to the directory of the
11899 project file. This directory must already exist, and be readable and writable.
11901 By default, when the attribute @code{Object_Dir} is not given an explicit value
11902 or when its value is the empty string, the object directory is the same as the
11903 directory containing the project file.
11905 @node Exec Directory
11906 @subsection Exec Directory
11909 The exec directory for a project is the directory containing the executables
11910 for the project's main subprograms.
11912 The exec directory is given by the value of the attribute @code{Exec_Dir}
11913 in the project file.
11915 @smallexample @c projectfile
11916 for Exec_Dir use "executables";
11920 The attribute @var{Exec_Dir} has a string value, the path name of the exec
11921 directory. The path name may be absolute or relative to the directory of the
11922 project file. This directory must already exist, and be writable.
11924 By default, when the attribute @code{Exec_Dir} is not given an explicit value
11925 or when its value is the empty string, the exec directory is the same as the
11926 object directory of the project file.
11928 @node Source Directories
11929 @subsection Source Directories
11932 The source directories of a project are specified by the project file
11933 attribute @code{Source_Dirs}.
11935 This attribute's value is a string list. If the attribute is not given an
11936 explicit value, then there is only one source directory, the one where the
11937 project file resides.
11939 A @code{Source_Dirs} attribute that is explicitly defined to be the empty list,
11942 @smallexample @c projectfile
11943 for Source_Dirs use ();
11947 indicates that the project contains no source files.
11949 Otherwise, each string in the string list designates one or more
11950 source directories.
11952 @smallexample @c projectfile
11953 for Source_Dirs use ("sources", "test/drivers");
11957 If a string in the list ends with @code{"/**"}, then the directory whose path
11958 name precedes the two asterisks, as well as all its subdirectories
11959 (recursively), are source directories.
11961 @smallexample @c projectfile
11962 for Source_Dirs use ("/system/sources/**");
11966 Here the directory @code{/system/sources} and all of its subdirectories
11967 (recursively) are source directories.
11969 To specify that the source directories are the directory of the project file
11970 and all of its subdirectories, you can declare @code{Source_Dirs} as follows:
11971 @smallexample @c projectfile
11972 for Source_Dirs use ("./**");
11976 Each of the source directories must exist and be readable.
11978 @node Source File Names
11979 @subsection Source File Names
11982 In a project that contains source files, their names may be specified by the
11983 attributes @code{Source_Files} (a string list) or @code{Source_List_File}
11984 (a string). Source file names never include any directory information.
11986 If the attribute @code{Source_Files} is given an explicit value, then each
11987 element of the list is a source file name.
11989 @smallexample @c projectfile
11990 for Source_Files use ("main.adb");
11991 for Source_Files use ("main.adb", "pack1.ads", "pack2.adb");
11995 If the attribute @code{Source_Files} is not given an explicit value,
11996 but the attribute @code{Source_List_File} is given a string value,
11997 then the source file names are contained in the text file whose path name
11998 (absolute or relative to the directory of the project file) is the
11999 value of the attribute @code{Source_List_File}.
12001 Each line in the file that is not empty or is not a comment
12002 contains a source file name.
12004 @smallexample @c projectfile
12005 for Source_List_File use "source_list.txt";
12009 By default, if neither the attribute @code{Source_Files} nor the attribute
12010 @code{Source_List_File} is given an explicit value, then each file in the
12011 source directories that conforms to the project's naming scheme
12012 (@pxref{Naming Schemes}) is an immediate source of the project.
12014 A warning is issued if both attributes @code{Source_Files} and
12015 @code{Source_List_File} are given explicit values. In this case, the attribute
12016 @code{Source_Files} prevails.
12018 Each source file name must be the name of one existing source file
12019 in one of the source directories.
12021 A @code{Source_Files} attribute whose value is an empty list
12022 indicates that there are no source files in the project.
12024 If the order of the source directories is known statically, that is if
12025 @code{"/**"} is not used in the string list @code{Source_Dirs}, then there may
12026 be several files with the same source file name. In this case, only the file
12027 in the first directory is considered as an immediate source of the project
12028 file. If the order of the source directories is not known statically, it is
12029 an error to have several files with the same source file name.
12031 Projects can be specified to have no Ada source
12032 files: the value of (@code{Source_Dirs} or @code{Source_Files} may be an empty
12033 list, or the @code{"Ada"} may be absent from @code{Languages}:
12035 @smallexample @c projectfile
12036 for Source_Dirs use ();
12037 for Source_Files use ();
12038 for Languages use ("C", "C++");
12042 Otherwise, a project must contain at least one immediate source.
12044 Projects with no source files are useful as template packages
12045 (@pxref{Packages in Project Files}) for other projects; in particular to
12046 define a package @code{Naming} (@pxref{Naming Schemes}).
12048 @c ****************************
12049 @c * Importing Projects *
12050 @c ****************************
12052 @node Importing Projects
12053 @section Importing Projects
12054 @cindex @code{ADA_PROJECT_PATH}
12057 An immediate source of a project P may depend on source files that
12058 are neither immediate sources of P nor in the predefined library.
12059 To get this effect, P must @emph{import} the projects that contain the needed
12062 @smallexample @c projectfile
12064 with "project1", "utilities.gpr";
12065 with "/namings/apex.gpr";
12072 As can be seen in this example, the syntax for importing projects is similar
12073 to the syntax for importing compilation units in Ada. However, project files
12074 use literal strings instead of names, and the @code{with} clause identifies
12075 project files rather than packages.
12077 Each literal string is the file name or path name (absolute or relative) of a
12078 project file. If a string corresponds to a file name, with no path or a
12079 relative path, then its location is determined by the @emph{project path}. The
12080 latter can be queried using @code{gnatls -v}. It contains:
12084 In first position, the directory containing the current project file.
12086 In last position, the default project directory. This default project directory
12087 is part of the GNAT installation and is the standard place to install project
12088 files giving access to standard support libraries.
12090 @ref{Installing a library}
12094 In between, all the directories referenced in the
12095 ^environment variable^logical name^ @env{ADA_PROJECT_PATH} if it exists.
12099 If a relative pathname is used, as in
12101 @smallexample @c projectfile
12106 then the full path for the project is constructed by concatenating this
12107 relative path to those in the project path, in order, until a matching file is
12108 found. Any symbolic link will be fully resolved in the directory of the
12109 importing project file before the imported project file is examined.
12111 If the @code{with}'ed project file name does not have an extension,
12112 the default is @file{^.gpr^.GPR^}. If a file with this extension is not found,
12113 then the file name as specified in the @code{with} clause (no extension) will
12114 be used. In the above example, if a file @code{project1.gpr} is found, then it
12115 will be used; otherwise, if a file @code{^project1^PROJECT1^} exists
12116 then it will be used; if neither file exists, this is an error.
12118 A warning is issued if the name of the project file does not match the
12119 name of the project; this check is case insensitive.
12121 Any source file that is an immediate source of the imported project can be
12122 used by the immediate sources of the importing project, transitively. Thus
12123 if @code{A} imports @code{B}, and @code{B} imports @code{C}, the immediate
12124 sources of @code{A} may depend on the immediate sources of @code{C}, even if
12125 @code{A} does not import @code{C} explicitly. However, this is not recommended,
12126 because if and when @code{B} ceases to import @code{C}, some sources in
12127 @code{A} will no longer compile.
12129 A side effect of this capability is that normally cyclic dependencies are not
12130 permitted: if @code{A} imports @code{B} (directly or indirectly) then @code{B}
12131 is not allowed to import @code{A}. However, there are cases when cyclic
12132 dependencies would be beneficial. For these cases, another form of import
12133 between projects exists, the @code{limited with}: a project @code{A} that
12134 imports a project @code{B} with a straigh @code{with} may also be imported,
12135 directly or indirectly, by @code{B} on the condition that imports from @code{B}
12136 to @code{A} include at least one @code{limited with}.
12138 @smallexample @c 0projectfile
12144 limited with "../a/a.gpr";
12152 limited with "../a/a.gpr";
12158 In the above legal example, there are two project cycles:
12161 @item A -> C -> D -> A
12165 In each of these cycle there is one @code{limited with}: import of @code{A}
12166 from @code{B} and import of @code{A} from @code{D}.
12168 The difference between straight @code{with} and @code{limited with} is that
12169 the name of a project imported with a @code{limited with} cannot be used in the
12170 project that imports it. In particular, its packages cannot be renamed and
12171 its variables cannot be referred to.
12173 An exception to the above rules for @code{limited with} is that for the main
12174 project specified to @command{gnatmake} or to the @command{GNAT} driver a
12175 @code{limited with} is equivalent to a straight @code{with}. For example,
12176 in the example above, projects @code{B} and @code{D} could not be main
12177 projects for @command{gnatmake} or to the @command{GNAT} driver, because they
12178 each have a @code{limited with} that is the only one in a cycle of importing
12181 @c *********************
12182 @c * Project Extension *
12183 @c *********************
12185 @node Project Extension
12186 @section Project Extension
12189 During development of a large system, it is sometimes necessary to use
12190 modified versions of some of the source files, without changing the original
12191 sources. This can be achieved through the @emph{project extension} facility.
12193 @smallexample @c projectfile
12194 project Modified_Utilities extends "/baseline/utilities.gpr" is ...
12198 A project extension declaration introduces an extending project
12199 (the @emph{child}) and a project being extended (the @emph{parent}).
12201 By default, a child project inherits all the sources of its parent.
12202 However, inherited sources can be overridden: a unit in a parent is hidden
12203 by a unit of the same name in the child.
12205 Inherited sources are considered to be sources (but not immediate sources)
12206 of the child project; see @ref{Project File Syntax}.
12208 An inherited source file retains any switches specified in the parent project.
12210 For example if the project @code{Utilities} contains the specification and the
12211 body of an Ada package @code{Util_IO}, then the project
12212 @code{Modified_Utilities} can contain a new body for package @code{Util_IO}.
12213 The original body of @code{Util_IO} will not be considered in program builds.
12214 However, the package specification will still be found in the project
12217 A child project can have only one parent but it may import any number of other
12220 A project is not allowed to import directly or indirectly at the same time a
12221 child project and any of its ancestors.
12223 @c *******************************
12224 @c * Project Hierarchy Extension *
12225 @c *******************************
12227 @node Project Hierarchy Extension
12228 @section Project Hierarchy Extension
12231 When extending a large system spanning multiple projects, it is often
12232 inconvenient to extend every project in the hierarchy that is impacted by a
12233 small change introduced. In such cases, it is possible to create a virtual
12234 extension of entire hierarchy using @code{extends all} relationship.
12236 When the project is extended using @code{extends all} inheritance, all projects
12237 that are imported by it, both directly and indirectly, are considered virtually
12238 extended. That is, the Project Manager creates "virtual projects"
12239 that extend every project in the hierarchy; all these virtual projects have
12240 no sources of their own and have as object directory the object directory of
12241 the root of "extending all" project.
12243 It is possible to explicitly extend one or more projects in the hierarchy
12244 in order to modify the sources. These extending projects must be imported by
12245 the "extending all" project, which will replace the corresponding virtual
12246 projects with the explicit ones.
12248 When building such a project hierarchy extension, the Project Manager will
12249 ensure that both modified sources and sources in virtual extending projects
12250 that depend on them, are recompiled.
12252 By means of example, consider the following hierarchy of projects.
12256 project A, containing package P1
12258 project B importing A and containing package P2 which depends on P1
12260 project C importing B and containing package P3 which depends on P2
12264 We want to modify packages P1 and P3.
12266 This project hierarchy will need to be extended as follows:
12270 Create project A1 that extends A, placing modified P1 there:
12272 @smallexample @c 0projectfile
12273 project A1 extends "(...)/A" is
12278 Create project C1 that "extends all" C and imports A1, placing modified
12281 @smallexample @c 0projectfile
12283 project C1 extends all "(...)/C" is
12288 When you build project C1, your entire modified project space will be
12289 recompiled, including the virtual project B1 that has been impacted by the
12290 "extending all" inheritance of project C.
12292 Note that if a Library Project in the hierarchy is virtually extended,
12293 the virtual project that extends the Library Project is not a Library Project.
12295 @c ****************************************
12296 @c * External References in Project Files *
12297 @c ****************************************
12299 @node External References in Project Files
12300 @section External References in Project Files
12303 A project file may contain references to external variables; such references
12304 are called @emph{external references}.
12306 An external variable is either defined as part of the environment (an
12307 environment variable in Unix, for example) or else specified on the command
12308 line via the @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
12309 If both, then the command line value is used.
12311 The value of an external reference is obtained by means of the built-in
12312 function @code{external}, which returns a string value.
12313 This function has two forms:
12315 @item @code{external (external_variable_name)}
12316 @item @code{external (external_variable_name, default_value)}
12320 Each parameter must be a string literal. For example:
12322 @smallexample @c projectfile
12324 external ("OS", "GNU/Linux")
12328 In the form with one parameter, the function returns the value of
12329 the external variable given as parameter. If this name is not present in the
12330 environment, the function returns an empty string.
12332 In the form with two string parameters, the second argument is
12333 the value returned when the variable given as the first argument is not
12334 present in the environment. In the example above, if @code{"OS"} is not
12335 the name of ^an environment variable^a logical name^ and is not passed on
12336 the command line, then the returned value is @code{"GNU/Linux"}.
12338 An external reference may be part of a string expression or of a string
12339 list expression, and can therefore appear in a variable declaration or
12340 an attribute declaration.
12342 @smallexample @c projectfile
12344 type Mode_Type is ("Debug", "Release");
12345 Mode : Mode_Type := external ("MODE");
12352 @c *****************************
12353 @c * Packages in Project Files *
12354 @c *****************************
12356 @node Packages in Project Files
12357 @section Packages in Project Files
12360 A @emph{package} defines the settings for project-aware tools within a
12362 For each such tool one can declare a package; the names for these
12363 packages are preset (@pxref{Packages}).
12364 A package may contain variable declarations, attribute declarations, and case
12367 @smallexample @c projectfile
12370 package Builder is -- used by gnatmake
12371 for ^Default_Switches^Default_Switches^ ("Ada")
12380 The syntax of package declarations mimics that of package in Ada.
12382 Most of the packages have an attribute
12383 @code{^Default_Switches^Default_Switches^}.
12384 This attribute is an associative array, and its value is a string list.
12385 The index of the associative array is the name of a programming language (case
12386 insensitive). This attribute indicates the ^switch^switch^
12387 or ^switches^switches^ to be used
12388 with the corresponding tool.
12390 Some packages also have another attribute, @code{^Switches^Switches^},
12391 an associative array whose value is a string list.
12392 The index is the name of a source file.
12393 This attribute indicates the ^switch^switch^
12394 or ^switches^switches^ to be used by the corresponding
12395 tool when dealing with this specific file.
12397 Further information on these ^switch^switch^-related attributes is found in
12398 @ref{^Switches^Switches^ and Project Files}.
12400 A package may be declared as a @emph{renaming} of another package; e.g., from
12401 the project file for an imported project.
12403 @smallexample @c projectfile
12405 with "/global/apex.gpr";
12407 package Naming renames Apex.Naming;
12414 Packages that are renamed in other project files often come from project files
12415 that have no sources: they are just used as templates. Any modification in the
12416 template will be reflected automatically in all the project files that rename
12417 a package from the template.
12419 In addition to the tool-oriented packages, you can also declare a package
12420 named @code{Naming} to establish specialized source file naming conventions
12421 (@pxref{Naming Schemes}).
12423 @c ************************************
12424 @c * Variables from Imported Projects *
12425 @c ************************************
12427 @node Variables from Imported Projects
12428 @section Variables from Imported Projects
12431 An attribute or variable defined in an imported or parent project can
12432 be used in expressions in the importing / extending project.
12433 Such an attribute or variable is denoted by an expanded name whose prefix
12434 is either the name of the project or the expanded name of a package within
12437 @smallexample @c projectfile
12440 project Main extends "base" is
12441 Var1 := Imported.Var;
12442 Var2 := Base.Var & ".new";
12447 for ^Default_Switches^Default_Switches^ ("Ada")
12448 use Imported.Builder.Ada_^Switches^Switches^ &
12449 "^-gnatg^-gnatg^" &
12455 package Compiler is
12456 for ^Default_Switches^Default_Switches^ ("Ada")
12457 use Base.Compiler.Ada_^Switches^Switches^;
12468 The value of @code{Var1} is a copy of the variable @code{Var} defined
12469 in the project file @file{"imported.gpr"}
12471 the value of @code{Var2} is a copy of the value of variable @code{Var}
12472 defined in the project file @file{base.gpr}, concatenated with @code{".new"}
12474 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
12475 @code{Builder} is a string list that includes in its value a copy of the value
12476 of @code{Ada_^Switches^Switches^} defined in the @code{Builder} package
12477 in project file @file{imported.gpr} plus two new elements:
12478 @option{"^-gnatg^-gnatg^"}
12479 and @option{"^-v^-v^"};
12481 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
12482 @code{Compiler} is a copy of the variable @code{Ada_^Switches^Switches^}
12483 defined in the @code{Compiler} package in project file @file{base.gpr},
12484 the project being extended.
12487 @c ******************
12488 @c * Naming Schemes *
12489 @c ******************
12491 @node Naming Schemes
12492 @section Naming Schemes
12495 Sometimes an Ada software system is ported from a foreign compilation
12496 environment to GNAT, and the file names do not use the default GNAT
12497 conventions. Instead of changing all the file names (which for a variety
12498 of reasons might not be possible), you can define the relevant file
12499 naming scheme in the @code{Naming} package in your project file.
12502 Note that the use of pragmas described in
12503 @ref{Alternative File Naming Schemes} by mean of a configuration
12504 pragmas file is not supported when using project files. You must use
12505 the features described in this paragraph. You can however use specify
12506 other configuration pragmas (@pxref{Specifying Configuration Pragmas}).
12509 For example, the following
12510 package models the Apex file naming rules:
12512 @smallexample @c projectfile
12515 for Casing use "lowercase";
12516 for Dot_Replacement use ".";
12517 for Spec_Suffix ("Ada") use ".1.ada";
12518 for Body_Suffix ("Ada") use ".2.ada";
12525 For example, the following package models the DEC Ada file naming rules:
12527 @smallexample @c projectfile
12530 for Casing use "lowercase";
12531 for Dot_Replacement use "__";
12532 for Spec_Suffix ("Ada") use "_.^ada^ada^";
12533 for Body_Suffix ("Ada") use ".^ada^ada^";
12539 (Note that @code{Casing} is @code{"lowercase"} because GNAT gets the file
12540 names in lower case)
12544 You can define the following attributes in package @code{Naming}:
12549 This must be a string with one of the three values @code{"lowercase"},
12550 @code{"uppercase"} or @code{"mixedcase"}; these strings are case insensitive.
12553 If @var{Casing} is not specified, then the default is @code{"lowercase"}.
12555 @item @var{Dot_Replacement}
12556 This must be a string whose value satisfies the following conditions:
12559 @item It must not be empty
12560 @item It cannot start or end with an alphanumeric character
12561 @item It cannot be a single underscore
12562 @item It cannot start with an underscore followed by an alphanumeric
12563 @item It cannot contain a dot @code{'.'} except if the entire string
12568 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
12570 @item @var{Spec_Suffix}
12571 This is an associative array (indexed by the programming language name, case
12572 insensitive) whose value is a string that must satisfy the following
12576 @item It must not be empty
12577 @item It must include at least one dot
12580 If @code{Spec_Suffix ("Ada")} is not specified, then the default is
12581 @code{"^.ads^.ADS^"}.
12583 @item @var{Body_Suffix}
12584 This is an associative array (indexed by the programming language name, case
12585 insensitive) whose value is a string that must satisfy the following
12589 @item It must not be empty
12590 @item It must include at least one dot
12591 @item It cannot end with the same string as @code{Spec_Suffix ("Ada")}
12594 If @code{Body_Suffix ("Ada")} is not specified, then the default is
12595 @code{"^.adb^.ADB^"}.
12597 @item @var{Separate_Suffix}
12598 This must be a string whose value satisfies the same conditions as
12599 @code{Body_Suffix}.
12602 If @code{Separate_Suffix ("Ada")} is not specified, then it defaults to same
12603 value as @code{Body_Suffix ("Ada")}.
12607 You can use the associative array attribute @code{Spec} to define
12608 the source file name for an individual Ada compilation unit's spec. The array
12609 index must be a string literal that identifies the Ada unit (case insensitive).
12610 The value of this attribute must be a string that identifies the file that
12611 contains this unit's spec (case sensitive or insensitive depending on the
12614 @smallexample @c projectfile
12615 for Spec ("MyPack.MyChild") use "mypack.mychild.spec";
12620 You can use the associative array attribute @code{Body} to
12621 define the source file name for an individual Ada compilation unit's body
12622 (possibly a subunit). The array index must be a string literal that identifies
12623 the Ada unit (case insensitive). The value of this attribute must be a string
12624 that identifies the file that contains this unit's body or subunit (case
12625 sensitive or insensitive depending on the operating system).
12627 @smallexample @c projectfile
12628 for Body ("MyPack.MyChild") use "mypack.mychild.body";
12632 @c ********************
12633 @c * Library Projects *
12634 @c ********************
12636 @node Library Projects
12637 @section Library Projects
12640 @emph{Library projects} are projects whose object code is placed in a library.
12641 (Note that this facility is not yet supported on all platforms)
12643 To create a library project, you need to define in its project file
12644 two project-level attributes: @code{Library_Name} and @code{Library_Dir}.
12645 Additionally, you may define the library-related attributes
12646 @code{Library_Kind}, @code{Library_Version}, @code{Library_Interface},
12647 @code{Library_Auto_Init}, @code{Library_Options} and @code{Library_GCC}.
12649 The @code{Library_Name} attribute has a string value. There is no restriction
12650 on the name of a library. It is the responsibility of the developer to
12651 choose a name that will be accepted by the platform. It is recommended to
12652 choose names that could be Ada identifiers; such names are almost guaranteed
12653 to be acceptable on all platforms.
12655 The @code{Library_Dir} attribute has a string value that designates the path
12656 (absolute or relative) of the directory where the library will reside.
12657 It must designate an existing directory, and this directory must be
12658 different from the project's object directory. It also needs to be writable.
12659 The directory should only be used for one library; the reason is that all
12660 files contained in this directory may be deleted by the Project Manager.
12662 If both @code{Library_Name} and @code{Library_Dir} are specified and
12663 are legal, then the project file defines a library project. The optional
12664 library-related attributes are checked only for such project files.
12666 The @code{Library_Kind} attribute has a string value that must be one of the
12667 following (case insensitive): @code{"static"}, @code{"dynamic"} or
12668 @code{"relocatable"} (which is a synonym for @code{"dynamic"}). If this
12669 attribute is not specified, the library is a static library, that is
12670 an archive of object files that can be potentially linked into an
12671 static executable. Otherwise, the library may be dynamic or
12672 relocatable, that is a library that is loaded only at the start of execution.
12674 If you need to build both a static and a dynamic library, you should use two
12675 different object directories, since in some cases some extra code needs to
12676 be generated for the latter. For such cases, it is recommended to either use
12677 two different project files, or a single one which uses external variables
12678 to indicate what kind of library should be build.
12680 The @code{Library_Version} attribute has a string value whose interpretation
12681 is platform dependent. It has no effect on VMS and Windows. On Unix, it is
12682 used only for dynamic/relocatable libraries as the internal name of the
12683 library (the @code{"soname"}). If the library file name (built from the
12684 @code{Library_Name}) is different from the @code{Library_Version}, then the
12685 library file will be a symbolic link to the actual file whose name will be
12686 @code{Library_Version}.
12690 @smallexample @c projectfile
12696 for Library_Dir use "lib_dir";
12697 for Library_Name use "dummy";
12698 for Library_Kind use "relocatable";
12699 for Library_Version use "libdummy.so." & Version;
12706 Directory @file{lib_dir} will contain the internal library file whose name
12707 will be @file{libdummy.so.1}, and @file{libdummy.so} will be a symbolic link to
12708 @file{libdummy.so.1}.
12710 When @command{gnatmake} detects that a project file
12711 is a library project file, it will check all immediate sources of the project
12712 and rebuild the library if any of the sources have been recompiled.
12714 Standard project files can import library project files. In such cases,
12715 the libraries will only be rebuild if some of its sources are recompiled
12716 because they are in the closure of some other source in an importing project.
12717 Sources of the library project files that are not in such a closure will
12718 not be checked, unless the full library is checked, because one of its sources
12719 needs to be recompiled.
12721 For instance, assume the project file @code{A} imports the library project file
12722 @code{L}. The immediate sources of A are @file{a1.adb}, @file{a2.ads} and
12723 @file{a2.adb}. The immediate sources of L are @file{l1.ads}, @file{l1.adb},
12724 @file{l2.ads}, @file{l2.adb}.
12726 If @file{l1.adb} has been modified, then the library associated with @code{L}
12727 will be rebuild when compiling all the immediate sources of @code{A} only
12728 if @file{a1.ads}, @file{a2.ads} or @file{a2.adb} includes a statement
12731 To be sure that all the sources in the library associated with @code{L} are
12732 up to date, and that all the sources of project @code{A} are also up to date,
12733 the following two commands needs to be used:
12740 When a library is built or rebuilt, an attempt is made first to delete all
12741 files in the library directory.
12742 All @file{ALI} files will also be copied from the object directory to the
12743 library directory. To build executables, @command{gnatmake} will use the
12744 library rather than the individual object files.
12747 It is also possible to create library project files for third-party libraries
12748 that are precompiled and cannot be compiled locally thanks to the
12749 @code{externally_built} attribute. (See @ref{Installing a library}).
12752 @c *******************************
12753 @c * Stand-alone Library Projects *
12754 @c *******************************
12756 @node Stand-alone Library Projects
12757 @section Stand-alone Library Projects
12760 A Stand-alone Library is a library that contains the necessary code to
12761 elaborate the Ada units that are included in the library. A Stand-alone
12762 Library is suitable to be used in an executable when the main is not
12763 in Ada. However, Stand-alone Libraries may also be used with an Ada main
12766 A Stand-alone Library Project is a Library Project where the library is
12767 a Stand-alone Library.
12769 To be a Stand-alone Library Project, in addition to the two attributes
12770 that make a project a Library Project (@code{Library_Name} and
12771 @code{Library_Dir}, see @ref{Library Projects}), the attribute
12772 @code{Library_Interface} must be defined.
12774 @smallexample @c projectfile
12776 for Library_Dir use "lib_dir";
12777 for Library_Name use "dummy";
12778 for Library_Interface use ("int1", "int1.child");
12782 Attribute @code{Library_Interface} has a non empty string list value,
12783 each string in the list designating a unit contained in an immediate source
12784 of the project file.
12786 When a Stand-alone Library is built, first the binder is invoked to build
12787 a package whose name depends on the library name
12788 (^b~dummy.ads/b^B$DUMMY.ADS/B^ in the example above).
12789 This binder-generated package includes initialization and
12790 finalization procedures whose
12791 names depend on the library name (dummyinit and dummyfinal in the example
12792 above). The object corresponding to this package is included in the library.
12794 A dynamic or relocatable Stand-alone Library is automatically initialized
12795 if automatic initialization of Stand-alone Libraries is supported on the
12796 platform and if attribute @code{Library_Auto_Init} is not specified or
12797 is specified with the value "true". A static Stand-alone Library is never
12798 automatically initialized.
12800 Single string attribute @code{Library_Auto_Init} may be specified with only
12801 two possible values: "false" or "true" (case-insensitive). Specifying
12802 "false" for attribute @code{Library_Auto_Init} will prevent automatic
12803 initialization of dynamic or relocatable libraries.
12805 When a non automatically initialized Stand-alone Library is used
12806 in an executable, its initialization procedure must be called before
12807 any service of the library is used.
12808 When the main subprogram is in Ada, it may mean that the initialization
12809 procedure has to be called during elaboration of another package.
12811 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
12812 (those that are listed in attribute @code{Library_Interface}) are copied to
12813 the Library Directory. As a consequence, only the Interface Units may be
12814 imported from Ada units outside of the library. If other units are imported,
12815 the binding phase will fail.
12817 When a Stand-Alone Library is bound, the switches that are specified in
12818 the attribute @code{Default_Switches ("Ada")} in package @code{Binder} are
12819 used in the call to @command{gnatbind}.
12821 The string list attribute @code{Library_Options} may be used to specified
12822 additional switches to the call to @command{gcc} to link the library.
12824 The attribute @code{Library_Src_Dir}, may be specified for a
12825 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
12826 single string value. Its value must be the path (absolute or relative to the
12827 project directory) of an existing directory. This directory cannot be the
12828 object directory or one of the source directories, but it can be the same as
12829 the library directory. The sources of the Interface
12830 Units of the library, necessary to an Ada client of the library, will be
12831 copied to the designated directory, called Interface Copy directory.
12832 These sources includes the specs of the Interface Units, but they may also
12833 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
12834 are used, or when there is a generic units in the spec. Before the sources
12835 are copied to the Interface Copy directory, an attempt is made to delete all
12836 files in the Interface Copy directory.
12838 @c *************************************
12839 @c * Switches Related to Project Files *
12840 @c *************************************
12841 @node Switches Related to Project Files
12842 @section Switches Related to Project Files
12845 The following switches are used by GNAT tools that support project files:
12849 @item ^-P^/PROJECT_FILE=^@var{project}
12850 @cindex @option{^-P^/PROJECT_FILE^} (any tool supporting project files)
12851 Indicates the name of a project file. This project file will be parsed with
12852 the verbosity indicated by @option{^-vP^MESSAGE_PROJECT_FILES=^@emph{x}},
12853 if any, and using the external references indicated
12854 by @option{^-X^/EXTERNAL_REFERENCE^} switches, if any.
12856 There may zero, one or more spaces between @option{-P} and @var{project}.
12860 There must be only one @option{^-P^/PROJECT_FILE^} switch on the command line.
12863 Since the Project Manager parses the project file only after all the switches
12864 on the command line are checked, the order of the switches
12865 @option{^-P^/PROJECT_FILE^},
12866 @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}}
12867 or @option{^-X^/EXTERNAL_REFERENCE^} is not significant.
12869 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
12870 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (any tool supporting project files)
12871 Indicates that external variable @var{name} has the value @var{value}.
12872 The Project Manager will use this value for occurrences of
12873 @code{external(name)} when parsing the project file.
12877 If @var{name} or @var{value} includes a space, then @var{name=value} should be
12878 put between quotes.
12886 Several @option{^-X^/EXTERNAL_REFERENCE^} switches can be used simultaneously.
12887 If several @option{^-X^/EXTERNAL_REFERENCE^} switches specify the same
12888 @var{name}, only the last one is used.
12891 An external variable specified with a @option{^-X^/EXTERNAL_REFERENCE^} switch
12892 takes precedence over the value of the same name in the environment.
12894 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
12895 @cindex @code{^-vP^/MESSAGES_PROJECT_FILE^} (any tool supporting project files)
12896 @c Previous line uses code vs option command, to stay less than 80 chars
12897 Indicates the verbosity of the parsing of GNAT project files.
12900 @option{-vP0} means Default;
12901 @option{-vP1} means Medium;
12902 @option{-vP2} means High.
12906 There are three possible options for this qualifier: DEFAULT, MEDIUM and
12911 The default is ^Default^DEFAULT^: no output for syntactically correct
12914 If several @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}} switches are present,
12915 only the last one is used.
12919 @c **********************************
12920 @c * Tools Supporting Project Files *
12921 @c **********************************
12923 @node Tools Supporting Project Files
12924 @section Tools Supporting Project Files
12927 * gnatmake and Project Files::
12928 * The GNAT Driver and Project Files::
12930 * Glide and Project Files::
12934 @node gnatmake and Project Files
12935 @subsection gnatmake and Project Files
12938 This section covers several topics related to @command{gnatmake} and
12939 project files: defining ^switches^switches^ for @command{gnatmake}
12940 and for the tools that it invokes; specifying configuration pragmas;
12941 the use of the @code{Main} attribute; building and rebuilding library project
12945 * ^Switches^Switches^ and Project Files::
12946 * Specifying Configuration Pragmas::
12947 * Project Files and Main Subprograms::
12948 * Library Project Files::
12951 @node ^Switches^Switches^ and Project Files
12952 @subsubsection ^Switches^Switches^ and Project Files
12955 It is not currently possible to specify VMS style qualifiers in the project
12956 files; only Unix style ^switches^switches^ may be specified.
12960 For each of the packages @code{Builder}, @code{Compiler}, @code{Binder}, and
12961 @code{Linker}, you can specify a @code{^Default_Switches^Default_Switches^}
12962 attribute, a @code{^Switches^Switches^} attribute, or both;
12963 as their names imply, these ^switch^switch^-related
12964 attributes affect the ^switches^switches^ that are used for each of these GNAT
12966 @command{gnatmake} is invoked. As will be explained below, these
12967 component-specific ^switches^switches^ precede
12968 the ^switches^switches^ provided on the @command{gnatmake} command line.
12970 The @code{^Default_Switches^Default_Switches^} attribute is an associative
12971 array indexed by language name (case insensitive) whose value is a string list.
12974 @smallexample @c projectfile
12976 package Compiler is
12977 for ^Default_Switches^Default_Switches^ ("Ada")
12978 use ("^-gnaty^-gnaty^",
12985 The @code{^Switches^Switches^} attribute is also an associative array,
12986 indexed by a file name (which may or may not be case sensitive, depending
12987 on the operating system) whose value is a string list. For example:
12989 @smallexample @c projectfile
12992 for ^Switches^Switches^ ("main1.adb")
12994 for ^Switches^Switches^ ("main2.adb")
13001 For the @code{Builder} package, the file names must designate source files
13002 for main subprograms. For the @code{Binder} and @code{Linker} packages, the
13003 file names must designate @file{ALI} or source files for main subprograms.
13004 In each case just the file name without an explicit extension is acceptable.
13006 For each tool used in a program build (@command{gnatmake}, the compiler, the
13007 binder, and the linker), the corresponding package @dfn{contributes} a set of
13008 ^switches^switches^ for each file on which the tool is invoked, based on the
13009 ^switch^switch^-related attributes defined in the package.
13010 In particular, the ^switches^switches^
13011 that each of these packages contributes for a given file @var{f} comprise:
13015 the value of attribute @code{^Switches^Switches^ (@var{f})},
13016 if it is specified in the package for the given file,
13018 otherwise, the value of @code{^Default_Switches^Default_Switches^ ("Ada")},
13019 if it is specified in the package.
13023 If neither of these attributes is defined in the package, then the package does
13024 not contribute any ^switches^switches^ for the given file.
13026 When @command{gnatmake} is invoked on a file, the ^switches^switches^ comprise
13027 two sets, in the following order: those contributed for the file
13028 by the @code{Builder} package;
13029 and the switches passed on the command line.
13031 When @command{gnatmake} invokes a tool (compiler, binder, linker) on a file,
13032 the ^switches^switches^ passed to the tool comprise three sets,
13033 in the following order:
13037 the applicable ^switches^switches^ contributed for the file
13038 by the @code{Builder} package in the project file supplied on the command line;
13041 those contributed for the file by the package (in the relevant project file --
13042 see below) corresponding to the tool; and
13045 the applicable switches passed on the command line.
13049 The term @emph{applicable ^switches^switches^} reflects the fact that
13050 @command{gnatmake} ^switches^switches^ may or may not be passed to individual
13051 tools, depending on the individual ^switch^switch^.
13053 @command{gnatmake} may invoke the compiler on source files from different
13054 projects. The Project Manager will use the appropriate project file to
13055 determine the @code{Compiler} package for each source file being compiled.
13056 Likewise for the @code{Binder} and @code{Linker} packages.
13058 As an example, consider the following package in a project file:
13060 @smallexample @c projectfile
13063 package Compiler is
13064 for ^Default_Switches^Default_Switches^ ("Ada")
13066 for ^Switches^Switches^ ("a.adb")
13068 for ^Switches^Switches^ ("b.adb")
13070 "^-gnaty^-gnaty^");
13077 If @command{gnatmake} is invoked with this project file, and it needs to
13078 compile, say, the files @file{a.adb}, @file{b.adb}, and @file{c.adb}, then
13079 @file{a.adb} will be compiled with the ^switch^switch^
13080 @option{^-O1^-O1^},
13081 @file{b.adb} with ^switches^switches^
13083 and @option{^-gnaty^-gnaty^},
13084 and @file{c.adb} with @option{^-g^-g^}.
13086 The following example illustrates the ordering of the ^switches^switches^
13087 contributed by different packages:
13089 @smallexample @c projectfile
13093 for ^Switches^Switches^ ("main.adb")
13101 package Compiler is
13102 for ^Switches^Switches^ ("main.adb")
13110 If you issue the command:
13113 gnatmake ^-Pproj2^/PROJECT_FILE=PROJ2^ -O0 main
13117 then the compiler will be invoked on @file{main.adb} with the following
13118 sequence of ^switches^switches^
13121 ^-g -O1 -O2 -O0^-g -O1 -O2 -O0^
13124 with the last @option{^-O^-O^}
13125 ^switch^switch^ having precedence over the earlier ones;
13126 several other ^switches^switches^
13127 (such as @option{^-c^-c^}) are added implicitly.
13129 The ^switches^switches^
13131 and @option{^-O1^-O1^} are contributed by package
13132 @code{Builder}, @option{^-O2^-O2^} is contributed
13133 by the package @code{Compiler}
13134 and @option{^-O0^-O0^} comes from the command line.
13136 The @option{^-g^-g^}
13137 ^switch^switch^ will also be passed in the invocation of
13138 @command{Gnatlink.}
13140 A final example illustrates switch contributions from packages in different
13143 @smallexample @c projectfile
13146 for Source_Files use ("pack.ads", "pack.adb");
13147 package Compiler is
13148 for ^Default_Switches^Default_Switches^ ("Ada")
13149 use ("^-gnata^-gnata^");
13157 for Source_Files use ("foo_main.adb", "bar_main.adb");
13159 for ^Switches^Switches^ ("foo_main.adb")
13167 -- Ada source file:
13169 procedure Foo_Main is
13177 gnatmake ^-PProj4^/PROJECT_FILE=PROJ4^ foo_main.adb -cargs -gnato
13181 then the ^switches^switches^ passed to the compiler for @file{foo_main.adb} are
13182 @option{^-g^-g^} (contributed by the package @code{Proj4.Builder}) and
13183 @option{^-gnato^-gnato^} (passed on the command line).
13184 When the imported package @code{Pack} is compiled, the ^switches^switches^ used
13185 are @option{^-g^-g^} from @code{Proj4.Builder},
13186 @option{^-gnata^-gnata^} (contributed from package @code{Proj3.Compiler},
13187 and @option{^-gnato^-gnato^} from the command line.
13190 When using @command{gnatmake} with project files, some ^switches^switches^ or
13191 arguments may be expressed as relative paths. As the working directory where
13192 compilation occurs may change, these relative paths are converted to absolute
13193 paths. For the ^switches^switches^ found in a project file, the relative paths
13194 are relative to the project file directory, for the switches on the command
13195 line, they are relative to the directory where @command{gnatmake} is invoked.
13196 The ^switches^switches^ for which this occurs are:
13202 ^-aI^-aI^, as well as all arguments that are not switches (arguments to
13204 ^-o^-o^, object files specified in package @code{Linker} or after
13205 -largs on the command line). The exception to this rule is the ^switch^switch^
13206 ^--RTS=^--RTS=^ for which a relative path argument is never converted.
13208 @node Specifying Configuration Pragmas
13209 @subsubsection Specifying Configuration Pragmas
13211 When using @command{gnatmake} with project files, if there exists a file
13212 @file{gnat.adc} that contains configuration pragmas, this file will be
13215 Configuration pragmas can be defined by means of the following attributes in
13216 project files: @code{Global_Configuration_Pragmas} in package @code{Builder}
13217 and @code{Local_Configuration_Pragmas} in package @code{Compiler}.
13219 Both these attributes are single string attributes. Their values is the path
13220 name of a file containing configuration pragmas. If a path name is relative,
13221 then it is relative to the project directory of the project file where the
13222 attribute is defined.
13224 When compiling a source, the configuration pragmas used are, in order,
13225 those listed in the file designated by attribute
13226 @code{Global_Configuration_Pragmas} in package @code{Builder} of the main
13227 project file, if it is specified, and those listed in the file designated by
13228 attribute @code{Local_Configuration_Pragmas} in package @code{Compiler} of
13229 the project file of the source, if it exists.
13231 @node Project Files and Main Subprograms
13232 @subsubsection Project Files and Main Subprograms
13235 When using a project file, you can invoke @command{gnatmake}
13236 with one or several main subprograms, by specifying their source files on the
13240 gnatmake ^-P^/PROJECT_FILE=^prj main1 main2 main3
13244 Each of these needs to be a source file of the same project, except
13245 when the switch ^-u^/UNIQUE^ is used.
13248 When ^-u^/UNIQUE^ is not used, all the mains need to be sources of the
13249 same project, one of the project in the tree rooted at the project specified
13250 on the command line. The package @code{Builder} of this common project, the
13251 "main project" is the one that is considered by @command{gnatmake}.
13254 When ^-u^/UNIQUE^ is used, the specified source files may be in projects
13255 imported directly or indirectly by the project specified on the command line.
13256 Note that if such a source file is not part of the project specified on the
13257 command line, the ^switches^switches^ found in package @code{Builder} of the
13258 project specified on the command line, if any, that are transmitted
13259 to the compiler will still be used, not those found in the project file of
13263 When using a project file, you can also invoke @command{gnatmake} without
13264 explicitly specifying any main, and the effect depends on whether you have
13265 defined the @code{Main} attribute. This attribute has a string list value,
13266 where each element in the list is the name of a source file (the file
13267 extension is optional) that contains a unit that can be a main subprogram.
13269 If the @code{Main} attribute is defined in a project file as a non-empty
13270 string list and the switch @option{^-u^/UNIQUE^} is not used on the command
13271 line, then invoking @command{gnatmake} with this project file but without any
13272 main on the command line is equivalent to invoking @command{gnatmake} with all
13273 the file names in the @code{Main} attribute on the command line.
13276 @smallexample @c projectfile
13279 for Main use ("main1", "main2", "main3");
13285 With this project file, @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^"}
13287 @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^ main1 main2 main3"}.
13289 When the project attribute @code{Main} is not specified, or is specified
13290 as an empty string list, or when the switch @option{-u} is used on the command
13291 line, then invoking @command{gnatmake} with no main on the command line will
13292 result in all immediate sources of the project file being checked, and
13293 potentially recompiled. Depending on the presence of the switch @option{-u},
13294 sources from other project files on which the immediate sources of the main
13295 project file depend are also checked and potentially recompiled. In other
13296 words, the @option{-u} switch is applied to all of the immediate sources of the
13299 When no main is specified on the command line and attribute @code{Main} exists
13300 and includes several mains, or when several mains are specified on the
13301 command line, the default ^switches^switches^ in package @code{Builder} will
13302 be used for all mains, even if there are specific ^switches^switches^
13303 specified for one or several mains.
13305 But the ^switches^switches^ from package @code{Binder} or @code{Linker} will be
13306 the specific ^switches^switches^ for each main, if they are specified.
13308 @node Library Project Files
13309 @subsubsection Library Project Files
13312 When @command{gnatmake} is invoked with a main project file that is a library
13313 project file, it is not allowed to specify one or more mains on the command
13317 When a library project file is specified, switches ^-b^/ACTION=BIND^ and
13318 ^-l^/ACTION=LINK^ have special meanings.
13321 @item ^-b^/ACTION=BIND^ is only allowed for stand-alone libraries. It indicates
13322 to @command{gnatmake} that @command{gnatbind} should be invoked for the
13325 @item ^-l^/ACTION=LINK^ may be used for all library projects. It indicates
13326 to @command{gnatmake} that the binder generated file should be compiled
13327 (in the case of a stand-alone library) and that the library should be built.
13331 @node The GNAT Driver and Project Files
13332 @subsection The GNAT Driver and Project Files
13335 A number of GNAT tools, other than @command{^gnatmake^gnatmake^}
13337 @command{^gnatbind^gnatbind^},
13338 @command{^gnatfind^gnatfind^},
13339 @command{^gnatlink^gnatlink^},
13340 @command{^gnatls^gnatls^},
13341 @command{^gnatelim^gnatelim^},
13342 @command{^gnatpp^gnatpp^},
13343 @command{^gnatmetric^gnatmetric^},
13344 @command{^gnatstub^gnatstub^},
13345 and @command{^gnatxref^gnatxref^}. However, none of these tools can be invoked
13346 directly with a project file switch (@option{^-P^/PROJECT_FILE=^}).
13347 They must be invoked through the @command{gnat} driver.
13349 The @command{gnat} driver is a front-end that accepts a number of commands and
13350 call the corresponding tool. It has been designed initially for VMS to convert
13351 VMS style qualifiers to Unix style switches, but it is now available to all
13352 the GNAT supported platforms.
13354 On non VMS platforms, the @command{gnat} driver accepts the following commands
13355 (case insensitive):
13359 BIND to invoke @command{^gnatbind^gnatbind^}
13361 CHOP to invoke @command{^gnatchop^gnatchop^}
13363 CLEAN to invoke @command{^gnatclean^gnatclean^}
13365 COMP or COMPILE to invoke the compiler
13367 ELIM to invoke @command{^gnatelim^gnatelim^}
13369 FIND to invoke @command{^gnatfind^gnatfind^}
13371 KR or KRUNCH to invoke @command{^gnatkr^gnatkr^}
13373 LINK to invoke @command{^gnatlink^gnatlink^}
13375 LS or LIST to invoke @command{^gnatls^gnatls^}
13377 MAKE to invoke @command{^gnatmake^gnatmake^}
13379 NAME to invoke @command{^gnatname^gnatname^}
13381 PREP or PREPROCESS to invoke @command{^gnatprep^gnatprep^}
13383 PP or PRETTY to invoke @command{^gnatpp^gnatpp^}
13385 METRIC to invoke @command{^gnatmetric^gnatmetric^}
13387 STUB to invoke @command{^gnatstub^gnatstub^}
13389 XREF to invoke @command{^gnatxref^gnatxref^}
13393 (note that the compiler is invoked using the command
13394 @command{^gnatmake -f -u -c^gnatmake -f -u -c^}).
13397 On non VMS platforms, between @command{gnat} and the command, two
13398 special switches may be used:
13402 @command{-v} to display the invocation of the tool.
13404 @command{-dn} to prevent the @command{gnat} driver from removing
13405 the temporary files it has created. These temporary files are
13406 configuration files and temporary file list files.
13410 The command may be followed by switches and arguments for the invoked
13414 gnat bind -C main.ali
13420 Switches may also be put in text files, one switch per line, and the text
13421 files may be specified with their path name preceded by '@@'.
13424 gnat bind @@args.txt main.ali
13428 In addition, for commands BIND, COMP or COMPILE, FIND, ELIM, LS or LIST, LINK,
13429 METRIC, PP or PRETTY, STUB and XREF, the project file related switches
13430 (@option{^-P^/PROJECT_FILE^},
13431 @option{^-X^/EXTERNAL_REFERENCE^} and
13432 @option{^-vP^/MESSAGES_PROJECT_FILE=^x}) may be used in addition to
13433 the switches of the invoking tool.
13436 When GNAT PP or GNAT PRETTY is used with a project file, but with no source
13437 specified on the command line, it invokes @command{^gnatpp^gnatpp^} with all
13438 the immediate sources of the specified project file.
13441 When GNAT METRIC is used with a project file, but with no source
13442 specified on the command line, it invokes @command{^gnatmetric^gnatmetric^}
13443 with all the immediate sources of the specified project file and with
13444 @option{^-d^/DIRECTORY^} with the parameter pointing to the object directory
13448 In addition, when GNAT PP, GNAT PRETTY or GNAT METRIC is used with
13449 a project file, no source is specified on the command line and
13450 switch ^-U^/ALL_PROJECTS^ is specified on the command line, then
13451 the underlying tool (^gnatpp^gnatpp^ or
13452 ^gnatmetric^gnatmetric^) is invoked for all sources of all projects,
13453 not only for the immediate sources of the main project.
13455 (-U stands for Universal or Union of the project files of the project tree)
13459 For each of the following commands, there is optionally a corresponding
13460 package in the main project.
13464 package @code{Binder} for command BIND (invoking @code{^gnatbind^gnatbind^})
13467 package @code{Compiler} for command COMP or COMPILE (invoking the compiler)
13470 package @code{Finder} for command FIND (invoking @code{^gnatfind^gnatfind^})
13473 package @code{Eliminate} for command ELIM (invoking
13474 @code{^gnatelim^gnatelim^})
13477 package @code{Gnatls} for command LS or LIST (invoking @code{^gnatls^gnatls^})
13480 package @code{Linker} for command LINK (invoking @code{^gnatlink^gnatlink^})
13483 package @code{Metrics} for command METRIC
13484 (invoking @code{^gnatmetric^gnatmetric^})
13487 package @code{Pretty_Printer} for command PP or PRETTY
13488 (invoking @code{^gnatpp^gnatpp^})
13491 package @code{Gnatstub} for command STUB
13492 (invoking @code{^gnatstub^gnatstub^})
13495 package @code{Cross_Reference} for command XREF (invoking
13496 @code{^gnatxref^gnatxref^})
13501 Package @code{Gnatls} has a unique attribute @code{^Switches^Switches^},
13502 a simple variable with a string list value. It contains ^switches^switches^
13503 for the invocation of @code{^gnatls^gnatls^}.
13505 @smallexample @c projectfile
13509 for ^Switches^Switches^
13518 All other packages have two attribute @code{^Switches^Switches^} and
13519 @code{^Default_Switches^Default_Switches^}.
13522 @code{^Switches^Switches^} is an associated array attribute, indexed by the
13523 source file name, that has a string list value: the ^switches^switches^ to be
13524 used when the tool corresponding to the package is invoked for the specific
13528 @code{^Default_Switches^Default_Switches^} is an associative array attribute,
13529 indexed by the programming language that has a string list value.
13530 @code{^Default_Switches^Default_Switches^ ("Ada")} contains the
13531 ^switches^switches^ for the invocation of the tool corresponding
13532 to the package, except if a specific @code{^Switches^Switches^} attribute
13533 is specified for the source file.
13535 @smallexample @c projectfile
13539 for Source_Dirs use ("./**");
13542 for ^Switches^Switches^ use
13549 package Compiler is
13550 for ^Default_Switches^Default_Switches^ ("Ada")
13551 use ("^-gnatv^-gnatv^",
13552 "^-gnatwa^-gnatwa^");
13558 for ^Default_Switches^Default_Switches^ ("Ada")
13566 for ^Default_Switches^Default_Switches^ ("Ada")
13568 for ^Switches^Switches^ ("main.adb")
13577 for ^Default_Switches^Default_Switches^ ("Ada")
13584 package Cross_Reference is
13585 for ^Default_Switches^Default_Switches^ ("Ada")
13590 end Cross_Reference;
13596 With the above project file, commands such as
13599 ^gnat comp -Pproj main^GNAT COMP /PROJECT_FILE=PROJ MAIN^
13600 ^gnat ls -Pproj main^GNAT LIST /PROJECT_FILE=PROJ MAIN^
13601 ^gnat xref -Pproj main^GNAT XREF /PROJECT_FILE=PROJ MAIN^
13602 ^gnat bind -Pproj main.ali^GNAT BIND /PROJECT_FILE=PROJ MAIN.ALI^
13603 ^gnat link -Pproj main.ali^GNAT LINK /PROJECT_FILE=PROJ MAIN.ALI^
13607 will set up the environment properly and invoke the tool with the switches
13608 found in the package corresponding to the tool:
13609 @code{^Default_Switches^Default_Switches^ ("Ada")} for all tools,
13610 except @code{^Switches^Switches^ ("main.adb")}
13611 for @code{^gnatlink^gnatlink^}.
13614 @node Glide and Project Files
13615 @subsection Glide and Project Files
13618 Glide will automatically recognize the @file{.gpr} extension for
13619 project files, and will
13620 convert them to its own internal format automatically. However, it
13621 doesn't provide a syntax-oriented editor for modifying these
13623 The project file will be loaded as text when you select the menu item
13624 @code{Ada} @result{} @code{Project} @result{} @code{Edit}.
13625 You can edit this text and save the @file{gpr} file;
13626 when you next select this project file in Glide it
13627 will be automatically reloaded.
13630 @c **********************
13631 @node An Extended Example
13632 @section An Extended Example
13635 Suppose that we have two programs, @var{prog1} and @var{prog2},
13636 whose sources are in corresponding directories. We would like
13637 to build them with a single @command{gnatmake} command, and we want to place
13638 their object files into @file{build} subdirectories of the source directories.
13639 Furthermore, we want to have to have two separate subdirectories
13640 in @file{build} -- @file{release} and @file{debug} -- which will contain
13641 the object files compiled with different set of compilation flags.
13643 In other words, we have the following structure:
13660 Here are the project files that we must place in a directory @file{main}
13661 to maintain this structure:
13665 @item We create a @code{Common} project with a package @code{Compiler} that
13666 specifies the compilation ^switches^switches^:
13671 @b{project} Common @b{is}
13673 @b{for} Source_Dirs @b{use} (); -- No source files
13677 @b{type} Build_Type @b{is} ("release", "debug");
13678 Build : Build_Type := External ("BUILD", "debug");
13681 @b{package} Compiler @b{is}
13682 @b{case} Build @b{is}
13683 @b{when} "release" =>
13684 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
13685 @b{use} ("^-O2^-O2^");
13686 @b{when} "debug" =>
13687 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
13688 @b{use} ("^-g^-g^");
13696 @item We create separate projects for the two programs:
13703 @b{project} Prog1 @b{is}
13705 @b{for} Source_Dirs @b{use} ("prog1");
13706 @b{for} Object_Dir @b{use} "prog1/build/" & Common.Build;
13708 @b{package} Compiler @b{renames} Common.Compiler;
13719 @b{project} Prog2 @b{is}
13721 @b{for} Source_Dirs @b{use} ("prog2");
13722 @b{for} Object_Dir @b{use} "prog2/build/" & Common.Build;
13724 @b{package} Compiler @b{renames} Common.Compiler;
13730 @item We create a wrapping project @code{Main}:
13739 @b{project} Main @b{is}
13741 @b{package} Compiler @b{renames} Common.Compiler;
13747 @item Finally we need to create a dummy procedure that @code{with}s (either
13748 explicitly or implicitly) all the sources of our two programs.
13753 Now we can build the programs using the command
13756 gnatmake ^-P^/PROJECT_FILE=^main dummy
13760 for the Debug mode, or
13764 gnatmake -Pmain -XBUILD=release
13770 GNAT MAKE /PROJECT_FILE=main /EXTERNAL_REFERENCE=BUILD=release
13775 for the Release mode.
13777 @c ********************************
13778 @c * Project File Complete Syntax *
13779 @c ********************************
13781 @node Project File Complete Syntax
13782 @section Project File Complete Syntax
13786 context_clause project_declaration
13792 @b{with} path_name @{ , path_name @} ;
13797 project_declaration ::=
13798 simple_project_declaration | project_extension
13800 simple_project_declaration ::=
13801 @b{project} <project_>simple_name @b{is}
13802 @{declarative_item@}
13803 @b{end} <project_>simple_name;
13805 project_extension ::=
13806 @b{project} <project_>simple_name @b{extends} path_name @b{is}
13807 @{declarative_item@}
13808 @b{end} <project_>simple_name;
13810 declarative_item ::=
13811 package_declaration |
13812 typed_string_declaration |
13813 other_declarative_item
13815 package_declaration ::=
13816 package_specification | package_renaming
13818 package_specification ::=
13819 @b{package} package_identifier @b{is}
13820 @{simple_declarative_item@}
13821 @b{end} package_identifier ;
13823 package_identifier ::=
13824 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
13825 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
13826 @code{^gnatls^gnatls^} | @code{IDE} | @code{Pretty_Printer}
13828 package_renaming ::==
13829 @b{package} package_identifier @b{renames}
13830 <project_>simple_name.package_identifier ;
13832 typed_string_declaration ::=
13833 @b{type} <typed_string_>_simple_name @b{is}
13834 ( string_literal @{, string_literal@} );
13836 other_declarative_item ::=
13837 attribute_declaration |
13838 typed_variable_declaration |
13839 variable_declaration |
13842 attribute_declaration ::=
13843 full_associative_array_declaration |
13844 @b{for} attribute_designator @b{use} expression ;
13846 full_associative_array_declaration ::=
13847 @b{for} <associative_array_attribute_>simple_name @b{use}
13848 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
13850 attribute_designator ::=
13851 <simple_attribute_>simple_name |
13852 <associative_array_attribute_>simple_name ( string_literal )
13854 typed_variable_declaration ::=
13855 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
13857 variable_declaration ::=
13858 <variable_>simple_name := expression;
13868 attribute_reference
13874 ( <string_>expression @{ , <string_>expression @} )
13877 @b{external} ( string_literal [, string_literal] )
13879 attribute_reference ::=
13880 attribute_prefix ' <simple_attribute_>simple_name [ ( literal_string ) ]
13882 attribute_prefix ::=
13884 <project_>simple_name | package_identifier |
13885 <project_>simple_name . package_identifier
13887 case_construction ::=
13888 @b{case} <typed_variable_>name @b{is}
13893 @b{when} discrete_choice_list =>
13894 @{case_construction | attribute_declaration@}
13896 discrete_choice_list ::=
13897 string_literal @{| string_literal@} |
13901 simple_name @{. simple_name@}
13904 identifier (same as Ada)
13908 @node The Cross-Referencing Tools gnatxref and gnatfind
13909 @chapter The Cross-Referencing Tools @code{gnatxref} and @code{gnatfind}
13914 The compiler generates cross-referencing information (unless
13915 you set the @samp{-gnatx} switch), which are saved in the @file{.ali} files.
13916 This information indicates where in the source each entity is declared and
13917 referenced. Note that entities in package Standard are not included, but
13918 entities in all other predefined units are included in the output.
13920 Before using any of these two tools, you need to compile successfully your
13921 application, so that GNAT gets a chance to generate the cross-referencing
13924 The two tools @code{gnatxref} and @code{gnatfind} take advantage of this
13925 information to provide the user with the capability to easily locate the
13926 declaration and references to an entity. These tools are quite similar,
13927 the difference being that @code{gnatfind} is intended for locating
13928 definitions and/or references to a specified entity or entities, whereas
13929 @code{gnatxref} is oriented to generating a full report of all
13932 To use these tools, you must not compile your application using the
13933 @option{-gnatx} switch on the @command{gnatmake} command line
13934 (@pxref{The GNAT Make Program gnatmake}). Otherwise, cross-referencing
13935 information will not be generated.
13938 * gnatxref Switches::
13939 * gnatfind Switches::
13940 * Project Files for gnatxref and gnatfind::
13941 * Regular Expressions in gnatfind and gnatxref::
13942 * Examples of gnatxref Usage::
13943 * Examples of gnatfind Usage::
13946 @node gnatxref Switches
13947 @section @code{gnatxref} Switches
13950 The command invocation for @code{gnatxref} is:
13952 $ gnatxref [switches] sourcefile1 [sourcefile2 ...]
13959 @item sourcefile1, sourcefile2
13960 identifies the source files for which a report is to be generated. The
13961 ``with''ed units will be processed too. You must provide at least one file.
13963 These file names are considered to be regular expressions, so for instance
13964 specifying @file{source*.adb} is the same as giving every file in the current
13965 directory whose name starts with @file{source} and whose extension is
13968 You shouldn't specify any directory name, just base names. @command{gnatxref}
13969 and @command{gnatfind} will be able to locate these files by themselves using
13970 the source path. If you specify directories, no result is produced.
13975 The switches can be :
13978 @item ^-a^/ALL_FILES^
13979 @cindex @option{^-a^/ALL_FILES^} (@command{gnatxref})
13980 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
13981 the read-only files found in the library search path. Otherwise, these files
13982 will be ignored. This option can be used to protect Gnat sources or your own
13983 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
13984 much faster, and their output much smaller. Read-only here refers to access
13985 or permissions status in the file system for the current user.
13988 @cindex @option{-aIDIR} (@command{gnatxref})
13989 When looking for source files also look in directory DIR. The order in which
13990 source file search is undertaken is the same as for @command{gnatmake}.
13993 @cindex @option{-aODIR} (@command{gnatxref})
13994 When searching for library and object files, look in directory
13995 DIR. The order in which library files are searched is the same as for
13996 @command{gnatmake}.
13999 @cindex @option{-nostdinc} (@command{gnatxref})
14000 Do not look for sources in the system default directory.
14003 @cindex @option{-nostdlib} (@command{gnatxref})
14004 Do not look for library files in the system default directory.
14006 @item --RTS=@var{rts-path}
14007 @cindex @option{--RTS} (@command{gnatxref})
14008 Specifies the default location of the runtime library. Same meaning as the
14009 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
14011 @item ^-d^/DERIVED_TYPES^
14012 @cindex @option{^-d^/DERIVED_TYPES^} (@command{gnatxref})
14013 If this switch is set @code{gnatxref} will output the parent type
14014 reference for each matching derived types.
14016 @item ^-f^/FULL_PATHNAME^
14017 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatxref})
14018 If this switch is set, the output file names will be preceded by their
14019 directory (if the file was found in the search path). If this switch is
14020 not set, the directory will not be printed.
14022 @item ^-g^/IGNORE_LOCALS^
14023 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatxref})
14024 If this switch is set, information is output only for library-level
14025 entities, ignoring local entities. The use of this switch may accelerate
14026 @code{gnatfind} and @code{gnatxref}.
14029 @cindex @option{-IDIR} (@command{gnatxref})
14030 Equivalent to @samp{-aODIR -aIDIR}.
14033 @cindex @option{-pFILE} (@command{gnatxref})
14034 Specify a project file to use @xref{Project Files}. These project files are
14035 the @file{.adp} files used by Glide. If you need to use the @file{.gpr}
14036 project files, you should use gnatxref through the GNAT driver
14037 (@command{gnat xref -Pproject}).
14039 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
14040 project file in the current directory.
