1 \input texinfo @c -*-texinfo-*-
4 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
6 @c GNAT DOCUMENTATION o
10 @c GNAT is maintained by Ada Core Technologies Inc (http://www.gnat.com). o
12 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
14 @setfilename gnat_ugn.info
17 Copyright @copyright{} 1995-2009 Free Software Foundation,
20 Permission is granted to copy, distribute and/or modify this document
21 under the terms of the GNU Free Documentation License, Version 1.3 or
22 any later version published by the Free Software Foundation; with no
23 Invariant Sections, with no Front-Cover Texts and with no Back-Cover
24 Texts. A copy of the license is included in the section entitled
25 ``GNU Free Documentation License''.
28 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
30 @c GNAT_UGN Style Guide
32 @c 1. Always put a @noindent on the line before the first paragraph
33 @c after any of these commands:
45 @c 2. DO NOT use @example. Use @smallexample instead.
46 @c a) DO NOT use highlighting commands (@b{}, @i{}) inside an @smallexample
47 @c context. These can interfere with the readability of the texi
48 @c source file. Instead, use one of the following annotated
49 @c @smallexample commands, and preprocess the texi file with the
50 @c ada2texi tool (which generates appropriate highlighting):
51 @c @smallexample @c ada
52 @c @smallexample @c adanocomment
53 @c @smallexample @c projectfile
54 @c b) The "@c ada" markup will result in boldface for reserved words
55 @c and italics for comments
56 @c c) The "@c adanocomment" markup will result only in boldface for
57 @c reserved words (comments are left alone)
58 @c d) The "@c projectfile" markup is like "@c ada" except that the set
59 @c of reserved words include the new reserved words for project files
61 @c 3. Each @chapter, @section, @subsection, @subsubsection, etc.
62 @c command must be preceded by two empty lines
64 @c 4. The @item command should be on a line of its own if it is in an
65 @c @itemize or @enumerate command.
67 @c 5. When talking about ALI files use "ALI" (all uppercase), not "Ali"
70 @c 6. DO NOT put trailing spaces at the end of a line. Such spaces will
71 @c cause the document build to fail.
73 @c 7. DO NOT use @cartouche for examples that are longer than around 10 lines.
74 @c This command inhibits page breaks, so long examples in a @cartouche can
75 @c lead to large, ugly patches of empty space on a page.
77 @c NOTE: This file should be submitted to xgnatugn with either the vms flag
78 @c or the unw flag set. The unw flag covers topics for both Unix and
81 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
84 @c This flag is used where the text refers to conditions that exist when the
85 @c text was entered into the document but which may change over time.
86 @c Update the setting for the flag, and (if necessary) the text surrounding,
87 @c the references to the flag, on future doc revisions:
88 @c search for @value{NOW}.
92 @set DEFAULTLANGUAGEVERSION Ada 2005
93 @set NONDEFAULTLANGUAGEVERSION Ada 95
100 @set PLATFORM OpenVMS
105 @c The ARG is an optional argument. To be used for macro arguments in
106 @c their documentation (@defmac).
108 @r{[}@var{\varname\}@r{]}@c
110 @c Status as of November 2009:
111 @c Unfortunately texi2pdf and texi2html treat the trailing "@c"
112 @c differently, and faulty output is produced by one or the other
113 @c depending on whether the "@c" is present or absent.
114 @c As a result, the @ovar macro is not used, and all invocations
115 @c of the @ovar macro have been expanded inline.
118 @settitle @value{EDITION} User's Guide @value{PLATFORM}
119 @dircategory GNU Ada tools
121 * @value{EDITION} User's Guide: (gnat_ugn). @value{PLATFORM}
124 @include gcc-common.texi
126 @setchapternewpage odd
131 @title @value{EDITION} User's Guide
135 @titlefont{@i{@value{PLATFORM}}}
141 @subtitle GNAT, The GNU Ada Compiler
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 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 * Stack Related Facilities::
195 * Verifying Properties Using gnatcheck::
196 * Creating Sample Bodies Using gnatstub::
197 * Generating Ada Bindings for C and C++ headers::
198 * Other Utility Programs::
199 * Running and Debugging Ada Programs::
201 * Code Coverage and Profiling::
204 * Compatibility with HP Ada::
206 * Platform-Specific Information for the Run-Time Libraries::
207 * Example of Binder Output File::
208 * Elaboration Order Handling in GNAT::
209 * Conditional Compilation::
211 * Compatibility and Porting Guide::
213 * Microsoft Windows Topics::
215 * GNU Free Documentation License::
218 --- The Detailed Node Listing ---
222 * What This Guide Contains::
223 * What You Should Know before Reading This Guide::
224 * Related Information::
227 Getting Started with GNAT
230 * Running a Simple Ada Program::
231 * Running a Program with Multiple Units::
232 * Using the gnatmake Utility::
234 * Editing with Emacs::
237 * Introduction to GPS::
240 The GNAT Compilation Model
242 * Source Representation::
243 * Foreign Language Representation::
244 * File Naming Rules::
245 * Using Other File Names::
246 * Alternative File Naming Schemes::
247 * Generating Object Files::
248 * Source Dependencies::
249 * The Ada Library Information Files::
250 * Binding an Ada Program::
251 * Mixed Language Programming::
253 * Building Mixed Ada & C++ Programs::
254 * Comparison between GNAT and C/C++ Compilation Models::
256 * Comparison between GNAT and Conventional Ada Library Models::
258 * Placement of temporary files::
261 Foreign Language Representation
264 * Other 8-Bit Codes::
265 * Wide Character Encodings::
267 Compiling Ada Programs With gcc
269 * Compiling Programs::
271 * Search Paths and the Run-Time Library (RTL)::
272 * Order of Compilation Issues::
277 * Output and Error Message Control::
278 * Warning Message Control::
279 * Debugging and Assertion Control::
280 * Validity Checking::
283 * Using gcc for Syntax Checking::
284 * Using gcc for Semantic Checking::
285 * Compiling Different Versions of Ada::
286 * Character Set Control::
287 * File Naming Control::
288 * Subprogram Inlining Control::
289 * Auxiliary Output Control::
290 * Debugging Control::
291 * Exception Handling Control::
292 * Units to Sources Mapping Files::
293 * Integrated Preprocessing::
298 Binding Ada Programs With gnatbind
301 * Switches for gnatbind::
302 * Command-Line Access::
303 * Search Paths for gnatbind::
304 * Examples of gnatbind Usage::
306 Switches for gnatbind
308 * Consistency-Checking Modes::
309 * Binder Error Message Control::
310 * Elaboration Control::
312 * Binding with Non-Ada Main Programs::
313 * Binding Programs with No Main Subprogram::
315 Linking Using gnatlink
318 * Switches for gnatlink::
320 The GNAT Make Program gnatmake
323 * Switches for gnatmake::
324 * Mode Switches for gnatmake::
325 * Notes on the Command Line::
326 * How gnatmake Works::
327 * Examples of gnatmake Usage::
329 Improving Performance
330 * Performance Considerations::
331 * Text_IO Suggestions::
332 * Reducing Size of Ada Executables with gnatelim::
333 * Reducing Size of Executables with unused subprogram/data elimination::
335 Performance Considerations
336 * Controlling Run-Time Checks::
337 * Use of Restrictions::
338 * Optimization Levels::
339 * Debugging Optimized Code::
340 * Inlining of Subprograms::
341 * Other Optimization Switches::
342 * Optimization and Strict Aliasing::
344 * Coverage Analysis::
347 Reducing Size of Ada Executables with gnatelim
350 * Processing Precompiled Libraries::
351 * Correcting the List of Eliminate Pragmas::
352 * Making Your Executables Smaller::
353 * Summary of the gnatelim Usage Cycle::
355 Reducing Size of Executables with unused subprogram/data elimination
356 * About unused subprogram/data elimination::
357 * Compilation options::
359 Renaming Files Using gnatchop
361 * Handling Files with Multiple Units::
362 * Operating gnatchop in Compilation Mode::
363 * Command Line for gnatchop::
364 * Switches for gnatchop::
365 * Examples of gnatchop Usage::
367 Configuration Pragmas
369 * Handling of Configuration Pragmas::
370 * The Configuration Pragmas Files::
372 Handling Arbitrary File Naming Conventions Using gnatname
374 * Arbitrary File Naming Conventions::
376 * Switches for gnatname::
377 * Examples of gnatname Usage::
382 * Examples of Project Files::
383 * Project File Syntax::
384 * Objects and Sources in Project Files::
385 * Importing Projects::
386 * Project Extension::
387 * Project Hierarchy Extension::
388 * External References in Project Files::
389 * Packages in Project Files::
390 * Variables from Imported Projects::
393 * Stand-alone Library Projects::
394 * Switches Related to Project Files::
395 * Tools Supporting Project Files::
396 * An Extended Example::
397 * Project File Complete Syntax::
399 The Cross-Referencing Tools gnatxref and gnatfind
401 * Switches for gnatxref::
402 * Switches for gnatfind::
403 * Project Files for gnatxref and gnatfind::
404 * Regular Expressions in gnatfind and gnatxref::
405 * Examples of gnatxref Usage::
406 * Examples of gnatfind Usage::
408 The GNAT Pretty-Printer gnatpp
410 * Switches for gnatpp::
413 The GNAT Metrics Tool gnatmetric
415 * Switches for gnatmetric::
417 File Name Krunching Using gnatkr
422 * Examples of gnatkr Usage::
424 Preprocessing Using gnatprep
425 * Preprocessing Symbols::
427 * Switches for gnatprep::
428 * Form of Definitions File::
429 * Form of Input Text for gnatprep::
432 The GNAT Run-Time Library Builder gnatlbr
435 * Switches for gnatlbr::
436 * Examples of gnatlbr Usage::
439 The GNAT Library Browser gnatls
442 * Switches for gnatls::
443 * Examples of gnatls Usage::
445 Cleaning Up Using gnatclean
447 * Running gnatclean::
448 * Switches for gnatclean::
449 @c * Examples of gnatclean Usage::
455 * Introduction to Libraries in GNAT::
456 * General Ada Libraries::
457 * Stand-alone Ada Libraries::
458 * Rebuilding the GNAT Run-Time Library::
460 Using the GNU make Utility
462 * Using gnatmake in a Makefile::
463 * Automatically Creating a List of Directories::
464 * Generating the Command Line Switches::
465 * Overcoming Command Line Length Limits::
468 Memory Management Issues
470 * Some Useful Memory Pools::
471 * The GNAT Debug Pool Facility::
476 Stack Related Facilities
478 * Stack Overflow Checking::
479 * Static Stack Usage Analysis::
480 * Dynamic Stack Usage Analysis::
482 Some Useful Memory Pools
484 The GNAT Debug Pool Facility
490 * Switches for gnatmem::
491 * Example of gnatmem Usage::
494 Verifying Properties Using gnatcheck
496 * Format of the Report File::
497 * General gnatcheck Switches::
498 * gnatcheck Rule Options::
499 * Adding the Results of Compiler Checks to gnatcheck Output::
500 * Project-Wide Checks::
503 * Example of gnatcheck Usage::
505 Sample Bodies Using gnatstub
508 * Switches for gnatstub::
510 Other Utility Programs
512 * Using Other Utility Programs with GNAT::
513 * The External Symbol Naming Scheme of GNAT::
514 * Converting Ada Files to html with gnathtml::
517 Code Coverage and Profiling
519 * Code Coverage of Ada Programs using gcov::
520 * Profiling an Ada Program using gprof::
523 Running and Debugging Ada Programs
525 * The GNAT Debugger GDB::
527 * Introduction to GDB Commands::
528 * Using Ada Expressions::
529 * Calling User-Defined Subprograms::
530 * Using the Next Command in a Function::
533 * Debugging Generic Units::
534 * Remote Debugging using gdbserver::
535 * GNAT Abnormal Termination or Failure to Terminate::
536 * Naming Conventions for GNAT Source Files::
537 * Getting Internal Debugging Information::
545 Compatibility with HP Ada
547 * Ada Language Compatibility::
548 * Differences in the Definition of Package System::
549 * Language-Related Features::
550 * The Package STANDARD::
551 * The Package SYSTEM::
552 * Tasking and Task-Related Features::
553 * Pragmas and Pragma-Related Features::
554 * Library of Predefined Units::
556 * Main Program Definition::
557 * Implementation-Defined Attributes::
558 * Compiler and Run-Time Interfacing::
559 * Program Compilation and Library Management::
561 * Implementation Limits::
562 * Tools and Utilities::
564 Language-Related Features
566 * Integer Types and Representations::
567 * Floating-Point Types and Representations::
568 * Pragmas Float_Representation and Long_Float::
569 * Fixed-Point Types and Representations::
570 * Record and Array Component Alignment::
572 * Other Representation Clauses::
574 Tasking and Task-Related Features
576 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
577 * Assigning Task IDs::
578 * Task IDs and Delays::
579 * Task-Related Pragmas::
580 * Scheduling and Task Priority::
582 * External Interrupts::
584 Pragmas and Pragma-Related Features
586 * Restrictions on the Pragma INLINE::
587 * Restrictions on the Pragma INTERFACE::
588 * Restrictions on the Pragma SYSTEM_NAME::
590 Library of Predefined Units
592 * Changes to DECLIB::
596 * Shared Libraries and Options Files::
600 Platform-Specific Information for the Run-Time Libraries
602 * Summary of Run-Time Configurations::
603 * Specifying a Run-Time Library::
604 * Choosing the Scheduling Policy::
605 * Solaris-Specific Considerations::
606 * Linux-Specific Considerations::
607 * AIX-Specific Considerations::
608 * Irix-Specific Considerations::
609 * RTX-Specific Considerations::
610 * HP-UX-Specific Considerations::
612 Example of Binder Output File
614 Elaboration Order Handling in GNAT
617 * Checking the Elaboration Order::
618 * Controlling the Elaboration Order::
619 * Controlling Elaboration in GNAT - Internal Calls::
620 * Controlling Elaboration in GNAT - External Calls::
621 * Default Behavior in GNAT - Ensuring Safety::
622 * Treatment of Pragma Elaborate::
623 * Elaboration Issues for Library Tasks::
624 * Mixing Elaboration Models::
625 * What to Do If the Default Elaboration Behavior Fails::
626 * Elaboration for Access-to-Subprogram Values::
627 * Summary of Procedures for Elaboration Control::
628 * Other Elaboration Order Considerations::
630 Conditional Compilation
631 * Use of Boolean Constants::
632 * Debugging - A Special Case::
633 * Conditionalizing Declarations::
634 * Use of Alternative Implementations::
639 * Basic Assembler Syntax::
640 * A Simple Example of Inline Assembler::
641 * Output Variables in Inline Assembler::
642 * Input Variables in Inline Assembler::
643 * Inlining Inline Assembler Code::
644 * Other Asm Functionality::
646 Compatibility and Porting Guide
648 * Compatibility with Ada 83::
649 * Compatibility between Ada 95 and Ada 2005::
650 * Implementation-dependent characteristics::
652 @c This brief section is only in the non-VMS version
653 @c The complete chapter on HP Ada issues is in the VMS version
654 * Compatibility with HP Ada 83::
656 * Compatibility with Other Ada Systems::
657 * Representation Clauses::
659 * Transitioning to 64-Bit GNAT for OpenVMS::
663 Microsoft Windows Topics
665 * Using GNAT on Windows::
666 * CONSOLE and WINDOWS subsystems::
668 * Mixed-Language Programming on Windows::
669 * Windows Calling Conventions::
670 * Introduction to Dynamic Link Libraries (DLLs)::
671 * Using DLLs with GNAT::
672 * Building DLLs with GNAT::
673 * GNAT and Windows Resources::
675 * Setting Stack Size from gnatlink::
676 * Setting Heap Size from gnatlink::
683 @node About This Guide
684 @unnumbered About This Guide
688 This guide describes the use of @value{EDITION},
689 a compiler and software development toolset for the full Ada
690 programming language, implemented on OpenVMS for HP's Alpha and
691 Integrity server (I64) platforms.
694 This guide describes the use of @value{EDITION},
695 a compiler and software development
696 toolset for the full Ada programming language.
698 It documents the features of the compiler and tools, and explains
699 how to use them to build Ada applications.
701 @value{EDITION} implements Ada 95 and Ada 2005, and it may also be invoked in
702 Ada 83 compatibility mode.
703 By default, @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
704 but you can override with a compiler switch
705 (@pxref{Compiling Different Versions of Ada})
706 to explicitly specify the language version.
707 Throughout this manual, references to ``Ada'' without a year suffix
708 apply to both the Ada 95 and Ada 2005 versions of the language.
712 For ease of exposition, ``@value{EDITION}'' will be referred to simply as
713 ``GNAT'' in the remainder of this document.
720 * What This Guide Contains::
721 * What You Should Know before Reading This Guide::
722 * Related Information::
726 @node What This Guide Contains
727 @unnumberedsec What This Guide Contains
730 This guide contains the following chapters:
734 @ref{Getting Started with GNAT}, describes how to get started compiling
735 and running Ada programs with the GNAT Ada programming environment.
737 @ref{The GNAT Compilation Model}, describes the compilation model used
741 @ref{Compiling Using gcc}, describes how to compile
742 Ada programs with @command{gcc}, the Ada compiler.
745 @ref{Binding Using gnatbind}, describes how to
746 perform binding of Ada programs with @code{gnatbind}, the GNAT binding
750 @ref{Linking Using gnatlink},
751 describes @command{gnatlink}, a
752 program that provides for linking using the GNAT run-time library to
753 construct a program. @command{gnatlink} can also incorporate foreign language
754 object units into the executable.
757 @ref{The GNAT Make Program gnatmake}, describes @command{gnatmake}, a
758 utility that automatically determines the set of sources
759 needed by an Ada compilation unit, and executes the necessary compilations
763 @ref{Improving Performance}, shows various techniques for making your
764 Ada program run faster or take less space.
765 It discusses the effect of the compiler's optimization switch and
766 also describes the @command{gnatelim} tool and unused subprogram/data
770 @ref{Renaming Files Using gnatchop}, describes
771 @code{gnatchop}, a utility that allows you to preprocess a file that
772 contains Ada source code, and split it into one or more new files, one
773 for each compilation unit.
776 @ref{Configuration Pragmas}, describes the configuration pragmas
780 @ref{Handling Arbitrary File Naming Conventions Using gnatname},
781 shows how to override the default GNAT file naming conventions,
782 either for an individual unit or globally.
785 @ref{GNAT Project Manager}, describes how to use project files
786 to organize large projects.
789 @ref{The Cross-Referencing Tools gnatxref and gnatfind}, discusses
790 @code{gnatxref} and @code{gnatfind}, two tools that provide an easy
791 way to navigate through sources.
794 @ref{The GNAT Pretty-Printer gnatpp}, shows how to produce a reformatted
795 version of an Ada source file with control over casing, indentation,
796 comment placement, and other elements of program presentation style.
799 @ref{The GNAT Metric Tool gnatmetric}, shows how to compute various
800 metrics for an Ada source file, such as the number of types and subprograms,
801 and assorted complexity measures.
804 @ref{File Name Krunching Using gnatkr}, describes the @code{gnatkr}
805 file name krunching utility, used to handle shortened
806 file names on operating systems with a limit on the length of names.
809 @ref{Preprocessing Using gnatprep}, describes @code{gnatprep}, a
810 preprocessor utility that allows a single source file to be used to
811 generate multiple or parameterized source files by means of macro
816 @ref{The GNAT Run-Time Library Builder gnatlbr}, describes @command{gnatlbr},
817 a tool for rebuilding the GNAT run time with user-supplied
818 configuration pragmas.
822 @ref{The GNAT Library Browser gnatls}, describes @code{gnatls}, a
823 utility that displays information about compiled units, including dependences
824 on the corresponding sources files, and consistency of compilations.
827 @ref{Cleaning Up Using gnatclean}, describes @code{gnatclean}, a utility
828 to delete files that are produced by the compiler, binder and linker.
832 @ref{GNAT and Libraries}, describes the process of creating and using
833 Libraries with GNAT. It also describes how to recompile the GNAT run-time
837 @ref{Using the GNU make Utility}, describes some techniques for using
838 the GNAT toolset in Makefiles.
842 @ref{Memory Management Issues}, describes some useful predefined storage pools
843 and in particular the GNAT Debug Pool facility, which helps detect incorrect
846 It also describes @command{gnatmem}, a utility that monitors dynamic
847 allocation and deallocation and helps detect ``memory leaks''.
851 @ref{Stack Related Facilities}, describes some useful tools associated with
852 stack checking and analysis.
855 @ref{Verifying Properties Using gnatcheck}, discusses @code{gnatcheck},
856 a utility that checks Ada code against a set of rules.
859 @ref{Creating Sample Bodies Using gnatstub}, discusses @code{gnatstub},
860 a utility that generates empty but compilable bodies for library units.
863 @ref{Generating Ada Bindings for C and C++ headers}, describes how to
864 generate automatically Ada bindings from C and C++ headers.
867 @ref{Other Utility Programs}, discusses several other GNAT utilities,
868 including @code{gnathtml}.
872 @ref{Code Coverage and Profiling}, describes how to perform a structural
873 coverage and profile the execution of Ada programs.
877 @ref{Running and Debugging Ada Programs}, describes how to run and debug
882 @ref{Compatibility with HP Ada}, details the compatibility of GNAT with
883 HP Ada 83 @footnote{``HP Ada'' refers to the legacy product originally
884 developed by Digital Equipment Corporation and currently supported by HP.}
885 for OpenVMS Alpha. This product was formerly known as DEC Ada,
888 historical compatibility reasons, the relevant libraries still use the
893 @ref{Platform-Specific Information for the Run-Time Libraries},
894 describes the various run-time
895 libraries supported by GNAT on various platforms and explains how to
896 choose a particular library.
899 @ref{Example of Binder Output File}, shows the source code for the binder
900 output file for a sample program.
903 @ref{Elaboration Order Handling in GNAT}, describes how GNAT helps
904 you deal with elaboration order issues.
907 @ref{Conditional Compilation}, describes how to model conditional compilation,
908 both with Ada in general and with GNAT facilities in particular.
911 @ref{Inline Assembler}, shows how to use the inline assembly facility
915 @ref{Compatibility and Porting Guide}, contains sections on compatibility
916 of GNAT with other Ada development environments (including Ada 83 systems),
917 to assist in porting code from those environments.
921 @ref{Microsoft Windows Topics}, presents information relevant to the
922 Microsoft Windows platform.
926 @c *************************************************
927 @node What You Should Know before Reading This Guide
928 @c *************************************************
929 @unnumberedsec What You Should Know before Reading This Guide
931 @cindex Ada 95 Language Reference Manual
932 @cindex Ada 2005 Language Reference Manual
934 This guide assumes a basic familiarity with the Ada 95 language, as
935 described in the International Standard ANSI/ISO/IEC-8652:1995, January
937 It does not require knowledge of the new features introduced by Ada 2005,
938 (officially known as ISO/IEC 8652:1995 with Technical Corrigendum 1
940 Both reference manuals are included in the GNAT documentation
943 @node Related Information
944 @unnumberedsec Related Information
947 For further information about related tools, refer to the following
952 @xref{Top, GNAT Reference Manual, About This Guide, gnat_rm, GNAT
953 Reference Manual}, which contains all reference material for the GNAT
954 implementation of Ada.
958 @cite{Using the GNAT Programming Studio}, which describes the GPS
959 Integrated Development Environment.
962 @cite{GNAT Programming Studio Tutorial}, which introduces the
963 main GPS features through examples.
967 @cite{Ada 95 Reference Manual}, which contains reference
968 material for the Ada 95 programming language.
971 @cite{Ada 2005 Reference Manual}, which contains reference
972 material for the Ada 2005 programming language.
975 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
977 in the GNU:[DOCS] directory,
979 for all details on the use of the GNU source-level debugger.
982 @xref{Top,, The extensible self-documenting text editor, emacs,
985 located in the GNU:[DOCS] directory if the EMACS kit is installed,
987 for full information on the extensible editor and programming
994 @unnumberedsec Conventions
996 @cindex Typographical conventions
999 Following are examples of the typographical and graphic conventions used
1004 @code{Functions}, @command{utility program names}, @code{standard names},
1008 @option{Option flags}
1011 @file{File names}, @samp{button names}, and @samp{field names}.
1014 @code{Variables}, @env{environment variables}, and @var{metasyntactic
1021 @r{[}optional information or parameters@r{]}
1024 Examples are described by text
1026 and then shown this way.
1031 Commands that are entered by the user are preceded in this manual by the
1032 characters @w{``@code{$ }''} (dollar sign followed by space). If your system
1033 uses this sequence as a prompt, then the commands will appear exactly as
1034 you see them in the manual. If your system uses some other prompt, then
1035 the command will appear with the @code{$} replaced by whatever prompt
1036 character you are using.
1039 Full file names are shown with the ``@code{/}'' character
1040 as the directory separator; e.g., @file{parent-dir/subdir/myfile.adb}.
1041 If you are using GNAT on a Windows platform, please note that
1042 the ``@code{\}'' character should be used instead.
1045 @c ****************************
1046 @node Getting Started with GNAT
1047 @chapter Getting Started with GNAT
1050 This chapter describes some simple ways of using GNAT to build
1051 executable Ada programs.
1053 @ref{Running GNAT}, through @ref{Using the gnatmake Utility},
1054 show how to use the command line environment.
1055 @ref{Introduction to GPS}, provides a brief
1056 introduction to the GNAT Programming Studio, a visually-oriented
1057 Integrated Development Environment for GNAT.
1058 GPS offers a graphical ``look and feel'', support for development in
1059 other programming languages, comprehensive browsing features, and
1060 many other capabilities.
1061 For information on GPS please refer to
1062 @cite{Using the GNAT Programming Studio}.
1067 * Running a Simple Ada Program::
1068 * Running a Program with Multiple Units::
1069 * Using the gnatmake Utility::
1071 * Editing with Emacs::
1074 * Introduction to GPS::
1079 @section Running GNAT
1082 Three steps are needed to create an executable file from an Ada source
1087 The source file(s) must be compiled.
1089 The file(s) must be bound using the GNAT binder.
1091 All appropriate object files must be linked to produce an executable.
1095 All three steps are most commonly handled by using the @command{gnatmake}
1096 utility program that, given the name of the main program, automatically
1097 performs the necessary compilation, binding and linking steps.
1099 @node Running a Simple Ada Program
1100 @section Running a Simple Ada Program
1103 Any text editor may be used to prepare an Ada program.
1105 used, the optional Ada mode may be helpful in laying out the program.)
1107 program text is a normal text file. We will assume in our initial
1108 example that you have used your editor to prepare the following
1109 standard format text file:
1111 @smallexample @c ada
1113 with Ada.Text_IO; use Ada.Text_IO;
1116 Put_Line ("Hello WORLD!");
1122 This file should be named @file{hello.adb}.
1123 With the normal default file naming conventions, GNAT requires
1125 contain a single compilation unit whose file name is the
1127 with periods replaced by hyphens; the
1128 extension is @file{ads} for a
1129 spec and @file{adb} for a body.
1130 You can override this default file naming convention by use of the
1131 special pragma @code{Source_File_Name} (@pxref{Using Other File Names}).
1132 Alternatively, if you want to rename your files according to this default
1133 convention, which is probably more convenient if you will be using GNAT
1134 for all your compilations, then the @code{gnatchop} utility
1135 can be used to generate correctly-named source files
1136 (@pxref{Renaming Files Using gnatchop}).
1138 You can compile the program using the following command (@code{$} is used
1139 as the command prompt in the examples in this document):
1146 @command{gcc} is the command used to run the compiler. This compiler is
1147 capable of compiling programs in several languages, including Ada and
1148 C. It assumes that you have given it an Ada program if the file extension is
1149 either @file{.ads} or @file{.adb}, and it will then call
1150 the GNAT compiler to compile the specified file.
1153 The @option{-c} switch is required. It tells @command{gcc} to only do a
1154 compilation. (For C programs, @command{gcc} can also do linking, but this
1155 capability is not used directly for Ada programs, so the @option{-c}
1156 switch must always be present.)
1159 This compile command generates a file
1160 @file{hello.o}, which is the object
1161 file corresponding to your Ada program. It also generates
1162 an ``Ada Library Information'' file @file{hello.ali},
1163 which contains additional information used to check
1164 that an Ada program is consistent.
1165 To build an executable file,
1166 use @code{gnatbind} to bind the program
1167 and @command{gnatlink} to link it. The
1168 argument to both @code{gnatbind} and @command{gnatlink} is the name of the
1169 @file{ALI} file, but the default extension of @file{.ali} can
1170 be omitted. This means that in the most common case, the argument
1171 is simply the name of the main program:
1179 A simpler method of carrying out these steps is to use
1181 a master program that invokes all the required
1182 compilation, binding and linking tools in the correct order. In particular,
1183 @command{gnatmake} automatically recompiles any sources that have been
1184 modified since they were last compiled, or sources that depend
1185 on such modified sources, so that ``version skew'' is avoided.
1186 @cindex Version skew (avoided by @command{gnatmake})
1189 $ gnatmake hello.adb
1193 The result is an executable program called @file{hello}, which can be
1201 assuming that the current directory is on the search path
1202 for executable programs.
1205 and, if all has gone well, you will see
1212 appear in response to this command.
1214 @c ****************************************
1215 @node Running a Program with Multiple Units
1216 @section Running a Program with Multiple Units
1219 Consider a slightly more complicated example that has three files: a
1220 main program, and the spec and body of a package:
1222 @smallexample @c ada
1225 package Greetings is
1230 with Ada.Text_IO; use Ada.Text_IO;
1231 package body Greetings is
1234 Put_Line ("Hello WORLD!");
1237 procedure Goodbye is
1239 Put_Line ("Goodbye WORLD!");
1256 Following the one-unit-per-file rule, place this program in the
1257 following three separate files:
1261 spec of package @code{Greetings}
1264 body of package @code{Greetings}
1267 body of main program
1271 To build an executable version of
1272 this program, we could use four separate steps to compile, bind, and link
1273 the program, as follows:
1277 $ gcc -c greetings.adb
1283 Note that there is no required order of compilation when using GNAT.
1284 In particular it is perfectly fine to compile the main program first.
1285 Also, it is not necessary to compile package specs in the case where
1286 there is an accompanying body; you only need to compile the body. If you want
1287 to submit these files to the compiler for semantic checking and not code
1288 generation, then use the
1289 @option{-gnatc} switch:
1292 $ gcc -c greetings.ads -gnatc
1296 Although the compilation can be done in separate steps as in the
1297 above example, in practice it is almost always more convenient
1298 to use the @command{gnatmake} tool. All you need to know in this case
1299 is the name of the main program's source file. The effect of the above four
1300 commands can be achieved with a single one:
1303 $ gnatmake gmain.adb
1307 In the next section we discuss the advantages of using @command{gnatmake} in
1310 @c *****************************
1311 @node Using the gnatmake Utility
1312 @section Using the @command{gnatmake} Utility
1315 If you work on a program by compiling single components at a time using
1316 @command{gcc}, you typically keep track of the units you modify. In order to
1317 build a consistent system, you compile not only these units, but also any
1318 units that depend on the units you have modified.
1319 For example, in the preceding case,
1320 if you edit @file{gmain.adb}, you only need to recompile that file. But if
1321 you edit @file{greetings.ads}, you must recompile both
1322 @file{greetings.adb} and @file{gmain.adb}, because both files contain
1323 units that depend on @file{greetings.ads}.
1325 @code{gnatbind} will warn you if you forget one of these compilation
1326 steps, so that it is impossible to generate an inconsistent program as a
1327 result of forgetting to do a compilation. Nevertheless it is tedious and
1328 error-prone to keep track of dependencies among units.
1329 One approach to handle the dependency-bookkeeping is to use a
1330 makefile. However, makefiles present maintenance problems of their own:
1331 if the dependencies change as you change the program, you must make
1332 sure that the makefile is kept up-to-date manually, which is also an
1333 error-prone process.
1335 The @command{gnatmake} utility takes care of these details automatically.
1336 Invoke it using either one of the following forms:
1339 $ gnatmake gmain.adb
1340 $ gnatmake ^gmain^GMAIN^
1344 The argument is the name of the file containing the main program;
1345 you may omit the extension. @command{gnatmake}
1346 examines the environment, automatically recompiles any files that need
1347 recompiling, and binds and links the resulting set of object files,
1348 generating the executable file, @file{^gmain^GMAIN.EXE^}.
1349 In a large program, it
1350 can be extremely helpful to use @command{gnatmake}, because working out by hand
1351 what needs to be recompiled can be difficult.
1353 Note that @command{gnatmake}
1354 takes into account all the Ada rules that
1355 establish dependencies among units. These include dependencies that result
1356 from inlining subprogram bodies, and from
1357 generic instantiation. Unlike some other
1358 Ada make tools, @command{gnatmake} does not rely on the dependencies that were
1359 found by the compiler on a previous compilation, which may possibly
1360 be wrong when sources change. @command{gnatmake} determines the exact set of
1361 dependencies from scratch each time it is run.
1364 @node Editing with Emacs
1365 @section Editing with Emacs
1369 Emacs is an extensible self-documenting text editor that is available in a
1370 separate VMSINSTAL kit.
1372 Invoke Emacs by typing @kbd{Emacs} at the command prompt. To get started,
1373 click on the Emacs Help menu and run the Emacs Tutorial.
1374 In a character cell terminal, Emacs help is invoked with @kbd{Ctrl-h} (also
1375 written as @kbd{C-h}), and the tutorial by @kbd{C-h t}.
1377 Documentation on Emacs and other tools is available in Emacs under the
1378 pull-down menu button: @code{Help - Info}. After selecting @code{Info},
1379 use the middle mouse button to select a topic (e.g.@: Emacs).
1381 In a character cell terminal, do @kbd{C-h i} to invoke info, and then @kbd{m}
1382 (stands for menu) followed by the menu item desired, as in @kbd{m Emacs}, to
1383 get to the Emacs manual.
1384 Help on Emacs is also available by typing @kbd{HELP EMACS} at the DCL command
1387 The tutorial is highly recommended in order to learn the intricacies of Emacs,
1388 which is sufficiently extensible to provide for a complete programming
1389 environment and shell for the sophisticated user.
1393 @node Introduction to GPS
1394 @section Introduction to GPS
1395 @cindex GPS (GNAT Programming Studio)
1396 @cindex GNAT Programming Studio (GPS)
1398 Although the command line interface (@command{gnatmake}, etc.) alone
1399 is sufficient, a graphical Interactive Development
1400 Environment can make it easier for you to compose, navigate, and debug
1401 programs. This section describes the main features of GPS
1402 (``GNAT Programming Studio''), the GNAT graphical IDE.
1403 You will see how to use GPS to build and debug an executable, and
1404 you will also learn some of the basics of the GNAT ``project'' facility.
1406 GPS enables you to do much more than is presented here;
1407 e.g., you can produce a call graph, interface to a third-party
1408 Version Control System, and inspect the generated assembly language
1410 Indeed, GPS also supports languages other than Ada.
1411 Such additional information, and an explanation of all of the GPS menu
1412 items. may be found in the on-line help, which includes
1413 a user's guide and a tutorial (these are also accessible from the GNAT
1417 * Building a New Program with GPS::
1418 * Simple Debugging with GPS::
1421 @node Building a New Program with GPS
1422 @subsection Building a New Program with GPS
1424 GPS invokes the GNAT compilation tools using information
1425 contained in a @emph{project} (also known as a @emph{project file}):
1426 a collection of properties such
1427 as source directories, identities of main subprograms, tool switches, etc.,
1428 and their associated values.
1429 See @ref{GNAT Project Manager} for details.
1430 In order to run GPS, you will need to either create a new project
1431 or else open an existing one.
1433 This section will explain how you can use GPS to create a project,
1434 to associate Ada source files with a project, and to build and run
1438 @item @emph{Creating a project}
1440 Invoke GPS, either from the command line or the platform's IDE.
1441 After it starts, GPS will display a ``Welcome'' screen with three
1446 @code{Start with default project in directory}
1449 @code{Create new project with wizard}
1452 @code{Open existing project}
1456 Select @code{Create new project with wizard} and press @code{OK}.
1457 A new window will appear. In the text box labeled with
1458 @code{Enter the name of the project to create}, type @file{sample}
1459 as the project name.
1460 In the next box, browse to choose the directory in which you
1461 would like to create the project file.
1462 After selecting an appropriate directory, press @code{Forward}.
1464 A window will appear with the title
1465 @code{Version Control System Configuration}.
1466 Simply press @code{Forward}.
1468 A window will appear with the title
1469 @code{Please select the source directories for this project}.
1470 The directory that you specified for the project file will be selected
1471 by default as the one to use for sources; simply press @code{Forward}.
1473 A window will appear with the title
1474 @code{Please select the build directory for this project}.
1475 The directory that you specified for the project file will be selected
1476 by default for object files and executables;
1477 simply press @code{Forward}.
1479 A window will appear with the title
1480 @code{Please select the main units for this project}.
1481 You will supply this information later, after creating the source file.
1482 Simply press @code{Forward} for now.
1484 A window will appear with the title
1485 @code{Please select the switches to build the project}.
1486 Press @code{Apply}. This will create a project file named
1487 @file{sample.prj} in the directory that you had specified.
1489 @item @emph{Creating and saving the source file}
1491 After you create the new project, a GPS window will appear, which is
1492 partitioned into two main sections:
1496 A @emph{Workspace area}, initially greyed out, which you will use for
1497 creating and editing source files
1500 Directly below, a @emph{Messages area}, which initially displays a
1501 ``Welcome'' message.
1502 (If the Messages area is not visible, drag its border upward to expand it.)
1506 Select @code{File} on the menu bar, and then the @code{New} command.
1507 The Workspace area will become white, and you can now
1508 enter the source program explicitly.
1509 Type the following text
1511 @smallexample @c ada
1513 with Ada.Text_IO; use Ada.Text_IO;
1516 Put_Line("Hello from GPS!");
1522 Select @code{File}, then @code{Save As}, and enter the source file name
1524 The file will be saved in the same directory you specified as the
1525 location of the default project file.
1527 @item @emph{Updating the project file}
1529 You need to add the new source file to the project.
1531 the @code{Project} menu and then @code{Edit project properties}.
1532 Click the @code{Main files} tab on the left, and then the
1534 Choose @file{hello.adb} from the list, and press @code{Open}.
1535 The project settings window will reflect this action.
1538 @item @emph{Building and running the program}
1540 In the main GPS window, now choose the @code{Build} menu, then @code{Make},
1541 and select @file{hello.adb}.
1542 The Messages window will display the resulting invocations of @command{gcc},
1543 @command{gnatbind}, and @command{gnatlink}
1544 (reflecting the default switch settings from the
1545 project file that you created) and then a ``successful compilation/build''
1548 To run the program, choose the @code{Build} menu, then @code{Run}, and
1549 select @command{hello}.
1550 An @emph{Arguments Selection} window will appear.
1551 There are no command line arguments, so just click @code{OK}.
1553 The Messages window will now display the program's output (the string
1554 @code{Hello from GPS}), and at the bottom of the GPS window a status
1555 update is displayed (@code{Run: hello}).
1556 Close the GPS window (or select @code{File}, then @code{Exit}) to
1557 terminate this GPS session.
1560 @node Simple Debugging with GPS
1561 @subsection Simple Debugging with GPS
1563 This section illustrates basic debugging techniques (setting breakpoints,
1564 examining/modifying variables, single stepping).
1567 @item @emph{Opening a project}
1569 Start GPS and select @code{Open existing project}; browse to
1570 specify the project file @file{sample.prj} that you had created in the
1573 @item @emph{Creating a source file}
1575 Select @code{File}, then @code{New}, and type in the following program:
1577 @smallexample @c ada
1579 with Ada.Text_IO; use Ada.Text_IO;
1580 procedure Example is
1581 Line : String (1..80);
1584 Put_Line("Type a line of text at each prompt; an empty line to exit");
1588 Put_Line (Line (1..N) );
1596 Select @code{File}, then @code{Save as}, and enter the file name
1599 @item @emph{Updating the project file}
1601 Add @code{Example} as a new main unit for the project:
1604 Select @code{Project}, then @code{Edit Project Properties}.
1607 Select the @code{Main files} tab, click @code{Add}, then
1608 select the file @file{example.adb} from the list, and
1610 You will see the file name appear in the list of main units
1616 @item @emph{Building/running the executable}
1618 To build the executable
1619 select @code{Build}, then @code{Make}, and then choose @file{example.adb}.
1621 Run the program to see its effect (in the Messages area).
1622 Each line that you enter is displayed; an empty line will
1623 cause the loop to exit and the program to terminate.
1625 @item @emph{Debugging the program}
1627 Note that the @option{-g} switches to @command{gcc} and @command{gnatlink},
1628 which are required for debugging, are on by default when you create
1630 Thus unless you intentionally remove these settings, you will be able
1631 to debug any program that you develop using GPS.
1634 @item @emph{Initializing}
1636 Select @code{Debug}, then @code{Initialize}, then @file{example}
1638 @item @emph{Setting a breakpoint}
1640 After performing the initialization step, you will observe a small
1641 icon to the right of each line number.
1642 This serves as a toggle for breakpoints; clicking the icon will
1643 set a breakpoint at the corresponding line (the icon will change to
1644 a red circle with an ``x''), and clicking it again
1645 will remove the breakpoint / reset the icon.
1647 For purposes of this example, set a breakpoint at line 10 (the
1648 statement @code{Put_Line@ (Line@ (1..N));}
1650 @item @emph{Starting program execution}
1652 Select @code{Debug}, then @code{Run}. When the
1653 @code{Program Arguments} window appears, click @code{OK}.
1654 A console window will appear; enter some line of text,
1655 e.g.@: @code{abcde}, at the prompt.
1656 The program will pause execution when it gets to the
1657 breakpoint, and the corresponding line is highlighted.
1659 @item @emph{Examining a variable}
1661 Move the mouse over one of the occurrences of the variable @code{N}.
1662 You will see the value (5) displayed, in ``tool tip'' fashion.
1663 Right click on @code{N}, select @code{Debug}, then select @code{Display N}.
1664 You will see information about @code{N} appear in the @code{Debugger Data}
1665 pane, showing the value as 5.
1667 @item @emph{Assigning a new value to a variable}
1669 Right click on the @code{N} in the @code{Debugger Data} pane, and
1670 select @code{Set value of N}.
1671 When the input window appears, enter the value @code{4} and click
1673 This value does not automatically appear in the @code{Debugger Data}
1674 pane; to see it, right click again on the @code{N} in the
1675 @code{Debugger Data} pane and select @code{Update value}.
1676 The new value, 4, will appear in red.
1678 @item @emph{Single stepping}
1680 Select @code{Debug}, then @code{Next}.
1681 This will cause the next statement to be executed, in this case the
1682 call of @code{Put_Line} with the string slice.
1683 Notice in the console window that the displayed string is simply
1684 @code{abcd} and not @code{abcde} which you had entered.
1685 This is because the upper bound of the slice is now 4 rather than 5.
1687 @item @emph{Removing a breakpoint}
1689 Toggle the breakpoint icon at line 10.
1691 @item @emph{Resuming execution from a breakpoint}
1693 Select @code{Debug}, then @code{Continue}.
1694 The program will reach the next iteration of the loop, and
1695 wait for input after displaying the prompt.
1696 This time, just hit the @kbd{Enter} key.
1697 The value of @code{N} will be 0, and the program will terminate.
1698 The console window will disappear.
1703 @node The GNAT Compilation Model
1704 @chapter The GNAT Compilation Model
1705 @cindex GNAT compilation model
1706 @cindex Compilation model
1709 * Source Representation::
1710 * Foreign Language Representation::
1711 * File Naming Rules::
1712 * Using Other File Names::
1713 * Alternative File Naming Schemes::
1714 * Generating Object Files::
1715 * Source Dependencies::
1716 * The Ada Library Information Files::
1717 * Binding an Ada Program::
1718 * Mixed Language Programming::
1720 * Building Mixed Ada & C++ Programs::
1721 * Comparison between GNAT and C/C++ Compilation Models::
1723 * Comparison between GNAT and Conventional Ada Library Models::
1725 * Placement of temporary files::
1730 This chapter describes the compilation model used by GNAT. Although
1731 similar to that used by other languages, such as C and C++, this model
1732 is substantially different from the traditional Ada compilation models,
1733 which are based on a library. The model is initially described without
1734 reference to the library-based model. If you have not previously used an
1735 Ada compiler, you need only read the first part of this chapter. The
1736 last section describes and discusses the differences between the GNAT
1737 model and the traditional Ada compiler models. If you have used other
1738 Ada compilers, this section will help you to understand those
1739 differences, and the advantages of the GNAT model.
1741 @node Source Representation
1742 @section Source Representation
1746 Ada source programs are represented in standard text files, using
1747 Latin-1 coding. Latin-1 is an 8-bit code that includes the familiar
1748 7-bit ASCII set, plus additional characters used for
1749 representing foreign languages (@pxref{Foreign Language Representation}
1750 for support of non-USA character sets). The format effector characters
1751 are represented using their standard ASCII encodings, as follows:
1756 Vertical tab, @code{16#0B#}
1760 Horizontal tab, @code{16#09#}
1764 Carriage return, @code{16#0D#}
1768 Line feed, @code{16#0A#}
1772 Form feed, @code{16#0C#}
1776 Source files are in standard text file format. In addition, GNAT will
1777 recognize a wide variety of stream formats, in which the end of
1778 physical lines is marked by any of the following sequences:
1779 @code{LF}, @code{CR}, @code{CR-LF}, or @code{LF-CR}. This is useful
1780 in accommodating files that are imported from other operating systems.
1782 @cindex End of source file
1783 @cindex Source file, end
1785 The end of a source file is normally represented by the physical end of
1786 file. However, the control character @code{16#1A#} (@code{SUB}) is also
1787 recognized as signalling the end of the source file. Again, this is
1788 provided for compatibility with other operating systems where this
1789 code is used to represent the end of file.
1791 Each file contains a single Ada compilation unit, including any pragmas
1792 associated with the unit. For example, this means you must place a
1793 package declaration (a package @dfn{spec}) and the corresponding body in
1794 separate files. An Ada @dfn{compilation} (which is a sequence of
1795 compilation units) is represented using a sequence of files. Similarly,
1796 you will place each subunit or child unit in a separate file.
1798 @node Foreign Language Representation
1799 @section Foreign Language Representation
1802 GNAT supports the standard character sets defined in Ada as well as
1803 several other non-standard character sets for use in localized versions
1804 of the compiler (@pxref{Character Set Control}).
1807 * Other 8-Bit Codes::
1808 * Wide Character Encodings::
1816 The basic character set is Latin-1. This character set is defined by ISO
1817 standard 8859, part 1. The lower half (character codes @code{16#00#}
1818 @dots{} @code{16#7F#)} is identical to standard ASCII coding, but the upper half
1819 is used to represent additional characters. These include extended letters
1820 used by European languages, such as French accents, the vowels with umlauts
1821 used in German, and the extra letter A-ring used in Swedish.
1823 @findex Ada.Characters.Latin_1
1824 For a complete list of Latin-1 codes and their encodings, see the source
1825 file of library unit @code{Ada.Characters.Latin_1} in file
1826 @file{a-chlat1.ads}.
1827 You may use any of these extended characters freely in character or
1828 string literals. In addition, the extended characters that represent
1829 letters can be used in identifiers.
1831 @node Other 8-Bit Codes
1832 @subsection Other 8-Bit Codes
1835 GNAT also supports several other 8-bit coding schemes:
1838 @item ISO 8859-2 (Latin-2)
1841 Latin-2 letters allowed in identifiers, with uppercase and lowercase
1844 @item ISO 8859-3 (Latin-3)
1847 Latin-3 letters allowed in identifiers, with uppercase and lowercase
1850 @item ISO 8859-4 (Latin-4)
1853 Latin-4 letters allowed in identifiers, with uppercase and lowercase
1856 @item ISO 8859-5 (Cyrillic)
1859 ISO 8859-5 letters (Cyrillic) allowed in identifiers, with uppercase and
1860 lowercase equivalence.
1862 @item ISO 8859-15 (Latin-9)
1865 ISO 8859-15 (Latin-9) letters allowed in identifiers, with uppercase and
1866 lowercase equivalence
1868 @item IBM PC (code page 437)
1869 @cindex code page 437
1870 This code page is the normal default for PCs in the U.S. It corresponds
1871 to the original IBM PC character set. This set has some, but not all, of
1872 the extended Latin-1 letters, but these letters do not have the same
1873 encoding as Latin-1. In this mode, these letters are allowed in
1874 identifiers with uppercase and lowercase equivalence.
1876 @item IBM PC (code page 850)
1877 @cindex code page 850
1878 This code page is a modification of 437 extended to include all the
1879 Latin-1 letters, but still not with the usual Latin-1 encoding. In this
1880 mode, all these letters are allowed in identifiers with uppercase and
1881 lowercase equivalence.
1883 @item Full Upper 8-bit
1884 Any character in the range 80-FF allowed in identifiers, and all are
1885 considered distinct. In other words, there are no uppercase and lowercase
1886 equivalences in this range. This is useful in conjunction with
1887 certain encoding schemes used for some foreign character sets (e.g.,
1888 the typical method of representing Chinese characters on the PC).
1891 No upper-half characters in the range 80-FF are allowed in identifiers.
1892 This gives Ada 83 compatibility for identifier names.
1896 For precise data on the encodings permitted, and the uppercase and lowercase
1897 equivalences that are recognized, see the file @file{csets.adb} in
1898 the GNAT compiler sources. You will need to obtain a full source release
1899 of GNAT to obtain this file.
1901 @node Wide Character Encodings
1902 @subsection Wide Character Encodings
1905 GNAT allows wide character codes to appear in character and string
1906 literals, and also optionally in identifiers, by means of the following
1907 possible encoding schemes:
1912 In this encoding, a wide character is represented by the following five
1920 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1921 characters (using uppercase letters) of the wide character code. For
1922 example, ESC A345 is used to represent the wide character with code
1924 This scheme is compatible with use of the full Wide_Character set.
1926 @item Upper-Half Coding
1927 @cindex Upper-Half Coding
1928 The wide character with encoding @code{16#abcd#} where the upper bit is on
1929 (in other words, ``a'' is in the range 8-F) is represented as two bytes,
1930 @code{16#ab#} and @code{16#cd#}. The second byte cannot be a format control
1931 character, but is not required to be in the upper half. This method can
1932 be also used for shift-JIS or EUC, where the internal coding matches the
1935 @item Shift JIS Coding
1936 @cindex Shift JIS Coding
1937 A wide character is represented by a two-character sequence,
1939 @code{16#cd#}, with the restrictions described for upper-half encoding as
1940 described above. The internal character code is the corresponding JIS
1941 character according to the standard algorithm for Shift-JIS
1942 conversion. Only characters defined in the JIS code set table can be
1943 used with this encoding method.
1947 A wide character is represented by a two-character sequence
1949 @code{16#cd#}, with both characters being in the upper half. The internal
1950 character code is the corresponding JIS character according to the EUC
1951 encoding algorithm. Only characters defined in the JIS code set table
1952 can be used with this encoding method.
1955 A wide character is represented using
1956 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
1957 10646-1/Am.2. Depending on the character value, the representation
1958 is a one, two, or three byte sequence:
1963 16#0000#-16#007f#: 2#0@var{xxxxxxx}#
1964 16#0080#-16#07ff#: 2#110@var{xxxxx}# 2#10@var{xxxxxx}#
1965 16#0800#-16#ffff#: 2#1110@var{xxxx}# 2#10@var{xxxxxx}# 2#10@var{xxxxxx}#
1970 where the @var{xxx} bits correspond to the left-padded bits of the
1971 16-bit character value. Note that all lower half ASCII characters
1972 are represented as ASCII bytes and all upper half characters and
1973 other wide characters are represented as sequences of upper-half
1974 (The full UTF-8 scheme allows for encoding 31-bit characters as
1975 6-byte sequences, but in this implementation, all UTF-8 sequences
1976 of four or more bytes length will be treated as illegal).
1977 @item Brackets Coding
1978 In this encoding, a wide character is represented by the following eight
1986 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1987 characters (using uppercase letters) of the wide character code. For
1988 example, [``A345''] is used to represent the wide character with code
1989 @code{16#A345#}. It is also possible (though not required) to use the
1990 Brackets coding for upper half characters. For example, the code
1991 @code{16#A3#} can be represented as @code{[``A3'']}.
1993 This scheme is compatible with use of the full Wide_Character set,
1994 and is also the method used for wide character encoding in the standard
1995 ACVC (Ada Compiler Validation Capability) test suite distributions.
2000 Note: Some of these coding schemes do not permit the full use of the
2001 Ada character set. For example, neither Shift JIS, nor EUC allow the
2002 use of the upper half of the Latin-1 set.
2004 @node File Naming Rules
2005 @section File Naming Rules
2008 The default file name is determined by the name of the unit that the
2009 file contains. The name is formed by taking the full expanded name of
2010 the unit and replacing the separating dots with hyphens and using
2011 ^lowercase^uppercase^ for all letters.
2013 An exception arises if the file name generated by the above rules starts
2014 with one of the characters
2016 @samp{A}, @samp{G}, @samp{I}, or @samp{S},
2019 @samp{a}, @samp{g}, @samp{i}, or @samp{s},
2021 and the second character is a
2022 minus. In this case, the character ^tilde^dollar sign^ is used in place
2023 of the minus. The reason for this special rule is to avoid clashes with
2024 the standard names for child units of the packages System, Ada,
2025 Interfaces, and GNAT, which use the prefixes
2027 @samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},
2030 @samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},
2034 The file extension is @file{.ads} for a spec and
2035 @file{.adb} for a body. The following list shows some
2036 examples of these rules.
2043 @item arith_functions.ads
2044 Arith_Functions (package spec)
2045 @item arith_functions.adb
2046 Arith_Functions (package body)
2048 Func.Spec (child package spec)
2050 Func.Spec (child package body)
2052 Sub (subunit of Main)
2053 @item ^a~bad.adb^A$BAD.ADB^
2054 A.Bad (child package body)
2058 Following these rules can result in excessively long
2059 file names if corresponding
2060 unit names are long (for example, if child units or subunits are
2061 heavily nested). An option is available to shorten such long file names
2062 (called file name ``krunching''). This may be particularly useful when
2063 programs being developed with GNAT are to be used on operating systems
2064 with limited file name lengths. @xref{Using gnatkr}.
2066 Of course, no file shortening algorithm can guarantee uniqueness over
2067 all possible unit names; if file name krunching is used, it is your
2068 responsibility to ensure no name clashes occur. Alternatively you
2069 can specify the exact file names that you want used, as described
2070 in the next section. Finally, if your Ada programs are migrating from a
2071 compiler with a different naming convention, you can use the gnatchop
2072 utility to produce source files that follow the GNAT naming conventions.
2073 (For details @pxref{Renaming Files Using gnatchop}.)
2075 Note: in the case of @code{Windows NT/XP} or @code{OpenVMS} operating
2076 systems, case is not significant. So for example on @code{Windows XP}
2077 if the canonical name is @code{main-sub.adb}, you can use the file name
2078 @code{Main-Sub.adb} instead. However, case is significant for other
2079 operating systems, so for example, if you want to use other than
2080 canonically cased file names on a Unix system, you need to follow
2081 the procedures described in the next section.
2083 @node Using Other File Names
2084 @section Using Other File Names
2088 In the previous section, we have described the default rules used by
2089 GNAT to determine the file name in which a given unit resides. It is
2090 often convenient to follow these default rules, and if you follow them,
2091 the compiler knows without being explicitly told where to find all
2094 However, in some cases, particularly when a program is imported from
2095 another Ada compiler environment, it may be more convenient for the
2096 programmer to specify which file names contain which units. GNAT allows
2097 arbitrary file names to be used by means of the Source_File_Name pragma.
2098 The form of this pragma is as shown in the following examples:
2099 @cindex Source_File_Name pragma
2101 @smallexample @c ada
2103 pragma Source_File_Name (My_Utilities.Stacks,
2104 Spec_File_Name => "myutilst_a.ada");
2105 pragma Source_File_name (My_Utilities.Stacks,
2106 Body_File_Name => "myutilst.ada");
2111 As shown in this example, the first argument for the pragma is the unit
2112 name (in this example a child unit). The second argument has the form
2113 of a named association. The identifier
2114 indicates whether the file name is for a spec or a body;
2115 the file name itself is given by a string literal.
2117 The source file name pragma is a configuration pragma, which means that
2118 normally it will be placed in the @file{gnat.adc}
2119 file used to hold configuration
2120 pragmas that apply to a complete compilation environment.
2121 For more details on how the @file{gnat.adc} file is created and used
2122 see @ref{Handling of Configuration Pragmas}.
2123 @cindex @file{gnat.adc}
2126 GNAT allows completely arbitrary file names to be specified using the
2127 source file name pragma. However, if the file name specified has an
2128 extension other than @file{.ads} or @file{.adb} it is necessary to use
2129 a special syntax when compiling the file. The name in this case must be
2130 preceded by the special sequence @option{-x} followed by a space and the name
2131 of the language, here @code{ada}, as in:
2134 $ gcc -c -x ada peculiar_file_name.sim
2139 @command{gnatmake} handles non-standard file names in the usual manner (the
2140 non-standard file name for the main program is simply used as the
2141 argument to gnatmake). Note that if the extension is also non-standard,
2142 then it must be included in the @command{gnatmake} command, it may not
2145 @node Alternative File Naming Schemes
2146 @section Alternative File Naming Schemes
2147 @cindex File naming schemes, alternative
2150 In the previous section, we described the use of the @code{Source_File_Name}
2151 pragma to allow arbitrary names to be assigned to individual source files.
2152 However, this approach requires one pragma for each file, and especially in
2153 large systems can result in very long @file{gnat.adc} files, and also create
2154 a maintenance problem.