14042 If a project file is either specified or found by the tools, then the content
14043 of the source directory and object directory lines are added as if they
14044 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^}
14045 and @samp{^-aO^OBJECT_SEARCH^}.
14047 Output only unused symbols. This may be really useful if you give your
14048 main compilation unit on the command line, as @code{gnatxref} will then
14049 display every unused entity and 'with'ed package.
14053 Instead of producing the default output, @code{gnatxref} will generate a
14054 @file{tags} file that can be used by vi. For examples how to use this
14055 feature, see @ref{Examples of gnatxref Usage}. The tags file is output
14056 to the standard output, thus you will have to redirect it to a file.
14062 All these switches may be in any order on the command line, and may even
14063 appear after the file names. They need not be separated by spaces, thus
14064 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
14065 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
14067 @node gnatfind Switches
14068 @section @code{gnatfind} Switches
14071 The command line for @code{gnatfind} is:
14074 $ gnatfind [switches] pattern[:sourcefile[:line[:column]]]
14083 An entity will be output only if it matches the regular expression found
14084 in @samp{pattern}, see @ref{Regular Expressions in gnatfind and gnatxref}.
14086 Omitting the pattern is equivalent to specifying @samp{*}, which
14087 will match any entity. Note that if you do not provide a pattern, you
14088 have to provide both a sourcefile and a line.
14090 Entity names are given in Latin-1, with uppercase/lowercase equivalence
14091 for matching purposes. At the current time there is no support for
14092 8-bit codes other than Latin-1, or for wide characters in identifiers.
14095 @code{gnatfind} will look for references, bodies or declarations
14096 of symbols referenced in @file{sourcefile}, at line @samp{line}
14097 and column @samp{column}. See @ref{Examples of gnatfind Usage}
14098 for syntax examples.
14101 is a decimal integer identifying the line number containing
14102 the reference to the entity (or entities) to be located.
14105 is a decimal integer identifying the exact location on the
14106 line of the first character of the identifier for the
14107 entity reference. Columns are numbered from 1.
14109 @item file1 file2 ...
14110 The search will be restricted to these source files. If none are given, then
14111 the search will be done for every library file in the search path.
14112 These file must appear only after the pattern or sourcefile.
14114 These file names are considered to be regular expressions, so for instance
14115 specifying 'source*.adb' is the same as giving every file in the current
14116 directory whose name starts with 'source' and whose extension is 'adb'.
14118 The location of the spec of the entity will always be displayed, even if it
14119 isn't in one of file1, file2,... The occurrences of the entity in the
14120 separate units of the ones given on the command line will also be displayed.
14122 Note that if you specify at least one file in this part, @code{gnatfind} may
14123 sometimes not be able to find the body of the subprograms...
14128 At least one of 'sourcefile' or 'pattern' has to be present on
14131 The following switches are available:
14135 @item ^-a^/ALL_FILES^
14136 @cindex @option{^-a^/ALL_FILES^} (@command{gnatfind})
14137 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
14138 the read-only files found in the library search path. Otherwise, these files
14139 will be ignored. This option can be used to protect Gnat sources or your own
14140 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
14141 much faster, and their output much smaller. Read-only here refers to access
14142 or permission status in the file system for the current user.
14145 @cindex @option{-aIDIR} (@command{gnatfind})
14146 When looking for source files also look in directory DIR. The order in which
14147 source file search is undertaken is the same as for @command{gnatmake}.
14150 @cindex @option{-aODIR} (@command{gnatfind})
14151 When searching for library and object files, look in directory
14152 DIR. The order in which library files are searched is the same as for
14153 @command{gnatmake}.
14156 @cindex @option{-nostdinc} (@command{gnatfind})
14157 Do not look for sources in the system default directory.
14160 @cindex @option{-nostdlib} (@command{gnatfind})
14161 Do not look for library files in the system default directory.
14163 @item --RTS=@var{rts-path}
14164 @cindex @option{--RTS} (@command{gnatfind})
14165 Specifies the default location of the runtime library. Same meaning as the
14166 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
14168 @item ^-d^/DERIVED_TYPE_INFORMATION^
14169 @cindex @option{^-d^/DERIVED_TYPE_INFORMATION^} (@code{gnatfind})
14170 If this switch is set, then @code{gnatfind} will output the parent type
14171 reference for each matching derived types.
14173 @item ^-e^/EXPRESSIONS^
14174 @cindex @option{^-e^/EXPRESSIONS^} (@command{gnatfind})
14175 By default, @code{gnatfind} accept the simple regular expression set for
14176 @samp{pattern}. If this switch is set, then the pattern will be
14177 considered as full Unix-style regular expression.
14179 @item ^-f^/FULL_PATHNAME^
14180 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatfind})
14181 If this switch is set, the output file names will be preceded by their
14182 directory (if the file was found in the search path). If this switch is
14183 not set, the directory will not be printed.
14185 @item ^-g^/IGNORE_LOCALS^
14186 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatfind})
14187 If this switch is set, information is output only for library-level
14188 entities, ignoring local entities. The use of this switch may accelerate
14189 @code{gnatfind} and @code{gnatxref}.
14192 @cindex @option{-IDIR} (@command{gnatfind})
14193 Equivalent to @samp{-aODIR -aIDIR}.
14196 @cindex @option{-pFILE} (@command{gnatfind})
14197 Specify a project file (@pxref{Project Files}) to use.
14198 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
14199 project file in the current directory.
14201 If a project file is either specified or found by the tools, then the content
14202 of the source directory and object directory lines are added as if they
14203 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^} and
14204 @samp{^-aO^/OBJECT_SEARCH^}.
14206 @item ^-r^/REFERENCES^
14207 @cindex @option{^-r^/REFERENCES^} (@command{gnatfind})
14208 By default, @code{gnatfind} will output only the information about the
14209 declaration, body or type completion of the entities. If this switch is
14210 set, the @code{gnatfind} will locate every reference to the entities in
14211 the files specified on the command line (or in every file in the search
14212 path if no file is given on the command line).
14214 @item ^-s^/PRINT_LINES^
14215 @cindex @option{^-s^/PRINT_LINES^} (@command{gnatfind})
14216 If this switch is set, then @code{gnatfind} will output the content
14217 of the Ada source file lines were the entity was found.
14219 @item ^-t^/TYPE_HIERARCHY^
14220 @cindex @option{^-t^/TYPE_HIERARCHY^} (@command{gnatfind})
14221 If this switch is set, then @code{gnatfind} will output the type hierarchy for
14222 the specified type. It act like -d option but recursively from parent
14223 type to parent type. When this switch is set it is not possible to
14224 specify more than one file.
14229 All these switches may be in any order on the command line, and may even
14230 appear after the file names. They need not be separated by spaces, thus
14231 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
14232 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
14234 As stated previously, gnatfind will search in every directory in the
14235 search path. You can force it to look only in the current directory if
14236 you specify @code{*} at the end of the command line.
14238 @node Project Files for gnatxref and gnatfind
14239 @section Project Files for @command{gnatxref} and @command{gnatfind}
14242 Project files allow a programmer to specify how to compile its
14243 application, where to find sources, etc. These files are used
14245 primarily by the Glide Ada mode, but they can also be used
14248 @code{gnatxref} and @code{gnatfind}.
14250 A project file name must end with @file{.gpr}. If a single one is
14251 present in the current directory, then @code{gnatxref} and @code{gnatfind} will
14252 extract the information from it. If multiple project files are found, none of
14253 them is read, and you have to use the @samp{-p} switch to specify the one
14256 The following lines can be included, even though most of them have default
14257 values which can be used in most cases.
14258 The lines can be entered in any order in the file.
14259 Except for @file{src_dir} and @file{obj_dir}, you can only have one instance of
14260 each line. If you have multiple instances, only the last one is taken into
14265 [default: @code{"^./^[]^"}]
14266 specifies a directory where to look for source files. Multiple @code{src_dir}
14267 lines can be specified and they will be searched in the order they
14271 [default: @code{"^./^[]^"}]
14272 specifies a directory where to look for object and library files. Multiple
14273 @code{obj_dir} lines can be specified, and they will be searched in the order
14276 @item comp_opt=SWITCHES
14277 [default: @code{""}]
14278 creates a variable which can be referred to subsequently by using
14279 the @code{$@{comp_opt@}} notation. This is intended to store the default
14280 switches given to @command{gnatmake} and @command{gcc}.
14282 @item bind_opt=SWITCHES
14283 [default: @code{""}]
14284 creates a variable which can be referred to subsequently by using
14285 the @samp{$@{bind_opt@}} notation. This is intended to store the default
14286 switches given to @command{gnatbind}.
14288 @item link_opt=SWITCHES
14289 [default: @code{""}]
14290 creates a variable which can be referred to subsequently by using
14291 the @samp{$@{link_opt@}} notation. This is intended to store the default
14292 switches given to @command{gnatlink}.
14294 @item main=EXECUTABLE
14295 [default: @code{""}]
14296 specifies the name of the executable for the application. This variable can
14297 be referred to in the following lines by using the @samp{$@{main@}} notation.
14300 @item comp_cmd=COMMAND
14301 [default: @code{"GNAT COMPILE /SEARCH=$@{src_dir@} /DEBUG /TRY_SEMANTICS"}]
14304 @item comp_cmd=COMMAND
14305 [default: @code{"gcc -c -I$@{src_dir@} -g -gnatq"}]
14307 specifies the command used to compile a single file in the application.
14310 @item make_cmd=COMMAND
14311 [default: @code{"GNAT MAKE $@{main@}
14312 /SOURCE_SEARCH=$@{src_dir@} /OBJECT_SEARCH=$@{obj_dir@}
14313 /DEBUG /TRY_SEMANTICS /COMPILER_QUALIFIERS $@{comp_opt@}
14314 /BINDER_QUALIFIERS $@{bind_opt@} /LINKER_QUALIFIERS $@{link_opt@}"}]
14317 @item make_cmd=COMMAND
14318 [default: @code{"gnatmake $@{main@} -aI$@{src_dir@}
14319 -aO$@{obj_dir@} -g -gnatq -cargs $@{comp_opt@}
14320 -bargs $@{bind_opt@} -largs $@{link_opt@}"}]
14322 specifies the command used to recompile the whole application.
14324 @item run_cmd=COMMAND
14325 [default: @code{"$@{main@}"}]
14326 specifies the command used to run the application.
14328 @item debug_cmd=COMMAND
14329 [default: @code{"gdb $@{main@}"}]
14330 specifies the command used to debug the application
14335 @command{gnatxref} and @command{gnatfind} only take into account the
14336 @code{src_dir} and @code{obj_dir} lines, and ignore the others.
14338 @node Regular Expressions in gnatfind and gnatxref
14339 @section Regular Expressions in @code{gnatfind} and @code{gnatxref}
14342 As specified in the section about @command{gnatfind}, the pattern can be a
14343 regular expression. Actually, there are to set of regular expressions
14344 which are recognized by the program :
14347 @item globbing patterns
14348 These are the most usual regular expression. They are the same that you
14349 generally used in a Unix shell command line, or in a DOS session.
14351 Here is a more formal grammar :
14358 term ::= elmt -- matches elmt
14359 term ::= elmt elmt -- concatenation (elmt then elmt)
14360 term ::= * -- any string of 0 or more characters
14361 term ::= ? -- matches any character
14362 term ::= [char @{char@}] -- matches any character listed
14363 term ::= [char - char] -- matches any character in range
14367 @item full regular expression
14368 The second set of regular expressions is much more powerful. This is the
14369 type of regular expressions recognized by utilities such a @file{grep}.
14371 The following is the form of a regular expression, expressed in Ada
14372 reference manual style BNF is as follows
14379 regexp ::= term @{| term@} -- alternation (term or term ...)
14381 term ::= item @{item@} -- concatenation (item then item)
14383 item ::= elmt -- match elmt
14384 item ::= elmt * -- zero or more elmt's
14385 item ::= elmt + -- one or more elmt's
14386 item ::= elmt ? -- matches elmt or nothing
14389 elmt ::= nschar -- matches given character
14390 elmt ::= [nschar @{nschar@}] -- matches any character listed
14391 elmt ::= [^^^ nschar @{nschar@}] -- matches any character not listed
14392 elmt ::= [char - char] -- matches chars in given range
14393 elmt ::= \ char -- matches given character
14394 elmt ::= . -- matches any single character
14395 elmt ::= ( regexp ) -- parens used for grouping
14397 char ::= any character, including special characters
14398 nschar ::= any character except ()[].*+?^^^
14402 Following are a few examples :
14406 will match any of the two strings 'abcde' and 'fghi'.
14409 will match any string like 'abd', 'abcd', 'abccd', 'abcccd', and so on
14412 will match any string which has only lowercase characters in it (and at
14413 least one character
14418 @node Examples of gnatxref Usage
14419 @section Examples of @code{gnatxref} Usage
14421 @subsection General Usage
14424 For the following examples, we will consider the following units :
14426 @smallexample @c ada
14432 3: procedure Foo (B : in Integer);
14439 1: package body Main is
14440 2: procedure Foo (B : in Integer) is
14451 2: procedure Print (B : Integer);
14460 The first thing to do is to recompile your application (for instance, in
14461 that case just by doing a @samp{gnatmake main}, so that GNAT generates
14462 the cross-referencing information.
14463 You can then issue any of the following commands:
14465 @item gnatxref main.adb
14466 @code{gnatxref} generates cross-reference information for main.adb
14467 and every unit 'with'ed by main.adb.
14469 The output would be:
14477 Decl: main.ads 3:20
14478 Body: main.adb 2:20
14479 Ref: main.adb 4:13 5:13 6:19
14482 Ref: main.adb 6:8 7:8
14492 Decl: main.ads 3:15
14493 Body: main.adb 2:15
14496 Body: main.adb 1:14
14499 Ref: main.adb 6:12 7:12
14503 that is the entity @code{Main} is declared in main.ads, line 2, column 9,
14504 its body is in main.adb, line 1, column 14 and is not referenced any where.
14506 The entity @code{Print} is declared in bar.ads, line 2, column 15 and it
14507 it referenced in main.adb, line 6 column 12 and line 7 column 12.
14509 @item gnatxref package1.adb package2.ads
14510 @code{gnatxref} will generates cross-reference information for
14511 package1.adb, package2.ads and any other package 'with'ed by any
14517 @subsection Using gnatxref with vi
14519 @code{gnatxref} can generate a tags file output, which can be used
14520 directly from @file{vi}. Note that the standard version of @file{vi}
14521 will not work properly with overloaded symbols. Consider using another
14522 free implementation of @file{vi}, such as @file{vim}.
14525 $ gnatxref -v gnatfind.adb > tags
14529 will generate the tags file for @code{gnatfind} itself (if the sources
14530 are in the search path!).
14532 From @file{vi}, you can then use the command @samp{:tag @i{entity}}
14533 (replacing @i{entity} by whatever you are looking for), and vi will
14534 display a new file with the corresponding declaration of entity.
14537 @node Examples of gnatfind Usage
14538 @section Examples of @code{gnatfind} Usage
14542 @item gnatfind ^-f^/FULL_PATHNAME^ xyz:main.adb
14543 Find declarations for all entities xyz referenced at least once in
14544 main.adb. The references are search in every library file in the search
14547 The directories will be printed as well (as the @samp{^-f^/FULL_PATHNAME^}
14550 The output will look like:
14552 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
14553 ^directory/^[directory]^main.adb:24:10: xyz <= body
14554 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
14558 that is to say, one of the entities xyz found in main.adb is declared at
14559 line 12 of main.ads (and its body is in main.adb), and another one is
14560 declared at line 45 of foo.ads
14562 @item gnatfind ^-fs^/FULL_PATHNAME/SOURCE_LINE^ xyz:main.adb
14563 This is the same command as the previous one, instead @code{gnatfind} will
14564 display the content of the Ada source file lines.
14566 The output will look like:
14569 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
14571 ^directory/^[directory]^main.adb:24:10: xyz <= body
14573 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
14578 This can make it easier to find exactly the location your are looking
14581 @item gnatfind ^-r^/REFERENCES^ "*x*":main.ads:123 foo.adb
14582 Find references to all entities containing an x that are
14583 referenced on line 123 of main.ads.
14584 The references will be searched only in main.ads and foo.adb.
14586 @item gnatfind main.ads:123
14587 Find declarations and bodies for all entities that are referenced on
14588 line 123 of main.ads.
14590 This is the same as @code{gnatfind "*":main.adb:123}.
14592 @item gnatfind ^mydir/^[mydir]^main.adb:123:45
14593 Find the declaration for the entity referenced at column 45 in
14594 line 123 of file main.adb in directory mydir. Note that it
14595 is usual to omit the identifier name when the column is given,
14596 since the column position identifies a unique reference.
14598 The column has to be the beginning of the identifier, and should not
14599 point to any character in the middle of the identifier.
14603 @c *********************************
14604 @node The GNAT Pretty-Printer gnatpp
14605 @chapter The GNAT Pretty-Printer @command{gnatpp}
14607 @cindex Pretty-Printer
14610 ^The @command{gnatpp} tool^GNAT PRETTY^ is an ASIS-based utility
14611 for source reformatting / pretty-printing.
14612 It takes an Ada source file as input and generates a reformatted
14614 You can specify various style directives via switches; e.g.,
14615 identifier case conventions, rules of indentation, and comment layout.
14617 To produce a reformatted file, @command{gnatpp} generates and uses the ASIS
14618 tree for the input source and thus requires the input to be syntactically and
14619 semantically legal.
14620 If this condition is not met, @command{gnatpp} will terminate with an
14621 error message; no output file will be generated.
14623 If the compilation unit
14624 contained in the input source depends semantically upon units located
14625 outside the current directory, you have to provide the source search path
14626 when invoking @command{gnatpp}, if these units are contained in files with
14627 names that do not follow the GNAT file naming rules, you have to provide
14628 the configuration file describing the corresponding naming scheme;
14629 see the description of the @command{gnatpp}
14630 switches below. Another possibility is to use a project file and to
14631 call @command{gnatpp} through the @command{gnat} driver
14633 The @command{gnatpp} command has the form
14636 $ gnatpp [@var{switches}] @var{filename}
14643 @var{switches} is an optional sequence of switches defining such properties as
14644 the formatting rules, the source search path, and the destination for the
14648 @var{filename} is the name (including the extension) of the source file to
14649 reformat; ``wildcards'' or several file names on the same gnatpp command are
14650 allowed. The file name may contain path information; it does not have to
14651 follow the GNAT file naming rules
14655 * Switches for gnatpp::
14656 * Formatting Rules::
14659 @node Switches for gnatpp
14660 @section Switches for @command{gnatpp}
14663 The following subsections describe the various switches accepted by
14664 @command{gnatpp}, organized by category.
14667 You specify a switch by supplying a name and generally also a value.
14668 In many cases the values for a switch with a given name are incompatible with
14670 (for example the switch that controls the casing of a reserved word may have
14671 exactly one value: upper case, lower case, or
14672 mixed case) and thus exactly one such switch can be in effect for an
14673 invocation of @command{gnatpp}.
14674 If more than one is supplied, the last one is used.
14675 However, some values for the same switch are mutually compatible.
14676 You may supply several such switches to @command{gnatpp}, but then
14677 each must be specified in full, with both the name and the value.
14678 Abbreviated forms (the name appearing once, followed by each value) are
14680 For example, to set
14681 the alignment of the assignment delimiter both in declarations and in
14682 assignment statements, you must write @option{-A2A3}
14683 (or @option{-A2 -A3}), but not @option{-A23}.
14687 In many cases the set of options for a given qualifier are incompatible with
14688 each other (for example the qualifier that controls the casing of a reserved
14689 word may have exactly one option, which specifies either upper case, lower
14690 case, or mixed case), and thus exactly one such option can be in effect for
14691 an invocation of @command{gnatpp}.
14692 If more than one is supplied, the last one is used.
14693 However, some qualifiers have options that are mutually compatible,
14694 and then you may then supply several such options when invoking
14698 In most cases, it is obvious whether or not the
14699 ^values for a switch with a given name^options for a given qualifier^
14700 are compatible with each other.
14701 When the semantics might not be evident, the summaries below explicitly
14702 indicate the effect.
14705 * Alignment Control::
14707 * Construct Layout Control::
14708 * General Text Layout Control::
14709 * Other Formatting Options::
14710 * Setting the Source Search Path::
14711 * Output File Control::
14712 * Other gnatpp Switches::
14715 @node Alignment Control
14716 @subsection Alignment Control
14717 @cindex Alignment control in @command{gnatpp}
14720 Programs can be easier to read if certain constructs are vertically aligned.
14721 By default all alignments are set ON.
14722 Through the @option{^-A0^/ALIGN=OFF^} switch you may reset the default to
14723 OFF, and then use one or more of the other
14724 ^@option{-A@var{n}} switches^@option{/ALIGN} options^
14725 to activate alignment for specific constructs.
14728 @cindex @option{^-A@var{n}^/ALIGN^} (@command{gnatpp})
14732 Set all alignments to ON
14735 @item ^-A0^/ALIGN=OFF^
14736 Set all alignments to OFF
14738 @item ^-A1^/ALIGN=COLONS^
14739 Align @code{:} in declarations
14741 @item ^-A2^/ALIGN=DECLARATIONS^
14742 Align @code{:=} in initializations in declarations
14744 @item ^-A3^/ALIGN=STATEMENTS^
14745 Align @code{:=} in assignment statements
14747 @item ^-A4^/ALIGN=ARROWS^
14748 Align @code{=>} in associations
14750 @item ^-A5^/ALIGN=COMPONENT_CLAUSES^
14751 Align @code{at} keywords in the component clauses in record
14752 representation clauses
14756 The @option{^-A^/ALIGN^} switches are mutually compatible; any combination
14759 @node Casing Control
14760 @subsection Casing Control
14761 @cindex Casing control in @command{gnatpp}
14764 @command{gnatpp} allows you to specify the casing for reserved words,
14765 pragma names, attribute designators and identifiers.
14766 For identifiers you may define a
14767 general rule for name casing but also override this rule
14768 via a set of dictionary files.
14770 Three types of casing are supported: lower case, upper case, and mixed case.
14771 Lower and upper case are self-explanatory (but since some letters in
14772 Latin1 and other GNAT-supported character sets
14773 exist only in lower-case form, an upper case conversion will have no
14775 ``Mixed case'' means that the first letter, and also each letter immediately
14776 following an underscore, are converted to their uppercase forms;
14777 all the other letters are converted to their lowercase forms.
14780 @cindex @option{^-a@var{x}^/ATTRIBUTE^} (@command{gnatpp})
14781 @item ^-aL^/ATTRIBUTE_CASING=LOWER_CASE^
14782 Attribute designators are lower case
14784 @item ^-aU^/ATTRIBUTE_CASING=UPPER_CASE^
14785 Attribute designators are upper case
14787 @item ^-aM^/ATTRIBUTE_CASING=MIXED_CASE^
14788 Attribute designators are mixed case (this is the default)
14790 @cindex @option{^-k@var{x}^/KEYWORD_CASING^} (@command{gnatpp})
14791 @item ^-kL^/KEYWORD_CASING=LOWER_CASE^
14792 Keywords (technically, these are known in Ada as @emph{reserved words}) are
14793 lower case (this is the default)
14795 @item ^-kU^/KEYWORD_CASING=UPPER_CASE^
14796 Keywords are upper case
14798 @cindex @option{^-n@var{x}^/NAME_CASING^} (@command{gnatpp})
14799 @item ^-nD^/NAME_CASING=AS_DECLARED^
14800 Name casing for defining occurrences are as they appear in the source file
14801 (this is the default)
14803 @item ^-nU^/NAME_CASING=UPPER_CASE^
14804 Names are in upper case
14806 @item ^-nL^/NAME_CASING=LOWER_CASE^
14807 Names are in lower case
14809 @item ^-nM^/NAME_CASING=MIXED_CASE^
14810 Names are in mixed case
14812 @cindex @option{^-p@var{x}^/PRAGMA_CASING^} (@command{gnatpp})
14813 @item ^-pL^/PRAGMA_CASING=LOWER_CASE^
14814 Pragma names are lower case
14816 @item ^-pU^/PRAGMA_CASING=UPPER_CASE^
14817 Pragma names are upper case
14819 @item ^-pM^/PRAGMA_CASING=MIXED_CASE^
14820 Pragma names are mixed case (this is the default)
14822 @item ^-D@var{file}^/DICTIONARY=@var{file}^
14823 @cindex @option{^-D^/DICTIONARY^} (@command{gnatpp})
14824 Use @var{file} as a @emph{dictionary file} that defines
14825 the casing for a set of specified names,
14826 thereby overriding the effect on these names by
14827 any explicit or implicit
14828 ^-n^/NAME_CASING^ switch.
14829 To supply more than one dictionary file,
14830 use ^several @option{-D} switches^a list of files as options^.
14833 @option{gnatpp} implicitly uses a @emph{default dictionary file}
14834 to define the casing for the Ada predefined names and
14835 the names declared in the GNAT libraries.
14837 @item ^-D-^/SPECIFIC_CASING^
14838 @cindex @option{^-D-^/SPECIFIC_CASING^} (@command{gnatpp})
14839 Do not use the default dictionary file;
14840 instead, use the casing
14841 defined by a @option{^-n^/NAME_CASING^} switch and any explicit
14846 The structure of a dictionary file, and details on the conventions
14847 used in the default dictionary file, are defined in @ref{Name Casing}.
14849 The @option{^-D-^/SPECIFIC_CASING^} and
14850 @option{^-D@var{file}^/DICTIONARY=@var{file}^} switches are mutually
14853 @node Construct Layout Control
14854 @subsection Construct Layout Control
14855 @cindex Layout control in @command{gnatpp}
14858 This group of @command{gnatpp} switches controls the layout of comments and
14859 complex syntactic constructs. See @ref{Formatting Comments} for details
14863 @cindex @option{^-c@var{n}^/COMMENTS_LAYOUT^} (@command{gnatpp})
14864 @item ^-c0^/COMMENTS_LAYOUT=UNTOUCHED^
14865 All the comments remain unchanged
14867 @item ^-c1^/COMMENTS_LAYOUT=DEFAULT^
14868 GNAT-style comment line indentation (this is the default).
14870 @item ^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^
14871 Reference-manual comment line indentation.
14873 @item ^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^
14874 GNAT-style comment beginning
14876 @item ^-c4^/COMMENTS_LAYOUT=REFORMAT^
14877 Reformat comment blocks
14879 @cindex @option{^-l@var{n}^/CONSTRUCT_LAYOUT^} (@command{gnatpp})
14880 @item ^-l1^/CONSTRUCT_LAYOUT=GNAT^
14881 GNAT-style layout (this is the default)
14883 @item ^-l2^/CONSTRUCT_LAYOUT=COMPACT^
14886 @item ^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^
14889 @cindex @option{^-N^/NOTABS^} (@command{gnatpp})
14891 All the VT characters are removed from the comment text. All the HT characters
14892 are expanded with the sequences of space characters to get to the next tab
14899 The @option{-c1} and @option{-c2} switches are incompatible.
14900 The @option{-c3} and @option{-c4} switches are compatible with each other and
14901 also with @option{-c1} and @option{-c2}. The @option{-c0} switch disables all
14902 the other comment formatting switches.
14904 The @option{-l1}, @option{-l2}, and @option{-l3} switches are incompatible.
14909 For the @option{/COMMENTS_LAYOUT} qualifier:
14912 The @option{DEFAULT} and @option{STANDARD_INDENT} options are incompatible.
14914 The @option{GNAT_BEGINNING} and @option{REFORMAT} options are compatible with
14915 each other and also with @option{DEFAULT} and @option{STANDARD_INDENT}.
14919 The @option{GNAT}, @option{COMPACT}, and @option{UNCOMPACT} options for the
14920 @option{/CONSTRUCT_LAYOUT} qualifier are incompatible.
14923 @node General Text Layout Control
14924 @subsection General Text Layout Control
14927 These switches allow control over line length and indentation.
14930 @item ^-M@i{nnn}^/LINE_LENGTH_MAX=@i{nnn}^
14931 @cindex @option{^-M^/LINE_LENGTH^} (@command{gnatpp})
14932 Maximum line length, @i{nnn} from 32 ..256, the default value is 79
14934 @item ^-i@i{nnn}^/INDENTATION_LEVEL=@i{nnn}^
14935 @cindex @option{^-i^/INDENTATION_LEVEL^} (@command{gnatpp})
14936 Indentation level, @i{nnn} from 1 .. 9, the default value is 3
14938 @item ^-cl@i{nnn}^/CONTINUATION_INDENT=@i{nnn}^
14939 @cindex @option{^-cl^/CONTINUATION_INDENT^} (@command{gnatpp})
14940 Indentation level for continuation lines (relative to the line being
14941 continued), @i{nnn} from 1 .. 9.
14943 value is one less then the (normal) indentation level, unless the
14944 indentation is set to 1 (in which case the default value for continuation
14945 line indentation is also 1)
14948 @node Other Formatting Options
14949 @subsection Other Formatting Options
14952 These switches control the inclusion of missing end/exit labels, and
14953 the indentation level in @b{case} statements.
14956 @item ^-e^/NO_MISSED_LABELS^
14957 @cindex @option{^-e^/NO_MISSED_LABELS^} (@command{gnatpp})
14958 Do not insert missing end/exit labels. An end label is the name of
14959 a construct that may optionally be repeated at the end of the
14960 construct's declaration;
14961 e.g., the names of packages, subprograms, and tasks.
14962 An exit label is the name of a loop that may appear as target
14963 of an exit statement within the loop.
14964 By default, @command{gnatpp} inserts these end/exit labels when
14965 they are absent from the original source. This option suppresses such
14966 insertion, so that the formatted source reflects the original.
14968 @item ^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^
14969 @cindex @option{^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^} (@command{gnatpp})
14970 Insert a Form Feed character after a pragma Page.
14972 @item ^-T@i{nnn}^/MAX_INDENT=@i{nnn}^
14973 @cindex @option{^-T^/MAX_INDENT^} (@command{gnatpp})
14974 Do not use an additional indentation level for @b{case} alternatives
14975 and variants if there are @i{nnn} or more (the default
14977 If @i{nnn} is 0, an additional indentation level is
14978 used for @b{case} alternatives and variants regardless of their number.
14981 @node Setting the Source Search Path
14982 @subsection Setting the Source Search Path
14985 To define the search path for the input source file, @command{gnatpp}
14986 uses the same switches as the GNAT compiler, with the same effects.
14989 @item ^-I^/SEARCH=^@var{dir}
14990 @cindex @option{^-I^/SEARCH^} (@code{gnatpp})
14991 The same as the corresponding gcc switch
14993 @item ^-I-^/NOCURRENT_DIRECTORY^
14994 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatpp})
14995 The same as the corresponding gcc switch
14997 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE^=@var{path}
14998 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@code{gnatpp})
14999 The same as the corresponding gcc switch
15001 @item ^--RTS^/RUNTIME_SYSTEM^=@var{path}
15002 @cindex @option{^--RTS^/RUNTIME_SYSTEM^} (@code{gnatpp})
15003 The same as the corresponding gcc switch
15007 @node Output File Control
15008 @subsection Output File Control
15011 By default the output is sent to the file whose name is obtained by appending
15012 the ^@file{.pp}^@file{$PP}^ suffix to the name of the input file
15013 (if the file with this name already exists, it is unconditionally overwritten).
15014 Thus if the input file is @file{^my_ada_proc.adb^MY_ADA_PROC.ADB^} then
15015 @command{gnatpp} will produce @file{^my_ada_proc.adb.pp^MY_ADA_PROC.ADB$PP^}
15017 The output may be redirected by the following switches:
15020 @item ^-pipe^/STANDARD_OUTPUT^
15021 @cindex @option{^-pipe^/STANDARD_OUTPUT^} (@code{gnatpp})
15022 Send the output to @code{Standard_Output}
15024 @item ^-o @var{output_file}^/OUTPUT=@var{output_file}^
15025 @cindex @option{^-o^/OUTPUT^} (@code{gnatpp})
15026 Write the output into @var{output_file}.
15027 If @var{output_file} already exists, @command{gnatpp} terminates without
15028 reading or processing the input file.
15030 @item ^-of ^/FORCED_OUTPUT=^@var{output_file}
15031 @cindex @option{^-of^/FORCED_OUTPUT^} (@code{gnatpp})
15032 Write the output into @var{output_file}, overwriting the existing file
15033 (if one is present).
15035 @item ^-r^/REPLACE^
15036 @cindex @option{^-r^/REPLACE^} (@code{gnatpp})
15037 Replace the input source file with the reformatted output, and copy the
15038 original input source into the file whose name is obtained by appending the
15039 ^@file{.npp}^@file{$NPP}^ suffix to the name of the input file.
15040 If a file with this name already exists, @command{gnatpp} terminates without
15041 reading or processing the input file.
15043 @item ^-rf^/OVERRIDING_REPLACE^
15044 @cindex @option{^-rf^/OVERRIDING_REPLACE^} (@code{gnatpp})
15045 Like @option{^-r^/REPLACE^} except that if the file with the specified name
15046 already exists, it is overwritten.
15048 @item ^-rnb^/NO_BACKUP^
15049 @cindex @option{^-rnb^/NO_BACKUP^} (@code{gnatpp})
15050 Replace the input source file with the reformatted output without
15051 creating any backup copy of the input source.
15055 Options @option{^-pipe^/STANDARD_OUTPUT^},
15056 @option{^-o^/OUTPUT^} and
15057 @option{^-of^/FORCED_OUTPUT^} are allowed only if the call to gnatpp
15058 contains only one file to reformat
15060 @node Other gnatpp Switches
15061 @subsection Other @code{gnatpp} Switches
15064 The additional @command{gnatpp} switches are defined in this subsection.
15067 @item ^-files @var{filename}^/FILES=@var{output_file}^
15068 @cindex @option{^-files^/FILES^} (@code{gnatpp})
15069 Take the argument source files from the specified file. This file should be an
15070 ordinary textual file containing file names separated by spaces or
15071 line breaks. You can use this switch more then once in the same call to
15072 @command{gnatpp}. You also can combine this switch with explicit list of
15075 @item ^-v^/VERBOSE^
15076 @cindex @option{^-v^/VERBOSE^} (@code{gnatpp})
15078 @command{gnatpp} generates version information and then
15079 a trace of the actions it takes to produce or obtain the ASIS tree.
15081 @item ^-w^/WARNINGS^
15082 @cindex @option{^-w^/WARNINGS^} (@code{gnatpp})
15084 @command{gnatpp} generates a warning whenever it can not provide
15085 a required layout in the result source.
15088 @node Formatting Rules
15089 @section Formatting Rules
15092 The following subsections show how @command{gnatpp} treats ``white space'',
15093 comments, program layout, and name casing.
15094 They provide the detailed descriptions of the switches shown above.
15097 * White Space and Empty Lines::
15098 * Formatting Comments::
15099 * Construct Layout::
15103 @node White Space and Empty Lines
15104 @subsection White Space and Empty Lines
15107 @command{gnatpp} does not have an option to control space characters.
15108 It will add or remove spaces according to the style illustrated by the
15109 examples in the @cite{Ada Reference Manual}.
15111 The only format effectors
15112 (see @cite{Ada Reference Manual}, paragraph 2.1(13))
15113 that will appear in the output file are platform-specific line breaks,
15114 and also format effectors within (but not at the end of) comments.
15115 In particular, each horizontal tab character that is not inside
15116 a comment will be treated as a space and thus will appear in the
15117 output file as zero or more spaces depending on
15118 the reformatting of the line in which it appears.
15119 The only exception is a Form Feed character, which is inserted after a
15120 pragma @code{Page} when @option{-ff} is set.
15122 The output file will contain no lines with trailing ``white space'' (spaces,
15125 Empty lines in the original source are preserved
15126 only if they separate declarations or statements.
15127 In such contexts, a
15128 sequence of two or more empty lines is replaced by exactly one empty line.
15129 Note that a blank line will be removed if it separates two ``comment blocks''
15130 (a comment block is a sequence of whole-line comments).
15131 In order to preserve a visual separation between comment blocks, use an
15132 ``empty comment'' (a line comprising only hyphens) rather than an empty line.
15133 Likewise, if for some reason you wish to have a sequence of empty lines,
15134 use a sequence of empty comments instead.
15136 @node Formatting Comments
15137 @subsection Formatting Comments
15140 Comments in Ada code are of two kinds:
15143 a @emph{whole-line comment}, which appears by itself (possibly preceded by
15144 ``white space'') on a line
15147 an @emph{end-of-line comment}, which follows some other Ada lexical element
15152 The indentation of a whole-line comment is that of either
15153 the preceding or following line in
15154 the formatted source, depending on switch settings as will be described below.
15156 For an end-of-line comment, @command{gnatpp} leaves the same number of spaces
15157 between the end of the preceding Ada lexical element and the beginning
15158 of the comment as appear in the original source,
15159 unless either the comment has to be split to
15160 satisfy the line length limitation, or else the next line contains a
15161 whole line comment that is considered a continuation of this end-of-line
15162 comment (because it starts at the same position).
15164 cases, the start of the end-of-line comment is moved right to the nearest
15165 multiple of the indentation level.
15166 This may result in a ``line overflow'' (the right-shifted comment extending
15167 beyond the maximum line length), in which case the comment is split as
15170 There is a difference between @option{^-c1^/COMMENTS_LAYOUT=DEFAULT^}
15171 (GNAT-style comment line indentation)
15172 and @option{^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^}
15173 (reference-manual comment line indentation).
15174 With reference-manual style, a whole-line comment is indented as if it
15175 were a declaration or statement at the same place
15176 (i.e., according to the indentation of the preceding line(s)).
15177 With GNAT style, a whole-line comment that is immediately followed by an
15178 @b{if} or @b{case} statement alternative, a record variant, or the reserved
15179 word @b{begin}, is indented based on the construct that follows it.
15182 @smallexample @c ada
15194 Reference-manual indentation produces:
15196 @smallexample @c ada
15208 while GNAT-style indentation produces:
15210 @smallexample @c ada
15222 The @option{^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^} switch
15223 (GNAT style comment beginning) has the following
15228 For each whole-line comment that does not end with two hyphens,
15229 @command{gnatpp} inserts spaces if necessary after the starting two hyphens
15230 to ensure that there are at least two spaces between these hyphens and the
15231 first non-blank character of the comment.
15235 For an end-of-line comment, if in the original source the next line is a
15236 whole-line comment that starts at the same position
15237 as the end-of-line comment,
15238 then the whole-line comment (and all whole-line comments
15239 that follow it and that start at the same position)
15240 will start at this position in the output file.
15243 That is, if in the original source we have:
15245 @smallexample @c ada
15248 A := B + C; -- B must be in the range Low1..High1
15249 -- C must be in the range Low2..High2
15250 --B+C will be in the range Low1+Low2..High1+High2
15256 Then in the formatted source we get
15258 @smallexample @c ada
15261 A := B + C; -- B must be in the range Low1..High1
15262 -- C must be in the range Low2..High2
15263 -- B+C will be in the range Low1+Low2..High1+High2
15269 A comment that exceeds the line length limit will be split.
15271 @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} (reformat comment blocks) is set and
15272 the line belongs to a reformattable block, splitting the line generates a
15273 @command{gnatpp} warning.
15274 The @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} switch specifies that whole-line
15275 comments may be reformatted in typical
15276 word processor style (that is, moving words between lines and putting as
15277 many words in a line as possible).
15279 @node Construct Layout
15280 @subsection Construct Layout
15283 In several cases the suggested layout in the Ada Reference Manual includes
15284 an extra level of indentation that many programmers prefer to avoid. The
15285 affected cases include:
15289 @item Record type declaration (RM 3.8)
15291 @item Record representation clause (RM 13.5.1)
15293 @item Loop statement in case if a loop has a statement identifier (RM 5.6)
15295 @item Block statement in case if a block has a statement identifier (RM 5.6)
15299 In compact mode (when GNAT style layout or compact layout is set),
15300 the pretty printer uses one level of indentation instead
15301 of two. This is achieved in the record definition and record representation
15302 clause cases by putting the @code{record} keyword on the same line as the
15303 start of the declaration or representation clause, and in the block and loop
15304 case by putting the block or loop header on the same line as the statement
15308 The difference between GNAT style @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^}
15309 and compact @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^}
15310 layout on the one hand, and uncompact layout
15311 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} on the other hand,
15312 can be illustrated by the following examples:
15316 @multitable @columnfractions .5 .5
15317 @item @i{GNAT style, compact layout} @tab @i{Uncompact layout}
15320 @smallexample @c ada
15327 @smallexample @c ada
15336 @smallexample @c ada
15338 a at 0 range 0 .. 31;
15339 b at 4 range 0 .. 31;
15343 @smallexample @c ada
15346 a at 0 range 0 .. 31;
15347 b at 4 range 0 .. 31;
15352 @smallexample @c ada
15360 @smallexample @c ada
15370 @smallexample @c ada
15371 Clear : for J in 1 .. 10 loop
15376 @smallexample @c ada
15378 for J in 1 .. 10 loop
15389 GNAT style, compact layout Uncompact layout
15391 type q is record type q is
15392 a : integer; record
15393 b : integer; a : integer;
15394 end record; b : integer;
15397 for q use record for q use
15398 a at 0 range 0 .. 31; record
15399 b at 4 range 0 .. 31; a at 0 range 0 .. 31;
15400 end record; b at 4 range 0 .. 31;
15403 Block : declare Block :
15404 A : Integer := 3; declare
15405 begin A : Integer := 3;
15407 end Block; Proc (A, A);
15410 Clear : for J in 1 .. 10 loop Clear :
15411 A (J) := 0; for J in 1 .. 10 loop
15412 end loop Clear; A (J) := 0;
15419 A further difference between GNAT style layout and compact layout is that
15420 GNAT style layout inserts empty lines as separation for
15421 compound statements, return statements and bodies.
15424 @subsection Name Casing
15427 @command{gnatpp} always converts the usage occurrence of a (simple) name to
15428 the same casing as the corresponding defining identifier.
15430 You control the casing for defining occurrences via the
15431 @option{^-n^/NAME_CASING^} switch.
15433 With @option{-nD} (``as declared'', which is the default),
15436 With @option{/NAME_CASING=AS_DECLARED}, which is the default,
15438 defining occurrences appear exactly as in the source file
15439 where they are declared.
15440 The other ^values for this switch^options for this qualifier^ ---
15441 @option{^-nU^UPPER_CASE^},
15442 @option{^-nL^LOWER_CASE^},
15443 @option{^-nM^MIXED_CASE^} ---
15445 ^upper, lower, or mixed case, respectively^the corresponding casing^.
15446 If @command{gnatpp} changes the casing of a defining
15447 occurrence, it analogously changes the casing of all the
15448 usage occurrences of this name.
15450 If the defining occurrence of a name is not in the source compilation unit
15451 currently being processed by @command{gnatpp}, the casing of each reference to
15452 this name is changed according to the value of the @option{^-n^/NAME_CASING^}
15453 switch (subject to the dictionary file mechanism described below).
15454 Thus @command{gnatpp} acts as though the @option{^-n^/NAME_CASING^} switch
15456 casing for the defining occurrence of the name.
15458 Some names may need to be spelled with casing conventions that are not
15459 covered by the upper-, lower-, and mixed-case transformations.
15460 You can arrange correct casing by placing such names in a
15461 @emph{dictionary file},
15462 and then supplying a @option{^-D^/DICTIONARY^} switch.
15463 The casing of names from dictionary files overrides
15464 any @option{^-n^/NAME_CASING^} switch.
15466 To handle the casing of Ada predefined names and the names from GNAT libraries,
15467 @command{gnatpp} assumes a default dictionary file.
15468 The name of each predefined entity is spelled with the same casing as is used
15469 for the entity in the @cite{Ada Reference Manual}.
15470 The name of each entity in the GNAT libraries is spelled with the same casing
15471 as is used in the declaration of that entity.
15473 The @w{@option{^-D-^/SPECIFIC_CASING^}} switch suppresses the use of the
15474 default dictionary file.
15475 Instead, the casing for predefined and GNAT-defined names will be established
15476 by the @option{^-n^/NAME_CASING^} switch or explicit dictionary files.
15477 For example, by default the names @code{Ada.Text_IO} and @code{GNAT.OS_Lib}
15478 will appear as just shown,
15479 even in the presence of a @option{^-nU^/NAME_CASING=UPPER_CASE^} switch.
15480 To ensure that even such names are rendered in uppercase,
15481 additionally supply the @w{@option{^-D-^/SPECIFIC_CASING^}} switch
15482 (or else, less conveniently, place these names in upper case in a dictionary
15485 A dictionary file is
15486 a plain text file; each line in this file can be either a blank line
15487 (containing only space characters and ASCII.HT characters), an Ada comment
15488 line, or the specification of exactly one @emph{casing schema}.
15490 A casing schema is a string that has the following syntax:
15494 @var{casing_schema} ::= @var{identifier} | *@var{simple_identifier}*
15496 @var{simple_identifier} ::= @var{letter}@{@var{letter_or_digit}@}
15501 (See @cite{Ada Reference Manual}, Section 2.3) for the definition of the
15502 @var{identifier} lexical element and the @var{letter_or_digit} category.)
15504 The casing schema string can be followed by white space and/or an Ada-style
15505 comment; any amount of white space is allowed before the string.
15507 If a dictionary file is passed as
15509 the value of a @option{-D@var{file}} switch
15512 an option to the @option{/DICTIONARY} qualifier
15515 simple name and every identifier, @command{gnatpp} checks if the dictionary
15516 defines the casing for the name or for some of its parts (the term ``subword''
15517 is used below to denote the part of a name which is delimited by ``_'' or by
15518 the beginning or end of the word and which does not contain any ``_'' inside):
15522 if the whole name is in the dictionary, @command{gnatpp} uses for this name
15523 the casing defined by the dictionary; no subwords are checked for this word
15526 for every subword @command{gnatpp} checks if the dictionary contains the
15527 corresponding string of the form @code{*@var{simple_identifier}*},
15528 and if it does, the casing of this @var{simple_identifier} is used
15532 if the whole name does not contain any ``_'' inside, and if for this name
15533 the dictionary contains two entries - one of the form @var{identifier},
15534 and another - of the form *@var{simple_identifier}*, then the first one
15535 is applied to define the casing of this name
15538 if more than one dictionary file is passed as @command{gnatpp} switches, each
15539 dictionary adds new casing exceptions and overrides all the existing casing
15540 exceptions set by the previous dictionaries
15543 when @command{gnatpp} checks if the word or subword is in the dictionary,
15544 this check is not case sensitive
15548 For example, suppose we have the following source to reformat:
15550 @smallexample @c ada
15553 name1 : integer := 1;
15554 name4_name3_name2 : integer := 2;
15555 name2_name3_name4 : Boolean;
15558 name2_name3_name4 := name4_name3_name2 > name1;
15564 And suppose we have two dictionaries:
15581 If @command{gnatpp} is called with the following switches:
15585 @command{gnatpp -nM -D dict1 -D dict2 test.adb}
15588 @command{gnatpp test.adb /NAME_CASING=MIXED_CASE /DICTIONARY=(dict1, dict2)}
15593 then we will get the following name casing in the @command{gnatpp} output:
15595 @smallexample @c ada
15598 NAME1 : Integer := 1;
15599 Name4_NAME3_Name2 : Integer := 2;
15600 Name2_NAME3_Name4 : Boolean;
15603 Name2_NAME3_Name4 := Name4_NAME3_Name2 > NAME1;
15608 @c *********************************
15609 @node The GNAT Metric Tool gnatmetric
15610 @chapter The GNAT Metric Tool @command{gnatmetric}
15612 @cindex Metric tool
15615 ^The @command{gnatmetric} tool^@command{GNAT METRIC}^ is an ASIS-based utility
15616 for computing various program metrics.
15617 It takes an Ada source file as input and generates a file containing the
15618 metrics data as output. Various switches control which
15619 metrics are computed and output.
15621 @command{gnatmetric} generates and uses the ASIS
15622 tree for the input source and thus requires the input to be syntactically and
15623 semantically legal.
15624 If this condition is not met, @command{gnatmetric} will generate
15625 an error message; no metric information for this file will be
15626 computed and reported.
15628 If the compilation unit contained in the input source depends semantically
15629 upon units in files located outside the current directory, you have to provide
15630 the source search path when invoking @command{gnatmetric}.
15631 If it depends semantically upon units that are contained
15632 in files with names that do not follow the GNAT file naming rules, you have to
15633 provide the configuration file describing the corresponding naming scheme (see
15634 the description of the @command{gnatmetric} switches below.)
15635 Alternatively, you may use a project file and invoke @command{gnatmetric}
15636 through the @command{gnat} driver.
15639 The @command{gnatmetric} command has the form
15642 $ gnatmetric [@i{switches}] @{@i{filename}@} [@i{-cargs gcc_switches}]
15649 @i{switches} specify the metrics to compute and define the destination for
15653 Each @i{filename} is the name (including the extension) of a source
15654 file to process. ``Wildcards'' are allowed, and
15655 the file name may contain path information.
15656 If no @i{filename} is supplied, then the @i{switches} list must contain
15658 @option{-files} switch (@pxref{Other gnatmetric Switches}).
15659 Including both a @option{-files} switch and one or more
15660 @i{filename} arguments is permitted.
15663 @i{-cargs gcc_switches} is a list of switches for
15664 @command{gcc}. They will be passed on to all compiler invocations made by
15665 @command{gnatmetric} to generate the ASIS trees. Here you can provide
15666 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
15667 and use the @option{-gnatec} switch to set the configuration file.
15671 * Switches for gnatmetric::
15674 @node Switches for gnatmetric
15675 @section Switches for @command{gnatmetric}
15678 The following subsections describe the various switches accepted by
15679 @command{gnatmetric}, organized by category.
15682 * Output Files Control::
15683 * Disable Metrics For Local Units::
15684 * Line Metrics Control::
15685 * Syntax Metrics Control::
15686 * Complexity Metrics Control::
15687 * Other gnatmetric Switches::
15690 @node Output Files Control
15691 @subsection Output File Control
15692 @cindex Output file control in @command{gnatmetric}
15695 @command{gnatmetric} has two output formats. It can generate a
15696 textual (human-readable) form, and also XML. By default only textual
15697 output is generated.
15699 When generating the output in textual form, @command{gnatmetric} creates
15700 for each Ada source file a corresponding text file
15701 containing the computed metrics. By default, this file
15702 is placed in the same directory as where the source file is located, and
15703 its name is obtained
15704 by appending the ^@file{.metrix}^@file{$METRIX}^ suffix to the name of the
15707 All the output information generated in XML format is placed in a single
15708 file. By default this file is placed in the current directory and has the
15709 name ^@file{metrix.xml}^@file{METRIX$XML}^.
15711 Some of the computed metrics are summed over the units passed to
15712 @command{gnatmetric}; for example, the total number of lines of code.
15713 By default this information is sent to @file{stdout}, but a file
15714 can be specified with the @option{-og} switch.
15716 The following switches control the @command{gnatmetric} output:
15719 @cindex @option{^-x^/XML^} (@command{gnatmetric})
15721 Generate the XML output
15723 @cindex @option{^-nt^/NO_TEXT^} (@command{gnatmetric})
15724 @item ^-nt^/NO_TEXT^
15725 Do not generate the output in text form (implies @option{^-x^/XML^})
15727 @cindex @option{^-d^/DIRECTORY^} (@command{gnatmetric})
15728 @item ^-d @var{output_dir}^/DIRECTORY=@var{output_dir}^
15729 Put textual files with detailed metrics into @var{output_dir}
15731 @cindex @option{^-o^/SUFFIX_DETAILS^} (@command{gnatmetric})
15732 @item ^-o @var{file_suffix}^/SUFFIX_DETAILS=@var{file_suffix}^
15733 Use @var{file_suffix}, instead of ^@file{.metrix}^@file{$METRIX}^
15734 in the name of the output file.
15736 @cindex @option{^-og^/GLOBAL_OUTPUT^} (@command{gnatmetric})
15737 @item ^-og @var{file_name}^/GLOBAL_OUTPUT=@var{file_name}^
15738 Put global metrics into @var{file_name}
15740 @cindex @option{^-ox^/XML_OUTPUT^} (@command{gnatmetric})
15741 @item ^-ox @var{file_name}^/XML_OUTPUT=@var{file_name}^
15742 Put the XML output into @var{file_name} (also implies @option{^-x^/XML^})
15744 @cindex @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} (@command{gnatmetric})
15745 @item ^-sfn^/SHORT_SOURCE_FILE_NAME^
15746 Use ``short'' source file names in the output. (The @command{gnatmetric}
15747 output includes the name(s) of the Ada source file(s) from which the metrics
15748 are computed. By default each name includes the absolute path. The
15749 @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} switch causes @command{gnatmetric}
15750 to exclude all directory information from the file names that are output.)
15754 @node Disable Metrics For Local Units
15755 @subsection Disable Metrics For Local Units
15756 @cindex Disable Metrics For Local Units in @command{gnatmetric}
15759 @command{gnatmetric} relies on the GNAT compilation model @minus{}
15761 unit per one source file. It computes line metrics for the whole source
15762 file, and it also computes syntax
15763 and complexity metrics for the file's outermost unit.
15765 By default, @command{gnatmetric} will also compute all metrics for certain
15766 kinds of locally declared program units:
15770 subprogram (and generic subprogram) bodies;
15773 package (and generic package) specifications and bodies;
15776 task object and type specifications and bodies;
15779 protected object and type specifications and bodies.
15783 These kinds of entities will be referred to as
15784 @emph{eligible local program units}, or simply @emph{eligible local units},
15785 @cindex Eligible local unit (for @command{gnatmetric})
15786 in the discussion below.
15788 Note that a subprogram declaration, generic instantiation,
15789 or renaming declaration only receives metrics
15790 computation when it appear as the outermost entity
15793 Suppression of metrics computation for eligible local units can be
15794 obtained via the following switch:
15797 @cindex @option{^-n@var{x}^/SUPPRESS^} (@command{gnatmetric})
15798 @item ^-nolocal^/SUPPRESS=LOCAL_DETAILS^
15799 Do not compute detailed metrics for eligible local program units
15803 @node Line Metrics Control
15804 @subsection Line Metrics Control
15805 @cindex Line metrics control in @command{gnatmetric}
15808 For any (legal) source file, and for each of its
15809 eligible local program units, @command{gnatmetric} computes the following
15814 the total number of lines;
15817 the total number of code lines (i.e., non-blank lines that are not comments)
15820 the number of comment lines
15823 the number of code lines containing end-of-line comments;
15826 the number of empty lines and lines containing only space characters and/or
15827 format effectors (blank lines)
15831 If @command{gnatmetric} is invoked on more than one source file, it sums the
15832 values of the line metrics for all the files being processed and then
15833 generates the cumulative results.
15835 By default, all the line metrics are computed and reported. You can use the
15836 following switches to select the specific line metrics to be computed and
15837 reported (if any of these parameters is set, only explicitly specified line
15838 metrics are computed).
15841 @cindex @option{^-la^/LINES_ALL^} (@command{gnatmetric})
15842 @item ^-la^/LINES_ALL^
15843 The number of all lines
15845 @cindex @option{^-lcode^/CODE_LINES^} (@command{gnatmetric})
15846 @item ^-lcode^/CODE_LINES^
15847 The number of code lines
15849 @cindex @option{^-lcomm^/COMENT_LINES^} (@command{gnatmetric})
15850 @item ^-lcomm^/COMENT_LINES^
15851 The number of comment lines
15853 @cindex @option{^-leol^/MIXED_CODE_COMMENTS^} (@command{gnatmetric})
15854 @item ^-leol^/MIXED_CODE_COMMENTS^
15855 The number of code lines containing
15856 end-of-line comments
15858 @cindex @option{^-lb^/BLANK_LINES^} (@command{gnatmetric})
15859 @item ^-lb^/BLANK_LINES^
15860 The number of blank lines
15865 @node Syntax Metrics Control
15866 @subsection Syntax Metrics Control
15867 @cindex Syntax metrics control in @command{gnatmetric}
15870 @command{gnatmetric} computes various syntactic metrics for the
15871 outermost unit and for each eligible local unit:
15874 @item LSLOC (``Logical Source Lines Of Code'')
15875 The total number of declarations and the total number of statements
15877 @item Maximal static nesting level of inner program units
15879 @cite{Ada 95 Language Reference Manual}, 10.1(1), ``A program unit is either a
15880 package, a task unit, a protected unit, a
15881 protected entry, a generic unit, or an explicitly declared subprogram other
15882 than an enumeration literal.''
15884 @item Maximal nesting level of composite syntactic constructs
15885 This corresponds to the notion of the
15886 maximum nesting level in the GNAT built-in style checks
15887 (@pxref{Style Checking})
15891 For the outermost unit in the file, @command{gnatmetric} additionally computes
15892 the following metrics:
15895 @item Public subprograms
15896 This metric is computed for package specifications. It is the
15897 number of subprograms and generic subprograms declared in the visible
15898 part (including in nested packages, protected objects, and
15901 @item All subprograms
15902 This metric is computed for bodies and subunits. The
15903 metric is equal to a total number of subprogram bodies in the compilation
15905 Neither generic instantiations nor renamings-as-a-body nor body stubs
15906 are counted. Any subprogram body is counted, independently of its nesting
15907 level and enclosing constructs. Generic bodies and bodies of protected
15908 subprograms are counted in the same way as ``usual'' subprogram bodies.
15911 This metric is computed for package specifications and
15912 generic package declarations. It is the total number of types
15913 that can be referenced from outside this compilation unit, plus the
15914 number of types from all the visible parts of all the visible generic packages.
15915 Generic formal types are not counted. Only types, not subtypes,
15919 Along with the total number of public types, the following
15920 types are counted and reported separately:
15927 Root tagged types (abstract, non-abstract, private, non-private). Type
15928 extensions are @emph{not} counted
15931 Private types (including private extensions)
15942 This metric is computed for any compilation unit. It is equal to the total
15943 number of the declarations of different types given in the compilation unit.
15944 The private and the corresponding full type declaration are counted as one
15945 type declaration. Incomplete type declarations and generic formal types
15947 No distinction is made among different kinds of types (abstract,
15948 private etc.); the total number of types is computed and reported.
15953 By default, all the syntax metrics are computed and reported. You can use the
15954 following switches to select specific syntax metrics;
15955 if any of these is set, only the explicitly specified metrics are computed.
15958 @cindex @option{^-ed^/DECLARATION_TOTAL^} (@command{gnatmetric})
15959 @item ^-ed^/DECLARATION_TOTAL^
15960 The total number of declarations
15962 @cindex @option{^-es^/STATEMENT_TOTAL^} (@command{gnatmetric})
15963 @item ^-es^/STATEMENT_TOTAL^
15964 The total number of statements
15966 @cindex @option{^-eps^/^} (@command{gnatmetric})
15967 @item ^-eps^/INT_SUBPROGRAMS^
15968 The number of public subprograms in a compilation unit
15970 @cindex @option{^-eas^/SUBPROGRAMS_ALL^} (@command{gnatmetric})
15971 @item ^-eas^/SUBPROGRAMS_ALL^
15972 The number of all the subprograms in a compilation unit
15974 @cindex @option{^-ept^/INT_TYPES^} (@command{gnatmetric})
15975 @item ^-ept^/INT_TYPES^
15976 The number of public types in a compilation unit
15978 @cindex @option{^-eat^/TYPES_ALL^} (@command{gnatmetric})
15979 @item ^-eat^/TYPES_ALL^
15980 The number of all the types in a compilation unit
15982 @cindex @option{^-enu^/PROGRAM_NESTING_MAX^} (@command{gnatmetric})
15983 @item ^-enu^/PROGRAM_NESTING_MAX^
15984 The maximal program unit nesting level
15986 @cindex @option{^-ec^/CONSTRUCT_NESTING_MAX^} (@command{gnatmetric})
15987 @item ^-ec^/CONSTRUCT_NESTING_MAX^
15988 The maximal construct nesting level
15992 @node Complexity Metrics Control
15993 @subsection Complexity Metrics Control
15994 @cindex Complexity metrics control in @command{gnatmetric}
15997 For a program unit that is an executable body (a subprogram body (including
15998 generic bodies), task body, entry body or a package body containing
15999 its own statement sequence ) @command{gnatmetric} computes the following
16000 complexity metrics:
16004 McCabe cyclomatic complexity;
16007 McCabe essential complexity;
16010 maximal loop nesting level
16015 The McCabe complexity metrics are defined
16016 in @url{www.mccabe.com/pdf/nist235r.pdf}
16018 According to McCabe, both control statements and short-circuit control forms
16019 should be taken into account when computing cyclomatic complexity. For each
16020 body, we compute three metric values:
16024 the complexity introduced by control
16025 statements only, without taking into account short-circuit forms,
16028 the complexity introduced by short-circuit control forms only, and
16032 cyclomatic complexity, which is the sum of these two values.
16036 When computing cyclomatic and essential complexity, @command{gnatmetric} skips
16037 the code in the exception handlers and in all the nested program units.
16039 By default, all the complexity metrics are computed and reported.
16040 For more finely-grained control you can use
16041 the following switches:
16044 @cindex @option{^-n@var{x}^/SUPPRESS^} (@command{gnatmetric})
16046 @item ^-nocc^/SUPPRESS=CYCLOMATIC_COMPLEXITY^
16047 Do not compute the McCabe Cyclomatic Complexity
16049 @item ^-noec^/SUPPRESS=ESSENTIAL_COMPLEXITY^
16050 Do not compute the Essential Complexity
16052 @item ^-nonl^/SUPPRESS=MAXIMAL_LOOP_NESTING^
16053 Do not compute maximal loop nesting level
16055 @item ^-ne^/SUPPRESS=EXITS_AS_GOTOS^
16056 Do not consider @code{exit} statements as @code{goto}s when
16057 computing Essential Complexity
16061 @node Other gnatmetric Switches
16062 @subsection Other @code{gnatmetric} Switches
16065 Additional @command{gnatmetric} switches are as follows:
16068 @item ^-files @var{filename}^/FILES=@var{filename}^
16069 @cindex @option{^-files^/FILES^} (@code{gnatmetric})
16070 Take the argument source files from the specified file. This file should be an
16071 ordinary textual file containing file names separated by spaces or
16072 line breaks. You can use this switch more then once in the same call to
16073 @command{gnatmetric}. You also can combine this switch with
16074 an explicit list of files.