2156 GNAT also provides a facility for specifying systematic file naming schemes
2157 other than the standard default naming scheme previously described. An
2158 alternative scheme for naming is specified by the use of
2159 @code{Source_File_Name} pragmas having the following format:
2160 @cindex Source_File_Name pragma
2162 @smallexample @c ada
2163 pragma Source_File_Name (
2164 Spec_File_Name => FILE_NAME_PATTERN
2165 @r{[},Casing => CASING_SPEC@r{]}
2166 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2168 pragma Source_File_Name (
2169 Body_File_Name => FILE_NAME_PATTERN
2170 @r{[},Casing => CASING_SPEC@r{]}
2171 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2173 pragma Source_File_Name (
2174 Subunit_File_Name => FILE_NAME_PATTERN
2175 @r{[},Casing => CASING_SPEC@r{]}
2176 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2178 FILE_NAME_PATTERN ::= STRING_LITERAL
2179 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
2183 The @code{FILE_NAME_PATTERN} string shows how the file name is constructed.
2184 It contains a single asterisk character, and the unit name is substituted
2185 systematically for this asterisk. The optional parameter
2186 @code{Casing} indicates
2187 whether the unit name is to be all upper-case letters, all lower-case letters,
2188 or mixed-case. If no
2189 @code{Casing} parameter is used, then the default is all
2190 ^lower-case^upper-case^.
2192 The optional @code{Dot_Replacement} string is used to replace any periods
2193 that occur in subunit or child unit names. If no @code{Dot_Replacement}
2194 argument is used then separating dots appear unchanged in the resulting
2196 Although the above syntax indicates that the
2197 @code{Casing} argument must appear
2198 before the @code{Dot_Replacement} argument, but it
2199 is also permissible to write these arguments in the opposite order.
2201 As indicated, it is possible to specify different naming schemes for
2202 bodies, specs, and subunits. Quite often the rule for subunits is the
2203 same as the rule for bodies, in which case, there is no need to give
2204 a separate @code{Subunit_File_Name} rule, and in this case the
2205 @code{Body_File_name} rule is used for subunits as well.
2207 The separate rule for subunits can also be used to implement the rather
2208 unusual case of a compilation environment (e.g.@: a single directory) which
2209 contains a subunit and a child unit with the same unit name. Although
2210 both units cannot appear in the same partition, the Ada Reference Manual
2211 allows (but does not require) the possibility of the two units coexisting
2212 in the same environment.
2214 The file name translation works in the following steps:
2219 If there is a specific @code{Source_File_Name} pragma for the given unit,
2220 then this is always used, and any general pattern rules are ignored.
2223 If there is a pattern type @code{Source_File_Name} pragma that applies to
2224 the unit, then the resulting file name will be used if the file exists. If
2225 more than one pattern matches, the latest one will be tried first, and the
2226 first attempt resulting in a reference to a file that exists will be used.
2229 If no pattern type @code{Source_File_Name} pragma that applies to the unit
2230 for which the corresponding file exists, then the standard GNAT default
2231 naming rules are used.
2236 As an example of the use of this mechanism, consider a commonly used scheme
2237 in which file names are all lower case, with separating periods copied
2238 unchanged to the resulting file name, and specs end with @file{.1.ada}, and
2239 bodies end with @file{.2.ada}. GNAT will follow this scheme if the following
2242 @smallexample @c ada
2243 pragma Source_File_Name
2244 (Spec_File_Name => "*.1.ada");
2245 pragma Source_File_Name
2246 (Body_File_Name => "*.2.ada");
2250 The default GNAT scheme is actually implemented by providing the following
2251 default pragmas internally:
2253 @smallexample @c ada
2254 pragma Source_File_Name
2255 (Spec_File_Name => "*.ads", Dot_Replacement => "-");
2256 pragma Source_File_Name
2257 (Body_File_Name => "*.adb", Dot_Replacement => "-");
2261 Our final example implements a scheme typically used with one of the
2262 Ada 83 compilers, where the separator character for subunits was ``__''
2263 (two underscores), specs were identified by adding @file{_.ADA}, bodies
2264 by adding @file{.ADA}, and subunits by
2265 adding @file{.SEP}. All file names were
2266 upper case. Child units were not present of course since this was an
2267 Ada 83 compiler, but it seems reasonable to extend this scheme to use
2268 the same double underscore separator for child units.
2270 @smallexample @c ada
2271 pragma Source_File_Name
2272 (Spec_File_Name => "*_.ADA",
2273 Dot_Replacement => "__",
2274 Casing = Uppercase);
2275 pragma Source_File_Name
2276 (Body_File_Name => "*.ADA",
2277 Dot_Replacement => "__",
2278 Casing = Uppercase);
2279 pragma Source_File_Name
2280 (Subunit_File_Name => "*.SEP",
2281 Dot_Replacement => "__",
2282 Casing = Uppercase);
2285 @node Generating Object Files
2286 @section Generating Object Files
2289 An Ada program consists of a set of source files, and the first step in
2290 compiling the program is to generate the corresponding object files.
2291 These are generated by compiling a subset of these source files.
2292 The files you need to compile are the following:
2296 If a package spec has no body, compile the package spec to produce the
2297 object file for the package.
2300 If a package has both a spec and a body, compile the body to produce the
2301 object file for the package. The source file for the package spec need
2302 not be compiled in this case because there is only one object file, which
2303 contains the code for both the spec and body of the package.
2306 For a subprogram, compile the subprogram body to produce the object file
2307 for the subprogram. The spec, if one is present, is as usual in a
2308 separate file, and need not be compiled.
2312 In the case of subunits, only compile the parent unit. A single object
2313 file is generated for the entire subunit tree, which includes all the
2317 Compile child units independently of their parent units
2318 (though, of course, the spec of all the ancestor unit must be present in order
2319 to compile a child unit).
2323 Compile generic units in the same manner as any other units. The object
2324 files in this case are small dummy files that contain at most the
2325 flag used for elaboration checking. This is because GNAT always handles generic
2326 instantiation by means of macro expansion. However, it is still necessary to
2327 compile generic units, for dependency checking and elaboration purposes.
2331 The preceding rules describe the set of files that must be compiled to
2332 generate the object files for a program. Each object file has the same
2333 name as the corresponding source file, except that the extension is
2336 You may wish to compile other files for the purpose of checking their
2337 syntactic and semantic correctness. For example, in the case where a
2338 package has a separate spec and body, you would not normally compile the
2339 spec. However, it is convenient in practice to compile the spec to make
2340 sure it is error-free before compiling clients of this spec, because such
2341 compilations will fail if there is an error in the spec.
2343 GNAT provides an option for compiling such files purely for the
2344 purposes of checking correctness; such compilations are not required as
2345 part of the process of building a program. To compile a file in this
2346 checking mode, use the @option{-gnatc} switch.
2348 @node Source Dependencies
2349 @section Source Dependencies
2352 A given object file clearly depends on the source file which is compiled
2353 to produce it. Here we are using @dfn{depends} in the sense of a typical
2354 @code{make} utility; in other words, an object file depends on a source
2355 file if changes to the source file require the object file to be
2357 In addition to this basic dependency, a given object may depend on
2358 additional source files as follows:
2362 If a file being compiled @code{with}'s a unit @var{X}, the object file
2363 depends on the file containing the spec of unit @var{X}. This includes
2364 files that are @code{with}'ed implicitly either because they are parents
2365 of @code{with}'ed child units or they are run-time units required by the
2366 language constructs used in a particular unit.
2369 If a file being compiled instantiates a library level generic unit, the
2370 object file depends on both the spec and body files for this generic
2374 If a file being compiled instantiates a generic unit defined within a
2375 package, the object file depends on the body file for the package as
2376 well as the spec file.
2380 @cindex @option{-gnatn} switch
2381 If a file being compiled contains a call to a subprogram for which
2382 pragma @code{Inline} applies and inlining is activated with the
2383 @option{-gnatn} switch, the object file depends on the file containing the
2384 body of this subprogram as well as on the file containing the spec. Note
2385 that for inlining to actually occur as a result of the use of this switch,
2386 it is necessary to compile in optimizing mode.
2388 @cindex @option{-gnatN} switch
2389 The use of @option{-gnatN} activates inlining optimization
2390 that is performed by the front end of the compiler. This inlining does
2391 not require that the code generation be optimized. Like @option{-gnatn},
2392 the use of this switch generates additional dependencies.
2394 When using a gcc-based back end (in practice this means using any version
2395 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
2396 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
2397 Historically front end inlining was more extensive than the gcc back end
2398 inlining, but that is no longer the case.
2401 If an object file @file{O} depends on the proper body of a subunit through
2402 inlining or instantiation, it depends on the parent unit of the subunit.
2403 This means that any modification of the parent unit or one of its subunits
2404 affects the compilation of @file{O}.
2407 The object file for a parent unit depends on all its subunit body files.
2410 The previous two rules meant that for purposes of computing dependencies and
2411 recompilation, a body and all its subunits are treated as an indivisible whole.
2414 These rules are applied transitively: if unit @code{A} @code{with}'s
2415 unit @code{B}, whose elaboration calls an inlined procedure in package
2416 @code{C}, the object file for unit @code{A} will depend on the body of
2417 @code{C}, in file @file{c.adb}.
2419 The set of dependent files described by these rules includes all the
2420 files on which the unit is semantically dependent, as dictated by the
2421 Ada language standard. However, it is a superset of what the
2422 standard describes, because it includes generic, inline, and subunit
2425 An object file must be recreated by recompiling the corresponding source
2426 file if any of the source files on which it depends are modified. For
2427 example, if the @code{make} utility is used to control compilation,
2428 the rule for an Ada object file must mention all the source files on
2429 which the object file depends, according to the above definition.
2430 The determination of the necessary
2431 recompilations is done automatically when one uses @command{gnatmake}.
2434 @node The Ada Library Information Files
2435 @section The Ada Library Information Files
2436 @cindex Ada Library Information files
2437 @cindex @file{ALI} files
2440 Each compilation actually generates two output files. The first of these
2441 is the normal object file that has a @file{.o} extension. The second is a
2442 text file containing full dependency information. It has the same
2443 name as the source file, but an @file{.ali} extension.
2444 This file is known as the Ada Library Information (@file{ALI}) file.
2445 The following information is contained in the @file{ALI} file.
2449 Version information (indicates which version of GNAT was used to compile
2450 the unit(s) in question)
2453 Main program information (including priority and time slice settings,
2454 as well as the wide character encoding used during compilation).
2457 List of arguments used in the @command{gcc} command for the compilation
2460 Attributes of the unit, including configuration pragmas used, an indication
2461 of whether the compilation was successful, exception model used etc.
2464 A list of relevant restrictions applying to the unit (used for consistency)
2468 Categorization information (e.g.@: use of pragma @code{Pure}).
2471 Information on all @code{with}'ed units, including presence of
2472 @code{Elaborate} or @code{Elaborate_All} pragmas.
2475 Information from any @code{Linker_Options} pragmas used in the unit
2478 Information on the use of @code{Body_Version} or @code{Version}
2479 attributes in the unit.
2482 Dependency information. This is a list of files, together with
2483 time stamp and checksum information. These are files on which
2484 the unit depends in the sense that recompilation is required
2485 if any of these units are modified.
2488 Cross-reference data. Contains information on all entities referenced
2489 in the unit. Used by tools like @code{gnatxref} and @code{gnatfind} to
2490 provide cross-reference information.
2495 For a full detailed description of the format of the @file{ALI} file,
2496 see the source of the body of unit @code{Lib.Writ}, contained in file
2497 @file{lib-writ.adb} in the GNAT compiler sources.
2499 @node Binding an Ada Program
2500 @section Binding an Ada Program
2503 When using languages such as C and C++, once the source files have been
2504 compiled the only remaining step in building an executable program
2505 is linking the object modules together. This means that it is possible to
2506 link an inconsistent version of a program, in which two units have
2507 included different versions of the same header.
2509 The rules of Ada do not permit such an inconsistent program to be built.
2510 For example, if two clients have different versions of the same package,
2511 it is illegal to build a program containing these two clients.
2512 These rules are enforced by the GNAT binder, which also determines an
2513 elaboration order consistent with the Ada rules.
2515 The GNAT binder is run after all the object files for a program have
2516 been created. It is given the name of the main program unit, and from
2517 this it determines the set of units required by the program, by reading the
2518 corresponding ALI files. It generates error messages if the program is
2519 inconsistent or if no valid order of elaboration exists.
2521 If no errors are detected, the binder produces a main program, in Ada by
2522 default, that contains calls to the elaboration procedures of those
2523 compilation unit that require them, followed by
2524 a call to the main program. This Ada program is compiled to generate the
2525 object file for the main program. The name of
2526 the Ada file is @file{b~@var{xxx}.adb} (with the corresponding spec
2527 @file{b~@var{xxx}.ads}) where @var{xxx} is the name of the
2530 Finally, the linker is used to build the resulting executable program,
2531 using the object from the main program from the bind step as well as the
2532 object files for the Ada units of the program.
2534 @node Mixed Language Programming
2535 @section Mixed Language Programming
2536 @cindex Mixed Language Programming
2539 This section describes how to develop a mixed-language program,
2540 specifically one that comprises units in both Ada and C.
2543 * Interfacing to C::
2544 * Calling Conventions::
2547 @node Interfacing to C
2548 @subsection Interfacing to C
2550 Interfacing Ada with a foreign language such as C involves using
2551 compiler directives to import and/or export entity definitions in each
2552 language---using @code{extern} statements in C, for instance, and the
2553 @code{Import}, @code{Export}, and @code{Convention} pragmas in Ada.
2554 A full treatment of these topics is provided in Appendix B, section 1
2555 of the Ada Reference Manual.
2557 There are two ways to build a program using GNAT that contains some Ada
2558 sources and some foreign language sources, depending on whether or not
2559 the main subprogram is written in Ada. Here is a source example with
2560 the main subprogram in Ada:
2566 void print_num (int num)
2568 printf ("num is %d.\n", num);
2574 /* num_from_Ada is declared in my_main.adb */
2575 extern int num_from_Ada;
2579 return num_from_Ada;
2583 @smallexample @c ada
2585 procedure My_Main is
2587 -- Declare then export an Integer entity called num_from_Ada
2588 My_Num : Integer := 10;
2589 pragma Export (C, My_Num, "num_from_Ada");
2591 -- Declare an Ada function spec for Get_Num, then use
2592 -- C function get_num for the implementation.
2593 function Get_Num return Integer;
2594 pragma Import (C, Get_Num, "get_num");
2596 -- Declare an Ada procedure spec for Print_Num, then use
2597 -- C function print_num for the implementation.
2598 procedure Print_Num (Num : Integer);
2599 pragma Import (C, Print_Num, "print_num");
2602 Print_Num (Get_Num);
2608 To build this example, first compile the foreign language files to
2609 generate object files:
2611 ^gcc -c file1.c^gcc -c FILE1.C^
2612 ^gcc -c file2.c^gcc -c FILE2.C^
2616 Then, compile the Ada units to produce a set of object files and ALI
2619 gnatmake ^-c^/ACTIONS=COMPILE^ my_main.adb
2623 Run the Ada binder on the Ada main program:
2625 gnatbind my_main.ali
2629 Link the Ada main program, the Ada objects and the other language
2632 gnatlink my_main.ali file1.o file2.o
2636 The last three steps can be grouped in a single command:
2638 gnatmake my_main.adb -largs file1.o file2.o
2641 @cindex Binder output file
2643 If the main program is in a language other than Ada, then you may have
2644 more than one entry point into the Ada subsystem. You must use a special
2645 binder option to generate callable routines that initialize and
2646 finalize the Ada units (@pxref{Binding with Non-Ada Main Programs}).
2647 Calls to the initialization and finalization routines must be inserted
2648 in the main program, or some other appropriate point in the code. The
2649 call to initialize the Ada units must occur before the first Ada
2650 subprogram is called, and the call to finalize the Ada units must occur
2651 after the last Ada subprogram returns. The binder will place the
2652 initialization and finalization subprograms into the
2653 @file{b~@var{xxx}.adb} file where they can be accessed by your C
2654 sources. To illustrate, we have the following example:
2658 extern void adainit (void);
2659 extern void adafinal (void);
2660 extern int add (int, int);
2661 extern int sub (int, int);
2663 int main (int argc, char *argv[])
2669 /* Should print "21 + 7 = 28" */
2670 printf ("%d + %d = %d\n", a, b, add (a, b));
2671 /* Should print "21 - 7 = 14" */
2672 printf ("%d - %d = %d\n", a, b, sub (a, b));
2678 @smallexample @c ada
2681 function Add (A, B : Integer) return Integer;
2682 pragma Export (C, Add, "add");
2686 package body Unit1 is
2687 function Add (A, B : Integer) return Integer is
2695 function Sub (A, B : Integer) return Integer;
2696 pragma Export (C, Sub, "sub");
2700 package body Unit2 is
2701 function Sub (A, B : Integer) return Integer is
2710 The build procedure for this application is similar to the last
2711 example's. First, compile the foreign language files to generate object
2714 ^gcc -c main.c^gcc -c main.c^
2718 Next, compile the Ada units to produce a set of object files and ALI
2721 gnatmake ^-c^/ACTIONS=COMPILE^ unit1.adb
2722 gnatmake ^-c^/ACTIONS=COMPILE^ unit2.adb
2726 Run the Ada binder on every generated ALI file. Make sure to use the
2727 @option{-n} option to specify a foreign main program:
2729 gnatbind ^-n^/NOMAIN^ unit1.ali unit2.ali
2733 Link the Ada main program, the Ada objects and the foreign language
2734 objects. You need only list the last ALI file here:
2736 gnatlink unit2.ali main.o -o exec_file
2739 This procedure yields a binary executable called @file{exec_file}.
2743 Depending on the circumstances (for example when your non-Ada main object
2744 does not provide symbol @code{main}), you may also need to instruct the
2745 GNAT linker not to include the standard startup objects by passing the
2746 @option{^-nostartfiles^/NOSTART_FILES^} switch to @command{gnatlink}.
2748 @node Calling Conventions
2749 @subsection Calling Conventions
2750 @cindex Foreign Languages
2751 @cindex Calling Conventions
2752 GNAT follows standard calling sequence conventions and will thus interface
2753 to any other language that also follows these conventions. The following
2754 Convention identifiers are recognized by GNAT:
2757 @cindex Interfacing to Ada
2758 @cindex Other Ada compilers
2759 @cindex Convention Ada
2761 This indicates that the standard Ada calling sequence will be
2762 used and all Ada data items may be passed without any limitations in the
2763 case where GNAT is used to generate both the caller and callee. It is also
2764 possible to mix GNAT generated code and code generated by another Ada
2765 compiler. In this case, the data types should be restricted to simple
2766 cases, including primitive types. Whether complex data types can be passed
2767 depends on the situation. Probably it is safe to pass simple arrays, such
2768 as arrays of integers or floats. Records may or may not work, depending
2769 on whether both compilers lay them out identically. Complex structures
2770 involving variant records, access parameters, tasks, or protected types,
2771 are unlikely to be able to be passed.
2773 Note that in the case of GNAT running
2774 on a platform that supports HP Ada 83, a higher degree of compatibility
2775 can be guaranteed, and in particular records are layed out in an identical
2776 manner in the two compilers. Note also that if output from two different
2777 compilers is mixed, the program is responsible for dealing with elaboration
2778 issues. Probably the safest approach is to write the main program in the
2779 version of Ada other than GNAT, so that it takes care of its own elaboration
2780 requirements, and then call the GNAT-generated adainit procedure to ensure
2781 elaboration of the GNAT components. Consult the documentation of the other
2782 Ada compiler for further details on elaboration.
2784 However, it is not possible to mix the tasking run time of GNAT and
2785 HP Ada 83, All the tasking operations must either be entirely within
2786 GNAT compiled sections of the program, or entirely within HP Ada 83
2787 compiled sections of the program.
2789 @cindex Interfacing to Assembly
2790 @cindex Convention Assembler
2792 Specifies assembler as the convention. In practice this has the
2793 same effect as convention Ada (but is not equivalent in the sense of being
2794 considered the same convention).
2796 @cindex Convention Asm
2799 Equivalent to Assembler.
2801 @cindex Interfacing to COBOL
2802 @cindex Convention COBOL
2805 Data will be passed according to the conventions described
2806 in section B.4 of the Ada Reference Manual.
2809 @cindex Interfacing to C
2810 @cindex Convention C
2812 Data will be passed according to the conventions described
2813 in section B.3 of the Ada Reference Manual.
2815 A note on interfacing to a C ``varargs'' function:
2816 @findex C varargs function
2817 @cindex Interfacing to C varargs function
2818 @cindex varargs function interfaces
2822 In C, @code{varargs} allows a function to take a variable number of
2823 arguments. There is no direct equivalent in this to Ada. One
2824 approach that can be used is to create a C wrapper for each
2825 different profile and then interface to this C wrapper. For
2826 example, to print an @code{int} value using @code{printf},
2827 create a C function @code{printfi} that takes two arguments, a
2828 pointer to a string and an int, and calls @code{printf}.
2829 Then in the Ada program, use pragma @code{Import} to
2830 interface to @code{printfi}.
2833 It may work on some platforms to directly interface to
2834 a @code{varargs} function by providing a specific Ada profile
2835 for a particular call. However, this does not work on
2836 all platforms, since there is no guarantee that the
2837 calling sequence for a two argument normal C function
2838 is the same as for calling a @code{varargs} C function with
2839 the same two arguments.
2842 @cindex Convention Default
2847 @cindex Convention External
2854 @cindex Interfacing to C++
2855 @cindex Convention C++
2856 @item C_Plus_Plus (or CPP)
2857 This stands for C++. For most purposes this is identical to C.
2858 See the separate description of the specialized GNAT pragmas relating to
2859 C++ interfacing for further details.
2863 @cindex Interfacing to Fortran
2864 @cindex Convention Fortran
2866 Data will be passed according to the conventions described
2867 in section B.5 of the Ada Reference Manual.
2870 This applies to an intrinsic operation, as defined in the Ada
2871 Reference Manual. If a pragma Import (Intrinsic) applies to a subprogram,
2872 this means that the body of the subprogram is provided by the compiler itself,
2873 usually by means of an efficient code sequence, and that the user does not
2874 supply an explicit body for it. In an application program, the pragma may
2875 be applied to the following sets of names:
2879 Rotate_Left, Rotate_Right, Shift_Left, Shift_Right,
2880 Shift_Right_Arithmetic. The corresponding subprogram declaration must have
2881 two formal parameters. The
2882 first one must be a signed integer type or a modular type with a binary
2883 modulus, and the second parameter must be of type Natural.
2884 The return type must be the same as the type of the first argument. The size
2885 of this type can only be 8, 16, 32, or 64.
2888 Binary arithmetic operators: ``+'', ``-'', ``*'', ``/''
2889 The corresponding operator declaration must have parameters and result type
2890 that have the same root numeric type (for example, all three are long_float
2891 types). This simplifies the definition of operations that use type checking
2892 to perform dimensional checks:
2894 @smallexample @c ada
2895 type Distance is new Long_Float;
2896 type Time is new Long_Float;
2897 type Velocity is new Long_Float;
2898 function "/" (D : Distance; T : Time)
2900 pragma Import (Intrinsic, "/");
2904 This common idiom is often programmed with a generic definition and an
2905 explicit body. The pragma makes it simpler to introduce such declarations.
2906 It incurs no overhead in compilation time or code size, because it is
2907 implemented as a single machine instruction.
2910 General subprogram entities, to bind an Ada subprogram declaration to
2911 a compiler builtin by name with back-ends where such interfaces are
2912 available. A typical example is the set of ``__builtin'' functions
2913 exposed by the GCC back-end, as in the following example:
2915 @smallexample @c ada
2916 function builtin_sqrt (F : Float) return Float;
2917 pragma Import (Intrinsic, builtin_sqrt, "__builtin_sqrtf");
2920 Most of the GCC builtins are accessible this way, and as for other
2921 import conventions (e.g. C), it is the user's responsibility to ensure
2922 that the Ada subprogram profile matches the underlying builtin
2930 @cindex Convention Stdcall
2932 This is relevant only to Windows XP/2000/NT implementations of GNAT,
2933 and specifies that the @code{Stdcall} calling sequence will be used,
2934 as defined by the NT API. Nevertheless, to ease building
2935 cross-platform bindings this convention will be handled as a @code{C} calling
2936 convention on non-Windows platforms.
2939 @cindex Convention DLL
2941 This is equivalent to @code{Stdcall}.
2944 @cindex Convention Win32
2946 This is equivalent to @code{Stdcall}.
2950 @cindex Convention Stubbed
2952 This is a special convention that indicates that the compiler
2953 should provide a stub body that raises @code{Program_Error}.
2957 GNAT additionally provides a useful pragma @code{Convention_Identifier}
2958 that can be used to parametrize conventions and allow additional synonyms
2959 to be specified. For example if you have legacy code in which the convention
2960 identifier Fortran77 was used for Fortran, you can use the configuration
2963 @smallexample @c ada
2964 pragma Convention_Identifier (Fortran77, Fortran);
2968 And from now on the identifier Fortran77 may be used as a convention
2969 identifier (for example in an @code{Import} pragma) with the same
2973 @node Building Mixed Ada & C++ Programs
2974 @section Building Mixed Ada and C++ Programs
2977 A programmer inexperienced with mixed-language development may find that
2978 building an application containing both Ada and C++ code can be a
2979 challenge. This section gives a few
2980 hints that should make this task easier. The first section addresses
2981 the differences between interfacing with C and interfacing with C++.
2983 looks into the delicate problem of linking the complete application from
2984 its Ada and C++ parts. The last section gives some hints on how the GNAT
2985 run-time library can be adapted in order to allow inter-language dispatching
2986 with a new C++ compiler.
2989 * Interfacing to C++::
2990 * Linking a Mixed C++ & Ada Program::
2991 * A Simple Example::
2992 * Interfacing with C++ constructors::
2993 * Interfacing with C++ at the Class Level::
2996 @node Interfacing to C++
2997 @subsection Interfacing to C++
3000 GNAT supports interfacing with the G++ compiler (or any C++ compiler
3001 generating code that is compatible with the G++ Application Binary
3002 Interface ---see http://www.codesourcery.com/archives/cxx-abi).
3005 Interfacing can be done at 3 levels: simple data, subprograms, and
3006 classes. In the first two cases, GNAT offers a specific @code{Convention
3007 C_Plus_Plus} (or @code{CPP}) that behaves exactly like @code{Convention C}.
3008 Usually, C++ mangles the names of subprograms. To generate proper mangled
3009 names automatically, see @ref{Generating Ada Bindings for C and C++ headers}).
3010 This problem can also be addressed manually in two ways:
3014 by modifying the C++ code in order to force a C convention using
3015 the @code{extern "C"} syntax.
3018 by figuring out the mangled name (using e.g. @command{nm}) and using it as the
3019 Link_Name argument of the pragma import.
3023 Interfacing at the class level can be achieved by using the GNAT specific
3024 pragmas such as @code{CPP_Constructor}. @xref{Interfacing to C++,,,
3025 gnat_rm, GNAT Reference Manual}, for additional information.
3027 @node Linking a Mixed C++ & Ada Program
3028 @subsection Linking a Mixed C++ & Ada Program
3031 Usually the linker of the C++ development system must be used to link
3032 mixed applications because most C++ systems will resolve elaboration
3033 issues (such as calling constructors on global class instances)
3034 transparently during the link phase. GNAT has been adapted to ease the
3035 use of a foreign linker for the last phase. Three cases can be
3040 Using GNAT and G++ (GNU C++ compiler) from the same GCC installation:
3041 The C++ linker can simply be called by using the C++ specific driver
3044 Note that if the C++ code uses inline functions, you will need to
3045 compile your C++ code with the @code{-fkeep-inline-functions} switch in
3046 order to provide an existing function implementation that the Ada code can
3050 $ g++ -c -fkeep-inline-functions file1.C
3051 $ g++ -c -fkeep-inline-functions file2.C
3052 $ gnatmake ada_unit -largs file1.o file2.o --LINK=g++
3056 Using GNAT and G++ from two different GCC installations: If both
3057 compilers are on the @env{PATH}, the previous method may be used. It is
3058 important to note that environment variables such as
3059 @env{C_INCLUDE_PATH}, @env{GCC_EXEC_PREFIX}, @env{BINUTILS_ROOT}, and
3060 @env{GCC_ROOT} will affect both compilers
3061 at the same time and may make one of the two compilers operate
3062 improperly if set during invocation of the wrong compiler. It is also
3063 very important that the linker uses the proper @file{libgcc.a} GCC
3064 library -- that is, the one from the C++ compiler installation. The
3065 implicit link command as suggested in the @command{gnatmake} command
3066 from the former example can be replaced by an explicit link command with
3067 the full-verbosity option in order to verify which library is used:
3070 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=c++
3072 If there is a problem due to interfering environment variables, it can
3073 be worked around by using an intermediate script. The following example
3074 shows the proper script to use when GNAT has not been installed at its
3075 default location and g++ has been installed at its default location:
3083 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=./my_script
3087 Using a non-GNU C++ compiler: The commands previously described can be
3088 used to insure that the C++ linker is used. Nonetheless, you need to add
3089 a few more parameters to the link command line, depending on the exception
3092 If the @code{setjmp/longjmp} exception mechanism is used, only the paths
3093 to the libgcc libraries are required:
3098 CC $* `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a`
3099 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3102 Where CC is the name of the non-GNU C++ compiler.
3104 If the @code{zero cost} exception mechanism is used, and the platform
3105 supports automatic registration of exception tables (e.g.@: Solaris or IRIX),
3106 paths to more objects are required:
3111 CC `gcc -print-file-name=crtbegin.o` $* \
3112 `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a` \
3113 `gcc -print-file-name=crtend.o`
3114 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3117 If the @code{zero cost} exception mechanism is used, and the platform
3118 doesn't support automatic registration of exception tables (e.g.@: HP-UX,
3119 Tru64 or AIX), the simple approach described above will not work and
3120 a pre-linking phase using GNAT will be necessary.
3124 Another alternative is to use the @command{gprbuild} multi-language builder
3125 which has a large knowledge base and knows how to link Ada and C++ code
3126 together automatically in most cases.
3128 @node A Simple Example
3129 @subsection A Simple Example
3131 The following example, provided as part of the GNAT examples, shows how
3132 to achieve procedural interfacing between Ada and C++ in both
3133 directions. The C++ class A has two methods. The first method is exported
3134 to Ada by the means of an extern C wrapper function. The second method
3135 calls an Ada subprogram. On the Ada side, The C++ calls are modelled by
3136 a limited record with a layout comparable to the C++ class. The Ada
3137 subprogram, in turn, calls the C++ method. So, starting from the C++
3138 main program, the process passes back and forth between the two
3142 Here are the compilation commands:
3144 $ gnatmake -c simple_cpp_interface
3147 $ gnatbind -n simple_cpp_interface
3148 $ gnatlink simple_cpp_interface -o cpp_main --LINK=g++
3149 -lstdc++ ex7.o cpp_main.o
3153 Here are the corresponding sources:
3161 void adainit (void);
3162 void adafinal (void);
3163 void method1 (A *t);
3185 class A : public Origin @{
3187 void method1 (void);
3188 void method2 (int v);
3198 extern "C" @{ void ada_method2 (A *t, int v);@}
3200 void A::method1 (void)
3203 printf ("in A::method1, a_value = %d \n",a_value);
3207 void A::method2 (int v)
3209 ada_method2 (this, v);
3210 printf ("in A::method2, a_value = %d \n",a_value);
3217 printf ("in A::A, a_value = %d \n",a_value);
3221 @smallexample @c ada
3223 package body Simple_Cpp_Interface is
3225 procedure Ada_Method2 (This : in out A; V : Integer) is
3231 end Simple_Cpp_Interface;
3234 package Simple_Cpp_Interface is
3237 Vptr : System.Address;
3241 pragma Convention (C, A);
3243 procedure Method1 (This : in out A);
3244 pragma Import (C, Method1);
3246 procedure Ada_Method2 (This : in out A; V : Integer);
3247 pragma Export (C, Ada_Method2);
3249 end Simple_Cpp_Interface;
3252 @node Interfacing with C++ constructors
3253 @subsection Interfacing with C++ constructors
3256 In order to interface with C++ constructors GNAT provides the
3257 @code{pragma CPP_Constructor} (@xref{Interfacing to C++,,,
3258 gnat_rm, GNAT Reference Manual}, for additional information).
3259 In this section we present some common uses of C++ constructors
3260 in mixed-languages programs in GNAT.
3262 Let us assume that we need to interface with the following
3270 @b{virtual} int Get_Value ();
3271 Root(); // Default constructor
3272 Root(int v); // 1st non-default constructor
3273 Root(int v, int w); // 2nd non-default constructor
3277 For this purpose we can write the following package spec (further
3278 information on how to build this spec is available in
3279 @ref{Interfacing with C++ at the Class Level} and
3280 @ref{Generating Ada Bindings for C and C++ headers}).
3282 @smallexample @c ada
3283 with Interfaces.C; use Interfaces.C;
3285 type Root is tagged limited record
3289 pragma Import (CPP, Root);
3291 function Get_Value (Obj : Root) return int;
3292 pragma Import (CPP, Get_Value);
3294 function Constructor return Root;
3295 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ev");
3297 function Constructor (v : Integer) return Root;
3298 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ei");
3300 function Constructor (v, w : Integer) return Root;
3301 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Eii");
3305 On the Ada side the constructor is represented by a function (whose
3306 name is arbitrary) that returns the classwide type corresponding to
3307 the imported C++ class. Although the constructor is described as a
3308 function, it is typically a procedure with an extra implicit argument
3309 (the object being initialized) at the implementation level. GNAT
3310 issues the appropriate call, whatever it is, to get the object
3311 properly initialized.
3313 Constructors can only appear in the following contexts:
3317 On the right side of an initialization of an object of type @var{T}.
3319 On the right side of an initialization of a record component of type @var{T}.
3321 In an Ada 2005 limited aggregate.
3323 In an Ada 2005 nested limited aggregate.
3325 In an Ada 2005 limited aggregate that initializes an object built in
3326 place by an extended return statement.
3330 In a declaration of an object whose type is a class imported from C++,
3331 either the default C++ constructor is implicitly called by GNAT, or
3332 else the required C++ constructor must be explicitly called in the
3333 expression that initializes the object. For example:
3335 @smallexample @c ada
3337 Obj2 : Root := Constructor;
3338 Obj3 : Root := Constructor (v => 10);
3339 Obj4 : Root := Constructor (30, 40);
3342 The first two declarations are equivalent: in both cases the default C++
3343 constructor is invoked (in the former case the call to the constructor is
3344 implicit, and in the latter case the call is explicit in the object
3345 declaration). @code{Obj3} is initialized by the C++ non-default constructor
3346 that takes an integer argument, and @code{Obj4} is initialized by the
3347 non-default C++ constructor that takes two integers.
3349 Let us derive the imported C++ class in the Ada side. For example:
3351 @smallexample @c ada
3352 type DT is new Root with record
3353 C_Value : Natural := 2009;
3357 In this case the components DT inherited from the C++ side must be
3358 initialized by a C++ constructor, and the additional Ada components
3359 of type DT are initialized by GNAT. The initialization of such an
3360 object is done either by default, or by means of a function returning
3361 an aggregate of type DT, or by means of an extension aggregate.
3363 @smallexample @c ada
3365 Obj6 : DT := Function_Returning_DT (50);
3366 Obj7 : DT := (Constructor (30,40) with C_Value => 50);
3369 The declaration of @code{Obj5} invokes the default constructors: the
3370 C++ default constructor of the parent type takes care of the initialization
3371 of the components inherited from Root, and GNAT takes care of the default
3372 initialization of the additional Ada components of type DT (that is,
3373 @code{C_Value} is initialized to value 2009). The order of invocation of
3374 the constructors is consistent with the order of elaboration required by
3375 Ada and C++. That is, the constructor of the parent type is always called
3376 before the constructor of the derived type.
3378 Let us now consider a record that has components whose type is imported
3379 from C++. For example:
3381 @smallexample @c ada
3382 type Rec1 is limited record
3383 Data1 : Root := Constructor (10);
3384 Value : Natural := 1000;
3387 type Rec2 (D : Integer := 20) is limited record
3389 Data2 : Root := Constructor (D, 30);
3393 The initialization of an object of type @code{Rec2} will call the
3394 non-default C++ constructors specified for the imported components.
3397 @smallexample @c ada
3401 Using Ada 2005 we can use limited aggregates to initialize an object
3402 invoking C++ constructors that differ from those specified in the type
3403 declarations. For example:
3405 @smallexample @c ada
3406 Obj9 : Rec2 := (Rec => (Data1 => Constructor (15, 16),
3411 The above declaration uses an Ada 2005 limited aggregate to
3412 initialize @code{Obj9}, and the C++ constructor that has two integer
3413 arguments is invoked to initialize the @code{Data1} component instead
3414 of the constructor specified in the declaration of type @code{Rec1}. In
3415 Ada 2005 the box in the aggregate indicates that unspecified components
3416 are initialized using the expression (if any) available in the component
3417 declaration. That is, in this case discriminant @code{D} is initialized
3418 to value @code{20}, @code{Value} is initialized to value 1000, and the
3419 non-default C++ constructor that handles two integers takes care of
3420 initializing component @code{Data2} with values @code{20,30}.
3422 In Ada 2005 we can use the extended return statement to build the Ada
3423 equivalent to C++ non-default constructors. For example:
3425 @smallexample @c ada
3426 function Constructor (V : Integer) return Rec2 is
3428 return Obj : Rec2 := (Rec => (Data1 => Constructor (V, 20),
3431 -- Further actions required for construction of
3432 -- objects of type Rec2
3438 In this example the extended return statement construct is used to
3439 build in place the returned object whose components are initialized
3440 by means of a limited aggregate. Any further action associated with
3441 the constructor can be placed inside the construct.
3443 @node Interfacing with C++ at the Class Level
3444 @subsection Interfacing with C++ at the Class Level
3446 In this section we demonstrate the GNAT features for interfacing with
3447 C++ by means of an example making use of Ada 2005 abstract interface
3448 types. This example consists of a classification of animals; classes
3449 have been used to model our main classification of animals, and
3450 interfaces provide support for the management of secondary
3451 classifications. We first demonstrate a case in which the types and
3452 constructors are defined on the C++ side and imported from the Ada
3453 side, and latter the reverse case.
3455 The root of our derivation will be the @code{Animal} class, with a
3456 single private attribute (the @code{Age} of the animal) and two public
3457 primitives to set and get the value of this attribute.
3462 @b{virtual} void Set_Age (int New_Age);
3463 @b{virtual} int Age ();
3469 Abstract interface types are defined in C++ by means of classes with pure
3470 virtual functions and no data members. In our example we will use two
3471 interfaces that provide support for the common management of @code{Carnivore}
3472 and @code{Domestic} animals:
3475 @b{class} Carnivore @{
3477 @b{virtual} int Number_Of_Teeth () = 0;
3480 @b{class} Domestic @{
3482 @b{virtual void} Set_Owner (char* Name) = 0;
3486 Using these declarations, we can now say that a @code{Dog} is an animal that is
3487 both Carnivore and Domestic, that is:
3490 @b{class} Dog : Animal, Carnivore, Domestic @{
3492 @b{virtual} int Number_Of_Teeth ();
3493 @b{virtual} void Set_Owner (char* Name);
3495 Dog(); // Constructor
3502 In the following examples we will assume that the previous declarations are
3503 located in a file named @code{animals.h}. The following package demonstrates
3504 how to import these C++ declarations from the Ada side:
3506 @smallexample @c ada
3507 with Interfaces.C.Strings; use Interfaces.C.Strings;
3509 type Carnivore is interface;
3510 pragma Convention (C_Plus_Plus, Carnivore);
3511 function Number_Of_Teeth (X : Carnivore)
3512 return Natural is abstract;
3514 type Domestic is interface;
3515 pragma Convention (C_Plus_Plus, Set_Owner);
3517 (X : in out Domestic;
3518 Name : Chars_Ptr) is abstract;
3520 type Animal is tagged record
3523 pragma Import (C_Plus_Plus, Animal);
3525 procedure Set_Age (X : in out Animal; Age : Integer);
3526 pragma Import (C_Plus_Plus, Set_Age);
3528 function Age (X : Animal) return Integer;
3529 pragma Import (C_Plus_Plus, Age);
3531 type Dog is new Animal and Carnivore and Domestic with record
3532 Tooth_Count : Natural;
3533 Owner : String (1 .. 30);
3535 pragma Import (C_Plus_Plus, Dog);
3537 function Number_Of_Teeth (A : Dog) return Integer;
3538 pragma Import (C_Plus_Plus, Number_Of_Teeth);
3540 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3541 pragma Import (C_Plus_Plus, Set_Owner);
3543 function New_Dog return Dog;
3544 pragma CPP_Constructor (New_Dog);
3545 pragma Import (CPP, New_Dog, "_ZN3DogC2Ev");
3549 Thanks to the compatibility between GNAT run-time structures and the C++ ABI,
3550 interfacing with these C++ classes is easy. The only requirement is that all
3551 the primitives and components must be declared exactly in the same order in
3554 Regarding the abstract interfaces, we must indicate to the GNAT compiler by
3555 means of a @code{pragma Convention (C_Plus_Plus)}, the convention used to pass
3556 the arguments to the called primitives will be the same as for C++. For the
3557 imported classes we use @code{pragma Import} with convention @code{C_Plus_Plus}
3558 to indicate that they have been defined on the C++ side; this is required
3559 because the dispatch table associated with these tagged types will be built
3560 in the C++ side and therefore will not contain the predefined Ada primitives
3561 which Ada would otherwise expect.
3563 As the reader can see there is no need to indicate the C++ mangled names
3564 associated with each subprogram because it is assumed that all the calls to
3565 these primitives will be dispatching calls. The only exception is the
3566 constructor, which must be registered with the compiler by means of
3567 @code{pragma CPP_Constructor} and needs to provide its associated C++
3568 mangled name because the Ada compiler generates direct calls to it.
3570 With the above packages we can now declare objects of type Dog on the Ada side
3571 and dispatch calls to the corresponding subprograms on the C++ side. We can
3572 also extend the tagged type Dog with further fields and primitives, and
3573 override some of its C++ primitives on the Ada side. For example, here we have
3574 a type derivation defined on the Ada side that inherits all the dispatching
3575 primitives of the ancestor from the C++ side.
3578 @b{with} Animals; @b{use} Animals;
3579 @b{package} Vaccinated_Animals @b{is}
3580 @b{type} Vaccinated_Dog @b{is new} Dog @b{with null record};
3581 @b{function} Vaccination_Expired (A : Vaccinated_Dog) @b{return} Boolean;
3582 @b{end} Vaccinated_Animals;
3585 It is important to note that, because of the ABI compatibility, the programmer
3586 does not need to add any further information to indicate either the object
3587 layout or the dispatch table entry associated with each dispatching operation.
3589 Now let us define all the types and constructors on the Ada side and export
3590 them to C++, using the same hierarchy of our previous example:
3592 @smallexample @c ada
3593 with Interfaces.C.Strings;
3594 use Interfaces.C.Strings;
3596 type Carnivore is interface;
3597 pragma Convention (C_Plus_Plus, Carnivore);
3598 function Number_Of_Teeth (X : Carnivore)
3599 return Natural is abstract;
3601 type Domestic is interface;
3602 pragma Convention (C_Plus_Plus, Set_Owner);
3604 (X : in out Domestic;
3605 Name : Chars_Ptr) is abstract;
3607 type Animal is tagged record
3610 pragma Convention (C_Plus_Plus, Animal);
3612 procedure Set_Age (X : in out Animal; Age : Integer);
3613 pragma Export (C_Plus_Plus, Set_Age);
3615 function Age (X : Animal) return Integer;
3616 pragma Export (C_Plus_Plus, Age);
3618 type Dog is new Animal and Carnivore and Domestic with record
3619 Tooth_Count : Natural;
3620 Owner : String (1 .. 30);
3622 pragma Convention (C_Plus_Plus, Dog);
3624 function Number_Of_Teeth (A : Dog) return Integer;
3625 pragma Export (C_Plus_Plus, Number_Of_Teeth);
3627 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3628 pragma Export (C_Plus_Plus, Set_Owner);
3630 function New_Dog return Dog'Class;
3631 pragma Export (C_Plus_Plus, New_Dog);
3635 Compared with our previous example the only difference is the use of
3636 @code{pragma Export} to indicate to the GNAT compiler that the primitives will
3637 be available to C++. Thanks to the ABI compatibility, on the C++ side there is
3638 nothing else to be done; as explained above, the only requirement is that all
3639 the primitives and components are declared in exactly the same order.
3641 For completeness, let us see a brief C++ main program that uses the
3642 declarations available in @code{animals.h} (presented in our first example) to
3643 import and use the declarations from the Ada side, properly initializing and
3644 finalizing the Ada run-time system along the way:
3647 @b{#include} "animals.h"
3648 @b{#include} <iostream>
3649 @b{using namespace} std;
3651 void Check_Carnivore (Carnivore *obj) @{@dots{}@}
3652 void Check_Domestic (Domestic *obj) @{@dots{}@}
3653 void Check_Animal (Animal *obj) @{@dots{}@}
3654 void Check_Dog (Dog *obj) @{@dots{}@}
3657 void adainit (void);
3658 void adafinal (void);
3664 Dog *obj = new_dog(); // Ada constructor
3665 Check_Carnivore (obj); // Check secondary DT
3666 Check_Domestic (obj); // Check secondary DT
3667 Check_Animal (obj); // Check primary DT
3668 Check_Dog (obj); // Check primary DT
3673 adainit (); test(); adafinal ();
3678 @node Comparison between GNAT and C/C++ Compilation Models
3679 @section Comparison between GNAT and C/C++ Compilation Models
3682 The GNAT model of compilation is close to the C and C++ models. You can
3683 think of Ada specs as corresponding to header files in C. As in C, you
3684 don't need to compile specs; they are compiled when they are used. The
3685 Ada @code{with} is similar in effect to the @code{#include} of a C
3688 One notable difference is that, in Ada, you may compile specs separately
3689 to check them for semantic and syntactic accuracy. This is not always
3690 possible with C headers because they are fragments of programs that have
3691 less specific syntactic or semantic rules.
3693 The other major difference is the requirement for running the binder,
3694 which performs two important functions. First, it checks for
3695 consistency. In C or C++, the only defense against assembling
3696 inconsistent programs lies outside the compiler, in a makefile, for
3697 example. The binder satisfies the Ada requirement that it be impossible
3698 to construct an inconsistent program when the compiler is used in normal
3701 @cindex Elaboration order control
3702 The other important function of the binder is to deal with elaboration
3703 issues. There are also elaboration issues in C++ that are handled
3704 automatically. This automatic handling has the advantage of being
3705 simpler to use, but the C++ programmer has no control over elaboration.
3706 Where @code{gnatbind} might complain there was no valid order of
3707 elaboration, a C++ compiler would simply construct a program that
3708 malfunctioned at run time.
3711 @node Comparison between GNAT and Conventional Ada Library Models
3712 @section Comparison between GNAT and Conventional Ada Library Models
3715 This section is intended for Ada programmers who have
3716 used an Ada compiler implementing the traditional Ada library
3717 model, as described in the Ada Reference Manual.
3719 @cindex GNAT library
3720 In GNAT, there is no ``library'' in the normal sense. Instead, the set of
3721 source files themselves acts as the library. Compiling Ada programs does
3722 not generate any centralized information, but rather an object file and
3723 a ALI file, which are of interest only to the binder and linker.
3724 In a traditional system, the compiler reads information not only from
3725 the source file being compiled, but also from the centralized library.
3726 This means that the effect of a compilation depends on what has been
3727 previously compiled. In particular:
3731 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3732 to the version of the unit most recently compiled into the library.
3735 Inlining is effective only if the necessary body has already been
3736 compiled into the library.
3739 Compiling a unit may obsolete other units in the library.
3743 In GNAT, compiling one unit never affects the compilation of any other
3744 units because the compiler reads only source files. Only changes to source
3745 files can affect the results of a compilation. In particular:
3749 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3750 to the source version of the unit that is currently accessible to the
3755 Inlining requires the appropriate source files for the package or
3756 subprogram bodies to be available to the compiler. Inlining is always
3757 effective, independent of the order in which units are complied.
3760 Compiling a unit never affects any other compilations. The editing of
3761 sources may cause previous compilations to be out of date if they
3762 depended on the source file being modified.
3766 The most important result of these differences is that order of compilation
3767 is never significant in GNAT. There is no situation in which one is
3768 required to do one compilation before another. What shows up as order of
3769 compilation requirements in the traditional Ada library becomes, in
3770 GNAT, simple source dependencies; in other words, there is only a set
3771 of rules saying what source files must be present when a file is
3775 @node Placement of temporary files
3776 @section Placement of temporary files
3777 @cindex Temporary files (user control over placement)
3780 GNAT creates temporary files in the directory designated by the environment
3781 variable @env{TMPDIR}.
3782 (See the HP @emph{C RTL Reference Manual} on the function @code{getenv()}
3783 for detailed information on how environment variables are resolved.
3784 For most users the easiest way to make use of this feature is to simply
3785 define @env{TMPDIR} as a job level logical name).
3786 For example, if you wish to use a Ramdisk (assuming DECRAM is installed)
3787 for compiler temporary files, then you can include something like the
3788 following command in your @file{LOGIN.COM} file:
3791 $ define/job TMPDIR "/disk$scratchram/000000/temp/"
3795 If @env{TMPDIR} is not defined, then GNAT uses the directory designated by
3796 @env{TMP}; if @env{TMP} is not defined, then GNAT uses the directory
3797 designated by @env{TEMP}.
3798 If none of these environment variables are defined then GNAT uses the
3799 directory designated by the logical name @code{SYS$SCRATCH:}
3800 (by default the user's home directory). If all else fails
3801 GNAT uses the current directory for temporary files.
3804 @c *************************
3805 @node Compiling Using gcc
3806 @chapter Compiling Using @command{gcc}
3809 This chapter discusses how to compile Ada programs using the @command{gcc}
3810 command. It also describes the set of switches
3811 that can be used to control the behavior of the compiler.
3813 * Compiling Programs::
3814 * Switches for gcc::
3815 * Search Paths and the Run-Time Library (RTL)::
3816 * Order of Compilation Issues::
3820 @node Compiling Programs
3821 @section Compiling Programs
3824 The first step in creating an executable program is to compile the units
3825 of the program using the @command{gcc} command. You must compile the
3830 the body file (@file{.adb}) for a library level subprogram or generic
3834 the spec file (@file{.ads}) for a library level package or generic
3835 package that has no body
3838 the body file (@file{.adb}) for a library level package
3839 or generic package that has a body
3844 You need @emph{not} compile the following files
3849 the spec of a library unit which has a body
3856 because they are compiled as part of compiling related units. GNAT
3858 when the corresponding body is compiled, and subunits when the parent is
3861 @cindex cannot generate code
3862 If you attempt to compile any of these files, you will get one of the
3863 following error messages (where @var{fff} is the name of the file you compiled):
3866 cannot generate code for file @var{fff} (package spec)
3867 to check package spec, use -gnatc
3869 cannot generate code for file @var{fff} (missing subunits)
3870 to check parent unit, use -gnatc
3872 cannot generate code for file @var{fff} (subprogram spec)
3873 to check subprogram spec, use -gnatc
3875 cannot generate code for file @var{fff} (subunit)
3876 to check subunit, use -gnatc
3880 As indicated by the above error messages, if you want to submit
3881 one of these files to the compiler to check for correct semantics
3882 without generating code, then use the @option{-gnatc} switch.
3884 The basic command for compiling a file containing an Ada unit is
3887 @c $ gcc -c @ovar{switches} @file{file name}
3888 @c Expanding @ovar macro inline (explanation in macro def comments)
3889 $ gcc -c @r{[}@var{switches}@r{]} @file{file name}
3893 where @var{file name} is the name of the Ada file (usually
3895 @file{.ads} for a spec or @file{.adb} for a body).
3898 @option{-c} switch to tell @command{gcc} to compile, but not link, the file.
3900 The result of a successful compilation is an object file, which has the
3901 same name as the source file but an extension of @file{.o} and an Ada
3902 Library Information (ALI) file, which also has the same name as the
3903 source file, but with @file{.ali} as the extension. GNAT creates these
3904 two output files in the current directory, but you may specify a source
3905 file in any directory using an absolute or relative path specification
3906 containing the directory information.
3909 @command{gcc} is actually a driver program that looks at the extensions of
3910 the file arguments and loads the appropriate compiler. For example, the
3911 GNU C compiler is @file{cc1}, and the Ada compiler is @file{gnat1}.
3912 These programs are in directories known to the driver program (in some
3913 configurations via environment variables you set), but need not be in
3914 your path. The @command{gcc} driver also calls the assembler and any other
3915 utilities needed to complete the generation of the required object
3918 It is possible to supply several file names on the same @command{gcc}
3919 command. This causes @command{gcc} to call the appropriate compiler for
3920 each file. For example, the following command lists three separate
3921 files to be compiled:
3924 $ gcc -c x.adb y.adb z.c
3928 calls @code{gnat1} (the Ada compiler) twice to compile @file{x.adb} and
3929 @file{y.adb}, and @code{cc1} (the C compiler) once to compile @file{z.c}.
3930 The compiler generates three object files @file{x.o}, @file{y.o} and
3931 @file{z.o} and the two ALI files @file{x.ali} and @file{y.ali} from the
3932 Ada compilations. Any switches apply to all the files ^listed,^listed.^
3935 @option{-gnat@var{x}} switches, which apply only to Ada compilations.
3938 @node Switches for gcc
3939 @section Switches for @command{gcc}
3942 The @command{gcc} command accepts switches that control the
3943 compilation process. These switches are fully described in this section.
3944 First we briefly list all the switches, in alphabetical order, then we
3945 describe the switches in more detail in functionally grouped sections.
3947 More switches exist for GCC than those documented here, especially
3948 for specific targets. However, their use is not recommended as
3949 they may change code generation in ways that are incompatible with
3950 the Ada run-time library, or can cause inconsistencies between
3954 * Output and Error Message Control::
3955 * Warning Message Control::
3956 * Debugging and Assertion Control::
3957 * Validity Checking::
3960 * Using gcc for Syntax Checking::
3961 * Using gcc for Semantic Checking::
3962 * Compiling Different Versions of Ada::
3963 * Character Set Control::
3964 * File Naming Control::
3965 * Subprogram Inlining Control::
3966 * Auxiliary Output Control::
3967 * Debugging Control::
3968 * Exception Handling Control::
3969 * Units to Sources Mapping Files::
3970 * Integrated Preprocessing::
3971 * Code Generation Control::
3980 @cindex @option{-b} (@command{gcc})
3981 @item -b @var{target}
3982 Compile your program to run on @var{target}, which is the name of a
3983 system configuration. You must have a GNAT cross-compiler built if
3984 @var{target} is not the same as your host system.
3987 @cindex @option{-B} (@command{gcc})
3988 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
3989 from @var{dir} instead of the default location. Only use this switch
3990 when multiple versions of the GNAT compiler are available.
3991 @xref{Directory Options,, Options for Directory Search, gcc, Using the
3992 GNU Compiler Collection (GCC)}, for further details. You would normally
3993 use the @option{-b} or @option{-V} switch instead.
3996 @cindex @option{-c} (@command{gcc})
3997 Compile. Always use this switch when compiling Ada programs.
3999 Note: for some other languages when using @command{gcc}, notably in
4000 the case of C and C++, it is possible to use
4001 use @command{gcc} without a @option{-c} switch to
4002 compile and link in one step. In the case of GNAT, you
4003 cannot use this approach, because the binder must be run
4004 and @command{gcc} cannot be used to run the GNAT binder.
4008 @cindex @option{-fno-inline} (@command{gcc})
4009 Suppresses all back-end inlining, even if other optimization or inlining
4011 This includes suppression of inlining that results
4012 from the use of the pragma @code{Inline_Always}.
4013 Any occurrences of pragma @code{Inline} or @code{Inline_Always}
4014 are ignored, and @option{-gnatn} and @option{-gnatN} have no
4015 effect if this switch is present.
4017 @item -fno-inline-functions
4018 @cindex @option{-fno-inline-functions} (@command{gcc})
4019 Suppresses automatic inlining of simple subprograms, which is enabled
4020 if @option{-O3} is used.
4022 @item -fno-inline-small-functions
4023 @cindex @option{-fno-inline-small-functions} (@command{gcc})
4024 Suppresses automatic inlining of small subprograms, which is enabled
4025 if @option{-O2} is used.
4027 @item -fno-inline-functions-called-once
4028 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
4029 Suppresses inlining of subprograms local to the unit and called once
4030 from within it, which is enabled if @option{-O1} is used.
4033 @cindex @option{-fno-ivopts} (@command{gcc})
4034 Suppresses high-level loop induction variable optimizations, which are
4035 enabled if @option{-O1} is used. These optimizations are generally
4036 profitable but, for some specific cases of loops with numerous uses
4037 of the iteration variable that follow a common pattern, they may end
4038 up destroying the regularity that could be exploited at a lower level
4039 and thus producing inferior code.
4041 @item -fno-strict-aliasing
4042 @cindex @option{-fno-strict-aliasing} (@command{gcc})
4043 Causes the compiler to avoid assumptions regarding non-aliasing
4044 of objects of different types. See
4045 @ref{Optimization and Strict Aliasing} for details.
4048 @cindex @option{-fstack-check} (@command{gcc})
4049 Activates stack checking.
4050 See @ref{Stack Overflow Checking} for details.
4053 @cindex @option{-fstack-usage} (@command{gcc})
4054 Makes the compiler output stack usage information for the program, on a
4055 per-function basis. See @ref{Static Stack Usage Analysis} for details.
4057 @item -fcallgraph-info@r{[}=su@r{]}
4058 @cindex @option{-fcallgraph-info} (@command{gcc})
4059 Makes the compiler output callgraph information for the program, on a
4060 per-file basis. The information is generated in the VCG format. It can
4061 be decorated with stack-usage per-node information.
4064 @cindex @option{^-g^/DEBUG^} (@command{gcc})
4065 Generate debugging information. This information is stored in the object
4066 file and copied from there to the final executable file by the linker,
4067 where it can be read by the debugger. You must use the
4068 @option{^-g^/DEBUG^} switch if you plan on using the debugger.