16076 @item ^-v^/VERBOSE^
16077 @cindex @option{^-v^/VERBOSE^} (@code{gnatmetric})
16079 @command{gnatmetric} generates version information and then
16080 a trace of sources being processed.
16082 @item ^-dv^/DEBUG_OUTPUT^
16083 @cindex @option{^-dv^/DEBUG_OUTPUT^} (@code{gnatmetric})
16085 @command{gnatmetric} generates various messages useful to understand what
16086 happens during the metrics computation
16089 @cindex @option{^-q^/QUIET^} (@code{gnatmetric})
16093 @c ***********************************
16094 @node File Name Krunching Using gnatkr
16095 @chapter File Name Krunching Using @code{gnatkr}
16099 This chapter discusses the method used by the compiler to shorten
16100 the default file names chosen for Ada units so that they do not
16101 exceed the maximum length permitted. It also describes the
16102 @code{gnatkr} utility that can be used to determine the result of
16103 applying this shortening.
16107 * Krunching Method::
16108 * Examples of gnatkr Usage::
16112 @section About @code{gnatkr}
16115 The default file naming rule in GNAT
16116 is that the file name must be derived from
16117 the unit name. The exact default rule is as follows:
16120 Take the unit name and replace all dots by hyphens.
16122 If such a replacement occurs in the
16123 second character position of a name, and the first character is
16124 ^a, g, s, or i^A, G, S, or I^ then replace the dot by the character
16125 ^~ (tilde)^$ (dollar sign)^
16126 instead of a minus.
16128 The reason for this exception is to avoid clashes
16129 with the standard names for children of System, Ada, Interfaces,
16130 and GNAT, which use the prefixes ^s- a- i- and g-^S- A- I- and G-^
16133 The @option{^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{nn}}
16134 switch of the compiler activates a ``krunching''
16135 circuit that limits file names to nn characters (where nn is a decimal
16136 integer). For example, using OpenVMS,
16137 where the maximum file name length is
16138 39, the value of nn is usually set to 39, but if you want to generate
16139 a set of files that would be usable if ported to a system with some
16140 different maximum file length, then a different value can be specified.
16141 The default value of 39 for OpenVMS need not be specified.
16143 The @code{gnatkr} utility can be used to determine the krunched name for
16144 a given file, when krunched to a specified maximum length.
16147 @section Using @code{gnatkr}
16150 The @code{gnatkr} command has the form
16154 $ gnatkr @var{name} [@var{length}]
16160 $ gnatkr @var{name} /COUNT=nn
16165 @var{name} is the uncrunched file name, derived from the name of the unit
16166 in the standard manner described in the previous section (i.e. in particular
16167 all dots are replaced by hyphens). The file name may or may not have an
16168 extension (defined as a suffix of the form period followed by arbitrary
16169 characters other than period). If an extension is present then it will
16170 be preserved in the output. For example, when krunching @file{hellofile.ads}
16171 to eight characters, the result will be hellofil.ads.
16173 Note: for compatibility with previous versions of @code{gnatkr} dots may
16174 appear in the name instead of hyphens, but the last dot will always be
16175 taken as the start of an extension. So if @code{gnatkr} is given an argument
16176 such as @file{Hello.World.adb} it will be treated exactly as if the first
16177 period had been a hyphen, and for example krunching to eight characters
16178 gives the result @file{hellworl.adb}.
16180 Note that the result is always all lower case (except on OpenVMS where it is
16181 all upper case). Characters of the other case are folded as required.
16183 @var{length} represents the length of the krunched name. The default
16184 when no argument is given is ^8^39^ characters. A length of zero stands for
16185 unlimited, in other words do not chop except for system files where the
16186 implied crunching length is always eight characters.
16189 The output is the krunched name. The output has an extension only if the
16190 original argument was a file name with an extension.
16192 @node Krunching Method
16193 @section Krunching Method
16196 The initial file name is determined by the name of the unit that the file
16197 contains. The name is formed by taking the full expanded name of the
16198 unit and replacing the separating dots with hyphens and
16199 using ^lowercase^uppercase^
16200 for all letters, except that a hyphen in the second character position is
16201 replaced by a ^tilde^dollar sign^ if the first character is
16202 ^a, i, g, or s^A, I, G, or S^.
16203 The extension is @code{.ads} for a
16204 specification and @code{.adb} for a body.
16205 Krunching does not affect the extension, but the file name is shortened to
16206 the specified length by following these rules:
16210 The name is divided into segments separated by hyphens, tildes or
16211 underscores and all hyphens, tildes, and underscores are
16212 eliminated. If this leaves the name short enough, we are done.
16215 If the name is too long, the longest segment is located (left-most
16216 if there are two of equal length), and shortened by dropping
16217 its last character. This is repeated until the name is short enough.
16219 As an example, consider the krunching of @*@file{our-strings-wide_fixed.adb}
16220 to fit the name into 8 characters as required by some operating systems.
16223 our-strings-wide_fixed 22
16224 our strings wide fixed 19
16225 our string wide fixed 18
16226 our strin wide fixed 17
16227 our stri wide fixed 16
16228 our stri wide fixe 15
16229 our str wide fixe 14
16230 our str wid fixe 13
16236 Final file name: oustwifi.adb
16240 The file names for all predefined units are always krunched to eight
16241 characters. The krunching of these predefined units uses the following
16242 special prefix replacements:
16246 replaced by @file{^a^A^-}
16249 replaced by @file{^g^G^-}
16252 replaced by @file{^i^I^-}
16255 replaced by @file{^s^S^-}
16258 These system files have a hyphen in the second character position. That
16259 is why normal user files replace such a character with a
16260 ^tilde^dollar sign^, to
16261 avoid confusion with system file names.
16263 As an example of this special rule, consider
16264 @*@file{ada-strings-wide_fixed.adb}, which gets krunched as follows:
16267 ada-strings-wide_fixed 22
16268 a- strings wide fixed 18
16269 a- string wide fixed 17
16270 a- strin wide fixed 16
16271 a- stri wide fixed 15
16272 a- stri wide fixe 14
16273 a- str wide fixe 13
16279 Final file name: a-stwifi.adb
16283 Of course no file shortening algorithm can guarantee uniqueness over all
16284 possible unit names, and if file name krunching is used then it is your
16285 responsibility to ensure that no name clashes occur. The utility
16286 program @code{gnatkr} is supplied for conveniently determining the
16287 krunched name of a file.
16289 @node Examples of gnatkr Usage
16290 @section Examples of @code{gnatkr} Usage
16297 $ gnatkr very_long_unit_name.ads --> velounna.ads
16298 $ gnatkr grandparent-parent-child.ads --> grparchi.ads
16299 $ gnatkr Grandparent.Parent.Child.ads --> grparchi.ads
16300 $ gnatkr grandparent-parent-child --> grparchi
16302 $ gnatkr very_long_unit_name.ads/count=6 --> vlunna.ads
16303 $ gnatkr very_long_unit_name.ads/count=0 --> very_long_unit_name.ads
16306 @node Preprocessing Using gnatprep
16307 @chapter Preprocessing Using @code{gnatprep}
16311 The @code{gnatprep} utility provides
16312 a simple preprocessing capability for Ada programs.
16313 It is designed for use with GNAT, but is not dependent on any special
16318 * Switches for gnatprep::
16319 * Form of Definitions File::
16320 * Form of Input Text for gnatprep::
16323 @node Using gnatprep
16324 @section Using @code{gnatprep}
16327 To call @code{gnatprep} use
16330 $ gnatprep [-bcrsu] [-Dsymbol=value] infile outfile [deffile]
16337 is the full name of the input file, which is an Ada source
16338 file containing preprocessor directives.
16341 is the full name of the output file, which is an Ada source
16342 in standard Ada form. When used with GNAT, this file name will
16343 normally have an ads or adb suffix.
16346 is the full name of a text file containing definitions of
16347 symbols to be referenced by the preprocessor. This argument is
16348 optional, and can be replaced by the use of the @option{-D} switch.
16351 is an optional sequence of switches as described in the next section.
16354 @node Switches for gnatprep
16355 @section Switches for @code{gnatprep}
16360 @item ^-b^/BLANK_LINES^
16361 @cindex @option{^-b^/BLANK_LINES^} (@command{gnatprep})
16362 Causes both preprocessor lines and the lines deleted by
16363 preprocessing to be replaced by blank lines in the output source file,
16364 preserving line numbers in the output file.
16366 @item ^-c^/COMMENTS^
16367 @cindex @option{^-c^/COMMENTS^} (@command{gnatprep})
16368 Causes both preprocessor lines and the lines deleted
16369 by preprocessing to be retained in the output source as comments marked
16370 with the special string @code{"--! "}. This option will result in line numbers
16371 being preserved in the output file.
16373 @item ^-Dsymbol=value^/ASSOCIATE="symbol=value"^
16374 @cindex @option{^-D^/ASSOCIATE^} (@command{gnatprep})
16375 Defines a new symbol, associated with value. If no value is given on the
16376 command line, then symbol is considered to be @code{True}. This switch
16377 can be used in place of a definition file.
16381 @cindex @option{/REMOVE} (@command{gnatprep})
16382 This is the default setting which causes lines deleted by preprocessing
16383 to be entirely removed from the output file.
16386 @item ^-r^/REFERENCE^
16387 @cindex @option{^-r^/REFERENCE^} (@command{gnatprep})
16388 Causes a @code{Source_Reference} pragma to be generated that
16389 references the original input file, so that error messages will use
16390 the file name of this original file. The use of this switch implies
16391 that preprocessor lines are not to be removed from the file, so its
16392 use will force @option{^-b^/BLANK_LINES^} mode if
16393 @option{^-c^/COMMENTS^}
16394 has not been specified explicitly.
16396 Note that if the file to be preprocessed contains multiple units, then
16397 it will be necessary to @code{gnatchop} the output file from
16398 @code{gnatprep}. If a @code{Source_Reference} pragma is present
16399 in the preprocessed file, it will be respected by
16400 @code{gnatchop ^-r^/REFERENCE^}
16401 so that the final chopped files will correctly refer to the original
16402 input source file for @code{gnatprep}.
16404 @item ^-s^/SYMBOLS^
16405 @cindex @option{^-s^/SYMBOLS^} (@command{gnatprep})
16406 Causes a sorted list of symbol names and values to be
16407 listed on the standard output file.
16409 @item ^-u^/UNDEFINED^
16410 @cindex @option{^-u^/UNDEFINED^} (@command{gnatprep})
16411 Causes undefined symbols to be treated as having the value FALSE in the context
16412 of a preprocessor test. In the absence of this option, an undefined symbol in
16413 a @code{#if} or @code{#elsif} test will be treated as an error.
16419 Note: if neither @option{-b} nor @option{-c} is present,
16420 then preprocessor lines and
16421 deleted lines are completely removed from the output, unless -r is
16422 specified, in which case -b is assumed.
16425 @node Form of Definitions File
16426 @section Form of Definitions File
16429 The definitions file contains lines of the form
16436 where symbol is an identifier, following normal Ada (case-insensitive)
16437 rules for its syntax, and value is one of the following:
16441 Empty, corresponding to a null substitution
16443 A string literal using normal Ada syntax
16445 Any sequence of characters from the set
16446 (letters, digits, period, underline).
16450 Comment lines may also appear in the definitions file, starting with
16451 the usual @code{--},
16452 and comments may be added to the definitions lines.
16454 @node Form of Input Text for gnatprep
16455 @section Form of Input Text for @code{gnatprep}
16458 The input text may contain preprocessor conditional inclusion lines,
16459 as well as general symbol substitution sequences.
16461 The preprocessor conditional inclusion commands have the form
16466 #if @i{expression} [then]
16468 #elsif @i{expression} [then]
16470 #elsif @i{expression} [then]
16481 In this example, @i{expression} is defined by the following grammar:
16483 @i{expression} ::= <symbol>
16484 @i{expression} ::= <symbol> = "<value>"
16485 @i{expression} ::= <symbol> = <symbol>
16486 @i{expression} ::= <symbol> 'Defined
16487 @i{expression} ::= not @i{expression}
16488 @i{expression} ::= @i{expression} and @i{expression}
16489 @i{expression} ::= @i{expression} or @i{expression}
16490 @i{expression} ::= @i{expression} and then @i{expression}
16491 @i{expression} ::= @i{expression} or else @i{expression}
16492 @i{expression} ::= ( @i{expression} )
16496 For the first test (@i{expression} ::= <symbol>) the symbol must have
16497 either the value true or false, that is to say the right-hand of the
16498 symbol definition must be one of the (case-insensitive) literals
16499 @code{True} or @code{False}. If the value is true, then the
16500 corresponding lines are included, and if the value is false, they are
16503 The test (@i{expression} ::= <symbol> @code{'Defined}) is true only if
16504 the symbol has been defined in the definition file or by a @option{-D}
16505 switch on the command line. Otherwise, the test is false.
16507 The equality tests are case insensitive, as are all the preprocessor lines.
16509 If the symbol referenced is not defined in the symbol definitions file,
16510 then the effect depends on whether or not switch @option{-u}
16511 is specified. If so, then the symbol is treated as if it had the value
16512 false and the test fails. If this switch is not specified, then
16513 it is an error to reference an undefined symbol. It is also an error to
16514 reference a symbol that is defined with a value other than @code{True}
16517 The use of the @code{not} operator inverts the sense of this logical test, so
16518 that the lines are included only if the symbol is not defined.
16519 The @code{then} keyword is optional as shown
16521 The @code{#} must be the first non-blank character on a line, but
16522 otherwise the format is free form. Spaces or tabs may appear between
16523 the @code{#} and the keyword. The keywords and the symbols are case
16524 insensitive as in normal Ada code. Comments may be used on a
16525 preprocessor line, but other than that, no other tokens may appear on a
16526 preprocessor line. Any number of @code{elsif} clauses can be present,
16527 including none at all. The @code{else} is optional, as in Ada.
16529 The @code{#} marking the start of a preprocessor line must be the first
16530 non-blank character on the line, i.e. it must be preceded only by
16531 spaces or horizontal tabs.
16533 Symbol substitution outside of preprocessor lines is obtained by using
16541 anywhere within a source line, except in a comment or within a
16542 string literal. The identifier
16543 following the @code{$} must match one of the symbols defined in the symbol
16544 definition file, and the result is to substitute the value of the
16545 symbol in place of @code{$symbol} in the output file.
16547 Note that although the substitution of strings within a string literal
16548 is not possible, it is possible to have a symbol whose defined value is
16549 a string literal. So instead of setting XYZ to @code{hello} and writing:
16552 Header : String := "$XYZ";
16556 you should set XYZ to @code{"hello"} and write:
16559 Header : String := $XYZ;
16563 and then the substitution will occur as desired.
16566 @node The GNAT Run-Time Library Builder gnatlbr
16567 @chapter The GNAT Run-Time Library Builder @code{gnatlbr}
16569 @cindex Library builder
16572 @code{gnatlbr} is a tool for rebuilding the GNAT run time with user
16573 supplied configuration pragmas.
16576 * Running gnatlbr::
16577 * Switches for gnatlbr::
16578 * Examples of gnatlbr Usage::
16581 @node Running gnatlbr
16582 @section Running @code{gnatlbr}
16585 The @code{gnatlbr} command has the form
16588 $ GNAT LIBRARY /[CREATE | SET | DELETE]=directory [/CONFIG=file]
16591 @node Switches for gnatlbr
16592 @section Switches for @code{gnatlbr}
16595 @code{gnatlbr} recognizes the following switches:
16599 @item /CREATE=directory
16600 @cindex @code{/CREATE} (@code{gnatlbr})
16601 Create the new run-time library in the specified directory.
16603 @item /SET=directory
16604 @cindex @code{/SET} (@code{gnatlbr})
16605 Make the library in the specified directory the current run-time
16608 @item /DELETE=directory
16609 @cindex @code{/DELETE} (@code{gnatlbr})
16610 Delete the run-time library in the specified directory.
16613 @cindex @code{/CONFIG} (@code{gnatlbr})
16615 Use the configuration pragmas in the specified file when building
16619 Use the configuration pragmas in the specified file when compiling.
16623 @node Examples of gnatlbr Usage
16624 @section Example of @code{gnatlbr} Usage
16627 Contents of VAXFLOAT.ADC:
16628 pragma Float_Representation (VAX_Float);
16630 $ GNAT LIBRARY /CREATE=[.VAXFLOAT] /CONFIG=VAXFLOAT.ADC
16632 GNAT LIBRARY rebuilds the run-time library in directory [.VAXFLOAT]
16637 @node The GNAT Library Browser gnatls
16638 @chapter The GNAT Library Browser @code{gnatls}
16640 @cindex Library browser
16643 @code{gnatls} is a tool that outputs information about compiled
16644 units. It gives the relationship between objects, unit names and source
16645 files. It can also be used to check the source dependencies of a unit
16646 as well as various characteristics.
16650 * Switches for gnatls::
16651 * Examples of gnatls Usage::
16654 @node Running gnatls
16655 @section Running @code{gnatls}
16658 The @code{gnatls} command has the form
16661 $ gnatls switches @var{object_or_ali_file}
16665 The main argument is the list of object or @file{ali} files
16666 (@pxref{The Ada Library Information Files})
16667 for which information is requested.
16669 In normal mode, without additional option, @code{gnatls} produces a
16670 four-column listing. Each line represents information for a specific
16671 object. The first column gives the full path of the object, the second
16672 column gives the name of the principal unit in this object, the third
16673 column gives the status of the source and the fourth column gives the
16674 full path of the source representing this unit.
16675 Here is a simple example of use:
16679 ^./^[]^demo1.o demo1 DIF demo1.adb
16680 ^./^[]^demo2.o demo2 OK demo2.adb
16681 ^./^[]^hello.o h1 OK hello.adb
16682 ^./^[]^instr-child.o instr.child MOK instr-child.adb
16683 ^./^[]^instr.o instr OK instr.adb
16684 ^./^[]^tef.o tef DIF tef.adb
16685 ^./^[]^text_io_example.o text_io_example OK text_io_example.adb
16686 ^./^[]^tgef.o tgef DIF tgef.adb
16690 The first line can be interpreted as follows: the main unit which is
16692 object file @file{demo1.o} is demo1, whose main source is in
16693 @file{demo1.adb}. Furthermore, the version of the source used for the
16694 compilation of demo1 has been modified (DIF). Each source file has a status
16695 qualifier which can be:
16698 @item OK (unchanged)
16699 The version of the source file used for the compilation of the
16700 specified unit corresponds exactly to the actual source file.
16702 @item MOK (slightly modified)
16703 The version of the source file used for the compilation of the
16704 specified unit differs from the actual source file but not enough to
16705 require recompilation. If you use gnatmake with the qualifier
16706 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}, a file marked
16707 MOK will not be recompiled.
16709 @item DIF (modified)
16710 No version of the source found on the path corresponds to the source
16711 used to build this object.
16713 @item ??? (file not found)
16714 No source file was found for this unit.
16716 @item HID (hidden, unchanged version not first on PATH)
16717 The version of the source that corresponds exactly to the source used
16718 for compilation has been found on the path but it is hidden by another
16719 version of the same source that has been modified.
16723 @node Switches for gnatls
16724 @section Switches for @code{gnatls}
16727 @code{gnatls} recognizes the following switches:
16731 @item ^-a^/ALL_UNITS^
16732 @cindex @option{^-a^/ALL_UNITS^} (@code{gnatls})
16733 Consider all units, including those of the predefined Ada library.
16734 Especially useful with @option{^-d^/DEPENDENCIES^}.
16736 @item ^-d^/DEPENDENCIES^
16737 @cindex @option{^-d^/DEPENDENCIES^} (@code{gnatls})
16738 List sources from which specified units depend on.
16740 @item ^-h^/OUTPUT=OPTIONS^
16741 @cindex @option{^-h^/OUTPUT=OPTIONS^} (@code{gnatls})
16742 Output the list of options.
16744 @item ^-o^/OUTPUT=OBJECTS^
16745 @cindex @option{^-o^/OUTPUT=OBJECTS^} (@code{gnatls})
16746 Only output information about object files.
16748 @item ^-s^/OUTPUT=SOURCES^
16749 @cindex @option{^-s^/OUTPUT=SOURCES^} (@code{gnatls})
16750 Only output information about source files.
16752 @item ^-u^/OUTPUT=UNITS^
16753 @cindex @option{^-u^/OUTPUT=UNITS^} (@code{gnatls})
16754 Only output information about compilation units.
16756 @item ^-files^/FILES^=@var{file}
16757 @cindex @option{^-files^/FILES^} (@code{gnatls})
16758 Take as arguments the files listed in text file @var{file}.
16759 Text file @var{file} may contain empty lines that are ignored.
16760 Each non empty line should contain the name of an existing file.
16761 Several such switches may be specified simultaneously.
16763 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
16764 @itemx ^-aI^/SOURCE_SEARCH=^@var{dir}
16765 @itemx ^-I^/SEARCH=^@var{dir}
16766 @itemx ^-I-^/NOCURRENT_DIRECTORY^
16768 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatls})
16769 @cindex @option{^-aI^/SOURCE_SEARCH^} (@code{gnatls})
16770 @cindex @option{^-I^/SEARCH^} (@code{gnatls})
16771 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatls})
16772 Source path manipulation. Same meaning as the equivalent @command{gnatmake}
16773 flags (@pxref{Switches for gnatmake}).
16775 @item --RTS=@var{rts-path}
16776 @cindex @option{--RTS} (@code{gnatls})
16777 Specifies the default location of the runtime library. Same meaning as the
16778 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
16780 @item ^-v^/OUTPUT=VERBOSE^
16781 @cindex @option{^-v^/OUTPUT=VERBOSE^} (@code{gnatls})
16782 Verbose mode. Output the complete source, object and project paths. Do not use
16783 the default column layout but instead use long format giving as much as
16784 information possible on each requested units, including special
16785 characteristics such as:
16788 @item Preelaborable
16789 The unit is preelaborable in the Ada 95 sense.
16792 No elaboration code has been produced by the compiler for this unit.
16795 The unit is pure in the Ada 95 sense.
16797 @item Elaborate_Body
16798 The unit contains a pragma Elaborate_Body.
16801 The unit contains a pragma Remote_Types.
16803 @item Shared_Passive
16804 The unit contains a pragma Shared_Passive.
16807 This unit is part of the predefined environment and cannot be modified
16810 @item Remote_Call_Interface
16811 The unit contains a pragma Remote_Call_Interface.
16817 @node Examples of gnatls Usage
16818 @section Example of @code{gnatls} Usage
16822 Example of using the verbose switch. Note how the source and
16823 object paths are affected by the -I switch.
16826 $ gnatls -v -I.. demo1.o
16828 GNATLS 5.03w (20041123-34)
16829 Copyright 1997-2004 Free Software Foundation, Inc.
16831 Source Search Path:
16832 <Current_Directory>
16834 /home/comar/local/adainclude/
16836 Object Search Path:
16837 <Current_Directory>
16839 /home/comar/local/lib/gcc-lib/x86-linux/3.4.3/adalib/
16841 Project Search Path:
16842 <Current_Directory>
16843 /home/comar/local/lib/gnat/
16848 Kind => subprogram body
16849 Flags => No_Elab_Code
16850 Source => demo1.adb modified
16854 The following is an example of use of the dependency list.
16855 Note the use of the -s switch
16856 which gives a straight list of source files. This can be useful for
16857 building specialized scripts.
16860 $ gnatls -d demo2.o
16861 ./demo2.o demo2 OK demo2.adb
16867 $ gnatls -d -s -a demo1.o
16869 /home/comar/local/adainclude/ada.ads
16870 /home/comar/local/adainclude/a-finali.ads
16871 /home/comar/local/adainclude/a-filico.ads
16872 /home/comar/local/adainclude/a-stream.ads
16873 /home/comar/local/adainclude/a-tags.ads
16876 /home/comar/local/adainclude/gnat.ads
16877 /home/comar/local/adainclude/g-io.ads
16879 /home/comar/local/adainclude/system.ads
16880 /home/comar/local/adainclude/s-exctab.ads
16881 /home/comar/local/adainclude/s-finimp.ads
16882 /home/comar/local/adainclude/s-finroo.ads
16883 /home/comar/local/adainclude/s-secsta.ads
16884 /home/comar/local/adainclude/s-stalib.ads
16885 /home/comar/local/adainclude/s-stoele.ads
16886 /home/comar/local/adainclude/s-stratt.ads
16887 /home/comar/local/adainclude/s-tasoli.ads
16888 /home/comar/local/adainclude/s-unstyp.ads
16889 /home/comar/local/adainclude/unchconv.ads
16895 GNAT LIST /DEPENDENCIES /OUTPUT=SOURCES /ALL_UNITS DEMO1.ADB
16897 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]ada.ads
16898 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-finali.ads
16899 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-filico.ads
16900 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-stream.ads
16901 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-tags.ads
16905 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]gnat.ads
16906 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]g-io.ads
16908 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]system.ads
16909 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-exctab.ads
16910 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finimp.ads
16911 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finroo.ads
16912 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-secsta.ads
16913 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stalib.ads
16914 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stoele.ads
16915 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stratt.ads
16916 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-tasoli.ads
16917 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-unstyp.ads
16918 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]unchconv.ads
16922 @node Cleaning Up Using gnatclean
16923 @chapter Cleaning Up Using @code{gnatclean}
16925 @cindex Cleaning tool
16928 @code{gnatclean} is a tool that allows the deletion of files produced by the
16929 compiler, binder and linker, including ALI files, object files, tree files,
16930 expanded source files, library files, interface copy source files, binder
16931 generated files and executable files.
16934 * Running gnatclean::
16935 * Switches for gnatclean::
16936 @c * Examples of gnatclean Usage::
16939 @node Running gnatclean
16940 @section Running @code{gnatclean}
16943 The @code{gnatclean} command has the form:
16946 $ gnatclean switches @var{names}
16950 @var{names} is a list of source file names. Suffixes @code{.^ads^ADS^} and
16951 @code{^adb^ADB^} may be omitted. If a project file is specified using switch
16952 @code{^-P^/PROJECT_FILE=^}, then @var{names} may be completely omitted.
16955 In normal mode, @code{gnatclean} delete the files produced by the compiler and,
16956 if switch @code{^-c^/COMPILER_FILES_ONLY^} is not specified, by the binder and
16957 the linker. In informative-only mode, specified by switch
16958 @code{^-n^/NODELETE^}, the list of files that would have been deleted in
16959 normal mode is listed, but no file is actually deleted.
16961 @node Switches for gnatclean
16962 @section Switches for @code{gnatclean}
16965 @code{gnatclean} recognizes the following switches:
16969 @item ^-c^/COMPILER_FILES_ONLY^
16970 @cindex @option{^-c^/COMPILER_FILES_ONLY^} (@code{gnatclean})
16971 Only attempt to delete the files produced by the compiler, not those produced
16972 by the binder or the linker. The files that are not to be deleted are library
16973 files, interface copy files, binder generated files and executable files.
16975 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
16976 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@code{gnatclean})
16977 Indicate that ALI and object files should normally be found in directory
16980 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
16981 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@code{gnatclean})
16982 When using project files, if some errors or warnings are detected during
16983 parsing and verbose mode is not in effect (no use of switch
16984 ^-v^/VERBOSE^), then error lines start with the full path name of the project
16985 file, rather than its simple file name.
16988 @cindex @option{^-h^/HELP^} (@code{gnatclean})
16989 Output a message explaining the usage of @code{^gnatclean^gnatclean^}.
16991 @item ^-n^/NODELETE^
16992 @cindex @option{^-n^/NODELETE^} (@code{gnatclean})
16993 Informative-only mode. Do not delete any files. Output the list of the files
16994 that would have been deleted if this switch was not specified.
16996 @item ^-P^/PROJECT_FILE=^@var{project}
16997 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatclean})
16998 Use project file @var{project}. Only one such switch can be used.
16999 When cleaning a project file, the files produced by the compilation of the
17000 immediate sources or inherited sources of the project files are to be
17001 deleted. This is not depending on the presence or not of executable names
17002 on the command line.
17005 @cindex @option{^-q^/QUIET^} (@code{gnatclean})
17006 Quiet output. If there are no error, do not ouuput anything, except in
17007 verbose mode (switch ^-v^/VERBOSE^) or in informative-only mode
17008 (switch ^-n^/NODELETE^).
17010 @item ^-r^/RECURSIVE^
17011 @cindex @option{^-r^/RECURSIVE^} (@code{gnatclean})
17012 When a project file is specified (using switch ^-P^/PROJECT_FILE=^),
17013 clean all imported and extended project files, recursively. If this switch
17014 is not specified, only the files related to the main project file are to be
17015 deleted. This switch has no effect if no project file is specified.
17017 @item ^-v^/VERBOSE^
17018 @cindex @option{^-v^/VERBOSE^} (@code{gnatclean})
17021 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
17022 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (@code{gnatclean})
17023 Indicates the verbosity of the parsing of GNAT project files.
17024 @xref{Switches Related to Project Files}.
17026 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
17027 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (@code{gnatclean})
17028 Indicates that external variable @var{name} has the value @var{value}.
17029 The Project Manager will use this value for occurrences of
17030 @code{external(name)} when parsing the project file.
17031 @xref{Switches Related to Project Files}.
17033 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
17034 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatclean})
17035 When searching for ALI and object files, look in directory
17038 @item ^-I^/SEARCH=^@var{dir}
17039 @cindex @option{^-I^/SEARCH^} (@code{gnatclean})
17040 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}}.
17042 @item ^-I-^/NOCURRENT_DIRECTORY^
17043 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatclean})
17044 @cindex Source files, suppressing search
17045 Do not look for ALI or object files in the directory
17046 where @code{gnatclean} was invoked.
17050 @c @node Examples of gnatclean Usage
17051 @c @section Examples of @code{gnatclean} Usage
17054 @node GNAT and Libraries
17055 @chapter GNAT and Libraries
17056 @cindex Library, building, installing, using
17059 This chapter describes how to build and use libraries with GNAT, and also shows
17060 how to recompile the GNAT run-time library. You should be familiar with the
17061 Project Manager facility (@pxref{GNAT Project Manager}) before reading this
17065 * Introduction to Libraries in GNAT::
17066 * General Ada Libraries::
17067 * Stand-alone Ada Libraries::
17068 * Rebuilding the GNAT Run-Time Library::
17071 @node Introduction to Libraries in GNAT
17072 @section Introduction to Libraries in GNAT
17075 A library is, conceptually, a collection of objects which does not have its
17076 own main thread of execution, but rather provides certain services to the
17077 applications that use it. A library can be either statically linked with the
17078 application, in which case its code is directly included in the application,
17079 or, on platforms that support it, be dynamically linked, in which case
17080 its code is shared by all applications making use of this library.
17082 GNAT supports both types of libraries.
17083 In the static case, the compiled code can be provided in different ways. The
17084 simplest approach is to provide directly the set of objects resulting from
17085 compilation of the library source files. Alternatively, you can group the
17086 objects into an archive using whatever commands are provided by the operating
17087 system. For the latter case, the objects are grouped into a shared library.
17089 In the GNAT environment, a library has three types of components:
17095 @xref{The Ada Library Information Files}.
17097 Object files, an archive or a shared library.
17101 A GNAT library may expose all its source files, which is useful for
17102 documentation purposes. Alternatively, it may expose only the units needed by
17103 an external user to make use of the library. That is to say, the specs
17104 reflecting the library services along with all the units needed to compile
17105 those specs, which can include generic bodies or any body implementing an
17106 inlined routine. In the case of @emph{stand-alone libraries} those exposed
17107 units are called @emph{interface units} (@pxref{Stand-alone Ada Libraries}).
17109 All compilation units comprising an application, including those in a library,
17110 need to be elaborated in an order partially defined by Ada's semantics. GNAT
17111 computes the elaboration order from the @file{ALI} files and this is why they
17112 constitute a mandatory part of GNAT libraries. Except in the case of
17113 @emph{stand-alone libraries}, where a specific library elaboration routine is
17114 produced independently of the application(s) using the library.
17116 @node General Ada Libraries
17117 @section General Ada Libraries
17120 * Building a library::
17121 * Installing a library::
17122 * Using a library::
17125 @node Building a library
17126 @subsection Building a library
17129 The easiest way to build a library is to use the Project Manager,
17130 which supports a special type of project called a @emph{Library Project}
17131 (@pxref{Library Projects}).
17133 A project is considered a library project, when two project-level attributes
17134 are defined in it: @code{Library_Name} and @code{Library_Dir}. In order to
17135 control different aspects of library configuration, additional optional
17136 project-level attributes can be specified:
17139 This attribute controls whether the library is to be static or dynamic
17141 @item Library_Version
17142 This attribute specifies the library version; this value is used
17143 during dynamic linking of shared libraries to determine if the currently
17144 installed versions of the binaries are compatible.
17146 @item Library_Options
17148 These attributes specify additional low-level options to be used during
17149 library generation, and redefine the actual application used to generate
17154 The GNAT Project Manager takes full care of the library maintenance task,
17155 including recompilation of the source files for which objects do not exist
17156 or are not up to date, assembly of the library archive, and installation of
17157 the library (i.e., copying associated source, object and @file{ALI} files
17158 to the specified location).
17160 Here is a simple library project file:
17161 @smallexample @c ada
17163 for Source_Dirs use ("src1", "src2");
17164 for Object_Dir use "obj";
17165 for Library_Name use "mylib";
17166 for Library_Dir use "lib";
17167 for Library_Kind use "dynamic";
17172 and the compilation command to build and install the library:
17174 @smallexample @c ada
17175 $ gnatmake -Pmy_lib
17179 It is not entirely trivial to perform manually all the steps required to
17180 produce a library. We recommend that you use the GNAT Project Manager
17181 for this task. In special cases where this is not desired, the necessary
17182 steps are discussed below.
17184 There are various possibilities for compiling the units that make up the
17185 library: for example with a Makefile (@pxref{Using the GNU make Utility}) or
17186 with a conventional script. For simple libraries, it is also possible to create
17187 a dummy main program which depends upon all the packages that comprise the
17188 interface of the library. This dummy main program can then be given to
17189 @command{gnatmake}, which will ensure that all necessary objects are built.
17191 After this task is accomplished, you should follow the standard procedure
17192 of the underlying operating system to produce the static or shared library.
17194 Here is an example of such a dummy program:
17195 @smallexample @c ada
17197 with My_Lib.Service1;
17198 with My_Lib.Service2;
17199 with My_Lib.Service3;
17200 procedure My_Lib_Dummy is
17208 Here are the generic commands that will build an archive or a shared library.
17211 # compiling the library
17212 $ gnatmake -c my_lib_dummy.adb
17214 # we don't need the dummy object itself
17215 $ rm my_lib_dummy.o my_lib_dummy.ali
17217 # create an archive with the remaining objects
17218 $ ar rc libmy_lib.a *.o
17219 # some systems may require "ranlib" to be run as well
17221 # or create a shared library
17222 $ gcc -shared -o libmy_lib.so *.o
17223 # some systems may require the code to have been compiled with -fPIC
17225 # remove the object files that are now in the library
17228 # Make the ALI files read-only so that gnatmake will not try to
17229 # regenerate the objects that are in the library
17234 Please note that the library must have a name of the form @file{libxxx.a} or
17235 @file{libxxx.so} (or @file{libxxx.dll} on Windows) in order to be accessed by
17236 the directive @option{-lxxx} at link time.
17238 @node Installing a library
17239 @subsection Installing a library
17240 @cindex @code{ADA_PROJECT_PATH}
17243 If you use project files, library installation is part of the library build
17244 process. Thus no further action is needed in order to make use of the
17245 libraries that are built as part of the general application build. A usable
17246 version of the library is installed in the directory specified by the
17247 @code{Library_Dir} attribute of the library project file.
17249 You may want to install a library in a context different from where the library
17250 is built. This situation arises with third party suppliers, who may want
17251 to distribute a library in binary form where the user is not expected to be
17252 able to recompile the library. The simplest option in this case is to provide
17253 a project file slightly different from the one used to build the library, by
17254 using the @code{externally_built} attribute. For instance, the project
17255 file used to build the library in the previous section can be changed into the
17256 following one when the library is installed:
17258 @smallexample @c projectfile
17260 for Source_Dirs use ("src1", "src2");
17261 for Library_Name use "mylib";
17262 for Library_Dir use "lib";
17263 for Library_Kind use "dynamic";
17264 for Externally_Built use "true";
17269 This project file assumes that the directories @file{src1},
17270 @file{src2}, and @file{lib} exist in
17271 the directory containing the project file. The @code{externally_built}
17272 attribute makes it clear to the GNAT builder that it should not attempt to
17273 recompile any of the units from this library. It allows the library provider to
17274 restrict the source set to the minimum necessary for clients to make use of the
17275 library as described in the first section of this chapter. It is the
17276 responsibility of the library provider to install the necessary sources, ALI
17277 files and libraries in the directories mentioned in the project file. For
17278 convenience, the user's library project file should be installed in a location
17279 that will be searched automatically by the GNAT
17280 builder. These are the directories referenced in the @code{ADA_PROJECT_PATH}
17281 environment variable (@pxref{Importing Projects}), and also the default GNAT
17282 library location that can be queried with @command{gnatls -v} and is usually of
17283 the form $gnat_install_root/lib/gnat.
17285 When project files are not an option, it is also possible, but not recommended,
17286 to install the library so that the sources needed to use the library are on the
17287 Ada source path and the ALI files & libraries be on the Ada Object path (see
17288 @ref{Search Paths and the Run-Time Library (RTL)}. Alternatively, the system
17289 administrator can place general-purpose libraries in the default compiler
17290 paths, by specifying the libraries' location in the configuration files
17291 @file{ada_source_path} and @file{ada_object_path}. These configuration files
17292 must be located in the GNAT installation tree at the same place as the gcc spec
17293 file. The location of the gcc spec file can be determined as follows:
17299 The configuration files mentioned above have a simple format: each line
17300 must contain one unique directory name.
17301 Those names are added to the corresponding path
17302 in their order of appearance in the file. The names can be either absolute
17303 or relative; in the latter case, they are relative to where theses files
17306 The files @file{ada_source_path} and @file{ada_object_path} might not be
17308 GNAT installation, in which case, GNAT will look for its run-time library in
17309 the directories @file{adainclude} (for the sources) and @file{adalib} (for the
17310 objects and @file{ALI} files). When the files exist, the compiler does not
17311 look in @file{adainclude} and @file{adalib}, and thus the
17312 @file{ada_source_path} file
17313 must contain the location for the GNAT run-time sources (which can simply
17314 be @file{adainclude}). In the same way, the @file{ada_object_path} file must
17315 contain the location for the GNAT run-time objects (which can simply
17318 You can also specify a new default path to the run-time library at compilation
17319 time with the switch @option{--RTS=rts-path}. You can thus choose / change
17320 the run-time library you want your program to be compiled with. This switch is
17321 recognized by @command{gcc}, @command{gnatmake}, @command{gnatbind},
17322 @command{gnatls}, @command{gnatfind} and @command{gnatxref}.
17324 It is possible to install a library before or after the standard GNAT
17325 library, by reordering the lines in the configuration files. In general, a
17326 library must be installed before the GNAT library if it redefines
17329 @node Using a library
17330 @subsection Using a library
17332 @noindent Once again, the project facility greatly simplifies the use of
17333 libraries. In this context, using a library is just a matter of adding a
17334 @code{with} clause in the user project. For instance, to make use of the
17335 library @code{My_Lib} shown in examples in earlier sections, you can
17338 @smallexample @c projectfile
17345 Even if you have a third-party, non-Ada library, you can still use GNAT's
17346 Project Manager facility to provide a wrapper for it. For example, the
17347 following project, when @code{with}ed by your main project, will link with the
17348 third-party library @file{liba.a}:
17350 @smallexample @c projectfile
17353 for Externally_Built use "true";
17354 for Library_Dir use "lib";
17355 for Library_Name use "a";
17356 for Library_Kind use "static";
17360 This is an alternative to the use of @code{pragma Linker_Options}. It is
17361 especially interesting in the context of systems with several interdependant
17362 static libraries where finding a proper linker order is not easy and best be
17363 left to the tools having visibility over project dependancy information.
17366 In order to use an Ada library manually, you need to make sure that this
17367 library is on both your source and object path
17368 (see @ref{Search Paths and the Run-Time Library (RTL)}
17369 and @ref{Search Paths for gnatbind}). Furthermore, when the objects are grouped
17370 in an archive or a shared library, you need to specify the desired
17371 library at link time.
17373 For example, you can use the library @file{mylib} installed in
17374 @file{/dir/my_lib_src} and @file{/dir/my_lib_obj} with the following commands:
17377 $ gnatmake -aI/dir/my_lib_src -aO/dir/my_lib_obj my_appl \
17382 This can be expressed more simply:
17387 when the following conditions are met:
17390 @file{/dir/my_lib_src} has been added by the user to the environment
17391 variable @code{ADA_INCLUDE_PATH}, or by the administrator to the file
17392 @file{ada_source_path}
17394 @file{/dir/my_lib_obj} has been added by the user to the environment
17395 variable @code{ADA_OBJECTS_PATH}, or by the administrator to the file
17396 @file{ada_object_path}
17398 a pragma @code{Linker_Options} has been added to one of the sources.
17401 @smallexample @c ada
17402 pragma Linker_Options ("-lmy_lib");
17406 @node Stand-alone Ada Libraries
17407 @section Stand-alone Ada Libraries
17408 @cindex Stand-alone library, building, using
17411 * Introduction to Stand-alone Libraries::
17412 * Building a Stand-alone Library::
17413 * Creating a Stand-alone Library to be used in a non-Ada context::
17414 * Restrictions in Stand-alone Libraries::
17417 @node Introduction to Stand-alone Libraries
17418 @subsection Introduction to Stand-alone Libraries
17421 A Stand-alone Library (abbreviated ``SAL'') is a library that contains the
17423 elaborate the Ada units that are included in the library. In contrast with
17424 an ordinary library, which consists of all sources, objects and @file{ALI}
17426 library, a SAL may specify a restricted subset of compilation units
17427 to serve as a library interface. In this case, the fully
17428 self-sufficient set of files will normally consist of an objects
17429 archive, the sources of interface units' specs, and the @file{ALI}
17430 files of interface units.
17431 If an interface spec contains a generic unit or an inlined subprogram,
17433 source must also be provided; if the units that must be provided in the source
17434 form depend on other units, the source and @file{ALI} files of those must
17437 The main purpose of a SAL is to minimize the recompilation overhead of client
17438 applications when a new version of the library is installed. Specifically,
17439 if the interface sources have not changed, client applications do not need to
17440 be recompiled. If, furthermore, a SAL is provided in the shared form and its
17441 version, controlled by @code{Library_Version} attribute, is not changed,
17442 then the clients do not need to be relinked.
17444 SALs also allow the library providers to minimize the amount of library source
17445 text exposed to the clients. Such ``information hiding'' might be useful or
17446 necessary for various reasons.
17448 Stand-alone libraries are also well suited to be used in an executable whose
17449 main routine is not written in Ada.
17451 @node Building a Stand-alone Library
17452 @subsection Building a Stand-alone Library
17455 GNAT's Project facility provides a simple way of building and installing
17456 stand-alone libraries; see @ref{Stand-alone Library Projects}.
17457 To be a Stand-alone Library Project, in addition to the two attributes
17458 that make a project a Library Project (@code{Library_Name} and
17459 @code{Library_Dir}; see @ref{Library Projects}), the attribute
17460 @code{Library_Interface} must be defined. For example:
17462 @smallexample @c projectfile
17464 for Library_Dir use "lib_dir";
17465 for Library_Name use "dummy";
17466 for Library_Interface use ("int1", "int1.child");
17471 Attribute @code{Library_Interface} has a non-empty string list value,
17472 each string in the list designating a unit contained in an immediate source
17473 of the project file.
17475 When a Stand-alone Library is built, first the binder is invoked to build
17476 a package whose name depends on the library name
17477 (@file{^b~dummy.ads/b^B$DUMMY.ADS/B^} in the example above).
17478 This binder-generated package includes initialization and
17479 finalization procedures whose
17480 names depend on the library name (@code{dummyinit} and @code{dummyfinal}
17482 above). The object corresponding to this package is included in the library.
17484 You must ensure timely (e.g., prior to any use of interfaces in the SAL)
17485 calling of these procedures if a static SAL is built, or if a shared SAL
17487 with the project-level attribute @code{Library_Auto_Init} set to
17490 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
17491 (those that are listed in attribute @code{Library_Interface}) are copied to
17492 the Library Directory. As a consequence, only the Interface Units may be
17493 imported from Ada units outside of the library. If other units are imported,
17494 the binding phase will fail.
17496 The attribute @code{Library_Src_Dir} may be specified for a
17497 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
17498 single string value. Its value must be the path (absolute or relative to the
17499 project directory) of an existing directory. This directory cannot be the
17500 object directory or one of the source directories, but it can be the same as
17501 the library directory. The sources of the Interface
17502 Units of the library that are needed by an Ada client of the library will be
17503 copied to the designated directory, called the Interface Copy directory.
17504 These sources include the specs of the Interface Units, but they may also
17505 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
17506 are used, or when there is a generic unit in the spec. Before the sources
17507 are copied to the Interface Copy directory, an attempt is made to delete all
17508 files in the Interface Copy directory.
17510 Building stand-alone libraries by hand is somewhat tedious, but for those
17511 occasions when it is necessary here are the steps that you need to perform:
17514 Compile all library sources.
17517 Invoke the binder with the switch @option{-n} (No Ada main program),
17518 with all the @file{ALI} files of the interfaces, and
17519 with the switch @option{-L} to give specific names to the @code{init}
17520 and @code{final} procedures. For example:
17522 gnatbind -n int1.ali int2.ali -Lsal1
17526 Compile the binder generated file:
17532 Link the dynamic library with all the necessary object files,
17533 indicating to the linker the names of the @code{init} (and possibly
17534 @code{final}) procedures for automatic initialization (and finalization).
17535 The built library should be placed in a directory different from
17536 the object directory.
17539 Copy the @code{ALI} files of the interface to the library directory,
17540 add in this copy an indication that it is an interface to a SAL
17541 (i.e. add a word @option{SL} on the line in the @file{ALI} file that starts
17542 with letter ``P'') and make the modified copy of the @file{ALI} file
17547 Using SALs is not different from using other libraries
17548 (see @ref{Using a library}).
17550 @node Creating a Stand-alone Library to be used in a non-Ada context
17551 @subsection Creating a Stand-alone Library to be used in a non-Ada context
17554 It is easy to adapt the SAL build procedure discussed above for use of a SAL in
17557 The only extra step required is to ensure that library interface subprograms
17558 are compatible with the main program, by means of @code{pragma Export}
17559 or @code{pragma Convention}.
17561 Here is an example of simple library interface for use with C main program:
17563 @smallexample @c ada
17564 package Interface is
17566 procedure Do_Something;
17567 pragma Export (C, Do_Something, "do_something");
17569 procedure Do_Something_Else;
17570 pragma Export (C, Do_Something_Else, "do_something_else");
17576 On the foreign language side, you must provide a ``foreign'' view of the
17577 library interface; remember that it should contain elaboration routines in
17578 addition to interface subprograms.
17580 The example below shows the content of @code{mylib_interface.h} (note
17581 that there is no rule for the naming of this file, any name can be used)
17583 /* the library elaboration procedure */
17584 extern void mylibinit (void);
17586 /* the library finalization procedure */
17587 extern void mylibfinal (void);
17589 /* the interface exported by the library */
17590 extern void do_something (void);
17591 extern void do_something_else (void);
17595 Libraries built as explained above can be used from any program, provided
17596 that the elaboration procedures (named @code{mylibinit} in the previous
17597 example) are called before the library services are used. Any number of
17598 libraries can be used simultaneously, as long as the elaboration
17599 procedure of each library is called.
17601 Below is an example of a C program that uses the @code{mylib} library.
17604 #include "mylib_interface.h"
17609 /* First, elaborate the library before using it */
17612 /* Main program, using the library exported entities */
17614 do_something_else ();
17616 /* Library finalization at the end of the program */
17623 Note that invoking any library finalization procedure generated by
17624 @code{gnatbind} shuts down the Ada run-time environment.
17626 finalization of all Ada libraries must be performed at the end of the program.
17627 No call to these libraries or to the Ada run-time library should be made
17628 after the finalization phase.
17630 @node Restrictions in Stand-alone Libraries
17631 @subsection Restrictions in Stand-alone Libraries
17634 The pragmas listed below should be used with caution inside libraries,
17635 as they can create incompatibilities with other Ada libraries:
17637 @item pragma @code{Locking_Policy}
17638 @item pragma @code{Queuing_Policy}
17639 @item pragma @code{Task_Dispatching_Policy}
17640 @item pragma @code{Unreserve_All_Interrupts}
17644 When using a library that contains such pragmas, the user must make sure
17645 that all libraries use the same pragmas with the same values. Otherwise,
17646 @code{Program_Error} will
17647 be raised during the elaboration of the conflicting
17648 libraries. The usage of these pragmas and its consequences for the user
17649 should therefore be well documented.
17651 Similarly, the traceback in the exception occurrence mechanism should be
17652 enabled or disabled in a consistent manner across all libraries.
17653 Otherwise, Program_Error will be raised during the elaboration of the
17654 conflicting libraries.
17656 If the @code{Version} or @code{Body_Version}
17657 attributes are used inside a library, then you need to
17658 perform a @code{gnatbind} step that specifies all @file{ALI} files in all
17659 libraries, so that version identifiers can be properly computed.
17660 In practice these attributes are rarely used, so this is unlikely
17661 to be a consideration.
17663 @node Rebuilding the GNAT Run-Time Library
17664 @section Rebuilding the GNAT Run-Time Library
17665 @cindex GNAT Run-Time Library, rebuilding
17668 It may be useful to recompile the GNAT library in various contexts, the
17669 most important one being the use of partition-wide configuration pragmas
17670 such as @code{Normalize_Scalars}. A special Makefile called
17671 @code{Makefile.adalib} is provided to that effect and can be found in
17672 the directory containing the GNAT library. The location of this
17673 directory depends on the way the GNAT environment has been installed and can
17674 be determined by means of the command:
17681 The last entry in the object search path usually contains the
17682 gnat library. This Makefile contains its own documentation and in
17683 particular the set of instructions needed to rebuild a new library and
17686 @node Using the GNU make Utility
17687 @chapter Using the GNU @code{make} Utility
17691 This chapter offers some examples of makefiles that solve specific
17692 problems. It does not explain how to write a makefile (see the GNU make
17693 documentation), nor does it try to replace the @command{gnatmake} utility
17694 (@pxref{The GNAT Make Program gnatmake}).
17696 All the examples in this section are specific to the GNU version of
17697 make. Although @code{make} is a standard utility, and the basic language
17698 is the same, these examples use some advanced features found only in
17702 * Using gnatmake in a Makefile::
17703 * Automatically Creating a List of Directories::
17704 * Generating the Command Line Switches::
17705 * Overcoming Command Line Length Limits::
17708 @node Using gnatmake in a Makefile
17709 @section Using gnatmake in a Makefile
17714 Complex project organizations can be handled in a very powerful way by
17715 using GNU make combined with gnatmake. For instance, here is a Makefile
17716 which allows you to build each subsystem of a big project into a separate
17717 shared library. Such a makefile allows you to significantly reduce the link
17718 time of very big applications while maintaining full coherence at
17719 each step of the build process.
17721 The list of dependencies are handled automatically by
17722 @command{gnatmake}. The Makefile is simply used to call gnatmake in each of
17723 the appropriate directories.
17725 Note that you should also read the example on how to automatically
17726 create the list of directories
17727 (@pxref{Automatically Creating a List of Directories})
17728 which might help you in case your project has a lot of subdirectories.
17733 @font@heightrm=cmr8
17736 ## This Makefile is intended to be used with the following directory
17738 ## - The sources are split into a series of csc (computer software components)
17739 ## Each of these csc is put in its own directory.
17740 ## Their name are referenced by the directory names.
17741 ## They will be compiled into shared library (although this would also work
17742 ## with static libraries
17743 ## - The main program (and possibly other packages that do not belong to any
17744 ## csc is put in the top level directory (where the Makefile is).
17745 ## toplevel_dir __ first_csc (sources) __ lib (will contain the library)
17746 ## \_ second_csc (sources) __ lib (will contain the library)
17748 ## Although this Makefile is build for shared library, it is easy to modify
17749 ## to build partial link objects instead (modify the lines with -shared and
17752 ## With this makefile, you can change any file in the system or add any new
17753 ## file, and everything will be recompiled correctly (only the relevant shared
17754 ## objects will be recompiled, and the main program will be re-linked).
17756 # The list of computer software component for your project. This might be
17757 # generated automatically.
17760 # Name of the main program (no extension)
17763 # If we need to build objects with -fPIC, uncomment the following line
17766 # The following variable should give the directory containing libgnat.so
17767 # You can get this directory through 'gnatls -v'. This is usually the last
17768 # directory in the Object_Path.
17771 # The directories for the libraries
17772 # (This macro expands the list of CSC to the list of shared libraries, you
17773 # could simply use the expanded form :
17774 # LIB_DIR=aa/lib/libaa.so bb/lib/libbb.so cc/lib/libcc.so
17775 LIB_DIR=$@{foreach dir,$@{CSC_LIST@},$@{dir@}/lib/lib$@{dir@}.so@}
17777 $@{MAIN@}: objects $@{LIB_DIR@}
17778 gnatbind $@{MAIN@} $@{CSC_LIST:%=-aO%/lib@} -shared
17779 gnatlink $@{MAIN@} $@{CSC_LIST:%=-l%@}
17782 # recompile the sources
17783 gnatmake -c -i $@{MAIN@}.adb $@{NEED_FPIC@} $@{CSC_LIST:%=-I%@}
17785 # Note: In a future version of GNAT, the following commands will be simplified
17786 # by a new tool, gnatmlib
17788 mkdir -p $@{dir $@@ @}
17789 cd $@{dir $@@ @}; gcc -shared -o $@{notdir $@@ @} ../*.o -L$@{GLIB@} -lgnat
17790 cd $@{dir $@@ @}; cp -f ../*.ali .
17792 # The dependencies for the modules
17793 # Note that we have to force the expansion of *.o, since in some cases
17794 # make won't be able to do it itself.
17795 aa/lib/libaa.so: $@{wildcard aa/*.o@}
17796 bb/lib/libbb.so: $@{wildcard bb/*.o@}
17797 cc/lib/libcc.so: $@{wildcard cc/*.o@}
17799 # Make sure all of the shared libraries are in the path before starting the
17802 LD_LIBRARY_PATH=`pwd`/aa/lib:`pwd`/bb/lib:`pwd`/cc/lib ./$@{MAIN@}
17805 $@{RM@} -rf $@{CSC_LIST:%=%/lib@}
17806 $@{RM@} $@{CSC_LIST:%=%/*.ali@}
17807 $@{RM@} $@{CSC_LIST:%=%/*.o@}
17808 $@{RM@} *.o *.ali $@{MAIN@}
17811 @node Automatically Creating a List of Directories
17812 @section Automatically Creating a List of Directories
17815 In most makefiles, you will have to specify a list of directories, and
17816 store it in a variable. For small projects, it is often easier to
17817 specify each of them by hand, since you then have full control over what
17818 is the proper order for these directories, which ones should be
17821 However, in larger projects, which might involve hundreds of
17822 subdirectories, it might be more convenient to generate this list
17825 The example below presents two methods. The first one, although less
17826 general, gives you more control over the list. It involves wildcard
17827 characters, that are automatically expanded by @code{make}. Its
17828 shortcoming is that you need to explicitly specify some of the
17829 organization of your project, such as for instance the directory tree
17830 depth, whether some directories are found in a separate tree,...
17832 The second method is the most general one. It requires an external
17833 program, called @code{find}, which is standard on all Unix systems. All
17834 the directories found under a given root directory will be added to the
17840 @font@heightrm=cmr8
17843 # The examples below are based on the following directory hierarchy:
17844 # All the directories can contain any number of files
17845 # ROOT_DIRECTORY -> a -> aa -> aaa
17848 # -> b -> ba -> baa
17851 # This Makefile creates a variable called DIRS, that can be reused any time
17852 # you need this list (see the other examples in this section)
17854 # The root of your project's directory hierarchy
17858 # First method: specify explicitly the list of directories
17859 # This allows you to specify any subset of all the directories you need.
17862 DIRS := a/aa/ a/ab/ b/ba/
17865 # Second method: use wildcards
17866 # Note that the argument(s) to wildcard below should end with a '/'.
17867 # Since wildcards also return file names, we have to filter them out
17868 # to avoid duplicate directory names.
17869 # We thus use make's @code{dir} and @code{sort} functions.
17870 # It sets DIRs to the following value (note that the directories aaa and baa
17871 # are not given, unless you change the arguments to wildcard).
17872 # DIRS= ./a/a/ ./b/ ./a/aa/ ./a/ab/ ./a/ac/ ./b/ba/ ./b/bb/ ./b/bc/
17875 DIRS := $@{sort $@{dir $@{wildcard $@{ROOT_DIRECTORY@}/*/
17876 $@{ROOT_DIRECTORY@}/*/*/@}@}@}
17879 # Third method: use an external program
17880 # This command is much faster if run on local disks, avoiding NFS slowdowns.
17881 # This is the most complete command: it sets DIRs to the following value:
17882 # DIRS= ./a ./a/aa ./a/aa/aaa ./a/ab ./a/ac ./b ./b/ba ./b/ba/baa ./b/bb ./b/bc
17885 DIRS := $@{shell find $@{ROOT_DIRECTORY@} -type d -print@}
17889 @node Generating the Command Line Switches
17890 @section Generating the Command Line Switches
17893 Once you have created the list of directories as explained in the
17894 previous section (@pxref{Automatically Creating a List of Directories}),
17895 you can easily generate the command line arguments to pass to gnatmake.
17897 For the sake of completeness, this example assumes that the source path
17898 is not the same as the object path, and that you have two separate lists
17902 # see "Automatically creating a list of directories" to create
17907 GNATMAKE_SWITCHES := $@{patsubst %,-aI%,$@{SOURCE_DIRS@}@}
17908 GNATMAKE_SWITCHES += $@{patsubst %,-aO%,$@{OBJECT_DIRS@}@}
17911 gnatmake $@{GNATMAKE_SWITCHES@} main_unit
17914 @node Overcoming Command Line Length Limits
17915 @section Overcoming Command Line Length Limits
17918 One problem that might be encountered on big projects is that many
17919 operating systems limit the length of the command line. It is thus hard to give
17920 gnatmake the list of source and object directories.
17922 This example shows how you can set up environment variables, which will
17923 make @command{gnatmake} behave exactly as if the directories had been
17924 specified on the command line, but have a much higher length limit (or
17925 even none on most systems).
17927 It assumes that you have created a list of directories in your Makefile,
17928 using one of the methods presented in
17929 @ref{Automatically Creating a List of Directories}.
17930 For the sake of completeness, we assume that the object
17931 path (where the ALI files are found) is different from the sources patch.
17933 Note a small trick in the Makefile below: for efficiency reasons, we
17934 create two temporary variables (SOURCE_LIST and OBJECT_LIST), that are
17935 expanded immediately by @code{make}. This way we overcome the standard
17936 make behavior which is to expand the variables only when they are
17939 On Windows, if you are using the standard Windows command shell, you must
17940 replace colons with semicolons in the assignments to these variables.
17945 @font@heightrm=cmr8
17948 # In this example, we create both ADA_INCLUDE_PATH and ADA_OBJECT_PATH.
17949 # This is the same thing as putting the -I arguments on the command line.
17950 # (the equivalent of using -aI on the command line would be to define
17951 # only ADA_INCLUDE_PATH, the equivalent of -aO is ADA_OBJECT_PATH).
17952 # You can of course have different values for these variables.
17954 # Note also that we need to keep the previous values of these variables, since
17955 # they might have been set before running 'make' to specify where the GNAT
17956 # library is installed.
17958 # see "Automatically creating a list of directories" to create these
17964 space:=$@{empty@} $@{empty@}
17965 SOURCE_LIST := $@{subst $@{space@},:,$@{SOURCE_DIRS@}@}
17966 OBJECT_LIST := $@{subst $@{space@},:,$@{OBJECT_DIRS@}@}
17967 ADA_INCLUDE_PATH += $@{SOURCE_LIST@}
17968 ADA_OBJECT_PATH += $@{OBJECT_LIST@}
17969 export ADA_INCLUDE_PATH
17970 export ADA_OBJECT_PATH
17977 @node Memory Management Issues
17978 @chapter Memory Management Issues
17981 This chapter describes some useful memory pools provided in the GNAT library
17982 and in particular the GNAT Debug Pool facility, which can be used to detect
17983 incorrect uses of access values (including ``dangling references'').
17985 It also describes the @command{gnatmem} tool, which can be used to track down
17990 * Some Useful Memory Pools::
17991 * The GNAT Debug Pool Facility::
17993 * The gnatmem Tool::
17997 @node Some Useful Memory Pools
17998 @section Some Useful Memory Pools
17999 @findex Memory Pool
18000 @cindex storage, pool
18003 The @code{System.Pool_Global} package offers the Unbounded_No_Reclaim_Pool
18004 storage pool. Allocations use the standard system call @code{malloc} while
18005 deallocations use the standard system call @code{free}. No reclamation is
18006 performed when the pool goes out of scope. For performance reasons, the
18007 standard default Ada allocators/deallocators do not use any explicit storage
18008 pools but if they did, they could use this storage pool without any change in
18009 behavior. That is why this storage pool is used when the user
18010 manages to make the default implicit allocator explicit as in this example:
18011 @smallexample @c ada
18012 type T1 is access Something;
18013 -- no Storage pool is defined for T2
18014 type T2 is access Something_Else;
18015 for T2'Storage_Pool use T1'Storage_Pool;
18016 -- the above is equivalent to
18017 for T2'Storage_Pool use System.Pool_Global.Global_Pool_Object;
18021 The @code{System.Pool_Local} package offers the Unbounded_Reclaim_Pool storage
18022 pool. The allocation strategy is similar to @code{Pool_Local}'s
18023 except that the all
18024 storage allocated with this pool is reclaimed when the pool object goes out of
18025 scope. This pool provides a explicit mechanism similar to the implicit one
18026 provided by several Ada 83 compilers for allocations performed through a local
18027 access type and whose purpose was to reclaim memory when exiting the
18028 scope of a given local access. As an example, the following program does not
18029 leak memory even though it does not perform explicit deallocation:
18031 @smallexample @c ada
18032 with System.Pool_Local;
18033 procedure Pooloc1 is
18034 procedure Internal is
18035 type A is access Integer;
18036 X : System.Pool_Local.Unbounded_Reclaim_Pool;
18037 for A'Storage_Pool use X;
18040 for I in 1 .. 50 loop
18045 for I in 1 .. 100 loop
18052 The @code{System.Pool_Size} package implements the Stack_Bounded_Pool used when
18053 @code{Storage_Size} is specified for an access type.