4071 @cindex @option{-gnat83} (@command{gcc})
4072 Enforce Ada 83 restrictions.
4075 @cindex @option{-gnat95} (@command{gcc})
4076 Enforce Ada 95 restrictions.
4079 @cindex @option{-gnat05} (@command{gcc})
4080 Allow full Ada 2005 features.
4083 @cindex @option{-gnata} (@command{gcc})
4084 Assertions enabled. @code{Pragma Assert} and @code{pragma Debug} to be
4085 activated. Note that these pragmas can also be controlled using the
4086 configuration pragmas @code{Assertion_Policy} and @code{Debug_Policy}.
4087 It also activates pragmas @code{Check}, @code{Precondition}, and
4088 @code{Postcondition}. Note that these pragmas can also be controlled
4089 using the configuration pragma @code{Check_Policy}.
4092 @cindex @option{-gnatA} (@command{gcc})
4093 Avoid processing @file{gnat.adc}. If a @file{gnat.adc} file is present,
4097 @cindex @option{-gnatb} (@command{gcc})
4098 Generate brief messages to @file{stderr} even if verbose mode set.
4101 @cindex @option{-gnatB} (@command{gcc})
4102 Assume no invalid (bad) values except for 'Valid attribute use
4103 (@pxref{Validity Checking}).
4106 @cindex @option{-gnatc} (@command{gcc})
4107 Check syntax and semantics only (no code generation attempted).
4110 @cindex @option{-gnatC} (@command{gcc})
4111 Generate CodePeer information (no code generation attempted).
4112 This switch will generate an intermediate representation suitable for
4113 use by CodePeer (@file{.scil} files). This switch is not compatible with
4114 code generation (it will, among other things, disable some switches such
4115 as -gnatn, and enable others such as -gnata).
4118 @cindex @option{-gnatd} (@command{gcc})
4119 Specify debug options for the compiler. The string of characters after
4120 the @option{-gnatd} specify the specific debug options. The possible
4121 characters are 0-9, a-z, A-Z, optionally preceded by a dot. See
4122 compiler source file @file{debug.adb} for details of the implemented
4123 debug options. Certain debug options are relevant to applications
4124 programmers, and these are documented at appropriate points in this
4129 @cindex @option{-gnatD[nn]} (@command{gcc})
4132 @item /XDEBUG /LXDEBUG=nnn
4134 Create expanded source files for source level debugging. This switch
4135 also suppress generation of cross-reference information
4136 (see @option{-gnatx}).
4138 @item -gnatec=@var{path}
4139 @cindex @option{-gnatec} (@command{gcc})
4140 Specify a configuration pragma file
4142 (the equal sign is optional)
4144 (@pxref{The Configuration Pragmas Files}).
4146 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=@var{value}@r{]}
4147 @cindex @option{-gnateD} (@command{gcc})
4148 Defines a symbol, associated with @var{value}, for preprocessing.
4149 (@pxref{Integrated Preprocessing}).
4152 @cindex @option{-gnatef} (@command{gcc})
4153 Display full source path name in brief error messages.
4156 @cindex @option{-gnateG} (@command{gcc})
4157 Save result of preprocessing in a text file.
4159 @item -gnatem=@var{path}
4160 @cindex @option{-gnatem} (@command{gcc})
4161 Specify a mapping file
4163 (the equal sign is optional)
4165 (@pxref{Units to Sources Mapping Files}).
4167 @item -gnatep=@var{file}
4168 @cindex @option{-gnatep} (@command{gcc})
4169 Specify a preprocessing data file
4171 (the equal sign is optional)
4173 (@pxref{Integrated Preprocessing}).
4176 @cindex @option{-gnateS} (@command{gcc})
4177 Generate SCO (Source Coverage Obligation) information in the ALI
4178 file. This information is used by advanced coverage tools. See
4179 unit @file{SCOs} in the compiler sources for details in files
4180 @file{scos.ads} and @file{scos.adb}.
4183 @cindex @option{-gnatE} (@command{gcc})
4184 Full dynamic elaboration checks.
4187 @cindex @option{-gnatf} (@command{gcc})
4188 Full errors. Multiple errors per line, all undefined references, do not
4189 attempt to suppress cascaded errors.
4192 @cindex @option{-gnatF} (@command{gcc})
4193 Externals names are folded to all uppercase.
4195 @item ^-gnatg^/GNAT_INTERNAL^
4196 @cindex @option{^-gnatg^/GNAT_INTERNAL^} (@command{gcc})
4197 Internal GNAT implementation mode. This should not be used for
4198 applications programs, it is intended only for use by the compiler
4199 and its run-time library. For documentation, see the GNAT sources.
4200 Note that @option{^-gnatg^/GNAT_INTERNAL^} implies
4201 @option{^-gnatwae^/WARNINGS=ALL,ERRORS^} and
4202 @option{^-gnatyg^/STYLE_CHECKS=GNAT^}
4203 so that all standard warnings and all standard style options are turned on.
4204 All warnings and style error messages are treated as errors.
4208 @cindex @option{-gnatG[nn]} (@command{gcc})
4211 @item /EXPAND_SOURCE, /LEXPAND_SOURCE=nnn
4213 List generated expanded code in source form.
4215 @item ^-gnath^/HELP^
4216 @cindex @option{^-gnath^/HELP^} (@command{gcc})
4217 Output usage information. The output is written to @file{stdout}.
4219 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
4220 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
4221 Identifier character set
4223 (@var{c}=1/2/3/4/8/9/p/f/n/w).
4225 For details of the possible selections for @var{c},
4226 see @ref{Character Set Control}.
4228 @item ^-gnatI^/IGNORE_REP_CLAUSES^
4229 @cindex @option{^-gnatI^IGNORE_REP_CLAUSES^} (@command{gcc})
4230 Ignore representation clauses. When this switch is used,
4231 representation clauses are treated as comments. This is useful
4232 when initially porting code where you want to ignore rep clause
4233 problems, and also for compiling foreign code (particularly
4234 for use with ASIS). The representation clauses that are ignored
4235 are: enumeration_representation_clause, record_representation_clause,
4236 and attribute_definition_clause for the following attributes:
4237 Address, Alignment, Bit_Order, Component_Size, Machine_Radix,
4238 Object_Size, Size, Small, Stream_Size, and Value_Size.
4239 Note that this option should be used only for compiling -- the
4240 code is likely to malfunction at run time.
4243 @cindex @option{-gnatjnn} (@command{gcc})
4244 Reformat error messages to fit on nn character lines
4246 @item -gnatk=@var{n}
4247 @cindex @option{-gnatk} (@command{gcc})
4248 Limit file names to @var{n} (1-999) characters ^(@code{k} = krunch)^^.
4251 @cindex @option{-gnatl} (@command{gcc})
4252 Output full source listing with embedded error messages.
4255 @cindex @option{-gnatL} (@command{gcc})
4256 Used in conjunction with -gnatG or -gnatD to intersperse original
4257 source lines (as comment lines with line numbers) in the expanded
4260 @item -gnatm=@var{n}
4261 @cindex @option{-gnatm} (@command{gcc})
4262 Limit number of detected error or warning messages to @var{n}
4263 where @var{n} is in the range 1..999999. The default setting if
4264 no switch is given is 9999. If the number of warnings reaches this
4265 limit, then a message is output and further warnings are suppressed,
4266 but the compilation is continued. If the number of error messages
4267 reaches this limit, then a message is output and the compilation
4268 is abandoned. The equal sign here is optional. A value of zero
4269 means that no limit applies.
4272 @cindex @option{-gnatn} (@command{gcc})
4273 Activate inlining for subprograms for which
4274 pragma @code{inline} is specified. This inlining is performed
4275 by the GCC back-end.
4278 @cindex @option{-gnatN} (@command{gcc})
4279 Activate front end inlining for subprograms for which
4280 pragma @code{Inline} is specified. This inlining is performed
4281 by the front end and will be visible in the
4282 @option{-gnatG} output.
4284 When using a gcc-based back end (in practice this means using any version
4285 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
4286 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
4287 Historically front end inlining was more extensive than the gcc back end
4288 inlining, but that is no longer the case.
4291 @cindex @option{-gnato} (@command{gcc})
4292 Enable numeric overflow checking (which is not normally enabled by
4293 default). Note that division by zero is a separate check that is not
4294 controlled by this switch (division by zero checking is on by default).
4297 @cindex @option{-gnatp} (@command{gcc})
4298 Suppress all checks. See @ref{Run-Time Checks} for details. This switch
4299 has no effect if cancelled by a subsequent @option{-gnat-p} switch.
4302 @cindex @option{-gnat-p} (@command{gcc})
4303 Cancel effect of previous @option{-gnatp} switch.
4306 @cindex @option{-gnatP} (@command{gcc})
4307 Enable polling. This is required on some systems (notably Windows NT) to
4308 obtain asynchronous abort and asynchronous transfer of control capability.
4309 @xref{Pragma Polling,,, gnat_rm, GNAT Reference Manual}, for full
4313 @cindex @option{-gnatq} (@command{gcc})
4314 Don't quit. Try semantics, even if parse errors.
4317 @cindex @option{-gnatQ} (@command{gcc})
4318 Don't quit. Generate @file{ALI} and tree files even if illegalities.
4321 @cindex @option{-gnatr} (@command{gcc})
4322 Treat pragma Restrictions as Restriction_Warnings.
4324 @item ^-gnatR@r{[}0@r{/}1@r{/}2@r{/}3@r{[}s@r{]]}^/REPRESENTATION_INFO^
4325 @cindex @option{-gnatR} (@command{gcc})
4326 Output representation information for declared types and objects.
4329 @cindex @option{-gnats} (@command{gcc})
4333 @cindex @option{-gnatS} (@command{gcc})
4334 Print package Standard.
4337 @cindex @option{-gnatt} (@command{gcc})
4338 Generate tree output file.
4340 @item ^-gnatT^/TABLE_MULTIPLIER=^@var{nnn}
4341 @cindex @option{^-gnatT^/TABLE_MULTIPLIER^} (@command{gcc})
4342 All compiler tables start at @var{nnn} times usual starting size.
4345 @cindex @option{-gnatu} (@command{gcc})
4346 List units for this compilation.
4349 @cindex @option{-gnatU} (@command{gcc})
4350 Tag all error messages with the unique string ``error:''
4353 @cindex @option{-gnatv} (@command{gcc})
4354 Verbose mode. Full error output with source lines to @file{stdout}.
4357 @cindex @option{-gnatV} (@command{gcc})
4358 Control level of validity checking (@pxref{Validity Checking}).
4360 @item ^-gnatw@var{xxx}^/WARNINGS=(@var{option}@r{[},@dots{}@r{]})^
4361 @cindex @option{^-gnatw^/WARNINGS^} (@command{gcc})
4363 ^@var{xxx} is a string of option letters that^the list of options^ denotes
4364 the exact warnings that
4365 are enabled or disabled (@pxref{Warning Message Control}).
4367 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
4368 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
4369 Wide character encoding method
4371 (@var{e}=n/h/u/s/e/8).
4374 (@var{e}=@code{BRACKETS, NONE, HEX, UPPER, SHIFT_JIS, EUC, UTF8})
4378 @cindex @option{-gnatx} (@command{gcc})
4379 Suppress generation of cross-reference information.
4381 @item ^-gnaty^/STYLE_CHECKS=(option,option@dots{})^
4382 @cindex @option{^-gnaty^/STYLE_CHECKS^} (@command{gcc})
4383 Enable built-in style checks (@pxref{Style Checking}).
4385 @item ^-gnatz^/DISTRIBUTION_STUBS=^@var{m}
4386 @cindex @option{^-gnatz^/DISTRIBUTION_STUBS^} (@command{gcc})
4387 Distribution stub generation and compilation
4389 (@var{m}=r/c for receiver/caller stubs).
4392 (@var{m}=@code{RECEIVER} or @code{CALLER} to specify the type of stubs
4393 to be generated and compiled).
4396 @item ^-I^/SEARCH=^@var{dir}
4397 @cindex @option{^-I^/SEARCH^} (@command{gcc})
4399 Direct GNAT to search the @var{dir} directory for source files needed by
4400 the current compilation
4401 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4403 @item ^-I-^/NOCURRENT_DIRECTORY^
4404 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gcc})
4406 Except for the source file named in the command line, do not look for source
4407 files in the directory containing the source file named in the command line
4408 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4412 @cindex @option{-mbig-switch} (@command{gcc})
4413 @cindex @code{case} statement (effect of @option{-mbig-switch} option)
4414 This standard gcc switch causes the compiler to use larger offsets in its
4415 jump table representation for @code{case} statements.
4416 This may result in less efficient code, but is sometimes necessary
4417 (for example on HP-UX targets)
4418 @cindex HP-UX and @option{-mbig-switch} option
4419 in order to compile large and/or nested @code{case} statements.
4422 @cindex @option{-o} (@command{gcc})
4423 This switch is used in @command{gcc} to redirect the generated object file
4424 and its associated ALI file. Beware of this switch with GNAT, because it may
4425 cause the object file and ALI file to have different names which in turn
4426 may confuse the binder and the linker.
4430 @cindex @option{-nostdinc} (@command{gcc})
4431 Inhibit the search of the default location for the GNAT Run Time
4432 Library (RTL) source files.
4435 @cindex @option{-nostdlib} (@command{gcc})
4436 Inhibit the search of the default location for the GNAT Run Time
4437 Library (RTL) ALI files.
4441 @c Expanding @ovar macro inline (explanation in macro def comments)
4442 @item -O@r{[}@var{n}@r{]}
4443 @cindex @option{-O} (@command{gcc})
4444 @var{n} controls the optimization level.
4448 No optimization, the default setting if no @option{-O} appears
4451 Normal optimization, the default if you specify @option{-O} without
4452 an operand. A good compromise between code quality and compilation
4456 Extensive optimization, may improve execution time, possibly at the cost of
4457 substantially increased compilation time.
4460 Same as @option{-O2}, and also includes inline expansion for small subprograms
4464 Optimize space usage
4468 See also @ref{Optimization Levels}.
4473 @cindex @option{/NOOPTIMIZE} (@code{GNAT COMPILE})
4474 Equivalent to @option{/OPTIMIZE=NONE}.
4475 This is the default behavior in the absence of an @option{/OPTIMIZE}
4478 @item /OPTIMIZE@r{[}=(keyword@r{[},@dots{}@r{]})@r{]}
4479 @cindex @option{/OPTIMIZE} (@code{GNAT COMPILE})
4480 Selects the level of optimization for your program. The supported
4481 keywords are as follows:
4484 Perform most optimizations, including those that
4486 This is the default if the @option{/OPTIMIZE} qualifier is supplied
4487 without keyword options.
4490 Do not do any optimizations. Same as @code{/NOOPTIMIZE}.
4493 Perform some optimizations, but omit ones that are costly.
4496 Same as @code{SOME}.
4499 Full optimization as in @option{/OPTIMIZE=ALL}, and also attempts
4500 automatic inlining of small subprograms within a unit
4503 Try to unroll loops. This keyword may be specified together with
4504 any keyword above other than @code{NONE}. Loop unrolling
4505 usually, but not always, improves the performance of programs.
4508 Optimize space usage
4512 See also @ref{Optimization Levels}.
4516 @item -pass-exit-codes
4517 @cindex @option{-pass-exit-codes} (@command{gcc})
4518 Catch exit codes from the compiler and use the most meaningful as
4522 @item --RTS=@var{rts-path}
4523 @cindex @option{--RTS} (@command{gcc})
4524 Specifies the default location of the runtime library. Same meaning as the
4525 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
4528 @cindex @option{^-S^/ASM^} (@command{gcc})
4529 ^Used in place of @option{-c} to^Used to^
4530 cause the assembler source file to be
4531 generated, using @file{^.s^.S^} as the extension,
4532 instead of the object file.
4533 This may be useful if you need to examine the generated assembly code.
4535 @item ^-fverbose-asm^/VERBOSE_ASM^
4536 @cindex @option{^-fverbose-asm^/VERBOSE_ASM^} (@command{gcc})
4537 ^Used in conjunction with @option{-S}^Used in place of @option{/ASM}^
4538 to cause the generated assembly code file to be annotated with variable
4539 names, making it significantly easier to follow.
4542 @cindex @option{^-v^/VERBOSE^} (@command{gcc})
4543 Show commands generated by the @command{gcc} driver. Normally used only for
4544 debugging purposes or if you need to be sure what version of the
4545 compiler you are executing.
4549 @cindex @option{-V} (@command{gcc})
4550 Execute @var{ver} version of the compiler. This is the @command{gcc}
4551 version, not the GNAT version.
4554 @item ^-w^/NO_BACK_END_WARNINGS^
4555 @cindex @option{-w} (@command{gcc})
4556 Turn off warnings generated by the back end of the compiler. Use of
4557 this switch also causes the default for front end warnings to be set
4558 to suppress (as though @option{-gnatws} had appeared at the start of
4564 @c Combining qualifiers does not work on VMS
4565 You may combine a sequence of GNAT switches into a single switch. For
4566 example, the combined switch
4568 @cindex Combining GNAT switches
4574 is equivalent to specifying the following sequence of switches:
4577 -gnato -gnatf -gnati3
4582 The following restrictions apply to the combination of switches
4587 The switch @option{-gnatc} if combined with other switches must come
4588 first in the string.
4591 The switch @option{-gnats} if combined with other switches must come
4592 first in the string.
4596 ^^@option{/DISTRIBUTION_STUBS=},^
4597 @option{-gnatzc} and @option{-gnatzr} may not be combined with any other
4598 switches, and only one of them may appear in the command line.
4601 The switch @option{-gnat-p} may not be combined with any other switch.
4605 Once a ``y'' appears in the string (that is a use of the @option{-gnaty}
4606 switch), then all further characters in the switch are interpreted
4607 as style modifiers (see description of @option{-gnaty}).
4610 Once a ``d'' appears in the string (that is a use of the @option{-gnatd}
4611 switch), then all further characters in the switch are interpreted
4612 as debug flags (see description of @option{-gnatd}).
4615 Once a ``w'' appears in the string (that is a use of the @option{-gnatw}
4616 switch), then all further characters in the switch are interpreted
4617 as warning mode modifiers (see description of @option{-gnatw}).
4620 Once a ``V'' appears in the string (that is a use of the @option{-gnatV}
4621 switch), then all further characters in the switch are interpreted
4622 as validity checking options (@pxref{Validity Checking}).
4625 Option ``em'', ``ec'', ``ep'', ``l='' and ``R'' must be the last options in
4626 a combined list of options.
4630 @node Output and Error Message Control
4631 @subsection Output and Error Message Control
4635 The standard default format for error messages is called ``brief format''.
4636 Brief format messages are written to @file{stderr} (the standard error
4637 file) and have the following form:
4640 e.adb:3:04: Incorrect spelling of keyword "function"
4641 e.adb:4:20: ";" should be "is"
4645 The first integer after the file name is the line number in the file,
4646 and the second integer is the column number within the line.
4648 @code{GPS} can parse the error messages
4649 and point to the referenced character.
4651 The following switches provide control over the error message
4657 @cindex @option{-gnatv} (@command{gcc})
4660 The v stands for verbose.
4662 The effect of this setting is to write long-format error
4663 messages to @file{stdout} (the standard output file.
4664 The same program compiled with the
4665 @option{-gnatv} switch would generate:
4669 3. funcion X (Q : Integer)
4671 >>> Incorrect spelling of keyword "function"
4674 >>> ";" should be "is"
4679 The vertical bar indicates the location of the error, and the @samp{>>>}
4680 prefix can be used to search for error messages. When this switch is
4681 used the only source lines output are those with errors.
4684 @cindex @option{-gnatl} (@command{gcc})
4686 The @code{l} stands for list.
4688 This switch causes a full listing of
4689 the file to be generated. In the case where a body is
4690 compiled, the corresponding spec is also listed, along
4691 with any subunits. Typical output from compiling a package
4692 body @file{p.adb} might look like:
4694 @smallexample @c ada
4698 1. package body p is
4700 3. procedure a is separate;
4711 2. pragma Elaborate_Body
4735 When you specify the @option{-gnatv} or @option{-gnatl} switches and
4736 standard output is redirected, a brief summary is written to
4737 @file{stderr} (standard error) giving the number of error messages and
4738 warning messages generated.
4740 @item -^gnatl^OUTPUT_FILE^=file
4741 @cindex @option{^-gnatl^OUTPUT_FILE^=fname} (@command{gcc})
4742 This has the same effect as @option{-gnatl} except that the output is
4743 written to a file instead of to standard output. If the given name
4744 @file{fname} does not start with a period, then it is the full name
4745 of the file to be written. If @file{fname} is an extension, it is
4746 appended to the name of the file being compiled. For example, if
4747 file @file{xyz.adb} is compiled with @option{^-gnatl^OUTPUT_FILE^=.lst},
4748 then the output is written to file ^xyz.adb.lst^xyz.adb_lst^.
4751 @cindex @option{-gnatU} (@command{gcc})
4752 This switch forces all error messages to be preceded by the unique
4753 string ``error:''. This means that error messages take a few more
4754 characters in space, but allows easy searching for and identification
4758 @cindex @option{-gnatb} (@command{gcc})
4760 The @code{b} stands for brief.
4762 This switch causes GNAT to generate the
4763 brief format error messages to @file{stderr} (the standard error
4764 file) as well as the verbose
4765 format message or full listing (which as usual is written to
4766 @file{stdout} (the standard output file).
4768 @item -gnatm=@var{n}
4769 @cindex @option{-gnatm} (@command{gcc})
4771 The @code{m} stands for maximum.
4773 @var{n} is a decimal integer in the
4774 range of 1 to 999999 and limits the number of error or warning
4775 messages to be generated. For example, using
4776 @option{-gnatm2} might yield
4779 e.adb:3:04: Incorrect spelling of keyword "function"
4780 e.adb:5:35: missing ".."
4781 fatal error: maximum number of errors detected
4782 compilation abandoned
4786 The default setting if
4787 no switch is given is 9999. If the number of warnings reaches this
4788 limit, then a message is output and further warnings are suppressed,
4789 but the compilation is continued. If the number of error messages
4790 reaches this limit, then a message is output and the compilation
4791 is abandoned. A value of zero means that no limit applies.
4794 Note that the equal sign is optional, so the switches
4795 @option{-gnatm2} and @option{-gnatm=2} are equivalent.
4798 @cindex @option{-gnatf} (@command{gcc})
4799 @cindex Error messages, suppressing
4801 The @code{f} stands for full.
4803 Normally, the compiler suppresses error messages that are likely to be
4804 redundant. This switch causes all error
4805 messages to be generated. In particular, in the case of
4806 references to undefined variables. If a given variable is referenced
4807 several times, the normal format of messages is
4809 e.adb:7:07: "V" is undefined (more references follow)
4813 where the parenthetical comment warns that there are additional
4814 references to the variable @code{V}. Compiling the same program with the
4815 @option{-gnatf} switch yields
4818 e.adb:7:07: "V" is undefined
4819 e.adb:8:07: "V" is undefined
4820 e.adb:8:12: "V" is undefined
4821 e.adb:8:16: "V" is undefined
4822 e.adb:9:07: "V" is undefined
4823 e.adb:9:12: "V" is undefined
4827 The @option{-gnatf} switch also generates additional information for
4828 some error messages. Some examples are:
4832 Details on possibly non-portable unchecked conversion
4834 List possible interpretations for ambiguous calls
4836 Additional details on incorrect parameters
4840 @cindex @option{-gnatjnn} (@command{gcc})
4841 In normal operation mode (or if @option{-gnatj0} is used, then error messages
4842 with continuation lines are treated as though the continuation lines were
4843 separate messages (and so a warning with two continuation lines counts as
4844 three warnings, and is listed as three separate messages).
4846 If the @option{-gnatjnn} switch is used with a positive value for nn, then
4847 messages are output in a different manner. A message and all its continuation
4848 lines are treated as a unit, and count as only one warning or message in the
4849 statistics totals. Furthermore, the message is reformatted so that no line
4850 is longer than nn characters.
4853 @cindex @option{-gnatq} (@command{gcc})
4855 The @code{q} stands for quit (really ``don't quit'').
4857 In normal operation mode, the compiler first parses the program and
4858 determines if there are any syntax errors. If there are, appropriate
4859 error messages are generated and compilation is immediately terminated.
4861 GNAT to continue with semantic analysis even if syntax errors have been
4862 found. This may enable the detection of more errors in a single run. On
4863 the other hand, the semantic analyzer is more likely to encounter some
4864 internal fatal error when given a syntactically invalid tree.
4867 @cindex @option{-gnatQ} (@command{gcc})
4868 In normal operation mode, the @file{ALI} file is not generated if any
4869 illegalities are detected in the program. The use of @option{-gnatQ} forces
4870 generation of the @file{ALI} file. This file is marked as being in
4871 error, so it cannot be used for binding purposes, but it does contain
4872 reasonably complete cross-reference information, and thus may be useful
4873 for use by tools (e.g., semantic browsing tools or integrated development
4874 environments) that are driven from the @file{ALI} file. This switch
4875 implies @option{-gnatq}, since the semantic phase must be run to get a
4876 meaningful ALI file.
4878 In addition, if @option{-gnatt} is also specified, then the tree file is
4879 generated even if there are illegalities. It may be useful in this case
4880 to also specify @option{-gnatq} to ensure that full semantic processing
4881 occurs. The resulting tree file can be processed by ASIS, for the purpose
4882 of providing partial information about illegal units, but if the error
4883 causes the tree to be badly malformed, then ASIS may crash during the
4886 When @option{-gnatQ} is used and the generated @file{ALI} file is marked as
4887 being in error, @command{gnatmake} will attempt to recompile the source when it
4888 finds such an @file{ALI} file, including with switch @option{-gnatc}.
4890 Note that @option{-gnatQ} has no effect if @option{-gnats} is specified,
4891 since ALI files are never generated if @option{-gnats} is set.
4895 @node Warning Message Control
4896 @subsection Warning Message Control
4897 @cindex Warning messages
4899 In addition to error messages, which correspond to illegalities as defined
4900 in the Ada Reference Manual, the compiler detects two kinds of warning
4903 First, the compiler considers some constructs suspicious and generates a
4904 warning message to alert you to a possible error. Second, if the
4905 compiler detects a situation that is sure to raise an exception at
4906 run time, it generates a warning message. The following shows an example
4907 of warning messages:
4909 e.adb:4:24: warning: creation of object may raise Storage_Error
4910 e.adb:10:17: warning: static value out of range
4911 e.adb:10:17: warning: "Constraint_Error" will be raised at run time
4915 GNAT considers a large number of situations as appropriate
4916 for the generation of warning messages. As always, warnings are not
4917 definite indications of errors. For example, if you do an out-of-range
4918 assignment with the deliberate intention of raising a
4919 @code{Constraint_Error} exception, then the warning that may be
4920 issued does not indicate an error. Some of the situations for which GNAT
4921 issues warnings (at least some of the time) are given in the following
4922 list. This list is not complete, and new warnings are often added to
4923 subsequent versions of GNAT. The list is intended to give a general idea
4924 of the kinds of warnings that are generated.
4928 Possible infinitely recursive calls
4931 Out-of-range values being assigned
4934 Possible order of elaboration problems
4937 Assertions (pragma Assert) that are sure to fail
4943 Address clauses with possibly unaligned values, or where an attempt is
4944 made to overlay a smaller variable with a larger one.
4947 Fixed-point type declarations with a null range
4950 Direct_IO or Sequential_IO instantiated with a type that has access values
4953 Variables that are never assigned a value
4956 Variables that are referenced before being initialized
4959 Task entries with no corresponding @code{accept} statement
4962 Duplicate accepts for the same task entry in a @code{select}
4965 Objects that take too much storage
4968 Unchecked conversion between types of differing sizes
4971 Missing @code{return} statement along some execution path in a function
4974 Incorrect (unrecognized) pragmas
4977 Incorrect external names
4980 Allocation from empty storage pool
4983 Potentially blocking operation in protected type
4986 Suspicious parenthesization of expressions
4989 Mismatching bounds in an aggregate
4992 Attempt to return local value by reference
4995 Premature instantiation of a generic body
4998 Attempt to pack aliased components
5001 Out of bounds array subscripts
5004 Wrong length on string assignment
5007 Violations of style rules if style checking is enabled
5010 Unused @code{with} clauses
5013 @code{Bit_Order} usage that does not have any effect
5016 @code{Standard.Duration} used to resolve universal fixed expression
5019 Dereference of possibly null value
5022 Declaration that is likely to cause storage error
5025 Internal GNAT unit @code{with}'ed by application unit
5028 Values known to be out of range at compile time
5031 Unreferenced labels and variables
5034 Address overlays that could clobber memory
5037 Unexpected initialization when address clause present
5040 Bad alignment for address clause
5043 Useless type conversions
5046 Redundant assignment statements and other redundant constructs
5049 Useless exception handlers
5052 Accidental hiding of name by child unit
5055 Access before elaboration detected at compile time
5058 A range in a @code{for} loop that is known to be null or might be null
5063 The following section lists compiler switches that are available
5064 to control the handling of warning messages. It is also possible
5065 to exercise much finer control over what warnings are issued and
5066 suppressed using the GNAT pragma Warnings, @xref{Pragma Warnings,,,
5067 gnat_rm, GNAT Reference manual}.
5072 @emph{Activate all optional errors.}
5073 @cindex @option{-gnatwa} (@command{gcc})
5074 This switch activates most optional warning messages, see remaining list
5075 in this section for details on optional warning messages that can be
5076 individually controlled. The warnings that are not turned on by this
5078 @option{-gnatwd} (implicit dereferencing),
5079 @option{-gnatwh} (hiding),
5080 @option{-gnatwl} (elaboration warnings),
5081 @option{-gnatw.o} (warn on values set by out parameters ignored)
5082 and @option{-gnatwt} (tracking of deleted conditional code).
5083 All other optional warnings are turned on.
5086 @emph{Suppress all optional errors.}
5087 @cindex @option{-gnatwA} (@command{gcc})
5088 This switch suppresses all optional warning messages, see remaining list
5089 in this section for details on optional warning messages that can be
5090 individually controlled.
5093 @emph{Activate warnings on failing assertions.}
5094 @cindex @option{-gnatw.a} (@command{gcc})
5095 @cindex Assert failures
5096 This switch activates warnings for assertions where the compiler can tell at
5097 compile time that the assertion will fail. Note that this warning is given
5098 even if assertions are disabled. The default is that such warnings are
5102 @emph{Suppress warnings on failing assertions.}
5103 @cindex @option{-gnatw.A} (@command{gcc})
5104 @cindex Assert failures
5105 This switch suppresses warnings for assertions where the compiler can tell at
5106 compile time that the assertion will fail.
5109 @emph{Activate warnings on bad fixed values.}
5110 @cindex @option{-gnatwb} (@command{gcc})
5111 @cindex Bad fixed values
5112 @cindex Fixed-point Small value
5114 This switch activates warnings for static fixed-point expressions whose
5115 value is not an exact multiple of Small. Such values are implementation
5116 dependent, since an implementation is free to choose either of the multiples
5117 that surround the value. GNAT always chooses the closer one, but this is not
5118 required behavior, and it is better to specify a value that is an exact
5119 multiple, ensuring predictable execution. The default is that such warnings
5123 @emph{Suppress warnings on bad fixed values.}
5124 @cindex @option{-gnatwB} (@command{gcc})
5125 This switch suppresses warnings for static fixed-point expressions whose
5126 value is not an exact multiple of Small.
5129 @emph{Activate warnings on biased representation.}
5130 @cindex @option{-gnatw.b} (@command{gcc})
5131 @cindex Biased representation
5132 This switch activates warnings when a size clause, value size clause, component
5133 clause, or component size clause forces the use of biased representation for an
5134 integer type (e.g. representing a range of 10..11 in a single bit by using 0/1
5135 to represent 10/11). The default is that such warnings are generated.
5138 @emph{Suppress warnings on biased representation.}
5139 @cindex @option{-gnatwB} (@command{gcc})
5140 This switch suppresses warnings for representation clauses that force the use
5141 of biased representation.
5144 @emph{Activate warnings on conditionals.}
5145 @cindex @option{-gnatwc} (@command{gcc})
5146 @cindex Conditionals, constant
5147 This switch activates warnings for conditional expressions used in
5148 tests that are known to be True or False at compile time. The default
5149 is that such warnings are not generated.
5150 Note that this warning does
5151 not get issued for the use of boolean variables or constants whose
5152 values are known at compile time, since this is a standard technique
5153 for conditional compilation in Ada, and this would generate too many
5154 false positive warnings.
5156 This warning option also activates a special test for comparisons using
5157 the operators ``>='' and`` <=''.
5158 If the compiler can tell that only the equality condition is possible,
5159 then it will warn that the ``>'' or ``<'' part of the test
5160 is useless and that the operator could be replaced by ``=''.
5161 An example would be comparing a @code{Natural} variable <= 0.
5163 This warning option also generates warnings if
5164 one or both tests is optimized away in a membership test for integer
5165 values if the result can be determined at compile time. Range tests on
5166 enumeration types are not included, since it is common for such tests
5167 to include an end point.
5169 This warning can also be turned on using @option{-gnatwa}.
5172 @emph{Suppress warnings on conditionals.}
5173 @cindex @option{-gnatwC} (@command{gcc})
5174 This switch suppresses warnings for conditional expressions used in
5175 tests that are known to be True or False at compile time.
5178 @emph{Activate warnings on missing component clauses.}
5179 @cindex @option{-gnatw.c} (@command{gcc})
5180 @cindex Component clause, missing
5181 This switch activates warnings for record components where a record
5182 representation clause is present and has component clauses for the
5183 majority, but not all, of the components. A warning is given for each
5184 component for which no component clause is present.
5186 This warning can also be turned on using @option{-gnatwa}.
5189 @emph{Suppress warnings on missing component clauses.}
5190 @cindex @option{-gnatwC} (@command{gcc})
5191 This switch suppresses warnings for record components that are
5192 missing a component clause in the situation described above.
5195 @emph{Activate warnings on implicit dereferencing.}
5196 @cindex @option{-gnatwd} (@command{gcc})
5197 If this switch is set, then the use of a prefix of an access type
5198 in an indexed component, slice, or selected component without an
5199 explicit @code{.all} will generate a warning. With this warning
5200 enabled, access checks occur only at points where an explicit
5201 @code{.all} appears in the source code (assuming no warnings are
5202 generated as a result of this switch). The default is that such
5203 warnings are not generated.
5204 Note that @option{-gnatwa} does not affect the setting of
5205 this warning option.
5208 @emph{Suppress warnings on implicit dereferencing.}
5209 @cindex @option{-gnatwD} (@command{gcc})
5210 @cindex Implicit dereferencing
5211 @cindex Dereferencing, implicit
5212 This switch suppresses warnings for implicit dereferences in
5213 indexed components, slices, and selected components.
5216 @emph{Treat warnings as errors.}
5217 @cindex @option{-gnatwe} (@command{gcc})
5218 @cindex Warnings, treat as error
5219 This switch causes warning messages to be treated as errors.
5220 The warning string still appears, but the warning messages are counted
5221 as errors, and prevent the generation of an object file.
5224 @emph{Activate every optional warning}
5225 @cindex @option{-gnatw.e} (@command{gcc})
5226 @cindex Warnings, activate every optional warning
5227 This switch activates all optional warnings, including those which
5228 are not activated by @code{-gnatwa}.
5231 @emph{Activate warnings on unreferenced formals.}
5232 @cindex @option{-gnatwf} (@command{gcc})
5233 @cindex Formals, unreferenced
5234 This switch causes a warning to be generated if a formal parameter
5235 is not referenced in the body of the subprogram. This warning can
5236 also be turned on using @option{-gnatwa} or @option{-gnatwu}. The
5237 default is that these warnings are not generated.
5240 @emph{Suppress warnings on unreferenced formals.}
5241 @cindex @option{-gnatwF} (@command{gcc})
5242 This switch suppresses warnings for unreferenced formal
5243 parameters. Note that the
5244 combination @option{-gnatwu} followed by @option{-gnatwF} has the
5245 effect of warning on unreferenced entities other than subprogram
5249 @emph{Activate warnings on unrecognized pragmas.}
5250 @cindex @option{-gnatwg} (@command{gcc})
5251 @cindex Pragmas, unrecognized
5252 This switch causes a warning to be generated if an unrecognized
5253 pragma is encountered. Apart from issuing this warning, the
5254 pragma is ignored and has no effect. This warning can
5255 also be turned on using @option{-gnatwa}. The default
5256 is that such warnings are issued (satisfying the Ada Reference
5257 Manual requirement that such warnings appear).
5260 @emph{Suppress warnings on unrecognized pragmas.}
5261 @cindex @option{-gnatwG} (@command{gcc})
5262 This switch suppresses warnings for unrecognized pragmas.
5265 @emph{Activate warnings on hiding.}
5266 @cindex @option{-gnatwh} (@command{gcc})
5267 @cindex Hiding of Declarations
5268 This switch activates warnings on hiding declarations.
5269 A declaration is considered hiding
5270 if it is for a non-overloadable entity, and it declares an entity with the
5271 same name as some other entity that is directly or use-visible. The default
5272 is that such warnings are not generated.
5273 Note that @option{-gnatwa} does not affect the setting of this warning option.
5276 @emph{Suppress warnings on hiding.}
5277 @cindex @option{-gnatwH} (@command{gcc})
5278 This switch suppresses warnings on hiding declarations.
5281 @emph{Activate warnings on implementation units.}
5282 @cindex @option{-gnatwi} (@command{gcc})
5283 This switch activates warnings for a @code{with} of an internal GNAT
5284 implementation unit, defined as any unit from the @code{Ada},
5285 @code{Interfaces}, @code{GNAT},
5286 ^^@code{DEC},^ or @code{System}
5287 hierarchies that is not
5288 documented in either the Ada Reference Manual or the GNAT
5289 Programmer's Reference Manual. Such units are intended only
5290 for internal implementation purposes and should not be @code{with}'ed
5291 by user programs. The default is that such warnings are generated
5292 This warning can also be turned on using @option{-gnatwa}.
5295 @emph{Disable warnings on implementation units.}
5296 @cindex @option{-gnatwI} (@command{gcc})
5297 This switch disables warnings for a @code{with} of an internal GNAT
5298 implementation unit.
5301 @emph{Activate warnings on overlapping actuals.}
5302 @cindex @option{-gnatw.i} (@command{gcc})
5303 This switch enables a warning on statically detectable overlapping actuals in
5304 a subprogram call, when one of the actuals is an in-out parameter, and the
5305 types of the actuals are not by-copy types. The warning is off by default,
5306 and is not included under -gnatwa.
5309 @emph{Disable warnings on overlapping actuals.}
5310 @cindex @option{-gnatw.I} (@command{gcc})
5311 This switch disables warnings on overlapping actuals in a call..
5314 @emph{Activate warnings on obsolescent features (Annex J).}
5315 @cindex @option{-gnatwj} (@command{gcc})
5316 @cindex Features, obsolescent
5317 @cindex Obsolescent features
5318 If this warning option is activated, then warnings are generated for
5319 calls to subprograms marked with @code{pragma Obsolescent} and
5320 for use of features in Annex J of the Ada Reference Manual. In the
5321 case of Annex J, not all features are flagged. In particular use
5322 of the renamed packages (like @code{Text_IO}) and use of package
5323 @code{ASCII} are not flagged, since these are very common and
5324 would generate many annoying positive warnings. The default is that
5325 such warnings are not generated. This warning is also turned on by
5326 the use of @option{-gnatwa}.
5328 In addition to the above cases, warnings are also generated for
5329 GNAT features that have been provided in past versions but which
5330 have been superseded (typically by features in the new Ada standard).
5331 For example, @code{pragma Ravenscar} will be flagged since its
5332 function is replaced by @code{pragma Profile(Ravenscar)}.
5334 Note that this warning option functions differently from the
5335 restriction @code{No_Obsolescent_Features} in two respects.
5336 First, the restriction applies only to annex J features.
5337 Second, the restriction does flag uses of package @code{ASCII}.
5340 @emph{Suppress warnings on obsolescent features (Annex J).}
5341 @cindex @option{-gnatwJ} (@command{gcc})
5342 This switch disables warnings on use of obsolescent features.
5345 @emph{Activate warnings on variables that could be constants.}
5346 @cindex @option{-gnatwk} (@command{gcc})
5347 This switch activates warnings for variables that are initialized but
5348 never modified, and then could be declared constants. The default is that
5349 such warnings are not given.
5350 This warning can also be turned on using @option{-gnatwa}.
5353 @emph{Suppress warnings on variables that could be constants.}
5354 @cindex @option{-gnatwK} (@command{gcc})
5355 This switch disables warnings on variables that could be declared constants.
5358 @emph{Activate warnings for elaboration pragmas.}
5359 @cindex @option{-gnatwl} (@command{gcc})
5360 @cindex Elaboration, warnings
5361 This switch activates warnings on missing
5362 @code{Elaborate_All} and @code{Elaborate} pragmas.
5363 See the section in this guide on elaboration checking for details on
5364 when such pragmas should be used. In dynamic elaboration mode, this switch
5365 generations warnings about the need to add elaboration pragmas. Note however,
5366 that if you blindly follow these warnings, and add @code{Elaborate_All}
5367 warnings wherever they are recommended, you basically end up with the
5368 equivalent of the static elaboration model, which may not be what you want for
5369 legacy code for which the static model does not work.
5371 For the static model, the messages generated are labeled "info:" (for
5372 information messages). They are not warnings to add elaboration pragmas,
5373 merely informational messages showing what implicit elaboration pragmas
5374 have been added, for use in analyzing elaboration circularity problems.
5376 Warnings are also generated if you
5377 are using the static mode of elaboration, and a @code{pragma Elaborate}
5378 is encountered. The default is that such warnings
5380 This warning is not automatically turned on by the use of @option{-gnatwa}.
5383 @emph{Suppress warnings for elaboration pragmas.}
5384 @cindex @option{-gnatwL} (@command{gcc})
5385 This switch suppresses warnings on missing Elaborate and Elaborate_All pragmas.
5386 See the section in this guide on elaboration checking for details on
5387 when such pragmas should be used.
5390 @emph{Activate warnings on modified but unreferenced variables.}
5391 @cindex @option{-gnatwm} (@command{gcc})
5392 This switch activates warnings for variables that are assigned (using
5393 an initialization value or with one or more assignment statements) but
5394 whose value is never read. The warning is suppressed for volatile
5395 variables and also for variables that are renamings of other variables
5396 or for which an address clause is given.
5397 This warning can also be turned on using @option{-gnatwa}.
5398 The default is that these warnings are not given.
5401 @emph{Disable warnings on modified but unreferenced variables.}
5402 @cindex @option{-gnatwM} (@command{gcc})
5403 This switch disables warnings for variables that are assigned or
5404 initialized, but never read.
5407 @emph{Activate warnings on suspicious modulus values.}
5408 @cindex @option{-gnatw.m} (@command{gcc})
5409 This switch activates warnings for modulus values that seem suspicious.
5410 The cases caught are where the size is the same as the modulus (e.g.
5411 a modulus of 7 with a size of 7 bits), and modulus values of 32 or 64
5412 with no size clause. The guess in both cases is that 2**x was intended
5413 rather than x. The default is that these warnings are given.
5416 @emph{Disable warnings on suspicious modulus values.}
5417 @cindex @option{-gnatw.M} (@command{gcc})
5418 This switch disables warnings for suspicious modulus values.
5421 @emph{Set normal warnings mode.}
5422 @cindex @option{-gnatwn} (@command{gcc})
5423 This switch sets normal warning mode, in which enabled warnings are
5424 issued and treated as warnings rather than errors. This is the default
5425 mode. the switch @option{-gnatwn} can be used to cancel the effect of
5426 an explicit @option{-gnatws} or
5427 @option{-gnatwe}. It also cancels the effect of the
5428 implicit @option{-gnatwe} that is activated by the
5429 use of @option{-gnatg}.
5432 @emph{Activate warnings on address clause overlays.}
5433 @cindex @option{-gnatwo} (@command{gcc})
5434 @cindex Address Clauses, warnings
5435 This switch activates warnings for possibly unintended initialization
5436 effects of defining address clauses that cause one variable to overlap
5437 another. The default is that such warnings are generated.
5438 This warning can also be turned on using @option{-gnatwa}.
5441 @emph{Suppress warnings on address clause overlays.}
5442 @cindex @option{-gnatwO} (@command{gcc})
5443 This switch suppresses warnings on possibly unintended initialization
5444 effects of defining address clauses that cause one variable to overlap
5448 @emph{Activate warnings on modified but unreferenced out parameters.}
5449 @cindex @option{-gnatw.o} (@command{gcc})
5450 This switch activates warnings for variables that are modified by using
5451 them as actuals for a call to a procedure with an out mode formal, where
5452 the resulting assigned value is never read. It is applicable in the case
5453 where there is more than one out mode formal. If there is only one out
5454 mode formal, the warning is issued by default (controlled by -gnatwu).
5455 The warning is suppressed for volatile
5456 variables and also for variables that are renamings of other variables
5457 or for which an address clause is given.
5458 The default is that these warnings are not given. Note that this warning
5459 is not included in -gnatwa, it must be activated explicitly.
5462 @emph{Disable warnings on modified but unreferenced out parameters.}
5463 @cindex @option{-gnatw.O} (@command{gcc})
5464 This switch suppresses warnings for variables that are modified by using
5465 them as actuals for a call to a procedure with an out mode formal, where
5466 the resulting assigned value is never read.
5469 @emph{Activate warnings on ineffective pragma Inlines.}
5470 @cindex @option{-gnatwp} (@command{gcc})
5471 @cindex Inlining, warnings
5472 This switch activates warnings for failure of front end inlining
5473 (activated by @option{-gnatN}) to inline a particular call. There are
5474 many reasons for not being able to inline a call, including most
5475 commonly that the call is too complex to inline. The default is
5476 that such warnings are not given.
5477 This warning can also be turned on using @option{-gnatwa}.
5478 Warnings on ineffective inlining by the gcc back-end can be activated
5479 separately, using the gcc switch -Winline.
5482 @emph{Suppress warnings on ineffective pragma Inlines.}
5483 @cindex @option{-gnatwP} (@command{gcc})
5484 This switch suppresses warnings on ineffective pragma Inlines. If the
5485 inlining mechanism cannot inline a call, it will simply ignore the
5489 @emph{Activate warnings on parameter ordering.}
5490 @cindex @option{-gnatw.p} (@command{gcc})
5491 @cindex Parameter order, warnings
5492 This switch activates warnings for cases of suspicious parameter
5493 ordering when the list of arguments are all simple identifiers that
5494 match the names of the formals, but are in a different order. The
5495 warning is suppressed if any use of named parameter notation is used,
5496 so this is the appropriate way to suppress a false positive (and
5497 serves to emphasize that the "misordering" is deliberate). The
5499 that such warnings are not given.
5500 This warning can also be turned on using @option{-gnatwa}.
5503 @emph{Suppress warnings on parameter ordering.}
5504 @cindex @option{-gnatw.P} (@command{gcc})
5505 This switch suppresses warnings on cases of suspicious parameter
5509 @emph{Activate warnings on questionable missing parentheses.}
5510 @cindex @option{-gnatwq} (@command{gcc})
5511 @cindex Parentheses, warnings
5512 This switch activates warnings for cases where parentheses are not used and
5513 the result is potential ambiguity from a readers point of view. For example
5514 (not a > b) when a and b are modular means ((not a) > b) and very likely the
5515 programmer intended (not (a > b)). Similarly (-x mod 5) means (-(x mod 5)) and
5516 quite likely ((-x) mod 5) was intended. In such situations it seems best to
5517 follow the rule of always parenthesizing to make the association clear, and
5518 this warning switch warns if such parentheses are not present. The default
5519 is that these warnings are given.
5520 This warning can also be turned on using @option{-gnatwa}.
5523 @emph{Suppress warnings on questionable missing parentheses.}
5524 @cindex @option{-gnatwQ} (@command{gcc})
5525 This switch suppresses warnings for cases where the association is not
5526 clear and the use of parentheses is preferred.
5529 @emph{Activate warnings on redundant constructs.}
5530 @cindex @option{-gnatwr} (@command{gcc})
5531 This switch activates warnings for redundant constructs. The following
5532 is the current list of constructs regarded as redundant:
5536 Assignment of an item to itself.
5538 Type conversion that converts an expression to its own type.
5540 Use of the attribute @code{Base} where @code{typ'Base} is the same
5543 Use of pragma @code{Pack} when all components are placed by a record
5544 representation clause.
5546 Exception handler containing only a reraise statement (raise with no
5547 operand) which has no effect.
5549 Use of the operator abs on an operand that is known at compile time
5552 Comparison of boolean expressions to an explicit True value.
5555 This warning can also be turned on using @option{-gnatwa}.
5556 The default is that warnings for redundant constructs are not given.
5559 @emph{Suppress warnings on redundant constructs.}
5560 @cindex @option{-gnatwR} (@command{gcc})
5561 This switch suppresses warnings for redundant constructs.
5564 @emph{Activate warnings for object renaming function.}
5565 @cindex @option{-gnatw.r} (@command{gcc})
5566 This switch activates warnings for an object renaming that renames a
5567 function call, which is equivalent to a constant declaration (as
5568 opposed to renaming the function itself). The default is that these
5569 warnings are given. This warning can also be turned on using
5573 @emph{Suppress warnings for object renaming function.}
5574 @cindex @option{-gnatwT} (@command{gcc})
5575 This switch suppresses warnings for object renaming function.
5578 @emph{Suppress all warnings.}
5579 @cindex @option{-gnatws} (@command{gcc})
5580 This switch completely suppresses the
5581 output of all warning messages from the GNAT front end.
5582 Note that it does not suppress warnings from the @command{gcc} back end.
5583 To suppress these back end warnings as well, use the switch @option{-w}
5584 in addition to @option{-gnatws}.
5587 @emph{Activate warnings for tracking of deleted conditional code.}
5588 @cindex @option{-gnatwt} (@command{gcc})
5589 @cindex Deactivated code, warnings
5590 @cindex Deleted code, warnings
5591 This switch activates warnings for tracking of code in conditionals (IF and
5592 CASE statements) that is detected to be dead code which cannot be executed, and
5593 which is removed by the front end. This warning is off by default, and is not
5594 turned on by @option{-gnatwa}, it has to be turned on explicitly. This may be
5595 useful for detecting deactivated code in certified applications.
5598 @emph{Suppress warnings for tracking of deleted conditional code.}
5599 @cindex @option{-gnatwT} (@command{gcc})
5600 This switch suppresses warnings for tracking of deleted conditional code.
5603 @emph{Activate warnings on unused entities.}
5604 @cindex @option{-gnatwu} (@command{gcc})
5605 This switch activates warnings to be generated for entities that
5606 are declared but not referenced, and for units that are @code{with}'ed
5608 referenced. In the case of packages, a warning is also generated if
5609 no entities in the package are referenced. This means that if the package
5610 is referenced but the only references are in @code{use}
5611 clauses or @code{renames}
5612 declarations, a warning is still generated. A warning is also generated
5613 for a generic package that is @code{with}'ed but never instantiated.
5614 In the case where a package or subprogram body is compiled, and there
5615 is a @code{with} on the corresponding spec
5616 that is only referenced in the body,
5617 a warning is also generated, noting that the
5618 @code{with} can be moved to the body. The default is that
5619 such warnings are not generated.
5620 This switch also activates warnings on unreferenced formals
5621 (it includes the effect of @option{-gnatwf}).
5622 This warning can also be turned on using @option{-gnatwa}.
5625 @emph{Suppress warnings on unused entities.}
5626 @cindex @option{-gnatwU} (@command{gcc})
5627 This switch suppresses warnings for unused entities and packages.
5628 It also turns off warnings on unreferenced formals (and thus includes
5629 the effect of @option{-gnatwF}).
5632 @emph{Activate warnings on unassigned variables.}
5633 @cindex @option{-gnatwv} (@command{gcc})
5634 @cindex Unassigned variable warnings
5635 This switch activates warnings for access to variables which
5636 may not be properly initialized. The default is that
5637 such warnings are generated.
5638 This warning can also be turned on using @option{-gnatwa}.
5641 @emph{Suppress warnings on unassigned variables.}
5642 @cindex @option{-gnatwV} (@command{gcc})
5643 This switch suppresses warnings for access to variables which
5644 may not be properly initialized.
5645 For variables of a composite type, the warning can also be suppressed in
5646 Ada 2005 by using a default initialization with a box. For example, if
5647 Table is an array of records whose components are only partially uninitialized,
5648 then the following code:
5650 @smallexample @c ada
5651 Tab : Table := (others => <>);
5654 will suppress warnings on subsequent statements that access components
5658 @emph{Activate warnings on wrong low bound assumption.}
5659 @cindex @option{-gnatww} (@command{gcc})
5660 @cindex String indexing warnings
5661 This switch activates warnings for indexing an unconstrained string parameter
5662 with a literal or S'Length. This is a case where the code is assuming that the
5663 low bound is one, which is in general not true (for example when a slice is
5664 passed). The default is that such warnings are generated.
5665 This warning can also be turned on using @option{-gnatwa}.
5668 @emph{Suppress warnings on wrong low bound assumption.}
5669 @cindex @option{-gnatwW} (@command{gcc})
5670 This switch suppresses warnings for indexing an unconstrained string parameter
5671 with a literal or S'Length. Note that this warning can also be suppressed
5672 in a particular case by adding an
5673 assertion that the lower bound is 1,
5674 as shown in the following example.
5676 @smallexample @c ada
5677 procedure K (S : String) is
5678 pragma Assert (S'First = 1);
5683 @emph{Activate warnings on unnecessary Warnings Off pragmas}
5684 @cindex @option{-gnatw.w} (@command{gcc})
5685 @cindex Warnings Off control
5686 This switch activates warnings for use of @code{pragma Warnings (Off, entity}
5687 where either the pragma is entirely useless (because it suppresses no
5688 warnings), or it could be replaced by @code{pragma Unreferenced} or
5689 @code{pragma Unmodified}.The default is that these warnings are not given.
5690 Note that this warning is not included in -gnatwa, it must be
5691 activated explicitly.
5694 @emph{Suppress warnings on unnecessary Warnings Off pragmas}
5695 @cindex @option{-gnatw.W} (@command{gcc})
5696 This switch suppresses warnings for use of @code{pragma Warnings (Off, entity}.
5699 @emph{Activate warnings on Export/Import pragmas.}
5700 @cindex @option{-gnatwx} (@command{gcc})
5701 @cindex Export/Import pragma warnings
5702 This switch activates warnings on Export/Import pragmas when
5703 the compiler detects a possible conflict between the Ada and
5704 foreign language calling sequences. For example, the use of
5705 default parameters in a convention C procedure is dubious
5706 because the C compiler cannot supply the proper default, so
5707 a warning is issued. The default is that such warnings are
5709 This warning can also be turned on using @option{-gnatwa}.
5712 @emph{Suppress warnings on Export/Import pragmas.}
5713 @cindex @option{-gnatwX} (@command{gcc})
5714 This switch suppresses warnings on Export/Import pragmas.
5715 The sense of this is that you are telling the compiler that
5716 you know what you are doing in writing the pragma, and it
5717 should not complain at you.
5720 @emph{Activate warnings for No_Exception_Propagation mode.}
5721 @cindex @option{-gnatwm} (@command{gcc})
5722 This switch activates warnings for exception usage when pragma Restrictions
5723 (No_Exception_Propagation) is in effect. Warnings are given for implicit or
5724 explicit exception raises which are not covered by a local handler, and for
5725 exception handlers which do not cover a local raise. The default is that these
5726 warnings are not given.
5729 @emph{Disable warnings for No_Exception_Propagation mode.}
5730 This switch disables warnings for exception usage when pragma Restrictions
5731 (No_Exception_Propagation) is in effect.
5734 @emph{Activate warnings for Ada 2005 compatibility issues.}
5735 @cindex @option{-gnatwy} (@command{gcc})
5736 @cindex Ada 2005 compatibility issues warnings
5737 For the most part Ada 2005 is upwards compatible with Ada 95,
5738 but there are some exceptions (for example the fact that
5739 @code{interface} is now a reserved word in Ada 2005). This
5740 switch activates several warnings to help in identifying
5741 and correcting such incompatibilities. The default is that
5742 these warnings are generated. Note that at one point Ada 2005
5743 was called Ada 0Y, hence the choice of character.
5744 This warning can also be turned on using @option{-gnatwa}.
5747 @emph{Disable warnings for Ada 2005 compatibility issues.}
5748 @cindex @option{-gnatwY} (@command{gcc})
5749 @cindex Ada 2005 compatibility issues warnings
5750 This switch suppresses several warnings intended to help in identifying
5751 incompatibilities between Ada 95 and Ada 2005.
5754 @emph{Activate warnings on unchecked conversions.}
5755 @cindex @option{-gnatwz} (@command{gcc})
5756 @cindex Unchecked_Conversion warnings
5757 This switch activates warnings for unchecked conversions
5758 where the types are known at compile time to have different
5760 is that such warnings are generated. Warnings are also
5761 generated for subprogram pointers with different conventions,
5762 and, on VMS only, for data pointers with different conventions.
5763 This warning can also be turned on using @option{-gnatwa}.
5766 @emph{Suppress warnings on unchecked conversions.}
5767 @cindex @option{-gnatwZ} (@command{gcc})
5768 This switch suppresses warnings for unchecked conversions
5769 where the types are known at compile time to have different
5770 sizes or conventions.
5772 @item ^-Wunused^WARNINGS=UNUSED^
5773 @cindex @option{-Wunused}
5774 The warnings controlled by the @option{-gnatw} switch are generated by
5775 the front end of the compiler. The @option{GCC} back end can provide
5776 additional warnings and they are controlled by the @option{-W} switch.
5777 For example, @option{^-Wunused^WARNINGS=UNUSED^} activates back end
5778 warnings for entities that are declared but not referenced.
5780 @item ^-Wuninitialized^WARNINGS=UNINITIALIZED^
5781 @cindex @option{-Wuninitialized}
5782 Similarly, @option{^-Wuninitialized^WARNINGS=UNINITIALIZED^} activates
5783 the back end warning for uninitialized variables. This switch must be
5784 used in conjunction with an optimization level greater than zero.
5786 @item ^-Wall^/ALL_BACK_END_WARNINGS^
5787 @cindex @option{-Wall}
5788 This switch enables all the above warnings from the @option{GCC} back end.
5789 The code generator detects a number of warning situations that are missed
5790 by the @option{GNAT} front end, and this switch can be used to activate them.