18054 The whole storage for the pool is
18055 allocated at once, usually on the stack at the point where the access type is
18056 elaborated. It is automatically reclaimed when exiting the scope where the
18057 access type is defined. This package is not intended to be used directly by the
18058 user and it is implicitly used for each such declaration:
18060 @smallexample @c ada
18061 type T1 is access Something;
18062 for T1'Storage_Size use 10_000;
18066 @node The GNAT Debug Pool Facility
18067 @section The GNAT Debug Pool Facility
18069 @cindex storage, pool, memory corruption
18072 The use of unchecked deallocation and unchecked conversion can easily
18073 lead to incorrect memory references. The problems generated by such
18074 references are usually difficult to tackle because the symptoms can be
18075 very remote from the origin of the problem. In such cases, it is
18076 very helpful to detect the problem as early as possible. This is the
18077 purpose of the Storage Pool provided by @code{GNAT.Debug_Pools}.
18079 In order to use the GNAT specific debugging pool, the user must
18080 associate a debug pool object with each of the access types that may be
18081 related to suspected memory problems. See Ada Reference Manual 13.11.
18082 @smallexample @c ada
18083 type Ptr is access Some_Type;
18084 Pool : GNAT.Debug_Pools.Debug_Pool;
18085 for Ptr'Storage_Pool use Pool;
18089 @code{GNAT.Debug_Pools} is derived from a GNAT-specific kind of
18090 pool: the @code{Checked_Pool}. Such pools, like standard Ada storage pools,
18091 allow the user to redefine allocation and deallocation strategies. They
18092 also provide a checkpoint for each dereference, through the use of
18093 the primitive operation @code{Dereference} which is implicitly called at
18094 each dereference of an access value.
18096 Once an access type has been associated with a debug pool, operations on
18097 values of the type may raise four distinct exceptions,
18098 which correspond to four potential kinds of memory corruption:
18101 @code{GNAT.Debug_Pools.Accessing_Not_Allocated_Storage}
18103 @code{GNAT.Debug_Pools.Accessing_Deallocated_Storage}
18105 @code{GNAT.Debug_Pools.Freeing_Not_Allocated_Storage}
18107 @code{GNAT.Debug_Pools.Freeing_Deallocated_Storage }
18111 For types associated with a Debug_Pool, dynamic allocation is performed using
18112 the standard GNAT allocation routine. References to all allocated chunks of
18113 memory are kept in an internal dictionary. Several deallocation strategies are
18114 provided, whereupon the user can choose to release the memory to the system,
18115 keep it allocated for further invalid access checks, or fill it with an easily
18116 recognizable pattern for debug sessions. The memory pattern is the old IBM
18117 hexadecimal convention: @code{16#DEADBEEF#}.
18119 See the documentation in the file g-debpoo.ads for more information on the
18120 various strategies.
18122 Upon each dereference, a check is made that the access value denotes a
18123 properly allocated memory location. Here is a complete example of use of
18124 @code{Debug_Pools}, that includes typical instances of memory corruption:
18125 @smallexample @c ada
18129 with Gnat.Io; use Gnat.Io;
18130 with Unchecked_Deallocation;
18131 with Unchecked_Conversion;
18132 with GNAT.Debug_Pools;
18133 with System.Storage_Elements;
18134 with Ada.Exceptions; use Ada.Exceptions;
18135 procedure Debug_Pool_Test is
18137 type T is access Integer;
18138 type U is access all T;
18140 P : GNAT.Debug_Pools.Debug_Pool;
18141 for T'Storage_Pool use P;
18143 procedure Free is new Unchecked_Deallocation (Integer, T);
18144 function UC is new Unchecked_Conversion (U, T);
18147 procedure Info is new GNAT.Debug_Pools.Print_Info(Put_Line);
18157 Put_Line (Integer'Image(B.all));
18159 when E : others => Put_Line ("raised: " & Exception_Name (E));
18164 when E : others => Put_Line ("raised: " & Exception_Name (E));
18168 Put_Line (Integer'Image(B.all));
18170 when E : others => Put_Line ("raised: " & Exception_Name (E));
18175 when E : others => Put_Line ("raised: " & Exception_Name (E));
18178 end Debug_Pool_Test;
18182 The debug pool mechanism provides the following precise diagnostics on the
18183 execution of this erroneous program:
18186 Total allocated bytes : 0
18187 Total deallocated bytes : 0
18188 Current Water Mark: 0
18192 Total allocated bytes : 8
18193 Total deallocated bytes : 0
18194 Current Water Mark: 8
18197 raised: GNAT.DEBUG_POOLS.ACCESSING_DEALLOCATED_STORAGE
18198 raised: GNAT.DEBUG_POOLS.FREEING_DEALLOCATED_STORAGE
18199 raised: GNAT.DEBUG_POOLS.ACCESSING_NOT_ALLOCATED_STORAGE
18200 raised: GNAT.DEBUG_POOLS.FREEING_NOT_ALLOCATED_STORAGE
18202 Total allocated bytes : 8
18203 Total deallocated bytes : 4
18204 Current Water Mark: 4
18209 @node The gnatmem Tool
18210 @section The @command{gnatmem} Tool
18214 The @code{gnatmem} utility monitors dynamic allocation and
18215 deallocation activity in a program, and displays information about
18216 incorrect deallocations and possible sources of memory leaks.
18217 It provides three type of information:
18220 General information concerning memory management, such as the total
18221 number of allocations and deallocations, the amount of allocated
18222 memory and the high water mark, i.e. the largest amount of allocated
18223 memory in the course of program execution.
18226 Backtraces for all incorrect deallocations, that is to say deallocations
18227 which do not correspond to a valid allocation.
18230 Information on each allocation that is potentially the origin of a memory
18235 * Running gnatmem::
18236 * Switches for gnatmem::
18237 * Example of gnatmem Usage::
18240 @node Running gnatmem
18241 @subsection Running @code{gnatmem}
18244 @code{gnatmem} makes use of the output created by the special version of
18245 allocation and deallocation routines that record call information. This
18246 allows to obtain accurate dynamic memory usage history at a minimal cost to
18247 the execution speed. Note however, that @code{gnatmem} is not supported on
18248 all platforms (currently, it is supported on AIX, HP-UX, GNU/Linux x86,
18249 32-bit Solaris (sparc and x86) and Windows NT/2000/XP (x86).
18252 The @code{gnatmem} command has the form
18255 $ gnatmem [switches] user_program
18259 The program must have been linked with the instrumented version of the
18260 allocation and deallocation routines. This is done by linking with the
18261 @file{libgmem.a} library. For correct symbolic backtrace information,
18262 the user program should be compiled with debugging options
18263 @ref{Switches for gcc}. For example to build @file{my_program}:
18266 $ gnatmake -g my_program -largs -lgmem
18270 When running @file{my_program} the file @file{gmem.out} is produced. This file
18271 contains information about all allocations and deallocations done by the
18272 program. It is produced by the instrumented allocations and
18273 deallocations routines and will be used by @code{gnatmem}.
18276 Gnatmem must be supplied with the @file{gmem.out} file and the executable to
18277 examine. If the location of @file{gmem.out} file was not explicitly supplied by
18278 @code{-i} switch, gnatmem will assume that this file can be found in the
18279 current directory. For example, after you have executed @file{my_program},
18280 @file{gmem.out} can be analyzed by @code{gnatmem} using the command:
18283 $ gnatmem my_program
18287 This will produce the output with the following format:
18289 *************** debut cc
18291 $ gnatmem my_program
18295 Total number of allocations : 45
18296 Total number of deallocations : 6
18297 Final Water Mark (non freed mem) : 11.29 Kilobytes
18298 High Water Mark : 11.40 Kilobytes
18303 Allocation Root # 2
18304 -------------------
18305 Number of non freed allocations : 11
18306 Final Water Mark (non freed mem) : 1.16 Kilobytes
18307 High Water Mark : 1.27 Kilobytes
18309 my_program.adb:23 my_program.alloc
18315 The first block of output gives general information. In this case, the
18316 Ada construct ``@code{@b{new}}'' was executed 45 times, and only 6 calls to an
18317 Unchecked_Deallocation routine occurred.
18320 Subsequent paragraphs display information on all allocation roots.
18321 An allocation root is a specific point in the execution of the program
18322 that generates some dynamic allocation, such as a ``@code{@b{new}}''
18323 construct. This root is represented by an execution backtrace (or subprogram
18324 call stack). By default the backtrace depth for allocations roots is 1, so
18325 that a root corresponds exactly to a source location. The backtrace can
18326 be made deeper, to make the root more specific.
18328 @node Switches for gnatmem
18329 @subsection Switches for @code{gnatmem}
18332 @code{gnatmem} recognizes the following switches:
18337 @cindex @option{-q} (@code{gnatmem})
18338 Quiet. Gives the minimum output needed to identify the origin of the
18339 memory leaks. Omits statistical information.
18342 @cindex @var{N} (@code{gnatmem})
18343 N is an integer literal (usually between 1 and 10) which controls the
18344 depth of the backtraces defining allocation root. The default value for
18345 N is 1. The deeper the backtrace, the more precise the localization of
18346 the root. Note that the total number of roots can depend on this
18347 parameter. This parameter must be specified @emph{before} the name of the
18348 executable to be analyzed, to avoid ambiguity.
18351 @cindex @option{-b} (@code{gnatmem})
18352 This switch has the same effect as just depth parameter.
18354 @item -i @var{file}
18355 @cindex @option{-i} (@code{gnatmem})
18356 Do the @code{gnatmem} processing starting from @file{file}, rather than
18357 @file{gmem.out} in the current directory.
18360 @cindex @option{-m} (@code{gnatmem})
18361 This switch causes @code{gnatmem} to mask the allocation roots that have less
18362 than n leaks. The default value is 1. Specifying the value of 0 will allow to
18363 examine even the roots that didn't result in leaks.
18366 @cindex @option{-s} (@code{gnatmem})
18367 This switch causes @code{gnatmem} to sort the allocation roots according to the
18368 specified order of sort criteria, each identified by a single letter. The
18369 currently supported criteria are @code{n, h, w} standing respectively for
18370 number of unfreed allocations, high watermark, and final watermark
18371 corresponding to a specific root. The default order is @code{nwh}.
18375 @node Example of gnatmem Usage
18376 @subsection Example of @code{gnatmem} Usage
18379 The following example shows the use of @code{gnatmem}
18380 on a simple memory-leaking program.
18381 Suppose that we have the following Ada program:
18383 @smallexample @c ada
18386 with Unchecked_Deallocation;
18387 procedure Test_Gm is
18389 type T is array (1..1000) of Integer;
18390 type Ptr is access T;
18391 procedure Free is new Unchecked_Deallocation (T, Ptr);
18394 procedure My_Alloc is
18399 procedure My_DeAlloc is
18407 for I in 1 .. 5 loop
18408 for J in I .. 5 loop
18419 The program needs to be compiled with debugging option and linked with
18420 @code{gmem} library:
18423 $ gnatmake -g test_gm -largs -lgmem
18427 Then we execute the program as usual:
18434 Then @code{gnatmem} is invoked simply with
18440 which produces the following output (result may vary on different platforms):
18445 Total number of allocations : 18
18446 Total number of deallocations : 5
18447 Final Water Mark (non freed mem) : 53.00 Kilobytes
18448 High Water Mark : 56.90 Kilobytes
18450 Allocation Root # 1
18451 -------------------
18452 Number of non freed allocations : 11
18453 Final Water Mark (non freed mem) : 42.97 Kilobytes
18454 High Water Mark : 46.88 Kilobytes
18456 test_gm.adb:11 test_gm.my_alloc
18458 Allocation Root # 2
18459 -------------------
18460 Number of non freed allocations : 1
18461 Final Water Mark (non freed mem) : 10.02 Kilobytes
18462 High Water Mark : 10.02 Kilobytes
18464 s-secsta.adb:81 system.secondary_stack.ss_init
18466 Allocation Root # 3
18467 -------------------
18468 Number of non freed allocations : 1
18469 Final Water Mark (non freed mem) : 12 Bytes
18470 High Water Mark : 12 Bytes
18472 s-secsta.adb:181 system.secondary_stack.ss_init
18476 Note that the GNAT run time contains itself a certain number of
18477 allocations that have no corresponding deallocation,
18478 as shown here for root #2 and root
18479 #3. This is a normal behavior when the number of non freed allocations
18480 is one, it allocates dynamic data structures that the run time needs for
18481 the complete lifetime of the program. Note also that there is only one
18482 allocation root in the user program with a single line back trace:
18483 test_gm.adb:11 test_gm.my_alloc, whereas a careful analysis of the
18484 program shows that 'My_Alloc' is called at 2 different points in the
18485 source (line 21 and line 24). If those two allocation roots need to be
18486 distinguished, the backtrace depth parameter can be used:
18489 $ gnatmem 3 test_gm
18493 which will give the following output:
18498 Total number of allocations : 18
18499 Total number of deallocations : 5
18500 Final Water Mark (non freed mem) : 53.00 Kilobytes
18501 High Water Mark : 56.90 Kilobytes
18503 Allocation Root # 1
18504 -------------------
18505 Number of non freed allocations : 10
18506 Final Water Mark (non freed mem) : 39.06 Kilobytes
18507 High Water Mark : 42.97 Kilobytes
18509 test_gm.adb:11 test_gm.my_alloc
18510 test_gm.adb:24 test_gm
18511 b_test_gm.c:52 main
18513 Allocation Root # 2
18514 -------------------
18515 Number of non freed allocations : 1
18516 Final Water Mark (non freed mem) : 10.02 Kilobytes
18517 High Water Mark : 10.02 Kilobytes
18519 s-secsta.adb:81 system.secondary_stack.ss_init
18520 s-secsta.adb:283 <system__secondary_stack___elabb>
18521 b_test_gm.c:33 adainit
18523 Allocation Root # 3
18524 -------------------
18525 Number of non freed allocations : 1
18526 Final Water Mark (non freed mem) : 3.91 Kilobytes
18527 High Water Mark : 3.91 Kilobytes
18529 test_gm.adb:11 test_gm.my_alloc
18530 test_gm.adb:21 test_gm
18531 b_test_gm.c:52 main
18533 Allocation Root # 4
18534 -------------------
18535 Number of non freed allocations : 1
18536 Final Water Mark (non freed mem) : 12 Bytes
18537 High Water Mark : 12 Bytes
18539 s-secsta.adb:181 system.secondary_stack.ss_init
18540 s-secsta.adb:283 <system__secondary_stack___elabb>
18541 b_test_gm.c:33 adainit
18545 The allocation root #1 of the first example has been split in 2 roots #1
18546 and #3 thanks to the more precise associated backtrace.
18550 @node Creating Sample Bodies Using gnatstub
18551 @chapter Creating Sample Bodies Using @command{gnatstub}
18555 @command{gnatstub} creates body stubs, that is, empty but compilable bodies
18556 for library unit declarations.
18558 To create a body stub, @command{gnatstub} has to compile the library
18559 unit declaration. Therefore, bodies can be created only for legal
18560 library units. Moreover, if a library unit depends semantically upon
18561 units located outside the current directory, you have to provide
18562 the source search path when calling @command{gnatstub}, see the description
18563 of @command{gnatstub} switches below.
18566 * Running gnatstub::
18567 * Switches for gnatstub::
18570 @node Running gnatstub
18571 @section Running @command{gnatstub}
18574 @command{gnatstub} has the command-line interface of the form
18577 $ gnatstub [switches] filename [directory]
18584 is the name of the source file that contains a library unit declaration
18585 for which a body must be created. The file name may contain the path
18587 The file name does not have to follow the GNAT file name conventions. If the
18589 does not follow GNAT file naming conventions, the name of the body file must
18591 explicitly as the value of the @option{^-o^/BODY=^@var{body-name}} option.
18592 If the file name follows the GNAT file naming
18593 conventions and the name of the body file is not provided,
18596 of the body file from the argument file name by replacing the @file{.ads}
18598 with the @file{.adb} suffix.
18601 indicates the directory in which the body stub is to be placed (the default
18606 is an optional sequence of switches as described in the next section
18609 @node Switches for gnatstub
18610 @section Switches for @command{gnatstub}
18616 @cindex @option{^-f^/FULL^} (@command{gnatstub})
18617 If the destination directory already contains a file with the name of the
18619 for the argument spec file, replace it with the generated body stub.
18621 @item ^-hs^/HEADER=SPEC^
18622 @cindex @option{^-hs^/HEADER=SPEC^} (@command{gnatstub})
18623 Put the comment header (i.e., all the comments preceding the
18624 compilation unit) from the source of the library unit declaration
18625 into the body stub.
18627 @item ^-hg^/HEADER=GENERAL^
18628 @cindex @option{^-hg^/HEADER=GENERAL^} (@command{gnatstub})
18629 Put a sample comment header into the body stub.
18633 @cindex @option{-IDIR} (@command{gnatstub})
18635 @cindex @option{-I-} (@command{gnatstub})
18638 @item /NOCURRENT_DIRECTORY
18639 @cindex @option{/NOCURRENT_DIRECTORY} (@command{gnatstub})
18641 ^These switches have ^This switch has^ the same meaning as in calls to
18643 ^They define ^It defines ^ the source search path in the call to
18644 @command{gcc} issued
18645 by @command{gnatstub} to compile an argument source file.
18647 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE=^@var{PATH}
18648 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@command{gnatstub})
18649 This switch has the same meaning as in calls to @command{gcc}.
18650 It defines the additional configuration file to be passed to the call to
18651 @command{gcc} issued
18652 by @command{gnatstub} to compile an argument source file.
18654 @item ^-gnatyM^/MAX_LINE_LENGTH=^@var{n}
18655 @cindex @option{^-gnatyM^/MAX_LINE_LENGTH^} (@command{gnatstub})
18656 (@var{n} is a non-negative integer). Set the maximum line length in the
18657 body stub to @var{n}; the default is 79. The maximum value that can be
18658 specified is 32767. Note that in the special case of configuration
18659 pragma files, the maximum is always 32767 regardless of whether or
18660 not this switch appears.
18662 @item ^-gnaty^/STYLE_CHECKS=^@var{n}
18663 @cindex @option{^-gnaty^/STYLE_CHECKS=^} (@command{gnatstub})
18664 (@var{n} is a non-negative integer from 1 to 9). Set the indentation level in
18665 the generated body sample to @var{n}.
18666 The default indentation is 3.
18668 @item ^-gnatyo^/ORDERED_SUBPROGRAMS^
18669 @cindex @option{^-gnato^/ORDERED_SUBPROGRAMS^} (@command{gnatstub})
18670 Order local bodies alphabetically. (By default local bodies are ordered
18671 in the same way as the corresponding local specs in the argument spec file.)
18673 @item ^-i^/INDENTATION=^@var{n}
18674 @cindex @option{^-i^/INDENTATION^} (@command{gnatstub})
18675 Same as @option{^-gnaty^/STYLE_CHECKS=^@var{n}}
18677 @item ^-k^/TREE_FILE=SAVE^
18678 @cindex @option{^-k^/TREE_FILE=SAVE^} (@command{gnatstub})
18679 Do not remove the tree file (i.e., the snapshot of the compiler internal
18680 structures used by @command{gnatstub}) after creating the body stub.
18682 @item ^-l^/LINE_LENGTH=^@var{n}
18683 @cindex @option{^-l^/LINE_LENGTH^} (@command{gnatstub})
18684 Same as @option{^-gnatyM^/MAX_LINE_LENGTH=^@var{n}}
18686 @item ^-o^/BODY=^@var{body-name}
18687 @cindex @option{^-o^/BODY^} (@command{gnatstub})
18688 Body file name. This should be set if the argument file name does not
18690 the GNAT file naming
18691 conventions. If this switch is omitted the default name for the body will be
18693 from the argument file name according to the GNAT file naming conventions.
18696 @cindex @option{^-q^/QUIET^} (@command{gnatstub})
18697 Quiet mode: do not generate a confirmation when a body is
18698 successfully created, and do not generate a message when a body is not
18702 @item ^-r^/TREE_FILE=REUSE^
18703 @cindex @option{^-r^/TREE_FILE=REUSE^} (@command{gnatstub})
18704 Reuse the tree file (if it exists) instead of creating it. Instead of
18705 creating the tree file for the library unit declaration, @command{gnatstub}
18706 tries to find it in the current directory and use it for creating
18707 a body. If the tree file is not found, no body is created. This option
18708 also implies @option{^-k^/SAVE^}, whether or not
18709 the latter is set explicitly.
18711 @item ^-t^/TREE_FILE=OVERWRITE^
18712 @cindex @option{^-t^/TREE_FILE=OVERWRITE^} (@command{gnatstub})
18713 Overwrite the existing tree file. If the current directory already
18714 contains the file which, according to the GNAT file naming rules should
18715 be considered as a tree file for the argument source file,
18717 will refuse to create the tree file needed to create a sample body
18718 unless this option is set.
18720 @item ^-v^/VERBOSE^
18721 @cindex @option{^-v^/VERBOSE^} (@command{gnatstub})
18722 Verbose mode: generate version information.
18726 @node Other Utility Programs
18727 @chapter Other Utility Programs
18730 This chapter discusses some other utility programs available in the Ada
18734 * Using Other Utility Programs with GNAT::
18735 * The External Symbol Naming Scheme of GNAT::
18737 * Ada Mode for Glide::
18739 * Converting Ada Files to html with gnathtml::
18740 * Installing gnathtml::
18747 @node Using Other Utility Programs with GNAT
18748 @section Using Other Utility Programs with GNAT
18751 The object files generated by GNAT are in standard system format and in
18752 particular the debugging information uses this format. This means
18753 programs generated by GNAT can be used with existing utilities that
18754 depend on these formats.
18757 In general, any utility program that works with C will also often work with
18758 Ada programs generated by GNAT. This includes software utilities such as
18759 gprof (a profiling program), @code{gdb} (the FSF debugger), and utilities such
18763 @node The External Symbol Naming Scheme of GNAT
18764 @section The External Symbol Naming Scheme of GNAT
18767 In order to interpret the output from GNAT, when using tools that are
18768 originally intended for use with other languages, it is useful to
18769 understand the conventions used to generate link names from the Ada
18772 All link names are in all lowercase letters. With the exception of library
18773 procedure names, the mechanism used is simply to use the full expanded
18774 Ada name with dots replaced by double underscores. For example, suppose
18775 we have the following package spec:
18777 @smallexample @c ada
18788 The variable @code{MN} has a full expanded Ada name of @code{QRS.MN}, so
18789 the corresponding link name is @code{qrs__mn}.
18791 Of course if a @code{pragma Export} is used this may be overridden:
18793 @smallexample @c ada
18798 pragma Export (Var1, C, External_Name => "var1_name");
18800 pragma Export (Var2, C, Link_Name => "var2_link_name");
18807 In this case, the link name for @var{Var1} is whatever link name the
18808 C compiler would assign for the C function @var{var1_name}. This typically
18809 would be either @var{var1_name} or @var{_var1_name}, depending on operating
18810 system conventions, but other possibilities exist. The link name for
18811 @var{Var2} is @var{var2_link_name}, and this is not operating system
18815 One exception occurs for library level procedures. A potential ambiguity
18816 arises between the required name @code{_main} for the C main program,
18817 and the name we would otherwise assign to an Ada library level procedure
18818 called @code{Main} (which might well not be the main program).
18820 To avoid this ambiguity, we attach the prefix @code{_ada_} to such
18821 names. So if we have a library level procedure such as
18823 @smallexample @c ada
18826 procedure Hello (S : String);
18832 the external name of this procedure will be @var{_ada_hello}.
18835 @node Ada Mode for Glide
18836 @section Ada Mode for @code{Glide}
18837 @cindex Ada mode (for Glide)
18840 The Glide mode for programming in Ada (both Ada83 and Ada95) helps the
18841 user to understand and navigate existing code, and facilitates writing
18842 new code. It furthermore provides some utility functions for easier
18843 integration of standard Emacs features when programming in Ada.
18845 Its general features include:
18849 An Integrated Development Environment with functionality such as the
18854 ``Project files'' for configuration-specific aspects
18855 (e.g. directories and compilation options)
18858 Compiling and stepping through error messages.
18861 Running and debugging an applications within Glide.
18868 User configurability
18871 Some of the specific Ada mode features are:
18875 Functions for easy and quick stepping through Ada code
18878 Getting cross reference information for identifiers (e.g., finding a
18879 defining occurrence)
18882 Displaying an index menu of types and subprograms, allowing
18883 direct selection for browsing
18886 Automatic color highlighting of the various Ada entities
18889 Glide directly supports writing Ada code, via several facilities:
18893 Switching between spec and body files with possible
18894 autogeneration of body files
18897 Automatic formating of subprogram parameter lists
18900 Automatic indentation according to Ada syntax
18903 Automatic completion of identifiers
18906 Automatic (and configurable) casing of identifiers, keywords, and attributes
18909 Insertion of syntactic templates
18912 Block commenting / uncommenting
18916 For more information, please refer to the online documentation
18917 available in the @code{Glide} @result{} @code{Help} menu.
18920 @node Converting Ada Files to html with gnathtml
18921 @section Converting Ada Files to HTML with @code{gnathtml}
18924 This @code{Perl} script allows Ada source files to be browsed using
18925 standard Web browsers. For installation procedure, see the section
18926 @xref{Installing gnathtml}.
18928 Ada reserved keywords are highlighted in a bold font and Ada comments in
18929 a blue font. Unless your program was compiled with the gcc @option{-gnatx}
18930 switch to suppress the generation of cross-referencing information, user
18931 defined variables and types will appear in a different color; you will
18932 be able to click on any identifier and go to its declaration.
18934 The command line is as follow:
18936 $ perl gnathtml.pl [switches] ada-files
18940 You can pass it as many Ada files as you want. @code{gnathtml} will generate
18941 an html file for every ada file, and a global file called @file{index.htm}.
18942 This file is an index of every identifier defined in the files.
18944 The available switches are the following ones :
18948 @cindex @option{-83} (@code{gnathtml})
18949 Only the subset on the Ada 83 keywords will be highlighted, not the full
18950 Ada 95 keywords set.
18952 @item -cc @var{color}
18953 @cindex @option{-cc} (@code{gnathtml})
18954 This option allows you to change the color used for comments. The default
18955 value is green. The color argument can be any name accepted by html.
18958 @cindex @option{-d} (@code{gnathtml})
18959 If the ada files depend on some other files (using for instance the
18960 @code{with} command, the latter will also be converted to html.
18961 Only the files in the user project will be converted to html, not the files
18962 in the run-time library itself.
18965 @cindex @option{-D} (@code{gnathtml})
18966 This command is the same as @option{-d} above, but @command{gnathtml} will
18967 also look for files in the run-time library, and generate html files for them.
18969 @item -ext @var{extension}
18970 @cindex @option{-ext} (@code{gnathtml})
18971 This option allows you to change the extension of the generated HTML files.
18972 If you do not specify an extension, it will default to @file{htm}.
18975 @cindex @option{-f} (@code{gnathtml})
18976 By default, gnathtml will generate html links only for global entities
18977 ('with'ed units, global variables and types,...). If you specify the
18978 @option{-f} on the command line, then links will be generated for local
18981 @item -l @var{number}
18982 @cindex @option{-l} (@code{gnathtml})
18983 If this switch is provided and @var{number} is not 0, then @code{gnathtml}
18984 will number the html files every @var{number} line.
18987 @cindex @option{-I} (@code{gnathtml})
18988 Specify a directory to search for library files (@file{.ALI} files) and
18989 source files. You can provide several -I switches on the command line,
18990 and the directories will be parsed in the order of the command line.
18993 @cindex @option{-o} (@code{gnathtml})
18994 Specify the output directory for html files. By default, gnathtml will
18995 saved the generated html files in a subdirectory named @file{html/}.
18997 @item -p @var{file}
18998 @cindex @option{-p} (@code{gnathtml})
18999 If you are using Emacs and the most recent Emacs Ada mode, which provides
19000 a full Integrated Development Environment for compiling, checking,
19001 running and debugging applications, you may use @file{.gpr} files
19002 to give the directories where Emacs can find sources and object files.
19004 Using this switch, you can tell gnathtml to use these files. This allows
19005 you to get an html version of your application, even if it is spread
19006 over multiple directories.
19008 @item -sc @var{color}
19009 @cindex @option{-sc} (@code{gnathtml})
19010 This option allows you to change the color used for symbol definitions.
19011 The default value is red. The color argument can be any name accepted by html.
19013 @item -t @var{file}
19014 @cindex @option{-t} (@code{gnathtml})
19015 This switch provides the name of a file. This file contains a list of
19016 file names to be converted, and the effect is exactly as though they had
19017 appeared explicitly on the command line. This
19018 is the recommended way to work around the command line length limit on some
19023 @node Installing gnathtml
19024 @section Installing @code{gnathtml}
19027 @code{Perl} needs to be installed on your machine to run this script.
19028 @code{Perl} is freely available for almost every architecture and
19029 Operating System via the Internet.
19031 On Unix systems, you may want to modify the first line of the script
19032 @code{gnathtml}, to explicitly tell the Operating system where Perl
19033 is. The syntax of this line is :
19035 #!full_path_name_to_perl
19039 Alternatively, you may run the script using the following command line:
19042 $ perl gnathtml.pl [switches] files
19051 The GNAT distribution provides an Ada 95 template for the Digital Language
19052 Sensitive Editor (LSE), a component of DECset. In order to
19053 access it, invoke LSE with the qualifier /ENVIRONMENT=GNU:[LIB]ADA95.ENV.
19060 GNAT supports The Digital Performance Coverage Analyzer (PCA), a component
19061 of DECset. To use it proceed as outlined under ``HELP PCA'', except for running
19062 the collection phase with the /DEBUG qualifier.
19065 $ GNAT MAKE /DEBUG <PROGRAM_NAME>
19066 $ DEFINE LIB$DEBUG PCA$COLLECTOR
19067 $ RUN/DEBUG <PROGRAM_NAME>
19072 @node Running and Debugging Ada Programs
19073 @chapter Running and Debugging Ada Programs
19077 This chapter discusses how to debug Ada programs.
19079 It applies to the Alpha OpenVMS platform;
19080 the debugger for Integrity OpenVMS is scheduled for a subsequent release.
19083 An incorrect Ada program may be handled in three ways by the GNAT compiler:
19087 The illegality may be a violation of the static semantics of Ada. In
19088 that case GNAT diagnoses the constructs in the program that are illegal.
19089 It is then a straightforward matter for the user to modify those parts of
19093 The illegality may be a violation of the dynamic semantics of Ada. In
19094 that case the program compiles and executes, but may generate incorrect
19095 results, or may terminate abnormally with some exception.
19098 When presented with a program that contains convoluted errors, GNAT
19099 itself may terminate abnormally without providing full diagnostics on
19100 the incorrect user program.
19104 * The GNAT Debugger GDB::
19106 * Introduction to GDB Commands::
19107 * Using Ada Expressions::
19108 * Calling User-Defined Subprograms::
19109 * Using the Next Command in a Function::
19112 * Debugging Generic Units::
19113 * GNAT Abnormal Termination or Failure to Terminate::
19114 * Naming Conventions for GNAT Source Files::
19115 * Getting Internal Debugging Information::
19116 * Stack Traceback::
19122 @node The GNAT Debugger GDB
19123 @section The GNAT Debugger GDB
19126 @code{GDB} is a general purpose, platform-independent debugger that
19127 can be used to debug mixed-language programs compiled with @command{gcc},
19128 and in particular is capable of debugging Ada programs compiled with
19129 GNAT. The latest versions of @code{GDB} are Ada-aware and can handle
19130 complex Ada data structures.
19132 The manual @cite{Debugging with GDB}
19134 , located in the GNU:[DOCS] directory,
19136 contains full details on the usage of @code{GDB}, including a section on
19137 its usage on programs. This manual should be consulted for full
19138 details. The section that follows is a brief introduction to the
19139 philosophy and use of @code{GDB}.
19141 When GNAT programs are compiled, the compiler optionally writes debugging
19142 information into the generated object file, including information on
19143 line numbers, and on declared types and variables. This information is
19144 separate from the generated code. It makes the object files considerably
19145 larger, but it does not add to the size of the actual executable that
19146 will be loaded into memory, and has no impact on run-time performance. The
19147 generation of debug information is triggered by the use of the
19148 ^-g^/DEBUG^ switch in the gcc or gnatmake command used to carry out
19149 the compilations. It is important to emphasize that the use of these
19150 options does not change the generated code.
19152 The debugging information is written in standard system formats that
19153 are used by many tools, including debuggers and profilers. The format
19154 of the information is typically designed to describe C types and
19155 semantics, but GNAT implements a translation scheme which allows full
19156 details about Ada types and variables to be encoded into these
19157 standard C formats. Details of this encoding scheme may be found in
19158 the file exp_dbug.ads in the GNAT source distribution. However, the
19159 details of this encoding are, in general, of no interest to a user,
19160 since @code{GDB} automatically performs the necessary decoding.
19162 When a program is bound and linked, the debugging information is
19163 collected from the object files, and stored in the executable image of
19164 the program. Again, this process significantly increases the size of
19165 the generated executable file, but it does not increase the size of
19166 the executable program itself. Furthermore, if this program is run in
19167 the normal manner, it runs exactly as if the debug information were
19168 not present, and takes no more actual memory.
19170 However, if the program is run under control of @code{GDB}, the
19171 debugger is activated. The image of the program is loaded, at which
19172 point it is ready to run. If a run command is given, then the program
19173 will run exactly as it would have if @code{GDB} were not present. This
19174 is a crucial part of the @code{GDB} design philosophy. @code{GDB} is
19175 entirely non-intrusive until a breakpoint is encountered. If no
19176 breakpoint is ever hit, the program will run exactly as it would if no
19177 debugger were present. When a breakpoint is hit, @code{GDB} accesses
19178 the debugging information and can respond to user commands to inspect
19179 variables, and more generally to report on the state of execution.
19183 @section Running GDB
19186 The debugger can be launched directly and simply from @code{glide} or
19187 through its graphical interface: @code{gvd}. It can also be used
19188 directly in text mode. Here is described the basic use of @code{GDB}
19189 in text mode. All the commands described below can be used in the
19190 @code{gvd} console window even though there is usually other more
19191 graphical ways to achieve the same goals.
19195 The command to run the graphical interface of the debugger is
19202 The command to run @code{GDB} in text mode is
19205 $ ^gdb program^$ GDB PROGRAM^
19209 where @code{^program^PROGRAM^} is the name of the executable file. This
19210 activates the debugger and results in a prompt for debugger commands.
19211 The simplest command is simply @code{run}, which causes the program to run
19212 exactly as if the debugger were not present. The following section
19213 describes some of the additional commands that can be given to @code{GDB}.
19215 @c *******************************
19216 @node Introduction to GDB Commands
19217 @section Introduction to GDB Commands
19220 @code{GDB} contains a large repertoire of commands. The manual
19221 @cite{Debugging with GDB}
19223 , located in the GNU:[DOCS] directory,
19225 includes extensive documentation on the use
19226 of these commands, together with examples of their use. Furthermore,
19227 the command @var{help} invoked from within @code{GDB} activates a simple help
19228 facility which summarizes the available commands and their options.
19229 In this section we summarize a few of the most commonly
19230 used commands to give an idea of what @code{GDB} is about. You should create
19231 a simple program with debugging information and experiment with the use of
19232 these @code{GDB} commands on the program as you read through the
19236 @item set args @var{arguments}
19237 The @var{arguments} list above is a list of arguments to be passed to
19238 the program on a subsequent run command, just as though the arguments
19239 had been entered on a normal invocation of the program. The @code{set args}
19240 command is not needed if the program does not require arguments.
19243 The @code{run} command causes execution of the program to start from
19244 the beginning. If the program is already running, that is to say if
19245 you are currently positioned at a breakpoint, then a prompt will ask
19246 for confirmation that you want to abandon the current execution and
19249 @item breakpoint @var{location}
19250 The breakpoint command sets a breakpoint, that is to say a point at which
19251 execution will halt and @code{GDB} will await further
19252 commands. @var{location} is
19253 either a line number within a file, given in the format @code{file:linenumber},
19254 or it is the name of a subprogram. If you request that a breakpoint be set on
19255 a subprogram that is overloaded, a prompt will ask you to specify on which of
19256 those subprograms you want to breakpoint. You can also
19257 specify that all of them should be breakpointed. If the program is run
19258 and execution encounters the breakpoint, then the program
19259 stops and @code{GDB} signals that the breakpoint was encountered by
19260 printing the line of code before which the program is halted.
19262 @item breakpoint exception @var{name}
19263 A special form of the breakpoint command which breakpoints whenever
19264 exception @var{name} is raised.
19265 If @var{name} is omitted,
19266 then a breakpoint will occur when any exception is raised.
19268 @item print @var{expression}
19269 This will print the value of the given expression. Most simple
19270 Ada expression formats are properly handled by @code{GDB}, so the expression
19271 can contain function calls, variables, operators, and attribute references.
19274 Continues execution following a breakpoint, until the next breakpoint or the
19275 termination of the program.
19278 Executes a single line after a breakpoint. If the next statement
19279 is a subprogram call, execution continues into (the first statement of)
19280 the called subprogram.
19283 Executes a single line. If this line is a subprogram call, executes and
19284 returns from the call.
19287 Lists a few lines around the current source location. In practice, it
19288 is usually more convenient to have a separate edit window open with the
19289 relevant source file displayed. Successive applications of this command
19290 print subsequent lines. The command can be given an argument which is a
19291 line number, in which case it displays a few lines around the specified one.
19294 Displays a backtrace of the call chain. This command is typically
19295 used after a breakpoint has occurred, to examine the sequence of calls that
19296 leads to the current breakpoint. The display includes one line for each
19297 activation record (frame) corresponding to an active subprogram.
19300 At a breakpoint, @code{GDB} can display the values of variables local
19301 to the current frame. The command @code{up} can be used to
19302 examine the contents of other active frames, by moving the focus up
19303 the stack, that is to say from callee to caller, one frame at a time.
19306 Moves the focus of @code{GDB} down from the frame currently being
19307 examined to the frame of its callee (the reverse of the previous command),
19309 @item frame @var{n}
19310 Inspect the frame with the given number. The value 0 denotes the frame
19311 of the current breakpoint, that is to say the top of the call stack.
19315 The above list is a very short introduction to the commands that
19316 @code{GDB} provides. Important additional capabilities, including conditional
19317 breakpoints, the ability to execute command sequences on a breakpoint,
19318 the ability to debug at the machine instruction level and many other
19319 features are described in detail in @cite{Debugging with GDB}.
19320 Note that most commands can be abbreviated
19321 (for example, c for continue, bt for backtrace).
19323 @node Using Ada Expressions
19324 @section Using Ada Expressions
19325 @cindex Ada expressions
19328 @code{GDB} supports a fairly large subset of Ada expression syntax, with some
19329 extensions. The philosophy behind the design of this subset is
19333 That @code{GDB} should provide basic literals and access to operations for
19334 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
19335 leaving more sophisticated computations to subprograms written into the
19336 program (which therefore may be called from @code{GDB}).
19339 That type safety and strict adherence to Ada language restrictions
19340 are not particularly important to the @code{GDB} user.
19343 That brevity is important to the @code{GDB} user.
19346 Thus, for brevity, the debugger acts as if there were
19347 implicit @code{with} and @code{use} clauses in effect for all user-written
19348 packages, thus making it unnecessary to fully qualify most names with
19349 their packages, regardless of context. Where this causes ambiguity,
19350 @code{GDB} asks the user's intent.
19352 For details on the supported Ada syntax, see @cite{Debugging with GDB}.
19354 @node Calling User-Defined Subprograms
19355 @section Calling User-Defined Subprograms
19358 An important capability of @code{GDB} is the ability to call user-defined
19359 subprograms while debugging. This is achieved simply by entering
19360 a subprogram call statement in the form:
19363 call subprogram-name (parameters)
19367 The keyword @code{call} can be omitted in the normal case where the
19368 @code{subprogram-name} does not coincide with any of the predefined
19369 @code{GDB} commands.
19371 The effect is to invoke the given subprogram, passing it the
19372 list of parameters that is supplied. The parameters can be expressions and
19373 can include variables from the program being debugged. The
19374 subprogram must be defined
19375 at the library level within your program, and @code{GDB} will call the
19376 subprogram within the environment of your program execution (which
19377 means that the subprogram is free to access or even modify variables
19378 within your program).
19380 The most important use of this facility is in allowing the inclusion of
19381 debugging routines that are tailored to particular data structures
19382 in your program. Such debugging routines can be written to provide a suitably
19383 high-level description of an abstract type, rather than a low-level dump
19384 of its physical layout. After all, the standard
19385 @code{GDB print} command only knows the physical layout of your
19386 types, not their abstract meaning. Debugging routines can provide information
19387 at the desired semantic level and are thus enormously useful.
19389 For example, when debugging GNAT itself, it is crucial to have access to
19390 the contents of the tree nodes used to represent the program internally.
19391 But tree nodes are represented simply by an integer value (which in turn
19392 is an index into a table of nodes).
19393 Using the @code{print} command on a tree node would simply print this integer
19394 value, which is not very useful. But the PN routine (defined in file
19395 treepr.adb in the GNAT sources) takes a tree node as input, and displays
19396 a useful high level representation of the tree node, which includes the
19397 syntactic category of the node, its position in the source, the integers
19398 that denote descendant nodes and parent node, as well as varied
19399 semantic information. To study this example in more detail, you might want to
19400 look at the body of the PN procedure in the stated file.
19402 @node Using the Next Command in a Function
19403 @section Using the Next Command in a Function
19406 When you use the @code{next} command in a function, the current source
19407 location will advance to the next statement as usual. A special case
19408 arises in the case of a @code{return} statement.
19410 Part of the code for a return statement is the ``epilog'' of the function.
19411 This is the code that returns to the caller. There is only one copy of
19412 this epilog code, and it is typically associated with the last return
19413 statement in the function if there is more than one return. In some
19414 implementations, this epilog is associated with the first statement
19417 The result is that if you use the @code{next} command from a return
19418 statement that is not the last return statement of the function you
19419 may see a strange apparent jump to the last return statement or to
19420 the start of the function. You should simply ignore this odd jump.
19421 The value returned is always that from the first return statement
19422 that was stepped through.
19424 @node Ada Exceptions
19425 @section Breaking on Ada Exceptions
19429 You can set breakpoints that trip when your program raises
19430 selected exceptions.
19433 @item break exception
19434 Set a breakpoint that trips whenever (any task in the) program raises
19437 @item break exception @var{name}
19438 Set a breakpoint that trips whenever (any task in the) program raises
19439 the exception @var{name}.
19441 @item break exception unhandled
19442 Set a breakpoint that trips whenever (any task in the) program raises an
19443 exception for which there is no handler.
19445 @item info exceptions
19446 @itemx info exceptions @var{regexp}
19447 The @code{info exceptions} command permits the user to examine all defined
19448 exceptions within Ada programs. With a regular expression, @var{regexp}, as
19449 argument, prints out only those exceptions whose name matches @var{regexp}.
19457 @code{GDB} allows the following task-related commands:
19461 This command shows a list of current Ada tasks, as in the following example:
19468 ID TID P-ID Thread Pri State Name
19469 1 8088000 0 807e000 15 Child Activation Wait main_task
19470 2 80a4000 1 80ae000 15 Accept/Select Wait b
19471 3 809a800 1 80a4800 15 Child Activation Wait a
19472 * 4 80ae800 3 80b8000 15 Running c
19476 In this listing, the asterisk before the first task indicates it to be the
19477 currently running task. The first column lists the task ID that is used
19478 to refer to tasks in the following commands.
19480 @item break @var{linespec} task @var{taskid}
19481 @itemx break @var{linespec} task @var{taskid} if @dots{}
19482 @cindex Breakpoints and tasks
19483 These commands are like the @code{break @dots{} thread @dots{}}.
19484 @var{linespec} specifies source lines.
19486 Use the qualifier @samp{task @var{taskid}} with a breakpoint command
19487 to specify that you only want @code{GDB} to stop the program when a
19488 particular Ada task reaches this breakpoint. @var{taskid} is one of the
19489 numeric task identifiers assigned by @code{GDB}, shown in the first
19490 column of the @samp{info tasks} display.
19492 If you do not specify @samp{task @var{taskid}} when you set a
19493 breakpoint, the breakpoint applies to @emph{all} tasks of your
19496 You can use the @code{task} qualifier on conditional breakpoints as
19497 well; in this case, place @samp{task @var{taskid}} before the
19498 breakpoint condition (before the @code{if}).
19500 @item task @var{taskno}
19501 @cindex Task switching
19503 This command allows to switch to the task referred by @var{taskno}. In
19504 particular, This allows to browse the backtrace of the specified
19505 task. It is advised to switch back to the original task before
19506 continuing execution otherwise the scheduling of the program may be
19511 For more detailed information on the tasking support,
19512 see @cite{Debugging with GDB}.
19514 @node Debugging Generic Units
19515 @section Debugging Generic Units
19516 @cindex Debugging Generic Units
19520 GNAT always uses code expansion for generic instantiation. This means that
19521 each time an instantiation occurs, a complete copy of the original code is
19522 made, with appropriate substitutions of formals by actuals.
19524 It is not possible to refer to the original generic entities in
19525 @code{GDB}, but it is always possible to debug a particular instance of
19526 a generic, by using the appropriate expanded names. For example, if we have
19528 @smallexample @c ada
19533 generic package k is
19534 procedure kp (v1 : in out integer);
19538 procedure kp (v1 : in out integer) is
19544 package k1 is new k;
19545 package k2 is new k;
19547 var : integer := 1;
19560 Then to break on a call to procedure kp in the k2 instance, simply
19564 (gdb) break g.k2.kp
19568 When the breakpoint occurs, you can step through the code of the
19569 instance in the normal manner and examine the values of local variables, as for
19572 @node GNAT Abnormal Termination or Failure to Terminate
19573 @section GNAT Abnormal Termination or Failure to Terminate
19574 @cindex GNAT Abnormal Termination or Failure to Terminate
19577 When presented with programs that contain serious errors in syntax
19579 GNAT may on rare occasions experience problems in operation, such
19581 segmentation fault or illegal memory access, raising an internal
19582 exception, terminating abnormally, or failing to terminate at all.
19583 In such cases, you can activate
19584 various features of GNAT that can help you pinpoint the construct in your
19585 program that is the likely source of the problem.
19587 The following strategies are presented in increasing order of
19588 difficulty, corresponding to your experience in using GNAT and your
19589 familiarity with compiler internals.
19593 Run @command{gcc} with the @option{-gnatf}. This first
19594 switch causes all errors on a given line to be reported. In its absence,
19595 only the first error on a line is displayed.
19597 The @option{-gnatdO} switch causes errors to be displayed as soon as they
19598 are encountered, rather than after compilation is terminated. If GNAT
19599 terminates prematurely or goes into an infinite loop, the last error
19600 message displayed may help to pinpoint the culprit.
19603 Run @command{gcc} with the @option{^-v (verbose)^/VERBOSE^} switch. In this
19604 mode, @command{gcc} produces ongoing information about the progress of the
19605 compilation and provides the name of each procedure as code is
19606 generated. This switch allows you to find which Ada procedure was being
19607 compiled when it encountered a code generation problem.
19610 @cindex @option{-gnatdc} switch
19611 Run @command{gcc} with the @option{-gnatdc} switch. This is a GNAT specific
19612 switch that does for the front-end what @option{^-v^VERBOSE^} does
19613 for the back end. The system prints the name of each unit,
19614 either a compilation unit or nested unit, as it is being analyzed.
19616 Finally, you can start
19617 @code{gdb} directly on the @code{gnat1} executable. @code{gnat1} is the
19618 front-end of GNAT, and can be run independently (normally it is just
19619 called from @command{gcc}). You can use @code{gdb} on @code{gnat1} as you
19620 would on a C program (but @pxref{The GNAT Debugger GDB} for caveats). The
19621 @code{where} command is the first line of attack; the variable
19622 @code{lineno} (seen by @code{print lineno}), used by the second phase of
19623 @code{gnat1} and by the @command{gcc} backend, indicates the source line at
19624 which the execution stopped, and @code{input_file name} indicates the name of
19628 @node Naming Conventions for GNAT Source Files
19629 @section Naming Conventions for GNAT Source Files
19632 In order to examine the workings of the GNAT system, the following
19633 brief description of its organization may be helpful:
19637 Files with prefix @file{^sc^SC^} contain the lexical scanner.
19640 All files prefixed with @file{^par^PAR^} are components of the parser. The
19641 numbers correspond to chapters of the Ada 95 Reference Manual. For example,
19642 parsing of select statements can be found in @file{par-ch9.adb}.
19645 All files prefixed with @file{^sem^SEM^} perform semantic analysis. The
19646 numbers correspond to chapters of the Ada standard. For example, all
19647 issues involving context clauses can be found in @file{sem_ch10.adb}. In
19648 addition, some features of the language require sufficient special processing
19649 to justify their own semantic files: sem_aggr for aggregates, sem_disp for
19650 dynamic dispatching, etc.
19653 All files prefixed with @file{^exp^EXP^} perform normalization and
19654 expansion of the intermediate representation (abstract syntax tree, or AST).
19655 these files use the same numbering scheme as the parser and semantics files.
19656 For example, the construction of record initialization procedures is done in
19657 @file{exp_ch3.adb}.
19660 The files prefixed with @file{^bind^BIND^} implement the binder, which
19661 verifies the consistency of the compilation, determines an order of
19662 elaboration, and generates the bind file.
19665 The files @file{atree.ads} and @file{atree.adb} detail the low-level
19666 data structures used by the front-end.
19669 The files @file{sinfo.ads} and @file{sinfo.adb} detail the structure of
19670 the abstract syntax tree as produced by the parser.
19673 The files @file{einfo.ads} and @file{einfo.adb} detail the attributes of
19674 all entities, computed during semantic analysis.
19677 Library management issues are dealt with in files with prefix
19683 Ada files with the prefix @file{^a-^A-^} are children of @code{Ada}, as
19684 defined in Annex A.
19689 Files with prefix @file{^i-^I-^} are children of @code{Interfaces}, as
19690 defined in Annex B.
19694 Files with prefix @file{^s-^S-^} are children of @code{System}. This includes
19695 both language-defined children and GNAT run-time routines.
19699 Files with prefix @file{^g-^G-^} are children of @code{GNAT}. These are useful
19700 general-purpose packages, fully documented in their specifications. All
19701 the other @file{.c} files are modifications of common @command{gcc} files.
19704 @node Getting Internal Debugging Information
19705 @section Getting Internal Debugging Information
19708 Most compilers have internal debugging switches and modes. GNAT
19709 does also, except GNAT internal debugging switches and modes are not
19710 secret. A summary and full description of all the compiler and binder
19711 debug flags are in the file @file{debug.adb}. You must obtain the
19712 sources of the compiler to see the full detailed effects of these flags.
19714 The switches that print the source of the program (reconstructed from
19715 the internal tree) are of general interest for user programs, as are the
19717 the full internal tree, and the entity table (the symbol table
19718 information). The reconstructed source provides a readable version of the
19719 program after the front-end has completed analysis and expansion,
19720 and is useful when studying the performance of specific constructs.
19721 For example, constraint checks are indicated, complex aggregates
19722 are replaced with loops and assignments, and tasking primitives
19723 are replaced with run-time calls.
19725 @node Stack Traceback
19726 @section Stack Traceback
19728 @cindex stack traceback
19729 @cindex stack unwinding
19732 Traceback is a mechanism to display the sequence of subprogram calls that
19733 leads to a specified execution point in a program. Often (but not always)
19734 the execution point is an instruction at which an exception has been raised.
19735 This mechanism is also known as @i{stack unwinding} because it obtains
19736 its information by scanning the run-time stack and recovering the activation
19737 records of all active subprograms. Stack unwinding is one of the most
19738 important tools for program debugging.
19740 The first entry stored in traceback corresponds to the deepest calling level,
19741 that is to say the subprogram currently executing the instruction
19742 from which we want to obtain the traceback.
19744 Note that there is no runtime performance penalty when stack traceback
19745 is enabled, and no exception is raised during program execution.
19748 * Non-Symbolic Traceback::
19749 * Symbolic Traceback::
19752 @node Non-Symbolic Traceback
19753 @subsection Non-Symbolic Traceback
19754 @cindex traceback, non-symbolic
19757 Note: this feature is not supported on all platforms. See
19758 @file{GNAT.Traceback spec in g-traceb.ads} for a complete list of supported
19762 * Tracebacks From an Unhandled Exception::
19763 * Tracebacks From Exception Occurrences (non-symbolic)::
19764 * Tracebacks From Anywhere in a Program (non-symbolic)::
19767 @node Tracebacks From an Unhandled Exception
19768 @subsubsection Tracebacks From an Unhandled Exception
19771 A runtime non-symbolic traceback is a list of addresses of call instructions.
19772 To enable this feature you must use the @option{-E}
19773 @code{gnatbind}'s option. With this option a stack traceback is stored as part
19774 of exception information. You can retrieve this information using the
19775 @code{addr2line} tool.
19777 Here is a simple example:
19779 @smallexample @c ada
19785 raise Constraint_Error;
19800 $ gnatmake stb -bargs -E
19803 Execution terminated by unhandled exception
19804 Exception name: CONSTRAINT_ERROR
19806 Call stack traceback locations:
19807 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
19811 As we see the traceback lists a sequence of addresses for the unhandled
19812 exception @code{CONSTRAINT_ERROR} raised in procedure P1. It is easy to
19813 guess that this exception come from procedure P1. To translate these
19814 addresses into the source lines where the calls appear, the
19815 @code{addr2line} tool, described below, is invaluable. The use of this tool
19816 requires the program to be compiled with debug information.
19819 $ gnatmake -g stb -bargs -E
19822 Execution terminated by unhandled exception
19823 Exception name: CONSTRAINT_ERROR
19825 Call stack traceback locations:
19826 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
19828 $ addr2line --exe=stb 0x401373 0x40138b 0x40139c 0x401335 0x4011c4
19829 0x4011f1 0x77e892a4
19831 00401373 at d:/stb/stb.adb:5
19832 0040138B at d:/stb/stb.adb:10
19833 0040139C at d:/stb/stb.adb:14
19834 00401335 at d:/stb/b~stb.adb:104
19835 004011C4 at /build/.../crt1.c:200
19836 004011F1 at /build/.../crt1.c:222
19837 77E892A4 in ?? at ??:0
19841 The @code{addr2line} tool has several other useful options:
19845 to get the function name corresponding to any location
19847 @item --demangle=gnat
19848 to use the gnat decoding mode for the function names. Note that
19849 for binutils version 2.9.x the option is simply @option{--demangle}.
19853 $ addr2line --exe=stb --functions --demangle=gnat 0x401373 0x40138b
19854 0x40139c 0x401335 0x4011c4 0x4011f1
19856 00401373 in stb.p1 at d:/stb/stb.adb:5
19857 0040138B in stb.p2 at d:/stb/stb.adb:10
19858 0040139C in stb at d:/stb/stb.adb:14
19859 00401335 in main at d:/stb/b~stb.adb:104
19860 004011C4 in <__mingw_CRTStartup> at /build/.../crt1.c:200
19861 004011F1 in <mainCRTStartup> at /build/.../crt1.c:222
19865 From this traceback we can see that the exception was raised in
19866 @file{stb.adb} at line 5, which was reached from a procedure call in
19867 @file{stb.adb} at line 10, and so on. The @file{b~std.adb} is the binder file,
19868 which contains the call to the main program.
19869 @xref{Running gnatbind}. The remaining entries are assorted runtime routines,
19870 and the output will vary from platform to platform.
19872 It is also possible to use @code{GDB} with these traceback addresses to debug
19873 the program. For example, we can break at a given code location, as reported
19874 in the stack traceback:
19880 Furthermore, this feature is not implemented inside Windows DLL. Only
19881 the non-symbolic traceback is reported in this case.
19884 (gdb) break *0x401373
19885 Breakpoint 1 at 0x401373: file stb.adb, line 5.
19889 It is important to note that the stack traceback addresses
19890 do not change when debug information is included. This is particularly useful
19891 because it makes it possible to release software without debug information (to
19892 minimize object size), get a field report that includes a stack traceback
19893 whenever an internal bug occurs, and then be able to retrieve the sequence
19894 of calls with the same program compiled with debug information.
19896 @node Tracebacks From Exception Occurrences (non-symbolic)
19897 @subsubsection Tracebacks From Exception Occurrences
19900 Non-symbolic tracebacks are obtained by using the @option{-E} binder argument.
19901 The stack traceback is attached to the exception information string, and can
19902 be retrieved in an exception handler within the Ada program, by means of the
19903 Ada95 facilities defined in @code{Ada.Exceptions}. Here is a simple example:
19905 @smallexample @c ada
19907 with Ada.Exceptions;
19912 use Ada.Exceptions;
19920 Text_IO.Put_Line (Exception_Information (E));
19934 This program will output:
19939 Exception name: CONSTRAINT_ERROR
19940 Message: stb.adb:12
19941 Call stack traceback locations:
19942 0x4015e4 0x401633 0x401644 0x401461 0x4011c4 0x4011f1 0x77e892a4
19945 @node Tracebacks From Anywhere in a Program (non-symbolic)
19946 @subsubsection Tracebacks From Anywhere in a Program
19949 It is also possible to retrieve a stack traceback from anywhere in a
19950 program. For this you need to
19951 use the @code{GNAT.Traceback} API. This package includes a procedure called
19952 @code{Call_Chain} that computes a complete stack traceback, as well as useful
19953 display procedures described below. It is not necessary to use the
19954 @option{-E gnatbind} option in this case, because the stack traceback mechanism
19955 is invoked explicitly.
19958 In the following example we compute a traceback at a specific location in
19959 the program, and we display it using @code{GNAT.Debug_Utilities.Image} to
19960 convert addresses to strings:
19962 @smallexample @c ada
19964 with GNAT.Traceback;
19965 with GNAT.Debug_Utilities;
19971 use GNAT.Traceback;
19974 TB : Tracebacks_Array (1 .. 10);
19975 -- We are asking for a maximum of 10 stack frames.
19977 -- Len will receive the actual number of stack frames returned.
19979 Call_Chain (TB, Len);
19981 Text_IO.Put ("In STB.P1 : ");
19983 for K in 1 .. Len loop
19984 Text_IO.Put (Debug_Utilities.Image (TB (K)));
20005 In STB.P1 : 16#0040_F1E4# 16#0040_14F2# 16#0040_170B# 16#0040_171C#
20006 16#0040_1461# 16#0040_11C4# 16#0040_11F1# 16#77E8_92A4#
20010 You can then get further information by invoking the @code{addr2line}
20011 tool as described earlier (note that the hexadecimal addresses
20012 need to be specified in C format, with a leading ``0x'').
20014 @node Symbolic Traceback
20015 @subsection Symbolic Traceback
20016 @cindex traceback, symbolic
20019 A symbolic traceback is a stack traceback in which procedure names are
20020 associated with each code location.
20023 Note that this feature is not supported on all platforms. See
20024 @file{GNAT.Traceback.Symbolic spec in g-trasym.ads} for a complete
20025 list of currently supported platforms.
20028 Note that the symbolic traceback requires that the program be compiled
20029 with debug information. If it is not compiled with debug information
20030 only the non-symbolic information will be valid.
20033 * Tracebacks From Exception Occurrences (symbolic)::
20034 * Tracebacks From Anywhere in a Program (symbolic)::
20037 @node Tracebacks From Exception Occurrences (symbolic)
20038 @subsubsection Tracebacks From Exception Occurrences
20040 @smallexample @c ada
20042 with GNAT.Traceback.Symbolic;
20048 raise Constraint_Error;
20065 Ada.Text_IO.Put_Line (GNAT.Traceback.Symbolic.Symbolic_Traceback (E));
20070 $ gnatmake -g .\stb -bargs -E -largs -lgnat -laddr2line -lintl
20073 0040149F in stb.p1 at stb.adb:8
20074 004014B7 in stb.p2 at stb.adb:13
20075 004014CF in stb.p3 at stb.adb:18
20076 004015DD in ada.stb at stb.adb:22
20077 00401461 in main at b~stb.adb:168
20078 004011C4 in __mingw_CRTStartup at crt1.c:200
20079 004011F1 in mainCRTStartup at crt1.c:222
20080 77E892A4 in ?? at ??:0
20084 In the above example the ``.\'' syntax in the @command{gnatmake} command
20085 is currently required by @command{addr2line} for files that are in
20086 the current working directory.
20087 Moreover, the exact sequence of linker options may vary from platform
20089 The above @option{-largs} section is for Windows platforms. By contrast,
20090 under Unix there is no need for the @option{-largs} section.
20091 Differences across platforms are due to details of linker implementation.
20093 @node Tracebacks From Anywhere in a Program (symbolic)
20094 @subsubsection Tracebacks From Anywhere in a Program
20097 It is possible to get a symbolic stack traceback
20098 from anywhere in a program, just as for non-symbolic tracebacks.
20099 The first step is to obtain a non-symbolic
20100 traceback, and then call @code{Symbolic_Traceback} to compute the symbolic
20101 information. Here is an example:
20103 @smallexample @c ada
20105 with GNAT.Traceback;
20106 with GNAT.Traceback.Symbolic;
20111 use GNAT.Traceback;
20112 use GNAT.Traceback.Symbolic;
20115 TB : Tracebacks_Array (1 .. 10);
20116 -- We are asking for a maximum of 10 stack frames.
20118 -- Len will receive the actual number of stack frames returned.
20120 Call_Chain (TB, Len);
20121 Text_IO.Put_Line (Symbolic_Traceback (TB (1 .. Len)));
20135 @node Compatibility with DEC Ada
20136 @chapter Compatibility with DEC Ada
20137 @cindex Compatibility
20140 This section of the manual compares DEC Ada for OpenVMS Alpha and GNAT
20141 OpenVMS Alpha. GNAT achieves a high level of compatibility
20142 with DEC Ada, and it should generally be straightforward to port code
20143 from the DEC Ada environment to GNAT. However, there are a few language
20144 and implementation differences of which the user must be aware. These
20145 differences are discussed in this section. In
20146 addition, the operating environment and command structure for the
20147 compiler are different, and these differences are also discussed.
20149 Note that this discussion addresses specifically the implementation
20150 of Ada 83 for DIGITAL OpenVMS Alpha Systems. In cases where the implementation
20151 of DEC Ada differs between OpenVMS Alpha Systems and OpenVMS VAX Systems,
20152 GNAT always follows the Alpha implementation.
20155 * Ada 95 Compatibility::
20156 * Differences in the Definition of Package System::
20157 * Language-Related Features::
20158 * The Package STANDARD::
20159 * The Package SYSTEM::
20160 * Tasking and Task-Related Features::
20161 * Implementation of Tasks in DEC Ada for OpenVMS Alpha Systems::
20162 * Pragmas and Pragma-Related Features::
20163 * Library of Predefined Units::
20165 * Main Program Definition::
20166 * Implementation-Defined Attributes::
20167 * Compiler and Run-Time Interfacing::
20168 * Program Compilation and Library Management::
20170 * Implementation Limits::
20174 @node Ada 95 Compatibility
20175 @section Ada 95 Compatibility
20178 GNAT is an Ada 95 compiler, and DEC Ada is an Ada 83
20179 compiler. Ada 95 is almost completely upwards compatible
20180 with Ada 83, and therefore Ada 83 programs will compile
20181 and run under GNAT with
20182 no changes or only minor changes. The Ada 95 Reference
20183 Manual (ANSI/ISO/IEC-8652:1995) provides details on specific
20186 GNAT provides the switch /83 on the GNAT COMPILE command,
20187 as well as the pragma ADA_83, to force the compiler to
20188 operate in Ada 83 mode. This mode does not guarantee complete
20189 conformance to Ada 83, but in practice is sufficient to
20190 eliminate most sources of incompatibilities.
20191 In particular, it eliminates the recognition of the
20192 additional Ada 95 keywords, so that their use as identifiers
20193 in Ada83 program is legal, and handles the cases of packages
20194 with optional bodies, and generics that instantiate unconstrained
20195 types without the use of @code{(<>)}.
20197 @node Differences in the Definition of Package System
20198 @section Differences in the Definition of Package System
20201 Both the Ada 95 and Ada 83 reference manuals permit a compiler to add
20202 implementation-dependent declarations to package System. In normal mode,
20203 GNAT does not take advantage of this permission, and the version of System
20204 provided by GNAT exactly matches that in the Ada 95 Reference Manual.
20206 However, DEC Ada adds an extensive set of declarations to package System,
20207 as fully documented in the DEC Ada manuals. To minimize changes required
20208 for programs that make use of these extensions, GNAT provides the pragma
20209 Extend_System for extending the definition of package System. By using:
20211 @smallexample @c ada
20214 pragma Extend_System (Aux_DEC);
20220 The set of definitions in System is extended to include those in package
20221 @code{System.Aux_DEC}.
20222 These definitions are incorporated directly into package
20223 System, as though they had been declared there in the first place. For a
20224 list of the declarations added, see the specification of this package,
20225 which can be found in the file @code{s-auxdec.ads} in the GNAT library.
20226 The pragma Extend_System is a configuration pragma, which means that
20227 it can be placed in the file @file{gnat.adc}, so that it will automatically
20228 apply to all subsequent compilations. See the section on Configuration
20229 Pragmas for further details.
20231 An alternative approach that avoids the use of the non-standard
20232 Extend_System pragma is to add a context clause to the unit that
20233 references these facilities:
20235 @smallexample @c ada
20238 with System.Aux_DEC;
20239 use System.Aux_DEC;
20245 The effect is not quite semantically identical to incorporating
20246 the declarations directly into package @code{System},
20247 but most programs will not notice a difference
20248 unless they use prefix notation (e.g. @code{System.Integer_8})
20250 entities directly in package @code{System}.
20251 For units containing such references,
20252 the prefixes must either be removed, or the pragma @code{Extend_System}
20255 @node Language-Related Features
20256 @section Language-Related Features
20259 The following sections highlight differences in types,
20260 representations of types, operations, alignment, and
20264 * Integer Types and Representations::
20265 * Floating-Point Types and Representations::
20266 * Pragmas Float_Representation and Long_Float::
20267 * Fixed-Point Types and Representations::
20268 * Record and Array Component Alignment::
20269 * Address Clauses::
20270 * Other Representation Clauses::
20273 @node Integer Types and Representations
20274 @subsection Integer Types and Representations
20277 The set of predefined integer types is identical in DEC Ada and GNAT.
20278 Furthermore the representation of these integer types is also identical,
20279 including the capability of size clauses forcing biased representation.
20282 DEC Ada for OpenVMS Alpha systems has defined the
20283 following additional integer types in package System:
20304 When using GNAT, the first four of these types may be obtained from the
20305 standard Ada 95 package @code{Interfaces}.
20306 Alternatively, by use of the pragma
20307 @code{Extend_System}, identical
20308 declarations can be referenced directly in package @code{System}.
20309 On both GNAT and DEC Ada, the maximum integer size is 64 bits.
20311 @node Floating-Point Types and Representations
20312 @subsection Floating-Point Types and Representations
20313 @cindex Floating-Point types
20316 The set of predefined floating-point types is identical in DEC Ada and GNAT.
20317 Furthermore the representation of these floating-point
20318 types is also identical. One important difference is that the default
20319 representation for DEC Ada is VAX_Float, but the default representation
20322 Specific types may be declared to be VAX_Float or IEEE, using the pragma
20323 @code{Float_Representation} as described in the DEC Ada documentation.
20324 For example, the declarations:
20326 @smallexample @c ada
20329 type F_Float is digits 6;
20330 pragma Float_Representation (VAX_Float, F_Float);
20336 declare a type F_Float that will be represented in VAX_Float format.
20337 This set of declarations actually appears in System.Aux_DEC, which provides
20338 the full set of additional floating-point declarations provided in
20339 the DEC Ada version of package
20340 System. This and similar declarations may be accessed in a user program
20341 by using pragma @code{Extend_System}. The use of this
20342 pragma, and the related pragma @code{Long_Float} is described in further
20343 detail in the following section.
20345 @node Pragmas Float_Representation and Long_Float
20346 @subsection Pragmas Float_Representation and Long_Float
20349 DEC Ada provides the pragma @code{Float_Representation}, which
20350 acts as a program library switch to allow control over
20351 the internal representation chosen for the predefined
20352 floating-point types declared in the package @code{Standard}.
20353 The format of this pragma is as follows:
20358 @b{pragma} @code{Float_Representation}(VAX_Float | IEEE_Float);
20364 This pragma controls the representation of floating-point
20369 @code{VAX_Float} specifies that floating-point
20370 types are represented by default with the VAX hardware types
20371 F-floating, D-floating, G-floating. Note that the H-floating
20372 type is available only on DIGITAL Vax systems, and is not available
20373 in either DEC Ada or GNAT for Alpha systems.
20376 @code{IEEE_Float} specifies that floating-point
20377 types are represented by default with the IEEE single and
20378 double floating-point types.
20382 GNAT provides an identical implementation of the pragma
20383 @code{Float_Representation}, except that it functions as a
20384 configuration pragma, as defined by Ada 95. Note that the
20385 notion of configuration pragma corresponds closely to the
20386 DEC Ada notion of a program library switch.
20388 When no pragma is used in GNAT, the default is IEEE_Float, which is different
20389 from DEC Ada 83, where the default is VAX_Float. In addition, the
20390 predefined libraries in GNAT are built using IEEE_Float, so it is not
20391 advisable to change the format of numbers passed to standard library
20392 routines, and if necessary explicit type conversions may be needed.
20394 The use of IEEE_Float is recommended in GNAT since it is more efficient,
20395 and (given that it conforms to an international standard) potentially more
20396 portable. The situation in which VAX_Float may be useful is in interfacing
20397 to existing code and data that expects the use of VAX_Float. There are
20398 two possibilities here. If the requirement for the use of VAX_Float is
20399 localized, then the best approach is to use the predefined VAX_Float
20400 types in package @code{System}, as extended by
20401 @code{Extend_System}. For example, use @code{System.F_Float}
20402 to specify the 32-bit @code{F-Float} format.
20404 Alternatively, if an entire program depends heavily on the use of
20405 the @code{VAX_Float} and in particular assumes that the types in
20406 package @code{Standard} are in @code{Vax_Float} format, then it
20407 may be desirable to reconfigure GNAT to assume Vax_Float by default.
20408 This is done by using the GNAT LIBRARY command to rebuild the library, and
20409 then using the general form of the @code{Float_Representation}
20410 pragma to ensure that this default format is used throughout.
20411 The form of the GNAT LIBRARY command is:
20414 GNAT LIBRARY /CONFIG=@i{file} /CREATE=@i{directory}
20418 where @i{file} contains the new configuration pragmas
20419 and @i{directory} is the directory to be created to contain
20423 On OpenVMS systems, DEC Ada provides the pragma @code{Long_Float}
20424 to allow control over the internal representation chosen
20425 for the predefined type @code{Long_Float} and for floating-point
20426 type declarations with digits specified in the range 7 .. 15.
20427 The format of this pragma is as follows:
20429 @smallexample @c ada
20431 pragma Long_Float (D_FLOAT | G_FLOAT);
20435 @node Fixed-Point Types and Representations
20436 @subsection Fixed-Point Types and Representations
20439 On DEC Ada for OpenVMS Alpha systems, rounding is
20440 away from zero for both positive and negative numbers.
20441 Therefore, +0.5 rounds to 1 and -0.5 rounds to -1.
20443 On GNAT for OpenVMS Alpha, the results of operations
20444 on fixed-point types are in accordance with the Ada 95
20445 rules. In particular, results of operations on decimal
20446 fixed-point types are truncated.
20448 @node Record and Array Component Alignment
20449 @subsection Record and Array Component Alignment
20452 On DEC Ada for OpenVMS Alpha, all non composite components
20453 are aligned on natural boundaries. For example, 1-byte
20454 components are aligned on byte boundaries, 2-byte
20455 components on 2-byte boundaries, 4-byte components on 4-byte
20456 byte boundaries, and so on. The OpenVMS Alpha hardware
20457 runs more efficiently with naturally aligned data.
20459 ON GNAT for OpenVMS Alpha, alignment rules are compatible
20460 with DEC Ada for OpenVMS Alpha.
20462 @node Address Clauses
20463 @subsection Address Clauses
20466 In DEC Ada and GNAT, address clauses are supported for
20467 objects and imported subprograms.
20468 The predefined type @code{System.Address} is a private type
20469 in both compilers, with the same representation (it is simply
20470 a machine pointer). Addition, subtraction, and comparison
20471 operations are available in the standard Ada 95 package
20472 @code{System.Storage_Elements}, or in package @code{System}
20473 if it is extended to include @code{System.Aux_DEC} using a
20474 pragma @code{Extend_System} as previously described.
20476 Note that code that with's both this extended package @code{System}
20477 and the package @code{System.Storage_Elements} should not @code{use}
20478 both packages, or ambiguities will result. In general it is better
20479 not to mix these two sets of facilities. The Ada 95 package was
20480 designed specifically to provide the kind of features that DEC Ada
20481 adds directly to package @code{System}.
20483 GNAT is compatible with DEC Ada in its handling of address
20484 clauses, except for some limitations in
20485 the form of address clauses for composite objects with
20486 initialization. Such address clauses are easily replaced
20487 by the use of an explicitly-defined constant as described
20488 in the Ada 95 Reference Manual (13.1(22)). For example, the sequence
20491 @smallexample @c ada
20493 X, Y : Integer := Init_Func;
20494 Q : String (X .. Y) := "abc";
20496 for Q'Address use Compute_Address;
20501 will be rejected by GNAT, since the address cannot be computed at the time
20502 that Q is declared. To achieve the intended effect, write instead:
20504 @smallexample @c ada
20507 X, Y : Integer := Init_Func;
20508 Q_Address : constant Address := Compute_Address;
20509 Q : String (X .. Y) := "abc";
20511 for Q'Address use Q_Address;
20517 which will be accepted by GNAT (and other Ada 95 compilers), and is also
20518 backwards compatible with Ada 83. A fuller description of the restrictions
20519 on address specifications is found in the GNAT Reference Manual.
20521 @node Other Representation Clauses
20522 @subsection Other Representation Clauses
20525 GNAT supports in a compatible manner all the representation
20526 clauses supported by DEC Ada. In addition, it
20527 supports representation clause forms that are new in Ada 95
20528 including COMPONENT_SIZE and SIZE clauses for objects.
20530 @node The Package STANDARD
20531 @section The Package STANDARD
20534 The package STANDARD, as implemented by DEC Ada, is fully
20535 described in the Reference Manual for the Ada Programming
20536 Language (ANSI/MIL-STD-1815A-1983) and in the DEC Ada
20537 Language Reference Manual. As implemented by GNAT, the
20538 package STANDARD is described in the Ada 95 Reference
20541 In addition, DEC Ada supports the Latin-1 character set in
20542 the type CHARACTER. GNAT supports the Latin-1 character set
20543 in the type CHARACTER and also Unicode (ISO 10646 BMP) in
20544 the type WIDE_CHARACTER.
20546 The floating-point types supported by GNAT are those
20547 supported by DEC Ada, but defaults are different, and are controlled by
20548 pragmas. See @ref{Floating-Point Types and Representations} for details.
20550 @node The Package SYSTEM
20551 @section The Package SYSTEM
20554 DEC Ada provides a system-specific version of the package
20555 SYSTEM for each platform on which the language ships.
20556 For the complete specification of the package SYSTEM, see
20557 Appendix F of the DEC Ada Language Reference Manual.
20559 On DEC Ada, the package SYSTEM includes the following conversion functions:
20561 @item TO_ADDRESS(INTEGER)
20563 @item TO_ADDRESS(UNSIGNED_LONGWORD)
20565 @item TO_ADDRESS(universal_integer)
20567 @item TO_INTEGER(ADDRESS)
20569 @item TO_UNSIGNED_LONGWORD(ADDRESS)
20571 @item Function IMPORT_VALUE return UNSIGNED_LONGWORD and the
20572 functions IMPORT_ADDRESS and IMPORT_LARGEST_VALUE
20576 By default, GNAT supplies a version of SYSTEM that matches
20577 the definition given in the Ada 95 Reference Manual.
20579 is a subset of the DIGITAL system definitions, which is as
20580 close as possible to the original definitions. The only difference
20581 is that the definition of SYSTEM_NAME is different:
20583 @smallexample @c ada
20586 type Name is (SYSTEM_NAME_GNAT);
20587 System_Name : constant Name := SYSTEM_NAME_GNAT;
20593 Also, GNAT adds the new Ada 95 declarations for
20594 BIT_ORDER and DEFAULT_BIT_ORDER.
20596 However, the use of the following pragma causes GNAT
20597 to extend the definition of package SYSTEM so that it
20598 encompasses the full set of DIGITAL-specific extensions,
20599 including the functions listed above:
20601 @smallexample @c ada
20603 pragma Extend_System (Aux_DEC);
20608 The pragma Extend_System is a configuration pragma that
20609 is most conveniently placed in the @file{gnat.adc} file. See the
20610 GNAT Reference Manual for further details.
20612 DEC Ada does not allow the recompilation of the package
20613 SYSTEM. Instead DEC Ada provides several pragmas (SYSTEM_
20614 NAME, STORAGE_UNIT, and MEMORY_SIZE) to modify values in
20615 the package SYSTEM. On OpenVMS Alpha systems, the pragma
20616 SYSTEM_NAME takes the enumeration literal OPENVMS_AXP as
20617 its single argument.
20619 GNAT does permit the recompilation of package SYSTEM using
20620 a special switch (@option{-gnatg}) and this switch can be used if
20621 it is necessary to modify the definitions in SYSTEM. GNAT does
20622 not permit the specification of SYSTEM_NAME, STORAGE_UNIT
20623 or MEMORY_SIZE by any other means.
20625 On GNAT systems, the pragma SYSTEM_NAME takes the
20626 enumeration literal SYSTEM_NAME_GNAT.
20628 The definitions provided by the use of
20630 @smallexample @c ada
20631 pragma Extend_System (AUX_Dec);
20635 are virtually identical to those provided by the DEC Ada 83 package
20636 System. One important difference is that the name of the TO_ADDRESS
20637 function for type UNSIGNED_LONGWORD is changed to TO_ADDRESS_LONG.
20638 See the GNAT Reference manual for a discussion of why this change was
20642 The version of TO_ADDRESS taking a universal integer argument is in fact
20643 an extension to Ada 83 not strictly compatible with the reference manual.
20644 In GNAT, we are constrained to be exactly compatible with the standard,
20645 and this means we cannot provide this capability. In DEC Ada 83, the
20646 point of this definition is to deal with a call like:
20648 @smallexample @c ada
20649 TO_ADDRESS (16#12777#);
20653 Normally, according to the Ada 83 standard, one would expect this to be
20654 ambiguous, since it matches both the INTEGER and UNSIGNED_LONGWORD forms
20655 of TO_ADDRESS. However, in DEC Ada 83, there is no ambiguity, since the
20656 definition using universal_integer takes precedence.
20658 In GNAT, since the version with universal_integer cannot be supplied, it is
20659 not possible to be 100% compatible. Since there are many programs using
20660 numeric constants for the argument to TO_ADDRESS, the decision in GNAT was
20661 to change the name of the function in the UNSIGNED_LONGWORD case, so the
20662 declarations provided in the GNAT version of AUX_Dec are:
20664 @smallexample @c ada
20665 function To_Address (X : Integer) return Address;
20666 pragma Pure_Function (To_Address);
20668 function To_Address_Long (X : Unsigned_Longword) return Address;
20669 pragma Pure_Function (To_Address_Long);
20673 This means that programs using TO_ADDRESS for UNSIGNED_LONGWORD must
20674 change the name to TO_ADDRESS_LONG.
20676 @node Tasking and Task-Related Features
20677 @section Tasking and Task-Related Features
20680 The concepts relevant to a comparison of tasking on GNAT
20681 and on DEC Ada for OpenVMS Alpha systems are discussed in
20682 the following sections.
20684 For detailed information on concepts related to tasking in
20685 DEC Ada, see the DEC Ada Language Reference Manual and the
20686 relevant run-time reference manual.
20688 @node Implementation of Tasks in DEC Ada for OpenVMS Alpha Systems
20689 @section Implementation of Tasks in DEC Ada for OpenVMS Alpha Systems
20692 On OpenVMS Alpha systems, each Ada task (except a passive
20693 task) is implemented as a single stream of execution
20694 that is created and managed by the kernel. On these
20695 systems, DEC Ada tasking support is based on DECthreads,
20696 an implementation of the POSIX standard for threads.
20698 Although tasks are implemented as threads, all tasks in
20699 an Ada program are part of the same process. As a result,
20700 resources such as open files and virtual memory can be
20701 shared easily among tasks. Having all tasks in one process
20702 allows better integration with the programming environment
20703 (the shell and the debugger, for example).
20705 Also, on OpenVMS Alpha systems, DEC Ada tasks and foreign
20706 code that calls DECthreads routines can be used together.
20707 The interaction between Ada tasks and DECthreads routines
20708 can have some benefits. For example when on OpenVMS Alpha,
20709 DEC Ada can call C code that is already threaded.
20710 GNAT on OpenVMS Alpha uses the facilities of DECthreads,
20711 and Ada tasks are mapped to threads.
20714 * Assigning Task IDs::
20715 * Task IDs and Delays::
20716 * Task-Related Pragmas::
20717 * Scheduling and Task Priority::
20719 * External Interrupts::
20722 @node Assigning Task IDs
20723 @subsection Assigning Task IDs
20726 The DEC Ada Run-Time Library always assigns %TASK 1 to
20727 the environment task that executes the main program. On
20728 OpenVMS Alpha systems, %TASK 0 is often used for tasks
20729 that have been created but are not yet activated.
20731 On OpenVMS Alpha systems, task IDs are assigned at
20732 activation. On GNAT systems, task IDs are also assigned at
20733 task creation but do not have the same form or values as
20734 task ID values in DEC Ada. There is no null task, and the
20735 environment task does not have a specific task ID value.
20737 @node Task IDs and Delays
20738 @subsection Task IDs and Delays
20741 On OpenVMS Alpha systems, tasking delays are implemented
20742 using Timer System Services. The Task ID is used for the
20743 identification of the timer request (the REQIDT parameter).
20744 If Timers are used in the application take care not to use
20745 0 for the identification, because cancelling such a timer
20746 will cancel all timers and may lead to unpredictable results.
20748 @node Task-Related Pragmas
20749 @subsection Task-Related Pragmas
20752 Ada supplies the pragma TASK_STORAGE, which allows
20753 specification of the size of the guard area for a task
20754 stack. (The guard area forms an area of memory that has no
20755 read or write access and thus helps in the detection of
20756 stack overflow.) On OpenVMS Alpha systems, if the pragma
20757 TASK_STORAGE specifies a value of zero, a minimal guard
20758 area is created. In the absence of a pragma TASK_STORAGE, a default guard
20761 GNAT supplies the following task-related pragmas:
20766 This pragma appears within a task definition and
20767 applies to the task in which it appears. The argument
20768 must be of type SYSTEM.TASK_INFO.TASK_INFO_TYPE.
20772 GNAT implements pragma TASK_STORAGE in the same way as
20774 Both DEC Ada and GNAT supply the pragmas PASSIVE,
20775 SUPPRESS, and VOLATILE.
20777 @node Scheduling and Task Priority
20778 @subsection Scheduling and Task Priority
20781 DEC Ada implements the Ada language requirement that
20782 when two tasks are eligible for execution and they have
20783 different priorities, the lower priority task does not
20784 execute while the higher priority task is waiting. The DEC
20785 Ada Run-Time Library keeps a task running until either the
20786 task is suspended or a higher priority task becomes ready.
20788 On OpenVMS Alpha systems, the default strategy is round-
20789 robin with preemption. Tasks of equal priority take turns
20790 at the processor. A task is run for a certain period of
20791 time and then placed at the rear of the ready queue for
20792 its priority level.
20794 DEC Ada provides the implementation-defined pragma TIME_SLICE,
20795 which can be used to enable or disable round-robin
20796 scheduling of tasks with the same priority.
20797 See the relevant DEC Ada run-time reference manual for
20798 information on using the pragmas to control DEC Ada task
20801 GNAT follows the scheduling rules of Annex D (real-time
20802 Annex) of the Ada 95 Reference Manual. In general, this
20803 scheduling strategy is fully compatible with DEC Ada
20804 although it provides some additional constraints (as
20805 fully documented in Annex D).
20806 GNAT implements time slicing control in a manner compatible with
20807 DEC Ada 83, by means of the pragma Time_Slice, whose semantics are identical
20808 to the DEC Ada 83 pragma of the same name.
20809 Note that it is not possible to mix GNAT tasking and
20810 DEC Ada 83 tasking in the same program, since the two run times are
20813 @node The Task Stack
20814 @subsection The Task Stack
20817 In DEC Ada, a task stack is allocated each time a
20818 non passive task is activated. As soon as the task is
20819 terminated, the storage for the task stack is deallocated.
20820 If you specify a size of zero (bytes) with T'STORAGE_SIZE,
20821 a default stack size is used. Also, regardless of the size
20822 specified, some additional space is allocated for task
20823 management purposes. On OpenVMS Alpha systems, at least
20824 one page is allocated.
20826 GNAT handles task stacks in a similar manner. According to
20827 the Ada 95 rules, it provides the pragma STORAGE_SIZE as
20828 an alternative method for controlling the task stack size.
20829 The specification of the attribute T'STORAGE_SIZE is also
20830 supported in a manner compatible with DEC Ada.
20832 @node External Interrupts
20833 @subsection External Interrupts
20836 On DEC Ada, external interrupts can be associated with task entries.
20837 GNAT is compatible with DEC Ada in its handling of external interrupts.
20839 @node Pragmas and Pragma-Related Features
20840 @section Pragmas and Pragma-Related Features
20843 Both DEC Ada and GNAT supply all language-defined pragmas
20844 as specified by the Ada 83 standard. GNAT also supplies all
20845 language-defined pragmas specified in the Ada 95 Reference Manual.
20846 In addition, GNAT implements the implementation-defined pragmas
20852 @item COMMON_OBJECT
20854 @item COMPONENT_ALIGNMENT
20856 @item EXPORT_EXCEPTION
20858 @item EXPORT_FUNCTION
20860 @item EXPORT_OBJECT
20862 @item EXPORT_PROCEDURE
20864 @item EXPORT_VALUED_PROCEDURE
20866 @item FLOAT_REPRESENTATION
20870 @item IMPORT_EXCEPTION
20872 @item IMPORT_FUNCTION
20874 @item IMPORT_OBJECT
20876 @item IMPORT_PROCEDURE
20878 @item IMPORT_VALUED_PROCEDURE
20880 @item INLINE_GENERIC
20882 @item INTERFACE_NAME
20892 @item SHARE_GENERIC
20904 These pragmas are all fully implemented, with the exception of @code{Title},
20905 @code{Passive}, and @code{Share_Generic}, which are
20906 recognized, but which have no
20907 effect in GNAT. The effect of @code{Passive} may be obtained by the
20908 use of protected objects in Ada 95. In GNAT, all generics are inlined.
20910 Unlike DEC Ada, the GNAT 'EXPORT_@i{subprogram}' pragmas require
20911 a separate subprogram specification which must appear before the
20914 GNAT also supplies a number of implementation-defined pragmas as follows:
20916 @item C_PASS_BY_COPY
20918 @item EXTEND_SYSTEM
20920 @item SOURCE_FILE_NAME
20940 @item CPP_CONSTRUCTOR
20942 @item CPP_DESTRUCTOR
20952 @item LINKER_SECTION
20954 @item MACHINE_ATTRIBUTE
20958 @item PURE_FUNCTION
20960 @item SOURCE_REFERENCE
20964 @item UNCHECKED_UNION
20966 @item UNIMPLEMENTED_UNIT
20968 @item UNIVERSAL_DATA
20970 @item WEAK_EXTERNAL
20974 For full details on these GNAT implementation-defined pragmas, see
20975 the GNAT Reference Manual.
20978 * Restrictions on the Pragma INLINE::
20979 * Restrictions on the Pragma INTERFACE::
20980 * Restrictions on the Pragma SYSTEM_NAME::
20983 @node Restrictions on the Pragma INLINE
20984 @subsection Restrictions on the Pragma INLINE
20987 DEC Ada applies the following restrictions to the pragma INLINE:
20989 @item Parameters cannot be a task type.
20991 @item Function results cannot be task types, unconstrained
20992 array types, or unconstrained types with discriminants.
20994 @item Bodies cannot declare the following:
20996 @item Subprogram body or stub (imported subprogram is allowed)
21000 @item Generic declarations
21002 @item Instantiations
21006 @item Access types (types derived from access types allowed)
21008 @item Array or record types
21010 @item Dependent tasks
21012 @item Direct recursive calls of subprogram or containing
21013 subprogram, directly or via a renaming
21019 In GNAT, the only restriction on pragma INLINE is that the
21020 body must occur before the call if both are in the same
21021 unit, and the size must be appropriately small. There are
21022 no other specific restrictions which cause subprograms to
21023 be incapable of being inlined.
21025 @node Restrictions on the Pragma INTERFACE
21026 @subsection Restrictions on the Pragma INTERFACE
21029 The following lists and describes the restrictions on the
21030 pragma INTERFACE on DEC Ada and GNAT:
21032 @item Languages accepted: Ada, Bliss, C, Fortran, Default.
21033 Default is the default on OpenVMS Alpha systems.
21035 @item Parameter passing: Language specifies default
21036 mechanisms but can be overridden with an EXPORT pragma.
21039 @item Ada: Use internal Ada rules.
21041 @item Bliss, C: Parameters must be mode @code{in}; cannot be
21042 record or task type. Result cannot be a string, an
21043 array, or a record.
21045 @item Fortran: Parameters cannot be a task. Result cannot
21046 be a string, an array, or a record.
21051 GNAT is entirely upwards compatible with DEC Ada, and in addition allows
21052 record parameters for all languages.
21054 @node Restrictions on the Pragma SYSTEM_NAME
21055 @subsection Restrictions on the Pragma SYSTEM_NAME
21058 For DEC Ada for OpenVMS Alpha, the enumeration literal
21059 for the type NAME is OPENVMS_AXP. In GNAT, the enumeration
21060 literal for the type NAME is SYSTEM_NAME_GNAT.
21062 @node Library of Predefined Units
21063 @section Library of Predefined Units
21066 A library of predefined units is provided as part of the
21067 DEC Ada and GNAT implementations. DEC Ada does not provide
21068 the package MACHINE_CODE but instead recommends importing
21071 The GNAT versions of the DEC Ada Run-Time Library (ADA$PREDEFINED:)
21072 units are taken from the OpenVMS Alpha version, not the OpenVMS VAX
21073 version. During GNAT installation, the DEC Ada Predefined
21074 Library units are copied into the GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
21075 (aka DECLIB) directory and patched to remove Ada 95 incompatibilities
21076 and to make them interoperable with GNAT, @pxref{Changes to DECLIB}
21079 The GNAT RTL is contained in
21080 the GNU:[LIB.OPENVMS7_x.2_8_x.ADALIB] (aka ADALIB) directory and
21081 the default search path is set up to find DECLIB units in preference
21082 to ADALIB units with the same name (TEXT_IO, SEQUENTIAL_IO, and DIRECT_IO,
21085 However, it is possible to change the default so that the
21086 reverse is true, or even to mix them using child package
21087 notation. The DEC Ada 83 units are available as DEC.xxx where xxx
21088 is the package name, and the Ada units are available in the
21089 standard manner defined for Ada 95, that is to say as Ada.xxx. To
21090 change the default, set ADA_INCLUDE_PATH and ADA_OBJECTS_PATH
21091 appropriately. For example, to change the default to use the Ada95
21095 $ DEFINE ADA_INCLUDE_PATH GNU:[LIB.OPENVMS7_1.2_8_1.ADAINCLUDE],-
21096 GNU:[LIB.OPENVMS7_1.2_8_1.DECLIB]
21097 $ DEFINE ADA_OBJECTS_PATH GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB],-
21098 GNU:[LIB.OPENVMS7_1.2_8_1.DECLIB]
21102 * Changes to DECLIB::
21105 @node Changes to DECLIB
21106 @subsection Changes to DECLIB
21109 The changes made to the DEC Ada predefined library for GNAT and Ada 95
21110 compatibility are minor and include the following:
21113 @item Adjusting the location of pragmas and record representation
21114 clauses to obey Ada 95 rules
21116 @item Adding the proper notation to generic formal parameters
21117 that take unconstrained types in instantiation
21119 @item Adding pragma ELABORATE_BODY to package specifications
21120 that have package bodies not otherwise allowed
21122 @item Occurrences of the identifier @code{"PROTECTED"} are renamed to
21124 Currently these are found only in the STARLET package spec.
21128 None of the above changes is visible to users.
21134 On OpenVMS Alpha, DEC Ada provides the following strongly-typed bindings:
21137 @item Command Language Interpreter (CLI interface)
21139 @item DECtalk Run-Time Library (DTK interface)
21141 @item Librarian utility routines (LBR interface)
21143 @item General Purpose Run-Time Library (LIB interface)
21145 @item Math Run-Time Library (MTH interface)
21147 @item National Character Set Run-Time Library (NCS interface)
21149 @item Compiled Code Support Run-Time Library (OTS interface)
21151 @item Parallel Processing Run-Time Library (PPL interface)
21153 @item Screen Management Run-Time Library (SMG interface)
21155 @item Sort Run-Time Library (SOR interface)
21157 @item String Run-Time Library (STR interface)
21159 @item STARLET System Library
21162 @item X Window System Version 11R4 and 11R5 (X, XLIB interface)
21164 @item X Windows Toolkit (XT interface)
21166 @item X/Motif Version 1.1.3 and 1.2 (XM interface)
21170 GNAT provides implementations of these DEC bindings in the DECLIB directory.
21172 The X/Motif bindings used to build DECLIB are whatever versions are in the
21173 DEC Ada @file{ADA$PREDEFINED} directory with extension @file{.ADC}.
21174 The build script will
21175 automatically add a pragma Linker_Options to packages @code{Xm}, @code{Xt},
21177 causing the default X/Motif sharable image libraries to be linked in. This
21178 is done via options files named @file{xm.opt}, @file{xt.opt}, and
21179 @file{x_lib.opt} (also located in the @file{DECLIB} directory).
21181 It may be necessary to edit these options files to update or correct the
21182 library names if, for example, the newer X/Motif bindings from
21183 @file{ADA$EXAMPLES}
21184 had been (previous to installing GNAT) copied and renamed to supersede the
21185 default @file{ADA$PREDEFINED} versions.
21188 * Shared Libraries and Options Files::
21189 * Interfaces to C::
21192 @node Shared Libraries and Options Files
21193 @subsection Shared Libraries and Options Files
21196 When using the DEC Ada
21197 predefined X and Motif bindings, the linking with their sharable images is
21198 done automatically by @command{GNAT LINK}.
21199 When using other X and Motif bindings, you need
21200 to add the corresponding sharable images to the command line for
21201 @code{GNAT LINK}. When linking with shared libraries, or with
21202 @file{.OPT} files, you must
21203 also add them to the command line for @command{GNAT LINK}.
21205 A shared library to be used with GNAT is built in the same way as other
21206 libraries under VMS. The VMS Link command can be used in standard fashion.
21208 @node Interfaces to C
21209 @subsection Interfaces to C
21213 provides the following Ada types and operations:
21216 @item C types package (C_TYPES)
21218 @item C strings (C_TYPES.NULL_TERMINATED)
21220 @item Other_types (SHORT_INT)
21224 Interfacing to C with GNAT, one can use the above approach
21225 described for DEC Ada or the facilities of Annex B of
21226 the Ada 95 Reference Manual (packages INTERFACES.C,
21227 INTERFACES.C.STRINGS and INTERFACES.C.POINTERS). For more
21228 information, see the section ``Interfacing to C'' in the
21229 @cite{GNAT Reference Manual}.
21231 The @option{-gnatF} qualifier forces default and explicit
21232 @code{External_Name} parameters in pragmas Import and Export
21233 to be uppercased for compatibility with the default behavior
21234 of Compaq C. The qualifier has no effect on @code{Link_Name} parameters.
21236 @node Main Program Definition
21237 @section Main Program Definition
21240 The following section discusses differences in the
21241 definition of main programs on DEC Ada and GNAT.
21242 On DEC Ada, main programs are defined to meet the
21243 following conditions:
21245 @item Procedure with no formal parameters (returns 0 upon
21248 @item Procedure with no formal parameters (returns 42 when
21249 unhandled exceptions are raised)
21251 @item Function with no formal parameters whose returned value
21252 is of a discrete type
21254 @item Procedure with one OUT formal of a discrete type for
21255 which a specification of pragma EXPORT_VALUED_PROCEDURE is given.
21260 When declared with the pragma EXPORT_VALUED_PROCEDURE,
21261 a main function or main procedure returns a discrete
21262 value whose size is less than 64 bits (32 on VAX systems),
21263 the value is zero- or sign-extended as appropriate.
21264 On GNAT, main programs are defined as follows:
21266 @item Must be a non-generic, parameter-less subprogram that
21267 is either a procedure or function returning an Ada
21268 STANDARD.INTEGER (the predefined type)
21270 @item Cannot be a generic subprogram or an instantiation of a
21274 @node Implementation-Defined Attributes
21275 @section Implementation-Defined Attributes
21278 GNAT provides all DEC Ada implementation-defined
21281 @node Compiler and Run-Time Interfacing
21282 @section Compiler and Run-Time Interfacing
21285 DEC Ada provides the following ways to pass options to the linker
21288 @item /WAIT and /SUBMIT qualifiers
21290 @item /COMMAND qualifier
21292 @item /[NO]MAP qualifier
21294 @item /OUTPUT=file-spec
21296 @item /[NO]DEBUG and /[NO]TRACEBACK qualifiers
21300 To pass options to the linker, GNAT provides the following
21304 @item @option{/EXECUTABLE=exec-name}
21306 @item @option{/VERBOSE qualifier}
21308 @item @option{/[NO]DEBUG} and @option{/[NO]TRACEBACK} qualifiers
21312 For more information on these switches, see
21313 @ref{Switches for gnatlink}.
21314 In DEC Ada, the command-line switch @option{/OPTIMIZE} is available
21315 to control optimization. DEC Ada also supplies the
21318 @item @code{OPTIMIZE}
21320 @item @code{INLINE}
21322 @item @code{INLINE_GENERIC}
21324 @item @code{SUPPRESS_ALL}
21326 @item @code{PASSIVE}
21330 In GNAT, optimization is controlled strictly by command
21331 line parameters, as described in the corresponding section of this guide.
21332 The DIGITAL pragmas for control of optimization are
21333 recognized but ignored.
21335 Note that in GNAT, the default is optimization off, whereas in DEC Ada 83,
21336 the default is that optimization is turned on.
21338 @node Program Compilation and Library Management
21339 @section Program Compilation and Library Management
21342 DEC Ada and GNAT provide a comparable set of commands to
21343 build programs. DEC Ada also provides a program library,
21344 which is a concept that does not exist on GNAT. Instead,
21345 GNAT provides directories of sources that are compiled as
21348 The following table summarizes
21349 the DEC Ada commands and provides
21350 equivalent GNAT commands. In this table, some GNAT
21351 equivalents reflect the fact that GNAT does not use the
21352 concept of a program library. Instead, it uses a model
21353 in which collections of source and object files are used
21354 in a manner consistent with other languages like C and
21355 Fortran. Therefore, standard system file commands are used
21356 to manipulate these elements. Those GNAT commands are marked with
21358 Note that, unlike DEC Ada, none of the GNAT commands accepts wild cards.
21361 @multitable @columnfractions .35 .65
21363 @item @emph{DEC Ada Command}
21364 @tab @emph{GNAT Equivalent / Description}
21366 @item @command{ADA}
21367 @tab @command{GNAT COMPILE}@*
21368 Invokes the compiler to compile one or more Ada source files.
21370 @item @command{ACS ATTACH}@*
21371 @tab [No equivalent]@*
21372 Switches control of terminal from current process running the program
21375 @item @command{ACS CHECK}
21376 @tab @command{GNAT MAKE /DEPENDENCY_LIST}@*
21377 Forms the execution closure of one
21378 or more compiled units and checks completeness and currency.
21380 @item @command{ACS COMPILE}
21381 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
21382 Forms the execution closure of one or
21383 more specified units, checks completeness and currency,
21384 identifies units that have revised source files, compiles same,
21385 and recompiles units that are or will become obsolete.
21386 Also completes incomplete generic instantiations.
21388 @item @command{ACS COPY FOREIGN}
21390 Copies a foreign object file into the program library as a
21393 @item @command{ACS COPY UNIT}
21395 Copies a compiled unit from one program library to another.
21397 @item @command{ACS CREATE LIBRARY}
21398 @tab Create /directory (*)@*
21399 Creates a program library.
21401 @item @command{ACS CREATE SUBLIBRARY}
21402 @tab Create /directory (*)@*
21403 Creates a program sublibrary.
21405 @item @command{ACS DELETE LIBRARY}
21407 Deletes a program library and its contents.
21409 @item @command{ACS DELETE SUBLIBRARY}
21411 Deletes a program sublibrary and its contents.
21413 @item @command{ACS DELETE UNIT}
21414 @tab Delete file (*)@*
21415 On OpenVMS systems, deletes one or more compiled units from
21416 the current program library.
21418 @item @command{ACS DIRECTORY}
21419 @tab Directory (*)@*
21420 On OpenVMS systems, lists units contained in the current
21423 @item @command{ACS ENTER FOREIGN}
21425 Allows the import of a foreign body as an Ada library
21426 specification and enters a reference to a pointer.
21428 @item @command{ACS ENTER UNIT}
21430 Enters a reference (pointer) from the current program library to
21431 a unit compiled into another program library.
21433 @item @command{ACS EXIT}
21434 @tab [No equivalent]@*
21435 Exits from the program library manager.
21437 @item @command{ACS EXPORT}
21439 Creates an object file that contains system-specific object code
21440 for one or more units. With GNAT, object files can simply be copied
21441 into the desired directory.
21443 @item @command{ACS EXTRACT SOURCE}
21445 Allows access to the copied source file for each Ada compilation unit
21447 @item @command{ACS HELP}
21448 @tab @command{HELP GNAT}@*
21449 Provides online help.
21451 @item @command{ACS LINK}
21452 @tab @command{GNAT LINK}@*
21453 Links an object file containing Ada units into an executable file.
21455 @item @command{ACS LOAD}
21457 Loads (partially compiles) Ada units into the program library.
21458 Allows loading a program from a collection of files into a library
21459 without knowing the relationship among units.
21461 @item @command{ACS MERGE}
21463 Merges into the current program library, one or more units from
21464 another library where they were modified.
21466 @item @command{ACS RECOMPILE}
21467 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
21468 Recompiles from external or copied source files any obsolete
21469 unit in the closure. Also, completes any incomplete generic
21472 @item @command{ACS REENTER}
21473 @tab @command{GNAT MAKE}@*
21474 Reenters current references to units compiled after last entered
21475 with the @command{ACS ENTER UNIT} command.
21477 @item @command{ACS SET LIBRARY}
21478 @tab Set default (*)@*
21479 Defines a program library to be the compilation context as well
21480 as the target library for compiler output and commands in general.
21482 @item @command{ACS SET PRAGMA}
21483 @tab Edit @file{gnat.adc} (*)@*
21484 Redefines specified values of the library characteristics
21485 @code{LONG_ FLOAT}, @code{MEMORY_SIZE}, @code{SYSTEM_NAME},
21486 and @code{Float_Representation}.
21488 @item @command{ACS SET SOURCE}
21489 @tab Define @code{ADA_INCLUDE_PATH} path (*)@*
21490 Defines the source file search list for the @command{ACS COMPILE} command.
21492 @item @command{ACS SHOW LIBRARY}
21493 @tab Directory (*)@*
21494 Lists information about one or more program libraries.
21496 @item @command{ACS SHOW PROGRAM}
21497 @tab [No equivalent]@*
21498 Lists information about the execution closure of one or
21499 more units in the program library.
21501 @item @command{ACS SHOW SOURCE}
21502 @tab Show logical @code{ADA_INCLUDE_PATH}@*
21503 Shows the source file search used when compiling units.
21505 @item @command{ACS SHOW VERSION}
21506 @tab Compile with @option{VERBOSE} option
21507 Displays the version number of the compiler and program library
21510 @item @command{ACS SPAWN}
21511 @tab [No equivalent]@*
21512 Creates a subprocess of the current process (same as @command{DCL SPAWN}
21515 @item @command{ACS VERIFY}
21516 @tab [No equivalent]@*
21517 Performs a series of consistency checks on a program library to
21518 determine whether the library structure and library files are in
21525 @section Input-Output
21528 On OpenVMS Alpha systems, DEC Ada uses OpenVMS Record
21529 Management Services (RMS) to perform operations on
21533 DEC Ada and GNAT predefine an identical set of input-
21534 output packages. To make the use of the
21535 generic TEXT_IO operations more convenient, DEC Ada
21536 provides predefined library packages that instantiate the
21537 integer and floating-point operations for the predefined
21538 integer and floating-point types as shown in the following table.
21540 @multitable @columnfractions .45 .55
21541 @item @emph{Package Name} @tab Instantiation
21543 @item @code{INTEGER_TEXT_IO}
21544 @tab @code{INTEGER_IO(INTEGER)}
21546 @item @code{SHORT_INTEGER_TEXT_IO}
21547 @tab @code{INTEGER_IO(SHORT_INTEGER)}
21549 @item @code{SHORT_SHORT_INTEGER_TEXT_IO}
21550 @tab @code{INTEGER_IO(SHORT_SHORT_INTEGER)}
21552 @item @code{FLOAT_TEXT_IO}
21553 @tab @code{FLOAT_IO(FLOAT)}
21555 @item @code{LONG_FLOAT_TEXT_IO}
21556 @tab @code{FLOAT_IO(LONG_FLOAT)}
21560 The DEC Ada predefined packages and their operations
21561 are implemented using OpenVMS Alpha files and input-
21562 output facilities. DEC Ada supports asynchronous input-
21563 output on OpenVMS Alpha. Familiarity with the following is
21566 @item RMS file organizations and access methods
21568 @item OpenVMS file specifications and directories
21570 @item OpenVMS File Definition Language (FDL)
21574 GNAT provides I/O facilities that are completely
21575 compatible with DEC Ada. The distribution includes the
21576 standard DEC Ada versions of all I/O packages, operating
21577 in a manner compatible with DEC Ada. In particular, the
21578 following packages are by default the DEC Ada (Ada 83)
21579 versions of these packages rather than the renamings
21580 suggested in annex J of the Ada 95 Reference Manual:
21582 @item @code{TEXT_IO}
21584 @item @code{SEQUENTIAL_IO}
21586 @item @code{DIRECT_IO}
21590 The use of the standard Ada 95 syntax for child packages (for
21591 example, @code{ADA.TEXT_IO}) retrieves the Ada 95 versions of these
21592 packages, as defined in the Ada 95 Reference Manual.
21593 GNAT provides DIGITAL-compatible predefined instantiations
21594 of the @code{TEXT_IO} packages, and also
21595 provides the standard predefined instantiations required
21596 by the Ada 95 Reference Manual.
21598 For further information on how GNAT interfaces to the file
21599 system or how I/O is implemented in programs written in
21600 mixed languages, see the chapter ``Implementation of the
21601 Standard I/O'' in the @cite{GNAT Reference Manual}.
21602 This chapter covers the following:
21604 @item Standard I/O packages
21606 @item @code{FORM} strings
21608 @item @code{ADA.DIRECT_IO}
21610 @item @code{ADA.SEQUENTIAL_IO}
21612 @item @code{ADA.TEXT_IO}
21614 @item Stream pointer positioning
21616 @item Reading and writing non-regular files
21618 @item @code{GET_IMMEDIATE}
21620 @item Treating @code{TEXT_IO} files as streams
21627 @node Implementation Limits
21628 @section Implementation Limits
21631 The following table lists implementation limits for DEC Ada
21633 @multitable @columnfractions .60 .20 .20
21635 @item @emph{Compilation Parameter}
21636 @tab @emph{DEC Ada}
21640 @item In a subprogram or entry declaration, maximum number of
21641 formal parameters that are of an unconstrained record type
21646 @item Maximum identifier length (number of characters)
21651 @item Maximum number of characters in a source line
21656 @item Maximum collection size (number of bytes)
21661 @item Maximum number of discriminants for a record type
21666 @item Maximum number of formal parameters in an entry or
21667 subprogram declaration
21672 @item Maximum number of dimensions in an array type
21677 @item Maximum number of library units and subunits in a compilation.
21682 @item Maximum number of library units and subunits in an execution.
21687 @item Maximum number of objects declared with the pragma @code{COMMON_OBJECT}
21688 or @code{PSECT_OBJECT}
21693 @item Maximum number of enumeration literals in an enumeration type
21699 @item Maximum number of lines in a source file
21704 @item Maximum number of bits in any object
21709 @item Maximum size of the static portion of a stack frame (approximate)
21719 @c **************************************
21720 @node Platform-Specific Information for the Run-Time Libraries
21721 @appendix Platform-Specific Information for the Run-Time Libraries
21722 @cindex Tasking and threads libraries
21723 @cindex Threads libraries and tasking
21724 @cindex Run-time libraries (platform-specific information)
21727 The GNAT run-time implementation may vary with respect to both the
21728 underlying threads library and the exception handling scheme.
21729 For threads support, one or more of the following are supplied:
21731 @item @b{native threads library}, a binding to the thread package from
21732 the underlying operating system
21734 @item @b{pthreads library} (Sparc Solaris only), a binding to the Solaris
21735 POSIX thread package
21739 For exception handling, either or both of two models are supplied:
21741 @item @b{Zero-Cost Exceptions} (``ZCX''),@footnote{
21742 Most programs should experience a substantial speed improvement by
21743 being compiled with a ZCX run-time.
21744 This is especially true for
21745 tasking applications or applications with many exception handlers.}
21746 @cindex Zero-Cost Exceptions
21747 @cindex ZCX (Zero-Cost Exceptions)
21748 which uses binder-generated tables that
21749 are interrogated at run time to locate a handler
21751 @item @b{setjmp / longjmp} (``SJLJ''),
21752 @cindex setjmp/longjmp Exception Model
21753 @cindex SJLJ (setjmp/longjmp Exception Model)
21754 which uses dynamically-set data to establish
21755 the set of handlers
21759 This appendix summarizes which combinations of threads and exception support
21760 are supplied on various GNAT platforms.
21761 It then shows how to select a particular library either
21762 permanently or temporarily,
21763 explains the properties of (and tradeoffs among) the various threads
21764 libraries, and provides some additional
21765 information about several specific platforms.
21768 * Summary of Run-Time Configurations::
21769 * Specifying a Run-Time Library::
21770 * Choosing the Scheduling Policy::
21771 * Solaris-Specific Considerations::
21772 * IRIX-Specific Considerations::
21773 * Linux-Specific Considerations::
21774 * AIX-Specific Considerations::
21777 @node Summary of Run-Time Configurations
21778 @section Summary of Run-Time Configurations
21780 @multitable @columnfractions .30 .70
21781 @item @b{alpha-openvms}
21782 @item @code{@ @ }@i{rts-native (default)}
21783 @item @code{@ @ @ @ }Tasking @tab native VMS threads
21784 @item @code{@ @ @ @ }Exceptions @tab ZCX
21787 @item @code{@ @ }@i{rts-native (default)}
21788 @item @code{@ @ @ @ }Tasking @tab native HP threads library
21789 @item @code{@ @ @ @ }Exceptions @tab ZCX
21791 @item @code{@ @ }@i{rts-sjlj}
21792 @item @code{@ @ @ @ }Tasking @tab native HP threads library
21793 @item @code{@ @ @ @ }Exceptions @tab SJLJ
21795 @item @b{sparc-solaris} @tab
21796 @item @code{@ @ }@i{rts-native (default)}
21797 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
21798 @item @code{@ @ @ @ }Exceptions @tab ZCX
21800 @item @code{@ @ }@i{rts-m64}
21801 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
21802 @item @code{@ @ @ @ }Exceptions @tab ZCX
21803 @item @code{@ @ @ @ }Constraints @tab Use only when compiling in 64-bit mode;
21804 @item @tab Use only on Solaris 8 or later.
21805 @item @tab @xref{Building and Debugging 64-bit Applications}, for details.
21807 @item @code{@ @ }@i{rts-pthread}
21808 @item @code{@ @ @ @ }Tasking @tab pthreads library
21809 @item @code{@ @ @ @ }Exceptions @tab ZCX
21811 @item @code{@ @ }@i{rts-sjlj}
21812 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
21813 @item @code{@ @ @ @ }Exceptions @tab SJLJ
21815 @item @b{x86-linux}
21816 @item @code{@ @ }@i{rts-native (default)}
21817 @item @code{@ @ @ @ }Tasking @tab pthread library
21818 @item @code{@ @ @ @ }Exceptions @tab ZCX
21820 @item @code{@ @ }@i{rts-sjlj}
21821 @item @code{@ @ @ @ }Tasking @tab pthread library
21822 @item @code{@ @ @ @ }Exceptions @tab SJLJ
21824 @item @b{x86-windows}
21825 @item @code{@ @ }@i{rts-native (default)}
21826 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
21827 @item @code{@ @ @ @ }Exceptions @tab SJLJ
21831 @node Specifying a Run-Time Library
21832 @section Specifying a Run-Time Library
21835 The @file{adainclude} subdirectory containing the sources of the GNAT
21836 run-time library, and the @file{adalib} subdirectory containing the
21837 @file{ALI} files and the static and/or shared GNAT library, are located
21838 in the gcc target-dependent area:
21841 target=$prefix/lib/gcc-lib/gcc-@i{dumpmachine}/gcc-@i{dumpversion}/
21845 As indicated above, on some platforms several run-time libraries are supplied.
21846 These libraries are installed in the target dependent area and
21847 contain a complete source and binary subdirectory. The detailed description
21848 below explains the differences between the different libraries in terms of
21849 their thread support.
21851 The default run-time library (when GNAT is installed) is @emph{rts-native}.
21852 This default run time is selected by the means of soft links.
21853 For example on x86-linux:
21859 +--- adainclude----------+
21861 +--- adalib-----------+ |
21863 +--- rts-native | |
21865 | +--- adainclude <---+
21867 | +--- adalib <----+
21878 If the @i{rts-sjlj} library is to be selected on a permanent basis,
21879 these soft links can be modified with the following commands:
21883 $ rm -f adainclude adalib
21884 $ ln -s rts-sjlj/adainclude adainclude
21885 $ ln -s rts-sjlj/adalib adalib
21889 Alternatively, you can specify @file{rts-sjlj/adainclude} in the file
21890 @file{$target/ada_source_path} and @file{rts-sjlj/adalib} in
21891 @file{$target/ada_object_path}.
21893 Selecting another run-time library temporarily can be
21894 achieved by the regular mechanism for GNAT object or source path selection:
21898 Set the environment variables:
21901 $ ADA_INCLUDE_PATH=$target/rts-sjlj/adainclude:$ADA_INCLUDE_PATH
21902 $ ADA_OBJECTS_PATH=$target/rts-sjlj/adalib:$ADA_OBJECTS_PATH
21903 $ export ADA_INCLUDE_PATH ADA_OBJECTS_PATH
21907 Use @option{-aI$target/rts-sjlj/adainclude}
21908 and @option{-aO$target/rts-sjlj/adalib}
21909 on the @command{gnatmake} command line
21912 Use the switch @option{--RTS}; e.g., @option{--RTS=sjlj}
21913 @cindex @option{--RTS} option
21916 @node Choosing the Scheduling Policy
21917 @section Choosing the Scheduling Policy
21920 When using a POSIX threads implementation, you have a choice of several
21921 scheduling policies: @code{SCHED_FIFO},
21922 @cindex @code{SCHED_FIFO} scheduling policy
21924 @cindex @code{SCHED_RR} scheduling policy
21925 and @code{SCHED_OTHER}.
21926 @cindex @code{SCHED_OTHER} scheduling policy
21927 Typically, the default is @code{SCHED_OTHER}, while using @code{SCHED_FIFO}
21928 or @code{SCHED_RR} requires special (e.g., root) privileges.
21930 By default, GNAT uses the @code{SCHED_OTHER} policy. To specify
21932 @cindex @code{SCHED_FIFO} scheduling policy
21933 you can use one of the following:
21937 @code{pragma Time_Slice (0.0)}
21938 @cindex pragma Time_Slice
21940 the corresponding binder option @option{-T0}
21941 @cindex @option{-T0} option
21943 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
21944 @cindex pragma Task_Dispatching_Policy
21948 To specify @code{SCHED_RR},
21949 @cindex @code{SCHED_RR} scheduling policy
21950 you should use @code{pragma Time_Slice} with a
21951 value greater than @code{0.0}, or else use the corresponding @option{-T}
21954 @node Solaris-Specific Considerations
21955 @section Solaris-Specific Considerations
21956 @cindex Solaris Sparc threads libraries
21959 This section addresses some topics related to the various threads libraries
21960 on Sparc Solaris and then provides some information on building and
21961 debugging 64-bit applications.
21964 * Solaris Threads Issues::
21965 * Building and Debugging 64-bit Applications::
21968 @node Solaris Threads Issues
21969 @subsection Solaris Threads Issues
21972 GNAT under Solaris comes with an alternate tasking run-time library
21973 based on POSIX threads --- @emph{rts-pthread}.
21974 @cindex rts-pthread threads library
21975 This run-time library has the advantage of being mostly shared across all
21976 POSIX-compliant thread implementations, and it also provides under
21977 @w{Solaris 8} the @code{PTHREAD_PRIO_INHERIT}
21978 @cindex @code{PTHREAD_PRIO_INHERIT} policy (under rts-pthread)
21979 and @code{PTHREAD_PRIO_PROTECT}
21980 @cindex @code{PTHREAD_PRIO_PROTECT} policy (under rts-pthread)
21981 semantics that can be selected using the predefined pragma
21982 @code{Locking_Policy}
21983 @cindex pragma Locking_Policy (under rts-pthread)
21985 @code{Inheritance_Locking} and @code{Ceiling_Locking} as the policy.
21986 @cindex @code{Inheritance_Locking} (under rts-pthread)
21987 @cindex @code{Ceiling_Locking} (under rts-pthread)
21989 As explained above, the native run-time library is based on the Solaris thread
21990 library (@code{libthread}) and is the default library.
21992 When the Solaris threads library is used (this is the default), programs
21993 compiled with GNAT can automatically take advantage of
21994 and can thus execute on multiple processors.
21995 The user can alternatively specify a processor on which the program should run
21996 to emulate a single-processor system. The multiprocessor / uniprocessor choice
21998 setting the environment variable @code{GNAT_PROCESSOR}
21999 @cindex @code{GNAT_PROCESSOR} environment variable (on Sparc Solaris)
22000 to one of the following:
22004 Use the default configuration (run the program on all
22005 available processors) - this is the same as having
22006 @code{GNAT_PROCESSOR} unset
22009 Let the run-time implementation choose one processor and run the program on
22012 @item 0 .. Last_Proc
22013 Run the program on the specified processor.
22014 @code{Last_Proc} is equal to @code{_SC_NPROCESSORS_CONF - 1}
22015 (where @code{_SC_NPROCESSORS_CONF} is a system variable).
22018 @node Building and Debugging 64-bit Applications
22019 @subsection Building and Debugging 64-bit Applications
22022 In a 64-bit application, all the sources involved must be compiled with the
22023 @option{-m64} command-line option, and a specific GNAT library (compiled with
22024 this option) is required.
22025 The easiest way to build a 64bit application is to add
22026 @option{-m64 --RTS=m64} to the @command{gnatmake} flags.
22028 To debug these applications, a special version of gdb called @command{gdb64}
22031 To summarize, building and debugging a ``Hello World'' program in 64-bit mode
22035 $ gnatmake -m64 -g --RTS=m64 hello.adb
22039 In addition, the following capabilities are not supported when using the
22040 @option{-m64} option:
22043 @item -fstack-check does not work together with -m64.
22044 Any application combining these options crashes at startup time.
22046 @item Call-chain backtrace computation does not work with -m64.
22047 Thus the gnatbind switch -E is not supported.
22050 @node IRIX-Specific Considerations
22051 @section IRIX-Specific Considerations
22052 @cindex IRIX thread library
22055 On SGI IRIX, the thread library depends on which compiler is used.
22056 The @emph{o32 ABI} compiler comes with a run-time library based on the
22057 user-level @code{athread}
22058 library. Thus kernel-level capabilities such as nonblocking system
22059 calls or time slicing can only be achieved reliably by specifying different
22060 @code{sprocs} via the pragma @code{Task_Info}
22061 @cindex pragma Task_Info (and IRIX threads)
22063 @code{System.Task_Info} package.
22064 @cindex @code{System.Task_Info} package (and IRIX threads)
22065 See the @cite{GNAT Reference Manual} for further information.
22067 The @emph{n32 ABI} compiler comes with a run-time library based on the
22068 kernel POSIX threads and thus does not have the limitations mentioned above.
22070 @node Linux-Specific Considerations
22071 @section Linux-Specific Considerations
22072 @cindex Linux threads libraries
22075 The default thread library under GNU/Linux has the following disadvantages
22076 compared to other native thread libraries:
22079 @item The size of the task's stack is limited to 2 megabytes.
22080 @item The signal model is not POSIX compliant, which means that to send a
22081 signal to the process, you need to send the signal to all threads,
22082 e.g. by using @code{killpg()}.
22085 @node AIX-Specific Considerations
22086 @section AIX-Specific Considerations
22087 @cindex AIX resolver library
22090 On AIX, the resolver library initializes some internal structure on
22091 the first call to @code{get*by*} functions, which are used to implement
22092 @code{GNAT.Sockets.Get_Host_By_Name} and
22093 @code{GNAT.Sockets.Get_Host_By_Addrss}.
22094 If such initialization occurs within an Ada task, and the stack size for
22095 the task is the default size, a stack overflow may occur.
22097 To avoid this overflow, the user should either ensure that the first call
22098 to @code{GNAT.Sockets.Get_Host_By_Name} or
22099 @code{GNAT.Sockets.Get_Host_By_Addrss}
22100 occurs in the environment task, or use @code{pragma Storage_Size} to
22101 specify a sufficiently large size for the stack of the task that contains
22104 @c *******************************
22105 @node Example of Binder Output File
22106 @appendix Example of Binder Output File
22109 This Appendix displays the source code for @command{gnatbind}'s output
22110 file generated for a simple ``Hello World'' program.
22111 Comments have been added for clarification purposes.
22113 @smallexample @c adanocomment
22117 -- The package is called Ada_Main unless this name is actually used
22118 -- as a unit name in the partition, in which case some other unique
22122 package ada_main is
22124 Elab_Final_Code : Integer;
22125 pragma Import (C, Elab_Final_Code, "__gnat_inside_elab_final_code");
22127 -- The main program saves the parameters (argument count,
22128 -- argument values, environment pointer) in global variables
22129 -- for later access by other units including
22130 -- Ada.Command_Line.