5791 The use of this switch also sets the default front end warning mode to
5792 @option{-gnatwa}, that is, most front end warnings activated as well.
5794 @item ^-w^/NO_BACK_END_WARNINGS^
5796 Conversely, this switch suppresses warnings from the @option{GCC} back end.
5797 The use of this switch also sets the default front end warning mode to
5798 @option{-gnatws}, that is, front end warnings suppressed as well.
5804 A string of warning parameters can be used in the same parameter. For example:
5811 will turn on all optional warnings except for elaboration pragma warnings,
5812 and also specify that warnings should be treated as errors.
5814 When no switch @option{^-gnatw^/WARNINGS^} is used, this is equivalent to:
5839 @node Debugging and Assertion Control
5840 @subsection Debugging and Assertion Control
5844 @cindex @option{-gnata} (@command{gcc})
5850 The pragmas @code{Assert} and @code{Debug} normally have no effect and
5851 are ignored. This switch, where @samp{a} stands for assert, causes
5852 @code{Assert} and @code{Debug} pragmas to be activated.
5854 The pragmas have the form:
5858 @b{pragma} Assert (@var{Boolean-expression} @r{[},
5859 @var{static-string-expression}@r{]})
5860 @b{pragma} Debug (@var{procedure call})
5865 The @code{Assert} pragma causes @var{Boolean-expression} to be tested.
5866 If the result is @code{True}, the pragma has no effect (other than
5867 possible side effects from evaluating the expression). If the result is
5868 @code{False}, the exception @code{Assert_Failure} declared in the package
5869 @code{System.Assertions} is
5870 raised (passing @var{static-string-expression}, if present, as the
5871 message associated with the exception). If no string expression is
5872 given the default is a string giving the file name and line number
5875 The @code{Debug} pragma causes @var{procedure} to be called. Note that
5876 @code{pragma Debug} may appear within a declaration sequence, allowing
5877 debugging procedures to be called between declarations.
5880 @item /DEBUG@r{[}=debug-level@r{]}
5882 Specifies how much debugging information is to be included in
5883 the resulting object file where 'debug-level' is one of the following:
5886 Include both debugger symbol records and traceback
5888 This is the default setting.
5890 Include both debugger symbol records and traceback in
5893 Excludes both debugger symbol records and traceback
5894 the object file. Same as /NODEBUG.
5896 Includes only debugger symbol records in the object
5897 file. Note that this doesn't include traceback information.
5902 @node Validity Checking
5903 @subsection Validity Checking
5904 @findex Validity Checking
5907 The Ada Reference Manual defines the concept of invalid values (see
5908 RM 13.9.1). The primary source of invalid values is uninitialized
5909 variables. A scalar variable that is left uninitialized may contain
5910 an invalid value; the concept of invalid does not apply to access or
5913 It is an error to read an invalid value, but the RM does not require
5914 run-time checks to detect such errors, except for some minimal
5915 checking to prevent erroneous execution (i.e. unpredictable
5916 behavior). This corresponds to the @option{-gnatVd} switch below,
5917 which is the default. For example, by default, if the expression of a
5918 case statement is invalid, it will raise Constraint_Error rather than
5919 causing a wild jump, and if an array index on the left-hand side of an
5920 assignment is invalid, it will raise Constraint_Error rather than
5921 overwriting an arbitrary memory location.
5923 The @option{-gnatVa} may be used to enable additional validity checks,
5924 which are not required by the RM. These checks are often very
5925 expensive (which is why the RM does not require them). These checks
5926 are useful in tracking down uninitialized variables, but they are
5927 not usually recommended for production builds.
5929 The other @option{-gnatV^@var{x}^^} switches below allow finer-grained
5930 control; you can enable whichever validity checks you desire. However,
5931 for most debugging purposes, @option{-gnatVa} is sufficient, and the
5932 default @option{-gnatVd} (i.e. standard Ada behavior) is usually
5933 sufficient for non-debugging use.
5935 The @option{-gnatB} switch tells the compiler to assume that all
5936 values are valid (that is, within their declared subtype range)
5937 except in the context of a use of the Valid attribute. This means
5938 the compiler can generate more efficient code, since the range
5939 of values is better known at compile time. However, an uninitialized
5940 variable can cause wild jumps and memory corruption in this mode.
5942 The @option{-gnatV^@var{x}^^} switch allows control over the validity
5943 checking mode as described below.
5945 The @code{x} argument is a string of letters that
5946 indicate validity checks that are performed or not performed in addition
5947 to the default checks required by Ada as described above.
5950 The options allowed for this qualifier
5951 indicate validity checks that are performed or not performed in addition
5952 to the default checks required by Ada as described above.
5958 @emph{All validity checks.}
5959 @cindex @option{-gnatVa} (@command{gcc})
5960 All validity checks are turned on.
5962 That is, @option{-gnatVa} is
5963 equivalent to @option{gnatVcdfimorst}.
5967 @emph{Validity checks for copies.}
5968 @cindex @option{-gnatVc} (@command{gcc})
5969 The right hand side of assignments, and the initializing values of
5970 object declarations are validity checked.
5973 @emph{Default (RM) validity checks.}
5974 @cindex @option{-gnatVd} (@command{gcc})
5975 Some validity checks are done by default following normal Ada semantics
5977 A check is done in case statements that the expression is within the range
5978 of the subtype. If it is not, Constraint_Error is raised.
5979 For assignments to array components, a check is done that the expression used
5980 as index is within the range. If it is not, Constraint_Error is raised.
5981 Both these validity checks may be turned off using switch @option{-gnatVD}.
5982 They are turned on by default. If @option{-gnatVD} is specified, a subsequent
5983 switch @option{-gnatVd} will leave the checks turned on.
5984 Switch @option{-gnatVD} should be used only if you are sure that all such
5985 expressions have valid values. If you use this switch and invalid values
5986 are present, then the program is erroneous, and wild jumps or memory
5987 overwriting may occur.
5990 @emph{Validity checks for elementary components.}
5991 @cindex @option{-gnatVe} (@command{gcc})
5992 In the absence of this switch, assignments to record or array components are
5993 not validity checked, even if validity checks for assignments generally
5994 (@option{-gnatVc}) are turned on. In Ada, assignment of composite values do not
5995 require valid data, but assignment of individual components does. So for
5996 example, there is a difference between copying the elements of an array with a
5997 slice assignment, compared to assigning element by element in a loop. This
5998 switch allows you to turn off validity checking for components, even when they
5999 are assigned component by component.
6002 @emph{Validity checks for floating-point values.}
6003 @cindex @option{-gnatVf} (@command{gcc})
6004 In the absence of this switch, validity checking occurs only for discrete
6005 values. If @option{-gnatVf} is specified, then validity checking also applies
6006 for floating-point values, and NaNs and infinities are considered invalid,
6007 as well as out of range values for constrained types. Note that this means
6008 that standard IEEE infinity mode is not allowed. The exact contexts
6009 in which floating-point values are checked depends on the setting of other
6010 options. For example,
6011 @option{^-gnatVif^VALIDITY_CHECKING=(IN_PARAMS,FLOATS)^} or
6012 @option{^-gnatVfi^VALIDITY_CHECKING=(FLOATS,IN_PARAMS)^}
6013 (the order does not matter) specifies that floating-point parameters of mode
6014 @code{in} should be validity checked.
6017 @emph{Validity checks for @code{in} mode parameters}
6018 @cindex @option{-gnatVi} (@command{gcc})
6019 Arguments for parameters of mode @code{in} are validity checked in function
6020 and procedure calls at the point of call.
6023 @emph{Validity checks for @code{in out} mode parameters.}
6024 @cindex @option{-gnatVm} (@command{gcc})
6025 Arguments for parameters of mode @code{in out} are validity checked in
6026 procedure calls at the point of call. The @code{'m'} here stands for
6027 modify, since this concerns parameters that can be modified by the call.
6028 Note that there is no specific option to test @code{out} parameters,
6029 but any reference within the subprogram will be tested in the usual
6030 manner, and if an invalid value is copied back, any reference to it
6031 will be subject to validity checking.
6034 @emph{No validity checks.}
6035 @cindex @option{-gnatVn} (@command{gcc})
6036 This switch turns off all validity checking, including the default checking
6037 for case statements and left hand side subscripts. Note that the use of
6038 the switch @option{-gnatp} suppresses all run-time checks, including
6039 validity checks, and thus implies @option{-gnatVn}. When this switch
6040 is used, it cancels any other @option{-gnatV} previously issued.
6043 @emph{Validity checks for operator and attribute operands.}
6044 @cindex @option{-gnatVo} (@command{gcc})
6045 Arguments for predefined operators and attributes are validity checked.
6046 This includes all operators in package @code{Standard},
6047 the shift operators defined as intrinsic in package @code{Interfaces}
6048 and operands for attributes such as @code{Pos}. Checks are also made
6049 on individual component values for composite comparisons, and on the
6050 expressions in type conversions and qualified expressions. Checks are
6051 also made on explicit ranges using @samp{..} (e.g.@: slices, loops etc).
6054 @emph{Validity checks for parameters.}
6055 @cindex @option{-gnatVp} (@command{gcc})
6056 This controls the treatment of parameters within a subprogram (as opposed
6057 to @option{-gnatVi} and @option{-gnatVm} which control validity testing
6058 of parameters on a call. If either of these call options is used, then
6059 normally an assumption is made within a subprogram that the input arguments
6060 have been validity checking at the point of call, and do not need checking
6061 again within a subprogram). If @option{-gnatVp} is set, then this assumption
6062 is not made, and parameters are not assumed to be valid, so their validity
6063 will be checked (or rechecked) within the subprogram.
6066 @emph{Validity checks for function returns.}
6067 @cindex @option{-gnatVr} (@command{gcc})
6068 The expression in @code{return} statements in functions is validity
6072 @emph{Validity checks for subscripts.}
6073 @cindex @option{-gnatVs} (@command{gcc})
6074 All subscripts expressions are checked for validity, whether they appear
6075 on the right side or left side (in default mode only left side subscripts
6076 are validity checked).
6079 @emph{Validity checks for tests.}
6080 @cindex @option{-gnatVt} (@command{gcc})
6081 Expressions used as conditions in @code{if}, @code{while} or @code{exit}
6082 statements are checked, as well as guard expressions in entry calls.
6087 The @option{-gnatV} switch may be followed by
6088 ^a string of letters^a list of options^
6089 to turn on a series of validity checking options.
6091 @option{^-gnatVcr^/VALIDITY_CHECKING=(COPIES, RETURNS)^}
6092 specifies that in addition to the default validity checking, copies and
6093 function return expressions are to be validity checked.
6094 In order to make it easier
6095 to specify the desired combination of effects,
6097 the upper case letters @code{CDFIMORST} may
6098 be used to turn off the corresponding lower case option.
6101 the prefix @code{NO} on an option turns off the corresponding validity
6104 @item @code{NOCOPIES}
6105 @item @code{NODEFAULT}
6106 @item @code{NOFLOATS}
6107 @item @code{NOIN_PARAMS}
6108 @item @code{NOMOD_PARAMS}
6109 @item @code{NOOPERANDS}
6110 @item @code{NORETURNS}
6111 @item @code{NOSUBSCRIPTS}
6112 @item @code{NOTESTS}
6116 @option{^-gnatVaM^/VALIDITY_CHECKING=(ALL, NOMOD_PARAMS)^}
6117 turns on all validity checking options except for
6118 checking of @code{@b{in out}} procedure arguments.
6120 The specification of additional validity checking generates extra code (and
6121 in the case of @option{-gnatVa} the code expansion can be substantial).
6122 However, these additional checks can be very useful in detecting
6123 uninitialized variables, incorrect use of unchecked conversion, and other
6124 errors leading to invalid values. The use of pragma @code{Initialize_Scalars}
6125 is useful in conjunction with the extra validity checking, since this
6126 ensures that wherever possible uninitialized variables have invalid values.
6128 See also the pragma @code{Validity_Checks} which allows modification of
6129 the validity checking mode at the program source level, and also allows for
6130 temporary disabling of validity checks.
6132 @node Style Checking
6133 @subsection Style Checking
6134 @findex Style checking
6137 The @option{-gnaty^x^(option,option,@dots{})^} switch
6138 @cindex @option{-gnaty} (@command{gcc})
6139 causes the compiler to
6140 enforce specified style rules. A limited set of style rules has been used
6141 in writing the GNAT sources themselves. This switch allows user programs
6142 to activate all or some of these checks. If the source program fails a
6143 specified style check, an appropriate warning message is given, preceded by
6144 the character sequence ``(style)''.
6146 @code{(option,option,@dots{})} is a sequence of keywords
6149 The string @var{x} is a sequence of letters or digits
6151 indicating the particular style
6152 checks to be performed. The following checks are defined:
6157 @emph{Specify indentation level.}
6158 If a digit from 1-9 appears
6159 ^in the string after @option{-gnaty}^as an option for /STYLE_CHECKS^
6160 then proper indentation is checked, with the digit indicating the
6161 indentation level required. A value of zero turns off this style check.
6162 The general style of required indentation is as specified by
6163 the examples in the Ada Reference Manual. Full line comments must be
6164 aligned with the @code{--} starting on a column that is a multiple of
6165 the alignment level, or they may be aligned the same way as the following
6166 non-blank line (this is useful when full line comments appear in the middle
6170 @emph{Check attribute casing.}
6171 Attribute names, including the case of keywords such as @code{digits}
6172 used as attributes names, must be written in mixed case, that is, the
6173 initial letter and any letter following an underscore must be uppercase.
6174 All other letters must be lowercase.
6176 @item ^A^ARRAY_INDEXES^
6177 @emph{Use of array index numbers in array attributes.}
6178 When using the array attributes First, Last, Range,
6179 or Length, the index number must be omitted for one-dimensional arrays
6180 and is required for multi-dimensional arrays.
6183 @emph{Blanks not allowed at statement end.}
6184 Trailing blanks are not allowed at the end of statements. The purpose of this
6185 rule, together with h (no horizontal tabs), is to enforce a canonical format
6186 for the use of blanks to separate source tokens.
6188 @item ^B^BOOLEAN_OPERATORS^
6189 @emph{Check Boolean operators.}
6190 The use of AND/OR operators is not permitted except in the cases of modular
6191 operands, array operands, and simple stand-alone boolean variables or
6192 boolean constants. In all other cases AND THEN/OR ELSE are required.
6195 @emph{Check comments.}
6196 Comments must meet the following set of rules:
6201 The ``@code{--}'' that starts the column must either start in column one,
6202 or else at least one blank must precede this sequence.
6205 Comments that follow other tokens on a line must have at least one blank
6206 following the ``@code{--}'' at the start of the comment.
6209 Full line comments must have two blanks following the ``@code{--}'' that
6210 starts the comment, with the following exceptions.
6213 A line consisting only of the ``@code{--}'' characters, possibly preceded
6214 by blanks is permitted.
6217 A comment starting with ``@code{--x}'' where @code{x} is a special character
6219 This allows proper processing of the output generated by specialized tools
6220 including @command{gnatprep} (where ``@code{--!}'' is used) and the SPARK
6222 language (where ``@code{--#}'' is used). For the purposes of this rule, a
6223 special character is defined as being in one of the ASCII ranges
6224 @code{16#21#@dots{}16#2F#} or @code{16#3A#@dots{}16#3F#}.
6225 Note that this usage is not permitted
6226 in GNAT implementation units (i.e., when @option{-gnatg} is used).
6229 A line consisting entirely of minus signs, possibly preceded by blanks, is
6230 permitted. This allows the construction of box comments where lines of minus
6231 signs are used to form the top and bottom of the box.
6234 A comment that starts and ends with ``@code{--}'' is permitted as long as at
6235 least one blank follows the initial ``@code{--}''. Together with the preceding
6236 rule, this allows the construction of box comments, as shown in the following
6239 ---------------------------
6240 -- This is a box comment --
6241 -- with two text lines. --
6242 ---------------------------
6246 @item ^d^DOS_LINE_ENDINGS^
6247 @emph{Check no DOS line terminators present.}
6248 All lines must be terminated by a single ASCII.LF
6249 character (in particular the DOS line terminator sequence CR/LF is not
6253 @emph{Check end/exit labels.}
6254 Optional labels on @code{end} statements ending subprograms and on
6255 @code{exit} statements exiting named loops, are required to be present.
6258 @emph{No form feeds or vertical tabs.}
6259 Neither form feeds nor vertical tab characters are permitted
6263 @emph{GNAT style mode}
6264 The set of style check switches is set to match that used by the GNAT sources.
6265 This may be useful when developing code that is eventually intended to be
6266 incorporated into GNAT. For further details, see GNAT sources.
6269 @emph{No horizontal tabs.}
6270 Horizontal tab characters are not permitted in the source text.
6271 Together with the b (no blanks at end of line) check, this
6272 enforces a canonical form for the use of blanks to separate
6276 @emph{Check if-then layout.}
6277 The keyword @code{then} must appear either on the same
6278 line as corresponding @code{if}, or on a line on its own, lined
6279 up under the @code{if} with at least one non-blank line in between
6280 containing all or part of the condition to be tested.
6283 @emph{check mode IN keywords}
6284 Mode @code{in} (the default mode) is not
6285 allowed to be given explicitly. @code{in out} is fine,
6286 but not @code{in} on its own.
6289 @emph{Check keyword casing.}
6290 All keywords must be in lower case (with the exception of keywords
6291 such as @code{digits} used as attribute names to which this check
6295 @emph{Check layout.}
6296 Layout of statement and declaration constructs must follow the
6297 recommendations in the Ada Reference Manual, as indicated by the
6298 form of the syntax rules. For example an @code{else} keyword must
6299 be lined up with the corresponding @code{if} keyword.
6301 There are two respects in which the style rule enforced by this check
6302 option are more liberal than those in the Ada Reference Manual. First
6303 in the case of record declarations, it is permissible to put the
6304 @code{record} keyword on the same line as the @code{type} keyword, and
6305 then the @code{end} in @code{end record} must line up under @code{type}.
6306 This is also permitted when the type declaration is split on two lines.
6307 For example, any of the following three layouts is acceptable:
6309 @smallexample @c ada
6332 Second, in the case of a block statement, a permitted alternative
6333 is to put the block label on the same line as the @code{declare} or
6334 @code{begin} keyword, and then line the @code{end} keyword up under
6335 the block label. For example both the following are permitted:
6337 @smallexample @c ada
6355 The same alternative format is allowed for loops. For example, both of
6356 the following are permitted:
6358 @smallexample @c ada
6360 Clear : while J < 10 loop
6371 @item ^Lnnn^MAX_NESTING=nnn^
6372 @emph{Set maximum nesting level}
6373 The maximum level of nesting of constructs (including subprograms, loops,
6374 blocks, packages, and conditionals) may not exceed the given value
6375 @option{nnn}. A value of zero disconnects this style check.
6377 @item ^m^LINE_LENGTH^
6378 @emph{Check maximum line length.}
6379 The length of source lines must not exceed 79 characters, including
6380 any trailing blanks. The value of 79 allows convenient display on an
6381 80 character wide device or window, allowing for possible special
6382 treatment of 80 character lines. Note that this count is of
6383 characters in the source text. This means that a tab character counts
6384 as one character in this count but a wide character sequence counts as
6385 a single character (however many bytes are needed in the encoding).
6387 @item ^Mnnn^MAX_LENGTH=nnn^
6388 @emph{Set maximum line length.}
6389 The length of lines must not exceed the
6390 given value @option{nnn}. The maximum value that can be specified is 32767.
6392 @item ^n^STANDARD_CASING^
6393 @emph{Check casing of entities in Standard.}
6394 Any identifier from Standard must be cased
6395 to match the presentation in the Ada Reference Manual (for example,
6396 @code{Integer} and @code{ASCII.NUL}).
6399 @emph{Turn off all style checks}
6400 All style check options are turned off.
6402 @item ^o^ORDERED_SUBPROGRAMS^
6403 @emph{Check order of subprogram bodies.}
6404 All subprogram bodies in a given scope
6405 (e.g.@: a package body) must be in alphabetical order. The ordering
6406 rule uses normal Ada rules for comparing strings, ignoring casing
6407 of letters, except that if there is a trailing numeric suffix, then
6408 the value of this suffix is used in the ordering (e.g.@: Junk2 comes
6411 @item ^O^OVERRIDING_INDICATORS^
6412 @emph{Check that overriding subprograms are explicitly marked as such.}
6413 The declaration of a primitive operation of a type extension that overrides
6414 an inherited operation must carry an overriding indicator.
6417 @emph{Check pragma casing.}
6418 Pragma names must be written in mixed case, that is, the
6419 initial letter and any letter following an underscore must be uppercase.
6420 All other letters must be lowercase.
6422 @item ^r^REFERENCES^
6423 @emph{Check references.}
6424 All identifier references must be cased in the same way as the
6425 corresponding declaration. No specific casing style is imposed on
6426 identifiers. The only requirement is for consistency of references
6429 @item ^S^STATEMENTS_AFTER_THEN_ELSE^
6430 @emph{Check no statements after THEN/ELSE.}
6431 No statements are allowed
6432 on the same line as a THEN or ELSE keyword following the
6433 keyword in an IF statement. OR ELSE and AND THEN are not affected,
6434 and a special exception allows a pragma to appear after ELSE.
6437 @emph{Check separate specs.}
6438 Separate declarations (``specs'') are required for subprograms (a
6439 body is not allowed to serve as its own declaration). The only
6440 exception is that parameterless library level procedures are
6441 not required to have a separate declaration. This exception covers
6442 the most frequent form of main program procedures.
6445 @emph{Check token spacing.}
6446 The following token spacing rules are enforced:
6451 The keywords @code{@b{abs}} and @code{@b{not}} must be followed by a space.
6454 The token @code{=>} must be surrounded by spaces.
6457 The token @code{<>} must be preceded by a space or a left parenthesis.
6460 Binary operators other than @code{**} must be surrounded by spaces.
6461 There is no restriction on the layout of the @code{**} binary operator.
6464 Colon must be surrounded by spaces.
6467 Colon-equal (assignment, initialization) must be surrounded by spaces.
6470 Comma must be the first non-blank character on the line, or be
6471 immediately preceded by a non-blank character, and must be followed
6475 If the token preceding a left parenthesis ends with a letter or digit, then
6476 a space must separate the two tokens.
6479 if the token following a right parenthesis starts with a letter or digit, then
6480 a space must separate the two tokens.
6483 A right parenthesis must either be the first non-blank character on
6484 a line, or it must be preceded by a non-blank character.
6487 A semicolon must not be preceded by a space, and must not be followed by
6488 a non-blank character.
6491 A unary plus or minus may not be followed by a space.
6494 A vertical bar must be surrounded by spaces.
6497 @item ^u^UNNECESSARY_BLANK_LINES^
6498 @emph{Check unnecessary blank lines.}
6499 Unnecessary blank lines are not allowed. A blank line is considered
6500 unnecessary if it appears at the end of the file, or if more than
6501 one blank line occurs in sequence.
6503 @item ^x^XTRA_PARENS^
6504 @emph{Check extra parentheses.}
6505 Unnecessary extra level of parentheses (C-style) are not allowed
6506 around conditions in @code{if} statements, @code{while} statements and
6507 @code{exit} statements.
6509 @item ^y^ALL_BUILTIN^
6510 @emph{Set all standard style check options}
6511 This is equivalent to @code{gnaty3aAbcefhiklmnprst}, that is all checking
6512 options enabled with the exception of @option{-gnatyo}, @option{-gnatyI},
6513 @option{-gnatyS}, @option{-gnatyLnnn},
6514 @option{-gnatyd}, @option{-gnatyu}, and @option{-gnatyx}.
6518 @emph{Remove style check options}
6519 This causes any subsequent options in the string to act as canceling the
6520 corresponding style check option. To cancel maximum nesting level control,
6521 use @option{L} parameter witout any integer value after that, because any
6522 digit following @option{-} in the parameter string of the @option{-gnaty}
6523 option will be threated as canceling indentation check. The same is true
6524 for @option{M} parameter. @option{y} and @option{N} parameters are not
6525 allowed after @option{-}.
6528 This causes any subsequent options in the string to enable the corresponding
6529 style check option. That is, it cancels the effect of a previous ^-^REMOVE^,
6535 @emph{Removing style check options}
6536 If the name of a style check is preceded by @option{NO} then the corresponding
6537 style check is turned off. For example @option{NOCOMMENTS} turns off style
6538 checking for comments.
6543 In the above rules, appearing in column one is always permitted, that is,
6544 counts as meeting either a requirement for a required preceding space,
6545 or as meeting a requirement for no preceding space.
6547 Appearing at the end of a line is also always permitted, that is, counts
6548 as meeting either a requirement for a following space, or as meeting
6549 a requirement for no following space.
6552 If any of these style rules is violated, a message is generated giving
6553 details on the violation. The initial characters of such messages are
6554 always ``@code{(style)}''. Note that these messages are treated as warning
6555 messages, so they normally do not prevent the generation of an object
6556 file. The @option{-gnatwe} switch can be used to treat warning messages,
6557 including style messages, as fatal errors.
6561 @option{-gnaty} on its own (that is not
6562 followed by any letters or digits), then the effect is equivalent
6563 to the use of @option{-gnatyy}, as described above, that is all
6564 built-in standard style check options are enabled.
6568 /STYLE_CHECKS=ALL_BUILTIN enables all checking options with
6569 the exception of ORDERED_SUBPROGRAMS, UNNECESSARY_BLANK_LINES,
6570 XTRA_PARENS, and DOS_LINE_ENDINGS. In addition
6580 clears any previously set style checks.
6582 @node Run-Time Checks
6583 @subsection Run-Time Checks
6584 @cindex Division by zero
6585 @cindex Access before elaboration
6586 @cindex Checks, division by zero
6587 @cindex Checks, access before elaboration
6588 @cindex Checks, stack overflow checking
6591 By default, the following checks are suppressed: integer overflow
6592 checks, stack overflow checks, and checks for access before
6593 elaboration on subprogram calls. All other checks, including range
6594 checks and array bounds checks, are turned on by default. The
6595 following @command{gcc} switches refine this default behavior.
6600 @cindex @option{-gnatp} (@command{gcc})
6601 @cindex Suppressing checks
6602 @cindex Checks, suppressing
6604 This switch causes the unit to be compiled
6605 as though @code{pragma Suppress (All_checks)}
6606 had been present in the source. Validity checks are also eliminated (in
6607 other words @option{-gnatp} also implies @option{-gnatVn}.
6608 Use this switch to improve the performance
6609 of the code at the expense of safety in the presence of invalid data or
6612 Note that when checks are suppressed, the compiler is allowed, but not
6613 required, to omit the checking code. If the run-time cost of the
6614 checking code is zero or near-zero, the compiler will generate it even
6615 if checks are suppressed. In particular, if the compiler can prove
6616 that a certain check will necessarily fail, it will generate code to
6617 do an unconditional ``raise'', even if checks are suppressed. The
6618 compiler warns in this case. Another case in which checks may not be
6619 eliminated is when they are embedded in certain run time routines such
6620 as math library routines.
6622 Of course, run-time checks are omitted whenever the compiler can prove
6623 that they will not fail, whether or not checks are suppressed.
6625 Note that if you suppress a check that would have failed, program
6626 execution is erroneous, which means the behavior is totally
6627 unpredictable. The program might crash, or print wrong answers, or
6628 do anything else. It might even do exactly what you wanted it to do
6629 (and then it might start failing mysteriously next week or next
6630 year). The compiler will generate code based on the assumption that
6631 the condition being checked is true, which can result in disaster if
6632 that assumption is wrong.
6634 The @option{-gnatp} switch has no effect if a subsequent
6635 @option{-gnat-p} switch appears.
6638 @cindex @option{-gnat-p} (@command{gcc})
6639 @cindex Suppressing checks
6640 @cindex Checks, suppressing
6642 This switch cancels the effect of a previous @option{gnatp} switch.
6645 @cindex @option{-gnato} (@command{gcc})
6646 @cindex Overflow checks
6647 @cindex Check, overflow
6648 Enables overflow checking for integer operations.
6649 This causes GNAT to generate slower and larger executable
6650 programs by adding code to check for overflow (resulting in raising
6651 @code{Constraint_Error} as required by standard Ada
6652 semantics). These overflow checks correspond to situations in which
6653 the true value of the result of an operation may be outside the base
6654 range of the result type. The following example shows the distinction:
6656 @smallexample @c ada
6657 X1 : Integer := "Integer'Last";
6658 X2 : Integer range 1 .. 5 := "5";
6659 X3 : Integer := "Integer'Last";
6660 X4 : Integer range 1 .. 5 := "5";
6661 F : Float := "2.0E+20";
6670 Note that if explicit values are assigned at compile time, the
6671 compiler may be able to detect overflow at compile time, in which case
6672 no actual run-time checking code is required, and Constraint_Error
6673 will be raised unconditionally, with or without
6674 @option{-gnato}. That's why the assigned values in the above fragment
6675 are in quotes, the meaning is "assign a value not known to the
6676 compiler that happens to be equal to ...". The remaining discussion
6677 assumes that the compiler cannot detect the values at compile time.
6679 Here the first addition results in a value that is outside the base range
6680 of Integer, and hence requires an overflow check for detection of the
6681 constraint error. Thus the first assignment to @code{X1} raises a
6682 @code{Constraint_Error} exception only if @option{-gnato} is set.
6684 The second increment operation results in a violation of the explicit
6685 range constraint; such range checks are performed by default, and are
6686 unaffected by @option{-gnato}.
6688 The two conversions of @code{F} both result in values that are outside
6689 the base range of type @code{Integer} and thus will raise
6690 @code{Constraint_Error} exceptions only if @option{-gnato} is used.
6691 The fact that the result of the second conversion is assigned to
6692 variable @code{X4} with a restricted range is irrelevant, since the problem
6693 is in the conversion, not the assignment.
6695 Basically the rule is that in the default mode (@option{-gnato} not
6696 used), the generated code assures that all integer variables stay
6697 within their declared ranges, or within the base range if there is
6698 no declared range. This prevents any serious problems like indexes
6699 out of range for array operations.
6701 What is not checked in default mode is an overflow that results in
6702 an in-range, but incorrect value. In the above example, the assignments
6703 to @code{X1}, @code{X2}, @code{X3} all give results that are within the
6704 range of the target variable, but the result is wrong in the sense that
6705 it is too large to be represented correctly. Typically the assignment
6706 to @code{X1} will result in wrap around to the largest negative number.
6707 The conversions of @code{F} will result in some @code{Integer} value
6708 and if that integer value is out of the @code{X4} range then the
6709 subsequent assignment would generate an exception.
6711 @findex Machine_Overflows
6712 Note that the @option{-gnato} switch does not affect the code generated
6713 for any floating-point operations; it applies only to integer
6715 For floating-point, GNAT has the @code{Machine_Overflows}
6716 attribute set to @code{False} and the normal mode of operation is to
6717 generate IEEE NaN and infinite values on overflow or invalid operations
6718 (such as dividing 0.0 by 0.0).
6720 The reason that we distinguish overflow checking from other kinds of
6721 range constraint checking is that a failure of an overflow check, unlike
6722 for example the failure of a range check, can result in an incorrect
6723 value, but cannot cause random memory destruction (like an out of range
6724 subscript), or a wild jump (from an out of range case value). Overflow
6725 checking is also quite expensive in time and space, since in general it
6726 requires the use of double length arithmetic.
6728 Note again that @option{-gnato} is off by default, so overflow checking is
6729 not performed in default mode. This means that out of the box, with the
6730 default settings, GNAT does not do all the checks expected from the
6731 language description in the Ada Reference Manual. If you want all constraint
6732 checks to be performed, as described in this Manual, then you must
6733 explicitly use the -gnato switch either on the @command{gnatmake} or
6734 @command{gcc} command.
6737 @cindex @option{-gnatE} (@command{gcc})
6738 @cindex Elaboration checks
6739 @cindex Check, elaboration
6740 Enables dynamic checks for access-before-elaboration
6741 on subprogram calls and generic instantiations.
6742 Note that @option{-gnatE} is not necessary for safety, because in the
6743 default mode, GNAT ensures statically that the checks would not fail.
6744 For full details of the effect and use of this switch,
6745 @xref{Compiling Using gcc}.
6748 @cindex @option{-fstack-check} (@command{gcc})
6749 @cindex Stack Overflow Checking
6750 @cindex Checks, stack overflow checking
6751 Activates stack overflow checking. For full details of the effect and use of
6752 this switch see @ref{Stack Overflow Checking}.
6757 The setting of these switches only controls the default setting of the
6758 checks. You may modify them using either @code{Suppress} (to remove
6759 checks) or @code{Unsuppress} (to add back suppressed checks) pragmas in
6762 @node Using gcc for Syntax Checking
6763 @subsection Using @command{gcc} for Syntax Checking
6766 @cindex @option{-gnats} (@command{gcc})
6770 The @code{s} stands for ``syntax''.
6773 Run GNAT in syntax checking only mode. For
6774 example, the command
6777 $ gcc -c -gnats x.adb
6781 compiles file @file{x.adb} in syntax-check-only mode. You can check a
6782 series of files in a single command
6784 , and can use wild cards to specify such a group of files.
6785 Note that you must specify the @option{-c} (compile
6786 only) flag in addition to the @option{-gnats} flag.
6789 You may use other switches in conjunction with @option{-gnats}. In
6790 particular, @option{-gnatl} and @option{-gnatv} are useful to control the
6791 format of any generated error messages.
6793 When the source file is empty or contains only empty lines and/or comments,
6794 the output is a warning:
6797 $ gcc -c -gnats -x ada toto.txt
6798 toto.txt:1:01: warning: empty file, contains no compilation units
6802 Otherwise, the output is simply the error messages, if any. No object file or
6803 ALI file is generated by a syntax-only compilation. Also, no units other
6804 than the one specified are accessed. For example, if a unit @code{X}
6805 @code{with}'s a unit @code{Y}, compiling unit @code{X} in syntax
6806 check only mode does not access the source file containing unit
6809 @cindex Multiple units, syntax checking
6810 Normally, GNAT allows only a single unit in a source file. However, this
6811 restriction does not apply in syntax-check-only mode, and it is possible
6812 to check a file containing multiple compilation units concatenated
6813 together. This is primarily used by the @code{gnatchop} utility
6814 (@pxref{Renaming Files Using gnatchop}).
6817 @node Using gcc for Semantic Checking
6818 @subsection Using @command{gcc} for Semantic Checking
6821 @cindex @option{-gnatc} (@command{gcc})
6825 The @code{c} stands for ``check''.
6827 Causes the compiler to operate in semantic check mode,
6828 with full checking for all illegalities specified in the
6829 Ada Reference Manual, but without generation of any object code
6830 (no object file is generated).
6832 Because dependent files must be accessed, you must follow the GNAT
6833 semantic restrictions on file structuring to operate in this mode:
6837 The needed source files must be accessible
6838 (@pxref{Search Paths and the Run-Time Library (RTL)}).
6841 Each file must contain only one compilation unit.
6844 The file name and unit name must match (@pxref{File Naming Rules}).
6847 The output consists of error messages as appropriate. No object file is
6848 generated. An @file{ALI} file is generated for use in the context of
6849 cross-reference tools, but this file is marked as not being suitable
6850 for binding (since no object file is generated).
6851 The checking corresponds exactly to the notion of
6852 legality in the Ada Reference Manual.
6854 Any unit can be compiled in semantics-checking-only mode, including
6855 units that would not normally be compiled (subunits,
6856 and specifications where a separate body is present).
6859 @node Compiling Different Versions of Ada
6860 @subsection Compiling Different Versions of Ada
6863 The switches described in this section allow you to explicitly specify
6864 the version of the Ada language that your programs are written in.
6865 By default @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
6866 but you can also specify @value{NONDEFAULTLANGUAGEVERSION} or
6867 indicate Ada 83 compatibility mode.
6870 @cindex Compatibility with Ada 83
6872 @item -gnat83 (Ada 83 Compatibility Mode)
6873 @cindex @option{-gnat83} (@command{gcc})
6874 @cindex ACVC, Ada 83 tests
6878 Although GNAT is primarily an Ada 95 / Ada 2005 compiler, this switch
6879 specifies that the program is to be compiled in Ada 83 mode. With
6880 @option{-gnat83}, GNAT rejects most post-Ada 83 extensions and applies Ada 83
6881 semantics where this can be done easily.
6882 It is not possible to guarantee this switch does a perfect
6883 job; some subtle tests, such as are
6884 found in earlier ACVC tests (and that have been removed from the ACATS suite
6885 for Ada 95), might not compile correctly.
6886 Nevertheless, this switch may be useful in some circumstances, for example
6887 where, due to contractual reasons, existing code needs to be maintained
6888 using only Ada 83 features.
6890 With few exceptions (most notably the need to use @code{<>} on
6891 @cindex Generic formal parameters
6892 unconstrained generic formal parameters, the use of the new Ada 95 / Ada 2005
6893 reserved words, and the use of packages
6894 with optional bodies), it is not necessary to specify the
6895 @option{-gnat83} switch when compiling Ada 83 programs, because, with rare
6896 exceptions, Ada 95 and Ada 2005 are upwardly compatible with Ada 83. Thus
6897 a correct Ada 83 program is usually also a correct program
6898 in these later versions of the language standard.
6899 For further information, please refer to @ref{Compatibility and Porting Guide}.
6901 @item -gnat95 (Ada 95 mode)
6902 @cindex @option{-gnat95} (@command{gcc})
6906 This switch directs the compiler to implement the Ada 95 version of the
6908 Since Ada 95 is almost completely upwards
6909 compatible with Ada 83, Ada 83 programs may generally be compiled using
6910 this switch (see the description of the @option{-gnat83} switch for further
6911 information about Ada 83 mode).
6912 If an Ada 2005 program is compiled in Ada 95 mode,
6913 uses of the new Ada 2005 features will cause error
6914 messages or warnings.
6916 This switch also can be used to cancel the effect of a previous
6917 @option{-gnat83} or @option{-gnat05} switch earlier in the command line.
6919 @item -gnat05 (Ada 2005 mode)
6920 @cindex @option{-gnat05} (@command{gcc})
6921 @cindex Ada 2005 mode
6924 This switch directs the compiler to implement the Ada 2005 version of the
6926 Since Ada 2005 is almost completely upwards
6927 compatible with Ada 95 (and thus also with Ada 83), Ada 83 and Ada 95 programs
6928 may generally be compiled using this switch (see the description of the
6929 @option{-gnat83} and @option{-gnat95} switches for further
6932 For information about the approved ``Ada Issues'' that have been incorporated
6933 into Ada 2005, see @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs}.
6934 Included with GNAT releases is a file @file{features-ada0y} that describes
6935 the set of implemented Ada 2005 features.
6939 @node Character Set Control
6940 @subsection Character Set Control
6942 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
6943 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
6946 Normally GNAT recognizes the Latin-1 character set in source program
6947 identifiers, as described in the Ada Reference Manual.
6949 GNAT to recognize alternate character sets in identifiers. @var{c} is a
6950 single character ^^or word^ indicating the character set, as follows:
6954 ISO 8859-1 (Latin-1) identifiers
6957 ISO 8859-2 (Latin-2) letters allowed in identifiers
6960 ISO 8859-3 (Latin-3) letters allowed in identifiers
6963 ISO 8859-4 (Latin-4) letters allowed in identifiers
6966 ISO 8859-5 (Cyrillic) letters allowed in identifiers
6969 ISO 8859-15 (Latin-9) letters allowed in identifiers
6972 IBM PC letters (code page 437) allowed in identifiers
6975 IBM PC letters (code page 850) allowed in identifiers
6977 @item ^f^FULL_UPPER^
6978 Full upper-half codes allowed in identifiers
6981 No upper-half codes allowed in identifiers
6984 Wide-character codes (that is, codes greater than 255)
6985 allowed in identifiers
6988 @xref{Foreign Language Representation}, for full details on the
6989 implementation of these character sets.
6991 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
6992 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
6993 Specify the method of encoding for wide characters.
6994 @var{e} is one of the following:
6999 Hex encoding (brackets coding also recognized)
7002 Upper half encoding (brackets encoding also recognized)
7005 Shift/JIS encoding (brackets encoding also recognized)
7008 EUC encoding (brackets encoding also recognized)
7011 UTF-8 encoding (brackets encoding also recognized)
7014 Brackets encoding only (default value)
7016 For full details on these encoding
7017 methods see @ref{Wide Character Encodings}.
7018 Note that brackets coding is always accepted, even if one of the other
7019 options is specified, so for example @option{-gnatW8} specifies that both
7020 brackets and UTF-8 encodings will be recognized. The units that are
7021 with'ed directly or indirectly will be scanned using the specified
7022 representation scheme, and so if one of the non-brackets scheme is
7023 used, it must be used consistently throughout the program. However,
7024 since brackets encoding is always recognized, it may be conveniently
7025 used in standard libraries, allowing these libraries to be used with
7026 any of the available coding schemes.
7029 If no @option{-gnatW?} parameter is present, then the default
7030 representation is normally Brackets encoding only. However, if the
7031 first three characters of the file are 16#EF# 16#BB# 16#BF# (the standard
7032 byte order mark or BOM for UTF-8), then these three characters are
7033 skipped and the default representation for the file is set to UTF-8.
7035 Note that the wide character representation that is specified (explicitly
7036 or by default) for the main program also acts as the default encoding used
7037 for Wide_Text_IO files if not specifically overridden by a WCEM form
7041 @node File Naming Control
7042 @subsection File Naming Control
7045 @item ^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{n}
7046 @cindex @option{-gnatk} (@command{gcc})
7047 Activates file name ``krunching''. @var{n}, a decimal integer in the range
7048 1-999, indicates the maximum allowable length of a file name (not
7049 including the @file{.ads} or @file{.adb} extension). The default is not
7050 to enable file name krunching.
7052 For the source file naming rules, @xref{File Naming Rules}.
7055 @node Subprogram Inlining Control
7056 @subsection Subprogram Inlining Control
7061 @cindex @option{-gnatn} (@command{gcc})
7063 The @code{n} here is intended to suggest the first syllable of the
7066 GNAT recognizes and processes @code{Inline} pragmas. However, for the
7067 inlining to actually occur, optimization must be enabled. To enable
7068 inlining of subprograms specified by pragma @code{Inline},
7069 you must also specify this switch.
7070 In the absence of this switch, GNAT does not attempt
7071 inlining and does not need to access the bodies of
7072 subprograms for which @code{pragma Inline} is specified if they are not
7073 in the current unit.
7075 If you specify this switch the compiler will access these bodies,
7076 creating an extra source dependency for the resulting object file, and
7077 where possible, the call will be inlined.
7078 For further details on when inlining is possible
7079 see @ref{Inlining of Subprograms}.
7082 @cindex @option{-gnatN} (@command{gcc})
7083 This switch activates front-end inlining which also
7084 generates additional dependencies.
7086 When using a gcc-based back end (in practice this means using any version
7087 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
7088 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
7089 Historically front end inlining was more extensive than the gcc back end
7090 inlining, but that is no longer the case.
7093 @node Auxiliary Output Control
7094 @subsection Auxiliary Output Control
7098 @cindex @option{-gnatt} (@command{gcc})
7099 @cindex Writing internal trees
7100 @cindex Internal trees, writing to file
7101 Causes GNAT to write the internal tree for a unit to a file (with the
7102 extension @file{.adt}.
7103 This not normally required, but is used by separate analysis tools.
7105 these tools do the necessary compilations automatically, so you should
7106 not have to specify this switch in normal operation.
7107 Note that the combination of switches @option{-gnatct}
7108 generates a tree in the form required by ASIS applications.
7111 @cindex @option{-gnatu} (@command{gcc})
7112 Print a list of units required by this compilation on @file{stdout}.
7113 The listing includes all units on which the unit being compiled depends
7114 either directly or indirectly.
7117 @item -pass-exit-codes
7118 @cindex @option{-pass-exit-codes} (@command{gcc})
7119 If this switch is not used, the exit code returned by @command{gcc} when
7120 compiling multiple files indicates whether all source files have
7121 been successfully used to generate object files or not.
7123 When @option{-pass-exit-codes} is used, @command{gcc} exits with an extended
7124 exit status and allows an integrated development environment to better
7125 react to a compilation failure. Those exit status are:
7129 There was an error in at least one source file.
7131 At least one source file did not generate an object file.
7133 The compiler died unexpectedly (internal error for example).
7135 An object file has been generated for every source file.
7140 @node Debugging Control
7141 @subsection Debugging Control
7145 @cindex Debugging options
7148 @cindex @option{-gnatd} (@command{gcc})
7149 Activate internal debugging switches. @var{x} is a letter or digit, or
7150 string of letters or digits, which specifies the type of debugging
7151 outputs desired. Normally these are used only for internal development
7152 or system debugging purposes. You can find full documentation for these
7153 switches in the body of the @code{Debug} unit in the compiler source
7154 file @file{debug.adb}.
7158 @cindex @option{-gnatG} (@command{gcc})
7159 This switch causes the compiler to generate auxiliary output containing
7160 a pseudo-source listing of the generated expanded code. Like most Ada
7161 compilers, GNAT works by first transforming the high level Ada code into
7162 lower level constructs. For example, tasking operations are transformed
7163 into calls to the tasking run-time routines. A unique capability of GNAT
7164 is to list this expanded code in a form very close to normal Ada source.
7165 This is very useful in understanding the implications of various Ada
7166 usage on the efficiency of the generated code. There are many cases in
7167 Ada (e.g.@: the use of controlled types), where simple Ada statements can
7168 generate a lot of run-time code. By using @option{-gnatG} you can identify
7169 these cases, and consider whether it may be desirable to modify the coding
7170 approach to improve efficiency.
7172 The optional parameter @code{nn} if present after -gnatG specifies an
7173 alternative maximum line length that overrides the normal default of 72.
7174 This value is in the range 40-999999, values less than 40 being silently
7175 reset to 40. The equal sign is optional.
7177 The format of the output is very similar to standard Ada source, and is
7178 easily understood by an Ada programmer. The following special syntactic
7179 additions correspond to low level features used in the generated code that
7180 do not have any exact analogies in pure Ada source form. The following
7181 is a partial list of these special constructions. See the spec
7182 of package @code{Sprint} in file @file{sprint.ads} for a full list.
7184 If the switch @option{-gnatL} is used in conjunction with
7185 @cindex @option{-gnatL} (@command{gcc})
7186 @option{-gnatG}, then the original source lines are interspersed
7187 in the expanded source (as comment lines with the original line number).
7190 @item new @var{xxx} @r{[}storage_pool = @var{yyy}@r{]}
7191 Shows the storage pool being used for an allocator.
7193 @item at end @var{procedure-name};
7194 Shows the finalization (cleanup) procedure for a scope.
7196 @item (if @var{expr} then @var{expr} else @var{expr})
7197 Conditional expression equivalent to the @code{x?y:z} construction in C.
7199 @item @var{target}^^^(@var{source})
7200 A conversion with floating-point truncation instead of rounding.
7202 @item @var{target}?(@var{source})
7203 A conversion that bypasses normal Ada semantic checking. In particular
7204 enumeration types and fixed-point types are treated simply as integers.
7206 @item @var{target}?^^^(@var{source})
7207 Combines the above two cases.
7209 @item @var{x} #/ @var{y}
7210 @itemx @var{x} #mod @var{y}
7211 @itemx @var{x} #* @var{y}
7212 @itemx @var{x} #rem @var{y}
7213 A division or multiplication of fixed-point values which are treated as
7214 integers without any kind of scaling.
7216 @item free @var{expr} @r{[}storage_pool = @var{xxx}@r{]}
7217 Shows the storage pool associated with a @code{free} statement.
7219 @item [subtype or type declaration]
7220 Used to list an equivalent declaration for an internally generated
7221 type that is referenced elsewhere in the listing.
7223 @c @item freeze @var{type-name} @ovar{actions}
7224 @c Expanding @ovar macro inline (explanation in macro def comments)
7225 @item freeze @var{type-name} @r{[}@var{actions}@r{]}
7226 Shows the point at which @var{type-name} is frozen, with possible
7227 associated actions to be performed at the freeze point.
7229 @item reference @var{itype}
7230 Reference (and hence definition) to internal type @var{itype}.
7232 @item @var{function-name}! (@var{arg}, @var{arg}, @var{arg})
7233 Intrinsic function call.
7235 @item @var{label-name} : label
7236 Declaration of label @var{labelname}.
7238 @item #$ @var{subprogram-name}
7239 An implicit call to a run-time support routine
7240 (to meet the requirement of H.3.1(9) in a
7243 @item @var{expr} && @var{expr} && @var{expr} @dots{} && @var{expr}
7244 A multiple concatenation (same effect as @var{expr} & @var{expr} &
7245 @var{expr}, but handled more efficiently).
7247 @item [constraint_error]
7248 Raise the @code{Constraint_Error} exception.
7250 @item @var{expression}'reference
7251 A pointer to the result of evaluating @var{expression}.
7253 @item @var{target-type}!(@var{source-expression})
7254 An unchecked conversion of @var{source-expression} to @var{target-type}.
7256 @item [@var{numerator}/@var{denominator}]
7257 Used to represent internal real literals (that) have no exact
7258 representation in base 2-16 (for example, the result of compile time
7259 evaluation of the expression 1.0/27.0).
7263 @cindex @option{-gnatD} (@command{gcc})
7264 When used in conjunction with @option{-gnatG}, this switch causes
7265 the expanded source, as described above for
7266 @option{-gnatG} to be written to files with names
7267 @file{^xxx.dg^XXX_DG^}, where @file{xxx} is the normal file name,
7268 instead of to the standard output file. For
7269 example, if the source file name is @file{hello.adb}, then a file
7270 @file{^hello.adb.dg^HELLO.ADB_DG^} will be written. The debugging
7271 information generated by the @command{gcc} @option{^-g^/DEBUG^} switch
7272 will refer to the generated @file{^xxx.dg^XXX_DG^} file. This allows
7273 you to do source level debugging using the generated code which is
7274 sometimes useful for complex code, for example to find out exactly
7275 which part of a complex construction raised an exception. This switch
7276 also suppress generation of cross-reference information (see
7277 @option{-gnatx}) since otherwise the cross-reference information
7278 would refer to the @file{^.dg^.DG^} file, which would cause
7279 confusion since this is not the original source file.
7281 Note that @option{-gnatD} actually implies @option{-gnatG}
7282 automatically, so it is not necessary to give both options.
7283 In other words @option{-gnatD} is equivalent to @option{-gnatDG}).
7285 If the switch @option{-gnatL} is used in conjunction with
7286 @cindex @option{-gnatL} (@command{gcc})
7287 @option{-gnatDG}, then the original source lines are interspersed
7288 in the expanded source (as comment lines with the original line number).
7290 The optional parameter @code{nn} if present after -gnatD specifies an
7291 alternative maximum line length that overrides the normal default of 72.
7292 This value is in the range 40-999999, values less than 40 being silently
7293 reset to 40. The equal sign is optional.
7296 @cindex @option{-gnatr} (@command{gcc})
7297 @cindex pragma Restrictions
7298 This switch causes pragma Restrictions to be treated as Restriction_Warnings
7299 so that violation of restrictions causes warnings rather than illegalities.
7300 This is useful during the development process when new restrictions are added
7301 or investigated. The switch also causes pragma Profile to be treated as
7302 Profile_Warnings, and pragma Restricted_Run_Time and pragma Ravenscar set
7303 restriction warnings rather than restrictions.
7306 @item -gnatR@r{[}0@r{|}1@r{|}2@r{|}3@r{[}s@r{]]}
7307 @cindex @option{-gnatR} (@command{gcc})
7308 This switch controls output from the compiler of a listing showing
7309 representation information for declared types and objects. For
7310 @option{-gnatR0}, no information is output (equivalent to omitting
7311 the @option{-gnatR} switch). For @option{-gnatR1} (which is the default,
7312 so @option{-gnatR} with no parameter has the same effect), size and alignment
7313 information is listed for declared array and record types. For
7314 @option{-gnatR2}, size and alignment information is listed for all
7315 declared types and objects. Finally @option{-gnatR3} includes symbolic
7316 expressions for values that are computed at run time for
7317 variant records. These symbolic expressions have a mostly obvious
7318 format with #n being used to represent the value of the n'th
7319 discriminant. See source files @file{repinfo.ads/adb} in the
7320 @code{GNAT} sources for full details on the format of @option{-gnatR3}
7321 output. If the switch is followed by an s (e.g.@: @option{-gnatR2s}), then
7322 the output is to a file with the name @file{^file.rep^file_REP^} where
7323 file is the name of the corresponding source file.
7326 @item /REPRESENTATION_INFO
7327 @cindex @option{/REPRESENTATION_INFO} (@command{gcc})
7328 This qualifier controls output from the compiler of a listing showing
7329 representation information for declared types and objects. For
7330 @option{/REPRESENTATION_INFO=NONE}, no information is output
7331 (equivalent to omitting the @option{/REPRESENTATION_INFO} qualifier).
7332 @option{/REPRESENTATION_INFO} without option is equivalent to
7333 @option{/REPRESENTATION_INFO=ARRAYS}.
7334 For @option{/REPRESENTATION_INFO=ARRAYS}, size and alignment
7335 information is listed for declared array and record types. For
7336 @option{/REPRESENTATION_INFO=OBJECTS}, size and alignment information
7337 is listed for all expression information for values that are computed
7338 at run time for variant records. These symbolic expressions have a mostly
7339 obvious format with #n being used to represent the value of the n'th
7340 discriminant. See source files @file{REPINFO.ADS/ADB} in the
7341 @code{GNAT} sources for full details on the format of
7342 @option{/REPRESENTATION_INFO=SYMBOLIC} output.
7343 If _FILE is added at the end of an option
7344 (e.g.@: @option{/REPRESENTATION_INFO=ARRAYS_FILE}),
7345 then the output is to a file with the name @file{file_REP} where
7346 file is the name of the corresponding source file.
7348 Note that it is possible for record components to have zero size. In
7349 this case, the component clause uses an obvious extension of permitted
7350 Ada syntax, for example @code{at 0 range 0 .. -1}.
7352 Representation information requires that code be generated (since it is the
7353 code generator that lays out complex data structures). If an attempt is made
7354 to output representation information when no code is generated, for example
7355 when a subunit is compiled on its own, then no information can be generated
7356 and the compiler outputs a message to this effect.
7359 @cindex @option{-gnatS} (@command{gcc})
7360 The use of the switch @option{-gnatS} for an
7361 Ada compilation will cause the compiler to output a
7362 representation of package Standard in a form very
7363 close to standard Ada. It is not quite possible to
7364 do this entirely in standard Ada (since new
7365 numeric base types cannot be created in standard
7366 Ada), but the output is easily
7367 readable to any Ada programmer, and is useful to
7368 determine the characteristics of target dependent
7369 types in package Standard.
7372 @cindex @option{-gnatx} (@command{gcc})
7373 Normally the compiler generates full cross-referencing information in
7374 the @file{ALI} file. This information is used by a number of tools,
7375 including @code{gnatfind} and @code{gnatxref}. The @option{-gnatx} switch
7376 suppresses this information. This saves some space and may slightly
7377 speed up compilation, but means that these tools cannot be used.
7380 @node Exception Handling Control
7381 @subsection Exception Handling Control
7384 GNAT uses two methods for handling exceptions at run-time. The
7385 @code{setjmp/longjmp} method saves the context when entering
7386 a frame with an exception handler. Then when an exception is
7387 raised, the context can be restored immediately, without the
7388 need for tracing stack frames. This method provides very fast
7389 exception propagation, but introduces significant overhead for
7390 the use of exception handlers, even if no exception is raised.
7392 The other approach is called ``zero cost'' exception handling.
7393 With this method, the compiler builds static tables to describe
7394 the exception ranges. No dynamic code is required when entering
7395 a frame containing an exception handler. When an exception is
7396 raised, the tables are used to control a back trace of the
7397 subprogram invocation stack to locate the required exception
7398 handler. This method has considerably poorer performance for
7399 the propagation of exceptions, but there is no overhead for
7400 exception handlers if no exception is raised. Note that in this
7401 mode and in the context of mixed Ada and C/C++ programming,
7402 to propagate an exception through a C/C++ code, the C/C++ code
7403 must be compiled with the @option{-funwind-tables} GCC's
7406 The following switches may be used to control which of the
7407 two exception handling methods is used.
7413 @cindex @option{--RTS=sjlj} (@command{gnatmake})
7414 This switch causes the setjmp/longjmp run-time (when available) to be used
7415 for exception handling. If the default
7416 mechanism for the target is zero cost exceptions, then
7417 this switch can be used to modify this default, and must be
7418 used for all units in the partition.
7419 This option is rarely used. One case in which it may be
7420 advantageous is if you have an application where exception
7421 raising is common and the overall performance of the
7422 application is improved by favoring exception propagation.
7425 @cindex @option{--RTS=zcx} (@command{gnatmake})
7426 @cindex Zero Cost Exceptions
7427 This switch causes the zero cost approach to be used
7428 for exception handling. If this is the default mechanism for the
7429 target (see below), then this switch is unneeded. If the default
7430 mechanism for the target is setjmp/longjmp exceptions, then
7431 this switch can be used to modify this default, and must be
7432 used for all units in the partition.
7433 This option can only be used if the zero cost approach
7434 is available for the target in use, otherwise it will generate an error.
7438 The same option @option{--RTS} must be used both for @command{gcc}
7439 and @command{gnatbind}. Passing this option to @command{gnatmake}
7440 (@pxref{Switches for gnatmake}) will ensure the required consistency
7441 through the compilation and binding steps.
7443 @node Units to Sources Mapping Files
7444 @subsection Units to Sources Mapping Files
7448 @item -gnatem=@var{path}
7449 @cindex @option{-gnatem} (@command{gcc})
7450 A mapping file is a way to communicate to the compiler two mappings:
7451 from unit names to file names (without any directory information) and from
7452 file names to path names (with full directory information). These mappings
7453 are used by the compiler to short-circuit the path search.