22132 gnat_argc : Integer;
22133 gnat_argv : System.Address;
22134 gnat_envp : System.Address;
22136 -- The actual variables are stored in a library routine. This
22137 -- is useful for some shared library situations, where there
22138 -- are problems if variables are not in the library.
22140 pragma Import (C, gnat_argc);
22141 pragma Import (C, gnat_argv);
22142 pragma Import (C, gnat_envp);
22144 -- The exit status is similarly an external location
22146 gnat_exit_status : Integer;
22147 pragma Import (C, gnat_exit_status);
22149 GNAT_Version : constant String :=
22150 "GNAT Version: 3.15w (20010315)";
22151 pragma Export (C, GNAT_Version, "__gnat_version");
22153 -- This is the generated adafinal routine that performs
22154 -- finalization at the end of execution. In the case where
22155 -- Ada is the main program, this main program makes a call
22156 -- to adafinal at program termination.
22158 procedure adafinal;
22159 pragma Export (C, adafinal, "adafinal");
22161 -- This is the generated adainit routine that performs
22162 -- initialization at the start of execution. In the case
22163 -- where Ada is the main program, this main program makes
22164 -- a call to adainit at program startup.
22167 pragma Export (C, adainit, "adainit");
22169 -- This routine is called at the start of execution. It is
22170 -- a dummy routine that is used by the debugger to breakpoint
22171 -- at the start of execution.
22173 procedure Break_Start;
22174 pragma Import (C, Break_Start, "__gnat_break_start");
22176 -- This is the actual generated main program (it would be
22177 -- suppressed if the no main program switch were used). As
22178 -- required by standard system conventions, this program has
22179 -- the external name main.
22183 argv : System.Address;
22184 envp : System.Address)
22186 pragma Export (C, main, "main");
22188 -- The following set of constants give the version
22189 -- identification values for every unit in the bound
22190 -- partition. This identification is computed from all
22191 -- dependent semantic units, and corresponds to the
22192 -- string that would be returned by use of the
22193 -- Body_Version or Version attributes.
22195 type Version_32 is mod 2 ** 32;
22196 u00001 : constant Version_32 := 16#7880BEB3#;
22197 u00002 : constant Version_32 := 16#0D24CBD0#;
22198 u00003 : constant Version_32 := 16#3283DBEB#;
22199 u00004 : constant Version_32 := 16#2359F9ED#;
22200 u00005 : constant Version_32 := 16#664FB847#;
22201 u00006 : constant Version_32 := 16#68E803DF#;
22202 u00007 : constant Version_32 := 16#5572E604#;
22203 u00008 : constant Version_32 := 16#46B173D8#;
22204 u00009 : constant Version_32 := 16#156A40CF#;
22205 u00010 : constant Version_32 := 16#033DABE0#;
22206 u00011 : constant Version_32 := 16#6AB38FEA#;
22207 u00012 : constant Version_32 := 16#22B6217D#;
22208 u00013 : constant Version_32 := 16#68A22947#;
22209 u00014 : constant Version_32 := 16#18CC4A56#;
22210 u00015 : constant Version_32 := 16#08258E1B#;
22211 u00016 : constant Version_32 := 16#367D5222#;
22212 u00017 : constant Version_32 := 16#20C9ECA4#;
22213 u00018 : constant Version_32 := 16#50D32CB6#;
22214 u00019 : constant Version_32 := 16#39A8BB77#;
22215 u00020 : constant Version_32 := 16#5CF8FA2B#;
22216 u00021 : constant Version_32 := 16#2F1EB794#;
22217 u00022 : constant Version_32 := 16#31AB6444#;
22218 u00023 : constant Version_32 := 16#1574B6E9#;
22219 u00024 : constant Version_32 := 16#5109C189#;
22220 u00025 : constant Version_32 := 16#56D770CD#;
22221 u00026 : constant Version_32 := 16#02F9DE3D#;
22222 u00027 : constant Version_32 := 16#08AB6B2C#;
22223 u00028 : constant Version_32 := 16#3FA37670#;
22224 u00029 : constant Version_32 := 16#476457A0#;
22225 u00030 : constant Version_32 := 16#731E1B6E#;
22226 u00031 : constant Version_32 := 16#23C2E789#;
22227 u00032 : constant Version_32 := 16#0F1BD6A1#;
22228 u00033 : constant Version_32 := 16#7C25DE96#;
22229 u00034 : constant Version_32 := 16#39ADFFA2#;
22230 u00035 : constant Version_32 := 16#571DE3E7#;
22231 u00036 : constant Version_32 := 16#5EB646AB#;
22232 u00037 : constant Version_32 := 16#4249379B#;
22233 u00038 : constant Version_32 := 16#0357E00A#;
22234 u00039 : constant Version_32 := 16#3784FB72#;
22235 u00040 : constant Version_32 := 16#2E723019#;
22236 u00041 : constant Version_32 := 16#623358EA#;
22237 u00042 : constant Version_32 := 16#107F9465#;
22238 u00043 : constant Version_32 := 16#6843F68A#;
22239 u00044 : constant Version_32 := 16#63305874#;
22240 u00045 : constant Version_32 := 16#31E56CE1#;
22241 u00046 : constant Version_32 := 16#02917970#;
22242 u00047 : constant Version_32 := 16#6CCBA70E#;
22243 u00048 : constant Version_32 := 16#41CD4204#;
22244 u00049 : constant Version_32 := 16#572E3F58#;
22245 u00050 : constant Version_32 := 16#20729FF5#;
22246 u00051 : constant Version_32 := 16#1D4F93E8#;
22247 u00052 : constant Version_32 := 16#30B2EC3D#;
22248 u00053 : constant Version_32 := 16#34054F96#;
22249 u00054 : constant Version_32 := 16#5A199860#;
22250 u00055 : constant Version_32 := 16#0E7F912B#;
22251 u00056 : constant Version_32 := 16#5760634A#;
22252 u00057 : constant Version_32 := 16#5D851835#;
22254 -- The following Export pragmas export the version numbers
22255 -- with symbolic names ending in B (for body) or S
22256 -- (for spec) so that they can be located in a link. The
22257 -- information provided here is sufficient to track down
22258 -- the exact versions of units used in a given build.
22260 pragma Export (C, u00001, "helloB");
22261 pragma Export (C, u00002, "system__standard_libraryB");
22262 pragma Export (C, u00003, "system__standard_libraryS");
22263 pragma Export (C, u00004, "adaS");
22264 pragma Export (C, u00005, "ada__text_ioB");
22265 pragma Export (C, u00006, "ada__text_ioS");
22266 pragma Export (C, u00007, "ada__exceptionsB");
22267 pragma Export (C, u00008, "ada__exceptionsS");
22268 pragma Export (C, u00009, "gnatS");
22269 pragma Export (C, u00010, "gnat__heap_sort_aB");
22270 pragma Export (C, u00011, "gnat__heap_sort_aS");
22271 pragma Export (C, u00012, "systemS");
22272 pragma Export (C, u00013, "system__exception_tableB");
22273 pragma Export (C, u00014, "system__exception_tableS");
22274 pragma Export (C, u00015, "gnat__htableB");
22275 pragma Export (C, u00016, "gnat__htableS");
22276 pragma Export (C, u00017, "system__exceptionsS");
22277 pragma Export (C, u00018, "system__machine_state_operationsB");
22278 pragma Export (C, u00019, "system__machine_state_operationsS");
22279 pragma Export (C, u00020, "system__machine_codeS");
22280 pragma Export (C, u00021, "system__storage_elementsB");
22281 pragma Export (C, u00022, "system__storage_elementsS");
22282 pragma Export (C, u00023, "system__secondary_stackB");
22283 pragma Export (C, u00024, "system__secondary_stackS");
22284 pragma Export (C, u00025, "system__parametersB");
22285 pragma Export (C, u00026, "system__parametersS");
22286 pragma Export (C, u00027, "system__soft_linksB");
22287 pragma Export (C, u00028, "system__soft_linksS");
22288 pragma Export (C, u00029, "system__stack_checkingB");
22289 pragma Export (C, u00030, "system__stack_checkingS");
22290 pragma Export (C, u00031, "system__tracebackB");
22291 pragma Export (C, u00032, "system__tracebackS");
22292 pragma Export (C, u00033, "ada__streamsS");
22293 pragma Export (C, u00034, "ada__tagsB");
22294 pragma Export (C, u00035, "ada__tagsS");
22295 pragma Export (C, u00036, "system__string_opsB");
22296 pragma Export (C, u00037, "system__string_opsS");
22297 pragma Export (C, u00038, "interfacesS");
22298 pragma Export (C, u00039, "interfaces__c_streamsB");
22299 pragma Export (C, u00040, "interfaces__c_streamsS");
22300 pragma Export (C, u00041, "system__file_ioB");
22301 pragma Export (C, u00042, "system__file_ioS");
22302 pragma Export (C, u00043, "ada__finalizationB");
22303 pragma Export (C, u00044, "ada__finalizationS");
22304 pragma Export (C, u00045, "system__finalization_rootB");
22305 pragma Export (C, u00046, "system__finalization_rootS");
22306 pragma Export (C, u00047, "system__finalization_implementationB");
22307 pragma Export (C, u00048, "system__finalization_implementationS");
22308 pragma Export (C, u00049, "system__string_ops_concat_3B");
22309 pragma Export (C, u00050, "system__string_ops_concat_3S");
22310 pragma Export (C, u00051, "system__stream_attributesB");
22311 pragma Export (C, u00052, "system__stream_attributesS");
22312 pragma Export (C, u00053, "ada__io_exceptionsS");
22313 pragma Export (C, u00054, "system__unsigned_typesS");
22314 pragma Export (C, u00055, "system__file_control_blockS");
22315 pragma Export (C, u00056, "ada__finalization__list_controllerB");
22316 pragma Export (C, u00057, "ada__finalization__list_controllerS");
22318 -- BEGIN ELABORATION ORDER
22321 -- gnat.heap_sort_a (spec)
22322 -- gnat.heap_sort_a (body)
22323 -- gnat.htable (spec)
22324 -- gnat.htable (body)
22325 -- interfaces (spec)
22327 -- system.machine_code (spec)
22328 -- system.parameters (spec)
22329 -- system.parameters (body)
22330 -- interfaces.c_streams (spec)
22331 -- interfaces.c_streams (body)
22332 -- system.standard_library (spec)
22333 -- ada.exceptions (spec)
22334 -- system.exception_table (spec)
22335 -- system.exception_table (body)
22336 -- ada.io_exceptions (spec)
22337 -- system.exceptions (spec)
22338 -- system.storage_elements (spec)
22339 -- system.storage_elements (body)
22340 -- system.machine_state_operations (spec)
22341 -- system.machine_state_operations (body)
22342 -- system.secondary_stack (spec)
22343 -- system.stack_checking (spec)
22344 -- system.soft_links (spec)
22345 -- system.soft_links (body)
22346 -- system.stack_checking (body)
22347 -- system.secondary_stack (body)
22348 -- system.standard_library (body)
22349 -- system.string_ops (spec)
22350 -- system.string_ops (body)
22353 -- ada.streams (spec)
22354 -- system.finalization_root (spec)
22355 -- system.finalization_root (body)
22356 -- system.string_ops_concat_3 (spec)
22357 -- system.string_ops_concat_3 (body)
22358 -- system.traceback (spec)
22359 -- system.traceback (body)
22360 -- ada.exceptions (body)
22361 -- system.unsigned_types (spec)
22362 -- system.stream_attributes (spec)
22363 -- system.stream_attributes (body)
22364 -- system.finalization_implementation (spec)
22365 -- system.finalization_implementation (body)
22366 -- ada.finalization (spec)
22367 -- ada.finalization (body)
22368 -- ada.finalization.list_controller (spec)
22369 -- ada.finalization.list_controller (body)
22370 -- system.file_control_block (spec)
22371 -- system.file_io (spec)
22372 -- system.file_io (body)
22373 -- ada.text_io (spec)
22374 -- ada.text_io (body)
22376 -- END ELABORATION ORDER
22380 -- The following source file name pragmas allow the generated file
22381 -- names to be unique for different main programs. They are needed
22382 -- since the package name will always be Ada_Main.
22384 pragma Source_File_Name (ada_main, Spec_File_Name => "b~hello.ads");
22385 pragma Source_File_Name (ada_main, Body_File_Name => "b~hello.adb");
22387 -- Generated package body for Ada_Main starts here
22389 package body ada_main is
22391 -- The actual finalization is performed by calling the
22392 -- library routine in System.Standard_Library.Adafinal
22394 procedure Do_Finalize;
22395 pragma Import (C, Do_Finalize, "system__standard_library__adafinal");
22402 procedure adainit is
22404 -- These booleans are set to True once the associated unit has
22405 -- been elaborated. It is also used to avoid elaborating the
22406 -- same unit twice.
22409 pragma Import (Ada, E040, "interfaces__c_streams_E");
22412 pragma Import (Ada, E008, "ada__exceptions_E");
22415 pragma Import (Ada, E014, "system__exception_table_E");
22418 pragma Import (Ada, E053, "ada__io_exceptions_E");
22421 pragma Import (Ada, E017, "system__exceptions_E");
22424 pragma Import (Ada, E024, "system__secondary_stack_E");
22427 pragma Import (Ada, E030, "system__stack_checking_E");
22430 pragma Import (Ada, E028, "system__soft_links_E");
22433 pragma Import (Ada, E035, "ada__tags_E");
22436 pragma Import (Ada, E033, "ada__streams_E");
22439 pragma Import (Ada, E046, "system__finalization_root_E");
22442 pragma Import (Ada, E048, "system__finalization_implementation_E");
22445 pragma Import (Ada, E044, "ada__finalization_E");
22448 pragma Import (Ada, E057, "ada__finalization__list_controller_E");
22451 pragma Import (Ada, E055, "system__file_control_block_E");
22454 pragma Import (Ada, E042, "system__file_io_E");
22457 pragma Import (Ada, E006, "ada__text_io_E");
22459 -- Set_Globals is a library routine that stores away the
22460 -- value of the indicated set of global values in global
22461 -- variables within the library.
22463 procedure Set_Globals
22464 (Main_Priority : Integer;
22465 Time_Slice_Value : Integer;
22466 WC_Encoding : Character;
22467 Locking_Policy : Character;
22468 Queuing_Policy : Character;
22469 Task_Dispatching_Policy : Character;
22470 Adafinal : System.Address;
22471 Unreserve_All_Interrupts : Integer;
22472 Exception_Tracebacks : Integer);
22473 @findex __gnat_set_globals
22474 pragma Import (C, Set_Globals, "__gnat_set_globals");
22476 -- SDP_Table_Build is a library routine used to build the
22477 -- exception tables. See unit Ada.Exceptions in files
22478 -- a-except.ads/adb for full details of how zero cost
22479 -- exception handling works. This procedure, the call to
22480 -- it, and the two following tables are all omitted if the
22481 -- build is in longjmp/setjump exception mode.
22483 @findex SDP_Table_Build
22484 @findex Zero Cost Exceptions
22485 procedure SDP_Table_Build
22486 (SDP_Addresses : System.Address;
22487 SDP_Count : Natural;
22488 Elab_Addresses : System.Address;
22489 Elab_Addr_Count : Natural);
22490 pragma Import (C, SDP_Table_Build, "__gnat_SDP_Table_Build");
22492 -- Table of Unit_Exception_Table addresses. Used for zero
22493 -- cost exception handling to build the top level table.
22495 ST : aliased constant array (1 .. 23) of System.Address := (
22497 Ada.Text_Io'UET_Address,
22498 Ada.Exceptions'UET_Address,
22499 Gnat.Heap_Sort_A'UET_Address,
22500 System.Exception_Table'UET_Address,
22501 System.Machine_State_Operations'UET_Address,
22502 System.Secondary_Stack'UET_Address,
22503 System.Parameters'UET_Address,
22504 System.Soft_Links'UET_Address,
22505 System.Stack_Checking'UET_Address,
22506 System.Traceback'UET_Address,
22507 Ada.Streams'UET_Address,
22508 Ada.Tags'UET_Address,
22509 System.String_Ops'UET_Address,
22510 Interfaces.C_Streams'UET_Address,
22511 System.File_Io'UET_Address,
22512 Ada.Finalization'UET_Address,
22513 System.Finalization_Root'UET_Address,
22514 System.Finalization_Implementation'UET_Address,
22515 System.String_Ops_Concat_3'UET_Address,
22516 System.Stream_Attributes'UET_Address,
22517 System.File_Control_Block'UET_Address,
22518 Ada.Finalization.List_Controller'UET_Address);
22520 -- Table of addresses of elaboration routines. Used for
22521 -- zero cost exception handling to make sure these
22522 -- addresses are included in the top level procedure
22525 EA : aliased constant array (1 .. 23) of System.Address := (
22526 adainit'Code_Address,
22527 Do_Finalize'Code_Address,
22528 Ada.Exceptions'Elab_Spec'Address,
22529 System.Exceptions'Elab_Spec'Address,
22530 Interfaces.C_Streams'Elab_Spec'Address,
22531 System.Exception_Table'Elab_Body'Address,
22532 Ada.Io_Exceptions'Elab_Spec'Address,
22533 System.Stack_Checking'Elab_Spec'Address,
22534 System.Soft_Links'Elab_Body'Address,
22535 System.Secondary_Stack'Elab_Body'Address,
22536 Ada.Tags'Elab_Spec'Address,
22537 Ada.Tags'Elab_Body'Address,
22538 Ada.Streams'Elab_Spec'Address,
22539 System.Finalization_Root'Elab_Spec'Address,
22540 Ada.Exceptions'Elab_Body'Address,
22541 System.Finalization_Implementation'Elab_Spec'Address,
22542 System.Finalization_Implementation'Elab_Body'Address,
22543 Ada.Finalization'Elab_Spec'Address,
22544 Ada.Finalization.List_Controller'Elab_Spec'Address,
22545 System.File_Control_Block'Elab_Spec'Address,
22546 System.File_Io'Elab_Body'Address,
22547 Ada.Text_Io'Elab_Spec'Address,
22548 Ada.Text_Io'Elab_Body'Address);
22550 -- Start of processing for adainit
22554 -- Call SDP_Table_Build to build the top level procedure
22555 -- table for zero cost exception handling (omitted in
22556 -- longjmp/setjump mode).
22558 SDP_Table_Build (ST'Address, 23, EA'Address, 23);
22560 -- Call Set_Globals to record various information for
22561 -- this partition. The values are derived by the binder
22562 -- from information stored in the ali files by the compiler.
22564 @findex __gnat_set_globals
22566 (Main_Priority => -1,
22567 -- Priority of main program, -1 if no pragma Priority used
22569 Time_Slice_Value => -1,
22570 -- Time slice from Time_Slice pragma, -1 if none used
22572 WC_Encoding => 'b',
22573 -- Wide_Character encoding used, default is brackets
22575 Locking_Policy => ' ',
22576 -- Locking_Policy used, default of space means not
22577 -- specified, otherwise it is the first character of
22578 -- the policy name.
22580 Queuing_Policy => ' ',
22581 -- Queuing_Policy used, default of space means not
22582 -- specified, otherwise it is the first character of
22583 -- the policy name.
22585 Task_Dispatching_Policy => ' ',
22586 -- Task_Dispatching_Policy used, default of space means
22587 -- not specified, otherwise first character of the
22590 Adafinal => System.Null_Address,
22591 -- Address of Adafinal routine, not used anymore
22593 Unreserve_All_Interrupts => 0,
22594 -- Set true if pragma Unreserve_All_Interrupts was used
22596 Exception_Tracebacks => 0);
22597 -- Indicates if exception tracebacks are enabled
22599 Elab_Final_Code := 1;
22601 -- Now we have the elaboration calls for all units in the partition.
22602 -- The Elab_Spec and Elab_Body attributes generate references to the
22603 -- implicit elaboration procedures generated by the compiler for
22604 -- each unit that requires elaboration.
22607 Interfaces.C_Streams'Elab_Spec;
22611 Ada.Exceptions'Elab_Spec;
22614 System.Exception_Table'Elab_Body;
22618 Ada.Io_Exceptions'Elab_Spec;
22622 System.Exceptions'Elab_Spec;
22626 System.Stack_Checking'Elab_Spec;
22629 System.Soft_Links'Elab_Body;
22634 System.Secondary_Stack'Elab_Body;
22638 Ada.Tags'Elab_Spec;
22641 Ada.Tags'Elab_Body;
22645 Ada.Streams'Elab_Spec;
22649 System.Finalization_Root'Elab_Spec;
22653 Ada.Exceptions'Elab_Body;
22657 System.Finalization_Implementation'Elab_Spec;
22660 System.Finalization_Implementation'Elab_Body;
22664 Ada.Finalization'Elab_Spec;
22668 Ada.Finalization.List_Controller'Elab_Spec;
22672 System.File_Control_Block'Elab_Spec;
22676 System.File_Io'Elab_Body;
22680 Ada.Text_Io'Elab_Spec;
22683 Ada.Text_Io'Elab_Body;
22687 Elab_Final_Code := 0;
22695 procedure adafinal is
22704 -- main is actually a function, as in the ANSI C standard,
22705 -- defined to return the exit status. The three parameters
22706 -- are the argument count, argument values and environment
22709 @findex Main Program
22712 argv : System.Address;
22713 envp : System.Address)
22716 -- The initialize routine performs low level system
22717 -- initialization using a standard library routine which
22718 -- sets up signal handling and performs any other
22719 -- required setup. The routine can be found in file
22722 @findex __gnat_initialize
22723 procedure initialize;
22724 pragma Import (C, initialize, "__gnat_initialize");
22726 -- The finalize routine performs low level system
22727 -- finalization using a standard library routine. The
22728 -- routine is found in file a-final.c and in the standard
22729 -- distribution is a dummy routine that does nothing, so
22730 -- really this is a hook for special user finalization.
22732 @findex __gnat_finalize
22733 procedure finalize;
22734 pragma Import (C, finalize, "__gnat_finalize");
22736 -- We get to the main program of the partition by using
22737 -- pragma Import because if we try to with the unit and
22738 -- call it Ada style, then not only do we waste time
22739 -- recompiling it, but also, we don't really know the right
22740 -- switches (e.g. identifier character set) to be used
22743 procedure Ada_Main_Program;
22744 pragma Import (Ada, Ada_Main_Program, "_ada_hello");
22746 -- Start of processing for main
22749 -- Save global variables
22755 -- Call low level system initialization
22759 -- Call our generated Ada initialization routine
22763 -- This is the point at which we want the debugger to get
22768 -- Now we call the main program of the partition
22772 -- Perform Ada finalization
22776 -- Perform low level system finalization
22780 -- Return the proper exit status
22781 return (gnat_exit_status);
22784 -- This section is entirely comments, so it has no effect on the
22785 -- compilation of the Ada_Main package. It provides the list of
22786 -- object files and linker options, as well as some standard
22787 -- libraries needed for the link. The gnatlink utility parses
22788 -- this b~hello.adb file to read these comment lines to generate
22789 -- the appropriate command line arguments for the call to the
22790 -- system linker. The BEGIN/END lines are used for sentinels for
22791 -- this parsing operation.
22793 -- The exact file names will of course depend on the environment,
22794 -- host/target and location of files on the host system.
22796 @findex Object file list
22797 -- BEGIN Object file/option list
22800 -- -L/usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/
22801 -- /usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/libgnat.a
22802 -- END Object file/option list
22808 The Ada code in the above example is exactly what is generated by the
22809 binder. We have added comments to more clearly indicate the function
22810 of each part of the generated @code{Ada_Main} package.
22812 The code is standard Ada in all respects, and can be processed by any
22813 tools that handle Ada. In particular, it is possible to use the debugger
22814 in Ada mode to debug the generated @code{Ada_Main} package. For example,
22815 suppose that for reasons that you do not understand, your program is crashing
22816 during elaboration of the body of @code{Ada.Text_IO}. To locate this bug,
22817 you can place a breakpoint on the call:
22819 @smallexample @c ada
22820 Ada.Text_Io'Elab_Body;
22824 and trace the elaboration routine for this package to find out where
22825 the problem might be (more usually of course you would be debugging
22826 elaboration code in your own application).
22828 @node Elaboration Order Handling in GNAT
22829 @appendix Elaboration Order Handling in GNAT
22830 @cindex Order of elaboration
22831 @cindex Elaboration control
22834 * Elaboration Code in Ada 95::
22835 * Checking the Elaboration Order in Ada 95::
22836 * Controlling the Elaboration Order in Ada 95::
22837 * Controlling Elaboration in GNAT - Internal Calls::
22838 * Controlling Elaboration in GNAT - External Calls::
22839 * Default Behavior in GNAT - Ensuring Safety::
22840 * Treatment of Pragma Elaborate::
22841 * Elaboration Issues for Library Tasks::
22842 * Mixing Elaboration Models::
22843 * What to Do If the Default Elaboration Behavior Fails::
22844 * Elaboration for Access-to-Subprogram Values::
22845 * Summary of Procedures for Elaboration Control::
22846 * Other Elaboration Order Considerations::
22850 This chapter describes the handling of elaboration code in Ada 95 and
22851 in GNAT, and discusses how the order of elaboration of program units can
22852 be controlled in GNAT, either automatically or with explicit programming
22855 @node Elaboration Code in Ada 95
22856 @section Elaboration Code in Ada 95
22859 Ada 95 provides rather general mechanisms for executing code at elaboration
22860 time, that is to say before the main program starts executing. Such code arises
22864 @item Initializers for variables.
22865 Variables declared at the library level, in package specs or bodies, can
22866 require initialization that is performed at elaboration time, as in:
22867 @smallexample @c ada
22869 Sqrt_Half : Float := Sqrt (0.5);
22873 @item Package initialization code
22874 Code in a @code{BEGIN-END} section at the outer level of a package body is
22875 executed as part of the package body elaboration code.
22877 @item Library level task allocators
22878 Tasks that are declared using task allocators at the library level
22879 start executing immediately and hence can execute at elaboration time.
22883 Subprogram calls are possible in any of these contexts, which means that
22884 any arbitrary part of the program may be executed as part of the elaboration
22885 code. It is even possible to write a program which does all its work at
22886 elaboration time, with a null main program, although stylistically this
22887 would usually be considered an inappropriate way to structure
22890 An important concern arises in the context of elaboration code:
22891 we have to be sure that it is executed in an appropriate order. What we
22892 have is a series of elaboration code sections, potentially one section
22893 for each unit in the program. It is important that these execute
22894 in the correct order. Correctness here means that, taking the above
22895 example of the declaration of @code{Sqrt_Half},
22896 if some other piece of
22897 elaboration code references @code{Sqrt_Half},
22898 then it must run after the
22899 section of elaboration code that contains the declaration of
22902 There would never be any order of elaboration problem if we made a rule
22903 that whenever you @code{with} a unit, you must elaborate both the spec and body
22904 of that unit before elaborating the unit doing the @code{with}'ing:
22906 @smallexample @c ada
22910 package Unit_2 is ...
22916 would require that both the body and spec of @code{Unit_1} be elaborated
22917 before the spec of @code{Unit_2}. However, a rule like that would be far too
22918 restrictive. In particular, it would make it impossible to have routines
22919 in separate packages that were mutually recursive.
22921 You might think that a clever enough compiler could look at the actual
22922 elaboration code and determine an appropriate correct order of elaboration,
22923 but in the general case, this is not possible. Consider the following
22926 In the body of @code{Unit_1}, we have a procedure @code{Func_1}
22928 the variable @code{Sqrt_1}, which is declared in the elaboration code
22929 of the body of @code{Unit_1}:
22931 @smallexample @c ada
22933 Sqrt_1 : Float := Sqrt (0.1);
22938 The elaboration code of the body of @code{Unit_1} also contains:
22940 @smallexample @c ada
22943 if expression_1 = 1 then
22944 Q := Unit_2.Func_2;
22951 @code{Unit_2} is exactly parallel,
22952 it has a procedure @code{Func_2} that references
22953 the variable @code{Sqrt_2}, which is declared in the elaboration code of
22954 the body @code{Unit_2}:
22956 @smallexample @c ada
22958 Sqrt_2 : Float := Sqrt (0.1);
22963 The elaboration code of the body of @code{Unit_2} also contains:
22965 @smallexample @c ada
22968 if expression_2 = 2 then
22969 Q := Unit_1.Func_1;
22976 Now the question is, which of the following orders of elaboration is
23001 If you carefully analyze the flow here, you will see that you cannot tell
23002 at compile time the answer to this question.
23003 If @code{expression_1} is not equal to 1,
23004 and @code{expression_2} is not equal to 2,
23005 then either order is acceptable, because neither of the function calls is
23006 executed. If both tests evaluate to true, then neither order is acceptable
23007 and in fact there is no correct order.
23009 If one of the two expressions is true, and the other is false, then one
23010 of the above orders is correct, and the other is incorrect. For example,
23011 if @code{expression_1} = 1 and @code{expression_2} /= 2,
23012 then the call to @code{Func_2}
23013 will occur, but not the call to @code{Func_1.}
23014 This means that it is essential
23015 to elaborate the body of @code{Unit_1} before
23016 the body of @code{Unit_2}, so the first
23017 order of elaboration is correct and the second is wrong.
23019 By making @code{expression_1} and @code{expression_2}
23020 depend on input data, or perhaps
23021 the time of day, we can make it impossible for the compiler or binder
23022 to figure out which of these expressions will be true, and hence it
23023 is impossible to guarantee a safe order of elaboration at run time.
23025 @node Checking the Elaboration Order in Ada 95
23026 @section Checking the Elaboration Order in Ada 95
23029 In some languages that involve the same kind of elaboration problems,
23030 e.g. Java and C++, the programmer is expected to worry about these
23031 ordering problems himself, and it is common to
23032 write a program in which an incorrect elaboration order gives
23033 surprising results, because it references variables before they
23035 Ada 95 is designed to be a safe language, and a programmer-beware approach is
23036 clearly not sufficient. Consequently, the language provides three lines
23040 @item Standard rules
23041 Some standard rules restrict the possible choice of elaboration
23042 order. In particular, if you @code{with} a unit, then its spec is always
23043 elaborated before the unit doing the @code{with}. Similarly, a parent
23044 spec is always elaborated before the child spec, and finally
23045 a spec is always elaborated before its corresponding body.
23047 @item Dynamic elaboration checks
23048 @cindex Elaboration checks
23049 @cindex Checks, elaboration
23050 Dynamic checks are made at run time, so that if some entity is accessed
23051 before it is elaborated (typically by means of a subprogram call)
23052 then the exception (@code{Program_Error}) is raised.
23054 @item Elaboration control
23055 Facilities are provided for the programmer to specify the desired order
23059 Let's look at these facilities in more detail. First, the rules for
23060 dynamic checking. One possible rule would be simply to say that the
23061 exception is raised if you access a variable which has not yet been
23062 elaborated. The trouble with this approach is that it could require
23063 expensive checks on every variable reference. Instead Ada 95 has two
23064 rules which are a little more restrictive, but easier to check, and
23068 @item Restrictions on calls
23069 A subprogram can only be called at elaboration time if its body
23070 has been elaborated. The rules for elaboration given above guarantee
23071 that the spec of the subprogram has been elaborated before the
23072 call, but not the body. If this rule is violated, then the
23073 exception @code{Program_Error} is raised.
23075 @item Restrictions on instantiations
23076 A generic unit can only be instantiated if the body of the generic
23077 unit has been elaborated. Again, the rules for elaboration given above
23078 guarantee that the spec of the generic unit has been elaborated
23079 before the instantiation, but not the body. If this rule is
23080 violated, then the exception @code{Program_Error} is raised.
23084 The idea is that if the body has been elaborated, then any variables
23085 it references must have been elaborated; by checking for the body being
23086 elaborated we guarantee that none of its references causes any
23087 trouble. As we noted above, this is a little too restrictive, because a
23088 subprogram that has no non-local references in its body may in fact be safe
23089 to call. However, it really would be unsafe to rely on this, because
23090 it would mean that the caller was aware of details of the implementation
23091 in the body. This goes against the basic tenets of Ada.
23093 A plausible implementation can be described as follows.
23094 A Boolean variable is associated with each subprogram
23095 and each generic unit. This variable is initialized to False, and is set to
23096 True at the point body is elaborated. Every call or instantiation checks the
23097 variable, and raises @code{Program_Error} if the variable is False.
23099 Note that one might think that it would be good enough to have one Boolean
23100 variable for each package, but that would not deal with cases of trying
23101 to call a body in the same package as the call
23102 that has not been elaborated yet.
23103 Of course a compiler may be able to do enough analysis to optimize away
23104 some of the Boolean variables as unnecessary, and @code{GNAT} indeed
23105 does such optimizations, but still the easiest conceptual model is to
23106 think of there being one variable per subprogram.
23108 @node Controlling the Elaboration Order in Ada 95
23109 @section Controlling the Elaboration Order in Ada 95
23112 In the previous section we discussed the rules in Ada 95 which ensure
23113 that @code{Program_Error} is raised if an incorrect elaboration order is
23114 chosen. This prevents erroneous executions, but we need mechanisms to
23115 specify a correct execution and avoid the exception altogether.
23116 To achieve this, Ada 95 provides a number of features for controlling
23117 the order of elaboration. We discuss these features in this section.
23119 First, there are several ways of indicating to the compiler that a given
23120 unit has no elaboration problems:
23123 @item packages that do not require a body
23124 In Ada 95, a library package that does not require a body does not permit
23125 a body. This means that if we have a such a package, as in:
23127 @smallexample @c ada
23130 package Definitions is
23132 type m is new integer;
23134 type a is array (1 .. 10) of m;
23135 type b is array (1 .. 20) of m;
23143 A package that @code{with}'s @code{Definitions} may safely instantiate
23144 @code{Definitions.Subp} because the compiler can determine that there
23145 definitely is no package body to worry about in this case
23148 @cindex pragma Pure
23150 Places sufficient restrictions on a unit to guarantee that
23151 no call to any subprogram in the unit can result in an
23152 elaboration problem. This means that the compiler does not need
23153 to worry about the point of elaboration of such units, and in
23154 particular, does not need to check any calls to any subprograms
23157 @item pragma Preelaborate
23158 @findex Preelaborate
23159 @cindex pragma Preelaborate
23160 This pragma places slightly less stringent restrictions on a unit than
23162 but these restrictions are still sufficient to ensure that there
23163 are no elaboration problems with any calls to the unit.
23165 @item pragma Elaborate_Body
23166 @findex Elaborate_Body
23167 @cindex pragma Elaborate_Body
23168 This pragma requires that the body of a unit be elaborated immediately
23169 after its spec. Suppose a unit @code{A} has such a pragma,
23170 and unit @code{B} does
23171 a @code{with} of unit @code{A}. Recall that the standard rules require
23172 the spec of unit @code{A}
23173 to be elaborated before the @code{with}'ing unit; given the pragma in
23174 @code{A}, we also know that the body of @code{A}
23175 will be elaborated before @code{B}, so
23176 that calls to @code{A} are safe and do not need a check.
23181 unlike pragma @code{Pure} and pragma @code{Preelaborate},
23183 @code{Elaborate_Body} does not guarantee that the program is
23184 free of elaboration problems, because it may not be possible
23185 to satisfy the requested elaboration order.
23186 Let's go back to the example with @code{Unit_1} and @code{Unit_2}.
23188 marks @code{Unit_1} as @code{Elaborate_Body},
23189 and not @code{Unit_2,} then the order of
23190 elaboration will be:
23202 Now that means that the call to @code{Func_1} in @code{Unit_2}
23203 need not be checked,
23204 it must be safe. But the call to @code{Func_2} in
23205 @code{Unit_1} may still fail if
23206 @code{Expression_1} is equal to 1,
23207 and the programmer must still take
23208 responsibility for this not being the case.
23210 If all units carry a pragma @code{Elaborate_Body}, then all problems are
23211 eliminated, except for calls entirely within a body, which are
23212 in any case fully under programmer control. However, using the pragma
23213 everywhere is not always possible.
23214 In particular, for our @code{Unit_1}/@code{Unit_2} example, if
23215 we marked both of them as having pragma @code{Elaborate_Body}, then
23216 clearly there would be no possible elaboration order.
23218 The above pragmas allow a server to guarantee safe use by clients, and
23219 clearly this is the preferable approach. Consequently a good rule in
23220 Ada 95 is to mark units as @code{Pure} or @code{Preelaborate} if possible,
23221 and if this is not possible,
23222 mark them as @code{Elaborate_Body} if possible.
23223 As we have seen, there are situations where neither of these
23224 three pragmas can be used.
23225 So we also provide methods for clients to control the
23226 order of elaboration of the servers on which they depend:
23229 @item pragma Elaborate (unit)
23231 @cindex pragma Elaborate
23232 This pragma is placed in the context clause, after a @code{with} clause,
23233 and it requires that the body of the named unit be elaborated before
23234 the unit in which the pragma occurs. The idea is to use this pragma
23235 if the current unit calls at elaboration time, directly or indirectly,
23236 some subprogram in the named unit.
23238 @item pragma Elaborate_All (unit)
23239 @findex Elaborate_All
23240 @cindex pragma Elaborate_All
23241 This is a stronger version of the Elaborate pragma. Consider the
23245 Unit A @code{with}'s unit B and calls B.Func in elab code
23246 Unit B @code{with}'s unit C, and B.Func calls C.Func
23250 Now if we put a pragma @code{Elaborate (B)}
23251 in unit @code{A}, this ensures that the
23252 body of @code{B} is elaborated before the call, but not the
23253 body of @code{C}, so
23254 the call to @code{C.Func} could still cause @code{Program_Error} to
23257 The effect of a pragma @code{Elaborate_All} is stronger, it requires
23258 not only that the body of the named unit be elaborated before the
23259 unit doing the @code{with}, but also the bodies of all units that the
23260 named unit uses, following @code{with} links transitively. For example,
23261 if we put a pragma @code{Elaborate_All (B)} in unit @code{A},
23263 not only that the body of @code{B} be elaborated before @code{A},
23265 body of @code{C}, because @code{B} @code{with}'s @code{C}.
23269 We are now in a position to give a usage rule in Ada 95 for avoiding
23270 elaboration problems, at least if dynamic dispatching and access to
23271 subprogram values are not used. We will handle these cases separately
23274 The rule is simple. If a unit has elaboration code that can directly or
23275 indirectly make a call to a subprogram in a @code{with}'ed unit, or instantiate
23276 a generic unit in a @code{with}'ed unit,
23277 then if the @code{with}'ed unit does not have
23278 pragma @code{Pure} or @code{Preelaborate}, then the client should have
23279 a pragma @code{Elaborate_All}
23280 for the @code{with}'ed unit. By following this rule a client is
23281 assured that calls can be made without risk of an exception.
23282 If this rule is not followed, then a program may be in one of four
23286 @item No order exists
23287 No order of elaboration exists which follows the rules, taking into
23288 account any @code{Elaborate}, @code{Elaborate_All},
23289 or @code{Elaborate_Body} pragmas. In
23290 this case, an Ada 95 compiler must diagnose the situation at bind
23291 time, and refuse to build an executable program.
23293 @item One or more orders exist, all incorrect
23294 One or more acceptable elaboration orders exists, and all of them
23295 generate an elaboration order problem. In this case, the binder
23296 can build an executable program, but @code{Program_Error} will be raised
23297 when the program is run.
23299 @item Several orders exist, some right, some incorrect
23300 One or more acceptable elaboration orders exists, and some of them
23301 work, and some do not. The programmer has not controlled
23302 the order of elaboration, so the binder may or may not pick one of
23303 the correct orders, and the program may or may not raise an
23304 exception when it is run. This is the worst case, because it means
23305 that the program may fail when moved to another compiler, or even
23306 another version of the same compiler.
23308 @item One or more orders exists, all correct
23309 One ore more acceptable elaboration orders exist, and all of them
23310 work. In this case the program runs successfully. This state of
23311 affairs can be guaranteed by following the rule we gave above, but
23312 may be true even if the rule is not followed.
23316 Note that one additional advantage of following our Elaborate_All rule
23317 is that the program continues to stay in the ideal (all orders OK) state
23318 even if maintenance
23319 changes some bodies of some subprograms. Conversely, if a program that does
23320 not follow this rule happens to be safe at some point, this state of affairs
23321 may deteriorate silently as a result of maintenance changes.
23323 You may have noticed that the above discussion did not mention
23324 the use of @code{Elaborate_Body}. This was a deliberate omission. If you
23325 @code{with} an @code{Elaborate_Body} unit, it still may be the case that
23326 code in the body makes calls to some other unit, so it is still necessary
23327 to use @code{Elaborate_All} on such units.
23329 @node Controlling Elaboration in GNAT - Internal Calls
23330 @section Controlling Elaboration in GNAT - Internal Calls
23333 In the case of internal calls, i.e. calls within a single package, the
23334 programmer has full control over the order of elaboration, and it is up
23335 to the programmer to elaborate declarations in an appropriate order. For
23338 @smallexample @c ada
23341 function One return Float;
23345 function One return Float is
23354 will obviously raise @code{Program_Error} at run time, because function
23355 One will be called before its body is elaborated. In this case GNAT will
23356 generate a warning that the call will raise @code{Program_Error}:
23362 2. function One return Float;
23364 4. Q : Float := One;
23366 >>> warning: cannot call "One" before body is elaborated
23367 >>> warning: Program_Error will be raised at run time
23370 6. function One return Float is
23383 Note that in this particular case, it is likely that the call is safe, because
23384 the function @code{One} does not access any global variables.
23385 Nevertheless in Ada 95, we do not want the validity of the check to depend on
23386 the contents of the body (think about the separate compilation case), so this
23387 is still wrong, as we discussed in the previous sections.
23389 The error is easily corrected by rearranging the declarations so that the
23390 body of One appears before the declaration containing the call
23391 (note that in Ada 95,
23392 declarations can appear in any order, so there is no restriction that
23393 would prevent this reordering, and if we write:
23395 @smallexample @c ada
23398 function One return Float;
23400 function One return Float is
23411 then all is well, no warning is generated, and no
23412 @code{Program_Error} exception
23414 Things are more complicated when a chain of subprograms is executed:
23416 @smallexample @c ada
23419 function A return Integer;
23420 function B return Integer;
23421 function C return Integer;
23423 function B return Integer is begin return A; end;
23424 function C return Integer is begin return B; end;
23428 function A return Integer is begin return 1; end;
23434 Now the call to @code{C}
23435 at elaboration time in the declaration of @code{X} is correct, because
23436 the body of @code{C} is already elaborated,
23437 and the call to @code{B} within the body of
23438 @code{C} is correct, but the call
23439 to @code{A} within the body of @code{B} is incorrect, because the body
23440 of @code{A} has not been elaborated, so @code{Program_Error}
23441 will be raised on the call to @code{A}.
23442 In this case GNAT will generate a
23443 warning that @code{Program_Error} may be
23444 raised at the point of the call. Let's look at the warning:
23450 2. function A return Integer;
23451 3. function B return Integer;
23452 4. function C return Integer;
23454 6. function B return Integer is begin return A; end;
23456 >>> warning: call to "A" before body is elaborated may
23457 raise Program_Error
23458 >>> warning: "B" called at line 7
23459 >>> warning: "C" called at line 9
23461 7. function C return Integer is begin return B; end;
23463 9. X : Integer := C;
23465 11. function A return Integer is begin return 1; end;
23475 Note that the message here says ``may raise'', instead of the direct case,
23476 where the message says ``will be raised''. That's because whether
23478 actually called depends in general on run-time flow of control.
23479 For example, if the body of @code{B} said
23481 @smallexample @c ada
23484 function B return Integer is
23486 if some-condition-depending-on-input-data then
23497 then we could not know until run time whether the incorrect call to A would
23498 actually occur, so @code{Program_Error} might
23499 or might not be raised. It is possible for a compiler to
23500 do a better job of analyzing bodies, to
23501 determine whether or not @code{Program_Error}
23502 might be raised, but it certainly
23503 couldn't do a perfect job (that would require solving the halting problem
23504 and is provably impossible), and because this is a warning anyway, it does
23505 not seem worth the effort to do the analysis. Cases in which it
23506 would be relevant are rare.
23508 In practice, warnings of either of the forms given
23509 above will usually correspond to
23510 real errors, and should be examined carefully and eliminated.
23511 In the rare case where a warning is bogus, it can be suppressed by any of
23512 the following methods:
23516 Compile with the @option{-gnatws} switch set
23519 Suppress @code{Elaboration_Check} for the called subprogram
23522 Use pragma @code{Warnings_Off} to turn warnings off for the call
23526 For the internal elaboration check case,
23527 GNAT by default generates the
23528 necessary run-time checks to ensure
23529 that @code{Program_Error} is raised if any
23530 call fails an elaboration check. Of course this can only happen if a
23531 warning has been issued as described above. The use of pragma
23532 @code{Suppress (Elaboration_Check)} may (but is not guaranteed to) suppress
23533 some of these checks, meaning that it may be possible (but is not
23534 guaranteed) for a program to be able to call a subprogram whose body
23535 is not yet elaborated, without raising a @code{Program_Error} exception.
23537 @node Controlling Elaboration in GNAT - External Calls
23538 @section Controlling Elaboration in GNAT - External Calls
23541 The previous section discussed the case in which the execution of a
23542 particular thread of elaboration code occurred entirely within a
23543 single unit. This is the easy case to handle, because a programmer
23544 has direct and total control over the order of elaboration, and
23545 furthermore, checks need only be generated in cases which are rare
23546 and which the compiler can easily detect.
23547 The situation is more complex when separate compilation is taken into account.
23548 Consider the following:
23550 @smallexample @c ada
23554 function Sqrt (Arg : Float) return Float;
23557 package body Math is
23558 function Sqrt (Arg : Float) return Float is
23567 X : Float := Math.Sqrt (0.5);
23580 where @code{Main} is the main program. When this program is executed, the
23581 elaboration code must first be executed, and one of the jobs of the
23582 binder is to determine the order in which the units of a program are
23583 to be elaborated. In this case we have four units: the spec and body
23585 the spec of @code{Stuff} and the body of @code{Main}).
23586 In what order should the four separate sections of elaboration code
23589 There are some restrictions in the order of elaboration that the binder
23590 can choose. In particular, if unit U has a @code{with}
23591 for a package @code{X}, then you
23592 are assured that the spec of @code{X}
23593 is elaborated before U , but you are
23594 not assured that the body of @code{X}
23595 is elaborated before U.
23596 This means that in the above case, the binder is allowed to choose the
23607 but that's not good, because now the call to @code{Math.Sqrt}
23608 that happens during
23609 the elaboration of the @code{Stuff}
23610 spec happens before the body of @code{Math.Sqrt} is
23611 elaborated, and hence causes @code{Program_Error} exception to be raised.
23612 At first glance, one might say that the binder is misbehaving, because
23613 obviously you want to elaborate the body of something you @code{with}
23615 that is not a general rule that can be followed in all cases. Consider
23617 @smallexample @c ada
23625 package body Y is ...
23628 package body X is ...
23634 This is a common arrangement, and, apart from the order of elaboration
23635 problems that might arise in connection with elaboration code, this works fine.
23636 A rule that says that you must first elaborate the body of anything you
23637 @code{with} cannot work in this case:
23638 the body of @code{X} @code{with}'s @code{Y},
23639 which means you would have to
23640 elaborate the body of @code{Y} first, but that @code{with}'s @code{X},
23642 you have to elaborate the body of @code{X} first, but ... and we have a
23643 loop that cannot be broken.
23645 It is true that the binder can in many cases guess an order of elaboration
23646 that is unlikely to cause a @code{Program_Error}
23647 exception to be raised, and it tries to do so (in the
23648 above example of @code{Math/Stuff/Spec}, the GNAT binder will
23650 elaborate the body of @code{Math} right after its spec, so all will be well).
23652 However, a program that blindly relies on the binder to be helpful can
23653 get into trouble, as we discussed in the previous sections, so
23655 provides a number of facilities for assisting the programmer in
23656 developing programs that are robust with respect to elaboration order.
23658 @node Default Behavior in GNAT - Ensuring Safety
23659 @section Default Behavior in GNAT - Ensuring Safety
23662 The default behavior in GNAT ensures elaboration safety. In its
23663 default mode GNAT implements the
23664 rule we previously described as the right approach. Let's restate it:
23668 @emph{If a unit has elaboration code that can directly or indirectly make a
23669 call to a subprogram in a @code{with}'ed unit, or instantiate a generic unit
23670 in a @code{with}'ed unit, then if the @code{with}'ed unit
23671 does not have pragma @code{Pure} or
23672 @code{Preelaborate}, then the client should have an
23673 @code{Elaborate_All} for the @code{with}'ed unit.}
23677 By following this rule a client is assured that calls and instantiations
23678 can be made without risk of an exception.
23680 In this mode GNAT traces all calls that are potentially made from
23681 elaboration code, and puts in any missing implicit @code{Elaborate_All}
23683 The advantage of this approach is that no elaboration problems
23684 are possible if the binder can find an elaboration order that is
23685 consistent with these implicit @code{Elaborate_All} pragmas. The
23686 disadvantage of this approach is that no such order may exist.
23688 If the binder does not generate any diagnostics, then it means that it
23689 has found an elaboration order that is guaranteed to be safe. However,
23690 the binder may still be relying on implicitly generated
23691 @code{Elaborate_All} pragmas so portability to other compilers than
23692 GNAT is not guaranteed.
23694 If it is important to guarantee portability, then the compilations should
23697 (warn on elaboration problems) switch. This will cause warning messages
23698 to be generated indicating the missing @code{Elaborate_All} pragmas.
23699 Consider the following source program:
23701 @smallexample @c ada
23706 m : integer := k.r;
23713 where it is clear that there
23714 should be a pragma @code{Elaborate_All}
23715 for unit @code{k}. An implicit pragma will be generated, and it is
23716 likely that the binder will be able to honor it. However, if you want
23717 to port this program to some other Ada compiler than GNAT.
23718 it is safer to include the pragma explicitly in the source. If this
23719 unit is compiled with the
23721 switch, then the compiler outputs a warning:
23728 3. m : integer := k.r;
23730 >>> warning: call to "r" may raise Program_Error
23731 >>> warning: missing pragma Elaborate_All for "k"
23739 and these warnings can be used as a guide for supplying manually
23740 the missing pragmas. It is usually a bad idea to use this warning
23741 option during development. That's because it will warn you when
23742 you need to put in a pragma, but cannot warn you when it is time
23743 to take it out. So the use of pragma Elaborate_All may lead to
23744 unnecessary dependencies and even false circularities.
23746 This default mode is more restrictive than the Ada Reference
23747 Manual, and it is possible to construct programs which will compile
23748 using the dynamic model described there, but will run into a
23749 circularity using the safer static model we have described.
23751 Of course any Ada compiler must be able to operate in a mode
23752 consistent with the requirements of the Ada Reference Manual,
23753 and in particular must have the capability of implementing the
23754 standard dynamic model of elaboration with run-time checks.
23756 In GNAT, this standard mode can be achieved either by the use of
23757 the @option{-gnatE} switch on the compiler (@command{gcc} or
23758 @command{gnatmake}) command, or by the use of the configuration pragma:
23760 @smallexample @c ada
23761 pragma Elaboration_Checks (RM);
23765 Either approach will cause the unit affected to be compiled using the
23766 standard dynamic run-time elaboration checks described in the Ada
23767 Reference Manual. The static model is generally preferable, since it
23768 is clearly safer to rely on compile and link time checks rather than
23769 run-time checks. However, in the case of legacy code, it may be
23770 difficult to meet the requirements of the static model. This
23771 issue is further discussed in
23772 @ref{What to Do If the Default Elaboration Behavior Fails}.
23774 Note that the static model provides a strict subset of the allowed
23775 behavior and programs of the Ada Reference Manual, so if you do
23776 adhere to the static model and no circularities exist,
23777 then you are assured that your program will
23778 work using the dynamic model, providing that you remove any
23779 pragma Elaborate statements from the source.
23781 @node Treatment of Pragma Elaborate
23782 @section Treatment of Pragma Elaborate
23783 @cindex Pragma Elaborate
23786 The use of @code{pragma Elaborate}
23787 should generally be avoided in Ada 95 programs.
23788 The reason for this is that there is no guarantee that transitive calls
23789 will be properly handled. Indeed at one point, this pragma was placed
23790 in Annex J (Obsolescent Features), on the grounds that it is never useful.
23792 Now that's a bit restrictive. In practice, the case in which
23793 @code{pragma Elaborate} is useful is when the caller knows that there
23794 are no transitive calls, or that the called unit contains all necessary
23795 transitive @code{pragma Elaborate} statements, and legacy code often
23796 contains such uses.
23798 Strictly speaking the static mode in GNAT should ignore such pragmas,
23799 since there is no assurance at compile time that the necessary safety
23800 conditions are met. In practice, this would cause GNAT to be incompatible
23801 with correctly written Ada 83 code that had all necessary
23802 @code{pragma Elaborate} statements in place. Consequently, we made the
23803 decision that GNAT in its default mode will believe that if it encounters
23804 a @code{pragma Elaborate} then the programmer knows what they are doing,
23805 and it will trust that no elaboration errors can occur.
23807 The result of this decision is two-fold. First to be safe using the
23808 static mode, you should remove all @code{pragma Elaborate} statements.
23809 Second, when fixing circularities in existing code, you can selectively
23810 use @code{pragma Elaborate} statements to convince the static mode of
23811 GNAT that it need not generate an implicit @code{pragma Elaborate_All}
23814 When using the static mode with @option{-gnatwl}, any use of
23815 @code{pragma Elaborate} will generate a warning about possible
23818 @node Elaboration Issues for Library Tasks
23819 @section Elaboration Issues for Library Tasks
23820 @cindex Library tasks, elaboration issues
23821 @cindex Elaboration of library tasks
23824 In this section we examine special elaboration issues that arise for
23825 programs that declare library level tasks.
23827 Generally the model of execution of an Ada program is that all units are
23828 elaborated, and then execution of the program starts. However, the
23829 declaration of library tasks definitely does not fit this model. The
23830 reason for this is that library tasks start as soon as they are declared
23831 (more precisely, as soon as the statement part of the enclosing package
23832 body is reached), that is to say before elaboration
23833 of the program is complete. This means that if such a task calls a
23834 subprogram, or an entry in another task, the callee may or may not be
23835 elaborated yet, and in the standard
23836 Reference Manual model of dynamic elaboration checks, you can even
23837 get timing dependent Program_Error exceptions, since there can be
23838 a race between the elaboration code and the task code.
23840 The static model of elaboration in GNAT seeks to avoid all such
23841 dynamic behavior, by being conservative, and the conservative
23842 approach in this particular case is to assume that all the code
23843 in a task body is potentially executed at elaboration time if
23844 a task is declared at the library level.
23846 This can definitely result in unexpected circularities. Consider
23847 the following example
23849 @smallexample @c ada
23855 type My_Int is new Integer;
23857 function Ident (M : My_Int) return My_Int;
23861 package body Decls is
23862 task body Lib_Task is
23868 function Ident (M : My_Int) return My_Int is
23876 procedure Put_Val (Arg : Decls.My_Int);
23880 package body Utils is
23881 procedure Put_Val (Arg : Decls.My_Int) is
23883 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
23890 Decls.Lib_Task.Start;
23895 If the above example is compiled in the default static elaboration
23896 mode, then a circularity occurs. The circularity comes from the call
23897 @code{Utils.Put_Val} in the task body of @code{Decls.Lib_Task}. Since
23898 this call occurs in elaboration code, we need an implicit pragma
23899 @code{Elaborate_All} for @code{Utils}. This means that not only must
23900 the spec and body of @code{Utils} be elaborated before the body
23901 of @code{Decls}, but also the spec and body of any unit that is
23902 @code{with'ed} by the body of @code{Utils} must also be elaborated before
23903 the body of @code{Decls}. This is the transitive implication of
23904 pragma @code{Elaborate_All} and it makes sense, because in general
23905 the body of @code{Put_Val} might have a call to something in a
23906 @code{with'ed} unit.
23908 In this case, the body of Utils (actually its spec) @code{with's}
23909 @code{Decls}. Unfortunately this means that the body of @code{Decls}
23910 must be elaborated before itself, in case there is a call from the
23911 body of @code{Utils}.
23913 Here is the exact chain of events we are worrying about:
23917 In the body of @code{Decls} a call is made from within the body of a library
23918 task to a subprogram in the package @code{Utils}. Since this call may
23919 occur at elaboration time (given that the task is activated at elaboration
23920 time), we have to assume the worst, i.e. that the
23921 call does happen at elaboration time.
23924 This means that the body and spec of @code{Util} must be elaborated before
23925 the body of @code{Decls} so that this call does not cause an access before
23929 Within the body of @code{Util}, specifically within the body of
23930 @code{Util.Put_Val} there may be calls to any unit @code{with}'ed
23934 One such @code{with}'ed package is package @code{Decls}, so there
23935 might be a call to a subprogram in @code{Decls} in @code{Put_Val}.
23936 In fact there is such a call in this example, but we would have to
23937 assume that there was such a call even if it were not there, since
23938 we are not supposed to write the body of @code{Decls} knowing what
23939 is in the body of @code{Utils}; certainly in the case of the
23940 static elaboration model, the compiler does not know what is in
23941 other bodies and must assume the worst.
23944 This means that the spec and body of @code{Decls} must also be
23945 elaborated before we elaborate the unit containing the call, but
23946 that unit is @code{Decls}! This means that the body of @code{Decls}
23947 must be elaborated before itself, and that's a circularity.
23951 Indeed, if you add an explicit pragma Elaborate_All for @code{Utils} in
23952 the body of @code{Decls} you will get a true Ada Reference Manual
23953 circularity that makes the program illegal.
23955 In practice, we have found that problems with the static model of
23956 elaboration in existing code often arise from library tasks, so
23957 we must address this particular situation.
23959 Note that if we compile and run the program above, using the dynamic model of
23960 elaboration (that is to say use the @option{-gnatE} switch),
23961 then it compiles, binds,
23962 links, and runs, printing the expected result of 2. Therefore in some sense
23963 the circularity here is only apparent, and we need to capture
23964 the properties of this program that distinguish it from other library-level
23965 tasks that have real elaboration problems.
23967 We have four possible answers to this question:
23972 Use the dynamic model of elaboration.
23974 If we use the @option{-gnatE} switch, then as noted above, the program works.
23975 Why is this? If we examine the task body, it is apparent that the task cannot
23977 @code{accept} statement until after elaboration has been completed, because
23978 the corresponding entry call comes from the main program, not earlier.
23979 This is why the dynamic model works here. But that's really giving
23980 up on a precise analysis, and we prefer to take this approach only if we cannot
23982 problem in any other manner. So let us examine two ways to reorganize
23983 the program to avoid the potential elaboration problem.
23986 Split library tasks into separate packages.
23988 Write separate packages, so that library tasks are isolated from
23989 other declarations as much as possible. Let us look at a variation on
23992 @smallexample @c ada
24000 package body Decls1 is
24001 task body Lib_Task is
24009 type My_Int is new Integer;
24010 function Ident (M : My_Int) return My_Int;
24014 package body Decls2 is
24015 function Ident (M : My_Int) return My_Int is
24023 procedure Put_Val (Arg : Decls2.My_Int);
24027 package body Utils is
24028 procedure Put_Val (Arg : Decls2.My_Int) is
24030 Text_IO.Put_Line (Decls2.My_Int'Image (Decls2.Ident (Arg)));
24037 Decls1.Lib_Task.Start;
24042 All we have done is to split @code{Decls} into two packages, one
24043 containing the library task, and one containing everything else. Now
24044 there is no cycle, and the program compiles, binds, links and executes
24045 using the default static model of elaboration.
24048 Declare separate task types.
24050 A significant part of the problem arises because of the use of the
24051 single task declaration form. This means that the elaboration of
24052 the task type, and the elaboration of the task itself (i.e. the
24053 creation of the task) happen at the same time. A good rule
24054 of style in Ada 95 is to always create explicit task types. By
24055 following the additional step of placing task objects in separate
24056 packages from the task type declaration, many elaboration problems
24057 are avoided. Here is another modified example of the example program:
24059 @smallexample @c ada
24061 task type Lib_Task_Type is
24065 type My_Int is new Integer;
24067 function Ident (M : My_Int) return My_Int;
24071 package body Decls is
24072 task body Lib_Task_Type is
24078 function Ident (M : My_Int) return My_Int is
24086 procedure Put_Val (Arg : Decls.My_Int);
24090 package body Utils is
24091 procedure Put_Val (Arg : Decls.My_Int) is
24093 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
24099 Lib_Task : Decls.Lib_Task_Type;
24105 Declst.Lib_Task.Start;
24110 What we have done here is to replace the @code{task} declaration in
24111 package @code{Decls} with a @code{task type} declaration. Then we
24112 introduce a separate package @code{Declst} to contain the actual
24113 task object. This separates the elaboration issues for
24114 the @code{task type}
24115 declaration, which causes no trouble, from the elaboration issues
24116 of the task object, which is also unproblematic, since it is now independent
24117 of the elaboration of @code{Utils}.
24118 This separation of concerns also corresponds to
24119 a generally sound engineering principle of separating declarations
24120 from instances. This version of the program also compiles, binds, links,
24121 and executes, generating the expected output.
24124 Use No_Entry_Calls_In_Elaboration_Code restriction.
24125 @cindex No_Entry_Calls_In_Elaboration_Code
24127 The previous two approaches described how a program can be restructured
24128 to avoid the special problems caused by library task bodies. in practice,
24129 however, such restructuring may be difficult to apply to existing legacy code,
24130 so we must consider solutions that do not require massive rewriting.
24132 Let us consider more carefully why our original sample program works
24133 under the dynamic model of elaboration. The reason is that the code
24134 in the task body blocks immediately on the @code{accept}
24135 statement. Now of course there is nothing to prohibit elaboration
24136 code from making entry calls (for example from another library level task),
24137 so we cannot tell in isolation that
24138 the task will not execute the accept statement during elaboration.