7455 The use of mapping files is not required for correct operation of the
7456 compiler, but mapping files can improve efficiency, particularly when
7457 sources are read over a slow network connection. In normal operation,
7458 you need not be concerned with the format or use of mapping files,
7459 and the @option{-gnatem} switch is not a switch that you would use
7460 explicitly. It is intended primarily for use by automatic tools such as
7461 @command{gnatmake} running under the project file facility. The
7462 description here of the format of mapping files is provided
7463 for completeness and for possible use by other tools.
7465 A mapping file is a sequence of sets of three lines. In each set, the
7466 first line is the unit name, in lower case, with @code{%s} appended
7467 for specs and @code{%b} appended for bodies; the second line is the
7468 file name; and the third line is the path name.
7474 /gnat/project1/sources/main.2.ada
7477 When the switch @option{-gnatem} is specified, the compiler will
7478 create in memory the two mappings from the specified file. If there is
7479 any problem (nonexistent file, truncated file or duplicate entries),
7480 no mapping will be created.
7482 Several @option{-gnatem} switches may be specified; however, only the
7483 last one on the command line will be taken into account.
7485 When using a project file, @command{gnatmake} creates a temporary
7486 mapping file and communicates it to the compiler using this switch.
7490 @node Integrated Preprocessing
7491 @subsection Integrated Preprocessing
7494 GNAT sources may be preprocessed immediately before compilation.
7495 In this case, the actual
7496 text of the source is not the text of the source file, but is derived from it
7497 through a process called preprocessing. Integrated preprocessing is specified
7498 through switches @option{-gnatep} and/or @option{-gnateD}. @option{-gnatep}
7499 indicates, through a text file, the preprocessing data to be used.
7500 @option{-gnateD} specifies or modifies the values of preprocessing symbol.
7503 Note that when integrated preprocessing is used, the output from the
7504 preprocessor is not written to any external file. Instead it is passed
7505 internally to the compiler. If you need to preserve the result of
7506 preprocessing in a file, then you should use @command{gnatprep}
7507 to perform the desired preprocessing in stand-alone mode.
7510 It is recommended that @command{gnatmake} switch ^-s^/SWITCH_CHECK^ should be
7511 used when Integrated Preprocessing is used. The reason is that preprocessing
7512 with another Preprocessing Data file without changing the sources will
7513 not trigger recompilation without this switch.
7516 Note that @command{gnatmake} switch ^-m^/MINIMAL_RECOMPILATION^ will almost
7517 always trigger recompilation for sources that are preprocessed,
7518 because @command{gnatmake} cannot compute the checksum of the source after
7522 The actual preprocessing function is described in details in section
7523 @ref{Preprocessing Using gnatprep}. This section only describes how integrated
7524 preprocessing is triggered and parameterized.
7528 @item -gnatep=@var{file}
7529 @cindex @option{-gnatep} (@command{gcc})
7530 This switch indicates to the compiler the file name (without directory
7531 information) of the preprocessor data file to use. The preprocessor data file
7532 should be found in the source directories.
7535 A preprocessing data file is a text file with significant lines indicating
7536 how should be preprocessed either a specific source or all sources not
7537 mentioned in other lines. A significant line is a nonempty, non-comment line.
7538 Comments are similar to Ada comments.
7541 Each significant line starts with either a literal string or the character '*'.
7542 A literal string is the file name (without directory information) of the source
7543 to preprocess. A character '*' indicates the preprocessing for all the sources
7544 that are not specified explicitly on other lines (order of the lines is not
7545 significant). It is an error to have two lines with the same file name or two
7546 lines starting with the character '*'.
7549 After the file name or the character '*', another optional literal string
7550 indicating the file name of the definition file to be used for preprocessing
7551 (@pxref{Form of Definitions File}). The definition files are found by the
7552 compiler in one of the source directories. In some cases, when compiling
7553 a source in a directory other than the current directory, if the definition
7554 file is in the current directory, it may be necessary to add the current
7555 directory as a source directory through switch ^-I.^/SEARCH=[]^, otherwise
7556 the compiler would not find the definition file.
7559 Then, optionally, ^switches^switches^ similar to those of @code{gnatprep} may
7560 be found. Those ^switches^switches^ are:
7565 Causes both preprocessor lines and the lines deleted by
7566 preprocessing to be replaced by blank lines, preserving the line number.
7567 This ^switch^switch^ is always implied; however, if specified after @option{-c}
7568 it cancels the effect of @option{-c}.
7571 Causes both preprocessor lines and the lines deleted
7572 by preprocessing to be retained as comments marked
7573 with the special string ``@code{--! }''.
7575 @item -Dsymbol=value
7576 Define or redefine a symbol, associated with value. A symbol is an Ada
7577 identifier, or an Ada reserved word, with the exception of @code{if},
7578 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7579 @code{value} is either a literal string, an Ada identifier or any Ada reserved
7580 word. A symbol declared with this ^switch^switch^ replaces a symbol with the
7581 same name defined in a definition file.
7584 Causes a sorted list of symbol names and values to be
7585 listed on the standard output file.
7588 Causes undefined symbols to be treated as having the value @code{FALSE}
7590 of a preprocessor test. In the absence of this option, an undefined symbol in
7591 a @code{#if} or @code{#elsif} test will be treated as an error.
7596 Examples of valid lines in a preprocessor data file:
7599 "toto.adb" "prep.def" -u
7600 -- preprocess "toto.adb", using definition file "prep.def",
7601 -- undefined symbol are False.
7604 -- preprocess all other sources without a definition file;
7605 -- suppressed lined are commented; symbol VERSION has the value V101.
7607 "titi.adb" "prep2.def" -s
7608 -- preprocess "titi.adb", using definition file "prep2.def";
7609 -- list all symbols with their values.
7612 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=value@r{]}
7613 @cindex @option{-gnateD} (@command{gcc})
7614 Define or redefine a preprocessing symbol, associated with value. If no value
7615 is given on the command line, then the value of the symbol is @code{True}.
7616 A symbol is an identifier, following normal Ada (case-insensitive)
7617 rules for its syntax, and value is any sequence (including an empty sequence)
7618 of characters from the set (letters, digits, period, underline).
7619 Ada reserved words may be used as symbols, with the exceptions of @code{if},
7620 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7623 A symbol declared with this ^switch^switch^ on the command line replaces a
7624 symbol with the same name either in a definition file or specified with a
7625 ^switch^switch^ -D in the preprocessor data file.
7628 This switch is similar to switch @option{^-D^/ASSOCIATE^} of @code{gnatprep}.
7631 When integrated preprocessing is performed and the preprocessor modifies
7632 the source text, write the result of this preprocessing into a file
7633 <source>^.prep^_prep^.
7637 @node Code Generation Control
7638 @subsection Code Generation Control
7642 The GCC technology provides a wide range of target dependent
7643 @option{-m} switches for controlling
7644 details of code generation with respect to different versions of
7645 architectures. This includes variations in instruction sets (e.g.@:
7646 different members of the power pc family), and different requirements
7647 for optimal arrangement of instructions (e.g.@: different members of
7648 the x86 family). The list of available @option{-m} switches may be
7649 found in the GCC documentation.
7651 Use of these @option{-m} switches may in some cases result in improved
7654 The GNAT Pro technology is tested and qualified without any
7655 @option{-m} switches,
7656 so generally the most reliable approach is to avoid the use of these
7657 switches. However, we generally expect most of these switches to work
7658 successfully with GNAT Pro, and many customers have reported successful
7659 use of these options.
7661 Our general advice is to avoid the use of @option{-m} switches unless
7662 special needs lead to requirements in this area. In particular,
7663 there is no point in using @option{-m} switches to improve performance
7664 unless you actually see a performance improvement.
7668 @subsection Return Codes
7669 @cindex Return Codes
7670 @cindex @option{/RETURN_CODES=VMS}
7673 On VMS, GNAT compiled programs return POSIX-style codes by default,
7674 e.g.@: @option{/RETURN_CODES=POSIX}.
7676 To enable VMS style return codes, use GNAT BIND and LINK with the option
7677 @option{/RETURN_CODES=VMS}. For example:
7680 GNAT BIND MYMAIN.ALI /RETURN_CODES=VMS
7681 GNAT LINK MYMAIN.ALI /RETURN_CODES=VMS
7685 Programs built with /RETURN_CODES=VMS are suitable to be called in
7686 VMS DCL scripts. Programs compiled with the default /RETURN_CODES=POSIX
7687 are suitable for spawning with appropriate GNAT RTL routines.
7691 @node Search Paths and the Run-Time Library (RTL)
7692 @section Search Paths and the Run-Time Library (RTL)
7695 With the GNAT source-based library system, the compiler must be able to
7696 find source files for units that are needed by the unit being compiled.
7697 Search paths are used to guide this process.
7699 The compiler compiles one source file whose name must be given
7700 explicitly on the command line. In other words, no searching is done
7701 for this file. To find all other source files that are needed (the most
7702 common being the specs of units), the compiler examines the following
7703 directories, in the following order:
7707 The directory containing the source file of the main unit being compiled
7708 (the file name on the command line).
7711 Each directory named by an @option{^-I^/SOURCE_SEARCH^} switch given on the
7712 @command{gcc} command line, in the order given.
7715 @findex ADA_PRJ_INCLUDE_FILE
7716 Each of the directories listed in the text file whose name is given
7717 by the @env{ADA_PRJ_INCLUDE_FILE} ^environment variable^logical name^.
7720 @env{ADA_PRJ_INCLUDE_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
7721 driver when project files are used. It should not normally be set
7725 @findex ADA_INCLUDE_PATH
7726 Each of the directories listed in the value of the
7727 @env{ADA_INCLUDE_PATH} ^environment variable^logical name^.
7729 Construct this value
7730 exactly as the @env{PATH} environment variable: a list of directory
7731 names separated by colons (semicolons when working with the NT version).
7734 Normally, define this value as a logical name containing a comma separated
7735 list of directory names.
7737 This variable can also be defined by means of an environment string
7738 (an argument to the HP C exec* set of functions).
7742 DEFINE ANOTHER_PATH FOO:[BAG]
7743 DEFINE ADA_INCLUDE_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
7746 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
7747 first, followed by the standard Ada
7748 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADAINCLUDE].
7749 If this is not redefined, the user will obtain the HP Ada 83 IO packages
7750 (Text_IO, Sequential_IO, etc)
7751 instead of the standard Ada packages. Thus, in order to get the standard Ada
7752 packages by default, ADA_INCLUDE_PATH must be redefined.
7756 The content of the @file{ada_source_path} file which is part of the GNAT
7757 installation tree and is used to store standard libraries such as the
7758 GNAT Run Time Library (RTL) source files.
7760 @ref{Installing a library}
7765 Specifying the switch @option{^-I-^/NOCURRENT_DIRECTORY^}
7766 inhibits the use of the directory
7767 containing the source file named in the command line. You can still
7768 have this directory on your search path, but in this case it must be
7769 explicitly requested with a @option{^-I^/SOURCE_SEARCH^} switch.
7771 Specifying the switch @option{-nostdinc}
7772 inhibits the search of the default location for the GNAT Run Time
7773 Library (RTL) source files.
7775 The compiler outputs its object files and ALI files in the current
7778 Caution: The object file can be redirected with the @option{-o} switch;
7779 however, @command{gcc} and @code{gnat1} have not been coordinated on this
7780 so the @file{ALI} file will not go to the right place. Therefore, you should
7781 avoid using the @option{-o} switch.
7785 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
7786 children make up the GNAT RTL, together with the simple @code{System.IO}
7787 package used in the @code{"Hello World"} example. The sources for these units
7788 are needed by the compiler and are kept together in one directory. Not
7789 all of the bodies are needed, but all of the sources are kept together
7790 anyway. In a normal installation, you need not specify these directory
7791 names when compiling or binding. Either the environment variables or
7792 the built-in defaults cause these files to be found.
7794 In addition to the language-defined hierarchies (@code{System}, @code{Ada} and
7795 @code{Interfaces}), the GNAT distribution provides a fourth hierarchy,
7796 consisting of child units of @code{GNAT}. This is a collection of generally
7797 useful types, subprograms, etc. @xref{Top, GNAT Reference Manual, About
7798 This Guid, gnat_rm, GNAT Reference Manual}, for further details.
7800 Besides simplifying access to the RTL, a major use of search paths is
7801 in compiling sources from multiple directories. This can make
7802 development environments much more flexible.
7804 @node Order of Compilation Issues
7805 @section Order of Compilation Issues
7808 If, in our earlier example, there was a spec for the @code{hello}
7809 procedure, it would be contained in the file @file{hello.ads}; yet this
7810 file would not have to be explicitly compiled. This is the result of the
7811 model we chose to implement library management. Some of the consequences
7812 of this model are as follows:
7816 There is no point in compiling specs (except for package
7817 specs with no bodies) because these are compiled as needed by clients. If
7818 you attempt a useless compilation, you will receive an error message.
7819 It is also useless to compile subunits because they are compiled as needed
7823 There are no order of compilation requirements: performing a
7824 compilation never obsoletes anything. The only way you can obsolete
7825 something and require recompilations is to modify one of the
7826 source files on which it depends.
7829 There is no library as such, apart from the ALI files
7830 (@pxref{The Ada Library Information Files}, for information on the format
7831 of these files). For now we find it convenient to create separate ALI files,
7832 but eventually the information therein may be incorporated into the object
7836 When you compile a unit, the source files for the specs of all units
7837 that it @code{with}'s, all its subunits, and the bodies of any generics it
7838 instantiates must be available (reachable by the search-paths mechanism
7839 described above), or you will receive a fatal error message.
7846 The following are some typical Ada compilation command line examples:
7849 @item $ gcc -c xyz.adb
7850 Compile body in file @file{xyz.adb} with all default options.
7853 @item $ gcc -c -O2 -gnata xyz-def.adb
7856 @item $ GNAT COMPILE /OPTIMIZE=ALL -gnata xyz-def.adb
7859 Compile the child unit package in file @file{xyz-def.adb} with extensive
7860 optimizations, and pragma @code{Assert}/@code{Debug} statements
7863 @item $ gcc -c -gnatc abc-def.adb
7864 Compile the subunit in file @file{abc-def.adb} in semantic-checking-only
7868 @node Binding Using gnatbind
7869 @chapter Binding Using @code{gnatbind}
7873 * Running gnatbind::
7874 * Switches for gnatbind::
7875 * Command-Line Access::
7876 * Search Paths for gnatbind::
7877 * Examples of gnatbind Usage::
7881 This chapter describes the GNAT binder, @code{gnatbind}, which is used
7882 to bind compiled GNAT objects.
7884 Note: to invoke @code{gnatbind} with a project file, use the @code{gnat}
7885 driver (see @ref{The GNAT Driver and Project Files}).
7887 The @code{gnatbind} program performs four separate functions:
7891 Checks that a program is consistent, in accordance with the rules in
7892 Chapter 10 of the Ada Reference Manual. In particular, error
7893 messages are generated if a program uses inconsistent versions of a
7897 Checks that an acceptable order of elaboration exists for the program
7898 and issues an error message if it cannot find an order of elaboration
7899 that satisfies the rules in Chapter 10 of the Ada Language Manual.
7902 Generates a main program incorporating the given elaboration order.
7903 This program is a small Ada package (body and spec) that
7904 must be subsequently compiled
7905 using the GNAT compiler. The necessary compilation step is usually
7906 performed automatically by @command{gnatlink}. The two most important
7907 functions of this program
7908 are to call the elaboration routines of units in an appropriate order
7909 and to call the main program.
7912 Determines the set of object files required by the given main program.
7913 This information is output in the forms of comments in the generated program,
7914 to be read by the @command{gnatlink} utility used to link the Ada application.
7917 @node Running gnatbind
7918 @section Running @code{gnatbind}
7921 The form of the @code{gnatbind} command is
7924 @c $ gnatbind @ovar{switches} @var{mainprog}@r{[}.ali@r{]} @ovar{switches}
7925 @c Expanding @ovar macro inline (explanation in macro def comments)
7926 $ gnatbind @r{[}@var{switches}@r{]} @var{mainprog}@r{[}.ali@r{]} @r{[}@var{switches}@r{]}
7930 where @file{@var{mainprog}.adb} is the Ada file containing the main program
7931 unit body. @code{gnatbind} constructs an Ada
7932 package in two files whose names are
7933 @file{b~@var{mainprog}.ads}, and @file{b~@var{mainprog}.adb}.
7934 For example, if given the
7935 parameter @file{hello.ali}, for a main program contained in file
7936 @file{hello.adb}, the binder output files would be @file{b~hello.ads}
7937 and @file{b~hello.adb}.
7939 When doing consistency checking, the binder takes into consideration
7940 any source files it can locate. For example, if the binder determines
7941 that the given main program requires the package @code{Pack}, whose
7943 file is @file{pack.ali} and whose corresponding source spec file is
7944 @file{pack.ads}, it attempts to locate the source file @file{pack.ads}
7945 (using the same search path conventions as previously described for the
7946 @command{gcc} command). If it can locate this source file, it checks that
7948 or source checksums of the source and its references to in @file{ALI} files
7949 match. In other words, any @file{ALI} files that mentions this spec must have
7950 resulted from compiling this version of the source file (or in the case
7951 where the source checksums match, a version close enough that the
7952 difference does not matter).
7954 @cindex Source files, use by binder
7955 The effect of this consistency checking, which includes source files, is
7956 that the binder ensures that the program is consistent with the latest
7957 version of the source files that can be located at bind time. Editing a
7958 source file without compiling files that depend on the source file cause
7959 error messages to be generated by the binder.
7961 For example, suppose you have a main program @file{hello.adb} and a
7962 package @code{P}, from file @file{p.ads} and you perform the following
7967 Enter @code{gcc -c hello.adb} to compile the main program.
7970 Enter @code{gcc -c p.ads} to compile package @code{P}.
7973 Edit file @file{p.ads}.
7976 Enter @code{gnatbind hello}.
7980 At this point, the file @file{p.ali} contains an out-of-date time stamp
7981 because the file @file{p.ads} has been edited. The attempt at binding
7982 fails, and the binder generates the following error messages:
7985 error: "hello.adb" must be recompiled ("p.ads" has been modified)
7986 error: "p.ads" has been modified and must be recompiled
7990 Now both files must be recompiled as indicated, and then the bind can
7991 succeed, generating a main program. You need not normally be concerned
7992 with the contents of this file, but for reference purposes a sample
7993 binder output file is given in @ref{Example of Binder Output File}.
7995 In most normal usage, the default mode of @command{gnatbind} which is to
7996 generate the main package in Ada, as described in the previous section.
7997 In particular, this means that any Ada programmer can read and understand
7998 the generated main program. It can also be debugged just like any other
7999 Ada code provided the @option{^-g^/DEBUG^} switch is used for
8000 @command{gnatbind} and @command{gnatlink}.
8002 @node Switches for gnatbind
8003 @section Switches for @command{gnatbind}
8006 The following switches are available with @code{gnatbind}; details will
8007 be presented in subsequent sections.
8010 * Consistency-Checking Modes::
8011 * Binder Error Message Control::
8012 * Elaboration Control::
8014 * Binding with Non-Ada Main Programs::
8015 * Binding Programs with No Main Subprogram::
8022 @cindex @option{--version} @command{gnatbind}
8023 Display Copyright and version, then exit disregarding all other options.
8026 @cindex @option{--help} @command{gnatbind}
8027 If @option{--version} was not used, display usage, then exit disregarding
8031 @cindex @option{-a} @command{gnatbind}
8032 Indicates that, if supported by the platform, the adainit procedure should
8033 be treated as an initialisation routine by the linker (a constructor). This
8034 is intended to be used by the Project Manager to automatically initialize
8035 shared Stand-Alone Libraries.
8037 @item ^-aO^/OBJECT_SEARCH^
8038 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatbind})
8039 Specify directory to be searched for ALI files.
8041 @item ^-aI^/SOURCE_SEARCH^
8042 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
8043 Specify directory to be searched for source file.
8045 @item ^-b^/REPORT_ERRORS=BRIEF^
8046 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@command{gnatbind})
8047 Generate brief messages to @file{stderr} even if verbose mode set.
8049 @item ^-c^/NOOUTPUT^
8050 @cindex @option{^-c^/NOOUTPUT^} (@command{gnatbind})
8051 Check only, no generation of binder output file.
8053 @item ^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
8054 @cindex @option{^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}} (@command{gnatbind})
8055 This switch can be used to change the default task stack size value
8056 to a specified size @var{nn}, which is expressed in bytes by default, or
8057 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
8059 In the absence of a @samp{@r{[}k@r{|}m@r{]}} suffix, this switch is equivalent,
8060 in effect, to completing all task specs with
8061 @smallexample @c ada
8062 pragma Storage_Size (nn);
8064 When they do not already have such a pragma.
8066 @item ^-D^/DEFAULT_SECONDARY_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
8067 @cindex @option{^-D^/DEFAULT_SECONDARY_STACK_SIZE=nnnnn^} (@command{gnatbind})
8068 This switch can be used to change the default secondary stack size value
8069 to a specified size @var{nn}, which is expressed in bytes by default, or
8070 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
8073 The secondary stack is used to deal with functions that return a variable
8074 sized result, for example a function returning an unconstrained
8075 String. There are two ways in which this secondary stack is allocated.
8077 For most targets, the secondary stack is growing on demand and is allocated
8078 as a chain of blocks in the heap. The -D option is not very
8079 relevant. It only give some control over the size of the allocated
8080 blocks (whose size is the minimum of the default secondary stack size value,
8081 and the actual size needed for the current allocation request).
8083 For certain targets, notably VxWorks 653,
8084 the secondary stack is allocated by carving off a fixed ratio chunk of the
8085 primary task stack. The -D option is used to define the
8086 size of the environment task's secondary stack.
8088 @item ^-e^/ELABORATION_DEPENDENCIES^
8089 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@command{gnatbind})
8090 Output complete list of elaboration-order dependencies.
8092 @item ^-E^/STORE_TRACEBACKS^
8093 @cindex @option{^-E^/STORE_TRACEBACKS^} (@command{gnatbind})
8094 Store tracebacks in exception occurrences when the target supports it.
8095 This is the default with the zero cost exception mechanism.
8097 @c The following may get moved to an appendix
8098 This option is currently supported on the following targets:
8099 all x86 ports, Solaris, Windows, HP-UX, AIX, PowerPC VxWorks and Alpha VxWorks.
8101 See also the packages @code{GNAT.Traceback} and
8102 @code{GNAT.Traceback.Symbolic} for more information.
8104 Note that on x86 ports, you must not use @option{-fomit-frame-pointer}
8105 @command{gcc} option.
8108 @item ^-F^/FORCE_ELABS_FLAGS^
8109 @cindex @option{^-F^/FORCE_ELABS_FLAGS^} (@command{gnatbind})
8110 Force the checks of elaboration flags. @command{gnatbind} does not normally
8111 generate checks of elaboration flags for the main executable, except when
8112 a Stand-Alone Library is used. However, there are cases when this cannot be
8113 detected by gnatbind. An example is importing an interface of a Stand-Alone
8114 Library through a pragma Import and only specifying through a linker switch
8115 this Stand-Alone Library. This switch is used to guarantee that elaboration
8116 flag checks are generated.
8119 @cindex @option{^-h^/HELP^} (@command{gnatbind})
8120 Output usage (help) information
8123 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8124 Specify directory to be searched for source and ALI files.
8126 @item ^-I-^/NOCURRENT_DIRECTORY^
8127 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatbind})
8128 Do not look for sources in the current directory where @code{gnatbind} was
8129 invoked, and do not look for ALI files in the directory containing the
8130 ALI file named in the @code{gnatbind} command line.
8132 @item ^-l^/ORDER_OF_ELABORATION^
8133 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@command{gnatbind})
8134 Output chosen elaboration order.
8136 @item ^-L@var{xxx}^/BUILD_LIBRARY=@var{xxx}^
8137 @cindex @option{^-L^/BUILD_LIBRARY^} (@command{gnatbind})
8138 Bind the units for library building. In this case the adainit and
8139 adafinal procedures (@pxref{Binding with Non-Ada Main Programs})
8140 are renamed to ^@var{xxx}init^@var{XXX}INIT^ and
8141 ^@var{xxx}final^@var{XXX}FINAL^.
8142 Implies ^-n^/NOCOMPILE^.
8144 (@xref{GNAT and Libraries}, for more details.)
8147 On OpenVMS, these init and final procedures are exported in uppercase
8148 letters. For example if /BUILD_LIBRARY=toto is used, the exported name of
8149 the init procedure will be "TOTOINIT" and the exported name of the final
8150 procedure will be "TOTOFINAL".
8153 @item ^-Mxyz^/RENAME_MAIN=xyz^
8154 @cindex @option{^-M^/RENAME_MAIN^} (@command{gnatbind})
8155 Rename generated main program from main to xyz. This option is
8156 supported on cross environments only.
8158 @item ^-m^/ERROR_LIMIT=^@var{n}
8159 @cindex @option{^-m^/ERROR_LIMIT^} (@command{gnatbind})
8160 Limit number of detected errors or warnings to @var{n}, where @var{n} is
8161 in the range 1..999999. The default value if no switch is
8162 given is 9999. If the number of warnings reaches this limit, then a
8163 message is output and further warnings are suppressed, the bind
8164 continues in this case. If the number of errors reaches this
8165 limit, then a message is output and the bind is abandoned.
8166 A value of zero means that no limit is enforced. The equal
8170 Furthermore, under Windows, the sources pointed to by the libraries path
8171 set in the registry are not searched for.
8175 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8179 @cindex @option{-nostdinc} (@command{gnatbind})
8180 Do not look for sources in the system default directory.
8183 @cindex @option{-nostdlib} (@command{gnatbind})
8184 Do not look for library files in the system default directory.
8186 @item --RTS=@var{rts-path}
8187 @cindex @option{--RTS} (@code{gnatbind})
8188 Specifies the default location of the runtime library. Same meaning as the
8189 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
8191 @item ^-o ^/OUTPUT=^@var{file}
8192 @cindex @option{^-o ^/OUTPUT^} (@command{gnatbind})
8193 Name the output file @var{file} (default is @file{b~@var{xxx}.adb}).
8194 Note that if this option is used, then linking must be done manually,
8195 gnatlink cannot be used.
8197 @item ^-O^/OBJECT_LIST^
8198 @cindex @option{^-O^/OBJECT_LIST^} (@command{gnatbind})
8201 @item ^-p^/PESSIMISTIC_ELABORATION^
8202 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@command{gnatbind})
8203 Pessimistic (worst-case) elaboration order
8206 @cindex @option{^-R^-R^} (@command{gnatbind})
8207 Output closure source list.
8209 @item ^-s^/READ_SOURCES=ALL^
8210 @cindex @option{^-s^/READ_SOURCES=ALL^} (@command{gnatbind})
8211 Require all source files to be present.
8213 @item ^-S@var{xxx}^/INITIALIZE_SCALARS=@var{xxx}^
8214 @cindex @option{^-S^/INITIALIZE_SCALARS^} (@command{gnatbind})
8215 Specifies the value to be used when detecting uninitialized scalar
8216 objects with pragma Initialize_Scalars.
8217 The @var{xxx} ^string specified with the switch^option^ may be either
8219 @item ``@option{^in^INVALID^}'' requesting an invalid value where possible
8220 @item ``@option{^lo^LOW^}'' for the lowest possible value
8221 @item ``@option{^hi^HIGH^}'' for the highest possible value
8222 @item ``@option{@var{xx}}'' for a value consisting of repeated bytes with the
8223 value @code{16#@var{xx}#} (i.e., @var{xx} is a string of two hexadecimal digits).
8226 In addition, you can specify @option{-Sev} to indicate that the value is
8227 to be set at run time. In this case, the program will look for an environment
8228 @cindex GNAT_INIT_SCALARS
8229 variable of the form @env{GNAT_INIT_SCALARS=@var{xx}}, where @var{xx} is one
8230 of @option{in/lo/hi/@var{xx}} with the same meanings as above.
8231 If no environment variable is found, or if it does not have a valid value,
8232 then the default is @option{in} (invalid values).
8236 @cindex @option{-static} (@code{gnatbind})
8237 Link against a static GNAT run time.
8240 @cindex @option{-shared} (@code{gnatbind})
8241 Link against a shared GNAT run time when available.
8244 @item ^-t^/NOTIME_STAMP_CHECK^
8245 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8246 Tolerate time stamp and other consistency errors
8248 @item ^-T@var{n}^/TIME_SLICE=@var{n}^
8249 @cindex @option{^-T^/TIME_SLICE^} (@code{gnatbind})
8250 Set the time slice value to @var{n} milliseconds. If the system supports
8251 the specification of a specific time slice value, then the indicated value
8252 is used. If the system does not support specific time slice values, but
8253 does support some general notion of round-robin scheduling, then any
8254 nonzero value will activate round-robin scheduling.
8256 A value of zero is treated specially. It turns off time
8257 slicing, and in addition, indicates to the tasking run time that the
8258 semantics should match as closely as possible the Annex D
8259 requirements of the Ada RM, and in particular sets the default
8260 scheduling policy to @code{FIFO_Within_Priorities}.
8262 @item ^-u@var{n}^/DYNAMIC_STACK_USAGE=@var{n}^
8263 @cindex @option{^-u^/DYNAMIC_STACK_USAGE^} (@code{gnatbind})
8264 Enable dynamic stack usage, with @var{n} results stored and displayed
8265 at program termination. A result is generated when a task
8266 terminates. Results that can't be stored are displayed on the fly, at
8267 task termination. This option is currently not supported on Itanium
8268 platforms. (See @ref{Dynamic Stack Usage Analysis} for details.)
8270 @item ^-v^/REPORT_ERRORS=VERBOSE^
8271 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8272 Verbose mode. Write error messages, header, summary output to
8277 @cindex @option{-w} (@code{gnatbind})
8278 Warning mode (@var{x}=s/e for suppress/treat as error)
8282 @item /WARNINGS=NORMAL
8283 @cindex @option{/WARNINGS} (@code{gnatbind})
8284 Normal warnings mode. Warnings are issued but ignored
8286 @item /WARNINGS=SUPPRESS
8287 @cindex @option{/WARNINGS} (@code{gnatbind})
8288 All warning messages are suppressed
8290 @item /WARNINGS=ERROR
8291 @cindex @option{/WARNINGS} (@code{gnatbind})
8292 Warning messages are treated as fatal errors
8295 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8296 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8297 Override default wide character encoding for standard Text_IO files.
8299 @item ^-x^/READ_SOURCES=NONE^
8300 @cindex @option{^-x^/READ_SOURCES^} (@code{gnatbind})
8301 Exclude source files (check object consistency only).
8304 @item /READ_SOURCES=AVAILABLE
8305 @cindex @option{/READ_SOURCES} (@code{gnatbind})
8306 Default mode, in which sources are checked for consistency only if
8310 @item ^-y^/ENABLE_LEAP_SECONDS^
8311 @cindex @option{^-y^/ENABLE_LEAP_SECONDS^} (@code{gnatbind})
8312 Enable leap seconds support in @code{Ada.Calendar} and its children.
8314 @item ^-z^/ZERO_MAIN^
8315 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8321 You may obtain this listing of switches by running @code{gnatbind} with
8325 @node Consistency-Checking Modes
8326 @subsection Consistency-Checking Modes
8329 As described earlier, by default @code{gnatbind} checks
8330 that object files are consistent with one another and are consistent
8331 with any source files it can locate. The following switches control binder
8336 @item ^-s^/READ_SOURCES=ALL^
8337 @cindex @option{^-s^/READ_SOURCES=ALL^} (@code{gnatbind})
8338 Require source files to be present. In this mode, the binder must be
8339 able to locate all source files that are referenced, in order to check
8340 their consistency. In normal mode, if a source file cannot be located it
8341 is simply ignored. If you specify this switch, a missing source
8344 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8345 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8346 Override default wide character encoding for standard Text_IO files.
8347 Normally the default wide character encoding method used for standard
8348 [Wide_[Wide_]]Text_IO files is taken from the encoding specified for
8349 the main source input (see description of switch
8350 @option{^-gnatWx^/WIDE_CHARACTER_ENCODING^} for the compiler). The
8351 use of this switch for the binder (which has the same set of
8352 possible arguments) overrides this default as specified.
8354 @item ^-x^/READ_SOURCES=NONE^
8355 @cindex @option{^-x^/READ_SOURCES=NONE^} (@code{gnatbind})
8356 Exclude source files. In this mode, the binder only checks that ALI
8357 files are consistent with one another. Source files are not accessed.
8358 The binder runs faster in this mode, and there is still a guarantee that
8359 the resulting program is self-consistent.
8360 If a source file has been edited since it was last compiled, and you
8361 specify this switch, the binder will not detect that the object
8362 file is out of date with respect to the source file. Note that this is the
8363 mode that is automatically used by @command{gnatmake} because in this
8364 case the checking against sources has already been performed by
8365 @command{gnatmake} in the course of compilation (i.e.@: before binding).
8368 @item /READ_SOURCES=AVAILABLE
8369 @cindex @code{/READ_SOURCES=AVAILABLE} (@code{gnatbind})
8370 This is the default mode in which source files are checked if they are
8371 available, and ignored if they are not available.
8375 @node Binder Error Message Control
8376 @subsection Binder Error Message Control
8379 The following switches provide control over the generation of error
8380 messages from the binder:
8384 @item ^-v^/REPORT_ERRORS=VERBOSE^
8385 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8386 Verbose mode. In the normal mode, brief error messages are generated to
8387 @file{stderr}. If this switch is present, a header is written
8388 to @file{stdout} and any error messages are directed to @file{stdout}.
8389 All that is written to @file{stderr} is a brief summary message.
8391 @item ^-b^/REPORT_ERRORS=BRIEF^
8392 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@code{gnatbind})
8393 Generate brief error messages to @file{stderr} even if verbose mode is
8394 specified. This is relevant only when used with the
8395 @option{^-v^/REPORT_ERRORS=VERBOSE^} switch.
8399 @cindex @option{-m} (@code{gnatbind})
8400 Limits the number of error messages to @var{n}, a decimal integer in the
8401 range 1-999. The binder terminates immediately if this limit is reached.
8404 @cindex @option{-M} (@code{gnatbind})
8405 Renames the generated main program from @code{main} to @code{xxx}.
8406 This is useful in the case of some cross-building environments, where
8407 the actual main program is separate from the one generated
8411 @item ^-ws^/WARNINGS=SUPPRESS^
8412 @cindex @option{^-ws^/WARNINGS=SUPPRESS^} (@code{gnatbind})
8414 Suppress all warning messages.
8416 @item ^-we^/WARNINGS=ERROR^
8417 @cindex @option{^-we^/WARNINGS=ERROR^} (@code{gnatbind})
8418 Treat any warning messages as fatal errors.
8421 @item /WARNINGS=NORMAL
8422 Standard mode with warnings generated, but warnings do not get treated
8426 @item ^-t^/NOTIME_STAMP_CHECK^
8427 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8428 @cindex Time stamp checks, in binder
8429 @cindex Binder consistency checks
8430 @cindex Consistency checks, in binder
8431 The binder performs a number of consistency checks including:
8435 Check that time stamps of a given source unit are consistent
8437 Check that checksums of a given source unit are consistent
8439 Check that consistent versions of @code{GNAT} were used for compilation
8441 Check consistency of configuration pragmas as required
8445 Normally failure of such checks, in accordance with the consistency
8446 requirements of the Ada Reference Manual, causes error messages to be
8447 generated which abort the binder and prevent the output of a binder
8448 file and subsequent link to obtain an executable.
8450 The @option{^-t^/NOTIME_STAMP_CHECK^} switch converts these error messages
8451 into warnings, so that
8452 binding and linking can continue to completion even in the presence of such
8453 errors. The result may be a failed link (due to missing symbols), or a
8454 non-functional executable which has undefined semantics.
8455 @emph{This means that
8456 @option{^-t^/NOTIME_STAMP_CHECK^} should be used only in unusual situations,
8460 @node Elaboration Control
8461 @subsection Elaboration Control
8464 The following switches provide additional control over the elaboration
8465 order. For full details see @ref{Elaboration Order Handling in GNAT}.
8468 @item ^-p^/PESSIMISTIC_ELABORATION^
8469 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@code{gnatbind})
8470 Normally the binder attempts to choose an elaboration order that is
8471 likely to minimize the likelihood of an elaboration order error resulting
8472 in raising a @code{Program_Error} exception. This switch reverses the
8473 action of the binder, and requests that it deliberately choose an order
8474 that is likely to maximize the likelihood of an elaboration error.
8475 This is useful in ensuring portability and avoiding dependence on
8476 accidental fortuitous elaboration ordering.
8478 Normally it only makes sense to use the @option{^-p^/PESSIMISTIC_ELABORATION^}
8480 elaboration checking is used (@option{-gnatE} switch used for compilation).
8481 This is because in the default static elaboration mode, all necessary
8482 @code{Elaborate} and @code{Elaborate_All} pragmas are implicitly inserted.
8483 These implicit pragmas are still respected by the binder in
8484 @option{^-p^/PESSIMISTIC_ELABORATION^} mode, so a
8485 safe elaboration order is assured.
8488 @node Output Control
8489 @subsection Output Control
8492 The following switches allow additional control over the output
8493 generated by the binder.
8498 @item ^-c^/NOOUTPUT^
8499 @cindex @option{^-c^/NOOUTPUT^} (@code{gnatbind})
8500 Check only. Do not generate the binder output file. In this mode the
8501 binder performs all error checks but does not generate an output file.
8503 @item ^-e^/ELABORATION_DEPENDENCIES^
8504 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@code{gnatbind})
8505 Output complete list of elaboration-order dependencies, showing the
8506 reason for each dependency. This output can be rather extensive but may
8507 be useful in diagnosing problems with elaboration order. The output is
8508 written to @file{stdout}.
8511 @cindex @option{^-h^/HELP^} (@code{gnatbind})
8512 Output usage information. The output is written to @file{stdout}.
8514 @item ^-K^/LINKER_OPTION_LIST^
8515 @cindex @option{^-K^/LINKER_OPTION_LIST^} (@code{gnatbind})
8516 Output linker options to @file{stdout}. Includes library search paths,
8517 contents of pragmas Ident and Linker_Options, and libraries added
8520 @item ^-l^/ORDER_OF_ELABORATION^
8521 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@code{gnatbind})
8522 Output chosen elaboration order. The output is written to @file{stdout}.
8524 @item ^-O^/OBJECT_LIST^
8525 @cindex @option{^-O^/OBJECT_LIST^} (@code{gnatbind})
8526 Output full names of all the object files that must be linked to provide
8527 the Ada component of the program. The output is written to @file{stdout}.
8528 This list includes the files explicitly supplied and referenced by the user
8529 as well as implicitly referenced run-time unit files. The latter are
8530 omitted if the corresponding units reside in shared libraries. The
8531 directory names for the run-time units depend on the system configuration.
8533 @item ^-o ^/OUTPUT=^@var{file}
8534 @cindex @option{^-o^/OUTPUT^} (@code{gnatbind})
8535 Set name of output file to @var{file} instead of the normal
8536 @file{b~@var{mainprog}.adb} default. Note that @var{file} denote the Ada
8537 binder generated body filename.
8538 Note that if this option is used, then linking must be done manually.
8539 It is not possible to use gnatlink in this case, since it cannot locate
8542 @item ^-r^/RESTRICTION_LIST^
8543 @cindex @option{^-r^/RESTRICTION_LIST^} (@code{gnatbind})
8544 Generate list of @code{pragma Restrictions} that could be applied to
8545 the current unit. This is useful for code audit purposes, and also may
8546 be used to improve code generation in some cases.
8550 @node Binding with Non-Ada Main Programs
8551 @subsection Binding with Non-Ada Main Programs
8554 In our description so far we have assumed that the main
8555 program is in Ada, and that the task of the binder is to generate a
8556 corresponding function @code{main} that invokes this Ada main
8557 program. GNAT also supports the building of executable programs where
8558 the main program is not in Ada, but some of the called routines are
8559 written in Ada and compiled using GNAT (@pxref{Mixed Language Programming}).
8560 The following switch is used in this situation:
8564 @cindex @option{^-n^/NOMAIN^} (@code{gnatbind})
8565 No main program. The main program is not in Ada.
8569 In this case, most of the functions of the binder are still required,
8570 but instead of generating a main program, the binder generates a file
8571 containing the following callable routines:
8576 You must call this routine to initialize the Ada part of the program by
8577 calling the necessary elaboration routines. A call to @code{adainit} is
8578 required before the first call to an Ada subprogram.
8580 Note that it is assumed that the basic execution environment must be setup
8581 to be appropriate for Ada execution at the point where the first Ada
8582 subprogram is called. In particular, if the Ada code will do any
8583 floating-point operations, then the FPU must be setup in an appropriate
8584 manner. For the case of the x86, for example, full precision mode is
8585 required. The procedure GNAT.Float_Control.Reset may be used to ensure
8586 that the FPU is in the right state.
8590 You must call this routine to perform any library-level finalization
8591 required by the Ada subprograms. A call to @code{adafinal} is required
8592 after the last call to an Ada subprogram, and before the program
8597 If the @option{^-n^/NOMAIN^} switch
8598 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8599 @cindex Binder, multiple input files
8600 is given, more than one ALI file may appear on
8601 the command line for @code{gnatbind}. The normal @dfn{closure}
8602 calculation is performed for each of the specified units. Calculating
8603 the closure means finding out the set of units involved by tracing
8604 @code{with} references. The reason it is necessary to be able to
8605 specify more than one ALI file is that a given program may invoke two or
8606 more quite separate groups of Ada units.
8608 The binder takes the name of its output file from the last specified ALI
8609 file, unless overridden by the use of the @option{^-o file^/OUTPUT=file^}.
8610 @cindex @option{^-o^/OUTPUT^} (@command{gnatbind})
8611 The output is an Ada unit in source form that can be compiled with GNAT.
8612 This compilation occurs automatically as part of the @command{gnatlink}
8615 Currently the GNAT run time requires a FPU using 80 bits mode
8616 precision. Under targets where this is not the default it is required to
8617 call GNAT.Float_Control.Reset before using floating point numbers (this
8618 include float computation, float input and output) in the Ada code. A
8619 side effect is that this could be the wrong mode for the foreign code
8620 where floating point computation could be broken after this call.
8622 @node Binding Programs with No Main Subprogram
8623 @subsection Binding Programs with No Main Subprogram
8626 It is possible to have an Ada program which does not have a main
8627 subprogram. This program will call the elaboration routines of all the
8628 packages, then the finalization routines.
8630 The following switch is used to bind programs organized in this manner:
8633 @item ^-z^/ZERO_MAIN^
8634 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8635 Normally the binder checks that the unit name given on the command line
8636 corresponds to a suitable main subprogram. When this switch is used,
8637 a list of ALI files can be given, and the execution of the program
8638 consists of elaboration of these units in an appropriate order. Note
8639 that the default wide character encoding method for standard Text_IO
8640 files is always set to Brackets if this switch is set (you can use
8642 @option{^-Wx^WIDE_CHARACTER_ENCODING^} to override this default).
8645 @node Command-Line Access
8646 @section Command-Line Access
8649 The package @code{Ada.Command_Line} provides access to the command-line
8650 arguments and program name. In order for this interface to operate
8651 correctly, the two variables
8663 are declared in one of the GNAT library routines. These variables must
8664 be set from the actual @code{argc} and @code{argv} values passed to the
8665 main program. With no @option{^n^/NOMAIN^} present, @code{gnatbind}
8666 generates the C main program to automatically set these variables.
8667 If the @option{^n^/NOMAIN^} switch is used, there is no automatic way to
8668 set these variables. If they are not set, the procedures in
8669 @code{Ada.Command_Line} will not be available, and any attempt to use
8670 them will raise @code{Constraint_Error}. If command line access is
8671 required, your main program must set @code{gnat_argc} and
8672 @code{gnat_argv} from the @code{argc} and @code{argv} values passed to
8675 @node Search Paths for gnatbind
8676 @section Search Paths for @code{gnatbind}
8679 The binder takes the name of an ALI file as its argument and needs to
8680 locate source files as well as other ALI files to verify object consistency.
8682 For source files, it follows exactly the same search rules as @command{gcc}
8683 (@pxref{Search Paths and the Run-Time Library (RTL)}). For ALI files the
8684 directories searched are:
8688 The directory containing the ALI file named in the command line, unless
8689 the switch @option{^-I-^/NOCURRENT_DIRECTORY^} is specified.
8692 All directories specified by @option{^-I^/SEARCH^}
8693 switches on the @code{gnatbind}
8694 command line, in the order given.
8697 @findex ADA_PRJ_OBJECTS_FILE
8698 Each of the directories listed in the text file whose name is given
8699 by the @env{ADA_PRJ_OBJECTS_FILE} ^environment variable^logical name^.
8702 @env{ADA_PRJ_OBJECTS_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
8703 driver when project files are used. It should not normally be set
8707 @findex ADA_OBJECTS_PATH
8708 Each of the directories listed in the value of the
8709 @env{ADA_OBJECTS_PATH} ^environment variable^logical name^.
8711 Construct this value
8712 exactly as the @env{PATH} environment variable: a list of directory
8713 names separated by colons (semicolons when working with the NT version
8717 Normally, define this value as a logical name containing a comma separated
8718 list of directory names.
8720 This variable can also be defined by means of an environment string
8721 (an argument to the HP C exec* set of functions).
8725 DEFINE ANOTHER_PATH FOO:[BAG]
8726 DEFINE ADA_OBJECTS_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
8729 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
8730 first, followed by the standard Ada
8731 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADALIB].
8732 If this is not redefined, the user will obtain the HP Ada 83 IO packages
8733 (Text_IO, Sequential_IO, etc)
8734 instead of the standard Ada packages. Thus, in order to get the standard Ada
8735 packages by default, ADA_OBJECTS_PATH must be redefined.
8739 The content of the @file{ada_object_path} file which is part of the GNAT
8740 installation tree and is used to store standard libraries such as the
8741 GNAT Run Time Library (RTL) unless the switch @option{-nostdlib} is
8744 @ref{Installing a library}
8749 In the binder the switch @option{^-I^/SEARCH^}
8750 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8751 is used to specify both source and
8752 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
8753 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
8754 instead if you want to specify
8755 source paths only, and @option{^-aO^/LIBRARY_SEARCH^}
8756 @cindex @option{^-aO^/LIBRARY_SEARCH^} (@command{gnatbind})
8757 if you want to specify library paths
8758 only. This means that for the binder
8759 @option{^-I^/SEARCH=^}@var{dir} is equivalent to
8760 @option{^-aI^/SOURCE_SEARCH=^}@var{dir}
8761 @option{^-aO^/OBJECT_SEARCH=^}@var{dir}.
8762 The binder generates the bind file (a C language source file) in the
8763 current working directory.
8769 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
8770 children make up the GNAT Run-Time Library, together with the package
8771 GNAT and its children, which contain a set of useful additional
8772 library functions provided by GNAT. The sources for these units are
8773 needed by the compiler and are kept together in one directory. The ALI
8774 files and object files generated by compiling the RTL are needed by the
8775 binder and the linker and are kept together in one directory, typically
8776 different from the directory containing the sources. In a normal
8777 installation, you need not specify these directory names when compiling
8778 or binding. Either the environment variables or the built-in defaults
8779 cause these files to be found.
8781 Besides simplifying access to the RTL, a major use of search paths is
8782 in compiling sources from multiple directories. This can make
8783 development environments much more flexible.
8785 @node Examples of gnatbind Usage
8786 @section Examples of @code{gnatbind} Usage
8789 This section contains a number of examples of using the GNAT binding
8790 utility @code{gnatbind}.
8793 @item gnatbind hello
8794 The main program @code{Hello} (source program in @file{hello.adb}) is
8795 bound using the standard switch settings. The generated main program is
8796 @file{b~hello.adb}. This is the normal, default use of the binder.
8799 @item gnatbind hello -o mainprog.adb
8802 @item gnatbind HELLO.ALI /OUTPUT=Mainprog.ADB
8804 The main program @code{Hello} (source program in @file{hello.adb}) is
8805 bound using the standard switch settings. The generated main program is
8806 @file{mainprog.adb} with the associated spec in
8807 @file{mainprog.ads}. Note that you must specify the body here not the
8808 spec. Note that if this option is used, then linking must be done manually,
8809 since gnatlink will not be able to find the generated file.
8812 @c ------------------------------------
8813 @node Linking Using gnatlink
8814 @chapter Linking Using @command{gnatlink}
8815 @c ------------------------------------
8819 This chapter discusses @command{gnatlink}, a tool that links
8820 an Ada program and builds an executable file. This utility
8821 invokes the system linker ^(via the @command{gcc} command)^^
8822 with a correct list of object files and library references.
8823 @command{gnatlink} automatically determines the list of files and
8824 references for the Ada part of a program. It uses the binder file
8825 generated by the @command{gnatbind} to determine this list.
8827 Note: to invoke @code{gnatlink} with a project file, use the @code{gnat}
8828 driver (see @ref{The GNAT Driver and Project Files}).
8831 * Running gnatlink::
8832 * Switches for gnatlink::
8835 @node Running gnatlink
8836 @section Running @command{gnatlink}
8839 The form of the @command{gnatlink} command is
8842 @c $ gnatlink @ovar{switches} @var{mainprog}@r{[}.ali@r{]}
8843 @c @ovar{non-Ada objects} @ovar{linker options}
8844 @c Expanding @ovar macro inline (explanation in macro def comments)
8845 $ gnatlink @r{[}@var{switches}@r{]} @var{mainprog}@r{[}.ali@r{]}
8846 @r{[}@var{non-Ada objects}@r{]} @r{[}@var{linker options}@r{]}
8851 The arguments of @command{gnatlink} (switches, main @file{ALI} file,
8853 or linker options) may be in any order, provided that no non-Ada object may
8854 be mistaken for a main @file{ALI} file.
8855 Any file name @file{F} without the @file{.ali}
8856 extension will be taken as the main @file{ALI} file if a file exists
8857 whose name is the concatenation of @file{F} and @file{.ali}.
8860 @file{@var{mainprog}.ali} references the ALI file of the main program.
8861 The @file{.ali} extension of this file can be omitted. From this
8862 reference, @command{gnatlink} locates the corresponding binder file
8863 @file{b~@var{mainprog}.adb} and, using the information in this file along
8864 with the list of non-Ada objects and linker options, constructs a
8865 linker command file to create the executable.
8867 The arguments other than the @command{gnatlink} switches and the main
8868 @file{ALI} file are passed to the linker uninterpreted.
8869 They typically include the names of
8870 object files for units written in other languages than Ada and any library
8871 references required to resolve references in any of these foreign language
8872 units, or in @code{Import} pragmas in any Ada units.
8874 @var{linker options} is an optional list of linker specific
8876 The default linker called by gnatlink is @command{gcc} which in
8877 turn calls the appropriate system linker.
8878 Standard options for the linker such as @option{-lmy_lib} or
8879 @option{-Ldir} can be added as is.
8880 For options that are not recognized by
8881 @command{gcc} as linker options, use the @command{gcc} switches
8882 @option{-Xlinker} or @option{-Wl,}.
8883 Refer to the GCC documentation for
8884 details. Here is an example showing how to generate a linker map:
8887 $ ^gnatlink my_prog -Wl,-Map,MAPFILE^GNAT LINK my_prog.ali /MAP^
8890 Using @var{linker options} it is possible to set the program stack and
8893 See @ref{Setting Stack Size from gnatlink} and
8894 @ref{Setting Heap Size from gnatlink}.
8897 @command{gnatlink} determines the list of objects required by the Ada
8898 program and prepends them to the list of objects passed to the linker.
8899 @command{gnatlink} also gathers any arguments set by the use of
8900 @code{pragma Linker_Options} and adds them to the list of arguments
8901 presented to the linker.
8904 @command{gnatlink} accepts the following types of extra files on the command
8905 line: objects (@file{.OBJ}), libraries (@file{.OLB}), sharable images
8906 (@file{.EXE}), and options files (@file{.OPT}). These are recognized and
8907 handled according to their extension.
8910 @node Switches for gnatlink
8911 @section Switches for @command{gnatlink}
8914 The following switches are available with the @command{gnatlink} utility:
8920 @cindex @option{--version} @command{gnatlink}
8921 Display Copyright and version, then exit disregarding all other options.
8924 @cindex @option{--help} @command{gnatlink}
8925 If @option{--version} was not used, display usage, then exit disregarding
8928 @item ^-f^/FORCE_OBJECT_FILE_LIST^
8929 @cindex Command line length
8930 @cindex @option{^-f^/FORCE_OBJECT_FILE_LIST^} (@command{gnatlink})
8931 On some targets, the command line length is limited, and @command{gnatlink}
8932 will generate a separate file for the linker if the list of object files
8934 The @option{^-f^/FORCE_OBJECT_FILE_LIST^} switch forces this file
8935 to be generated even if
8936 the limit is not exceeded. This is useful in some cases to deal with
8937 special situations where the command line length is exceeded.
8940 @cindex Debugging information, including
8941 @cindex @option{^-g^/DEBUG^} (@command{gnatlink})
8942 The option to include debugging information causes the Ada bind file (in
8943 other words, @file{b~@var{mainprog}.adb}) to be compiled with
8944 @option{^-g^/DEBUG^}.
8945 In addition, the binder does not delete the @file{b~@var{mainprog}.adb},
8946 @file{b~@var{mainprog}.o} and @file{b~@var{mainprog}.ali} files.
8947 Without @option{^-g^/DEBUG^}, the binder removes these files by
8948 default. The same procedure apply if a C bind file was generated using
8949 @option{^-C^/BIND_FILE=C^} @code{gnatbind} option, in this case the filenames
8950 are @file{b_@var{mainprog}.c} and @file{b_@var{mainprog}.o}.
8952 @item ^-n^/NOCOMPILE^
8953 @cindex @option{^-n^/NOCOMPILE^} (@command{gnatlink})
8954 Do not compile the file generated by the binder. This may be used when
8955 a link is rerun with different options, but there is no need to recompile
8959 @cindex @option{^-v^/VERBOSE^} (@command{gnatlink})
8960 Causes additional information to be output, including a full list of the
8961 included object files. This switch option is most useful when you want
8962 to see what set of object files are being used in the link step.
8964 @item ^-v -v^/VERBOSE/VERBOSE^
8965 @cindex @option{^-v -v^/VERBOSE/VERBOSE^} (@command{gnatlink})
8966 Very verbose mode. Requests that the compiler operate in verbose mode when
8967 it compiles the binder file, and that the system linker run in verbose mode.
8969 @item ^-o ^/EXECUTABLE=^@var{exec-name}
8970 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatlink})
8971 @var{exec-name} specifies an alternate name for the generated
8972 executable program. If this switch is omitted, the executable has the same
8973 name as the main unit. For example, @code{gnatlink try.ali} creates
8974 an executable called @file{^try^TRY.EXE^}.
8977 @item -b @var{target}
8978 @cindex @option{-b} (@command{gnatlink})
8979 Compile your program to run on @var{target}, which is the name of a
8980 system configuration. You must have a GNAT cross-compiler built if
8981 @var{target} is not the same as your host system.
8984 @cindex @option{-B} (@command{gnatlink})
8985 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
8986 from @var{dir} instead of the default location. Only use this switch
8987 when multiple versions of the GNAT compiler are available.
8988 @xref{Directory Options,,, gcc, The GNU Compiler Collection},
8989 for further details. You would normally use the @option{-b} or
8990 @option{-V} switch instead.
8992 @item --GCC=@var{compiler_name}
8993 @cindex @option{--GCC=compiler_name} (@command{gnatlink})
8994 Program used for compiling the binder file. The default is
8995 @command{gcc}. You need to use quotes around @var{compiler_name} if
8996 @code{compiler_name} contains spaces or other separator characters.
8997 As an example @option{--GCC="foo -x -y"} will instruct @command{gnatlink} to
8998 use @code{foo -x -y} as your compiler. Note that switch @option{-c} is always
8999 inserted after your command name. Thus in the above example the compiler
9000 command that will be used by @command{gnatlink} will be @code{foo -c -x -y}.
9001 A limitation of this syntax is that the name and path name of the executable
9002 itself must not include any embedded spaces. If the compiler executable is
9003 different from the default one (gcc or <prefix>-gcc), then the back-end
9004 switches in the ALI file are not used to compile the binder generated source.
9005 For example, this is the case with @option{--GCC="foo -x -y"}. But the back end
9006 switches will be used for @option{--GCC="gcc -gnatv"}. If several
9007 @option{--GCC=compiler_name} are used, only the last @var{compiler_name}
9008 is taken into account. However, all the additional switches are also taken
9010 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
9011 @option{--GCC="bar -x -y -z -t"}.
9013 @item --LINK=@var{name}
9014 @cindex @option{--LINK=} (@command{gnatlink})
9015 @var{name} is the name of the linker to be invoked. This is especially
9016 useful in mixed language programs since languages such as C++ require
9017 their own linker to be used. When this switch is omitted, the default
9018 name for the linker is @command{gcc}. When this switch is used, the
9019 specified linker is called instead of @command{gcc} with exactly the same
9020 parameters that would have been passed to @command{gcc} so if the desired
9021 linker requires different parameters it is necessary to use a wrapper
9022 script that massages the parameters before invoking the real linker. It
9023 may be useful to control the exact invocation by using the verbose
9029 @item /DEBUG=TRACEBACK
9030 @cindex @code{/DEBUG=TRACEBACK} (@command{gnatlink})
9031 This qualifier causes sufficient information to be included in the
9032 executable file to allow a traceback, but does not include the full
9033 symbol information needed by the debugger.
9035 @item /IDENTIFICATION="<string>"
9036 @code{"<string>"} specifies the string to be stored in the image file
9037 identification field in the image header.
9038 It overrides any pragma @code{Ident} specified string.
9040 @item /NOINHIBIT-EXEC
9041 Generate the executable file even if there are linker warnings.
9043 @item /NOSTART_FILES
9044 Don't link in the object file containing the ``main'' transfer address.
9045 Used when linking with a foreign language main program compiled with an
9049 Prefer linking with object libraries over sharable images, even without
9055 @node The GNAT Make Program gnatmake
9056 @chapter The GNAT Make Program @command{gnatmake}
9060 * Running gnatmake::
9061 * Switches for gnatmake::
9062 * Mode Switches for gnatmake::
9063 * Notes on the Command Line::
9064 * How gnatmake Works::
9065 * Examples of gnatmake Usage::
9068 A typical development cycle when working on an Ada program consists of
9069 the following steps:
9073 Edit some sources to fix bugs.
9079 Compile all sources affected.
9089 The third step can be tricky, because not only do the modified files
9090 @cindex Dependency rules
9091 have to be compiled, but any files depending on these files must also be
9092 recompiled. The dependency rules in Ada can be quite complex, especially
9093 in the presence of overloading, @code{use} clauses, generics and inlined
9096 @command{gnatmake} automatically takes care of the third and fourth steps
9097 of this process. It determines which sources need to be compiled,
9098 compiles them, and binds and links the resulting object files.