24140 However, in practice it is very unusual to see elaboration code
24141 make any entry calls, and the pattern of tasks starting
24142 at elaboration time and then immediately blocking on @code{accept} or
24143 @code{select} statements is very common. What this means is that
24144 the compiler is being too pessimistic when it analyzes the
24145 whole package body as though it might be executed at elaboration
24148 If we know that the elaboration code contains no entry calls, (a very safe
24149 assumption most of the time, that could almost be made the default
24150 behavior), then we can compile all units of the program under control
24151 of the following configuration pragma:
24154 pragma Restrictions (No_Entry_Calls_In_Elaboration_Code);
24158 This pragma can be placed in the @file{gnat.adc} file in the usual
24159 manner. If we take our original unmodified program and compile it
24160 in the presence of a @file{gnat.adc} containing the above pragma,
24161 then once again, we can compile, bind, link, and execute, obtaining
24162 the expected result. In the presence of this pragma, the compiler does
24163 not trace calls in a task body, that appear after the first @code{accept}
24164 or @code{select} statement, and therefore does not report a potential
24165 circularity in the original program.
24167 The compiler will check to the extent it can that the above
24168 restriction is not violated, but it is not always possible to do a
24169 complete check at compile time, so it is important to use this
24170 pragma only if the stated restriction is in fact met, that is to say
24171 no task receives an entry call before elaboration of all units is completed.
24175 @node Mixing Elaboration Models
24176 @section Mixing Elaboration Models
24178 So far, we have assumed that the entire program is either compiled
24179 using the dynamic model or static model, ensuring consistency. It
24180 is possible to mix the two models, but rules have to be followed
24181 if this mixing is done to ensure that elaboration checks are not
24184 The basic rule is that @emph{a unit compiled with the static model cannot
24185 be @code{with'ed} by a unit compiled with the dynamic model}. The
24186 reason for this is that in the static model, a unit assumes that
24187 its clients guarantee to use (the equivalent of) pragma
24188 @code{Elaborate_All} so that no elaboration checks are required
24189 in inner subprograms, and this assumption is violated if the
24190 client is compiled with dynamic checks.
24192 The precise rule is as follows. A unit that is compiled with dynamic
24193 checks can only @code{with} a unit that meets at least one of the
24194 following criteria:
24199 The @code{with'ed} unit is itself compiled with dynamic elaboration
24200 checks (that is with the @option{-gnatE} switch.
24203 The @code{with'ed} unit is an internal GNAT implementation unit from
24204 the System, Interfaces, Ada, or GNAT hierarchies.
24207 The @code{with'ed} unit has pragma Preelaborate or pragma Pure.
24210 The @code{with'ing} unit (that is the client) has an explicit pragma
24211 @code{Elaborate_All} for the @code{with'ed} unit.
24216 If this rule is violated, that is if a unit with dynamic elaboration
24217 checks @code{with's} a unit that does not meet one of the above four
24218 criteria, then the binder (@code{gnatbind}) will issue a warning
24219 similar to that in the following example:
24222 warning: "x.ads" has dynamic elaboration checks and with's
24223 warning: "y.ads" which has static elaboration checks
24227 These warnings indicate that the rule has been violated, and that as a result
24228 elaboration checks may be missed in the resulting executable file.
24229 This warning may be suppressed using the @option{-ws} binder switch
24230 in the usual manner.
24232 One useful application of this mixing rule is in the case of a subsystem
24233 which does not itself @code{with} units from the remainder of the
24234 application. In this case, the entire subsystem can be compiled with
24235 dynamic checks to resolve a circularity in the subsystem, while
24236 allowing the main application that uses this subsystem to be compiled
24237 using the more reliable default static model.
24239 @node What to Do If the Default Elaboration Behavior Fails
24240 @section What to Do If the Default Elaboration Behavior Fails
24243 If the binder cannot find an acceptable order, it outputs detailed
24244 diagnostics. For example:
24250 error: elaboration circularity detected
24251 info: "proc (body)" must be elaborated before "pack (body)"
24252 info: reason: Elaborate_All probably needed in unit "pack (body)"
24253 info: recompile "pack (body)" with -gnatwl
24254 info: for full details
24255 info: "proc (body)"
24256 info: is needed by its spec:
24257 info: "proc (spec)"
24258 info: which is withed by:
24259 info: "pack (body)"
24260 info: "pack (body)" must be elaborated before "proc (body)"
24261 info: reason: pragma Elaborate in unit "proc (body)"
24267 In this case we have a cycle that the binder cannot break. On the one
24268 hand, there is an explicit pragma Elaborate in @code{proc} for
24269 @code{pack}. This means that the body of @code{pack} must be elaborated
24270 before the body of @code{proc}. On the other hand, there is elaboration
24271 code in @code{pack} that calls a subprogram in @code{proc}. This means
24272 that for maximum safety, there should really be a pragma
24273 Elaborate_All in @code{pack} for @code{proc} which would require that
24274 the body of @code{proc} be elaborated before the body of
24275 @code{pack}. Clearly both requirements cannot be satisfied.
24276 Faced with a circularity of this kind, you have three different options.
24279 @item Fix the program
24280 The most desirable option from the point of view of long-term maintenance
24281 is to rearrange the program so that the elaboration problems are avoided.
24282 One useful technique is to place the elaboration code into separate
24283 child packages. Another is to move some of the initialization code to
24284 explicitly called subprograms, where the program controls the order
24285 of initialization explicitly. Although this is the most desirable option,
24286 it may be impractical and involve too much modification, especially in
24287 the case of complex legacy code.
24289 @item Perform dynamic checks
24290 If the compilations are done using the
24292 (dynamic elaboration check) switch, then GNAT behaves in
24293 a quite different manner. Dynamic checks are generated for all calls
24294 that could possibly result in raising an exception. With this switch,
24295 the compiler does not generate implicit @code{Elaborate_All} pragmas.
24296 The behavior then is exactly as specified in the Ada 95 Reference Manual.
24297 The binder will generate an executable program that may or may not
24298 raise @code{Program_Error}, and then it is the programmer's job to ensure
24299 that it does not raise an exception. Note that it is important to
24300 compile all units with the switch, it cannot be used selectively.
24302 @item Suppress checks
24303 The drawback of dynamic checks is that they generate a
24304 significant overhead at run time, both in space and time. If you
24305 are absolutely sure that your program cannot raise any elaboration
24306 exceptions, and you still want to use the dynamic elaboration model,
24307 then you can use the configuration pragma
24308 @code{Suppress (Elaboration_Check)} to suppress all such checks. For
24309 example this pragma could be placed in the @file{gnat.adc} file.
24311 @item Suppress checks selectively
24312 When you know that certain calls in elaboration code cannot possibly
24313 lead to an elaboration error, and the binder nevertheless generates warnings
24314 on those calls and inserts Elaborate_All pragmas that lead to elaboration
24315 circularities, it is possible to remove those warnings locally and obtain
24316 a program that will bind. Clearly this can be unsafe, and it is the
24317 responsibility of the programmer to make sure that the resulting program has
24318 no elaboration anomalies. The pragma @code{Suppress (Elaboration_Check)} can
24319 be used with different granularity to suppress warnings and break
24320 elaboration circularities:
24324 Place the pragma that names the called subprogram in the declarative part
24325 that contains the call.
24328 Place the pragma in the declarative part, without naming an entity. This
24329 disables warnings on all calls in the corresponding declarative region.
24332 Place the pragma in the package spec that declares the called subprogram,
24333 and name the subprogram. This disables warnings on all elaboration calls to
24337 Place the pragma in the package spec that declares the called subprogram,
24338 without naming any entity. This disables warnings on all elaboration calls to
24339 all subprograms declared in this spec.
24341 @item Use Pragma Elaborate
24342 As previously described in section @xref{Treatment of Pragma Elaborate},
24343 GNAT in static mode assumes that a @code{pragma} Elaborate indicates correctly
24344 that no elaboration checks are required on calls to the designated unit.
24345 There may be cases in which the caller knows that no transitive calls
24346 can occur, so that a @code{pragma Elaborate} will be sufficient in a
24347 case where @code{pragma Elaborate_All} would cause a circularity.
24351 These five cases are listed in order of decreasing safety, and therefore
24352 require increasing programmer care in their application. Consider the
24355 @smallexample @c adanocomment
24357 function F1 return Integer;
24362 function F2 return Integer;
24363 function Pure (x : integer) return integer;
24364 -- pragma Suppress (Elaboration_Check, On => Pure); -- (3)
24365 -- pragma Suppress (Elaboration_Check); -- (4)
24369 package body Pack1 is
24370 function F1 return Integer is
24374 Val : integer := Pack2.Pure (11); -- Elab. call (1)
24377 -- pragma Suppress(Elaboration_Check, Pack2.F2); -- (1)
24378 -- pragma Suppress(Elaboration_Check); -- (2)
24380 X1 := Pack2.F2 + 1; -- Elab. call (2)
24385 package body Pack2 is
24386 function F2 return Integer is
24390 function Pure (x : integer) return integer is
24392 return x ** 3 - 3 * x;
24396 with Pack1, Ada.Text_IO;
24399 Ada.Text_IO.Put_Line(Pack1.X1'Img); -- 101
24402 In the absence of any pragmas, an attempt to bind this program produces
24403 the following diagnostics:
24409 error: elaboration circularity detected
24410 info: "pack1 (body)" must be elaborated before "pack1 (body)"
24411 info: reason: Elaborate_All probably needed in unit "pack1 (body)"
24412 info: recompile "pack1 (body)" with -gnatwl for full details
24413 info: "pack1 (body)"
24414 info: must be elaborated along with its spec:
24415 info: "pack1 (spec)"
24416 info: which is withed by:
24417 info: "pack2 (body)"
24418 info: which must be elaborated along with its spec:
24419 info: "pack2 (spec)"
24420 info: which is withed by:
24421 info: "pack1 (body)"
24424 The sources of the circularity are the two calls to @code{Pack2.Pure} and
24425 @code{Pack2.F2} in the body of @code{Pack1}. We can see that the call to
24426 F2 is safe, even though F2 calls F1, because the call appears after the
24427 elaboration of the body of F1. Therefore the pragma (1) is safe, and will
24428 remove the warning on the call. It is also possible to use pragma (2)
24429 because there are no other potentially unsafe calls in the block.
24432 The call to @code{Pure} is safe because this function does not depend on the
24433 state of @code{Pack2}. Therefore any call to this function is safe, and it
24434 is correct to place pragma (3) in the corresponding package spec.
24437 Finally, we could place pragma (4) in the spec of @code{Pack2} to disable
24438 warnings on all calls to functions declared therein. Note that this is not
24439 necessarily safe, and requires more detailed examination of the subprogram
24440 bodies involved. In particular, a call to @code{F2} requires that @code{F1}
24441 be already elaborated.
24445 It is hard to generalize on which of these four approaches should be
24446 taken. Obviously if it is possible to fix the program so that the default
24447 treatment works, this is preferable, but this may not always be practical.
24448 It is certainly simple enough to use
24450 but the danger in this case is that, even if the GNAT binder
24451 finds a correct elaboration order, it may not always do so,
24452 and certainly a binder from another Ada compiler might not. A
24453 combination of testing and analysis (for which the warnings generated
24456 switch can be useful) must be used to ensure that the program is free
24457 of errors. One switch that is useful in this testing is the
24458 @option{^-p (pessimistic elaboration order)^/PESSIMISTIC_ELABORATION_ORDER^}
24461 Normally the binder tries to find an order that has the best chance of
24462 of avoiding elaboration problems. With this switch, the binder
24463 plays a devil's advocate role, and tries to choose the order that
24464 has the best chance of failing. If your program works even with this
24465 switch, then it has a better chance of being error free, but this is still
24468 For an example of this approach in action, consider the C-tests (executable
24469 tests) from the ACVC suite. If these are compiled and run with the default
24470 treatment, then all but one of them succeed without generating any error
24471 diagnostics from the binder. However, there is one test that fails, and
24472 this is not surprising, because the whole point of this test is to ensure
24473 that the compiler can handle cases where it is impossible to determine
24474 a correct order statically, and it checks that an exception is indeed
24475 raised at run time.
24477 This one test must be compiled and run using the
24479 switch, and then it passes. Alternatively, the entire suite can
24480 be run using this switch. It is never wrong to run with the dynamic
24481 elaboration switch if your code is correct, and we assume that the
24482 C-tests are indeed correct (it is less efficient, but efficiency is
24483 not a factor in running the ACVC tests.)
24485 @node Elaboration for Access-to-Subprogram Values
24486 @section Elaboration for Access-to-Subprogram Values
24487 @cindex Access-to-subprogram
24490 The introduction of access-to-subprogram types in Ada 95 complicates
24491 the handling of elaboration. The trouble is that it becomes
24492 impossible to tell at compile time which procedure
24493 is being called. This means that it is not possible for the binder
24494 to analyze the elaboration requirements in this case.
24496 If at the point at which the access value is created
24497 (i.e., the evaluation of @code{P'Access} for a subprogram @code{P}),
24498 the body of the subprogram is
24499 known to have been elaborated, then the access value is safe, and its use
24500 does not require a check. This may be achieved by appropriate arrangement
24501 of the order of declarations if the subprogram is in the current unit,
24502 or, if the subprogram is in another unit, by using pragma
24503 @code{Pure}, @code{Preelaborate}, or @code{Elaborate_Body}
24504 on the referenced unit.
24506 If the referenced body is not known to have been elaborated at the point
24507 the access value is created, then any use of the access value must do a
24508 dynamic check, and this dynamic check will fail and raise a
24509 @code{Program_Error} exception if the body has not been elaborated yet.
24510 GNAT will generate the necessary checks, and in addition, if the
24512 switch is set, will generate warnings that such checks are required.
24514 The use of dynamic dispatching for tagged types similarly generates
24515 a requirement for dynamic checks, and premature calls to any primitive
24516 operation of a tagged type before the body of the operation has been
24517 elaborated, will result in the raising of @code{Program_Error}.
24519 @node Summary of Procedures for Elaboration Control
24520 @section Summary of Procedures for Elaboration Control
24521 @cindex Elaboration control
24524 First, compile your program with the default options, using none of
24525 the special elaboration control switches. If the binder successfully
24526 binds your program, then you can be confident that, apart from issues
24527 raised by the use of access-to-subprogram types and dynamic dispatching,
24528 the program is free of elaboration errors. If it is important that the
24529 program be portable, then use the
24531 switch to generate warnings about missing @code{Elaborate_All}
24532 pragmas, and supply the missing pragmas.
24534 If the program fails to bind using the default static elaboration
24535 handling, then you can fix the program to eliminate the binder
24536 message, or recompile the entire program with the
24537 @option{-gnatE} switch to generate dynamic elaboration checks,
24538 and, if you are sure there really are no elaboration problems,
24539 use a global pragma @code{Suppress (Elaboration_Check)}.
24541 @node Other Elaboration Order Considerations
24542 @section Other Elaboration Order Considerations
24544 This section has been entirely concerned with the issue of finding a valid
24545 elaboration order, as defined by the Ada Reference Manual. In a case
24546 where several elaboration orders are valid, the task is to find one
24547 of the possible valid elaboration orders (and the static model in GNAT
24548 will ensure that this is achieved).
24550 The purpose of the elaboration rules in the Ada Reference Manual is to
24551 make sure that no entity is accessed before it has been elaborated. For
24552 a subprogram, this means that the spec and body must have been elaborated
24553 before the subprogram is called. For an object, this means that the object
24554 must have been elaborated before its value is read or written. A violation
24555 of either of these two requirements is an access before elaboration order,
24556 and this section has been all about avoiding such errors.
24558 In the case where more than one order of elaboration is possible, in the
24559 sense that access before elaboration errors are avoided, then any one of
24560 the orders is ``correct'' in the sense that it meets the requirements of
24561 the Ada Reference Manual, and no such error occurs.
24563 However, it may be the case for a given program, that there are
24564 constraints on the order of elaboration that come not from consideration
24565 of avoiding elaboration errors, but rather from extra-lingual logic
24566 requirements. Consider this example:
24568 @smallexample @c ada
24569 with Init_Constants;
24570 package Constants is
24575 package Init_Constants is
24576 procedure P; -- require a body
24577 end Init_Constants;
24580 package body Init_Constants is
24581 procedure P is begin null; end;
24585 end Init_Constants;
24589 Z : Integer := Constants.X + Constants.Y;
24593 with Text_IO; use Text_IO;
24596 Put_Line (Calc.Z'Img);
24601 In this example, there is more than one valid order of elaboration. For
24602 example both the following are correct orders:
24605 Init_Constants spec
24608 Init_Constants body
24613 Init_Constants spec
24614 Init_Constants body
24621 There is no language rule to prefer one or the other, both are correct
24622 from an order of elaboration point of view. But the programmatic effects
24623 of the two orders are very different. In the first, the elaboration routine
24624 of @code{Calc} initializes @code{Z} to zero, and then the main program
24625 runs with this value of zero. But in the second order, the elaboration
24626 routine of @code{Calc} runs after the body of Init_Constants has set
24627 @code{X} and @code{Y} and thus @code{Z} is set to 7 before @code{Main}
24630 One could perhaps by applying pretty clever non-artificial intelligence
24631 to the situation guess that it is more likely that the second order of
24632 elaboration is the one desired, but there is no formal linguistic reason
24633 to prefer one over the other. In fact in this particular case, GNAT will
24634 prefer the second order, because of the rule that bodies are elaborated
24635 as soon as possible, but it's just luck that this is what was wanted
24636 (if indeed the second order was preferred).
24638 If the program cares about the order of elaboration routines in a case like
24639 this, it is important to specify the order required. In this particular
24640 case, that could have been achieved by adding to the spec of Calc:
24642 @smallexample @c ada
24643 pragma Elaborate_All (Constants);
24647 which requires that the body (if any) and spec of @code{Constants},
24648 as well as the body and spec of any unit @code{with}'ed by
24649 @code{Constants} be elaborated before @code{Calc} is elaborated.
24651 Clearly no automatic method can always guess which alternative you require,
24652 and if you are working with legacy code that had constraints of this kind
24653 which were not properly specified by adding @code{Elaborate} or
24654 @code{Elaborate_All} pragmas, then indeed it is possible that two different
24655 compilers can choose different orders.
24657 The @code{gnatbind}
24658 @option{^-p^/PESSIMISTIC_ELABORATION^} switch may be useful in smoking
24659 out problems. This switch causes bodies to be elaborated as late as possible
24660 instead of as early as possible. In the example above, it would have forced
24661 the choice of the first elaboration order. If you get different results
24662 when using this switch, and particularly if one set of results is right,
24663 and one is wrong as far as you are concerned, it shows that you have some
24664 missing @code{Elaborate} pragmas. For the example above, we have the
24668 gnatmake -f -q main
24671 gnatmake -f -q main -bargs -p
24677 It is of course quite unlikely that both these results are correct, so
24678 it is up to you in a case like this to investigate the source of the
24679 difference, by looking at the two elaboration orders that are chosen,
24680 and figuring out which is correct, and then adding the necessary
24681 @code{Elaborate_All} pragmas to ensure the desired order.
24683 @node Inline Assembler
24684 @appendix Inline Assembler
24687 If you need to write low-level software that interacts directly
24688 with the hardware, Ada provides two ways to incorporate assembly
24689 language code into your program. First, you can import and invoke
24690 external routines written in assembly language, an Ada feature fully
24691 supported by GNAT. However, for small sections of code it may be simpler
24692 or more efficient to include assembly language statements directly
24693 in your Ada source program, using the facilities of the implementation-defined
24694 package @code{System.Machine_Code}, which incorporates the gcc
24695 Inline Assembler. The Inline Assembler approach offers a number of advantages,
24696 including the following:
24699 @item No need to use non-Ada tools
24700 @item Consistent interface over different targets
24701 @item Automatic usage of the proper calling conventions
24702 @item Access to Ada constants and variables
24703 @item Definition of intrinsic routines
24704 @item Possibility of inlining a subprogram comprising assembler code
24705 @item Code optimizer can take Inline Assembler code into account
24708 This chapter presents a series of examples to show you how to use
24709 the Inline Assembler. Although it focuses on the Intel x86,
24710 the general approach applies also to other processors.
24711 It is assumed that you are familiar with Ada
24712 and with assembly language programming.
24715 * Basic Assembler Syntax::
24716 * A Simple Example of Inline Assembler::
24717 * Output Variables in Inline Assembler::
24718 * Input Variables in Inline Assembler::
24719 * Inlining Inline Assembler Code::
24720 * Other Asm Functionality::
24723 @c ---------------------------------------------------------------------------
24724 @node Basic Assembler Syntax
24725 @section Basic Assembler Syntax
24728 The assembler used by GNAT and gcc is based not on the Intel assembly
24729 language, but rather on a language that descends from the AT&T Unix
24730 assembler @emph{as} (and which is often referred to as ``AT&T syntax'').
24731 The following table summarizes the main features of @emph{as} syntax
24732 and points out the differences from the Intel conventions.
24733 See the gcc @emph{as} and @emph{gas} (an @emph{as} macro
24734 pre-processor) documentation for further information.
24737 @item Register names
24738 gcc / @emph{as}: Prefix with ``%''; for example @code{%eax}
24740 Intel: No extra punctuation; for example @code{eax}
24742 @item Immediate operand
24743 gcc / @emph{as}: Prefix with ``$''; for example @code{$4}
24745 Intel: No extra punctuation; for example @code{4}
24748 gcc / @emph{as}: Prefix with ``$''; for example @code{$loc}
24750 Intel: No extra punctuation; for example @code{loc}
24752 @item Memory contents
24753 gcc / @emph{as}: No extra punctuation; for example @code{loc}
24755 Intel: Square brackets; for example @code{[loc]}
24757 @item Register contents
24758 gcc / @emph{as}: Parentheses; for example @code{(%eax)}
24760 Intel: Square brackets; for example @code{[eax]}
24762 @item Hexadecimal numbers
24763 gcc / @emph{as}: Leading ``0x'' (C language syntax); for example @code{0xA0}
24765 Intel: Trailing ``h''; for example @code{A0h}
24768 gcc / @emph{as}: Explicit in op code; for example @code{movw} to move
24771 Intel: Implicit, deduced by assembler; for example @code{mov}
24773 @item Instruction repetition
24774 gcc / @emph{as}: Split into two lines; for example
24780 Intel: Keep on one line; for example @code{rep stosl}
24782 @item Order of operands
24783 gcc / @emph{as}: Source first; for example @code{movw $4, %eax}
24785 Intel: Destination first; for example @code{mov eax, 4}
24788 @c ---------------------------------------------------------------------------
24789 @node A Simple Example of Inline Assembler
24790 @section A Simple Example of Inline Assembler
24793 The following example will generate a single assembly language statement,
24794 @code{nop}, which does nothing. Despite its lack of run-time effect,
24795 the example will be useful in illustrating the basics of
24796 the Inline Assembler facility.
24798 @smallexample @c ada
24800 with System.Machine_Code; use System.Machine_Code;
24801 procedure Nothing is
24808 @code{Asm} is a procedure declared in package @code{System.Machine_Code};
24809 here it takes one parameter, a @emph{template string} that must be a static
24810 expression and that will form the generated instruction.
24811 @code{Asm} may be regarded as a compile-time procedure that parses
24812 the template string and additional parameters (none here),
24813 from which it generates a sequence of assembly language instructions.
24815 The examples in this chapter will illustrate several of the forms
24816 for invoking @code{Asm}; a complete specification of the syntax
24817 is found in the @cite{GNAT Reference Manual}.
24819 Under the standard GNAT conventions, the @code{Nothing} procedure
24820 should be in a file named @file{nothing.adb}.
24821 You can build the executable in the usual way:
24825 However, the interesting aspect of this example is not its run-time behavior
24826 but rather the generated assembly code.
24827 To see this output, invoke the compiler as follows:
24829 gcc -c -S -fomit-frame-pointer -gnatp @file{nothing.adb}
24831 where the options are:
24835 compile only (no bind or link)
24837 generate assembler listing
24838 @item -fomit-frame-pointer
24839 do not set up separate stack frames
24841 do not add runtime checks
24844 This gives a human-readable assembler version of the code. The resulting
24845 file will have the same name as the Ada source file, but with a @code{.s}
24846 extension. In our example, the file @file{nothing.s} has the following
24851 .file "nothing.adb"
24853 ___gnu_compiled_ada:
24856 .globl __ada_nothing
24868 The assembly code you included is clearly indicated by
24869 the compiler, between the @code{#APP} and @code{#NO_APP}
24870 delimiters. The character before the 'APP' and 'NOAPP'
24871 can differ on different targets. For example, GNU/Linux uses '#APP' while
24872 on NT you will see '/APP'.
24874 If you make a mistake in your assembler code (such as using the
24875 wrong size modifier, or using a wrong operand for the instruction) GNAT
24876 will report this error in a temporary file, which will be deleted when
24877 the compilation is finished. Generating an assembler file will help
24878 in such cases, since you can assemble this file separately using the
24879 @emph{as} assembler that comes with gcc.
24881 Assembling the file using the command
24884 as @file{nothing.s}
24887 will give you error messages whose lines correspond to the assembler
24888 input file, so you can easily find and correct any mistakes you made.
24889 If there are no errors, @emph{as} will generate an object file
24890 @file{nothing.out}.
24892 @c ---------------------------------------------------------------------------
24893 @node Output Variables in Inline Assembler
24894 @section Output Variables in Inline Assembler
24897 The examples in this section, showing how to access the processor flags,
24898 illustrate how to specify the destination operands for assembly language
24901 @smallexample @c ada
24903 with Interfaces; use Interfaces;
24904 with Ada.Text_IO; use Ada.Text_IO;
24905 with System.Machine_Code; use System.Machine_Code;
24906 procedure Get_Flags is
24907 Flags : Unsigned_32;
24910 Asm ("pushfl" & LF & HT & -- push flags on stack
24911 "popl %%eax" & LF & HT & -- load eax with flags
24912 "movl %%eax, %0", -- store flags in variable
24913 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
24914 Put_Line ("Flags register:" & Flags'Img);
24919 In order to have a nicely aligned assembly listing, we have separated
24920 multiple assembler statements in the Asm template string with linefeed
24921 (ASCII.LF) and horizontal tab (ASCII.HT) characters.
24922 The resulting section of the assembly output file is:
24929 movl %eax, -40(%ebp)
24934 It would have been legal to write the Asm invocation as:
24937 Asm ("pushfl popl %%eax movl %%eax, %0")
24940 but in the generated assembler file, this would come out as:
24944 pushfl popl %eax movl %eax, -40(%ebp)
24948 which is not so convenient for the human reader.
24950 We use Ada comments
24951 at the end of each line to explain what the assembler instructions
24952 actually do. This is a useful convention.
24954 When writing Inline Assembler instructions, you need to precede each register
24955 and variable name with a percent sign. Since the assembler already requires
24956 a percent sign at the beginning of a register name, you need two consecutive
24957 percent signs for such names in the Asm template string, thus @code{%%eax}.
24958 In the generated assembly code, one of the percent signs will be stripped off.
24960 Names such as @code{%0}, @code{%1}, @code{%2}, etc., denote input or output
24961 variables: operands you later define using @code{Input} or @code{Output}
24962 parameters to @code{Asm}.
24963 An output variable is illustrated in
24964 the third statement in the Asm template string:
24968 The intent is to store the contents of the eax register in a variable that can
24969 be accessed in Ada. Simply writing @code{movl %%eax, Flags} would not
24970 necessarily work, since the compiler might optimize by using a register
24971 to hold Flags, and the expansion of the @code{movl} instruction would not be
24972 aware of this optimization. The solution is not to store the result directly
24973 but rather to advise the compiler to choose the correct operand form;
24974 that is the purpose of the @code{%0} output variable.
24976 Information about the output variable is supplied in the @code{Outputs}
24977 parameter to @code{Asm}:
24979 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
24982 The output is defined by the @code{Asm_Output} attribute of the target type;
24983 the general format is
24985 Type'Asm_Output (constraint_string, variable_name)
24988 The constraint string directs the compiler how
24989 to store/access the associated variable. In the example
24991 Unsigned_32'Asm_Output ("=m", Flags);
24993 the @code{"m"} (memory) constraint tells the compiler that the variable
24994 @code{Flags} should be stored in a memory variable, thus preventing
24995 the optimizer from keeping it in a register. In contrast,
24997 Unsigned_32'Asm_Output ("=r", Flags);
24999 uses the @code{"r"} (register) constraint, telling the compiler to
25000 store the variable in a register.
25002 If the constraint is preceded by the equal character (@strong{=}), it tells
25003 the compiler that the variable will be used to store data into it.
25005 In the @code{Get_Flags} example, we used the @code{"g"} (global) constraint,
25006 allowing the optimizer to choose whatever it deems best.
25008 There are a fairly large number of constraints, but the ones that are
25009 most useful (for the Intel x86 processor) are the following:
25015 global (i.e. can be stored anywhere)
25033 use one of eax, ebx, ecx or edx
25035 use one of eax, ebx, ecx, edx, esi or edi
25038 The full set of constraints is described in the gcc and @emph{as}
25039 documentation; note that it is possible to combine certain constraints
25040 in one constraint string.
25042 You specify the association of an output variable with an assembler operand
25043 through the @code{%}@emph{n} notation, where @emph{n} is a non-negative
25045 @smallexample @c ada
25047 Asm ("pushfl" & LF & HT & -- push flags on stack
25048 "popl %%eax" & LF & HT & -- load eax with flags
25049 "movl %%eax, %0", -- store flags in variable
25050 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
25054 @code{%0} will be replaced in the expanded code by the appropriate operand,
25056 the compiler decided for the @code{Flags} variable.
25058 In general, you may have any number of output variables:
25061 Count the operands starting at 0; thus @code{%0}, @code{%1}, etc.
25063 Specify the @code{Outputs} parameter as a parenthesized comma-separated list
25064 of @code{Asm_Output} attributes
25068 @smallexample @c ada
25070 Asm ("movl %%eax, %0" & LF & HT &
25071 "movl %%ebx, %1" & LF & HT &
25073 Outputs => (Unsigned_32'Asm_Output ("=g", Var_A), -- %0 = Var_A
25074 Unsigned_32'Asm_Output ("=g", Var_B), -- %1 = Var_B
25075 Unsigned_32'Asm_Output ("=g", Var_C))); -- %2 = Var_C
25079 where @code{Var_A}, @code{Var_B}, and @code{Var_C} are variables
25080 in the Ada program.
25082 As a variation on the @code{Get_Flags} example, we can use the constraints
25083 string to direct the compiler to store the eax register into the @code{Flags}
25084 variable, instead of including the store instruction explicitly in the
25085 @code{Asm} template string:
25087 @smallexample @c ada
25089 with Interfaces; use Interfaces;
25090 with Ada.Text_IO; use Ada.Text_IO;
25091 with System.Machine_Code; use System.Machine_Code;
25092 procedure Get_Flags_2 is
25093 Flags : Unsigned_32;
25096 Asm ("pushfl" & LF & HT & -- push flags on stack
25097 "popl %%eax", -- save flags in eax
25098 Outputs => Unsigned_32'Asm_Output ("=a", Flags));
25099 Put_Line ("Flags register:" & Flags'Img);
25105 The @code{"a"} constraint tells the compiler that the @code{Flags}
25106 variable will come from the eax register. Here is the resulting code:
25114 movl %eax,-40(%ebp)
25119 The compiler generated the store of eax into Flags after
25120 expanding the assembler code.
25122 Actually, there was no need to pop the flags into the eax register;
25123 more simply, we could just pop the flags directly into the program variable:
25125 @smallexample @c ada
25127 with Interfaces; use Interfaces;
25128 with Ada.Text_IO; use Ada.Text_IO;
25129 with System.Machine_Code; use System.Machine_Code;
25130 procedure Get_Flags_3 is
25131 Flags : Unsigned_32;
25134 Asm ("pushfl" & LF & HT & -- push flags on stack
25135 "pop %0", -- save flags in Flags
25136 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
25137 Put_Line ("Flags register:" & Flags'Img);
25142 @c ---------------------------------------------------------------------------
25143 @node Input Variables in Inline Assembler
25144 @section Input Variables in Inline Assembler
25147 The example in this section illustrates how to specify the source operands
25148 for assembly language statements.
25149 The program simply increments its input value by 1:
25151 @smallexample @c ada
25153 with Interfaces; use Interfaces;
25154 with Ada.Text_IO; use Ada.Text_IO;
25155 with System.Machine_Code; use System.Machine_Code;
25156 procedure Increment is
25158 function Incr (Value : Unsigned_32) return Unsigned_32 is
25159 Result : Unsigned_32;
25162 Inputs => Unsigned_32'Asm_Input ("a", Value),
25163 Outputs => Unsigned_32'Asm_Output ("=a", Result));
25167 Value : Unsigned_32;
25171 Put_Line ("Value before is" & Value'Img);
25172 Value := Incr (Value);
25173 Put_Line ("Value after is" & Value'Img);
25178 The @code{Outputs} parameter to @code{Asm} specifies
25179 that the result will be in the eax register and that it is to be stored
25180 in the @code{Result} variable.
25182 The @code{Inputs} parameter looks much like the @code{Outputs} parameter,
25183 but with an @code{Asm_Input} attribute.
25184 The @code{"="} constraint, indicating an output value, is not present.
25186 You can have multiple input variables, in the same way that you can have more
25187 than one output variable.
25189 The parameter count (%0, %1) etc, now starts at the first input
25190 statement, and continues with the output statements.
25191 When both parameters use the same variable, the
25192 compiler will treat them as the same %n operand, which is the case here.
25194 Just as the @code{Outputs} parameter causes the register to be stored into the
25195 target variable after execution of the assembler statements, so does the
25196 @code{Inputs} parameter cause its variable to be loaded into the register
25197 before execution of the assembler statements.
25199 Thus the effect of the @code{Asm} invocation is:
25201 @item load the 32-bit value of @code{Value} into eax
25202 @item execute the @code{incl %eax} instruction
25203 @item store the contents of eax into the @code{Result} variable
25206 The resulting assembler file (with @option{-O2} optimization) contains:
25209 _increment__incr.1:
25222 @c ---------------------------------------------------------------------------
25223 @node Inlining Inline Assembler Code
25224 @section Inlining Inline Assembler Code
25227 For a short subprogram such as the @code{Incr} function in the previous
25228 section, the overhead of the call and return (creating / deleting the stack
25229 frame) can be significant, compared to the amount of code in the subprogram
25230 body. A solution is to apply Ada's @code{Inline} pragma to the subprogram,
25231 which directs the compiler to expand invocations of the subprogram at the
25232 point(s) of call, instead of setting up a stack frame for out-of-line calls.
25233 Here is the resulting program:
25235 @smallexample @c ada
25237 with Interfaces; use Interfaces;
25238 with Ada.Text_IO; use Ada.Text_IO;
25239 with System.Machine_Code; use System.Machine_Code;
25240 procedure Increment_2 is
25242 function Incr (Value : Unsigned_32) return Unsigned_32 is
25243 Result : Unsigned_32;
25246 Inputs => Unsigned_32'Asm_Input ("a", Value),
25247 Outputs => Unsigned_32'Asm_Output ("=a", Result));
25250 pragma Inline (Increment);
25252 Value : Unsigned_32;
25256 Put_Line ("Value before is" & Value'Img);
25257 Value := Increment (Value);
25258 Put_Line ("Value after is" & Value'Img);
25263 Compile the program with both optimization (@option{-O2}) and inlining
25264 enabled (@option{-gnatpn} instead of @option{-gnatp}).
25266 The @code{Incr} function is still compiled as usual, but at the
25267 point in @code{Increment} where our function used to be called:
25272 call _increment__incr.1
25277 the code for the function body directly appears:
25290 thus saving the overhead of stack frame setup and an out-of-line call.
25292 @c ---------------------------------------------------------------------------
25293 @node Other Asm Functionality
25294 @section Other @code{Asm} Functionality
25297 This section describes two important parameters to the @code{Asm}
25298 procedure: @code{Clobber}, which identifies register usage;
25299 and @code{Volatile}, which inhibits unwanted optimizations.
25302 * The Clobber Parameter::
25303 * The Volatile Parameter::
25306 @c ---------------------------------------------------------------------------
25307 @node The Clobber Parameter
25308 @subsection The @code{Clobber} Parameter
25311 One of the dangers of intermixing assembly language and a compiled language
25312 such as Ada is that the compiler needs to be aware of which registers are
25313 being used by the assembly code. In some cases, such as the earlier examples,
25314 the constraint string is sufficient to indicate register usage (e.g.,
25316 the eax register). But more generally, the compiler needs an explicit
25317 identification of the registers that are used by the Inline Assembly
25320 Using a register that the compiler doesn't know about
25321 could be a side effect of an instruction (like @code{mull}
25322 storing its result in both eax and edx).
25323 It can also arise from explicit register usage in your
25324 assembly code; for example:
25327 Asm ("movl %0, %%ebx" & LF & HT &
25329 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
25330 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out));
25334 where the compiler (since it does not analyze the @code{Asm} template string)
25335 does not know you are using the ebx register.
25337 In such cases you need to supply the @code{Clobber} parameter to @code{Asm},
25338 to identify the registers that will be used by your assembly code:
25342 Asm ("movl %0, %%ebx" & LF & HT &
25344 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
25345 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
25350 The Clobber parameter is a static string expression specifying the
25351 register(s) you are using. Note that register names are @emph{not} prefixed
25352 by a percent sign. Also, if more than one register is used then their names
25353 are separated by commas; e.g., @code{"eax, ebx"}
25355 The @code{Clobber} parameter has several additional uses:
25357 @item Use ``register'' name @code{cc} to indicate that flags might have changed
25358 @item Use ``register'' name @code{memory} if you changed a memory location
25361 @c ---------------------------------------------------------------------------
25362 @node The Volatile Parameter
25363 @subsection The @code{Volatile} Parameter
25364 @cindex Volatile parameter
25367 Compiler optimizations in the presence of Inline Assembler may sometimes have
25368 unwanted effects. For example, when an @code{Asm} invocation with an input
25369 variable is inside a loop, the compiler might move the loading of the input
25370 variable outside the loop, regarding it as a one-time initialization.
25372 If this effect is not desired, you can disable such optimizations by setting
25373 the @code{Volatile} parameter to @code{True}; for example:
25375 @smallexample @c ada
25377 Asm ("movl %0, %%ebx" & LF & HT &
25379 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
25380 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
25386 By default, @code{Volatile} is set to @code{False} unless there is no
25387 @code{Outputs} parameter.
25389 Although setting @code{Volatile} to @code{True} prevents unwanted
25390 optimizations, it will also disable other optimizations that might be
25391 important for efficiency. In general, you should set @code{Volatile}
25392 to @code{True} only if the compiler's optimizations have created
25394 @c END OF INLINE ASSEMBLER CHAPTER
25395 @c ===============================
25397 @c ***********************************
25398 @c * Compatibility and Porting Guide *
25399 @c ***********************************
25400 @node Compatibility and Porting Guide
25401 @appendix Compatibility and Porting Guide
25404 This chapter describes the compatibility issues that may arise between
25405 GNAT and other Ada 83 and Ada 95 compilation systems, and shows how GNAT
25406 can expedite porting
25407 applications developed in other Ada environments.
25410 * Compatibility with Ada 83::
25411 * Implementation-dependent characteristics::
25412 * Compatibility with Other Ada 95 Systems::
25413 * Representation Clauses::
25414 * Compatibility with DEC Ada 83::
25416 * Transitioning from Alpha to Integrity OpenVMS::
25420 @node Compatibility with Ada 83
25421 @section Compatibility with Ada 83
25422 @cindex Compatibility (between Ada 83 and Ada 95)
25425 Ada 95 is designed to be highly upwards compatible with Ada 83. In
25426 particular, the design intention is that the difficulties associated
25427 with moving from Ada 83 to Ada 95 should be no greater than those
25428 that occur when moving from one Ada 83 system to another.
25430 However, there are a number of points at which there are minor
25431 incompatibilities. The @cite{Ada 95 Annotated Reference Manual} contains
25432 full details of these issues,
25433 and should be consulted for a complete treatment.
25435 following subsections treat the most likely issues to be encountered.
25438 * Legal Ada 83 programs that are illegal in Ada 95::
25439 * More deterministic semantics::
25440 * Changed semantics::
25441 * Other language compatibility issues::
25444 @node Legal Ada 83 programs that are illegal in Ada 95
25445 @subsection Legal Ada 83 programs that are illegal in Ada 95
25448 @item Character literals
25449 Some uses of character literals are ambiguous. Since Ada 95 has introduced
25450 @code{Wide_Character} as a new predefined character type, some uses of
25451 character literals that were legal in Ada 83 are illegal in Ada 95.
25453 @smallexample @c ada
25454 for Char in 'A' .. 'Z' loop ... end loop;
25457 The problem is that @code{'A'} and @code{'Z'} could be from either
25458 @code{Character} or @code{Wide_Character}. The simplest correction
25459 is to make the type explicit; e.g.:
25460 @smallexample @c ada
25461 for Char in Character range 'A' .. 'Z' loop ... end loop;
25464 @item New reserved words
25465 The identifiers @code{abstract}, @code{aliased}, @code{protected},
25466 @code{requeue}, @code{tagged}, and @code{until} are reserved in Ada 95.
25467 Existing Ada 83 code using any of these identifiers must be edited to
25468 use some alternative name.
25470 @item Freezing rules
25471 The rules in Ada 95 are slightly different with regard to the point at
25472 which entities are frozen, and representation pragmas and clauses are
25473 not permitted past the freeze point. This shows up most typically in
25474 the form of an error message complaining that a representation item
25475 appears too late, and the appropriate corrective action is to move
25476 the item nearer to the declaration of the entity to which it refers.
25478 A particular case is that representation pragmas
25481 extended DEC Ada 83 compatibility pragmas such as @code{Export_Procedure})
25483 cannot be applied to a subprogram body. If necessary, a separate subprogram
25484 declaration must be introduced to which the pragma can be applied.
25486 @item Optional bodies for library packages
25487 In Ada 83, a package that did not require a package body was nevertheless
25488 allowed to have one. This lead to certain surprises in compiling large
25489 systems (situations in which the body could be unexpectedly ignored by the
25490 binder). In Ada 95, if a package does not require a body then it is not
25491 permitted to have a body. To fix this problem, simply remove a redundant
25492 body if it is empty, or, if it is non-empty, introduce a dummy declaration
25493 into the spec that makes the body required. One approach is to add a private
25494 part to the package declaration (if necessary), and define a parameterless
25495 procedure called @code{Requires_Body}, which must then be given a dummy
25496 procedure body in the package body, which then becomes required.
25497 Another approach (assuming that this does not introduce elaboration
25498 circularities) is to add an @code{Elaborate_Body} pragma to the package spec,
25499 since one effect of this pragma is to require the presence of a package body.
25501 @item @code{Numeric_Error} is now the same as @code{Constraint_Error}
25502 In Ada 95, the exception @code{Numeric_Error} is a renaming of
25503 @code{Constraint_Error}.
25504 This means that it is illegal to have separate exception handlers for
25505 the two exceptions. The fix is simply to remove the handler for the
25506 @code{Numeric_Error} case (since even in Ada 83, a compiler was free to raise
25507 @code{Constraint_Error} in place of @code{Numeric_Error} in all cases).
25509 @item Indefinite subtypes in generics
25510 In Ada 83, it was permissible to pass an indefinite type (e.g.@: @code{String})
25511 as the actual for a generic formal private type, but then the instantiation
25512 would be illegal if there were any instances of declarations of variables
25513 of this type in the generic body. In Ada 95, to avoid this clear violation
25514 of the methodological principle known as the ``contract model'',
25515 the generic declaration explicitly indicates whether
25516 or not such instantiations are permitted. If a generic formal parameter
25517 has explicit unknown discriminants, indicated by using @code{(<>)} after the
25518 type name, then it can be instantiated with indefinite types, but no
25519 stand-alone variables can be declared of this type. Any attempt to declare
25520 such a variable will result in an illegality at the time the generic is
25521 declared. If the @code{(<>)} notation is not used, then it is illegal
25522 to instantiate the generic with an indefinite type.
25523 This is the potential incompatibility issue when porting Ada 83 code to Ada 95.
25524 It will show up as a compile time error, and
25525 the fix is usually simply to add the @code{(<>)} to the generic declaration.
25528 @node More deterministic semantics
25529 @subsection More deterministic semantics
25533 Conversions from real types to integer types round away from 0. In Ada 83
25534 the conversion Integer(2.5) could deliver either 2 or 3 as its value. This
25535 implementation freedom was intended to support unbiased rounding in
25536 statistical applications, but in practice it interfered with portability.
25537 In Ada 95 the conversion semantics are unambiguous, and rounding away from 0
25538 is required. Numeric code may be affected by this change in semantics.
25539 Note, though, that this issue is no worse than already existed in Ada 83
25540 when porting code from one vendor to another.
25543 The Real-Time Annex introduces a set of policies that define the behavior of
25544 features that were implementation dependent in Ada 83, such as the order in
25545 which open select branches are executed.
25548 @node Changed semantics
25549 @subsection Changed semantics
25552 The worst kind of incompatibility is one where a program that is legal in
25553 Ada 83 is also legal in Ada 95 but can have an effect in Ada 95 that was not
25554 possible in Ada 83. Fortunately this is extremely rare, but the one
25555 situation that you should be alert to is the change in the predefined type
25556 @code{Character} from 7-bit ASCII to 8-bit Latin-1.
25559 @item range of @code{Character}
25560 The range of @code{Standard.Character} is now the full 256 characters
25561 of Latin-1, whereas in most Ada 83 implementations it was restricted
25562 to 128 characters. Although some of the effects of
25563 this change will be manifest in compile-time rejection of legal
25564 Ada 83 programs it is possible for a working Ada 83 program to have
25565 a different effect in Ada 95, one that was not permitted in Ada 83.
25566 As an example, the expression
25567 @code{Character'Pos(Character'Last)} returned @code{127} in Ada 83 and now
25568 delivers @code{255} as its value.
25569 In general, you should look at the logic of any
25570 character-processing Ada 83 program and see whether it needs to be adapted
25571 to work correctly with Latin-1. Note that the predefined Ada 95 API has a
25572 character handling package that may be relevant if code needs to be adapted
25573 to account for the additional Latin-1 elements.
25574 The desirable fix is to
25575 modify the program to accommodate the full character set, but in some cases
25576 it may be convenient to define a subtype or derived type of Character that
25577 covers only the restricted range.
25581 @node Other language compatibility issues
25582 @subsection Other language compatibility issues
25584 @item @option{-gnat83 switch}
25585 All implementations of GNAT provide a switch that causes GNAT to operate
25586 in Ada 83 mode. In this mode, some but not all compatibility problems
25587 of the type described above are handled automatically. For example, the
25588 new Ada 95 reserved words are treated simply as identifiers as in Ada 83.
25590 in practice, it is usually advisable to make the necessary modifications
25591 to the program to remove the need for using this switch.
25592 See @ref{Compiling Different Versions of Ada}.
25594 @item Support for removed Ada 83 pragmas and attributes
25595 A number of pragmas and attributes from Ada 83 have been removed from Ada 95,
25596 generally because they have been replaced by other mechanisms. Ada 95
25597 compilers are allowed, but not required, to implement these missing
25598 elements. In contrast with some other Ada 95 compilers, GNAT implements all
25599 such pragmas and attributes, eliminating this compatibility concern. These
25600 include @code{pragma Interface} and the floating point type attributes
25601 (@code{Emax}, @code{Mantissa}, etc.), among other items.
25604 @node Implementation-dependent characteristics
25605 @section Implementation-dependent characteristics
25607 Although the Ada language defines the semantics of each construct as
25608 precisely as practical, in some situations (for example for reasons of
25609 efficiency, or where the effect is heavily dependent on the host or target
25610 platform) the implementation is allowed some freedom. In porting Ada 83
25611 code to GNAT, you need to be aware of whether / how the existing code
25612 exercised such implementation dependencies. Such characteristics fall into
25613 several categories, and GNAT offers specific support in assisting the
25614 transition from certain Ada 83 compilers.
25617 * Implementation-defined pragmas::
25618 * Implementation-defined attributes::
25620 * Elaboration order::
25621 * Target-specific aspects::
25624 @node Implementation-defined pragmas
25625 @subsection Implementation-defined pragmas
25628 Ada compilers are allowed to supplement the language-defined pragmas, and
25629 these are a potential source of non-portability. All GNAT-defined pragmas
25630 are described in the GNAT Reference Manual, and these include several that
25631 are specifically intended to correspond to other vendors' Ada 83 pragmas.
25632 For migrating from VADS, the pragma @code{Use_VADS_Size} may be useful.
25634 compatibility with DEC Ada 83, GNAT supplies the pragmas
25635 @code{Extend_System}, @code{Ident}, @code{Inline_Generic},
25636 @code{Interface_Name}, @code{Passive}, @code{Suppress_All},
25637 and @code{Volatile}.
25638 Other relevant pragmas include @code{External} and @code{Link_With}.
25639 Some vendor-specific
25640 Ada 83 pragmas (@code{Share_Generic}, @code{Subtitle}, and @code{Title}) are
25642 avoiding compiler rejection of units that contain such pragmas; they are not
25643 relevant in a GNAT context and hence are not otherwise implemented.
25645 @node Implementation-defined attributes
25646 @subsection Implementation-defined attributes
25648 Analogous to pragmas, the set of attributes may be extended by an
25649 implementation. All GNAT-defined attributes are described in the
25650 @cite{GNAT Reference Manual}, and these include several that are specifically
25652 to correspond to other vendors' Ada 83 attributes. For migrating from VADS,
25653 the attribute @code{VADS_Size} may be useful. For compatibility with DEC
25654 Ada 83, GNAT supplies the attributes @code{Bit}, @code{Machine_Size} and
25658 @subsection Libraries
25660 Vendors may supply libraries to supplement the standard Ada API. If Ada 83
25661 code uses vendor-specific libraries then there are several ways to manage
25665 If the source code for the libraries (specifications and bodies) are
25666 available, then the libraries can be migrated in the same way as the
25669 If the source code for the specifications but not the bodies are
25670 available, then you can reimplement the bodies.
25672 Some new Ada 95 features obviate the need for library support. For
25673 example most Ada 83 vendors supplied a package for unsigned integers. The
25674 Ada 95 modular type feature is the preferred way to handle this need, so
25675 instead of migrating or reimplementing the unsigned integer package it may
25676 be preferable to retrofit the application using modular types.
25679 @node Elaboration order
25680 @subsection Elaboration order
25682 The implementation can choose any elaboration order consistent with the unit
25683 dependency relationship. This freedom means that some orders can result in
25684 Program_Error being raised due to an ``Access Before Elaboration'': an attempt
25685 to invoke a subprogram its body has been elaborated, or to instantiate a
25686 generic before the generic body has been elaborated. By default GNAT
25687 attempts to choose a safe order (one that will not encounter access before
25688 elaboration problems) by implicitly inserting Elaborate_All pragmas where
25689 needed. However, this can lead to the creation of elaboration circularities
25690 and a resulting rejection of the program by gnatbind. This issue is
25691 thoroughly described in @ref{Elaboration Order Handling in GNAT}.
25692 In brief, there are several
25693 ways to deal with this situation:
25697 Modify the program to eliminate the circularities, e.g. by moving
25698 elaboration-time code into explicitly-invoked procedures
25700 Constrain the elaboration order by including explicit @code{Elaborate_Body} or
25701 @code{Elaborate} pragmas, and then inhibit the generation of implicit
25702 @code{Elaborate_All}
25703 pragmas either globally (as an effect of the @option{-gnatE} switch) or locally
25704 (by selectively suppressing elaboration checks via pragma
25705 @code{Suppress(Elaboration_Check)} when it is safe to do so).
25708 @node Target-specific aspects
25709 @subsection Target-specific aspects
25711 Low-level applications need to deal with machine addresses, data
25712 representations, interfacing with assembler code, and similar issues. If
25713 such an Ada 83 application is being ported to different target hardware (for
25714 example where the byte endianness has changed) then you will need to
25715 carefully examine the program logic; the porting effort will heavily depend
25716 on the robustness of the original design. Moreover, Ada 95 is sometimes
25717 incompatible with typical Ada 83 compiler practices regarding implicit
25718 packing, the meaning of the Size attribute, and the size of access values.
25719 GNAT's approach to these issues is described in @ref{Representation Clauses}.
25721 @node Compatibility with Other Ada 95 Systems
25722 @section Compatibility with Other Ada 95 Systems
25725 Providing that programs avoid the use of implementation dependent and
25726 implementation defined features of Ada 95, as documented in the Ada 95
25727 reference manual, there should be a high degree of portability between
25728 GNAT and other Ada 95 systems. The following are specific items which
25729 have proved troublesome in moving GNAT programs to other Ada 95
25730 compilers, but do not affect porting code to GNAT@.
25733 @item Ada 83 Pragmas and Attributes
25734 Ada 95 compilers are allowed, but not required, to implement the missing
25735 Ada 83 pragmas and attributes that are no longer defined in Ada 95.
25736 GNAT implements all such pragmas and attributes, eliminating this as
25737 a compatibility concern, but some other Ada 95 compilers reject these
25738 pragmas and attributes.
25740 @item Special-needs Annexes
25741 GNAT implements the full set of special needs annexes. At the
25742 current time, it is the only Ada 95 compiler to do so. This means that
25743 programs making use of these features may not be portable to other Ada
25744 95 compilation systems.
25746 @item Representation Clauses
25747 Some other Ada 95 compilers implement only the minimal set of
25748 representation clauses required by the Ada 95 reference manual. GNAT goes
25749 far beyond this minimal set, as described in the next section.
25752 @node Representation Clauses
25753 @section Representation Clauses
25756 The Ada 83 reference manual was quite vague in describing both the minimal
25757 required implementation of representation clauses, and also their precise
25758 effects. The Ada 95 reference manual is much more explicit, but the minimal
25759 set of capabilities required in Ada 95 is quite limited.
25761 GNAT implements the full required set of capabilities described in the
25762 Ada 95 reference manual, but also goes much beyond this, and in particular
25763 an effort has been made to be compatible with existing Ada 83 usage to the
25764 greatest extent possible.
25766 A few cases exist in which Ada 83 compiler behavior is incompatible with
25767 requirements in the Ada 95 reference manual. These are instances of
25768 intentional or accidental dependence on specific implementation dependent
25769 characteristics of these Ada 83 compilers. The following is a list of
25770 the cases most likely to arise in existing legacy Ada 83 code.
25773 @item Implicit Packing
25774 Some Ada 83 compilers allowed a Size specification to cause implicit
25775 packing of an array or record. This could cause expensive implicit
25776 conversions for change of representation in the presence of derived
25777 types, and the Ada design intends to avoid this possibility.
25778 Subsequent AI's were issued to make it clear that such implicit
25779 change of representation in response to a Size clause is inadvisable,
25780 and this recommendation is represented explicitly in the Ada 95 RM
25781 as implementation advice that is followed by GNAT@.
25782 The problem will show up as an error
25783 message rejecting the size clause. The fix is simply to provide
25784 the explicit pragma @code{Pack}, or for more fine tuned control, provide
25785 a Component_Size clause.
25787 @item Meaning of Size Attribute
25788 The Size attribute in Ada 95 for discrete types is defined as being the
25789 minimal number of bits required to hold values of the type. For example,
25790 on a 32-bit machine, the size of Natural will typically be 31 and not
25791 32 (since no sign bit is required). Some Ada 83 compilers gave 31, and
25792 some 32 in this situation. This problem will usually show up as a compile
25793 time error, but not always. It is a good idea to check all uses of the
25794 'Size attribute when porting Ada 83 code. The GNAT specific attribute
25795 Object_Size can provide a useful way of duplicating the behavior of
25796 some Ada 83 compiler systems.
25798 @item Size of Access Types
25799 A common assumption in Ada 83 code is that an access type is in fact a pointer,
25800 and that therefore it will be the same size as a System.Address value. This
25801 assumption is true for GNAT in most cases with one exception. For the case of
25802 a pointer to an unconstrained array type (where the bounds may vary from one
25803 value of the access type to another), the default is to use a ``fat pointer'',
25804 which is represented as two separate pointers, one to the bounds, and one to
25805 the array. This representation has a number of advantages, including improved
25806 efficiency. However, it may cause some difficulties in porting existing Ada 83
25807 code which makes the assumption that, for example, pointers fit in 32 bits on
25808 a machine with 32-bit addressing.
25810 To get around this problem, GNAT also permits the use of ``thin pointers'' for
25811 access types in this case (where the designated type is an unconstrained array
25812 type). These thin pointers are indeed the same size as a System.Address value.
25813 To specify a thin pointer, use a size clause for the type, for example:
25815 @smallexample @c ada
25816 type X is access all String;
25817 for X'Size use Standard'Address_Size;
25821 which will cause the type X to be represented using a single pointer.
25822 When using this representation, the bounds are right behind the array.
25823 This representation is slightly less efficient, and does not allow quite
25824 such flexibility in the use of foreign pointers or in using the
25825 Unrestricted_Access attribute to create pointers to non-aliased objects.
25826 But for any standard portable use of the access type it will work in
25827 a functionally correct manner and allow porting of existing code.
25828 Note that another way of forcing a thin pointer representation
25829 is to use a component size clause for the element size in an array,
25830 or a record representation clause for an access field in a record.
25833 @node Compatibility with DEC Ada 83
25834 @section Compatibility with DEC Ada 83
25837 The VMS version of GNAT fully implements all the pragmas and attributes
25838 provided by DEC Ada 83, as well as providing the standard DEC Ada 83
25839 libraries, including Starlet. In addition, data layouts and parameter
25840 passing conventions are highly compatible. This means that porting
25841 existing DEC Ada 83 code to GNAT in VMS systems should be easier than
25842 most other porting efforts. The following are some of the most
25843 significant differences between GNAT and DEC Ada 83.
25846 @item Default floating-point representation
25847 In GNAT, the default floating-point format is IEEE, whereas in DEC Ada 83,
25848 it is VMS format. GNAT does implement the necessary pragmas
25849 (Long_Float, Float_Representation) for changing this default.
25852 The package System in GNAT exactly corresponds to the definition in the
25853 Ada 95 reference manual, which means that it excludes many of the
25854 DEC Ada 83 extensions. However, a separate package Aux_DEC is provided
25855 that contains the additional definitions, and a special pragma,
25856 Extend_System allows this package to be treated transparently as an
25857 extension of package System.
25860 The definitions provided by Aux_DEC are exactly compatible with those
25861 in the DEC Ada 83 version of System, with one exception.
25862 DEC Ada provides the following declarations:
25864 @smallexample @c ada
25865 TO_ADDRESS (INTEGER)
25866 TO_ADDRESS (UNSIGNED_LONGWORD)
25867 TO_ADDRESS (universal_integer)
25871 The version of TO_ADDRESS taking a universal integer argument is in fact
25872 an extension to Ada 83 not strictly compatible with the reference manual.
25873 In GNAT, we are constrained to be exactly compatible with the standard,
25874 and this means we cannot provide this capability. In DEC Ada 83, the
25875 point of this definition is to deal with a call like:
25877 @smallexample @c ada
25878 TO_ADDRESS (16#12777#);
25882 Normally, according to the Ada 83 standard, one would expect this to be
25883 ambiguous, since it matches both the INTEGER and UNSIGNED_LONGWORD forms
25884 of TO_ADDRESS@. However, in DEC Ada 83, there is no ambiguity, since the
25885 definition using universal_integer takes precedence.
25887 In GNAT, since the version with universal_integer cannot be supplied, it is
25888 not possible to be 100% compatible. Since there are many programs using
25889 numeric constants for the argument to TO_ADDRESS, the decision in GNAT was
25890 to change the name of the function in the UNSIGNED_LONGWORD case, so the
25891 declarations provided in the GNAT version of AUX_Dec are:
25893 @smallexample @c ada
25894 function To_Address (X : Integer) return Address;
25895 pragma Pure_Function (To_Address);
25897 function To_Address_Long (X : Unsigned_Longword)
25899 pragma Pure_Function (To_Address_Long);
25903 This means that programs using TO_ADDRESS for UNSIGNED_LONGWORD must
25904 change the name to TO_ADDRESS_LONG@.
25906 @item Task_Id values
25907 The Task_Id values assigned will be different in the two systems, and GNAT
25908 does not provide a specified value for the Task_Id of the environment task,
25909 which in GNAT is treated like any other declared task.
25912 For full details on these and other less significant compatibility issues,
25913 see appendix E of the Digital publication entitled @cite{DEC Ada, Technical
25914 Overview and Comparison on DIGITAL Platforms}.
25916 For GNAT running on other than VMS systems, all the DEC Ada 83 pragmas and
25917 attributes are recognized, although only a subset of them can sensibly
25918 be implemented. The description of pragmas in this reference manual
25919 indicates whether or not they are applicable to non-VMS systems.
25923 @node Transitioning from Alpha to Integrity OpenVMS
25924 @section Transitioning from Alpha to Integrity OpenVMS
25927 * Introduction to transitioning::
25928 * Migration of 32 bit code::
25929 * Taking advantage of 64 bit addressing::
25930 * Technical details::
25933 @node Introduction to transitioning
25934 @subsection Introduction to transitioning
25937 This guide is meant to assist users of GNAT Pro
25938 for Alpha OpenVMS who are planning to transition to the IA64 architecture.
25939 GNAT Pro for Open VMS Integrity has been designed to meet
25944 Providing a full conforming implementation of the Ada 95 language
25947 Allowing maximum backward compatibility, thus easing migration of existing
25951 Supplying a path for exploiting the full IA64 address range
25955 Ada's strong typing semantics has made it
25956 impractical to have different 32-bit and 64-bit modes. As soon as
25957 one object could possibly be outside the 32-bit address space, this
25958 would make it necessary for the @code{System.Address} type to be 64 bits.
25959 In particular, this would cause inconsistencies if 32-bit code is
25960 called from 64-bit code that raises an exception.
25962 This issue has been resolved by always using 64-bit addressing
25963 at the system level, but allowing for automatic conversions between
25964 32-bit and 64-bit addresses where required. Thus users who
25965 do not currently require 64-bit addressing capabilities, can
25966 recompile their code with only minimal changes (and indeed
25967 if the code is written in portable Ada, with no assumptions about
25968 the size of the @code{Address} type, then no changes at all are necessary).