9100 Unlike some other Ada make programs, the dependencies are always
9101 accurately recomputed from the new sources. The source based approach of
9102 the GNAT compilation model makes this possible. This means that if
9103 changes to the source program cause corresponding changes in
9104 dependencies, they will always be tracked exactly correctly by
9107 @node Running gnatmake
9108 @section Running @command{gnatmake}
9111 The usual form of the @command{gnatmake} command is
9114 @c $ gnatmake @ovar{switches} @var{file_name}
9115 @c @ovar{file_names} @ovar{mode_switches}
9116 @c Expanding @ovar macro inline (explanation in macro def comments)
9117 $ gnatmake @r{[}@var{switches}@r{]} @var{file_name}
9118 @r{[}@var{file_names}@r{]} @r{[}@var{mode_switches}@r{]}
9122 The only required argument is one @var{file_name}, which specifies
9123 a compilation unit that is a main program. Several @var{file_names} can be
9124 specified: this will result in several executables being built.
9125 If @code{switches} are present, they can be placed before the first
9126 @var{file_name}, between @var{file_names} or after the last @var{file_name}.
9127 If @var{mode_switches} are present, they must always be placed after
9128 the last @var{file_name} and all @code{switches}.
9130 If you are using standard file extensions (@file{.adb} and @file{.ads}), then the
9131 extension may be omitted from the @var{file_name} arguments. However, if
9132 you are using non-standard extensions, then it is required that the
9133 extension be given. A relative or absolute directory path can be
9134 specified in a @var{file_name}, in which case, the input source file will
9135 be searched for in the specified directory only. Otherwise, the input
9136 source file will first be searched in the directory where
9137 @command{gnatmake} was invoked and if it is not found, it will be search on
9138 the source path of the compiler as described in
9139 @ref{Search Paths and the Run-Time Library (RTL)}.
9141 All @command{gnatmake} output (except when you specify
9142 @option{^-M^/DEPENDENCIES_LIST^}) is to
9143 @file{stderr}. The output produced by the
9144 @option{^-M^/DEPENDENCIES_LIST^} switch is send to
9147 @node Switches for gnatmake
9148 @section Switches for @command{gnatmake}
9151 You may specify any of the following switches to @command{gnatmake}:
9157 @cindex @option{--version} @command{gnatmake}
9158 Display Copyright and version, then exit disregarding all other options.
9161 @cindex @option{--help} @command{gnatmake}
9162 If @option{--version} was not used, display usage, then exit disregarding
9166 @item --GCC=@var{compiler_name}
9167 @cindex @option{--GCC=compiler_name} (@command{gnatmake})
9168 Program used for compiling. The default is `@command{gcc}'. You need to use
9169 quotes around @var{compiler_name} if @code{compiler_name} contains
9170 spaces or other separator characters. As an example @option{--GCC="foo -x
9171 -y"} will instruct @command{gnatmake} to use @code{foo -x -y} as your
9172 compiler. A limitation of this syntax is that the name and path name of
9173 the executable itself must not include any embedded spaces. Note that
9174 switch @option{-c} is always inserted after your command name. Thus in the
9175 above example the compiler command that will be used by @command{gnatmake}
9176 will be @code{foo -c -x -y}. If several @option{--GCC=compiler_name} are
9177 used, only the last @var{compiler_name} is taken into account. However,
9178 all the additional switches are also taken into account. Thus,
9179 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
9180 @option{--GCC="bar -x -y -z -t"}.
9182 @item --GNATBIND=@var{binder_name}
9183 @cindex @option{--GNATBIND=binder_name} (@command{gnatmake})
9184 Program used for binding. The default is `@code{gnatbind}'. You need to
9185 use quotes around @var{binder_name} if @var{binder_name} contains spaces
9186 or other separator characters. As an example @option{--GNATBIND="bar -x
9187 -y"} will instruct @command{gnatmake} to use @code{bar -x -y} as your
9188 binder. Binder switches that are normally appended by @command{gnatmake}
9189 to `@code{gnatbind}' are now appended to the end of @code{bar -x -y}.
9190 A limitation of this syntax is that the name and path name of the executable
9191 itself must not include any embedded spaces.
9193 @item --GNATLINK=@var{linker_name}
9194 @cindex @option{--GNATLINK=linker_name} (@command{gnatmake})
9195 Program used for linking. The default is `@command{gnatlink}'. You need to
9196 use quotes around @var{linker_name} if @var{linker_name} contains spaces
9197 or other separator characters. As an example @option{--GNATLINK="lan -x
9198 -y"} will instruct @command{gnatmake} to use @code{lan -x -y} as your
9199 linker. Linker switches that are normally appended by @command{gnatmake} to
9200 `@command{gnatlink}' are now appended to the end of @code{lan -x -y}.
9201 A limitation of this syntax is that the name and path name of the executable
9202 itself must not include any embedded spaces.
9206 @item ^-a^/ALL_FILES^
9207 @cindex @option{^-a^/ALL_FILES^} (@command{gnatmake})
9208 Consider all files in the make process, even the GNAT internal system
9209 files (for example, the predefined Ada library files), as well as any
9210 locked files. Locked files are files whose ALI file is write-protected.
9212 @command{gnatmake} does not check these files,
9213 because the assumption is that the GNAT internal files are properly up
9214 to date, and also that any write protected ALI files have been properly
9215 installed. Note that if there is an installation problem, such that one
9216 of these files is not up to date, it will be properly caught by the
9218 You may have to specify this switch if you are working on GNAT
9219 itself. The switch @option{^-a^/ALL_FILES^} is also useful
9220 in conjunction with @option{^-f^/FORCE_COMPILE^}
9221 if you need to recompile an entire application,
9222 including run-time files, using special configuration pragmas,
9223 such as a @code{Normalize_Scalars} pragma.
9226 @code{gnatmake ^-a^/ALL_FILES^} compiles all GNAT
9229 @code{gcc -c -gnatpg} rather than @code{gcc -c}.
9232 the @code{/CHECKS=SUPPRESS_ALL /STYLE_CHECKS=GNAT} switch.
9235 @item ^-b^/ACTIONS=BIND^
9236 @cindex @option{^-b^/ACTIONS=BIND^} (@command{gnatmake})
9237 Bind only. Can be combined with @option{^-c^/ACTIONS=COMPILE^} to do
9238 compilation and binding, but no link.
9239 Can be combined with @option{^-l^/ACTIONS=LINK^}
9240 to do binding and linking. When not combined with
9241 @option{^-c^/ACTIONS=COMPILE^}
9242 all the units in the closure of the main program must have been previously
9243 compiled and must be up to date. The root unit specified by @var{file_name}
9244 may be given without extension, with the source extension or, if no GNAT
9245 Project File is specified, with the ALI file extension.
9247 @item ^-c^/ACTIONS=COMPILE^
9248 @cindex @option{^-c^/ACTIONS=COMPILE^} (@command{gnatmake})
9249 Compile only. Do not perform binding, except when @option{^-b^/ACTIONS=BIND^}
9250 is also specified. Do not perform linking, except if both
9251 @option{^-b^/ACTIONS=BIND^} and
9252 @option{^-l^/ACTIONS=LINK^} are also specified.
9253 If the root unit specified by @var{file_name} is not a main unit, this is the
9254 default. Otherwise @command{gnatmake} will attempt binding and linking
9255 unless all objects are up to date and the executable is more recent than
9259 @cindex @option{^-C^/MAPPING^} (@command{gnatmake})
9260 Use a temporary mapping file. A mapping file is a way to communicate
9261 to the compiler two mappings: from unit names to file names (without
9262 any directory information) and from file names to path names (with
9263 full directory information). A mapping file can make the compiler's
9264 file searches faster, especially if there are many source directories,
9265 or the sources are read over a slow network connection. If
9266 @option{^-P^/PROJECT_FILE^} is used, a mapping file is always used, so
9267 @option{^-C^/MAPPING^} is unnecessary; in this case the mapping file
9268 is initially populated based on the project file. If
9269 @option{^-C^/MAPPING^} is used without
9270 @option{^-P^/PROJECT_FILE^},
9271 the mapping file is initially empty. Each invocation of the compiler
9272 will add any newly accessed sources to the mapping file.
9274 @item ^-C=^/USE_MAPPING_FILE=^@var{file}
9275 @cindex @option{^-C=^/USE_MAPPING^} (@command{gnatmake})
9276 Use a specific mapping file. The file, specified as a path name (absolute or
9277 relative) by this switch, should already exist, otherwise the switch is
9278 ineffective. The specified mapping file will be communicated to the compiler.
9279 This switch is not compatible with a project file
9280 (^-P^/PROJECT_FILE=^@var{file}) or with multiple compiling processes
9281 (^-j^/PROCESSES=^nnn, when nnn is greater than 1).
9283 @item ^-d^/DISPLAY_PROGRESS^
9284 @cindex @option{^-d^/DISPLAY_PROGRESS^} (@command{gnatmake})
9285 Display progress for each source, up to date or not, as a single line
9288 completed x out of y (zz%)
9291 If the file needs to be compiled this is displayed after the invocation of
9292 the compiler. These lines are displayed even in quiet output mode.
9294 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
9295 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@command{gnatmake})
9296 Put all object files and ALI file in directory @var{dir}.
9297 If the @option{^-D^/DIRECTORY_OBJECTS^} switch is not used, all object files
9298 and ALI files go in the current working directory.
9300 This switch cannot be used when using a project file.
9304 @cindex @option{-eL} (@command{gnatmake})
9305 @cindex symbolic links
9306 Follow all symbolic links when processing project files.
9307 This should be used if your project uses symbolic links for files or
9308 directories, but is not needed in other cases.
9310 @cindex naming scheme
9311 This also assumes that no directory matches the naming scheme for files (for
9312 instance that you do not have a directory called "sources.ads" when using the
9313 default GNAT naming scheme).
9315 When you do not have to use this switch (ie by default), gnatmake is able to
9316 save a lot of system calls (several per source file and object file), which
9317 can result in a significant speed up to load and manipulate a project file,
9318 especially when using source files from a remote system.
9322 @item ^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^
9323 @cindex @option{^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^} (@command{gnatmake})
9324 Output the commands for the compiler, the binder and the linker
9325 on ^standard output^SYS$OUTPUT^,
9326 instead of ^standard error^SYS$ERROR^.
9328 @item ^-f^/FORCE_COMPILE^
9329 @cindex @option{^-f^/FORCE_COMPILE^} (@command{gnatmake})
9330 Force recompilations. Recompile all sources, even though some object
9331 files may be up to date, but don't recompile predefined or GNAT internal
9332 files or locked files (files with a write-protected ALI file),
9333 unless the @option{^-a^/ALL_FILES^} switch is also specified.
9335 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
9336 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@command{gnatmake})
9337 When using project files, if some errors or warnings are detected during
9338 parsing and verbose mode is not in effect (no use of switch
9339 ^-v^/VERBOSE^), then error lines start with the full path name of the project
9340 file, rather than its simple file name.
9343 @cindex @option{^-g^/DEBUG^} (@command{gnatmake})
9344 Enable debugging. This switch is simply passed to the compiler and to the
9347 @item ^-i^/IN_PLACE^
9348 @cindex @option{^-i^/IN_PLACE^} (@command{gnatmake})
9349 In normal mode, @command{gnatmake} compiles all object files and ALI files
9350 into the current directory. If the @option{^-i^/IN_PLACE^} switch is used,
9351 then instead object files and ALI files that already exist are overwritten
9352 in place. This means that once a large project is organized into separate
9353 directories in the desired manner, then @command{gnatmake} will automatically
9354 maintain and update this organization. If no ALI files are found on the
9355 Ada object path (@ref{Search Paths and the Run-Time Library (RTL)}),
9356 the new object and ALI files are created in the
9357 directory containing the source being compiled. If another organization
9358 is desired, where objects and sources are kept in different directories,
9359 a useful technique is to create dummy ALI files in the desired directories.
9360 When detecting such a dummy file, @command{gnatmake} will be forced to
9361 recompile the corresponding source file, and it will be put the resulting
9362 object and ALI files in the directory where it found the dummy file.
9364 @item ^-j^/PROCESSES=^@var{n}
9365 @cindex @option{^-j^/PROCESSES^} (@command{gnatmake})
9366 @cindex Parallel make
9367 Use @var{n} processes to carry out the (re)compilations. On a
9368 multiprocessor machine compilations will occur in parallel. In the
9369 event of compilation errors, messages from various compilations might
9370 get interspersed (but @command{gnatmake} will give you the full ordered
9371 list of failing compiles at the end). If this is problematic, rerun
9372 the make process with n set to 1 to get a clean list of messages.
9374 @item ^-k^/CONTINUE_ON_ERROR^
9375 @cindex @option{^-k^/CONTINUE_ON_ERROR^} (@command{gnatmake})
9376 Keep going. Continue as much as possible after a compilation error. To
9377 ease the programmer's task in case of compilation errors, the list of
9378 sources for which the compile fails is given when @command{gnatmake}
9381 If @command{gnatmake} is invoked with several @file{file_names} and with this
9382 switch, if there are compilation errors when building an executable,
9383 @command{gnatmake} will not attempt to build the following executables.
9385 @item ^-l^/ACTIONS=LINK^
9386 @cindex @option{^-l^/ACTIONS=LINK^} (@command{gnatmake})
9387 Link only. Can be combined with @option{^-b^/ACTIONS=BIND^} to binding
9388 and linking. Linking will not be performed if combined with
9389 @option{^-c^/ACTIONS=COMPILE^}
9390 but not with @option{^-b^/ACTIONS=BIND^}.
9391 When not combined with @option{^-b^/ACTIONS=BIND^}
9392 all the units in the closure of the main program must have been previously
9393 compiled and must be up to date, and the main program needs to have been bound.
9394 The root unit specified by @var{file_name}
9395 may be given without extension, with the source extension or, if no GNAT
9396 Project File is specified, with the ALI file extension.
9398 @item ^-m^/MINIMAL_RECOMPILATION^
9399 @cindex @option{^-m^/MINIMAL_RECOMPILATION^} (@command{gnatmake})
9400 Specify that the minimum necessary amount of recompilations
9401 be performed. In this mode @command{gnatmake} ignores time
9402 stamp differences when the only
9403 modifications to a source file consist in adding/removing comments,
9404 empty lines, spaces or tabs. This means that if you have changed the
9405 comments in a source file or have simply reformatted it, using this
9406 switch will tell @command{gnatmake} not to recompile files that depend on it
9407 (provided other sources on which these files depend have undergone no
9408 semantic modifications). Note that the debugging information may be
9409 out of date with respect to the sources if the @option{-m} switch causes
9410 a compilation to be switched, so the use of this switch represents a
9411 trade-off between compilation time and accurate debugging information.
9413 @item ^-M^/DEPENDENCIES_LIST^
9414 @cindex Dependencies, producing list
9415 @cindex @option{^-M^/DEPENDENCIES_LIST^} (@command{gnatmake})
9416 Check if all objects are up to date. If they are, output the object
9417 dependences to @file{stdout} in a form that can be directly exploited in
9418 a @file{Makefile}. By default, each source file is prefixed with its
9419 (relative or absolute) directory name. This name is whatever you
9420 specified in the various @option{^-aI^/SOURCE_SEARCH^}
9421 and @option{^-I^/SEARCH^} switches. If you use
9422 @code{gnatmake ^-M^/DEPENDENCIES_LIST^}
9423 @option{^-q^/QUIET^}
9424 (see below), only the source file names,
9425 without relative paths, are output. If you just specify the
9426 @option{^-M^/DEPENDENCIES_LIST^}
9427 switch, dependencies of the GNAT internal system files are omitted. This
9428 is typically what you want. If you also specify
9429 the @option{^-a^/ALL_FILES^} switch,
9430 dependencies of the GNAT internal files are also listed. Note that
9431 dependencies of the objects in external Ada libraries (see switch
9432 @option{^-aL^/SKIP_MISSING=^}@var{dir} in the following list)
9435 @item ^-n^/DO_OBJECT_CHECK^
9436 @cindex @option{^-n^/DO_OBJECT_CHECK^} (@command{gnatmake})
9437 Don't compile, bind, or link. Checks if all objects are up to date.
9438 If they are not, the full name of the first file that needs to be
9439 recompiled is printed.
9440 Repeated use of this option, followed by compiling the indicated source
9441 file, will eventually result in recompiling all required units.
9443 @item ^-o ^/EXECUTABLE=^@var{exec_name}
9444 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatmake})
9445 Output executable name. The name of the final executable program will be
9446 @var{exec_name}. If the @option{^-o^/EXECUTABLE^} switch is omitted the default
9447 name for the executable will be the name of the input file in appropriate form
9448 for an executable file on the host system.
9450 This switch cannot be used when invoking @command{gnatmake} with several
9453 @item ^-p or --create-missing-dirs^/CREATE_MISSING_DIRS^
9454 @cindex @option{^-p^/CREATE_MISSING_DIRS^} (@command{gnatmake})
9455 When using project files (^-P^/PROJECT_FILE=^@var{project}), create
9456 automatically missing object directories, library directories and exec
9459 @item ^-P^/PROJECT_FILE=^@var{project}
9460 @cindex @option{^-P^/PROJECT_FILE^} (@command{gnatmake})
9461 Use project file @var{project}. Only one such switch can be used.
9462 @xref{gnatmake and Project Files}.
9465 @cindex @option{^-q^/QUIET^} (@command{gnatmake})
9466 Quiet. When this flag is not set, the commands carried out by
9467 @command{gnatmake} are displayed.
9469 @item ^-s^/SWITCH_CHECK/^
9470 @cindex @option{^-s^/SWITCH_CHECK^} (@command{gnatmake})
9471 Recompile if compiler switches have changed since last compilation.
9472 All compiler switches but -I and -o are taken into account in the
9474 orders between different ``first letter'' switches are ignored, but
9475 orders between same switches are taken into account. For example,
9476 @option{-O -O2} is different than @option{-O2 -O}, but @option{-g -O}
9477 is equivalent to @option{-O -g}.
9479 This switch is recommended when Integrated Preprocessing is used.
9482 @cindex @option{^-u^/UNIQUE^} (@command{gnatmake})
9483 Unique. Recompile at most the main files. It implies -c. Combined with
9484 -f, it is equivalent to calling the compiler directly. Note that using
9485 ^-u^/UNIQUE^ with a project file and no main has a special meaning
9486 (@pxref{Project Files and Main Subprograms}).
9488 @item ^-U^/ALL_PROJECTS^
9489 @cindex @option{^-U^/ALL_PROJECTS^} (@command{gnatmake})
9490 When used without a project file or with one or several mains on the command
9491 line, is equivalent to ^-u^/UNIQUE^. When used with a project file and no main
9492 on the command line, all sources of all project files are checked and compiled
9493 if not up to date, and libraries are rebuilt, if necessary.
9496 @cindex @option{^-v^/REASONS^} (@command{gnatmake})
9497 Verbose. Display the reason for all recompilations @command{gnatmake}
9498 decides are necessary, with the highest verbosity level.
9500 @item ^-vl^/LOW_VERBOSITY^
9501 @cindex @option{^-vl^/LOW_VERBOSITY^} (@command{gnatmake})
9502 Verbosity level Low. Display fewer lines than in verbosity Medium.
9504 @item ^-vm^/MEDIUM_VERBOSITY^
9505 @cindex @option{^-vm^/MEDIUM_VERBOSITY^} (@command{gnatmake})
9506 Verbosity level Medium. Potentially display fewer lines than in verbosity High.
9508 @item ^-vh^/HIGH_VERBOSITY^
9509 @cindex @option{^-vm^/HIGH_VERBOSITY^} (@command{gnatmake})
9510 Verbosity level High. Equivalent to ^-v^/REASONS^.
9512 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
9513 Indicate the verbosity of the parsing of GNAT project files.
9514 @xref{Switches Related to Project Files}.
9516 @item ^-x^/NON_PROJECT_UNIT_COMPILATION^
9517 @cindex @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} (@command{gnatmake})
9518 Indicate that sources that are not part of any Project File may be compiled.
9519 Normally, when using Project Files, only sources that are part of a Project
9520 File may be compile. When this switch is used, a source outside of all Project
9521 Files may be compiled. The ALI file and the object file will be put in the
9522 object directory of the main Project. The compilation switches used will only
9523 be those specified on the command line. Even when
9524 @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} is used, mains specified on the
9525 command line need to be sources of a project file.
9527 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
9528 Indicate that external variable @var{name} has the value @var{value}.
9529 The Project Manager will use this value for occurrences of
9530 @code{external(name)} when parsing the project file.
9531 @xref{Switches Related to Project Files}.
9534 @cindex @option{^-z^/NOMAIN^} (@command{gnatmake})
9535 No main subprogram. Bind and link the program even if the unit name
9536 given on the command line is a package name. The resulting executable
9537 will execute the elaboration routines of the package and its closure,
9538 then the finalization routines.
9543 @item @command{gcc} @asis{switches}
9545 Any uppercase or multi-character switch that is not a @command{gnatmake} switch
9546 is passed to @command{gcc} (e.g.@: @option{-O}, @option{-gnato,} etc.)
9549 Any qualifier that cannot be recognized as a qualifier for @code{GNAT MAKE}
9550 but is recognizable as a valid qualifier for @code{GNAT COMPILE} is
9551 automatically treated as a compiler switch, and passed on to all
9552 compilations that are carried out.
9557 Source and library search path switches:
9561 @item ^-aI^/SOURCE_SEARCH=^@var{dir}
9562 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatmake})
9563 When looking for source files also look in directory @var{dir}.
9564 The order in which source files search is undertaken is
9565 described in @ref{Search Paths and the Run-Time Library (RTL)}.
9567 @item ^-aL^/SKIP_MISSING=^@var{dir}
9568 @cindex @option{^-aL^/SKIP_MISSING^} (@command{gnatmake})
9569 Consider @var{dir} as being an externally provided Ada library.
9570 Instructs @command{gnatmake} to skip compilation units whose @file{.ALI}
9571 files have been located in directory @var{dir}. This allows you to have
9572 missing bodies for the units in @var{dir} and to ignore out of date bodies
9573 for the same units. You still need to specify
9574 the location of the specs for these units by using the switches
9575 @option{^-aI^/SOURCE_SEARCH=^@var{dir}}
9576 or @option{^-I^/SEARCH=^@var{dir}}.
9577 Note: this switch is provided for compatibility with previous versions
9578 of @command{gnatmake}. The easier method of causing standard libraries
9579 to be excluded from consideration is to write-protect the corresponding
9582 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
9583 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatmake})
9584 When searching for library and object files, look in directory
9585 @var{dir}. The order in which library files are searched is described in
9586 @ref{Search Paths for gnatbind}.
9588 @item ^-A^/CONDITIONAL_SOURCE_SEARCH=^@var{dir}
9589 @cindex Search paths, for @command{gnatmake}
9590 @cindex @option{^-A^/CONDITIONAL_SOURCE_SEARCH^} (@command{gnatmake})
9591 Equivalent to @option{^-aL^/SKIP_MISSING=^@var{dir}
9592 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9594 @item ^-I^/SEARCH=^@var{dir}
9595 @cindex @option{^-I^/SEARCH^} (@command{gnatmake})
9596 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}
9597 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9599 @item ^-I-^/NOCURRENT_DIRECTORY^
9600 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatmake})
9601 @cindex Source files, suppressing search
9602 Do not look for source files in the directory containing the source
9603 file named in the command line.
9604 Do not look for ALI or object files in the directory
9605 where @command{gnatmake} was invoked.
9607 @item ^-L^/LIBRARY_SEARCH=^@var{dir}
9608 @cindex @option{^-L^/LIBRARY_SEARCH^} (@command{gnatmake})
9609 @cindex Linker libraries
9610 Add directory @var{dir} to the list of directories in which the linker
9611 will search for libraries. This is equivalent to
9612 @option{-largs ^-L^/LIBRARY_SEARCH=^}@var{dir}.
9614 Furthermore, under Windows, the sources pointed to by the libraries path
9615 set in the registry are not searched for.
9619 @cindex @option{-nostdinc} (@command{gnatmake})
9620 Do not look for source files in the system default directory.
9623 @cindex @option{-nostdlib} (@command{gnatmake})
9624 Do not look for library files in the system default directory.
9626 @item --RTS=@var{rts-path}
9627 @cindex @option{--RTS} (@command{gnatmake})
9628 Specifies the default location of the runtime library. GNAT looks for the
9630 in the following directories, and stops as soon as a valid runtime is found
9631 (@file{adainclude} or @file{ada_source_path}, and @file{adalib} or
9632 @file{ada_object_path} present):
9635 @item <current directory>/$rts_path
9637 @item <default-search-dir>/$rts_path
9639 @item <default-search-dir>/rts-$rts_path
9643 The selected path is handled like a normal RTS path.
9647 @node Mode Switches for gnatmake
9648 @section Mode Switches for @command{gnatmake}
9651 The mode switches (referred to as @code{mode_switches}) allow the
9652 inclusion of switches that are to be passed to the compiler itself, the
9653 binder or the linker. The effect of a mode switch is to cause all
9654 subsequent switches up to the end of the switch list, or up to the next
9655 mode switch, to be interpreted as switches to be passed on to the
9656 designated component of GNAT.
9660 @item -cargs @var{switches}
9661 @cindex @option{-cargs} (@command{gnatmake})
9662 Compiler switches. Here @var{switches} is a list of switches
9663 that are valid switches for @command{gcc}. They will be passed on to
9664 all compile steps performed by @command{gnatmake}.
9666 @item -bargs @var{switches}
9667 @cindex @option{-bargs} (@command{gnatmake})
9668 Binder switches. Here @var{switches} is a list of switches
9669 that are valid switches for @code{gnatbind}. They will be passed on to
9670 all bind steps performed by @command{gnatmake}.
9672 @item -largs @var{switches}
9673 @cindex @option{-largs} (@command{gnatmake})
9674 Linker switches. Here @var{switches} is a list of switches
9675 that are valid switches for @command{gnatlink}. They will be passed on to
9676 all link steps performed by @command{gnatmake}.
9678 @item -margs @var{switches}
9679 @cindex @option{-margs} (@command{gnatmake})
9680 Make switches. The switches are directly interpreted by @command{gnatmake},
9681 regardless of any previous occurrence of @option{-cargs}, @option{-bargs}
9685 @node Notes on the Command Line
9686 @section Notes on the Command Line
9689 This section contains some additional useful notes on the operation
9690 of the @command{gnatmake} command.
9694 @cindex Recompilation, by @command{gnatmake}
9695 If @command{gnatmake} finds no ALI files, it recompiles the main program
9696 and all other units required by the main program.
9697 This means that @command{gnatmake}
9698 can be used for the initial compile, as well as during subsequent steps of
9699 the development cycle.
9702 If you enter @code{gnatmake @var{file}.adb}, where @file{@var{file}.adb}
9703 is a subunit or body of a generic unit, @command{gnatmake} recompiles
9704 @file{@var{file}.adb} (because it finds no ALI) and stops, issuing a
9708 In @command{gnatmake} the switch @option{^-I^/SEARCH^}
9709 is used to specify both source and
9710 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
9711 instead if you just want to specify
9712 source paths only and @option{^-aO^/OBJECT_SEARCH^}
9713 if you want to specify library paths
9717 @command{gnatmake} will ignore any files whose ALI file is write-protected.
9718 This may conveniently be used to exclude standard libraries from
9719 consideration and in particular it means that the use of the
9720 @option{^-f^/FORCE_COMPILE^} switch will not recompile these files
9721 unless @option{^-a^/ALL_FILES^} is also specified.
9724 @command{gnatmake} has been designed to make the use of Ada libraries
9725 particularly convenient. Assume you have an Ada library organized
9726 as follows: @i{^obj-dir^[OBJ_DIR]^} contains the objects and ALI files for
9727 of your Ada compilation units,
9728 whereas @i{^include-dir^[INCLUDE_DIR]^} contains the
9729 specs of these units, but no bodies. Then to compile a unit
9730 stored in @code{main.adb}, which uses this Ada library you would just type
9734 $ gnatmake -aI@var{include-dir} -aL@var{obj-dir} main
9737 $ gnatmake /SOURCE_SEARCH=@i{[INCLUDE_DIR]}
9738 /SKIP_MISSING=@i{[OBJ_DIR]} main
9743 Using @command{gnatmake} along with the
9744 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}
9745 switch provides a mechanism for avoiding unnecessary recompilations. Using
9747 you can update the comments/format of your
9748 source files without having to recompile everything. Note, however, that
9749 adding or deleting lines in a source files may render its debugging
9750 info obsolete. If the file in question is a spec, the impact is rather
9751 limited, as that debugging info will only be useful during the
9752 elaboration phase of your program. For bodies the impact can be more
9753 significant. In all events, your debugger will warn you if a source file
9754 is more recent than the corresponding object, and alert you to the fact
9755 that the debugging information may be out of date.
9758 @node How gnatmake Works
9759 @section How @command{gnatmake} Works
9762 Generally @command{gnatmake} automatically performs all necessary
9763 recompilations and you don't need to worry about how it works. However,
9764 it may be useful to have some basic understanding of the @command{gnatmake}
9765 approach and in particular to understand how it uses the results of
9766 previous compilations without incorrectly depending on them.
9768 First a definition: an object file is considered @dfn{up to date} if the
9769 corresponding ALI file exists and if all the source files listed in the
9770 dependency section of this ALI file have time stamps matching those in
9771 the ALI file. This means that neither the source file itself nor any
9772 files that it depends on have been modified, and hence there is no need
9773 to recompile this file.
9775 @command{gnatmake} works by first checking if the specified main unit is up
9776 to date. If so, no compilations are required for the main unit. If not,
9777 @command{gnatmake} compiles the main program to build a new ALI file that
9778 reflects the latest sources. Then the ALI file of the main unit is
9779 examined to find all the source files on which the main program depends,
9780 and @command{gnatmake} recursively applies the above procedure on all these
9783 This process ensures that @command{gnatmake} only trusts the dependencies
9784 in an existing ALI file if they are known to be correct. Otherwise it
9785 always recompiles to determine a new, guaranteed accurate set of
9786 dependencies. As a result the program is compiled ``upside down'' from what may
9787 be more familiar as the required order of compilation in some other Ada
9788 systems. In particular, clients are compiled before the units on which
9789 they depend. The ability of GNAT to compile in any order is critical in
9790 allowing an order of compilation to be chosen that guarantees that
9791 @command{gnatmake} will recompute a correct set of new dependencies if
9794 When invoking @command{gnatmake} with several @var{file_names}, if a unit is
9795 imported by several of the executables, it will be recompiled at most once.
9797 Note: when using non-standard naming conventions
9798 (@pxref{Using Other File Names}), changing through a configuration pragmas
9799 file the version of a source and invoking @command{gnatmake} to recompile may
9800 have no effect, if the previous version of the source is still accessible
9801 by @command{gnatmake}. It may be necessary to use the switch
9802 ^-f^/FORCE_COMPILE^.
9804 @node Examples of gnatmake Usage
9805 @section Examples of @command{gnatmake} Usage
9808 @item gnatmake hello.adb
9809 Compile all files necessary to bind and link the main program
9810 @file{hello.adb} (containing unit @code{Hello}) and bind and link the
9811 resulting object files to generate an executable file @file{^hello^HELLO.EXE^}.
9813 @item gnatmake main1 main2 main3
9814 Compile all files necessary to bind and link the main programs
9815 @file{main1.adb} (containing unit @code{Main1}), @file{main2.adb}
9816 (containing unit @code{Main2}) and @file{main3.adb}
9817 (containing unit @code{Main3}) and bind and link the resulting object files
9818 to generate three executable files @file{^main1^MAIN1.EXE^},
9819 @file{^main2^MAIN2.EXE^}
9820 and @file{^main3^MAIN3.EXE^}.
9823 @item gnatmake -q Main_Unit -cargs -O2 -bargs -l
9827 @item gnatmake Main_Unit /QUIET
9828 /COMPILER_QUALIFIERS /OPTIMIZE=ALL
9829 /BINDER_QUALIFIERS /ORDER_OF_ELABORATION
9831 Compile all files necessary to bind and link the main program unit
9832 @code{Main_Unit} (from file @file{main_unit.adb}). All compilations will
9833 be done with optimization level 2 and the order of elaboration will be
9834 listed by the binder. @command{gnatmake} will operate in quiet mode, not
9835 displaying commands it is executing.
9838 @c *************************
9839 @node Improving Performance
9840 @chapter Improving Performance
9841 @cindex Improving performance
9844 This chapter presents several topics related to program performance.
9845 It first describes some of the tradeoffs that need to be considered
9846 and some of the techniques for making your program run faster.
9847 It then documents the @command{gnatelim} tool and unused subprogram/data
9848 elimination feature, which can reduce the size of program executables.
9850 Note: to invoke @command{gnatelim} with a project file, use the @code{gnat}
9851 driver (see @ref{The GNAT Driver and Project Files}).
9855 * Performance Considerations::
9856 * Text_IO Suggestions::
9857 * Reducing Size of Ada Executables with gnatelim::
9858 * Reducing Size of Executables with unused subprogram/data elimination::
9862 @c *****************************
9863 @node Performance Considerations
9864 @section Performance Considerations
9867 The GNAT system provides a number of options that allow a trade-off
9872 performance of the generated code
9875 speed of compilation
9878 minimization of dependences and recompilation
9881 the degree of run-time checking.
9885 The defaults (if no options are selected) aim at improving the speed
9886 of compilation and minimizing dependences, at the expense of performance
9887 of the generated code:
9894 no inlining of subprogram calls
9897 all run-time checks enabled except overflow and elaboration checks
9901 These options are suitable for most program development purposes. This
9902 chapter describes how you can modify these choices, and also provides
9903 some guidelines on debugging optimized code.
9906 * Controlling Run-Time Checks::
9907 * Use of Restrictions::
9908 * Optimization Levels::
9909 * Debugging Optimized Code::
9910 * Inlining of Subprograms::
9911 * Other Optimization Switches::
9912 * Optimization and Strict Aliasing::
9915 * Coverage Analysis::
9919 @node Controlling Run-Time Checks
9920 @subsection Controlling Run-Time Checks
9923 By default, GNAT generates all run-time checks, except integer overflow
9924 checks, stack overflow checks, and checks for access before elaboration on
9925 subprogram calls. The latter are not required in default mode, because all
9926 necessary checking is done at compile time.
9927 @cindex @option{-gnatp} (@command{gcc})
9928 @cindex @option{-gnato} (@command{gcc})
9929 Two gnat switches, @option{-gnatp} and @option{-gnato} allow this default to
9930 be modified. @xref{Run-Time Checks}.
9932 Our experience is that the default is suitable for most development
9935 We treat integer overflow specially because these
9936 are quite expensive and in our experience are not as important as other
9937 run-time checks in the development process. Note that division by zero
9938 is not considered an overflow check, and divide by zero checks are
9939 generated where required by default.
9941 Elaboration checks are off by default, and also not needed by default, since
9942 GNAT uses a static elaboration analysis approach that avoids the need for
9943 run-time checking. This manual contains a full chapter discussing the issue
9944 of elaboration checks, and if the default is not satisfactory for your use,
9945 you should read this chapter.
9947 For validity checks, the minimal checks required by the Ada Reference
9948 Manual (for case statements and assignments to array elements) are on
9949 by default. These can be suppressed by use of the @option{-gnatVn} switch.
9950 Note that in Ada 83, there were no validity checks, so if the Ada 83 mode
9951 is acceptable (or when comparing GNAT performance with an Ada 83 compiler),
9952 it may be reasonable to routinely use @option{-gnatVn}. Validity checks
9953 are also suppressed entirely if @option{-gnatp} is used.
9955 @cindex Overflow checks
9956 @cindex Checks, overflow
9959 @cindex pragma Suppress
9960 @cindex pragma Unsuppress
9961 Note that the setting of the switches controls the default setting of
9962 the checks. They may be modified using either @code{pragma Suppress} (to
9963 remove checks) or @code{pragma Unsuppress} (to add back suppressed
9964 checks) in the program source.
9966 @node Use of Restrictions
9967 @subsection Use of Restrictions
9970 The use of pragma Restrictions allows you to control which features are
9971 permitted in your program. Apart from the obvious point that if you avoid
9972 relatively expensive features like finalization (enforceable by the use
9973 of pragma Restrictions (No_Finalization), the use of this pragma does not
9974 affect the generated code in most cases.
9976 One notable exception to this rule is that the possibility of task abort
9977 results in some distributed overhead, particularly if finalization or
9978 exception handlers are used. The reason is that certain sections of code
9979 have to be marked as non-abortable.
9981 If you use neither the @code{abort} statement, nor asynchronous transfer
9982 of control (@code{select @dots{} then abort}), then this distributed overhead
9983 is removed, which may have a general positive effect in improving
9984 overall performance. Especially code involving frequent use of tasking
9985 constructs and controlled types will show much improved performance.
9986 The relevant restrictions pragmas are
9988 @smallexample @c ada
9989 pragma Restrictions (No_Abort_Statements);
9990 pragma Restrictions (Max_Asynchronous_Select_Nesting => 0);
9994 It is recommended that these restriction pragmas be used if possible. Note
9995 that this also means that you can write code without worrying about the
9996 possibility of an immediate abort at any point.
9998 @node Optimization Levels
9999 @subsection Optimization Levels
10000 @cindex @option{^-O^/OPTIMIZE^} (@command{gcc})
10003 Without any optimization ^option,^qualifier,^
10004 the compiler's goal is to reduce the cost of
10005 compilation and to make debugging produce the expected results.
10006 Statements are independent: if you stop the program with a breakpoint between
10007 statements, you can then assign a new value to any variable or change
10008 the program counter to any other statement in the subprogram and get exactly
10009 the results you would expect from the source code.
10011 Turning on optimization makes the compiler attempt to improve the
10012 performance and/or code size at the expense of compilation time and
10013 possibly the ability to debug the program.
10015 If you use multiple
10016 ^-O options, with or without level numbers,^/OPTIMIZE qualifiers,^
10017 the last such option is the one that is effective.
10020 The default is optimization off. This results in the fastest compile
10021 times, but GNAT makes absolutely no attempt to optimize, and the
10022 generated programs are considerably larger and slower than when
10023 optimization is enabled. You can use the
10025 @option{-O} switch (the permitted forms are @option{-O0}, @option{-O1}
10026 @option{-O2}, @option{-O3}, and @option{-Os})
10029 @code{OPTIMIZE} qualifier
10031 to @command{gcc} to control the optimization level:
10034 @item ^-O0^/OPTIMIZE=NONE^
10035 No optimization (the default);
10036 generates unoptimized code but has
10037 the fastest compilation time.
10039 Note that many other compilers do fairly extensive optimization
10040 even if ``no optimization'' is specified. With gcc, it is
10041 very unusual to use ^-O0^/OPTIMIZE=NONE^ for production if
10042 execution time is of any concern, since ^-O0^/OPTIMIZE=NONE^
10043 really does mean no optimization at all. This difference between
10044 gcc and other compilers should be kept in mind when doing
10045 performance comparisons.
10047 @item ^-O1^/OPTIMIZE=SOME^
10048 Moderate optimization;
10049 optimizes reasonably well but does not
10050 degrade compilation time significantly.
10052 @item ^-O2^/OPTIMIZE=ALL^
10054 @itemx /OPTIMIZE=DEVELOPMENT
10057 generates highly optimized code and has
10058 the slowest compilation time.
10060 @item ^-O3^/OPTIMIZE=INLINING^
10061 Full optimization as in @option{-O2},
10062 and also attempts automatic inlining of small
10063 subprograms within a unit (@pxref{Inlining of Subprograms}).
10065 @item ^-Os^/OPTIMIZE=SPACE^
10066 Optimize space usage of resulting program.
10070 Higher optimization levels perform more global transformations on the
10071 program and apply more expensive analysis algorithms in order to generate
10072 faster and more compact code. The price in compilation time, and the
10073 resulting improvement in execution time,
10074 both depend on the particular application and the hardware environment.
10075 You should experiment to find the best level for your application.
10077 Since the precise set of optimizations done at each level will vary from
10078 release to release (and sometime from target to target), it is best to think
10079 of the optimization settings in general terms.
10080 @xref{Optimize Options,, Options That Control Optimization, gcc, Using
10081 the GNU Compiler Collection (GCC)}, for details about
10082 ^the @option{-O} settings and a number of @option{-f} options that^how to^
10083 individually enable or disable specific optimizations.
10085 Unlike some other compilation systems, ^@command{gcc}^GNAT^ has
10086 been tested extensively at all optimization levels. There are some bugs
10087 which appear only with optimization turned on, but there have also been
10088 bugs which show up only in @emph{unoptimized} code. Selecting a lower
10089 level of optimization does not improve the reliability of the code
10090 generator, which in practice is highly reliable at all optimization
10093 Note regarding the use of @option{-O3}: The use of this optimization level
10094 is generally discouraged with GNAT, since it often results in larger
10095 executables which run more slowly. See further discussion of this point
10096 in @ref{Inlining of Subprograms}.
10098 @node Debugging Optimized Code
10099 @subsection Debugging Optimized Code
10100 @cindex Debugging optimized code
10101 @cindex Optimization and debugging
10104 Although it is possible to do a reasonable amount of debugging at
10106 nonzero optimization levels,
10107 the higher the level the more likely that
10110 @option{/OPTIMIZE} settings other than @code{NONE},
10111 such settings will make it more likely that
10113 source-level constructs will have been eliminated by optimization.
10114 For example, if a loop is strength-reduced, the loop
10115 control variable may be completely eliminated and thus cannot be
10116 displayed in the debugger.
10117 This can only happen at @option{-O2} or @option{-O3}.
10118 Explicit temporary variables that you code might be eliminated at
10119 ^level^setting^ @option{-O1} or higher.
10121 The use of the @option{^-g^/DEBUG^} switch,
10122 @cindex @option{^-g^/DEBUG^} (@command{gcc})
10123 which is needed for source-level debugging,
10124 affects the size of the program executable on disk,
10125 and indeed the debugging information can be quite large.
10126 However, it has no effect on the generated code (and thus does not
10127 degrade performance)
10129 Since the compiler generates debugging tables for a compilation unit before
10130 it performs optimizations, the optimizing transformations may invalidate some
10131 of the debugging data. You therefore need to anticipate certain
10132 anomalous situations that may arise while debugging optimized code.
10133 These are the most common cases:
10137 @i{The ``hopping Program Counter'':} Repeated @code{step} or @code{next}
10139 the PC bouncing back and forth in the code. This may result from any of
10140 the following optimizations:
10144 @i{Common subexpression elimination:} using a single instance of code for a
10145 quantity that the source computes several times. As a result you
10146 may not be able to stop on what looks like a statement.
10149 @i{Invariant code motion:} moving an expression that does not change within a
10150 loop, to the beginning of the loop.
10153 @i{Instruction scheduling:} moving instructions so as to
10154 overlap loads and stores (typically) with other code, or in
10155 general to move computations of values closer to their uses. Often
10156 this causes you to pass an assignment statement without the assignment
10157 happening and then later bounce back to the statement when the
10158 value is actually needed. Placing a breakpoint on a line of code
10159 and then stepping over it may, therefore, not always cause all the
10160 expected side-effects.
10164 @i{The ``big leap'':} More commonly known as @emph{cross-jumping}, in which
10165 two identical pieces of code are merged and the program counter suddenly
10166 jumps to a statement that is not supposed to be executed, simply because
10167 it (and the code following) translates to the same thing as the code
10168 that @emph{was} supposed to be executed. This effect is typically seen in
10169 sequences that end in a jump, such as a @code{goto}, a @code{return}, or
10170 a @code{break} in a C @code{^switch^switch^} statement.
10173 @i{The ``roving variable'':} The symptom is an unexpected value in a variable.
10174 There are various reasons for this effect:
10178 In a subprogram prologue, a parameter may not yet have been moved to its
10182 A variable may be dead, and its register re-used. This is
10183 probably the most common cause.
10186 As mentioned above, the assignment of a value to a variable may
10190 A variable may be eliminated entirely by value propagation or
10191 other means. In this case, GCC may incorrectly generate debugging
10192 information for the variable
10196 In general, when an unexpected value appears for a local variable or parameter
10197 you should first ascertain if that value was actually computed by
10198 your program, as opposed to being incorrectly reported by the debugger.
10200 array elements in an object designated by an access value
10201 are generally less of a problem, once you have ascertained that the access
10203 Typically, this means checking variables in the preceding code and in the
10204 calling subprogram to verify that the value observed is explainable from other
10205 values (one must apply the procedure recursively to those
10206 other values); or re-running the code and stopping a little earlier
10207 (perhaps before the call) and stepping to better see how the variable obtained
10208 the value in question; or continuing to step @emph{from} the point of the
10209 strange value to see if code motion had simply moved the variable's
10214 In light of such anomalies, a recommended technique is to use @option{-O0}
10215 early in the software development cycle, when extensive debugging capabilities
10216 are most needed, and then move to @option{-O1} and later @option{-O2} as
10217 the debugger becomes less critical.
10218 Whether to use the @option{^-g^/DEBUG^} switch in the release version is
10219 a release management issue.
10221 Note that if you use @option{-g} you can then use the @command{strip} program
10222 on the resulting executable,
10223 which removes both debugging information and global symbols.
10226 @node Inlining of Subprograms
10227 @subsection Inlining of Subprograms
10230 A call to a subprogram in the current unit is inlined if all the
10231 following conditions are met:
10235 The optimization level is at least @option{-O1}.
10238 The called subprogram is suitable for inlining: It must be small enough
10239 and not contain something that @command{gcc} cannot support in inlined
10243 @cindex pragma Inline
10245 Either @code{pragma Inline} applies to the subprogram, or it is local
10246 to the unit and called once from within it, or it is small and automatic
10247 inlining (optimization level @option{-O3}) is specified.
10251 Calls to subprograms in @code{with}'ed units are normally not inlined.
10252 To achieve actual inlining (that is, replacement of the call by the code
10253 in the body of the subprogram), the following conditions must all be true.
10257 The optimization level is at least @option{-O1}.
10260 The called subprogram is suitable for inlining: It must be small enough
10261 and not contain something that @command{gcc} cannot support in inlined
10265 The call appears in a body (not in a package spec).
10268 There is a @code{pragma Inline} for the subprogram.
10271 @cindex @option{-gnatn} (@command{gcc})
10272 The @option{^-gnatn^/INLINE^} switch
10273 is used in the @command{gcc} command line
10276 Even if all these conditions are met, it may not be possible for
10277 the compiler to inline the call, due to the length of the body,
10278 or features in the body that make it impossible for the compiler
10279 to do the inlining.
10281 Note that specifying the @option{-gnatn} switch causes additional
10282 compilation dependencies. Consider the following:
10284 @smallexample @c ada
10304 With the default behavior (no @option{-gnatn} switch specified), the
10305 compilation of the @code{Main} procedure depends only on its own source,
10306 @file{main.adb}, and the spec of the package in file @file{r.ads}. This
10307 means that editing the body of @code{R} does not require recompiling
10310 On the other hand, the call @code{R.Q} is not inlined under these
10311 circumstances. If the @option{-gnatn} switch is present when @code{Main}
10312 is compiled, the call will be inlined if the body of @code{Q} is small
10313 enough, but now @code{Main} depends on the body of @code{R} in
10314 @file{r.adb} as well as on the spec. This means that if this body is edited,
10315 the main program must be recompiled. Note that this extra dependency
10316 occurs whether or not the call is in fact inlined by @command{gcc}.
10318 The use of front end inlining with @option{-gnatN} generates similar
10319 additional dependencies.
10321 @cindex @option{^-fno-inline^/INLINE=SUPPRESS^} (@command{gcc})
10322 Note: The @option{^-fno-inline^/INLINE=SUPPRESS^} switch
10323 can be used to prevent
10324 all inlining. This switch overrides all other conditions and ensures
10325 that no inlining occurs. The extra dependences resulting from
10326 @option{-gnatn} will still be active, even if
10327 this switch is used to suppress the resulting inlining actions.
10329 @cindex @option{-fno-inline-functions} (@command{gcc})
10330 Note: The @option{-fno-inline-functions} switch can be used to prevent
10331 automatic inlining of small subprograms if @option{-O3} is used.
10333 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
10334 Note: The @option{-fno-inline-functions-called-once} switch
10335 can be used to prevent inlining of subprograms local to the unit
10336 and called once from within it if @option{-O1} is used.
10338 Note regarding the use of @option{-O3}: There is no difference in inlining
10339 behavior between @option{-O2} and @option{-O3} for subprograms with an explicit
10340 pragma @code{Inline} assuming the use of @option{-gnatn}
10341 or @option{-gnatN} (the switches that activate inlining). If you have used
10342 pragma @code{Inline} in appropriate cases, then it is usually much better
10343 to use @option{-O2} and @option{-gnatn} and avoid the use of @option{-O3} which
10344 in this case only has the effect of inlining subprograms you did not
10345 think should be inlined. We often find that the use of @option{-O3} slows
10346 down code by performing excessive inlining, leading to increased instruction
10347 cache pressure from the increased code size. So the bottom line here is
10348 that you should not automatically assume that @option{-O3} is better than
10349 @option{-O2}, and indeed you should use @option{-O3} only if tests show that
10350 it actually improves performance.
10352 @node Other Optimization Switches
10353 @subsection Other Optimization Switches
10354 @cindex Optimization Switches
10356 Since @code{GNAT} uses the @command{gcc} back end, all the specialized
10357 @command{gcc} optimization switches are potentially usable. These switches
10358 have not been extensively tested with GNAT but can generally be expected
10359 to work. Examples of switches in this category are
10360 @option{-funroll-loops} and
10361 the various target-specific @option{-m} options (in particular, it has been
10362 observed that @option{-march=pentium4} can significantly improve performance
10363 on appropriate machines). For full details of these switches, see
10364 @ref{Submodel Options,, Hardware Models and Configurations, gcc, Using
10365 the GNU Compiler Collection (GCC)}.
10367 @node Optimization and Strict Aliasing
10368 @subsection Optimization and Strict Aliasing
10370 @cindex Strict Aliasing
10371 @cindex No_Strict_Aliasing
10374 The strong typing capabilities of Ada allow an optimizer to generate
10375 efficient code in situations where other languages would be forced to
10376 make worst case assumptions preventing such optimizations. Consider
10377 the following example:
10379 @smallexample @c ada
10382 type Int1 is new Integer;
10383 type Int2 is new Integer;
10384 type Int1A is access Int1;
10385 type Int2A is access Int2;
10392 for J in Data'Range loop
10393 if Data (J) = Int1V.all then
10394 Int2V.all := Int2V.all + 1;
10403 In this example, since the variable @code{Int1V} can only access objects
10404 of type @code{Int1}, and @code{Int2V} can only access objects of type
10405 @code{Int2}, there is no possibility that the assignment to
10406 @code{Int2V.all} affects the value of @code{Int1V.all}. This means that
10407 the compiler optimizer can "know" that the value @code{Int1V.all} is constant
10408 for all iterations of the loop and avoid the extra memory reference
10409 required to dereference it each time through the loop.
10411 This kind of optimization, called strict aliasing analysis, is
10412 triggered by specifying an optimization level of @option{-O2} or
10413 higher or @option{-Os} and allows @code{GNAT} to generate more efficient code
10414 when access values are involved.
10416 However, although this optimization is always correct in terms of
10417 the formal semantics of the Ada Reference Manual, difficulties can
10418 arise if features like @code{Unchecked_Conversion} are used to break
10419 the typing system. Consider the following complete program example:
10421 @smallexample @c ada
10424 type int1 is new integer;
10425 type int2 is new integer;
10426 type a1 is access int1;
10427 type a2 is access int2;
10432 function to_a2 (Input : a1) return a2;
10435 with Unchecked_Conversion;
10437 function to_a2 (Input : a1) return a2 is
10439 new Unchecked_Conversion (a1, a2);
10441 return to_a2u (Input);
10447 with Text_IO; use Text_IO;
10449 v1 : a1 := new int1;
10450 v2 : a2 := to_a2 (v1);
10454 put_line (int1'image (v1.all));
10460 This program prints out 0 in @option{-O0} or @option{-O1}
10461 mode, but it prints out 1 in @option{-O2} mode. That's
10462 because in strict aliasing mode, the compiler can and
10463 does assume that the assignment to @code{v2.all} could not
10464 affect the value of @code{v1.all}, since different types
10467 This behavior is not a case of non-conformance with the standard, since
10468 the Ada RM specifies that an unchecked conversion where the resulting
10469 bit pattern is not a correct value of the target type can result in an
10470 abnormal value and attempting to reference an abnormal value makes the
10471 execution of a program erroneous. That's the case here since the result
10472 does not point to an object of type @code{int2}. This means that the
10473 effect is entirely unpredictable.
10475 However, although that explanation may satisfy a language
10476 lawyer, in practice an applications programmer expects an
10477 unchecked conversion involving pointers to create true
10478 aliases and the behavior of printing 1 seems plain wrong.
10479 In this case, the strict aliasing optimization is unwelcome.
10481 Indeed the compiler recognizes this possibility, and the
10482 unchecked conversion generates a warning:
10485 p2.adb:5:07: warning: possible aliasing problem with type "a2"
10486 p2.adb:5:07: warning: use -fno-strict-aliasing switch for references
10487 p2.adb:5:07: warning: or use "pragma No_Strict_Aliasing (a2);"
10491 Unfortunately the problem is recognized when compiling the body of
10492 package @code{p2}, but the actual "bad" code is generated while
10493 compiling the body of @code{m} and this latter compilation does not see
10494 the suspicious @code{Unchecked_Conversion}.
10496 As implied by the warning message, there are approaches you can use to
10497 avoid the unwanted strict aliasing optimization in a case like this.
10499 One possibility is to simply avoid the use of @option{-O2}, but
10500 that is a bit drastic, since it throws away a number of useful
10501 optimizations that do not involve strict aliasing assumptions.
10503 A less drastic approach is to compile the program using the
10504 option @option{-fno-strict-aliasing}. Actually it is only the
10505 unit containing the dereferencing of the suspicious pointer
10506 that needs to be compiled. So in this case, if we compile
10507 unit @code{m} with this switch, then we get the expected
10508 value of zero printed. Analyzing which units might need
10509 the switch can be painful, so a more reasonable approach
10510 is to compile the entire program with options @option{-O2}
10511 and @option{-fno-strict-aliasing}. If the performance is
10512 satisfactory with this combination of options, then the
10513 advantage is that the entire issue of possible "wrong"
10514 optimization due to strict aliasing is avoided.
10516 To avoid the use of compiler switches, the configuration
10517 pragma @code{No_Strict_Aliasing} with no parameters may be
10518 used to specify that for all access types, the strict
10519 aliasing optimization should be suppressed.
10521 However, these approaches are still overkill, in that they causes
10522 all manipulations of all access values to be deoptimized. A more
10523 refined approach is to concentrate attention on the specific
10524 access type identified as problematic.
10526 First, if a careful analysis of uses of the pointer shows
10527 that there are no possible problematic references, then
10528 the warning can be suppressed by bracketing the
10529 instantiation of @code{Unchecked_Conversion} to turn
10532 @smallexample @c ada
10533 pragma Warnings (Off);
10535 new Unchecked_Conversion (a1, a2);
10536 pragma Warnings (On);
10540 Of course that approach is not appropriate for this particular
10541 example, since indeed there is a problematic reference. In this
10542 case we can take one of two other approaches.
10544 The first possibility is to move the instantiation of unchecked
10545 conversion to the unit in which the type is declared. In
10546 this example, we would move the instantiation of
10547 @code{Unchecked_Conversion} from the body of package
10548 @code{p2} to the spec of package @code{p1}. Now the
10549 warning disappears. That's because any use of the
10550 access type knows there is a suspicious unchecked
10551 conversion, and the strict aliasing optimization
10552 is automatically suppressed for the type.
10554 If it is not practical to move the unchecked conversion to the same unit
10555 in which the destination access type is declared (perhaps because the
10556 source type is not visible in that unit), you may use pragma
10557 @code{No_Strict_Aliasing} for the type. This pragma must occur in the
10558 same declarative sequence as the declaration of the access type:
10560 @smallexample @c ada
10561 type a2 is access int2;
10562 pragma No_Strict_Aliasing (a2);
10566 Here again, the compiler now knows that the strict aliasing optimization
10567 should be suppressed for any reference to type @code{a2} and the
10568 expected behavior is obtained.
10570 Finally, note that although the compiler can generate warnings for
10571 simple cases of unchecked conversions, there are tricker and more
10572 indirect ways of creating type incorrect aliases which the compiler
10573 cannot detect. Examples are the use of address overlays and unchecked
10574 conversions involving composite types containing access types as
10575 components. In such cases, no warnings are generated, but there can
10576 still be aliasing problems. One safe coding practice is to forbid the
10577 use of address clauses for type overlaying, and to allow unchecked
10578 conversion only for primitive types. This is not really a significant
10579 restriction since any possible desired effect can be achieved by
10580 unchecked conversion of access values.
10582 The aliasing analysis done in strict aliasing mode can certainly
10583 have significant benefits. We have seen cases of large scale
10584 application code where the time is increased by up to 5% by turning
10585 this optimization off. If you have code that includes significant
10586 usage of unchecked conversion, you might want to just stick with
10587 @option{-O1} and avoid the entire issue. If you get adequate
10588 performance at this level of optimization level, that's probably
10589 the safest approach. If tests show that you really need higher
10590 levels of optimization, then you can experiment with @option{-O2}
10591 and @option{-O2 -fno-strict-aliasing} to see how much effect this
10592 has on size and speed of the code. If you really need to use
10593 @option{-O2} with strict aliasing in effect, then you should
10594 review any uses of unchecked conversion of access types,
10595 particularly if you are getting the warnings described above.
10598 @node Coverage Analysis
10599 @subsection Coverage Analysis
10602 GNAT supports the HP Performance Coverage Analyzer (PCA), which allows
10603 the user to determine the distribution of execution time across a program,
10604 @pxref{Profiling} for details of usage.
10608 @node Text_IO Suggestions
10609 @section @code{Text_IO} Suggestions
10610 @cindex @code{Text_IO} and performance
10613 The @code{Ada.Text_IO} package has fairly high overheads due in part to
10614 the requirement of maintaining page and line counts. If performance
10615 is critical, a recommendation is to use @code{Stream_IO} instead of
10616 @code{Text_IO} for volume output, since this package has less overhead.
10618 If @code{Text_IO} must be used, note that by default output to the standard
10619 output and standard error files is unbuffered (this provides better
10620 behavior when output statements are used for debugging, or if the
10621 progress of a program is observed by tracking the output, e.g. by
10622 using the Unix @command{tail -f} command to watch redirected output.
10624 If you are generating large volumes of output with @code{Text_IO} and
10625 performance is an important factor, use a designated file instead
10626 of the standard output file, or change the standard output file to
10627 be buffered using @code{Interfaces.C_Streams.setvbuf}.
10631 @node Reducing Size of Ada Executables with gnatelim
10632 @section Reducing Size of Ada Executables with @code{gnatelim}
10636 This section describes @command{gnatelim}, a tool which detects unused
10637 subprograms and helps the compiler to create a smaller executable for your
10642 * Running gnatelim::
10643 * Processing Precompiled Libraries::
10644 * Correcting the List of Eliminate Pragmas::
10645 * Making Your Executables Smaller::
10646 * Summary of the gnatelim Usage Cycle::
10649 @node About gnatelim
10650 @subsection About @code{gnatelim}
10653 When a program shares a set of Ada
10654 packages with other programs, it may happen that this program uses
10655 only a fraction of the subprograms defined in these packages. The code
10656 created for these unused subprograms increases the size of the executable.
10658 @code{gnatelim} tracks unused subprograms in an Ada program and
10659 outputs a list of GNAT-specific pragmas @code{Eliminate} marking all the
10660 subprograms that are declared but never called. By placing the list of
10661 @code{Eliminate} pragmas in the GNAT configuration file @file{gnat.adc} and
10662 recompiling your program, you may decrease the size of its executable,
10663 because the compiler will not generate the code for 'eliminated' subprograms.
10664 @xref{Pragma Eliminate,,, gnat_rm, GNAT Reference Manual}, for more
10665 information about this pragma.
10667 @code{gnatelim} needs as its input data the name of the main subprogram.
10669 If a set of source files is specified as @code{gnatelim} arguments, it
10670 treats these files as a complete set of sources making up a program to
10671 analyse, and analyses only these sources.
10673 After a full successful build of the main subprogram @code{gnatelim} can be
10674 called without specifying sources to analyse, in this case it computes
10675 the source closure of the main unit from the @file{ALI} files.
10677 The following command will create the set of @file{ALI} files needed for
10681 $ gnatmake ^-c Main_Prog^/ACTIONS=COMPILE MAIN_PROG^
10684 Note that @code{gnatelim} does not need object files.
10686 @node Running gnatelim
10687 @subsection Running @code{gnatelim}
10690 @code{gnatelim} has the following command-line interface:
10693 $ gnatelim [@var{switches}] ^-main^?MAIN^=@var{main_unit_name} @{@var{filename}@} @r{[}-cargs @var{gcc_switches}@r{]}
10697 @var{main_unit_name} should be a name of a source file that contains the main
10698 subprogram of a program (partition).
10700 Each @var{filename} is the name (including the extension) of a source
10701 file to process. ``Wildcards'' are allowed, and
10702 the file name may contain path information.
10704 @samp{@var{gcc_switches}} is a list of switches for
10705 @command{gcc}. They will be passed on to all compiler invocations made by
10706 @command{gnatelim} to generate the ASIS trees. Here you can provide
10707 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
10708 use the @option{-gnatec} switch to set the configuration file etc.
10710 @code{gnatelim} has the following switches:
10714 @item ^-files^/FILES^=@var{filename}
10715 @cindex @option{^-files^/FILES^} (@code{gnatelim})
10716 Take the argument source files from the specified file. This file should be an
10717 ordinary text file containing file names separated by spaces or
10718 line breaks. You can use this switch more than once in the same call to
10719 @command{gnatelim}. You also can combine this switch with
10720 an explicit list of files.
10723 @cindex @option{^-log^/LOG^} (@command{gnatelim})
10724 Duplicate all the output sent to @file{stderr} into a log file. The log file
10725 is named @file{gnatelim.log} and is located in the current directory.
10727 @item ^-log^/LOGFILE^=@var{filename}
10728 @cindex @option{^-log^/LOGFILE^} (@command{gnatelim})
10729 Duplicate all the output sent to @file{stderr} into a specified log file.
10731 @cindex @option{^--no-elim-dispatch^/NO_DISPATCH^} (@command{gnatelim})
10732 @item ^--no-elim-dispatch^/NO_DISPATCH^
10733 Do not generate pragmas for dispatching operations.
10735 @cindex @option{^-o^/OUTPUT^} (@command{gnatelim})
10736 @item ^-o^/OUTPUT^=@var{report_file}
10737 Put @command{gnatelim} output into a specified file. If this file already exists,
10738 it is overridden. If this switch is not used, @command{gnatelim} outputs its results
10742 @cindex @option{^-q^/QUIET^} (@command{gnatelim})
10743 Quiet mode: by default @code{gnatelim} outputs to the standard error
10744 stream the number of program units left to be processed. This option turns
10747 @cindex @option{^-t^/TIME^} (@command{gnatelim})
10749 Print out execution time.
10751 @item ^-v^/VERBOSE^
10752 @cindex @option{^-v^/VERBOSE^} (@command{gnatelim})
10753 Verbose mode: @code{gnatelim} version information is printed as Ada
10754 comments to the standard output stream. Also, in addition to the number of
10755 program units left @code{gnatelim} will output the name of the current unit
10758 @item ^-wq^/WARNINGS=QUIET^
10759 @cindex @option{^-wq^/WARNINGS=QUIET^} (@command{gnatelim})
10760 Quet warning mode - some warnings are suppressed. In particular warnings that
10761 indicate that the analysed set of sources is incomplete to make up a
10762 partition and that some subprogram bodies are missing are not generated.
10765 @node Processing Precompiled Libraries
10766 @subsection Processing Precompiled Libraries
10769 If some program uses a precompiled Ada library, it can be processed by
10770 @code{gnatelim} in a usual way. @code{gnatelim} will newer generate an
10771 Eliminate pragma for a subprogram if the body of this subprogram has not
10772 been analysed, this is a typical case for subprograms from precompiled
10773 libraries. Switch @option{^-wq^/WARNINGS=QUIET^} may be used to suppress
10774 warnings about missing source files and non-analyzed subprogram bodies
10775 that can be generated when processing precompiled Ada libraries.
10777 @node Correcting the List of Eliminate Pragmas
10778 @subsection Correcting the List of Eliminate Pragmas
10781 In some rare cases @code{gnatelim} may try to eliminate
10782 subprograms that are actually called in the program. In this case, the
10783 compiler will generate an error message of the form:
10786 main.adb:4:08: cannot reference subprogram "P" eliminated at elim.out:5
10790 You will need to manually remove the wrong @code{Eliminate} pragmas from
10791 the configuration file indicated in the error message. You should recompile
10792 your program from scratch after that, because you need a consistent
10793 configuration file(s) during the entire compilation.
10795 @node Making Your Executables Smaller
10796 @subsection Making Your Executables Smaller
10799 In order to get a smaller executable for your program you now have to
10800 recompile the program completely with the configuration file containing
10801 pragmas Eliminate generated by gnatelim. If these pragmas are placed in
10802 @file{gnat.adc} file located in your current directory, just do:
10805 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10809 (Use the @option{^-f^/FORCE_COMPILE^} option for @command{gnatmake} to
10810 recompile everything
10811 with the set of pragmas @code{Eliminate} that you have obtained with
10812 @command{gnatelim}).
10814 Be aware that the set of @code{Eliminate} pragmas is specific to each
10815 program. It is not recommended to merge sets of @code{Eliminate}
10816 pragmas created for different programs in one configuration file.
10818 @node Summary of the gnatelim Usage Cycle
10819 @subsection Summary of the @code{gnatelim} Usage Cycle
10822 Here is a quick summary of the steps to be taken in order to reduce
10823 the size of your executables with @code{gnatelim}. You may use
10824 other GNAT options to control the optimization level,
10825 to produce the debugging information, to set search path, etc.
10829 Create a complete set of @file{ALI} files (if the program has not been
10833 $ gnatmake ^-c main_prog^/ACTIONS=COMPILE MAIN_PROG^
10837 Generate a list of @code{Eliminate} pragmas in default configuration file
10838 @file{gnat.adc} in the current directory
10841 $ PIPE GNAT ELIM MAIN_PROG > GNAT.ADC
10844 $ gnatelim main_prog >@r{[}>@r{]} gnat.adc
10849 Recompile the application
10852 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10857 @node Reducing Size of Executables with unused subprogram/data elimination
10858 @section Reducing Size of Executables with Unused Subprogram/Data Elimination
10859 @findex unused subprogram/data elimination
10862 This section describes how you can eliminate unused subprograms and data from
10863 your executable just by setting options at compilation time.
10866 * About unused subprogram/data elimination::
10867 * Compilation options::
10868 * Example of unused subprogram/data elimination::
10871 @node About unused subprogram/data elimination
10872 @subsection About unused subprogram/data elimination
10875 By default, an executable contains all code and data of its composing objects
10876 (directly linked or coming from statically linked libraries), even data or code
10877 never used by this executable.
10879 This feature will allow you to eliminate such unused code from your
10880 executable, making it smaller (in disk and in memory).
10882 This functionality is available on all Linux platforms except for the IA-64
10883 architecture and on all cross platforms using the ELF binary file format.
10884 In both cases GNU binutils version 2.16 or later are required to enable it.
10886 @node Compilation options
10887 @subsection Compilation options
10890 The operation of eliminating the unused code and data from the final executable
10891 is directly performed by the linker.
10893 In order to do this, it has to work with objects compiled with the
10895 @option{-ffunction-sections} @option{-fdata-sections}.
10896 @cindex @option{-ffunction-sections} (@command{gcc})
10897 @cindex @option{-fdata-sections} (@command{gcc})
10898 These options are usable with C and Ada files.
10899 They will place respectively each
10900 function or data in a separate section in the resulting object file.
10902 Once the objects and static libraries are created with these options, the
10903 linker can perform the dead code elimination. You can do this by setting
10904 the @option{-Wl,--gc-sections} option to gcc command or in the
10905 @option{-largs} section of @command{gnatmake}. This will perform a
10906 garbage collection of code and data never referenced.
10908 If the linker performs a partial link (@option{-r} ld linker option), then you
10909 will need to provide one or several entry point using the
10910 @option{-e} / @option{--entry} ld option.
10912 Note that objects compiled without the @option{-ffunction-sections} and
10913 @option{-fdata-sections} options can still be linked with the executable.
10914 However, no dead code elimination will be performed on those objects (they will
10917 The GNAT static library is now compiled with -ffunction-sections and
10918 -fdata-sections on some platforms. This allows you to eliminate the unused code
10919 and data of the GNAT library from your executable.
10921 @node Example of unused subprogram/data elimination
10922 @subsection Example of unused subprogram/data elimination
10925 Here is a simple example:
10927 @smallexample @c ada
10936 Used_Data : Integer;
10937 Unused_Data : Integer;
10939 procedure Used (Data : Integer);
10940 procedure Unused (Data : Integer);
10943 package body Aux is
10944 procedure Used (Data : Integer) is
10949 procedure Unused (Data : Integer) is
10951 Unused_Data := Data;
10957 @code{Unused} and @code{Unused_Data} are never referenced in this code
10958 excerpt, and hence they may be safely removed from the final executable.
10963 $ nm test | grep used
10964 020015f0 T aux__unused
10965 02005d88 B aux__unused_data
10966 020015cc T aux__used
10967 02005d84 B aux__used_data
10969 $ gnatmake test -cargs -fdata-sections -ffunction-sections \
10970 -largs -Wl,--gc-sections
10972 $ nm test | grep used
10973 02005350 T aux__used
10974 0201ffe0 B aux__used_data
10978 It can be observed that the procedure @code{Unused} and the object
10979 @code{Unused_Data} are removed by the linker when using the
10980 appropriate options.
10982 @c ********************************
10983 @node Renaming Files Using gnatchop
10984 @chapter Renaming Files Using @code{gnatchop}
10988 This chapter discusses how to handle files with multiple units by using
10989 the @code{gnatchop} utility. This utility is also useful in renaming
10990 files to meet the standard GNAT default file naming conventions.
10993 * Handling Files with Multiple Units::
10994 * Operating gnatchop in Compilation Mode::
10995 * Command Line for gnatchop::
10996 * Switches for gnatchop::
10997 * Examples of gnatchop Usage::
11000 @node Handling Files with Multiple Units
11001 @section Handling Files with Multiple Units
11004 The basic compilation model of GNAT requires that a file submitted to the
11005 compiler have only one unit and there be a strict correspondence
11006 between the file name and the unit name.
11008 The @code{gnatchop} utility allows both of these rules to be relaxed,
11009 allowing GNAT to process files which contain multiple compilation units
11010 and files with arbitrary file names. @code{gnatchop}
11011 reads the specified file and generates one or more output files,
11012 containing one unit per file. The unit and the file name correspond,
11013 as required by GNAT.
11015 If you want to permanently restructure a set of ``foreign'' files so that
11016 they match the GNAT rules, and do the remaining development using the
11017 GNAT structure, you can simply use @command{gnatchop} once, generate the
11018 new set of files and work with them from that point on.
11020 Alternatively, if you want to keep your files in the ``foreign'' format,
11021 perhaps to maintain compatibility with some other Ada compilation
11022 system, you can set up a procedure where you use @command{gnatchop} each
11023 time you compile, regarding the source files that it writes as temporary
11024 files that you throw away.
11026 Note that if your file containing multiple units starts with a byte order
11027 mark (BOM) specifying UTF-8 encoding, then the files generated by gnatchop
11028 will each start with a copy of this BOM, meaning that they can be compiled
11029 automatically in UTF-8 mode without needing to specify an explicit encoding.
11031 @node Operating gnatchop in Compilation Mode
11032 @section Operating gnatchop in Compilation Mode
11035 The basic function of @code{gnatchop} is to take a file with multiple units
11036 and split it into separate files. The boundary between files is reasonably
11037 clear, except for the issue of comments and pragmas. In default mode, the
11038 rule is that any pragmas between units belong to the previous unit, except
11039 that configuration pragmas always belong to the following unit. Any comments
11040 belong to the following unit. These rules
11041 almost always result in the right choice of
11042 the split point without needing to mark it explicitly and most users will
11043 find this default to be what they want. In this default mode it is incorrect to
11044 submit a file containing only configuration pragmas, or one that ends in
11045 configuration pragmas, to @code{gnatchop}.
11047 However, using a special option to activate ``compilation mode'',
11049 can perform another function, which is to provide exactly the semantics
11050 required by the RM for handling of configuration pragmas in a compilation.
11051 In the absence of configuration pragmas (at the main file level), this
11052 option has no effect, but it causes such configuration pragmas to be handled
11053 in a quite different manner.
11055 First, in compilation mode, if @code{gnatchop} is given a file that consists of
11056 only configuration pragmas, then this file is appended to the
11057 @file{gnat.adc} file in the current directory. This behavior provides
11058 the required behavior described in the RM for the actions to be taken
11059 on submitting such a file to the compiler, namely that these pragmas
11060 should apply to all subsequent compilations in the same compilation
11061 environment. Using GNAT, the current directory, possibly containing a
11062 @file{gnat.adc} file is the representation
11063 of a compilation environment. For more information on the
11064 @file{gnat.adc} file, see @ref{Handling of Configuration Pragmas}.
11066 Second, in compilation mode, if @code{gnatchop}
11067 is given a file that starts with
11068 configuration pragmas, and contains one or more units, then these
11069 configuration pragmas are prepended to each of the chopped files. This
11070 behavior provides the required behavior described in the RM for the
11071 actions to be taken on compiling such a file, namely that the pragmas
11072 apply to all units in the compilation, but not to subsequently compiled
11075 Finally, if configuration pragmas appear between units, they are appended
11076 to the previous unit. This results in the previous unit being illegal,
11077 since the compiler does not accept configuration pragmas that follow
11078 a unit. This provides the required RM behavior that forbids configuration
11079 pragmas other than those preceding the first compilation unit of a
11082 For most purposes, @code{gnatchop} will be used in default mode. The
11083 compilation mode described above is used only if you need exactly
11084 accurate behavior with respect to compilations, and you have files
11085 that contain multiple units and configuration pragmas. In this
11086 circumstance the use of @code{gnatchop} with the compilation mode
11087 switch provides the required behavior, and is for example the mode
11088 in which GNAT processes the ACVC tests.
11090 @node Command Line for gnatchop
11091 @section Command Line for @code{gnatchop}
11094 The @code{gnatchop} command has the form:
11097 @c $ gnatchop switches @var{file name} @r{[}@var{file name} @dots{}@r{]}
11098 @c @ovar{directory}
11099 @c Expanding @ovar macro inline (explanation in macro def comments)
11100 $ gnatchop switches @var{file name} @r{[}@var{file name} @dots{}@r{]}
11101 @r{[}@var{directory}@r{]}
11105 The only required argument is the file name of the file to be chopped.
11106 There are no restrictions on the form of this file name. The file itself
11107 contains one or more Ada units, in normal GNAT format, concatenated
11108 together. As shown, more than one file may be presented to be chopped.
11110 When run in default mode, @code{gnatchop} generates one output file in
11111 the current directory for each unit in each of the files.
11113 @var{directory}, if specified, gives the name of the directory to which
11114 the output files will be written. If it is not specified, all files are
11115 written to the current directory.
11117 For example, given a
11118 file called @file{hellofiles} containing
11120 @smallexample @c ada
11125 with Text_IO; use Text_IO;
11128 Put_Line ("Hello");
11138 $ gnatchop ^hellofiles^HELLOFILES.^
11142 generates two files in the current directory, one called
11143 @file{hello.ads} containing the single line that is the procedure spec,
11144 and the other called @file{hello.adb} containing the remaining text. The
11145 original file is not affected. The generated files can be compiled in
11149 When gnatchop is invoked on a file that is empty or that contains only empty
11150 lines and/or comments, gnatchop will not fail, but will not produce any
11153 For example, given a
11154 file called @file{toto.txt} containing
11156 @smallexample @c ada
11168 $ gnatchop ^toto.txt^TOT.TXT^
11172 will not produce any new file and will result in the following warnings:
11175 toto.txt:1:01: warning: empty file, contains no compilation units
11176 no compilation units found
11177 no source files written
11180 @node Switches for gnatchop
11181 @section Switches for @code{gnatchop}
11184 @command{gnatchop} recognizes the following switches:
11190 @cindex @option{--version} @command{gnatchop}
11191 Display Copyright and version, then exit disregarding all other options.
11194 @cindex @option{--help} @command{gnatchop}
11195 If @option{--version} was not used, display usage, then exit disregarding
11198 @item ^-c^/COMPILATION^
11199 @cindex @option{^-c^/COMPILATION^} (@code{gnatchop})
11200 Causes @code{gnatchop} to operate in compilation mode, in which
11201 configuration pragmas are handled according to strict RM rules. See
11202 previous section for a full description of this mode.
11205 @item -gnat@var{xxx}
11206 This passes the given @option{-gnat@var{xxx}} switch to @code{gnat} which is
11207 used to parse the given file. Not all @var{xxx} options make sense,
11208 but for example, the use of @option{-gnati2} allows @code{gnatchop} to
11209 process a source file that uses Latin-2 coding for identifiers.
11213 Causes @code{gnatchop} to generate a brief help summary to the standard
11214 output file showing usage information.
11216 @item ^-k@var{mm}^/FILE_NAME_MAX_LENGTH=@var{mm}^
11217 @cindex @option{^-k^/FILE_NAME_MAX_LENGTH^} (@code{gnatchop})
11218 Limit generated file names to the specified number @code{mm}
11220 This is useful if the
11221 resulting set of files is required to be interoperable with systems
11222 which limit the length of file names.
11224 If no value is given, or
11225 if no @code{/FILE_NAME_MAX_LENGTH} qualifier is given,
11226 a default of 39, suitable for OpenVMS Alpha
11227 Systems, is assumed
11230 No space is allowed between the @option{-k} and the numeric value. The numeric
11231 value may be omitted in which case a default of @option{-k8},
11233 with DOS-like file systems, is used. If no @option{-k} switch
11235 there is no limit on the length of file names.
11238 @item ^-p^/PRESERVE^
11239 @cindex @option{^-p^/PRESERVE^} (@code{gnatchop})
11240 Causes the file ^modification^creation^ time stamp of the input file to be
11241 preserved and used for the time stamp of the output file(s). This may be
11242 useful for preserving coherency of time stamps in an environment where
11243 @code{gnatchop} is used as part of a standard build process.
11246 @cindex @option{^-q^/QUIET^} (@code{gnatchop})
11247 Causes output of informational messages indicating the set of generated
11248 files to be suppressed. Warnings and error messages are unaffected.
11250 @item ^-r^/REFERENCE^
11251 @cindex @option{^-r^/REFERENCE^} (@code{gnatchop})
11252 @findex Source_Reference
11253 Generate @code{Source_Reference} pragmas. Use this switch if the output
11254 files are regarded as temporary and development is to be done in terms
11255 of the original unchopped file. This switch causes
11256 @code{Source_Reference} pragmas to be inserted into each of the
11257 generated files to refers back to the original file name and line number.
11258 The result is that all error messages refer back to the original
11260 In addition, the debugging information placed into the object file (when
11261 the @option{^-g^/DEBUG^} switch of @command{gcc} or @command{gnatmake} is
11263 also refers back to this original file so that tools like profilers and
11264 debuggers will give information in terms of the original unchopped file.
11266 If the original file to be chopped itself contains
11267 a @code{Source_Reference}
11268 pragma referencing a third file, then gnatchop respects
11269 this pragma, and the generated @code{Source_Reference} pragmas
11270 in the chopped file refer to the original file, with appropriate
11271 line numbers. This is particularly useful when @code{gnatchop}
11272 is used in conjunction with @code{gnatprep} to compile files that
11273 contain preprocessing statements and multiple units.
11275 @item ^-v^/VERBOSE^
11276 @cindex @option{^-v^/VERBOSE^} (@code{gnatchop})
11277 Causes @code{gnatchop} to operate in verbose mode. The version
11278 number and copyright notice are output, as well as exact copies of
11279 the gnat1 commands spawned to obtain the chop control information.
11281 @item ^-w^/OVERWRITE^
11282 @cindex @option{^-w^/OVERWRITE^} (@code{gnatchop})
11283 Overwrite existing file names. Normally @code{gnatchop} regards it as a
11284 fatal error if there is already a file with the same name as a
11285 file it would otherwise output, in other words if the files to be
11286 chopped contain duplicated units. This switch bypasses this
11287 check, and causes all but the last instance of such duplicated
11288 units to be skipped.
11291 @item --GCC=@var{xxxx}
11292 @cindex @option{--GCC=} (@code{gnatchop})
11293 Specify the path of the GNAT parser to be used. When this switch is used,
11294 no attempt is made to add the prefix to the GNAT parser executable.
11298 @node Examples of gnatchop Usage
11299 @section Examples of @code{gnatchop} Usage
11303 @item gnatchop /OVERWRITE HELLO_S.ADA [PRERELEASE.FILES]
11306 @item gnatchop -w hello_s.ada prerelease/files
11309 Chops the source file @file{hello_s.ada}. The output files will be
11310 placed in the directory @file{^prerelease/files^[PRERELEASE.FILES]^},
11312 files with matching names in that directory (no files in the current
11313 directory are modified).
11315 @item gnatchop ^archive^ARCHIVE.^
11316 Chops the source file @file{^archive^ARCHIVE.^}
11317 into the current directory. One
11318 useful application of @code{gnatchop} is in sending sets of sources
11319 around, for example in email messages. The required sources are simply
11320 concatenated (for example, using a ^Unix @code{cat}^VMS @code{APPEND/NEW}^
11322 @command{gnatchop} is used at the other end to reconstitute the original
11325 @item gnatchop file1 file2 file3 direc
11326 Chops all units in files @file{file1}, @file{file2}, @file{file3}, placing
11327 the resulting files in the directory @file{direc}. Note that if any units
11328 occur more than once anywhere within this set of files, an error message
11329 is generated, and no files are written. To override this check, use the
11330 @option{^-w^/OVERWRITE^} switch,
11331 in which case the last occurrence in the last file will
11332 be the one that is output, and earlier duplicate occurrences for a given
11333 unit will be skipped.
11336 @node Configuration Pragmas
11337 @chapter Configuration Pragmas
11338 @cindex Configuration pragmas
11339 @cindex Pragmas, configuration
11342 Configuration pragmas include those pragmas described as
11343 such in the Ada Reference Manual, as well as
11344 implementation-dependent pragmas that are configuration pragmas.
11345 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
11346 for details on these additional GNAT-specific configuration pragmas.
11347 Most notably, the pragma @code{Source_File_Name}, which allows
11348 specifying non-default names for source files, is a configuration
11349 pragma. The following is a complete list of configuration pragmas
11350 recognized by GNAT:
11358 Assume_No_Invalid_Values
11363 Compile_Time_Warning
11365 Component_Alignment
11366 Convention_Identifier
11374 External_Name_Casing
11377 Float_Representation
11390 Priority_Specific_Dispatching
11393 Propagate_Exceptions
11396 Restricted_Run_Time
11398 Restrictions_Warnings
11401 Source_File_Name_Project
11404 Suppress_Exception_Locations
11405 Task_Dispatching_Policy
11411 Wide_Character_Encoding
11416 * Handling of Configuration Pragmas::
11417 * The Configuration Pragmas Files::
11420 @node Handling of Configuration Pragmas
11421 @section Handling of Configuration Pragmas
11423 Configuration pragmas may either appear at the start of a compilation
11424 unit, in which case they apply only to that unit, or they may apply to
11425 all compilations performed in a given compilation environment.
11427 GNAT also provides the @code{gnatchop} utility to provide an automatic
11428 way to handle configuration pragmas following the semantics for
11429 compilations (that is, files with multiple units), described in the RM.
11430 See @ref{Operating gnatchop in Compilation Mode} for details.
11431 However, for most purposes, it will be more convenient to edit the
11432 @file{gnat.adc} file that contains configuration pragmas directly,
11433 as described in the following section.
11435 @node The Configuration Pragmas Files
11436 @section The Configuration Pragmas Files
11437 @cindex @file{gnat.adc}
11440 In GNAT a compilation environment is defined by the current
11441 directory at the time that a compile command is given. This current
11442 directory is searched for a file whose name is @file{gnat.adc}. If
11443 this file is present, it is expected to contain one or more
11444 configuration pragmas that will be applied to the current compilation.
11445 However, if the switch @option{-gnatA} is used, @file{gnat.adc} is not
11448 Configuration pragmas may be entered into the @file{gnat.adc} file
11449 either by running @code{gnatchop} on a source file that consists only of
11450 configuration pragmas, or more conveniently by
11451 direct editing of the @file{gnat.adc} file, which is a standard format
11454 In addition to @file{gnat.adc}, additional files containing configuration
11455 pragmas may be applied to the current compilation using the switch
11456 @option{-gnatec}@var{path}. @var{path} must designate an existing file that
11457 contains only configuration pragmas. These configuration pragmas are
11458 in addition to those found in @file{gnat.adc} (provided @file{gnat.adc}
11459 is present and switch @option{-gnatA} is not used).
11461 It is allowed to specify several switches @option{-gnatec}, all of which
11462 will be taken into account.
11464 If you are using project file, a separate mechanism is provided using
11465 project attributes, see @ref{Specifying Configuration Pragmas} for more
11469 Of special interest to GNAT OpenVMS Alpha is the following
11470 configuration pragma:
11472 @smallexample @c ada
11474 pragma Extend_System (Aux_DEC);
11479 In the presence of this pragma, GNAT adds to the definition of the
11480 predefined package SYSTEM all the additional types and subprograms that are
11481 defined in HP Ada. See @ref{Compatibility with HP Ada} for details.
11484 @node Handling Arbitrary File Naming Conventions Using gnatname
11485 @chapter Handling Arbitrary File Naming Conventions Using @code{gnatname}
11486 @cindex Arbitrary File Naming Conventions
11489 * Arbitrary File Naming Conventions::
11490 * Running gnatname::
11491 * Switches for gnatname::
11492 * Examples of gnatname Usage::
11495 @node Arbitrary File Naming Conventions
11496 @section Arbitrary File Naming Conventions
11499 The GNAT compiler must be able to know the source file name of a compilation
11500 unit. When using the standard GNAT default file naming conventions
11501 (@code{.ads} for specs, @code{.adb} for bodies), the GNAT compiler
11502 does not need additional information.
11505 When the source file names do not follow the standard GNAT default file naming
11506 conventions, the GNAT compiler must be given additional information through
11507 a configuration pragmas file (@pxref{Configuration Pragmas})
11509 When the non-standard file naming conventions are well-defined,
11510 a small number of pragmas @code{Source_File_Name} specifying a naming pattern
11511 (@pxref{Alternative File Naming Schemes}) may be sufficient. However,
11512 if the file naming conventions are irregular or arbitrary, a number
11513 of pragma @code{Source_File_Name} for individual compilation units
11515 To help maintain the correspondence between compilation unit names and
11516 source file names within the compiler,
11517 GNAT provides a tool @code{gnatname} to generate the required pragmas for a
11520 @node Running gnatname
11521 @section Running @code{gnatname}
11524 The usual form of the @code{gnatname} command is
11527 @c $ gnatname @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}
11528 @c @r{[}--and @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}@r{]}
11529 @c Expanding @ovar macro inline (explanation in macro def comments)
11530 $ gnatname @r{[}@var{switches}@r{]} @var{naming_pattern} @r{[}@var{naming_patterns}@r{]}
11531 @r{[}--and @r{[}@var{switches}@r{]} @var{naming_pattern} @r{[}@var{naming_patterns}@r{]}@r{]}
11535 All of the arguments are optional. If invoked without any argument,
11536 @code{gnatname} will display its usage.
11539 When used with at least one naming pattern, @code{gnatname} will attempt to
11540 find all the compilation units in files that follow at least one of the
11541 naming patterns. To find these compilation units,
11542 @code{gnatname} will use the GNAT compiler in syntax-check-only mode on all
11546 One or several Naming Patterns may be given as arguments to @code{gnatname}.
11547 Each Naming Pattern is enclosed between double quotes.
11548 A Naming Pattern is a regular expression similar to the wildcard patterns
11549 used in file names by the Unix shells or the DOS prompt.
11552 @code{gnatname} may be called with several sections of directories/patterns.
11553 Sections are separated by switch @code{--and}. In each section, there must be
11554 at least one pattern. If no directory is specified in a section, the current
11555 directory (or the project directory is @code{-P} is used) is implied.
11556 The options other that the directory switches and the patterns apply globally
11557 even if they are in different sections.
11560 Examples of Naming Patterns are
11569 For a more complete description of the syntax of Naming Patterns,
11570 see the second kind of regular expressions described in @file{g-regexp.ads}
11571 (the ``Glob'' regular expressions).
11574 When invoked with no switch @code{-P}, @code{gnatname} will create a
11575 configuration pragmas file @file{gnat.adc} in the current working directory,
11576 with pragmas @code{Source_File_Name} for each file that contains a valid Ada
11579 @node Switches for gnatname
11580 @section Switches for @code{gnatname}
11583 Switches for @code{gnatname} must precede any specified Naming Pattern.
11586 You may specify any of the following switches to @code{gnatname}:
11592 @cindex @option{--version} @command{gnatname}
11593 Display Copyright and version, then exit disregarding all other options.
11596 @cindex @option{--help} @command{gnatname}
11597 If @option{--version} was not used, display usage, then exit disregarding
11601 Start another section of directories/patterns.
11603 @item ^-c^/CONFIG_FILE=^@file{file}
11604 @cindex @option{^-c^/CONFIG_FILE^} (@code{gnatname})
11605 Create a configuration pragmas file @file{file} (instead of the default
11608 There may be zero, one or more space between @option{-c} and
11611 @file{file} may include directory information. @file{file} must be
11612 writable. There may be only one switch @option{^-c^/CONFIG_FILE^}.
11613 When a switch @option{^-c^/CONFIG_FILE^} is
11614 specified, no switch @option{^-P^/PROJECT_FILE^} may be specified (see below).
11616 @item ^-d^/SOURCE_DIRS=^@file{dir}
11617 @cindex @option{^-d^/SOURCE_DIRS^} (@code{gnatname})
11618 Look for source files in directory @file{dir}. There may be zero, one or more
11619 spaces between @option{^-d^/SOURCE_DIRS=^} and @file{dir}.
11620 When a switch @option{^-d^/SOURCE_DIRS^}
11621 is specified, the current working directory will not be searched for source
11622 files, unless it is explicitly specified with a @option{^-d^/SOURCE_DIRS^}
11623 or @option{^-D^/DIR_FILES^} switch.
11624 Several switches @option{^-d^/SOURCE_DIRS^} may be specified.
11625 If @file{dir} is a relative path, it is relative to the directory of
11626 the configuration pragmas file specified with switch
11627 @option{^-c^/CONFIG_FILE^},
11628 or to the directory of the project file specified with switch
11629 @option{^-P^/PROJECT_FILE^} or,
11630 if neither switch @option{^-c^/CONFIG_FILE^}
11631 nor switch @option{^-P^/PROJECT_FILE^} are specified, it is relative to the
11632 current working directory. The directory
11633 specified with switch @option{^-d^/SOURCE_DIRS^} must exist and be readable.
11635 @item ^-D^/DIRS_FILE=^@file{file}
11636 @cindex @option{^-D^/DIRS_FILE^} (@code{gnatname})
11637 Look for source files in all directories listed in text file @file{file}.
11638 There may be zero, one or more spaces between @option{^-D^/DIRS_FILE=^}
11640 @file{file} must be an existing, readable text file.
11641 Each nonempty line in @file{file} must be a directory.
11642 Specifying switch @option{^-D^/DIRS_FILE^} is equivalent to specifying as many
11643 switches @option{^-d^/SOURCE_DIRS^} as there are nonempty lines in
11646 @item ^-f^/FOREIGN_PATTERN=^@file{pattern}
11647 @cindex @option{^-f^/FOREIGN_PATTERN^} (@code{gnatname})
11648 Foreign patterns. Using this switch, it is possible to add sources of languages
11649 other than Ada to the list of sources of a project file.
11650 It is only useful if a ^-P^/PROJECT_FILE^ switch is used.
11653 gnatname ^-Pprj -f"*.c"^/PROJECT_FILE=PRJ /FOREIGN_PATTERN=*.C^ "*.ada"
11656 will look for Ada units in all files with the @file{.ada} extension,
11657 and will add to the list of file for project @file{prj.gpr} the C files
11658 with extension @file{.^c^C^}.
11661 @cindex @option{^-h^/HELP^} (@code{gnatname})
11662 Output usage (help) information. The output is written to @file{stdout}.
11664 @item ^-P^/PROJECT_FILE=^@file{proj}
11665 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatname})
11666 Create or update project file @file{proj}. There may be zero, one or more space
11667 between @option{-P} and @file{proj}. @file{proj} may include directory
11668 information. @file{proj} must be writable.
11669 There may be only one switch @option{^-P^/PROJECT_FILE^}.
11670 When a switch @option{^-P^/PROJECT_FILE^} is specified,
11671 no switch @option{^-c^/CONFIG_FILE^} may be specified.
11673 @item ^-v^/VERBOSE^
11674 @cindex @option{^-v^/VERBOSE^} (@code{gnatname})
11675 Verbose mode. Output detailed explanation of behavior to @file{stdout}.
11676 This includes name of the file written, the name of the directories to search
11677 and, for each file in those directories whose name matches at least one of
11678 the Naming Patterns, an indication of whether the file contains a unit,
11679 and if so the name of the unit.
11681 @item ^-v -v^/VERBOSE /VERBOSE^
11682 @cindex @option{^-v -v^/VERBOSE /VERBOSE^} (@code{gnatname})
11683 Very Verbose mode. In addition to the output produced in verbose mode,
11684 for each file in the searched directories whose name matches none of
11685 the Naming Patterns, an indication is given that there is no match.
11687 @item ^-x^/EXCLUDED_PATTERN=^@file{pattern}
11688 @cindex @option{^-x^/EXCLUDED_PATTERN^} (@code{gnatname})
11689 Excluded patterns. Using this switch, it is possible to exclude some files
11690 that would match the name patterns. For example,
11692 gnatname ^-x "*_nt.ada"^/EXCLUDED_PATTERN=*_nt.ada^ "*.ada"
11695 will look for Ada units in all files with the @file{.ada} extension,
11696 except those whose names end with @file{_nt.ada}.
11700 @node Examples of gnatname Usage
11701 @section Examples of @code{gnatname} Usage
11705 $ gnatname /CONFIG_FILE=[HOME.ME]NAMES.ADC /SOURCE_DIRS=SOURCES "[a-z]*.ada*"
11711 $ gnatname -c /home/me/names.adc -d sources "[a-z]*.ada*"
11716 In this example, the directory @file{^/home/me^[HOME.ME]^} must already exist
11717 and be writable. In addition, the directory
11718 @file{^/home/me/sources^[HOME.ME.SOURCES]^} (specified by
11719 @option{^-d sources^/SOURCE_DIRS=SOURCES^}) must exist and be readable.
11722 Note the optional spaces after @option{-c} and @option{-d}.
11727 $ gnatname -P/home/me/proj -x "*_nt_body.ada"
11728 -dsources -dsources/plus -Dcommon_dirs.txt "body_*" "spec_*"
11731 $ gnatname /PROJECT_FILE=[HOME.ME]PROJ
11732 /EXCLUDED_PATTERN=*_nt_body.ada
11733 /SOURCE_DIRS=(SOURCES,[SOURCES.PLUS])
11734 /DIRS_FILE=COMMON_DIRS.TXT "body_*" "spec_*"
11738 Note that several switches @option{^-d^/SOURCE_DIRS^} may be used,
11739 even in conjunction with one or several switches
11740 @option{^-D^/DIRS_FILE^}. Several Naming Patterns and one excluded pattern
11741 are used in this example.
11743 @c *****************************************
11744 @c * G N A T P r o j e c t M a n a g e r *
11745 @c *****************************************
11746 @node GNAT Project Manager
11747 @chapter GNAT Project Manager
11751 * Examples of Project Files::
11752 * Project File Syntax::
11753 * Objects and Sources in Project Files::
11754 * Importing Projects::
11755 * Project Extension::
11756 * Project Hierarchy Extension::
11757 * External References in Project Files::
11758 * Packages in Project Files::
11759 * Variables from Imported Projects::
11761 * Library Projects::
11762 * Stand-alone Library Projects::
11763 * Switches Related to Project Files::
11764 * Tools Supporting Project Files::
11765 * An Extended Example::
11766 * Project File Complete Syntax::
11769 @c ****************
11770 @c * Introduction *
11771 @c ****************
11774 @section Introduction
11777 This chapter describes GNAT's @emph{Project Manager}, a facility that allows
11778 you to manage complex builds involving a number of source files, directories,
11779 and compilation options for different system configurations. In particular,
11780 project files allow you to specify:
11783 The directory or set of directories containing the source files, and/or the
11784 names of the specific source files themselves
11786 The directory in which the compiler's output
11787 (@file{ALI} files, object files, tree files) is to be placed
11789 The directory in which the executable programs is to be placed
11791 ^Switch^Switch^ settings for any of the project-enabled tools
11792 (@command{gnatmake}, compiler, binder, linker, @code{gnatls}, @code{gnatxref},
11793 @code{gnatfind}); you can apply these settings either globally or to individual
11796 The source files containing the main subprogram(s) to be built
11798 The source programming language(s) (currently Ada and/or C)
11800 Source file naming conventions; you can specify these either globally or for
11801 individual compilation units
11808 @node Project Files
11809 @subsection Project Files
11812 Project files are written in a syntax close to that of Ada, using familiar
11813 notions such as packages, context clauses, declarations, default values,
11814 assignments, and inheritance. Finally, project files can be built
11815 hierarchically from other project files, simplifying complex system
11816 integration and project reuse.
11818 A @dfn{project} is a specific set of values for various compilation properties.
11819 The settings for a given project are described by means of
11820 a @dfn{project file}, which is a text file written in an Ada-like syntax.
11821 Property values in project files are either strings or lists of strings.
11822 Properties that are not explicitly set receive default values. A project
11823 file may interrogate the values of @dfn{external variables} (user-defined
11824 command-line switches or environment variables), and it may specify property
11825 settings conditionally, based on the value of such variables.
11827 In simple cases, a project's source files depend only on other source files
11828 in the same project, or on the predefined libraries. (@emph{Dependence} is
11830 the Ada technical sense; as in one Ada unit @code{with}ing another.) However,
11831 the Project Manager also allows more sophisticated arrangements,
11832 where the source files in one project depend on source files in other
11836 One project can @emph{import} other projects containing needed source files.
11838 You can organize GNAT projects in a hierarchy: a @emph{child} project
11839 can extend a @emph{parent} project, inheriting the parent's source files and
11840 optionally overriding any of them with alternative versions
11844 More generally, the Project Manager lets you structure large development
11845 efforts into hierarchical subsystems, where build decisions are delegated
11846 to the subsystem level, and thus different compilation environments
11847 (^switch^switch^ settings) used for different subsystems.
11849 The Project Manager is invoked through the
11850 @option{^-P^/PROJECT_FILE=^@emph{projectfile}}
11851 switch to @command{gnatmake} or to the @command{^gnat^GNAT^} front driver.
11853 There may be zero, one or more spaces between @option{-P} and
11854 @option{@emph{projectfile}}.
11856 If you want to define (on the command line) an external variable that is
11857 queried by the project file, you must use the
11858 @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
11859 The Project Manager parses and interprets the project file, and drives the
11860 invoked tool based on the project settings.
11862 The Project Manager supports a wide range of development strategies,
11863 for systems of all sizes. Here are some typical practices that are
11867 Using a common set of source files, but generating object files in different
11868 directories via different ^switch^switch^ settings
11870 Using a mostly-shared set of source files, but with different versions of
11875 The destination of an executable can be controlled inside a project file
11876 using the @option{^-o^-o^}
11878 In the absence of such a ^switch^switch^ either inside
11879 the project file or on the command line, any executable files generated by
11880 @command{gnatmake} are placed in the directory @code{Exec_Dir} specified
11881 in the project file. If no @code{Exec_Dir} is specified, they will be placed
11882 in the object directory of the project.
11884 You can use project files to achieve some of the effects of a source
11885 versioning system (for example, defining separate projects for
11886 the different sets of sources that comprise different releases) but the
11887 Project Manager is independent of any source configuration management tools
11888 that might be used by the developers.
11890 The next section introduces the main features of GNAT's project facility
11891 through a sequence of examples; subsequent sections will present the syntax
11892 and semantics in more detail. A more formal description of the project
11893 facility appears in @ref{Project File Reference,,, gnat_rm, GNAT
11896 @c *****************************
11897 @c * Examples of Project Files *
11898 @c *****************************
11900 @node Examples of Project Files
11901 @section Examples of Project Files
11903 This section illustrates some of the typical uses of project files and
11904 explains their basic structure and behavior.
11907 * Common Sources with Different ^Switches^Switches^ and Directories::
11908 * Using External Variables::
11909 * Importing Other Projects::
11910 * Extending a Project::
11913 @node Common Sources with Different ^Switches^Switches^ and Directories
11914 @subsection Common Sources with Different ^Switches^Switches^ and Directories
11918 * Specifying the Object Directory::
11919 * Specifying the Exec Directory::
11920 * Project File Packages::
11921 * Specifying ^Switch^Switch^ Settings::
11922 * Main Subprograms::
11923 * Executable File Names::
11924 * Source File Naming Conventions::
11925 * Source Language(s)::
11929 Suppose that the Ada source files @file{pack.ads}, @file{pack.adb}, and
11930 @file{proc.adb} are in the @file{/common} directory. The file
11931 @file{proc.adb} contains an Ada main subprogram @code{Proc} that @code{with}s
11932 package @code{Pack}. We want to compile these source files under two sets
11933 of ^switches^switches^:
11936 When debugging, we want to pass the @option{-g} switch to @command{gnatmake},
11937 and the @option{^-gnata^-gnata^},
11938 @option{^-gnato^-gnato^},
11939 and @option{^-gnatE^-gnatE^} switches to the
11940 compiler; the compiler's output is to appear in @file{/common/debug}
11942 When preparing a release version, we want to pass the @option{^-O2^O2^} switch
11943 to the compiler; the compiler's output is to appear in @file{/common/release}
11947 The GNAT project files shown below, respectively @file{debug.gpr} and
11948 @file{release.gpr} in the @file{/common} directory, achieve these effects.
11961 ^/common/debug^[COMMON.DEBUG]^
11966 ^/common/release^[COMMON.RELEASE]^
11971 Here are the corresponding project files:
11973 @smallexample @c projectfile
11976 for Object_Dir use "debug";
11977 for Main use ("proc");
11980 for ^Default_Switches^Default_Switches^ ("Ada")
11982 for Executable ("proc.adb") use "proc1";
11987 package Compiler is
11988 for ^Default_Switches^Default_Switches^ ("Ada")
11989 use ("-fstack-check",
11992 "^-gnatE^-gnatE^");
11998 @smallexample @c projectfile
12001 for Object_Dir use "release";
12002 for Exec_Dir use ".";
12003 for Main use ("proc");
12005 package Compiler is
12006 for ^Default_Switches^Default_Switches^ ("Ada")
12014 The name of the project defined by @file{debug.gpr} is @code{"Debug"} (case
12015 insensitive), and analogously the project defined by @file{release.gpr} is
12016 @code{"Release"}. For consistency the file should have the same name as the
12017 project, and the project file's extension should be @code{"gpr"}. These
12018 conventions are not required, but a warning is issued if they are not followed.
12020 If the current directory is @file{^/temp^[TEMP]^}, then the command
12022 gnatmake ^-P/common/debug.gpr^/PROJECT_FILE=[COMMON]DEBUG^
12026 generates object and ALI files in @file{^/common/debug^[COMMON.DEBUG]^},
12027 as well as the @code{^proc1^PROC1.EXE^} executable,
12028 using the ^switch^switch^ settings defined in the project file.
12030 Likewise, the command
12032 gnatmake ^-P/common/release.gpr^/PROJECT_FILE=[COMMON]RELEASE^
12036 generates object and ALI files in @file{^/common/release^[COMMON.RELEASE]^},
12037 and the @code{^proc^PROC.EXE^}
12038 executable in @file{^/common^[COMMON]^},
12039 using the ^switch^switch^ settings from the project file.
12042 @unnumberedsubsubsec Source Files
12045 If a project file does not explicitly specify a set of source directories or
12046 a set of source files, then by default the project's source files are the
12047 Ada source files in the project file directory. Thus @file{pack.ads},
12048 @file{pack.adb}, and @file{proc.adb} are the source files for both projects.
12050 @node Specifying the Object Directory
12051 @unnumberedsubsubsec Specifying the Object Directory
12054 Several project properties are modeled by Ada-style @emph{attributes};
12055 a property is defined by supplying the equivalent of an Ada attribute
12056 definition clause in the project file.
12057 A project's object directory is another such a property; the corresponding
12058 attribute is @code{Object_Dir}, and its value is also a string expression,
12059 specified either as absolute or relative. In the later case,
12060 it is relative to the project file directory. Thus the compiler's
12061 output is directed to @file{^/common/debug^[COMMON.DEBUG]^}
12062 (for the @code{Debug} project)
12063 and to @file{^/common/release^[COMMON.RELEASE]^}
12064 (for the @code{Release} project).
12065 If @code{Object_Dir} is not specified, then the default is the project file
12068 @node Specifying the Exec Directory
12069 @unnumberedsubsubsec Specifying the Exec Directory
12072 A project's exec directory is another property; the corresponding
12073 attribute is @code{Exec_Dir}, and its value is also a string expression,
12074 either specified as relative or absolute. If @code{Exec_Dir} is not specified,
12075 then the default is the object directory (which may also be the project file
12076 directory if attribute @code{Object_Dir} is not specified). Thus the executable
12077 is placed in @file{^/common/debug^[COMMON.DEBUG]^}
12078 for the @code{Debug} project (attribute @code{Exec_Dir} not specified)
12079 and in @file{^/common^[COMMON]^} for the @code{Release} project.
12081 @node Project File Packages
12082 @unnumberedsubsubsec Project File Packages
12085 A GNAT tool that is integrated with the Project Manager is modeled by a
12086 corresponding package in the project file. In the example above,
12087 The @code{Debug} project defines the packages @code{Builder}
12088 (for @command{gnatmake}) and @code{Compiler};
12089 the @code{Release} project defines only the @code{Compiler} package.
12091 The Ada-like package syntax is not to be taken literally. Although packages in
12092 project files bear a surface resemblance to packages in Ada source code, the
12093 notation is simply a way to convey a grouping of properties for a named
12094 entity. Indeed, the package names permitted in project files are restricted
12095 to a predefined set, corresponding to the project-aware tools, and the contents
12096 of packages are limited to a small set of constructs.
12097 The packages in the example above contain attribute definitions.
12099 @node Specifying ^Switch^Switch^ Settings
12100 @unnumberedsubsubsec Specifying ^Switch^Switch^ Settings
12103 ^Switch^Switch^ settings for a project-aware tool can be specified through
12104 attributes in the package that corresponds to the tool.
12105 The example above illustrates one of the relevant attributes,
12106 @code{^Default_Switches^Default_Switches^}, which is defined in packages
12107 in both project files.
12108 Unlike simple attributes like @code{Source_Dirs},
12109 @code{^Default_Switches^Default_Switches^} is
12110 known as an @emph{associative array}. When you define this attribute, you must
12111 supply an ``index'' (a literal string), and the effect of the attribute
12112 definition is to set the value of the array at the specified index.
12113 For the @code{^Default_Switches^Default_Switches^} attribute,
12114 the index is a programming language (in our case, Ada),
12115 and the value specified (after @code{use}) must be a list
12116 of string expressions.
12118 The attributes permitted in project files are restricted to a predefined set.
12119 Some may appear at project level, others in packages.
12120 For any attribute that is an associative array, the index must always be a
12121 literal string, but the restrictions on this string (e.g., a file name or a
12122 language name) depend on the individual attribute.
12123 Also depending on the attribute, its specified value will need to be either a
12124 string or a string list.
12126 In the @code{Debug} project, we set the switches for two tools,
12127 @command{gnatmake} and the compiler, and thus we include the two corresponding
12128 packages; each package defines the @code{^Default_Switches^Default_Switches^}
12129 attribute with index @code{"Ada"}.
12130 Note that the package corresponding to
12131 @command{gnatmake} is named @code{Builder}. The @code{Release} project is
12132 similar, but only includes the @code{Compiler} package.
12134 In project @code{Debug} above, the ^switches^switches^ starting with
12135 @option{-gnat} that are specified in package @code{Compiler}
12136 could have been placed in package @code{Builder}, since @command{gnatmake}
12137 transmits all such ^switches^switches^ to the compiler.
12139 @node Main Subprograms
12140 @unnumberedsubsubsec Main Subprograms
12143 One of the specifiable properties of a project is a list of files that contain
12144 main subprograms. This property is captured in the @code{Main} attribute,
12145 whose value is a list of strings. If a project defines the @code{Main}
12146 attribute, it is not necessary to identify the main subprogram(s) when
12147 invoking @command{gnatmake} (@pxref{gnatmake and Project Files}).
12149 @node Executable File Names
12150 @unnumberedsubsubsec Executable File Names
12153 By default, the executable file name corresponding to a main source is
12154 deduced from the main source file name. Through the attributes
12155 @code{Executable} and @code{Executable_Suffix} of package @code{Builder},
12156 it is possible to change this default.
12157 In project @code{Debug} above, the executable file name
12158 for main source @file{^proc.adb^PROC.ADB^} is
12159 @file{^proc1^PROC1.EXE^}.
12160 Attribute @code{Executable_Suffix}, when specified, may change the suffix
12161 of the executable files, when no attribute @code{Executable} applies:
12162 its value replace the platform-specific executable suffix.
12163 Attributes @code{Executable} and @code{Executable_Suffix} are the only ways to
12164 specify a non-default executable file name when several mains are built at once
12165 in a single @command{gnatmake} command.
12167 @node Source File Naming Conventions
12168 @unnumberedsubsubsec Source File Naming Conventions
12171 Since the project files above do not specify any source file naming
12172 conventions, the GNAT defaults are used. The mechanism for defining source
12173 file naming conventions -- a package named @code{Naming} --
12174 is described below (@pxref{Naming Schemes}).
12176 @node Source Language(s)
12177 @unnumberedsubsubsec Source Language(s)
12180 Since the project files do not specify a @code{Languages} attribute, by
12181 default the GNAT tools assume that the language of the project file is Ada.
12182 More generally, a project can comprise source files
12183 in Ada, C, and/or other languages.
12185 @node Using External Variables
12186 @subsection Using External Variables
12189 Instead of supplying different project files for debug and release, we can
12190 define a single project file that queries an external variable (set either
12191 on the command line or via an ^environment variable^logical name^) in order to
12192 conditionally define the appropriate settings. Again, assume that the
12193 source files @file{pack.ads}, @file{pack.adb}, and @file{proc.adb} are
12194 located in directory @file{^/common^[COMMON]^}. The following project file,
12195 @file{build.gpr}, queries the external variable named @code{STYLE} and
12196 defines an object directory and ^switch^switch^ settings based on whether
12197 the value is @code{"deb"} (debug) or @code{"rel"} (release), and where
12198 the default is @code{"deb"}.
12200 @smallexample @c projectfile
12203 for Main use ("proc");
12205 type Style_Type is ("deb", "rel");
12206 Style : Style_Type := external ("STYLE", "deb");
12210 for Object_Dir use "debug";
12213 for Object_Dir use "release";
12214 for Exec_Dir use ".";
12223 for ^Default_Switches^Default_Switches^ ("Ada")
12225 for Executable ("proc") use "proc1";
12234 package Compiler is
12238 for ^Default_Switches^Default_Switches^ ("Ada")
12239 use ("^-gnata^-gnata^",
12241 "^-gnatE^-gnatE^");
12244 for ^Default_Switches^Default_Switches^ ("Ada")
12255 @code{Style_Type} is an example of a @emph{string type}, which is the project
12256 file analog of an Ada enumeration type but whose components are string literals
12257 rather than identifiers. @code{Style} is declared as a variable of this type.
12259 The form @code{external("STYLE", "deb")} is known as an
12260 @emph{external reference}; its first argument is the name of an
12261 @emph{external variable}, and the second argument is a default value to be
12262 used if the external variable doesn't exist. You can define an external
12263 variable on the command line via the @option{^-X^/EXTERNAL_REFERENCE^} switch,
12264 or you can use ^an environment variable^a logical name^
12265 as an external variable.
12267 Each @code{case} construct is expanded by the Project Manager based on the
12268 value of @code{Style}. Thus the command
12271 gnatmake -P/common/build.gpr -XSTYLE=deb
12277 gnatmake /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=deb
12282 is equivalent to the @command{gnatmake} invocation using the project file
12283 @file{debug.gpr} in the earlier example. So is the command
12285 gnatmake ^-P/common/build.gpr^/PROJECT_FILE=[COMMON]BUILD.GPR^
12289 since @code{"deb"} is the default for @code{STYLE}.
12295 gnatmake -P/common/build.gpr -XSTYLE=rel
12301 GNAT MAKE /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=rel
12306 is equivalent to the @command{gnatmake} invocation using the project file
12307 @file{release.gpr} in the earlier example.
12309 @node Importing Other Projects
12310 @subsection Importing Other Projects
12311 @cindex @code{ADA_PROJECT_PATH}
12312 @cindex @code{GPR_PROJECT_PATH}
12315 A compilation unit in a source file in one project may depend on compilation
12316 units in source files in other projects. To compile this unit under
12317 control of a project file, the
12318 dependent project must @emph{import} the projects containing the needed source
12320 This effect is obtained using syntax similar to an Ada @code{with} clause,
12321 but where @code{with}ed entities are strings that denote project files.
12323 As an example, suppose that the two projects @code{GUI_Proj} and
12324 @code{Comm_Proj} are defined in the project files @file{gui_proj.gpr} and
12325 @file{comm_proj.gpr} in directories @file{^/gui^[GUI]^}
12326 and @file{^/comm^[COMM]^}, respectively.
12327 Suppose that the source files for @code{GUI_Proj} are
12328 @file{gui.ads} and @file{gui.adb}, and that the source files for
12329 @code{Comm_Proj} are @file{comm.ads} and @file{comm.adb}, where each set of
12330 files is located in its respective project file directory. Schematically:
12349 We want to develop an application in directory @file{^/app^[APP]^} that
12350 @code{with} the packages @code{GUI} and @code{Comm}, using the properties of
12351 the corresponding project files (e.g.@: the ^switch^switch^ settings
12352 and object directory).
12353 Skeletal code for a main procedure might be something like the following:
12355 @smallexample @c ada
12358 procedure App_Main is
12367 Here is a project file, @file{app_proj.gpr}, that achieves the desired
12370 @smallexample @c projectfile
12372 with "/gui/gui_proj", "/comm/comm_proj";
12373 project App_Proj is
12374 for Main use ("app_main");
12380 Building an executable is achieved through the command:
12382 gnatmake ^-P/app/app_proj^/PROJECT_FILE=[APP]APP_PROJ^
12385 which will generate the @code{^app_main^APP_MAIN.EXE^} executable
12386 in the directory where @file{app_proj.gpr} resides.
12388 If an imported project file uses the standard extension (@code{^gpr^GPR^}) then
12389 (as illustrated above) the @code{with} clause can omit the extension.