25970 this approach provides a simple, gradual upgrade path to future
25971 use of larger memories than available for 32-bit systems.
25972 Also, newly written applications or libraries will by default
25973 be fully compatible with future systems exploiting 64-bit
25974 addressing capabilities present in IA64.
25976 @ref{Migration of 32 bit code}, will focus on porting applications
25977 that do not require more than 2 GB of
25978 addressable memory. This code will be referred to as
25979 @emph{32-bit code}.
25980 For applications intending to exploit the full ia64 address space,
25981 @ref{Taking advantage of 64 bit addressing},
25982 will consider further changes that may be required.
25983 Such code is called @emph{64-bit code} in the
25984 remainder of this guide.
25987 @node Migration of 32 bit code
25988 @subsection Migration of 32-bit code
25993 * Unchecked conversions::
25994 * Predefined constants::
25995 * Single source compatibility::
25996 * Experience with source compatibility::
25999 @node Address types
26000 @subsubsection Address types
26003 To solve the problem of mixing 64-bit and 32-bit addressing,
26004 while maintaining maximum backward compatibility, the following
26005 approach has been taken:
26009 @code{System.Address} always has a size of 64 bits
26012 @code{System.Short_Address} is a 32-bit subtype of @code{System.Address}
26017 Since @code{System.Short_Address} is a subtype of @code{System.Address},
26018 a @code{Short_Address}
26019 may be used where an @code{Address} is required, and vice versa, without
26020 needing explicit type conversions.
26021 By virtue of the Open VMS Integrity parameter passing conventions,
26023 and exported subprograms that have 32-bit address parameters are
26024 compatible with those that have 64-bit address parameters.
26025 (See @ref{Making code 64 bit clean} for details.)
26027 The areas that may need attention are those where record types have
26028 been defined that contain components of the type @code{System.Address}, and
26029 where objects of this type are passed to code expecting a record layout with
26032 Different compilers on different platforms cannot be
26033 expected to represent the same type in the same way,
26034 since alignment constraints
26035 and other system-dependent properties affect the compiler's decision.
26036 For that reason, Ada code
26037 generally uses representation clauses to specify the expected
26038 layout where required.
26040 If such a representation clause uses 32 bits for a component having
26041 the type @code{System.Address}, GNAT Pro for OpenVMS Integrity will detect
26042 that error and produce a specific diagnostic message.
26043 The developer should then determine whether the representation
26044 should be 64 bits or not and make either of two changes:
26045 change the size to 64 bits and leave the type as @code{System.Address}, or
26046 leave the size as 32 bits and change the type to @code{System.Short_Address}.
26047 Since @code{Short_Address} is a subtype of @code{Address}, no changes are
26048 required in any code setting or accessing the field; the compiler will
26049 automatically perform any needed conversions between address
26053 @subsubsection Access types
26056 By default, objects designated by access values are always
26057 allocated in the 32-bit
26058 address space. Thus legacy code will never contain
26059 any objects that are not addressable with 32-bit addresses, and
26060 the compiler will never raise exceptions as result of mixing
26061 32-bit and 64-bit addresses.
26063 However, the access values themselves are represented in 64 bits, for optimum
26064 performance and future compatibility with 64-bit code. As was
26065 the case with @code{System.Address}, the compiler will give an error message
26066 if an object or record component has a representation clause that
26067 requires the access value to fit in 32 bits. In such a situation,
26068 an explicit size clause for the access type, specifying 32 bits,
26069 will have the desired effect.
26071 General access types (declared with @code{access all}) can never be
26072 32 bits, as values of such types must be able to refer to any object
26073 of the designated type,
26074 including objects residing outside the 32-bit address range.
26075 Existing Ada 83 code will not contain such type definitions,
26076 however, since general access types were introduced in Ada 95.
26078 @node Unchecked conversions
26079 @subsubsection Unchecked conversions
26082 In the case of an @code{Unchecked_Conversion} where the source type is a
26083 64-bit access type or the type @code{System.Address}, and the target
26084 type is a 32-bit type, the compiler will generate a warning.
26085 Even though the generated code will still perform the required
26086 conversions, it is highly recommended in these cases to use
26087 respectively a 32-bit access type or @code{System.Short_Address}
26088 as the source type.
26090 @node Predefined constants
26091 @subsubsection Predefined constants
26094 The following predefined constants have changed:
26096 @multitable {@code{System.Address_Size}} {2**32} {2**64}
26097 @item @b{Constant} @tab @b{Old} @tab @b{New}
26098 @item @code{System.Word_Size} @tab 32 @tab 64
26099 @item @code{System.Memory_Size} @tab 2**32 @tab 2**64
26100 @item @code{System.Address_Size} @tab 32 @tab 64
26104 If you need to refer to the specific
26105 memory size of a 32-bit implementation, instead of the
26106 actual memory size, use @code{System.Short_Memory_Size}
26107 rather than @code{System.Memory_Size}.
26108 Similarly, references to @code{System.Address_Size} may need
26109 to be replaced by @code{System.Short_Address'Size}.
26110 The program @command{gnatfind} may be useful for locating
26111 references to the above constants, so that you can verify that they
26114 @node Single source compatibility
26115 @subsubsection Single source compatibility
26118 In order to allow the same source code to be compiled on
26119 both Alpha and IA64 platforms, GNAT Pro for Alpha/OpenVMS
26120 defines @code{System.Short_Address} and System.Short_Memory_Size
26121 as aliases of respectively @code{System.Address} and
26122 @code{System.Memory_Size}.
26123 (These aliases also leave the door open for a possible
26124 future ``upgrade'' of OpenVMS Alpha to a 64-bit address space.)
26126 @node Experience with source compatibility
26127 @subsubsection Experience with source compatibility
26130 The Security Server and STARLET provide an interesting ``test case''
26131 for source compatibility issues, since it is in such system code
26132 where assumptions about @code{Address} size might be expected to occur.
26133 Indeed, there were a small number of occasions in the Security Server
26134 file @file{jibdef.ads}
26135 where a representation clause for a record type specified
26136 32 bits for a component of type @code{Address}.
26137 All of these errors were detected by the compiler.
26138 The repair was obvious and immediate; to simply replace @code{Address} by
26139 @code{Short_Address}.
26141 In the case of STARLET, there were several record types that should
26142 have had representation clauses but did not. In these record types
26143 there was an implicit assumption that an @code{Address} value occupied
26145 These compiled without error, but their usage resulted in run-time error
26146 returns from STARLET system calls.
26147 To assist in the compile-time detection of such situations, we
26148 plan to include a switch to generate a warning message when a
26149 record component is of type @code{Address}.
26152 @c ****************************************
26153 @node Taking advantage of 64 bit addressing
26154 @subsection Taking advantage of 64-bit addressing
26157 * Making code 64 bit clean::
26158 * Allocating memory from the 64 bit storage pool::
26159 * Restrictions on use of 64 bit objects::
26160 * Using 64 bit storage pools by default::
26161 * General access types::
26162 * STARLET and other predefined libraries::
26165 @node Making code 64 bit clean
26166 @subsubsection Making code 64-bit clean
26169 In order to prevent problems that may occur when (parts of) a
26170 system start using memory outside the 32-bit address range,
26171 we recommend some additional guidelines:
26175 For imported subprograms that take parameters of the
26176 type @code{System.Address}, ensure that these subprograms can
26177 indeed handle 64-bit addresses. If not, or when in doubt,
26178 change the subprogram declaration to specify
26179 @code{System.Short_Address} instead.
26182 Resolve all warnings related to size mismatches in
26183 unchecked conversions. Failing to do so causes
26184 erroneous execution if the source object is outside
26185 the 32-bit address space.
26188 (optional) Explicitly use the 32-bit storage pool
26189 for access types used in a 32-bit context, or use
26190 generic access types where possible
26191 (@pxref{Restrictions on use of 64 bit objects}).
26195 If these rules are followed, the compiler will automatically insert
26196 any necessary checks to ensure that no addresses or access values
26197 passed to 32-bit code ever refer to objects outside the 32-bit
26199 Any attempt to do this will raise @code{Constraint_Error}.
26201 @node Allocating memory from the 64 bit storage pool
26202 @subsubsection Allocating memory from the 64-bit storage pool
26205 For any access type @code{T} that potentially requires memory allocations
26206 beyond the 32-bit address space,
26207 use the following representation clause:
26209 @smallexample @c ada
26210 for T'Storage_Pool use System.Pool_64;
26214 @node Restrictions on use of 64 bit objects
26215 @subsubsection Restrictions on use of 64-bit objects
26218 Taking the address of an object allocated from a 64-bit storage pool,
26219 and then passing this address to a subprogram expecting
26220 @code{System.Short_Address},
26221 or assigning it to a variable of type @code{Short_Address}, will cause
26222 @code{Constraint_Error} to be raised. In case the code is not 64-bit clean
26223 (@pxref{Making code 64 bit clean}), or checks are suppressed,
26224 no exception is raised and execution
26225 will become erroneous.
26227 @node Using 64 bit storage pools by default
26228 @subsubsection Using 64-bit storage pools by default
26231 In some cases it may be desirable to have the compiler allocate
26232 from 64-bit storage pools by default. This may be the case for
26233 libraries that are 64-bit clean, but may be used in both 32-bit
26234 and 64-bit contexts. For these cases the following configuration
26235 pragma may be specified:
26237 @smallexample @c ada
26238 pragma Pool_64_Default;
26242 Any code compiled in the context of this pragma will by default
26243 use the @code{System.Pool_64} storage pool. This default may be overridden
26244 for a specific access type @code{T} by the representation clause:
26246 @smallexample @c ada
26247 for T'Storage_Pool use System.Pool_32;
26251 Any object whose address may be passed to a subprogram with a
26252 @code{Short_Address} argument, or assigned to a variable of type
26253 @code{Short_Address}, needs to be allocated from this pool.
26255 @node General access types
26256 @subsubsection General access types
26259 Objects designated by access values from a
26260 general access type (declared with @code{access all}) are never allocated
26261 from a 64-bit storage pool. Code that uses general access types will
26262 accept objects allocated in either 32-bit or 64-bit address spaces,
26263 but never allocate objects outside the 32-bit address space.
26264 Using general access types ensures maximum compatibility with both
26265 32-bit and 64-bit code.
26268 @node STARLET and other predefined libraries
26269 @subsubsection STARLET and other predefined libraries
26272 All code that comes as part of GNAT is 64-bit clean, but the
26273 restrictions given in @ref{Restrictions on use of 64 bit objects},
26274 still apply. Look at the package
26275 specifications to see in which contexts objects allocated
26276 in 64-bit address space are acceptable.
26278 @node Technical details
26279 @subsection Technical details
26282 GNAT Pro for Open VMS Integrity takes advantage of the freedom given in the Ada
26283 standard with respect to the type of @code{System.Address}. Previous versions
26284 of GNAT Pro have defined this type as private and implemented it as
26287 In order to allow defining @code{System.Short_Address} as a proper subtype,
26288 and to match the implicit sign extension in parameter passing,
26289 in GNAT Pro for Open VMS Integrity, @code{System.Address} is defined as a
26290 visible (i.e., non-private) integer type.
26291 Standard operations on the type, such as the binary operators ``+'', ``-'',
26292 etc., that take @code{Address} operands and return an @code{Address} result,
26293 have been hidden by declaring these
26294 @code{abstract}, an Ada 95 feature that helps avoid the potential ambiguities
26295 that would otherwise result from overloading.
26296 (Note that, although @code{Address} is a visible integer type,
26297 good programming practice dictates against exploiting the type's
26298 integer properties such as literals, since this will compromise
26301 Defining @code{Address} as a visible integer type helps achieve
26302 maximum compatibility for existing Ada code,
26303 without sacrificing the capabilities of the IA64 architecture.
26307 @c ************************************************
26309 @node Microsoft Windows Topics
26310 @appendix Microsoft Windows Topics
26316 This chapter describes topics that are specific to the Microsoft Windows
26317 platforms (NT, 2000, and XP Professional).
26320 * Using GNAT on Windows::
26321 * Using a network installation of GNAT::
26322 * CONSOLE and WINDOWS subsystems::
26323 * Temporary Files::
26324 * Mixed-Language Programming on Windows::
26325 * Windows Calling Conventions::
26326 * Introduction to Dynamic Link Libraries (DLLs)::
26327 * Using DLLs with GNAT::
26328 * Building DLLs with GNAT::
26329 * Building DLLs with GNAT Project files::
26330 * Building DLLs with gnatdll::
26331 * GNAT and Windows Resources::
26332 * Debugging a DLL::
26333 * GNAT and COM/DCOM Objects::
26336 @node Using GNAT on Windows
26337 @section Using GNAT on Windows
26340 One of the strengths of the GNAT technology is that its tool set
26341 (@command{gcc}, @command{gnatbind}, @command{gnatlink}, @command{gnatmake}, the
26342 @code{gdb} debugger, etc.) is used in the same way regardless of the
26345 On Windows this tool set is complemented by a number of Microsoft-specific
26346 tools that have been provided to facilitate interoperability with Windows
26347 when this is required. With these tools:
26352 You can build applications using the @code{CONSOLE} or @code{WINDOWS}
26356 You can use any Dynamically Linked Library (DLL) in your Ada code (both
26357 relocatable and non-relocatable DLLs are supported).
26360 You can build Ada DLLs for use in other applications. These applications
26361 can be written in a language other than Ada (e.g., C, C++, etc). Again both
26362 relocatable and non-relocatable Ada DLLs are supported.
26365 You can include Windows resources in your Ada application.
26368 You can use or create COM/DCOM objects.
26372 Immediately below are listed all known general GNAT-for-Windows restrictions.
26373 Other restrictions about specific features like Windows Resources and DLLs
26374 are listed in separate sections below.
26379 It is not possible to use @code{GetLastError} and @code{SetLastError}
26380 when tasking, protected records, or exceptions are used. In these
26381 cases, in order to implement Ada semantics, the GNAT run-time system
26382 calls certain Win32 routines that set the last error variable to 0 upon
26383 success. It should be possible to use @code{GetLastError} and
26384 @code{SetLastError} when tasking, protected record, and exception
26385 features are not used, but it is not guaranteed to work.
26388 It is not possible to link against Microsoft libraries except for
26389 import libraries. The library must be built to be compatible with
26390 @file{MSVCRT.LIB} (/MD Microsoft compiler option), @file{LIBC.LIB} and
26391 @file{LIBCMT.LIB} (/ML or /MT Microsoft compiler options) are known to
26392 not be compatible with the GNAT runtime. Even if the library is
26393 compatible with @file{MSVCRT.LIB} it is not guaranteed to work.
26396 When the compilation environment is located on FAT32 drives, users may
26397 experience recompilations of the source files that have not changed if
26398 Daylight Saving Time (DST) state has changed since the last time files
26399 were compiled. NTFS drives do not have this problem.
26402 No components of the GNAT toolset use any entries in the Windows
26403 registry. The only entries that can be created are file associations and
26404 PATH settings, provided the user has chosen to create them at installation
26405 time, as well as some minimal book-keeping information needed to correctly
26406 uninstall or integrate different GNAT products.
26409 @node Using a network installation of GNAT
26410 @section Using a network installation of GNAT
26413 Make sure the system on which GNAT is installed is accessible from the
26414 current machine, i.e. the install location is shared over the network.
26415 Shared resources are accessed on Windows by means of UNC paths, which
26416 have the format @code{\\server\sharename\path}
26418 In order to use such a network installation, simply add the UNC path of the
26419 @file{bin} directory of your GNAT installation in front of your PATH. For
26420 example, if GNAT is installed in @file{\GNAT} directory of a share location
26421 called @file{c-drive} on a machine @file{LOKI}, the following command will
26424 @code{@ @ @ path \\loki\c-drive\gnat\bin;%path%}
26426 Be aware that every compilation using the network installation results in the
26427 transfer of large amounts of data across the network and will likely cause
26428 serious performance penalty.
26430 @node CONSOLE and WINDOWS subsystems
26431 @section CONSOLE and WINDOWS subsystems
26432 @cindex CONSOLE Subsystem
26433 @cindex WINDOWS Subsystem
26437 There are two main subsystems under Windows. The @code{CONSOLE} subsystem
26438 (which is the default subsystem) will always create a console when
26439 launching the application. This is not something desirable when the
26440 application has a Windows GUI. To get rid of this console the
26441 application must be using the @code{WINDOWS} subsystem. To do so
26442 the @option{-mwindows} linker option must be specified.
26445 $ gnatmake winprog -largs -mwindows
26448 @node Temporary Files
26449 @section Temporary Files
26450 @cindex Temporary files
26453 It is possible to control where temporary files gets created by setting
26454 the TMP environment variable. The file will be created:
26457 @item Under the directory pointed to by the TMP environment variable if
26458 this directory exists.
26460 @item Under c:\temp, if the TMP environment variable is not set (or not
26461 pointing to a directory) and if this directory exists.
26463 @item Under the current working directory otherwise.
26467 This allows you to determine exactly where the temporary
26468 file will be created. This is particularly useful in networked
26469 environments where you may not have write access to some
26472 @node Mixed-Language Programming on Windows
26473 @section Mixed-Language Programming on Windows
26476 Developing pure Ada applications on Windows is no different than on
26477 other GNAT-supported platforms. However, when developing or porting an
26478 application that contains a mix of Ada and C/C++, the choice of your
26479 Windows C/C++ development environment conditions your overall
26480 interoperability strategy.
26482 If you use @command{gcc} to compile the non-Ada part of your application,
26483 there are no Windows-specific restrictions that affect the overall
26484 interoperability with your Ada code. If you plan to use
26485 Microsoft tools (e.g. Microsoft Visual C/C++), you should be aware of
26486 the following limitations:
26490 You cannot link your Ada code with an object or library generated with
26491 Microsoft tools if these use the @code{.tls} section (Thread Local
26492 Storage section) since the GNAT linker does not yet support this section.
26495 You cannot link your Ada code with an object or library generated with
26496 Microsoft tools if these use I/O routines other than those provided in
26497 the Microsoft DLL: @code{msvcrt.dll}. This is because the GNAT run time
26498 uses the services of @code{msvcrt.dll} for its I/Os. Use of other I/O
26499 libraries can cause a conflict with @code{msvcrt.dll} services. For
26500 instance Visual C++ I/O stream routines conflict with those in
26505 If you do want to use the Microsoft tools for your non-Ada code and hit one
26506 of the above limitations, you have two choices:
26510 Encapsulate your non Ada code in a DLL to be linked with your Ada
26511 application. In this case, use the Microsoft or whatever environment to
26512 build the DLL and use GNAT to build your executable
26513 (@pxref{Using DLLs with GNAT}).
26516 Or you can encapsulate your Ada code in a DLL to be linked with the
26517 other part of your application. In this case, use GNAT to build the DLL
26518 (@pxref{Building DLLs with GNAT}) and use the Microsoft or whatever
26519 environment to build your executable.
26522 @node Windows Calling Conventions
26523 @section Windows Calling Conventions
26528 * C Calling Convention::
26529 * Stdcall Calling Convention::
26530 * DLL Calling Convention::
26534 When a subprogram @code{F} (caller) calls a subprogram @code{G}
26535 (callee), there are several ways to push @code{G}'s parameters on the
26536 stack and there are several possible scenarios to clean up the stack
26537 upon @code{G}'s return. A calling convention is an agreed upon software
26538 protocol whereby the responsibilities between the caller (@code{F}) and
26539 the callee (@code{G}) are clearly defined. Several calling conventions
26540 are available for Windows:
26544 @code{C} (Microsoft defined)
26547 @code{Stdcall} (Microsoft defined)
26550 @code{DLL} (GNAT specific)
26553 @node C Calling Convention
26554 @subsection @code{C} Calling Convention
26557 This is the default calling convention used when interfacing to C/C++
26558 routines compiled with either @command{gcc} or Microsoft Visual C++.
26560 In the @code{C} calling convention subprogram parameters are pushed on the
26561 stack by the caller from right to left. The caller itself is in charge of
26562 cleaning up the stack after the call. In addition, the name of a routine
26563 with @code{C} calling convention is mangled by adding a leading underscore.
26565 The name to use on the Ada side when importing (or exporting) a routine
26566 with @code{C} calling convention is the name of the routine. For
26567 instance the C function:
26570 int get_val (long);
26574 should be imported from Ada as follows:
26576 @smallexample @c ada
26578 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
26579 pragma Import (C, Get_Val, External_Name => "get_val");
26584 Note that in this particular case the @code{External_Name} parameter could
26585 have been omitted since, when missing, this parameter is taken to be the
26586 name of the Ada entity in lower case. When the @code{Link_Name} parameter
26587 is missing, as in the above example, this parameter is set to be the
26588 @code{External_Name} with a leading underscore.
26590 When importing a variable defined in C, you should always use the @code{C}
26591 calling convention unless the object containing the variable is part of a
26592 DLL (in which case you should use the @code{DLL} calling convention,
26593 @pxref{DLL Calling Convention}).
26595 @node Stdcall Calling Convention
26596 @subsection @code{Stdcall} Calling Convention
26599 This convention, which was the calling convention used for Pascal
26600 programs, is used by Microsoft for all the routines in the Win32 API for
26601 efficiency reasons. It must be used to import any routine for which this
26602 convention was specified.
26604 In the @code{Stdcall} calling convention subprogram parameters are pushed
26605 on the stack by the caller from right to left. The callee (and not the
26606 caller) is in charge of cleaning the stack on routine exit. In addition,
26607 the name of a routine with @code{Stdcall} calling convention is mangled by
26608 adding a leading underscore (as for the @code{C} calling convention) and a
26609 trailing @code{@@}@code{@i{nn}}, where @i{nn} is the overall size (in
26610 bytes) of the parameters passed to the routine.
26612 The name to use on the Ada side when importing a C routine with a
26613 @code{Stdcall} calling convention is the name of the C routine. The leading
26614 underscore and trailing @code{@@}@code{@i{nn}} are added automatically by
26615 the compiler. For instance the Win32 function:
26618 @b{APIENTRY} int get_val (long);
26622 should be imported from Ada as follows:
26624 @smallexample @c ada
26626 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
26627 pragma Import (Stdcall, Get_Val);
26628 -- On the x86 a long is 4 bytes, so the Link_Name is "_get_val@@4"
26633 As for the @code{C} calling convention, when the @code{External_Name}
26634 parameter is missing, it is taken to be the name of the Ada entity in lower
26635 case. If instead of writing the above import pragma you write:
26637 @smallexample @c ada
26639 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
26640 pragma Import (Stdcall, Get_Val, External_Name => "retrieve_val");
26645 then the imported routine is @code{_retrieve_val@@4}. However, if instead
26646 of specifying the @code{External_Name} parameter you specify the
26647 @code{Link_Name} as in the following example:
26649 @smallexample @c ada
26651 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
26652 pragma Import (Stdcall, Get_Val, Link_Name => "retrieve_val");
26657 then the imported routine is @code{retrieve_val@@4}, that is, there is no
26658 trailing underscore but the appropriate @code{@@}@code{@i{nn}} is always
26659 added at the end of the @code{Link_Name} by the compiler.
26662 Note, that in some special cases a DLL's entry point name lacks a trailing
26663 @code{@@}@code{@i{nn}} while the exported name generated for a call has it.
26664 The @code{gnatdll} tool, which creates the import library for the DLL, is able
26665 to handle those cases (@pxref{Using gnatdll} for the description of
26668 @node DLL Calling Convention
26669 @subsection @code{DLL} Calling Convention
26672 This convention, which is GNAT-specific, must be used when you want to
26673 import in Ada a variables defined in a DLL. For functions and procedures
26674 this convention is equivalent to the @code{Stdcall} convention. As an
26675 example, if a DLL contains a variable defined as:
26682 then, to access this variable from Ada you should write:
26684 @smallexample @c ada
26686 My_Var : Interfaces.C.int;
26687 pragma Import (DLL, My_Var);
26691 The remarks concerning the @code{External_Name} and @code{Link_Name}
26692 parameters given in the previous sections equally apply to the @code{DLL}
26693 calling convention.
26695 @node Introduction to Dynamic Link Libraries (DLLs)
26696 @section Introduction to Dynamic Link Libraries (DLLs)
26700 A Dynamically Linked Library (DLL) is a library that can be shared by
26701 several applications running under Windows. A DLL can contain any number of
26702 routines and variables.
26704 One advantage of DLLs is that you can change and enhance them without
26705 forcing all the applications that depend on them to be relinked or
26706 recompiled. However, you should be aware than all calls to DLL routines are
26707 slower since, as you will understand below, such calls are indirect.
26709 To illustrate the remainder of this section, suppose that an application
26710 wants to use the services of a DLL @file{API.dll}. To use the services
26711 provided by @file{API.dll} you must statically link against the DLL or
26712 an import library which contains a jump table with an entry for each
26713 routine and variable exported by the DLL. In the Microsoft world this
26714 import library is called @file{API.lib}. When using GNAT this import
26715 library is called either @file{libAPI.a} or @file{libapi.a} (names are
26718 After you have linked your application with the DLL or the import library
26719 and you run your application, here is what happens:
26723 Your application is loaded into memory.
26726 The DLL @file{API.dll} is mapped into the address space of your
26727 application. This means that:
26731 The DLL will use the stack of the calling thread.
26734 The DLL will use the virtual address space of the calling process.
26737 The DLL will allocate memory from the virtual address space of the calling
26741 Handles (pointers) can be safely exchanged between routines in the DLL
26742 routines and routines in the application using the DLL.
26746 The entries in the jump table (from the import library @file{libAPI.a}
26747 or @file{API.lib} or automatically created when linking against a DLL)
26748 which is part of your application are initialized with the addresses
26749 of the routines and variables in @file{API.dll}.
26752 If present in @file{API.dll}, routines @code{DllMain} or
26753 @code{DllMainCRTStartup} are invoked. These routines typically contain
26754 the initialization code needed for the well-being of the routines and
26755 variables exported by the DLL.
26759 There is an additional point which is worth mentioning. In the Windows
26760 world there are two kind of DLLs: relocatable and non-relocatable
26761 DLLs. Non-relocatable DLLs can only be loaded at a very specific address
26762 in the target application address space. If the addresses of two
26763 non-relocatable DLLs overlap and these happen to be used by the same
26764 application, a conflict will occur and the application will run
26765 incorrectly. Hence, when possible, it is always preferable to use and
26766 build relocatable DLLs. Both relocatable and non-relocatable DLLs are
26767 supported by GNAT. Note that the @option{-s} linker option (see GNU Linker
26768 User's Guide) removes the debugging symbols from the DLL but the DLL can
26769 still be relocated.
26771 As a side note, an interesting difference between Microsoft DLLs and
26772 Unix shared libraries, is the fact that on most Unix systems all public
26773 routines are exported by default in a Unix shared library, while under
26774 Windows it is possible (but not required) to list exported routines in
26775 a definition file (@pxref{The Definition File}).
26777 @node Using DLLs with GNAT
26778 @section Using DLLs with GNAT
26781 * Creating an Ada Spec for the DLL Services::
26782 * Creating an Import Library::
26786 To use the services of a DLL, say @file{API.dll}, in your Ada application
26791 The Ada spec for the routines and/or variables you want to access in
26792 @file{API.dll}. If not available this Ada spec must be built from the C/C++
26793 header files provided with the DLL.
26796 The import library (@file{libAPI.a} or @file{API.lib}). As previously
26797 mentioned an import library is a statically linked library containing the
26798 import table which will be filled at load time to point to the actual
26799 @file{API.dll} routines. Sometimes you don't have an import library for the
26800 DLL you want to use. The following sections will explain how to build
26801 one. Note that this is optional.
26804 The actual DLL, @file{API.dll}.
26808 Once you have all the above, to compile an Ada application that uses the
26809 services of @file{API.dll} and whose main subprogram is @code{My_Ada_App},
26810 you simply issue the command
26813 $ gnatmake my_ada_app -largs -lAPI
26817 The argument @option{-largs -lAPI} at the end of the @command{gnatmake} command
26818 tells the GNAT linker to look first for a library named @file{API.lib}
26819 (Microsoft-style name) and if not found for a library named @file{libAPI.a}
26820 (GNAT-style name). Note that if the Ada package spec for @file{API.dll}
26821 contains the following pragma
26823 @smallexample @c ada
26824 pragma Linker_Options ("-lAPI");
26828 you do not have to add @option{-largs -lAPI} at the end of the
26829 @command{gnatmake} command.
26831 If any one of the items above is missing you will have to create it
26832 yourself. The following sections explain how to do so using as an
26833 example a fictitious DLL called @file{API.dll}.
26835 @node Creating an Ada Spec for the DLL Services
26836 @subsection Creating an Ada Spec for the DLL Services
26839 A DLL typically comes with a C/C++ header file which provides the
26840 definitions of the routines and variables exported by the DLL. The Ada
26841 equivalent of this header file is a package spec that contains definitions
26842 for the imported entities. If the DLL you intend to use does not come with
26843 an Ada spec you have to generate one such spec yourself. For example if
26844 the header file of @file{API.dll} is a file @file{api.h} containing the
26845 following two definitions:
26857 then the equivalent Ada spec could be:
26859 @smallexample @c ada
26862 with Interfaces.C.Strings;
26867 function Get (Str : C.Strings.Chars_Ptr) return C.int;
26870 pragma Import (C, Get);
26871 pragma Import (DLL, Some_Var);
26878 Note that a variable is @strong{always imported with a DLL convention}. A
26879 function can have @code{C}, @code{Stdcall} or @code{DLL} convention. For
26880 subprograms, the @code{DLL} convention is a synonym of @code{Stdcall}
26881 (@pxref{Windows Calling Conventions}).
26883 @node Creating an Import Library
26884 @subsection Creating an Import Library
26885 @cindex Import library
26888 * The Definition File::
26889 * GNAT-Style Import Library::
26890 * Microsoft-Style Import Library::
26894 If a Microsoft-style import library @file{API.lib} or a GNAT-style
26895 import library @file{libAPI.a} is available with @file{API.dll} you
26896 can skip this section. You can also skip this section if
26897 @file{API.dll} is built with GNU tools as in this case it is possible
26898 to link directly against the DLL. Otherwise read on.
26900 @node The Definition File
26901 @subsubsection The Definition File
26902 @cindex Definition file
26906 As previously mentioned, and unlike Unix systems, the list of symbols
26907 that are exported from a DLL must be provided explicitly in Windows.
26908 The main goal of a definition file is precisely that: list the symbols
26909 exported by a DLL. A definition file (usually a file with a @code{.def}
26910 suffix) has the following structure:
26916 [DESCRIPTION @i{string}]
26926 @item LIBRARY @i{name}
26927 This section, which is optional, gives the name of the DLL.
26929 @item DESCRIPTION @i{string}
26930 This section, which is optional, gives a description string that will be
26931 embedded in the import library.
26934 This section gives the list of exported symbols (procedures, functions or
26935 variables). For instance in the case of @file{API.dll} the @code{EXPORTS}
26936 section of @file{API.def} looks like:
26950 Note that you must specify the correct suffix (@code{@@}@code{@i{nn}})
26951 (@pxref{Windows Calling Conventions}) for a Stdcall
26952 calling convention function in the exported symbols list.
26955 There can actually be other sections in a definition file, but these
26956 sections are not relevant to the discussion at hand.
26958 @node GNAT-Style Import Library
26959 @subsubsection GNAT-Style Import Library
26962 To create a static import library from @file{API.dll} with the GNAT tools
26963 you should proceed as follows:
26967 Create the definition file @file{API.def} (@pxref{The Definition File}).
26968 For that use the @code{dll2def} tool as follows:
26971 $ dll2def API.dll > API.def
26975 @code{dll2def} is a very simple tool: it takes as input a DLL and prints
26976 to standard output the list of entry points in the DLL. Note that if
26977 some routines in the DLL have the @code{Stdcall} convention
26978 (@pxref{Windows Calling Conventions}) with stripped @code{@@}@i{nn}
26979 suffix then you'll have to edit @file{api.def} to add it, and specify
26980 @code{-k} to @code{gnatdll} when creating the import library.
26983 Here are some hints to find the right @code{@@}@i{nn} suffix.
26987 If you have the Microsoft import library (.lib), it is possible to get
26988 the right symbols by using Microsoft @code{dumpbin} tool (see the
26989 corresponding Microsoft documentation for further details).
26992 $ dumpbin /exports api.lib
26996 If you have a message about a missing symbol at link time the compiler
26997 tells you what symbol is expected. You just have to go back to the
26998 definition file and add the right suffix.
27002 Build the import library @code{libAPI.a}, using @code{gnatdll}
27003 (@pxref{Using gnatdll}) as follows:
27006 $ gnatdll -e API.def -d API.dll
27010 @code{gnatdll} takes as input a definition file @file{API.def} and the
27011 name of the DLL containing the services listed in the definition file
27012 @file{API.dll}. The name of the static import library generated is
27013 computed from the name of the definition file as follows: if the
27014 definition file name is @i{xyz}@code{.def}, the import library name will
27015 be @code{lib}@i{xyz}@code{.a}. Note that in the previous example option
27016 @option{-e} could have been removed because the name of the definition
27017 file (before the ``@code{.def}'' suffix) is the same as the name of the
27018 DLL (@pxref{Using gnatdll} for more information about @code{gnatdll}).
27021 @node Microsoft-Style Import Library
27022 @subsubsection Microsoft-Style Import Library
27025 With GNAT you can either use a GNAT-style or Microsoft-style import
27026 library. A Microsoft import library is needed only if you plan to make an
27027 Ada DLL available to applications developed with Microsoft
27028 tools (@pxref{Mixed-Language Programming on Windows}).
27030 To create a Microsoft-style import library for @file{API.dll} you
27031 should proceed as follows:
27035 Create the definition file @file{API.def} from the DLL. For this use either
27036 the @code{dll2def} tool as described above or the Microsoft @code{dumpbin}
27037 tool (see the corresponding Microsoft documentation for further details).
27040 Build the actual import library using Microsoft's @code{lib} utility:
27043 $ lib -machine:IX86 -def:API.def -out:API.lib
27047 If you use the above command the definition file @file{API.def} must
27048 contain a line giving the name of the DLL:
27055 See the Microsoft documentation for further details about the usage of
27059 @node Building DLLs with GNAT
27060 @section Building DLLs with GNAT
27061 @cindex DLLs, building
27064 This section explain how to build DLLs using the GNAT built-in DLL
27065 support. With the following procedure it is straight forward to build
27066 and use DLLs with GNAT.
27070 @item building object files
27072 The first step is to build all objects files that are to be included
27073 into the DLL. This is done by using the standard @command{gnatmake} tool.
27075 @item building the DLL
27077 To build the DLL you must use @command{gcc}'s @code{-shared}
27078 option. It is quite simple to use this method:
27081 $ gcc -shared -o api.dll obj1.o obj2.o ...
27084 It is important to note that in this case all symbols found in the
27085 object files are automatically exported. It is possible to restrict
27086 the set of symbols to export by passing to @command{gcc} a definition
27087 file, @pxref{The Definition File}. For example:
27090 $ gcc -shared -o api.dll api.def obj1.o obj2.o ...
27093 If you use a definition file you must export the elaboration procedures
27094 for every package that required one. Elaboration procedures are named
27095 using the package name followed by "_E".
27097 @item preparing DLL to be used
27099 For the DLL to be used by client programs the bodies must be hidden
27100 from it and the .ali set with read-only attribute. This is very important
27101 otherwise GNAT will recompile all packages and will not actually use
27102 the code in the DLL. For example:
27106 $ copy *.ads *.ali api.dll apilib
27107 $ attrib +R apilib\*.ali
27112 At this point it is possible to use the DLL by directly linking
27113 against it. Note that you must use the GNAT shared runtime when using
27114 GNAT shared libraries. This is achieved by using @code{-shared} binder's
27118 $ gnatmake main -Iapilib -bargs -shared -largs -Lapilib -lAPI
27121 @node Building DLLs with GNAT Project files
27122 @section Building DLLs with GNAT Project files
27123 @cindex DLLs, building
27126 There is nothing specific to Windows in this area. @pxref{Library Projects}.
27128 @node Building DLLs with gnatdll
27129 @section Building DLLs with gnatdll
27130 @cindex DLLs, building
27133 * Limitations When Using Ada DLLs from Ada::
27134 * Exporting Ada Entities::
27135 * Ada DLLs and Elaboration::
27136 * Ada DLLs and Finalization::
27137 * Creating a Spec for Ada DLLs::
27138 * Creating the Definition File::
27143 Note that it is preferred to use the built-in GNAT DLL support
27144 (@pxref{Building DLLs with GNAT}) or GNAT Project files
27145 (@pxref{Building DLLs with GNAT Project files}) to build DLLs.
27147 This section explains how to build DLLs containing Ada code using
27148 @code{gnatdll}. These DLLs will be referred to as Ada DLLs in the
27149 remainder of this section.
27151 The steps required to build an Ada DLL that is to be used by Ada as well as
27152 non-Ada applications are as follows:
27156 You need to mark each Ada @i{entity} exported by the DLL with a @code{C} or
27157 @code{Stdcall} calling convention to avoid any Ada name mangling for the
27158 entities exported by the DLL (@pxref{Exporting Ada Entities}). You can
27159 skip this step if you plan to use the Ada DLL only from Ada applications.
27162 Your Ada code must export an initialization routine which calls the routine
27163 @code{adainit} generated by @command{gnatbind} to perform the elaboration of
27164 the Ada code in the DLL (@pxref{Ada DLLs and Elaboration}). The initialization
27165 routine exported by the Ada DLL must be invoked by the clients of the DLL
27166 to initialize the DLL.
27169 When useful, the DLL should also export a finalization routine which calls
27170 routine @code{adafinal} generated by @command{gnatbind} to perform the
27171 finalization of the Ada code in the DLL (@pxref{Ada DLLs and Finalization}).
27172 The finalization routine exported by the Ada DLL must be invoked by the
27173 clients of the DLL when the DLL services are no further needed.
27176 You must provide a spec for the services exported by the Ada DLL in each
27177 of the programming languages to which you plan to make the DLL available.
27180 You must provide a definition file listing the exported entities
27181 (@pxref{The Definition File}).
27184 Finally you must use @code{gnatdll} to produce the DLL and the import
27185 library (@pxref{Using gnatdll}).
27189 Note that a relocatable DLL stripped using the @code{strip} binutils
27190 tool will not be relocatable anymore. To build a DLL without debug
27191 information pass @code{-largs -s} to @code{gnatdll}.
27193 @node Limitations When Using Ada DLLs from Ada
27194 @subsection Limitations When Using Ada DLLs from Ada
27197 When using Ada DLLs from Ada applications there is a limitation users
27198 should be aware of. Because on Windows the GNAT run time is not in a DLL of
27199 its own, each Ada DLL includes a part of the GNAT run time. Specifically,
27200 each Ada DLL includes the services of the GNAT run time that are necessary
27201 to the Ada code inside the DLL. As a result, when an Ada program uses an
27202 Ada DLL there are two independent GNAT run times: one in the Ada DLL and
27203 one in the main program.
27205 It is therefore not possible to exchange GNAT run-time objects between the
27206 Ada DLL and the main Ada program. Example of GNAT run-time objects are file
27207 handles (e.g. @code{Text_IO.File_Type}), tasks types, protected objects
27210 It is completely safe to exchange plain elementary, array or record types,
27211 Windows object handles, etc.
27213 @node Exporting Ada Entities
27214 @subsection Exporting Ada Entities
27215 @cindex Export table
27218 Building a DLL is a way to encapsulate a set of services usable from any
27219 application. As a result, the Ada entities exported by a DLL should be
27220 exported with the @code{C} or @code{Stdcall} calling conventions to avoid
27221 any Ada name mangling. Please note that the @code{Stdcall} convention
27222 should only be used for subprograms, not for variables. As an example here
27223 is an Ada package @code{API}, spec and body, exporting two procedures, a
27224 function, and a variable:
27226 @smallexample @c ada
27229 with Interfaces.C; use Interfaces;
27231 Count : C.int := 0;
27232 function Factorial (Val : C.int) return C.int;
27234 procedure Initialize_API;
27235 procedure Finalize_API;
27236 -- Initialization & Finalization routines. More in the next section.
27238 pragma Export (C, Initialize_API);
27239 pragma Export (C, Finalize_API);
27240 pragma Export (C, Count);
27241 pragma Export (C, Factorial);
27247 @smallexample @c ada
27250 package body API is
27251 function Factorial (Val : C.int) return C.int is
27254 Count := Count + 1;
27255 for K in 1 .. Val loop
27261 procedure Initialize_API is
27263 pragma Import (C, Adainit);
27266 end Initialize_API;
27268 procedure Finalize_API is
27269 procedure Adafinal;
27270 pragma Import (C, Adafinal);
27280 If the Ada DLL you are building will only be used by Ada applications
27281 you do not have to export Ada entities with a @code{C} or @code{Stdcall}
27282 convention. As an example, the previous package could be written as
27285 @smallexample @c ada
27289 Count : Integer := 0;
27290 function Factorial (Val : Integer) return Integer;
27292 procedure Initialize_API;
27293 procedure Finalize_API;
27294 -- Initialization and Finalization routines.
27300 @smallexample @c ada
27303 package body API is
27304 function Factorial (Val : Integer) return Integer is
27305 Fact : Integer := 1;
27307 Count := Count + 1;
27308 for K in 1 .. Val loop
27315 -- The remainder of this package body is unchanged.
27322 Note that if you do not export the Ada entities with a @code{C} or
27323 @code{Stdcall} convention you will have to provide the mangled Ada names
27324 in the definition file of the Ada DLL
27325 (@pxref{Creating the Definition File}).
27327 @node Ada DLLs and Elaboration
27328 @subsection Ada DLLs and Elaboration
27329 @cindex DLLs and elaboration
27332 The DLL that you are building contains your Ada code as well as all the
27333 routines in the Ada library that are needed by it. The first thing a
27334 user of your DLL must do is elaborate the Ada code
27335 (@pxref{Elaboration Order Handling in GNAT}).
27337 To achieve this you must export an initialization routine
27338 (@code{Initialize_API} in the previous example), which must be invoked
27339 before using any of the DLL services. This elaboration routine must call
27340 the Ada elaboration routine @code{adainit} generated by the GNAT binder
27341 (@pxref{Binding with Non-Ada Main Programs}). See the body of
27342 @code{Initialize_Api} for an example. Note that the GNAT binder is
27343 automatically invoked during the DLL build process by the @code{gnatdll}
27344 tool (@pxref{Using gnatdll}).
27346 When a DLL is loaded, Windows systematically invokes a routine called
27347 @code{DllMain}. It would therefore be possible to call @code{adainit}
27348 directly from @code{DllMain} without having to provide an explicit
27349 initialization routine. Unfortunately, it is not possible to call
27350 @code{adainit} from the @code{DllMain} if your program has library level
27351 tasks because access to the @code{DllMain} entry point is serialized by
27352 the system (that is, only a single thread can execute ``through'' it at a
27353 time), which means that the GNAT run time will deadlock waiting for the
27354 newly created task to complete its initialization.
27356 @node Ada DLLs and Finalization
27357 @subsection Ada DLLs and Finalization
27358 @cindex DLLs and finalization
27361 When the services of an Ada DLL are no longer needed, the client code should
27362 invoke the DLL finalization routine, if available. The DLL finalization
27363 routine is in charge of releasing all resources acquired by the DLL. In the
27364 case of the Ada code contained in the DLL, this is achieved by calling
27365 routine @code{adafinal} generated by the GNAT binder
27366 (@pxref{Binding with Non-Ada Main Programs}).
27367 See the body of @code{Finalize_Api} for an
27368 example. As already pointed out the GNAT binder is automatically invoked
27369 during the DLL build process by the @code{gnatdll} tool
27370 (@pxref{Using gnatdll}).
27372 @node Creating a Spec for Ada DLLs
27373 @subsection Creating a Spec for Ada DLLs
27376 To use the services exported by the Ada DLL from another programming
27377 language (e.g. C), you have to translate the specs of the exported Ada
27378 entities in that language. For instance in the case of @code{API.dll},
27379 the corresponding C header file could look like:
27384 extern int *_imp__count;
27385 #define count (*_imp__count)
27386 int factorial (int);
27392 It is important to understand that when building an Ada DLL to be used by
27393 other Ada applications, you need two different specs for the packages
27394 contained in the DLL: one for building the DLL and the other for using
27395 the DLL. This is because the @code{DLL} calling convention is needed to
27396 use a variable defined in a DLL, but when building the DLL, the variable
27397 must have either the @code{Ada} or @code{C} calling convention. As an
27398 example consider a DLL comprising the following package @code{API}:
27400 @smallexample @c ada
27404 Count : Integer := 0;
27406 -- Remainder of the package omitted.
27413 After producing a DLL containing package @code{API}, the spec that
27414 must be used to import @code{API.Count} from Ada code outside of the
27417 @smallexample @c ada
27422 pragma Import (DLL, Count);
27428 @node Creating the Definition File
27429 @subsection Creating the Definition File
27432 The definition file is the last file needed to build the DLL. It lists
27433 the exported symbols. As an example, the definition file for a DLL
27434 containing only package @code{API} (where all the entities are exported
27435 with a @code{C} calling convention) is:
27450 If the @code{C} calling convention is missing from package @code{API},
27451 then the definition file contains the mangled Ada names of the above
27452 entities, which in this case are:
27461 api__initialize_api
27466 @node Using gnatdll
27467 @subsection Using @code{gnatdll}
27471 * gnatdll Example::
27472 * gnatdll behind the Scenes::
27477 @code{gnatdll} is a tool to automate the DLL build process once all the Ada
27478 and non-Ada sources that make up your DLL have been compiled.
27479 @code{gnatdll} is actually in charge of two distinct tasks: build the
27480 static import library for the DLL and the actual DLL. The form of the
27481 @code{gnatdll} command is
27485 $ gnatdll [@var{switches}] @var{list-of-files} [-largs @var{opts}]
27490 where @i{list-of-files} is a list of ALI and object files. The object
27491 file list must be the exact list of objects corresponding to the non-Ada
27492 sources whose services are to be included in the DLL. The ALI file list
27493 must be the exact list of ALI files for the corresponding Ada sources
27494 whose services are to be included in the DLL. If @i{list-of-files} is
27495 missing, only the static import library is generated.
27498 You may specify any of the following switches to @code{gnatdll}:
27501 @item -a[@var{address}]
27502 @cindex @option{-a} (@code{gnatdll})
27503 Build a non-relocatable DLL at @var{address}. If @var{address} is not
27504 specified the default address @var{0x11000000} will be used. By default,
27505 when this switch is missing, @code{gnatdll} builds relocatable DLL. We
27506 advise the reader to build relocatable DLL.
27508 @item -b @var{address}
27509 @cindex @option{-b} (@code{gnatdll})
27510 Set the relocatable DLL base address. By default the address is
27513 @item -bargs @var{opts}
27514 @cindex @option{-bargs} (@code{gnatdll})
27515 Binder options. Pass @var{opts} to the binder.
27517 @item -d @var{dllfile}
27518 @cindex @option{-d} (@code{gnatdll})
27519 @var{dllfile} is the name of the DLL. This switch must be present for
27520 @code{gnatdll} to do anything. The name of the generated import library is
27521 obtained algorithmically from @var{dllfile} as shown in the following
27522 example: if @var{dllfile} is @code{xyz.dll}, the import library name is
27523 @code{libxyz.a}. The name of the definition file to use (if not specified
27524 by option @option{-e}) is obtained algorithmically from @var{dllfile}
27525 as shown in the following example:
27526 if @var{dllfile} is @code{xyz.dll}, the definition
27527 file used is @code{xyz.def}.
27529 @item -e @var{deffile}
27530 @cindex @option{-e} (@code{gnatdll})
27531 @var{deffile} is the name of the definition file.
27534 @cindex @option{-g} (@code{gnatdll})
27535 Generate debugging information. This information is stored in the object
27536 file and copied from there to the final DLL file by the linker,
27537 where it can be read by the debugger. You must use the
27538 @option{-g} switch if you plan on using the debugger or the symbolic
27542 @cindex @option{-h} (@code{gnatdll})
27543 Help mode. Displays @code{gnatdll} switch usage information.
27546 @cindex @option{-I} (@code{gnatdll})
27547 Direct @code{gnatdll} to search the @var{dir} directory for source and
27548 object files needed to build the DLL.
27549 (@pxref{Search Paths and the Run-Time Library (RTL)}).
27552 @cindex @option{-k} (@code{gnatdll})
27553 Removes the @code{@@}@i{nn} suffix from the import library's exported
27554 names, but keeps them for the link names. You must specify this
27555 option if you want to use a @code{Stdcall} function in a DLL for which
27556 the @code{@@}@i{nn} suffix has been removed. This is the case for most
27557 of the Windows NT DLL for example. This option has no effect when
27558 @option{-n} option is specified.
27560 @item -l @var{file}
27561 @cindex @option{-l} (@code{gnatdll})
27562 The list of ALI and object files used to build the DLL are listed in
27563 @var{file}, instead of being given in the command line. Each line in
27564 @var{file} contains the name of an ALI or object file.
27567 @cindex @option{-n} (@code{gnatdll})
27568 No Import. Do not create the import library.
27571 @cindex @option{-q} (@code{gnatdll})
27572 Quiet mode. Do not display unnecessary messages.
27575 @cindex @option{-v} (@code{gnatdll})
27576 Verbose mode. Display extra information.
27578 @item -largs @var{opts}
27579 @cindex @option{-largs} (@code{gnatdll})
27580 Linker options. Pass @var{opts} to the linker.
27583 @node gnatdll Example
27584 @subsubsection @code{gnatdll} Example
27587 As an example the command to build a relocatable DLL from @file{api.adb}
27588 once @file{api.adb} has been compiled and @file{api.def} created is
27591 $ gnatdll -d api.dll api.ali
27595 The above command creates two files: @file{libapi.a} (the import
27596 library) and @file{api.dll} (the actual DLL). If you want to create
27597 only the DLL, just type:
27600 $ gnatdll -d api.dll -n api.ali
27604 Alternatively if you want to create just the import library, type:
27607 $ gnatdll -d api.dll
27610 @node gnatdll behind the Scenes
27611 @subsubsection @code{gnatdll} behind the Scenes
27614 This section details the steps involved in creating a DLL. @code{gnatdll}
27615 does these steps for you. Unless you are interested in understanding what
27616 goes on behind the scenes, you should skip this section.
27618 We use the previous example of a DLL containing the Ada package @code{API},
27619 to illustrate the steps necessary to build a DLL. The starting point is a
27620 set of objects that will make up the DLL and the corresponding ALI
27621 files. In the case of this example this means that @file{api.o} and
27622 @file{api.ali} are available. To build a relocatable DLL, @code{gnatdll} does
27627 @code{gnatdll} builds the base file (@file{api.base}). A base file gives
27628 the information necessary to generate relocation information for the
27634 $ gnatlink api -o api.jnk -mdll -Wl,--base-file,api.base
27639 In addition to the base file, the @command{gnatlink} command generates an
27640 output file @file{api.jnk} which can be discarded. The @option{-mdll} switch
27641 asks @command{gnatlink} to generate the routines @code{DllMain} and
27642 @code{DllMainCRTStartup} that are called by the Windows loader when the DLL
27643 is loaded into memory.
27646 @code{gnatdll} uses @code{dlltool} (@pxref{Using dlltool}) to build the
27647 export table (@file{api.exp}). The export table contains the relocation
27648 information in a form which can be used during the final link to ensure
27649 that the Windows loader is able to place the DLL anywhere in memory.
27653 $ dlltool --dllname api.dll --def api.def --base-file api.base \
27654 --output-exp api.exp
27659 @code{gnatdll} builds the base file using the new export table. Note that
27660 @command{gnatbind} must be called once again since the binder generated file
27661 has been deleted during the previous call to @command{gnatlink}.
27666 $ gnatlink api -o api.jnk api.exp -mdll
27667 -Wl,--base-file,api.base
27672 @code{gnatdll} builds the new export table using the new base file and
27673 generates the DLL import library @file{libAPI.a}.
27677 $ dlltool --dllname api.dll --def api.def --base-file api.base \
27678 --output-exp api.exp --output-lib libAPI.a
27683 Finally @code{gnatdll} builds the relocatable DLL using the final export
27689 $ gnatlink api api.exp -o api.dll -mdll
27694 @node Using dlltool
27695 @subsubsection Using @code{dlltool}
27698 @code{dlltool} is the low-level tool used by @code{gnatdll} to build
27699 DLLs and static import libraries. This section summarizes the most
27700 common @code{dlltool} switches. The form of the @code{dlltool} command
27704 $ dlltool [@var{switches}]
27708 @code{dlltool} switches include:
27711 @item --base-file @var{basefile}
27712 @cindex @option{--base-file} (@command{dlltool})
27713 Read the base file @var{basefile} generated by the linker. This switch
27714 is used to create a relocatable DLL.
27716 @item --def @var{deffile}
27717 @cindex @option{--def} (@command{dlltool})
27718 Read the definition file.
27720 @item --dllname @var{name}
27721 @cindex @option{--dllname} (@command{dlltool})
27722 Gives the name of the DLL. This switch is used to embed the name of the
27723 DLL in the static import library generated by @code{dlltool} with switch
27724 @option{--output-lib}.
27727 @cindex @option{-k} (@command{dlltool})
27728 Kill @code{@@}@i{nn} from exported names
27729 (@pxref{Windows Calling Conventions}
27730 for a discussion about @code{Stdcall}-style symbols.
27733 @cindex @option{--help} (@command{dlltool})
27734 Prints the @code{dlltool} switches with a concise description.
27736 @item --output-exp @var{exportfile}
27737 @cindex @option{--output-exp} (@command{dlltool})
27738 Generate an export file @var{exportfile}. The export file contains the
27739 export table (list of symbols in the DLL) and is used to create the DLL.
27741 @item --output-lib @i{libfile}
27742 @cindex @option{--output-lib} (@command{dlltool})
27743 Generate a static import library @var{libfile}.
27746 @cindex @option{-v} (@command{dlltool})
27749 @item --as @i{assembler-name}
27750 @cindex @option{--as} (@command{dlltool})
27751 Use @i{assembler-name} as the assembler. The default is @code{as}.
27754 @node GNAT and Windows Resources
27755 @section GNAT and Windows Resources
27756 @cindex Resources, windows
27759 * Building Resources::
27760 * Compiling Resources::
27761 * Using Resources::
27765 Resources are an easy way to add Windows specific objects to your
27766 application. The objects that can be added as resources include:
27795 This section explains how to build, compile and use resources.
27797 @node Building Resources
27798 @subsection Building Resources
27799 @cindex Resources, building
27802 A resource file is an ASCII file. By convention resource files have an
27803 @file{.rc} extension.
27804 The easiest way to build a resource file is to use Microsoft tools
27805 such as @code{imagedit.exe} to build bitmaps, icons and cursors and
27806 @code{dlgedit.exe} to build dialogs.
27807 It is always possible to build an @file{.rc} file yourself by writing a
27810 It is not our objective to explain how to write a resource file. A
27811 complete description of the resource script language can be found in the
27812 Microsoft documentation.
27814 @node Compiling Resources
27815 @subsection Compiling Resources
27818 @cindex Resources, compiling
27821 This section describes how to build a GNAT-compatible (COFF) object file
27822 containing the resources. This is done using the Resource Compiler
27823 @code{windres} as follows:
27826 $ windres -i myres.rc -o myres.o
27830 By default @code{windres} will run @command{gcc} to preprocess the @file{.rc}
27831 file. You can specify an alternate preprocessor (usually named
27832 @file{cpp.exe}) using the @code{windres} @option{--preprocessor}
27833 parameter. A list of all possible options may be obtained by entering
27834 the command @code{windres} @option{--help}.
27836 It is also possible to use the Microsoft resource compiler @code{rc.exe}
27837 to produce a @file{.res} file (binary resource file). See the
27838 corresponding Microsoft documentation for further details. In this case
27839 you need to use @code{windres} to translate the @file{.res} file to a
27840 GNAT-compatible object file as follows:
27843 $ windres -i myres.res -o myres.o
27846 @node Using Resources
27847 @subsection Using Resources
27848 @cindex Resources, using
27851 To include the resource file in your program just add the
27852 GNAT-compatible object file for the resource(s) to the linker
27853 arguments. With @command{gnatmake} this is done by using the @option{-largs}
27857 $ gnatmake myprog -largs myres.o
27860 @node Debugging a DLL
27861 @section Debugging a DLL
27862 @cindex DLL debugging
27865 * Program and DLL Both Built with GCC/GNAT::
27866 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
27870 Debugging a DLL is similar to debugging a standard program. But
27871 we have to deal with two different executable parts: the DLL and the
27872 program that uses it. We have the following four possibilities:
27876 The program and the DLL are built with @code{GCC/GNAT}.
27878 The program is built with foreign tools and the DLL is built with
27881 The program is built with @code{GCC/GNAT} and the DLL is built with
27887 In this section we address only cases one and two above.
27888 There is no point in trying to debug
27889 a DLL with @code{GNU/GDB}, if there is no GDB-compatible debugging
27890 information in it. To do so you must use a debugger compatible with the
27891 tools suite used to build the DLL.
27893 @node Program and DLL Both Built with GCC/GNAT
27894 @subsection Program and DLL Both Built with GCC/GNAT
27897 This is the simplest case. Both the DLL and the program have @code{GDB}
27898 compatible debugging information. It is then possible to break anywhere in
27899 the process. Let's suppose here that the main procedure is named
27900 @code{ada_main} and that in the DLL there is an entry point named
27904 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) and
27905 program must have been built with the debugging information (see GNAT -g
27906 switch). Here are the step-by-step instructions for debugging it:
27909 @item Launch @code{GDB} on the main program.
27915 @item Break on the main procedure and run the program.
27918 (gdb) break ada_main
27923 This step is required to be able to set a breakpoint inside the DLL. As long
27924 as the program is not run, the DLL is not loaded. This has the
27925 consequence that the DLL debugging information is also not loaded, so it is not
27926 possible to set a breakpoint in the DLL.
27928 @item Set a breakpoint inside the DLL
27931 (gdb) break ada_dll
27938 At this stage a breakpoint is set inside the DLL. From there on
27939 you can use the standard approach to debug the whole program
27940 (@pxref{Running and Debugging Ada Programs}).
27942 To break on the @code{DllMain} routine it is not possible to follow
27943 the procedure above. At the time the program stop on @code{ada_main}
27944 the @code{DllMain} routine as already been called. Either you can use
27945 the procedure below @pxref{Debugging the DLL Directly} or this procedure:
27948 @item Launch @code{GDB} on the main program.
27954 @item Load DLL symbols
27957 (gdb) add-sym api.dll
27960 @item Set a breakpoint inside the DLL
27963 (gdb) break ada_dll.adb:45
27966 Note that at this point it is not possible to break using the routine symbol
27967 directly as the program is not yet running. The solution is to break
27968 on the proper line (break in @file{ada_dll.adb} line 45).
27970 @item Start the program
27978 @node Program Built with Foreign Tools and DLL Built with GCC/GNAT
27979 @subsection Program Built with Foreign Tools and DLL Built with GCC/GNAT
27982 * Debugging the DLL Directly::
27983 * Attaching to a Running Process::
27987 In this case things are slightly more complex because it is not possible to
27988 start the main program and then break at the beginning to load the DLL and the
27989 associated DLL debugging information. It is not possible to break at the
27990 beginning of the program because there is no @code{GDB} debugging information,
27991 and therefore there is no direct way of getting initial control. This
27992 section addresses this issue by describing some methods that can be used
27993 to break somewhere in the DLL to debug it.
27996 First suppose that the main procedure is named @code{main} (this is for
27997 example some C code built with Microsoft Visual C) and that there is a
27998 DLL named @code{test.dll} containing an Ada entry point named
28002 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) must have
28003 been built with debugging information (see GNAT -g option).
28005 @node Debugging the DLL Directly
28006 @subsubsection Debugging the DLL Directly
28010 Launch the debugger on the DLL.
28016 @item Set a breakpoint on a DLL subroutine.
28019 (gdb) break ada_dll.adb:45
28022 Note that at this point it is not possible to break using the routine symbol
28023 directly as the program is not yet running. The solution is to break
28024 on the proper line (break in @file{ada_dll.adb} line 45).
28027 Specify the executable file to @code{GDB}.
28030 (gdb) exec-file main.exe
28041 This will run the program until it reaches the breakpoint that has been
28042 set. From that point you can use the standard way to debug a program
28043 as described in (@pxref{Running and Debugging Ada Programs}).
28048 It is also possible to debug the DLL by attaching to a running process.
28050 @node Attaching to a Running Process
28051 @subsubsection Attaching to a Running Process
28052 @cindex DLL debugging, attach to process
28055 With @code{GDB} it is always possible to debug a running process by
28056 attaching to it. It is possible to debug a DLL this way. The limitation
28057 of this approach is that the DLL must run long enough to perform the
28058 attach operation. It may be useful for instance to insert a time wasting
28059 loop in the code of the DLL to meet this criterion.
28063 @item Launch the main program @file{main.exe}.
28069 @item Use the Windows @i{Task Manager} to find the process ID. Let's say
28070 that the process PID for @file{main.exe} is 208.
28078 @item Attach to the running process to be debugged.
28084 @item Load the process debugging information.
28087 (gdb) symbol-file main.exe
28090 @item Break somewhere in the DLL.
28093 (gdb) break ada_dll
28096 @item Continue process execution.
28105 This last step will resume the process execution, and stop at
28106 the breakpoint we have set. From there you can use the standard
28107 approach to debug a program as described in
28108 (@pxref{Running and Debugging Ada Programs}).
28110 @node GNAT and COM/DCOM Objects
28111 @section GNAT and COM/DCOM Objects
28116 This section is temporarily left blank.
28120 @c **********************************
28121 @c * GNU Free Documentation License *
28122 @c **********************************
28124 @c GNU Free Documentation License
28126 @node Index,,GNU Free Documentation License, Top
28132 @c Put table of contents at end, otherwise it precedes the "title page" in
28133 @c the .txt version
28134 @c Edit the pdf file to move the contents to the beginning, after the title