12391 Our example specified an absolute path for each imported project file.
12392 Alternatively, the directory name of an imported object can be omitted
12396 The imported project file is in the same directory as the importing project
12399 You have defined one or two ^environment variables^logical names^
12400 that includes the directory containing
12401 the needed project file. The syntax of @code{GPR_PROJECT_PATH} and
12402 @code{ADA_PROJECT_PATH} is the same as
12403 the syntax of @code{ADA_INCLUDE_PATH} and @code{ADA_OBJECTS_PATH}: a list of
12404 directory names separated by colons (semicolons on Windows).
12408 Thus, if we define @code{ADA_PROJECT_PATH} or @code{GPR_PROJECT_PATH}
12409 to include @file{^/gui^[GUI]^} and
12410 @file{^/comm^[COMM]^}, then our project file @file{app_proj.gpr} can be written
12413 @smallexample @c projectfile
12415 with "gui_proj", "comm_proj";
12416 project App_Proj is
12417 for Main use ("app_main");
12423 Importing other projects can create ambiguities.
12424 For example, the same unit might be present in different imported projects, or
12425 it might be present in both the importing project and in an imported project.
12426 Both of these conditions are errors. Note that in the current version of
12427 the Project Manager, it is illegal to have an ambiguous unit even if the
12428 unit is never referenced by the importing project. This restriction may be
12429 relaxed in a future release.
12431 @node Extending a Project
12432 @subsection Extending a Project
12435 In large software systems it is common to have multiple
12436 implementations of a common interface; in Ada terms, multiple versions of a
12437 package body for the same spec. For example, one implementation
12438 might be safe for use in tasking programs, while another might only be used
12439 in sequential applications. This can be modeled in GNAT using the concept
12440 of @emph{project extension}. If one project (the ``child'') @emph{extends}
12441 another project (the ``parent'') then by default all source files of the
12442 parent project are inherited by the child, but the child project can
12443 override any of the parent's source files with new versions, and can also
12444 add new files. This facility is the project analog of a type extension in
12445 Object-Oriented Programming. Project hierarchies are permitted (a child
12446 project may be the parent of yet another project), and a project that
12447 inherits one project can also import other projects.
12449 As an example, suppose that directory @file{^/seq^[SEQ]^} contains the project
12450 file @file{seq_proj.gpr} as well as the source files @file{pack.ads},
12451 @file{pack.adb}, and @file{proc.adb}:
12464 Note that the project file can simply be empty (that is, no attribute or
12465 package is defined):
12467 @smallexample @c projectfile
12469 project Seq_Proj is
12475 implying that its source files are all the Ada source files in the project
12478 Suppose we want to supply an alternate version of @file{pack.adb}, in
12479 directory @file{^/tasking^[TASKING]^}, but use the existing versions of
12480 @file{pack.ads} and @file{proc.adb}. We can define a project
12481 @code{Tasking_Proj} that inherits @code{Seq_Proj}:
12485 ^/tasking^[TASKING]^
12491 project Tasking_Proj extends "/seq/seq_proj" is
12497 The version of @file{pack.adb} used in a build depends on which project file
12500 Note that we could have obtained the desired behavior using project import
12501 rather than project inheritance; a @code{base} project would contain the
12502 sources for @file{pack.ads} and @file{proc.adb}, a sequential project would
12503 import @code{base} and add @file{pack.adb}, and likewise a tasking project
12504 would import @code{base} and add a different version of @file{pack.adb}. The
12505 choice depends on whether other sources in the original project need to be
12506 overridden. If they do, then project extension is necessary, otherwise,
12507 importing is sufficient.
12510 In a project file that extends another project file, it is possible to
12511 indicate that an inherited source is not part of the sources of the extending
12512 project. This is necessary sometimes when a package spec has been overloaded
12513 and no longer requires a body: in this case, it is necessary to indicate that
12514 the inherited body is not part of the sources of the project, otherwise there
12515 will be a compilation error when compiling the spec.
12517 For that purpose, the attribute @code{Excluded_Source_Files} is used.
12518 Its value is a string list: a list of file names. It is also possible to use
12519 attribute @code{Excluded_Source_List_File}. Its value is a single string:
12520 the file name of a text file containing a list of file names, one per line.
12522 @smallexample @c @projectfile
12523 project B extends "a" is
12524 for Source_Files use ("pkg.ads");
12525 -- New spec of Pkg does not need a completion
12526 for Excluded_Source_Files use ("pkg.adb");
12530 Attribute @code{Excluded_Source_Files} may also be used to check if a source
12531 is still needed: if it is possible to build using @command{gnatmake} when such
12532 a source is put in attribute @code{Excluded_Source_Files} of a project P, then
12533 it is possible to remove the source completely from a system that includes
12536 @c ***********************
12537 @c * Project File Syntax *
12538 @c ***********************
12540 @node Project File Syntax
12541 @section Project File Syntax
12545 * Qualified Projects::
12551 * Associative Array Attributes::
12552 * case Constructions::
12556 This section describes the structure of project files.
12558 A project may be an @emph{independent project}, entirely defined by a single
12559 project file. Any Ada source file in an independent project depends only
12560 on the predefined library and other Ada source files in the same project.
12563 A project may also @dfn{depend on} other projects, in either or both of
12564 the following ways:
12566 @item It may import any number of projects
12567 @item It may extend at most one other project
12571 The dependence relation is a directed acyclic graph (the subgraph reflecting
12572 the ``extends'' relation is a tree).
12574 A project's @dfn{immediate sources} are the source files directly defined by
12575 that project, either implicitly by residing in the project file's directory,
12576 or explicitly through any of the source-related attributes described below.
12577 More generally, a project @var{proj}'s @dfn{sources} are the immediate sources
12578 of @var{proj} together with the immediate sources (unless overridden) of any
12579 project on which @var{proj} depends (either directly or indirectly).
12582 @subsection Basic Syntax
12585 As seen in the earlier examples, project files have an Ada-like syntax.
12586 The minimal project file is:
12587 @smallexample @c projectfile
12596 The identifier @code{Empty} is the name of the project.
12597 This project name must be present after the reserved
12598 word @code{end} at the end of the project file, followed by a semi-colon.
12600 Any name in a project file, such as the project name or a variable name,
12601 has the same syntax as an Ada identifier.
12603 The reserved words of project files are the Ada 95 reserved words plus
12604 @code{extends}, @code{external}, and @code{project}. Note that the only Ada
12605 reserved words currently used in project file syntax are:
12641 Comments in project files have the same syntax as in Ada, two consecutive
12642 hyphens through the end of the line.
12644 @node Qualified Projects
12645 @subsection Qualified Projects
12648 Before the reserved @code{project}, there may be one or two "qualifiers", that
12649 is identifiers or other reserved words, to qualify the project.
12651 The current list of qualifiers is:
12655 @code{abstract}: qualify a project with no sources. A qualified abstract
12656 project must either have no declaration of attributes @code{Source_Dirs},
12657 @code{Source_Files}, @code{Languages} or @code{Source_List_File}, or one of
12658 @code{Source_Dirs}, @code{Source_Files}, or @code{Languages} must be declared
12659 as empty. If it extends another project, the project it extends must also be a
12660 qualified abstract project.
12663 @code{standard}: a standard project is a non library project with sources.
12666 @code{aggregate}: for future extension
12669 @code{aggregate library}: for future extension
12672 @code{library}: a library project must declare both attributes
12673 @code{Library_Name} and @code{Library_Dir}.
12676 @code{configuration}: a configuration project cannot be in a project tree.
12680 @subsection Packages
12683 A project file may contain @emph{packages}. The name of a package must be one
12684 of the identifiers from the following list. A package
12685 with a given name may only appear once in a project file. Package names are
12686 case insensitive. The following package names are legal:
12702 @code{Cross_Reference}
12708 @code{Pretty_Printer}
12718 @code{Language_Processing}
12722 In its simplest form, a package may be empty:
12724 @smallexample @c projectfile
12734 A package may contain @emph{attribute declarations},
12735 @emph{variable declarations} and @emph{case constructions}, as will be
12738 When there is ambiguity between a project name and a package name,
12739 the name always designates the project. To avoid possible confusion, it is
12740 always a good idea to avoid naming a project with one of the
12741 names allowed for packages or any name that starts with @code{gnat}.
12744 @subsection Expressions
12747 An @emph{expression} is either a @emph{string expression} or a
12748 @emph{string list expression}.
12750 A @emph{string expression} is either a @emph{simple string expression} or a
12751 @emph{compound string expression}.
12753 A @emph{simple string expression} is one of the following:
12755 @item A literal string; e.g.@: @code{"comm/my_proj.gpr"}
12756 @item A string-valued variable reference (@pxref{Variables})
12757 @item A string-valued attribute reference (@pxref{Attributes})
12758 @item An external reference (@pxref{External References in Project Files})
12762 A @emph{compound string expression} is a concatenation of string expressions,
12763 using the operator @code{"&"}
12765 Path & "/" & File_Name & ".ads"
12769 A @emph{string list expression} is either a
12770 @emph{simple string list expression} or a
12771 @emph{compound string list expression}.
12773 A @emph{simple string list expression} is one of the following:
12775 @item A parenthesized list of zero or more string expressions,
12776 separated by commas
12778 File_Names := (File_Name, "gnat.adc", File_Name & ".orig");
12781 @item A string list-valued variable reference
12782 @item A string list-valued attribute reference
12786 A @emph{compound string list expression} is the concatenation (using
12787 @code{"&"}) of a simple string list expression and an expression. Note that
12788 each term in a compound string list expression, except the first, may be
12789 either a string expression or a string list expression.
12791 @smallexample @c projectfile
12793 File_Name_List := () & File_Name; -- One string in this list
12794 Extended_File_Name_List := File_Name_List & (File_Name & ".orig");
12796 Big_List := File_Name_List & Extended_File_Name_List;
12797 -- Concatenation of two string lists: three strings
12798 Illegal_List := "gnat.adc" & Extended_File_Name_List;
12799 -- Illegal: must start with a string list
12804 @subsection String Types
12807 A @emph{string type declaration} introduces a discrete set of string literals.
12808 If a string variable is declared to have this type, its value
12809 is restricted to the given set of literals.
12811 Here is an example of a string type declaration:
12813 @smallexample @c projectfile
12814 type OS is ("NT", "nt", "Unix", "GNU/Linux", "other OS");
12818 Variables of a string type are called @emph{typed variables}; all other
12819 variables are called @emph{untyped variables}. Typed variables are
12820 particularly useful in @code{case} constructions, to support conditional
12821 attribute declarations.
12822 (@pxref{case Constructions}).
12824 The string literals in the list are case sensitive and must all be different.
12825 They may include any graphic characters allowed in Ada, including spaces.
12827 A string type may only be declared at the project level, not inside a package.
12829 A string type may be referenced by its name if it has been declared in the same
12830 project file, or by an expanded name whose prefix is the name of the project
12831 in which it is declared.
12834 @subsection Variables
12837 A variable may be declared at the project file level, or within a package.
12838 Here are some examples of variable declarations:
12840 @smallexample @c projectfile
12842 This_OS : OS := external ("OS"); -- a typed variable declaration
12843 That_OS := "GNU/Linux"; -- an untyped variable declaration
12848 The syntax of a @emph{typed variable declaration} is identical to the Ada
12849 syntax for an object declaration. By contrast, the syntax of an untyped
12850 variable declaration is identical to an Ada assignment statement. In fact,
12851 variable declarations in project files have some of the characteristics of
12852 an assignment, in that successive declarations for the same variable are
12853 allowed. Untyped variable declarations do establish the expected kind of the
12854 variable (string or string list), and successive declarations for it must
12855 respect the initial kind.
12858 A string variable declaration (typed or untyped) declares a variable
12859 whose value is a string. This variable may be used as a string expression.
12860 @smallexample @c projectfile
12861 File_Name := "readme.txt";
12862 Saved_File_Name := File_Name & ".saved";
12866 A string list variable declaration declares a variable whose value is a list
12867 of strings. The list may contain any number (zero or more) of strings.
12869 @smallexample @c projectfile
12871 List_With_One_Element := ("^-gnaty^-gnaty^");
12872 List_With_Two_Elements := List_With_One_Element & "^-gnatg^-gnatg^";
12873 Long_List := ("main.ada", "pack1_.ada", "pack1.ada", "pack2_.ada"
12874 "pack2.ada", "util_.ada", "util.ada");
12878 The same typed variable may not be declared more than once at project level,
12879 and it may not be declared more than once in any package; it is in effect
12882 The same untyped variable may be declared several times. Declarations are
12883 elaborated in the order in which they appear, so the new value replaces
12884 the old one, and any subsequent reference to the variable uses the new value.
12885 However, as noted above, if a variable has been declared as a string, all
12887 declarations must give it a string value. Similarly, if a variable has
12888 been declared as a string list, all subsequent declarations
12889 must give it a string list value.
12891 A @emph{variable reference} may take several forms:
12894 @item The simple variable name, for a variable in the current package (if any)
12895 or in the current project
12896 @item An expanded name, whose prefix is a context name.
12900 A @emph{context} may be one of the following:
12903 @item The name of an existing package in the current project
12904 @item The name of an imported project of the current project
12905 @item The name of an ancestor project (i.e., a project extended by the current
12906 project, either directly or indirectly)
12907 @item An expanded name whose prefix is an imported/parent project name, and
12908 whose selector is a package name in that project.
12912 A variable reference may be used in an expression.
12915 @subsection Attributes
12918 A project (and its packages) may have @emph{attributes} that define
12919 the project's properties. Some attributes have values that are strings;
12920 others have values that are string lists.
12922 There are two categories of attributes: @emph{simple attributes}
12923 and @emph{associative arrays} (@pxref{Associative Array Attributes}).
12925 Legal project attribute names, and attribute names for each legal package are
12926 listed below. Attributes names are case-insensitive.
12928 The following attributes are defined on projects (all are simple attributes):
12930 @multitable @columnfractions .4 .3
12931 @item @emph{Attribute Name}
12933 @item @code{Source_Files}
12935 @item @code{Source_Dirs}
12937 @item @code{Source_List_File}
12939 @item @code{Object_Dir}
12941 @item @code{Exec_Dir}
12943 @item @code{Excluded_Source_Dirs}
12945 @item @code{Excluded_Source_Files}
12947 @item @code{Excluded_Source_List_File}
12949 @item @code{Languages}
12953 @item @code{Library_Dir}
12955 @item @code{Library_Name}
12957 @item @code{Library_Kind}
12959 @item @code{Library_Version}
12961 @item @code{Library_Interface}
12963 @item @code{Library_Auto_Init}
12965 @item @code{Library_Options}
12967 @item @code{Library_Src_Dir}
12969 @item @code{Library_ALI_Dir}
12971 @item @code{Library_GCC}
12973 @item @code{Library_Symbol_File}
12975 @item @code{Library_Symbol_Policy}
12977 @item @code{Library_Reference_Symbol_File}
12979 @item @code{Externally_Built}
12984 The following attributes are defined for package @code{Naming}
12985 (@pxref{Naming Schemes}):
12987 @multitable @columnfractions .4 .2 .2 .2
12988 @item Attribute Name @tab Category @tab Index @tab Value
12989 @item @code{Spec_Suffix}
12990 @tab associative array
12993 @item @code{Body_Suffix}
12994 @tab associative array
12997 @item @code{Separate_Suffix}
12998 @tab simple attribute
13001 @item @code{Casing}
13002 @tab simple attribute
13005 @item @code{Dot_Replacement}
13006 @tab simple attribute
13010 @tab associative array
13014 @tab associative array
13017 @item @code{Specification_Exceptions}
13018 @tab associative array
13021 @item @code{Implementation_Exceptions}
13022 @tab associative array
13028 The following attributes are defined for packages @code{Builder},
13029 @code{Compiler}, @code{Binder},
13030 @code{Linker}, @code{Cross_Reference}, and @code{Finder}
13031 (@pxref{^Switches^Switches^ and Project Files}).
13033 @multitable @columnfractions .4 .2 .2 .2
13034 @item Attribute Name @tab Category @tab Index @tab Value
13035 @item @code{^Default_Switches^Default_Switches^}
13036 @tab associative array
13039 @item @code{^Switches^Switches^}
13040 @tab associative array
13046 In addition, package @code{Compiler} has a single string attribute
13047 @code{Local_Configuration_Pragmas} and package @code{Builder} has a single
13048 string attribute @code{Global_Configuration_Pragmas}.
13051 Each simple attribute has a default value: the empty string (for string-valued
13052 attributes) and the empty list (for string list-valued attributes).
13054 An attribute declaration defines a new value for an attribute.
13056 Examples of simple attribute declarations:
13058 @smallexample @c projectfile
13059 for Object_Dir use "objects";
13060 for Source_Dirs use ("units", "test/drivers");
13064 The syntax of a @dfn{simple attribute declaration} is similar to that of an
13065 attribute definition clause in Ada.
13067 Attributes references may be appear in expressions.
13068 The general form for such a reference is @code{<entity>'<attribute>}:
13069 Associative array attributes are functions. Associative
13070 array attribute references must have an argument that is a string literal.
13074 @smallexample @c projectfile
13076 Naming'Dot_Replacement
13077 Imported_Project'Source_Dirs
13078 Imported_Project.Naming'Casing
13079 Builder'^Default_Switches^Default_Switches^("Ada")
13083 The prefix of an attribute may be:
13085 @item @code{project} for an attribute of the current project
13086 @item The name of an existing package of the current project
13087 @item The name of an imported project
13088 @item The name of a parent project that is extended by the current project
13089 @item An expanded name whose prefix is imported/parent project name,
13090 and whose selector is a package name
13095 @smallexample @c projectfile
13098 for Source_Dirs use project'Source_Dirs & "units";
13099 for Source_Dirs use project'Source_Dirs & "test/drivers"
13105 In the first attribute declaration, initially the attribute @code{Source_Dirs}
13106 has the default value: an empty string list. After this declaration,
13107 @code{Source_Dirs} is a string list of one element: @code{"units"}.
13108 After the second attribute declaration @code{Source_Dirs} is a string list of
13109 two elements: @code{"units"} and @code{"test/drivers"}.
13111 Note: this example is for illustration only. In practice,
13112 the project file would contain only one attribute declaration:
13114 @smallexample @c projectfile
13115 for Source_Dirs use ("units", "test/drivers");
13118 @node Associative Array Attributes
13119 @subsection Associative Array Attributes
13122 Some attributes are defined as @emph{associative arrays}. An associative
13123 array may be regarded as a function that takes a string as a parameter
13124 and delivers a string or string list value as its result.
13126 Here are some examples of single associative array attribute associations:
13128 @smallexample @c projectfile
13129 for Body ("main") use "Main.ada";
13130 for ^Switches^Switches^ ("main.ada")
13132 "^-gnatv^-gnatv^");
13133 for ^Switches^Switches^ ("main.ada")
13134 use Builder'^Switches^Switches^ ("main.ada")
13139 Like untyped variables and simple attributes, associative array attributes
13140 may be declared several times. Each declaration supplies a new value for the
13141 attribute, and replaces the previous setting.
13144 An associative array attribute may be declared as a full associative array
13145 declaration, with the value of the same attribute in an imported or extended
13148 @smallexample @c projectfile
13150 for Default_Switches use Default.Builder'Default_Switches;
13155 In this example, @code{Default} must be either a project imported by the
13156 current project, or the project that the current project extends. If the
13157 attribute is in a package (in this case, in package @code{Builder}), the same
13158 package needs to be specified.
13161 A full associative array declaration replaces any other declaration for the
13162 attribute, including other full associative array declaration. Single
13163 associative array associations may be declare after a full associative
13164 declaration, modifying the value for a single association of the attribute.
13166 @node case Constructions
13167 @subsection @code{case} Constructions
13170 A @code{case} construction is used in a project file to effect conditional
13172 Here is a typical example:
13174 @smallexample @c projectfile
13177 type OS_Type is ("GNU/Linux", "Unix", "NT", "VMS");
13179 OS : OS_Type := external ("OS", "GNU/Linux");
13183 package Compiler is
13185 when "GNU/Linux" | "Unix" =>
13186 for ^Default_Switches^Default_Switches^ ("Ada")
13187 use ("^-gnath^-gnath^");
13189 for ^Default_Switches^Default_Switches^ ("Ada")
13190 use ("^-gnatP^-gnatP^");
13199 The syntax of a @code{case} construction is based on the Ada case statement
13200 (although there is no @code{null} construction for empty alternatives).
13202 The case expression must be a typed string variable.
13203 Each alternative comprises the reserved word @code{when}, either a list of
13204 literal strings separated by the @code{"|"} character or the reserved word
13205 @code{others}, and the @code{"=>"} token.
13206 Each literal string must belong to the string type that is the type of the
13208 An @code{others} alternative, if present, must occur last.
13210 After each @code{=>}, there are zero or more constructions. The only
13211 constructions allowed in a case construction are other case constructions,
13212 attribute declarations and variable declarations. String type declarations and
13213 package declarations are not allowed. Variable declarations are restricted to
13214 variables that have already been declared before the case construction.
13216 The value of the case variable is often given by an external reference
13217 (@pxref{External References in Project Files}).
13219 @c ****************************************
13220 @c * Objects and Sources in Project Files *
13221 @c ****************************************
13223 @node Objects and Sources in Project Files
13224 @section Objects and Sources in Project Files
13227 * Object Directory::
13229 * Source Directories::
13230 * Source File Names::
13234 Each project has exactly one object directory and one or more source
13235 directories. The source directories must contain at least one source file,
13236 unless the project file explicitly specifies that no source files are present
13237 (@pxref{Source File Names}).
13239 @node Object Directory
13240 @subsection Object Directory
13243 The object directory for a project is the directory containing the compiler's
13244 output (such as @file{ALI} files and object files) for the project's immediate
13247 The object directory is given by the value of the attribute @code{Object_Dir}
13248 in the project file.
13250 @smallexample @c projectfile
13251 for Object_Dir use "objects";
13255 The attribute @code{Object_Dir} has a string value, the path name of the object
13256 directory. The path name may be absolute or relative to the directory of the
13257 project file. This directory must already exist, and be readable and writable.
13259 By default, when the attribute @code{Object_Dir} is not given an explicit value
13260 or when its value is the empty string, the object directory is the same as the
13261 directory containing the project file.
13263 @node Exec Directory
13264 @subsection Exec Directory
13267 The exec directory for a project is the directory containing the executables
13268 for the project's main subprograms.
13270 The exec directory is given by the value of the attribute @code{Exec_Dir}
13271 in the project file.
13273 @smallexample @c projectfile
13274 for Exec_Dir use "executables";
13278 The attribute @code{Exec_Dir} has a string value, the path name of the exec
13279 directory. The path name may be absolute or relative to the directory of the
13280 project file. This directory must already exist, and be writable.
13282 By default, when the attribute @code{Exec_Dir} is not given an explicit value
13283 or when its value is the empty string, the exec directory is the same as the
13284 object directory of the project file.
13286 @node Source Directories
13287 @subsection Source Directories
13290 The source directories of a project are specified by the project file
13291 attribute @code{Source_Dirs}.
13293 This attribute's value is a string list. If the attribute is not given an
13294 explicit value, then there is only one source directory, the one where the
13295 project file resides.
13297 A @code{Source_Dirs} attribute that is explicitly defined to be the empty list,
13300 @smallexample @c projectfile
13301 for Source_Dirs use ();
13305 indicates that the project contains no source files.
13307 Otherwise, each string in the string list designates one or more
13308 source directories.
13310 @smallexample @c projectfile
13311 for Source_Dirs use ("sources", "test/drivers");
13315 If a string in the list ends with @code{"/**"}, then the directory whose path
13316 name precedes the two asterisks, as well as all its subdirectories
13317 (recursively), are source directories.
13319 @smallexample @c projectfile
13320 for Source_Dirs use ("/system/sources/**");
13324 Here the directory @code{/system/sources} and all of its subdirectories
13325 (recursively) are source directories.
13327 To specify that the source directories are the directory of the project file
13328 and all of its subdirectories, you can declare @code{Source_Dirs} as follows:
13329 @smallexample @c projectfile
13330 for Source_Dirs use ("./**");
13334 Each of the source directories must exist and be readable.
13336 @node Source File Names
13337 @subsection Source File Names
13340 In a project that contains source files, their names may be specified by the
13341 attributes @code{Source_Files} (a string list) or @code{Source_List_File}
13342 (a string). Source file names never include any directory information.
13344 If the attribute @code{Source_Files} is given an explicit value, then each
13345 element of the list is a source file name.
13347 @smallexample @c projectfile
13348 for Source_Files use ("main.adb");
13349 for Source_Files use ("main.adb", "pack1.ads", "pack2.adb");
13353 If the attribute @code{Source_Files} is not given an explicit value,
13354 but the attribute @code{Source_List_File} is given a string value,
13355 then the source file names are contained in the text file whose path name
13356 (absolute or relative to the directory of the project file) is the
13357 value of the attribute @code{Source_List_File}.
13359 Each line in the file that is not empty or is not a comment
13360 contains a source file name.
13362 @smallexample @c projectfile
13363 for Source_List_File use "source_list.txt";
13367 By default, if neither the attribute @code{Source_Files} nor the attribute
13368 @code{Source_List_File} is given an explicit value, then each file in the
13369 source directories that conforms to the project's naming scheme
13370 (@pxref{Naming Schemes}) is an immediate source of the project.
13372 A warning is issued if both attributes @code{Source_Files} and
13373 @code{Source_List_File} are given explicit values. In this case, the attribute
13374 @code{Source_Files} prevails.
13376 Each source file name must be the name of one existing source file
13377 in one of the source directories.
13379 A @code{Source_Files} attribute whose value is an empty list
13380 indicates that there are no source files in the project.
13382 If the order of the source directories is known statically, that is if
13383 @code{"/**"} is not used in the string list @code{Source_Dirs}, then there may
13384 be several files with the same source file name. In this case, only the file
13385 in the first directory is considered as an immediate source of the project
13386 file. If the order of the source directories is not known statically, it is
13387 an error to have several files with the same source file name.
13389 Projects can be specified to have no Ada source
13390 files: the value of @code{Source_Dirs} or @code{Source_Files} may be an empty
13391 list, or the @code{"Ada"} may be absent from @code{Languages}:
13393 @smallexample @c projectfile
13394 for Source_Dirs use ();
13395 for Source_Files use ();
13396 for Languages use ("C", "C++");
13400 Otherwise, a project must contain at least one immediate source.
13402 Projects with no source files are useful as template packages
13403 (@pxref{Packages in Project Files}) for other projects; in particular to
13404 define a package @code{Naming} (@pxref{Naming Schemes}).
13406 @c ****************************
13407 @c * Importing Projects *
13408 @c ****************************
13410 @node Importing Projects
13411 @section Importing Projects
13412 @cindex @code{ADA_PROJECT_PATH}
13413 @cindex @code{GPR_PROJECT_PATH}
13416 An immediate source of a project P may depend on source files that
13417 are neither immediate sources of P nor in the predefined library.
13418 To get this effect, P must @emph{import} the projects that contain the needed
13421 @smallexample @c projectfile
13423 with "project1", "utilities.gpr";
13424 with "/namings/apex.gpr";
13431 As can be seen in this example, the syntax for importing projects is similar
13432 to the syntax for importing compilation units in Ada. However, project files
13433 use literal strings instead of names, and the @code{with} clause identifies
13434 project files rather than packages.
13436 Each literal string is the file name or path name (absolute or relative) of a
13437 project file. If a string corresponds to a file name, with no path or a
13438 relative path, then its location is determined by the @emph{project path}. The
13439 latter can be queried using @code{gnatls -v}. It contains:
13443 In first position, the directory containing the current project file.
13445 In last position, the default project directory. This default project directory
13446 is part of the GNAT installation and is the standard place to install project
13447 files giving access to standard support libraries.
13449 @ref{Installing a library}
13453 In between, all the directories referenced in the
13454 ^environment variables^logical names^ @env{GPR_PROJECT_PATH}
13455 and @env{ADA_PROJECT_PATH} if they exist, and in that order.
13459 If a relative pathname is used, as in
13461 @smallexample @c projectfile
13466 then the full path for the project is constructed by concatenating this
13467 relative path to those in the project path, in order, until a matching file is
13468 found. Any symbolic link will be fully resolved in the directory of the
13469 importing project file before the imported project file is examined.
13471 If the @code{with}'ed project file name does not have an extension,
13472 the default is @file{^.gpr^.GPR^}. If a file with this extension is not found,
13473 then the file name as specified in the @code{with} clause (no extension) will
13474 be used. In the above example, if a file @code{project1.gpr} is found, then it
13475 will be used; otherwise, if a file @code{^project1^PROJECT1^} exists
13476 then it will be used; if neither file exists, this is an error.
13478 A warning is issued if the name of the project file does not match the
13479 name of the project; this check is case insensitive.
13481 Any source file that is an immediate source of the imported project can be
13482 used by the immediate sources of the importing project, transitively. Thus
13483 if @code{A} imports @code{B}, and @code{B} imports @code{C}, the immediate
13484 sources of @code{A} may depend on the immediate sources of @code{C}, even if
13485 @code{A} does not import @code{C} explicitly. However, this is not recommended,
13486 because if and when @code{B} ceases to import @code{C}, some sources in
13487 @code{A} will no longer compile.
13489 A side effect of this capability is that normally cyclic dependencies are not
13490 permitted: if @code{A} imports @code{B} (directly or indirectly) then @code{B}
13491 is not allowed to import @code{A}. However, there are cases when cyclic
13492 dependencies would be beneficial. For these cases, another form of import
13493 between projects exists, the @code{limited with}: a project @code{A} that
13494 imports a project @code{B} with a straight @code{with} may also be imported,
13495 directly or indirectly, by @code{B} on the condition that imports from @code{B}
13496 to @code{A} include at least one @code{limited with}.
13498 @smallexample @c 0projectfile
13504 limited with "../a/a.gpr";
13512 limited with "../a/a.gpr";
13518 In the above legal example, there are two project cycles:
13521 @item A -> C -> D -> A
13525 In each of these cycle there is one @code{limited with}: import of @code{A}
13526 from @code{B} and import of @code{A} from @code{D}.
13528 The difference between straight @code{with} and @code{limited with} is that
13529 the name of a project imported with a @code{limited with} cannot be used in the
13530 project that imports it. In particular, its packages cannot be renamed and
13531 its variables cannot be referred to.
13533 An exception to the above rules for @code{limited with} is that for the main
13534 project specified to @command{gnatmake} or to the @command{GNAT} driver a
13535 @code{limited with} is equivalent to a straight @code{with}. For example,
13536 in the example above, projects @code{B} and @code{D} could not be main
13537 projects for @command{gnatmake} or to the @command{GNAT} driver, because they
13538 each have a @code{limited with} that is the only one in a cycle of importing
13541 @c *********************
13542 @c * Project Extension *
13543 @c *********************
13545 @node Project Extension
13546 @section Project Extension
13549 During development of a large system, it is sometimes necessary to use
13550 modified versions of some of the source files, without changing the original
13551 sources. This can be achieved through the @emph{project extension} facility.
13553 @smallexample @c projectfile
13554 project Modified_Utilities extends "/baseline/utilities.gpr" is @dots{}
13558 A project extension declaration introduces an extending project
13559 (the @emph{child}) and a project being extended (the @emph{parent}).
13561 By default, a child project inherits all the sources of its parent.
13562 However, inherited sources can be overridden: a unit in a parent is hidden
13563 by a unit of the same name in the child.
13565 Inherited sources are considered to be sources (but not immediate sources)
13566 of the child project; see @ref{Project File Syntax}.
13568 An inherited source file retains any switches specified in the parent project.
13570 For example if the project @code{Utilities} contains the spec and the
13571 body of an Ada package @code{Util_IO}, then the project
13572 @code{Modified_Utilities} can contain a new body for package @code{Util_IO}.
13573 The original body of @code{Util_IO} will not be considered in program builds.
13574 However, the package spec will still be found in the project
13577 A child project can have only one parent, except when it is qualified as
13578 abstract. But it may import any number of other projects.
13580 A project is not allowed to import directly or indirectly at the same time a
13581 child project and any of its ancestors.
13583 @c *******************************
13584 @c * Project Hierarchy Extension *
13585 @c *******************************
13587 @node Project Hierarchy Extension
13588 @section Project Hierarchy Extension
13591 When extending a large system spanning multiple projects, it is often
13592 inconvenient to extend every project in the hierarchy that is impacted by a
13593 small change introduced. In such cases, it is possible to create a virtual
13594 extension of entire hierarchy using @code{extends all} relationship.
13596 When the project is extended using @code{extends all} inheritance, all projects
13597 that are imported by it, both directly and indirectly, are considered virtually
13598 extended. That is, the Project Manager creates "virtual projects"
13599 that extend every project in the hierarchy; all these virtual projects have
13600 no sources of their own and have as object directory the object directory of
13601 the root of "extending all" project.
13603 It is possible to explicitly extend one or more projects in the hierarchy
13604 in order to modify the sources. These extending projects must be imported by
13605 the "extending all" project, which will replace the corresponding virtual
13606 projects with the explicit ones.
13608 When building such a project hierarchy extension, the Project Manager will
13609 ensure that both modified sources and sources in virtual extending projects
13610 that depend on them, are recompiled.
13612 By means of example, consider the following hierarchy of projects.
13616 project A, containing package P1
13618 project B importing A and containing package P2 which depends on P1
13620 project C importing B and containing package P3 which depends on P2
13624 We want to modify packages P1 and P3.
13626 This project hierarchy will need to be extended as follows:
13630 Create project A1 that extends A, placing modified P1 there:
13632 @smallexample @c 0projectfile
13633 project A1 extends "(@dots{})/A" is
13638 Create project C1 that "extends all" C and imports A1, placing modified
13641 @smallexample @c 0projectfile
13642 with "(@dots{})/A1";
13643 project C1 extends all "(@dots{})/C" is
13648 When you build project C1, your entire modified project space will be
13649 recompiled, including the virtual project B1 that has been impacted by the
13650 "extending all" inheritance of project C.
13652 Note that if a Library Project in the hierarchy is virtually extended,
13653 the virtual project that extends the Library Project is not a Library Project.
13655 @c ****************************************
13656 @c * External References in Project Files *
13657 @c ****************************************
13659 @node External References in Project Files
13660 @section External References in Project Files
13663 A project file may contain references to external variables; such references
13664 are called @emph{external references}.
13666 An external variable is either defined as part of the environment (an
13667 environment variable in Unix, for example) or else specified on the command
13668 line via the @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
13669 If both, then the command line value is used.
13671 The value of an external reference is obtained by means of the built-in
13672 function @code{external}, which returns a string value.
13673 This function has two forms:
13675 @item @code{external (external_variable_name)}
13676 @item @code{external (external_variable_name, default_value)}
13680 Each parameter must be a string literal. For example:
13682 @smallexample @c projectfile
13684 external ("OS", "GNU/Linux")
13688 In the form with one parameter, the function returns the value of
13689 the external variable given as parameter. If this name is not present in the
13690 environment, the function returns an empty string.
13692 In the form with two string parameters, the second argument is
13693 the value returned when the variable given as the first argument is not
13694 present in the environment. In the example above, if @code{"OS"} is not
13695 the name of ^an environment variable^a logical name^ and is not passed on
13696 the command line, then the returned value is @code{"GNU/Linux"}.
13698 An external reference may be part of a string expression or of a string
13699 list expression, and can therefore appear in a variable declaration or
13700 an attribute declaration.
13702 @smallexample @c projectfile
13704 type Mode_Type is ("Debug", "Release");
13705 Mode : Mode_Type := external ("MODE");
13712 @c *****************************
13713 @c * Packages in Project Files *
13714 @c *****************************
13716 @node Packages in Project Files
13717 @section Packages in Project Files
13720 A @emph{package} defines the settings for project-aware tools within a
13722 For each such tool one can declare a package; the names for these
13723 packages are preset (@pxref{Packages}).
13724 A package may contain variable declarations, attribute declarations, and case
13727 @smallexample @c projectfile
13730 package Builder is -- used by gnatmake
13731 for ^Default_Switches^Default_Switches^ ("Ada")
13740 The syntax of package declarations mimics that of package in Ada.
13742 Most of the packages have an attribute
13743 @code{^Default_Switches^Default_Switches^}.
13744 This attribute is an associative array, and its value is a string list.
13745 The index of the associative array is the name of a programming language (case
13746 insensitive). This attribute indicates the ^switch^switch^
13747 or ^switches^switches^ to be used
13748 with the corresponding tool.
13750 Some packages also have another attribute, @code{^Switches^Switches^},
13751 an associative array whose value is a string list.
13752 The index is the name of a source file.
13753 This attribute indicates the ^switch^switch^
13754 or ^switches^switches^ to be used by the corresponding
13755 tool when dealing with this specific file.
13757 Further information on these ^switch^switch^-related attributes is found in
13758 @ref{^Switches^Switches^ and Project Files}.
13760 A package may be declared as a @emph{renaming} of another package; e.g., from
13761 the project file for an imported project.
13763 @smallexample @c projectfile
13765 with "/global/apex.gpr";
13767 package Naming renames Apex.Naming;
13774 Packages that are renamed in other project files often come from project files
13775 that have no sources: they are just used as templates. Any modification in the
13776 template will be reflected automatically in all the project files that rename
13777 a package from the template.
13779 In addition to the tool-oriented packages, you can also declare a package
13780 named @code{Naming} to establish specialized source file naming conventions
13781 (@pxref{Naming Schemes}).
13783 @c ************************************
13784 @c * Variables from Imported Projects *
13785 @c ************************************
13787 @node Variables from Imported Projects
13788 @section Variables from Imported Projects
13791 An attribute or variable defined in an imported or parent project can
13792 be used in expressions in the importing / extending project.
13793 Such an attribute or variable is denoted by an expanded name whose prefix
13794 is either the name of the project or the expanded name of a package within
13797 @smallexample @c projectfile
13800 project Main extends "base" is
13801 Var1 := Imported.Var;
13802 Var2 := Base.Var & ".new";
13807 for ^Default_Switches^Default_Switches^ ("Ada")
13808 use Imported.Builder'Ada_^Switches^Switches^ &
13809 "^-gnatg^-gnatg^" &
13815 package Compiler is
13816 for ^Default_Switches^Default_Switches^ ("Ada")
13817 use Base.Compiler'Ada_^Switches^Switches^;
13828 The value of @code{Var1} is a copy of the variable @code{Var} defined
13829 in the project file @file{"imported.gpr"}
13831 the value of @code{Var2} is a copy of the value of variable @code{Var}
13832 defined in the project file @file{base.gpr}, concatenated with @code{".new"}
13834 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13835 @code{Builder} is a string list that includes in its value a copy of the value
13836 of @code{Ada_^Switches^Switches^} defined in the @code{Builder} package
13837 in project file @file{imported.gpr} plus two new elements:
13838 @option{"^-gnatg^-gnatg^"}
13839 and @option{"^-v^-v^"};
13841 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13842 @code{Compiler} is a copy of the variable @code{Ada_^Switches^Switches^}
13843 defined in the @code{Compiler} package in project file @file{base.gpr},
13844 the project being extended.
13847 @c ******************
13848 @c * Naming Schemes *
13849 @c ******************
13851 @node Naming Schemes
13852 @section Naming Schemes
13855 Sometimes an Ada software system is ported from a foreign compilation
13856 environment to GNAT, and the file names do not use the default GNAT
13857 conventions. Instead of changing all the file names (which for a variety
13858 of reasons might not be possible), you can define the relevant file
13859 naming scheme in the @code{Naming} package in your project file.
13862 Note that the use of pragmas described in
13863 @ref{Alternative File Naming Schemes} by mean of a configuration
13864 pragmas file is not supported when using project files. You must use
13865 the features described in this paragraph. You can however use specify
13866 other configuration pragmas (@pxref{Specifying Configuration Pragmas}).
13869 For example, the following
13870 package models the Apex file naming rules:
13872 @smallexample @c projectfile
13875 for Casing use "lowercase";
13876 for Dot_Replacement use ".";
13877 for Spec_Suffix ("Ada") use ".1.ada";
13878 for Body_Suffix ("Ada") use ".2.ada";
13885 For example, the following package models the HP Ada file naming rules:
13887 @smallexample @c projectfile
13890 for Casing use "lowercase";
13891 for Dot_Replacement use "__";
13892 for Spec_Suffix ("Ada") use "_.^ada^ada^";
13893 for Body_Suffix ("Ada") use ".^ada^ada^";
13899 (Note that @code{Casing} is @code{"lowercase"} because GNAT gets the file
13900 names in lower case)
13904 You can define the following attributes in package @code{Naming}:
13908 @item @code{Casing}
13909 This must be a string with one of the three values @code{"lowercase"},
13910 @code{"uppercase"} or @code{"mixedcase"}; these strings are case insensitive.
13913 If @code{Casing} is not specified, then the default is @code{"lowercase"}.
13915 @item @code{Dot_Replacement}
13916 This must be a string whose value satisfies the following conditions:
13919 @item It must not be empty
13920 @item It cannot start or end with an alphanumeric character
13921 @item It cannot be a single underscore
13922 @item It cannot start with an underscore followed by an alphanumeric
13923 @item It cannot contain a dot @code{'.'} except if the entire string
13928 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
13930 @item @code{Spec_Suffix}
13931 This is an associative array (indexed by the programming language name, case
13932 insensitive) whose value is a string that must satisfy the following
13936 @item It must not be empty
13937 @item It must include at least one dot
13940 If @code{Spec_Suffix ("Ada")} is not specified, then the default is
13941 @code{"^.ads^.ADS^"}.
13943 @item @code{Body_Suffix}
13944 This is an associative array (indexed by the programming language name, case
13945 insensitive) whose value is a string that must satisfy the following
13949 @item It must not be empty
13950 @item It must include at least one dot
13951 @item It cannot be the same as @code{Spec_Suffix ("Ada")}
13954 If @code{Body_Suffix ("Ada")} and @code{Spec_Suffix ("Ada")} end with the
13955 same string, then a file name that ends with the longest of these two suffixes
13956 will be a body if the longest suffix is @code{Body_Suffix ("Ada")} or a spec
13957 if the longest suffix is @code{Spec_Suffix ("Ada")}.
13959 If the suffix does not start with a '.', a file with a name exactly equal
13960 to the suffix will also be part of the project (for instance if you define
13961 the suffix as @code{Makefile}, a file called @file{Makefile} will be part
13962 of the project. This is not interesting in general when using projects to
13963 compile. However, it might become useful when a project is also used to
13964 find the list of source files in an editor, like the GNAT Programming System
13967 If @code{Body_Suffix ("Ada")} is not specified, then the default is
13968 @code{"^.adb^.ADB^"}.
13970 @item @code{Separate_Suffix}
13971 This must be a string whose value satisfies the same conditions as
13972 @code{Body_Suffix}. The same "longest suffix" rules apply.
13975 If @code{Separate_Suffix ("Ada")} is not specified, then it defaults to same
13976 value as @code{Body_Suffix ("Ada")}.
13980 You can use the associative array attribute @code{Spec} to define
13981 the source file name for an individual Ada compilation unit's spec. The array
13982 index must be a string literal that identifies the Ada unit (case insensitive).
13983 The value of this attribute must be a string that identifies the file that
13984 contains this unit's spec (case sensitive or insensitive depending on the
13987 @smallexample @c projectfile
13988 for Spec ("MyPack.MyChild") use "mypack.mychild.spec";
13991 When the source file contains several units, you can indicate at what
13992 position the unit occurs in the file, with the following. The first unit
13993 in the file has index 1
13995 @smallexample @c projectfile
13996 for Body ("top") use "foo.a" at 1;
13997 for Body ("foo") use "foo.a" at 2;
14002 You can use the associative array attribute @code{Body} to
14003 define the source file name for an individual Ada compilation unit's body
14004 (possibly a subunit). The array index must be a string literal that identifies
14005 the Ada unit (case insensitive). The value of this attribute must be a string
14006 that identifies the file that contains this unit's body or subunit (case
14007 sensitive or insensitive depending on the operating system).
14009 @smallexample @c projectfile
14010 for Body ("MyPack.MyChild") use "mypack.mychild.body";
14014 @c ********************
14015 @c * Library Projects *
14016 @c ********************
14018 @node Library Projects
14019 @section Library Projects
14022 @emph{Library projects} are projects whose object code is placed in a library.
14023 (Note that this facility is not yet supported on all platforms).
14025 @code{gnatmake} or @code{gprbuild} will collect all object files into a
14026 single archive, which might either be a shared or a static library. This
14027 library can later on be linked with multiple executables, potentially
14028 reducing their sizes.
14030 If your project file specifies languages other than Ada, but you are still
14031 using @code{gnatmake} to compile and link, the latter will not try to
14032 compile your sources other than Ada (you should use @code{gprbuild} if that
14033 is your intent). However, @code{gnatmake} will automatically link all object
14034 files found in the object directory, whether or not they were compiled from
14035 an Ada source file. This specific behavior only applies when multiple
14036 languages are specified.
14038 To create a library project, you need to define in its project file
14039 two project-level attributes: @code{Library_Name} and @code{Library_Dir}.
14040 Additionally, you may define other library-related attributes such as
14041 @code{Library_Kind}, @code{Library_Version}, @code{Library_Interface},
14042 @code{Library_Auto_Init}, @code{Library_Options} and @code{Library_GCC}.
14044 The @code{Library_Name} attribute has a string value. There is no restriction
14045 on the name of a library. It is the responsibility of the developer to
14046 choose a name that will be accepted by the platform. It is recommended to
14047 choose names that could be Ada identifiers; such names are almost guaranteed
14048 to be acceptable on all platforms.
14050 The @code{Library_Dir} attribute has a string value that designates the path
14051 (absolute or relative) of the directory where the library will reside.
14052 It must designate an existing directory. When the project is not externally
14053 built, this directory must be writable, different from the project's object
14054 directory and from any source directory in the project tree.
14056 If both @code{Library_Name} and @code{Library_Dir} are specified and
14057 are legal, then the project file defines a library project. The optional
14058 library-related attributes are checked only for such project files.
14060 The @code{Library_Kind} attribute has a string value that must be one of the
14061 following (case insensitive): @code{"static"}, @code{"dynamic"} or
14062 @code{"relocatable"} (which is a synonym for @code{"dynamic"}). If this
14063 attribute is not specified, the library is a static library, that is
14064 an archive of object files that can be potentially linked into a
14065 static executable. Otherwise, the library may be dynamic or
14066 relocatable, that is a library that is loaded only at the start of execution.
14068 If you need to build both a static and a dynamic library, you should use two
14069 different object directories, since in some cases some extra code needs to
14070 be generated for the latter. For such cases, it is recommended to either use
14071 two different project files, or a single one which uses external variables
14072 to indicate what kind of library should be build.
14074 The @code{Library_ALI_Dir} attribute may be specified to indicate the
14075 directory where the ALI files of the library will be copied. When it is
14076 not specified, the ALI files are copied to the directory specified in
14077 attribute @code{Library_Dir}. Except when the project is externally built, the
14078 directory specified by @code{Library_ALI_Dir} must be writable and different
14079 from the project's object directory and from any source directory in the
14082 The @code{Library_Version} attribute has a string value whose interpretation
14083 is platform dependent. It has no effect on VMS and Windows. On Unix, it is
14084 used only for dynamic/relocatable libraries as the internal name of the
14085 library (the @code{"soname"}). If the library file name (built from the
14086 @code{Library_Name}) is different from the @code{Library_Version}, then the
14087 library file will be a symbolic link to the actual file whose name will be
14088 @code{Library_Version}.
14092 @smallexample @c projectfile
14098 for Library_Dir use "lib_dir";
14099 for Library_Name use "dummy";
14100 for Library_Kind use "relocatable";
14101 for Library_Version use "libdummy.so." & Version;
14108 Directory @file{lib_dir} will contain the internal library file whose name
14109 will be @file{libdummy.so.1}, and @file{libdummy.so} will be a symbolic link to
14110 @file{libdummy.so.1}.
14112 When @command{gnatmake} detects that a project file
14113 is a library project file, it will check all immediate sources of the project
14114 and rebuild the library if any of the sources have been recompiled.
14116 Standard project files can import library project files. In such cases,
14117 the libraries will only be rebuilt if some of its sources are recompiled
14118 because they are in the closure of some other source in an importing project.
14119 Sources of the library project files that are not in such a closure will
14120 not be checked, unless the full library is checked, because one of its sources
14121 needs to be recompiled.
14123 For instance, assume the project file @code{A} imports the library project file
14124 @code{L}. The immediate sources of A are @file{a1.adb}, @file{a2.ads} and
14125 @file{a2.adb}. The immediate sources of L are @file{l1.ads}, @file{l1.adb},
14126 @file{l2.ads}, @file{l2.adb}.
14128 If @file{l1.adb} has been modified, then the library associated with @code{L}
14129 will be rebuilt when compiling all the immediate sources of @code{A} only
14130 if @file{a1.ads}, @file{a2.ads} or @file{a2.adb} includes a statement
14133 To be sure that all the sources in the library associated with @code{L} are
14134 up to date, and that all the sources of project @code{A} are also up to date,
14135 the following two commands needs to be used:
14142 When a library is built or rebuilt, an attempt is made first to delete all
14143 files in the library directory.
14144 All @file{ALI} files will also be copied from the object directory to the
14145 library directory. To build executables, @command{gnatmake} will use the
14146 library rather than the individual object files.
14149 It is also possible to create library project files for third-party libraries
14150 that are precompiled and cannot be compiled locally thanks to the
14151 @code{externally_built} attribute. (See @ref{Installing a library}).
14154 @c *******************************
14155 @c * Stand-alone Library Projects *
14156 @c *******************************
14158 @node Stand-alone Library Projects
14159 @section Stand-alone Library Projects
14162 A Stand-alone Library is a library that contains the necessary code to
14163 elaborate the Ada units that are included in the library. A Stand-alone
14164 Library is suitable to be used in an executable when the main is not
14165 in Ada. However, Stand-alone Libraries may also be used with an Ada main
14168 A Stand-alone Library Project is a Library Project where the library is
14169 a Stand-alone Library.
14171 To be a Stand-alone Library Project, in addition to the two attributes
14172 that make a project a Library Project (@code{Library_Name} and
14173 @code{Library_Dir}, see @ref{Library Projects}), the attribute
14174 @code{Library_Interface} must be defined.
14176 @smallexample @c projectfile
14178 for Library_Dir use "lib_dir";
14179 for Library_Name use "dummy";
14180 for Library_Interface use ("int1", "int1.child");
14184 Attribute @code{Library_Interface} has a nonempty string list value,
14185 each string in the list designating a unit contained in an immediate source
14186 of the project file.
14188 When a Stand-alone Library is built, first the binder is invoked to build
14189 a package whose name depends on the library name
14190 (^b~dummy.ads/b^B$DUMMY.ADS/B^ in the example above).
14191 This binder-generated package includes initialization and
14192 finalization procedures whose
14193 names depend on the library name (dummyinit and dummyfinal in the example
14194 above). The object corresponding to this package is included in the library.
14196 A dynamic or relocatable Stand-alone Library is automatically initialized
14197 if automatic initialization of Stand-alone Libraries is supported on the
14198 platform and if attribute @code{Library_Auto_Init} is not specified or
14199 is specified with the value "true". A static Stand-alone Library is never
14200 automatically initialized.
14202 Single string attribute @code{Library_Auto_Init} may be specified with only
14203 two possible values: "false" or "true" (case-insensitive). Specifying
14204 "false" for attribute @code{Library_Auto_Init} will prevent automatic
14205 initialization of dynamic or relocatable libraries.
14207 When a non-automatically initialized Stand-alone Library is used
14208 in an executable, its initialization procedure must be called before
14209 any service of the library is used.
14210 When the main subprogram is in Ada, it may mean that the initialization
14211 procedure has to be called during elaboration of another package.
14213 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
14214 (those that are listed in attribute @code{Library_Interface}) are copied to
14215 the Library Directory. As a consequence, only the Interface Units may be
14216 imported from Ada units outside of the library. If other units are imported,
14217 the binding phase will fail.
14219 When a Stand-Alone Library is bound, the switches that are specified in
14220 the attribute @code{Default_Switches ("Ada")} in package @code{Binder} are
14221 used in the call to @command{gnatbind}.
14223 The string list attribute @code{Library_Options} may be used to specified
14224 additional switches to the call to @command{gcc} to link the library.
14226 The attribute @code{Library_Src_Dir}, may be specified for a
14227 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
14228 single string value. Its value must be the path (absolute or relative to the
14229 project directory) of an existing directory. This directory cannot be the
14230 object directory or one of the source directories, but it can be the same as
14231 the library directory. The sources of the Interface
14232 Units of the library, necessary to an Ada client of the library, will be
14233 copied to the designated directory, called Interface Copy directory.
14234 These sources includes the specs of the Interface Units, but they may also
14235 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
14236 are used, or when there is a generic units in the spec. Before the sources
14237 are copied to the Interface Copy directory, an attempt is made to delete all
14238 files in the Interface Copy directory.
14240 @c *************************************
14241 @c * Switches Related to Project Files *
14242 @c *************************************
14243 @node Switches Related to Project Files
14244 @section Switches Related to Project Files
14247 The following switches are used by GNAT tools that support project files:
14251 @item ^-P^/PROJECT_FILE=^@var{project}
14252 @cindex @option{^-P^/PROJECT_FILE^} (any project-aware tool)
14253 Indicates the name of a project file. This project file will be parsed with
14254 the verbosity indicated by @option{^-vP^MESSAGE_PROJECT_FILES=^@emph{x}},
14255 if any, and using the external references indicated
14256 by @option{^-X^/EXTERNAL_REFERENCE^} switches, if any.
14258 There may zero, one or more spaces between @option{-P} and @var{project}.
14262 There must be only one @option{^-P^/PROJECT_FILE^} switch on the command line.
14265 Since the Project Manager parses the project file only after all the switches
14266 on the command line are checked, the order of the switches
14267 @option{^-P^/PROJECT_FILE^},
14268 @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}}
14269 or @option{^-X^/EXTERNAL_REFERENCE^} is not significant.
14271 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
14272 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (any project-aware tool)
14273 Indicates that external variable @var{name} has the value @var{value}.
14274 The Project Manager will use this value for occurrences of
14275 @code{external(name)} when parsing the project file.
14279 If @var{name} or @var{value} includes a space, then @var{name=value} should be
14280 put between quotes.
14288 Several @option{^-X^/EXTERNAL_REFERENCE^} switches can be used simultaneously.
14289 If several @option{^-X^/EXTERNAL_REFERENCE^} switches specify the same
14290 @var{name}, only the last one is used.
14293 An external variable specified with a @option{^-X^/EXTERNAL_REFERENCE^} switch
14294 takes precedence over the value of the same name in the environment.
14296 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
14297 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (any project-aware tool)
14298 Indicates the verbosity of the parsing of GNAT project files.
14301 @option{-vP0} means Default;
14302 @option{-vP1} means Medium;
14303 @option{-vP2} means High.
14307 There are three possible options for this qualifier: DEFAULT, MEDIUM and
14312 The default is ^Default^DEFAULT^: no output for syntactically correct
14315 If several @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}} switches are present,
14316 only the last one is used.
14318 @item ^-aP^/ADD_PROJECT_SEARCH_DIR=^<dir>
14319 @cindex @option{^-aP^/ADD_PROJECT_SEARCH_DIR=^} (any project-aware tool)
14320 Add directory <dir> at the beginning of the project search path, in order,
14321 after the current working directory.
14325 @cindex @option{-eL} (any project-aware tool)
14326 Follow all symbolic links when processing project files.
14329 @item ^--subdirs^/SUBDIRS^=<subdir>
14330 @cindex @option{^--subdirs^/SUBDIRS^=} (gnatmake and gnatclean)
14331 This switch is recognized by gnatmake and gnatclean. It indicate that the real
14332 directories (except the source directories) are the subdirectories <subdir>
14333 of the directories specified in the project files. This applies in particular
14334 to object directories, library directories and exec directories. If the
14335 subdirectories do not exist, they are created automatically.
14339 @c **********************************
14340 @c * Tools Supporting Project Files *
14341 @c **********************************
14343 @node Tools Supporting Project Files
14344 @section Tools Supporting Project Files
14347 * gnatmake and Project Files::
14348 * The GNAT Driver and Project Files::
14351 @node gnatmake and Project Files
14352 @subsection gnatmake and Project Files
14355 This section covers several topics related to @command{gnatmake} and
14356 project files: defining ^switches^switches^ for @command{gnatmake}
14357 and for the tools that it invokes; specifying configuration pragmas;
14358 the use of the @code{Main} attribute; building and rebuilding library project
14362 * ^Switches^Switches^ and Project Files::
14363 * Specifying Configuration Pragmas::
14364 * Project Files and Main Subprograms::
14365 * Library Project Files::
14368 @node ^Switches^Switches^ and Project Files
14369 @subsubsection ^Switches^Switches^ and Project Files
14372 It is not currently possible to specify VMS style qualifiers in the project
14373 files; only Unix style ^switches^switches^ may be specified.
14377 For each of the packages @code{Builder}, @code{Compiler}, @code{Binder}, and
14378 @code{Linker}, you can specify a @code{^Default_Switches^Default_Switches^}
14379 attribute, a @code{^Switches^Switches^} attribute, or both;
14380 as their names imply, these ^switch^switch^-related
14381 attributes affect the ^switches^switches^ that are used for each of these GNAT
14383 @command{gnatmake} is invoked. As will be explained below, these
14384 component-specific ^switches^switches^ precede
14385 the ^switches^switches^ provided on the @command{gnatmake} command line.
14387 The @code{^Default_Switches^Default_Switches^} attribute is an associative
14388 array indexed by language name (case insensitive) whose value is a string list.
14391 @smallexample @c projectfile
14393 package Compiler is
14394 for ^Default_Switches^Default_Switches^ ("Ada")
14395 use ("^-gnaty^-gnaty^",
14402 The @code{^Switches^Switches^} attribute is also an associative array,
14403 indexed by a file name (which may or may not be case sensitive, depending
14404 on the operating system) whose value is a string list. For example:
14406 @smallexample @c projectfile
14409 for ^Switches^Switches^ ("main1.adb")
14411 for ^Switches^Switches^ ("main2.adb")
14418 For the @code{Builder} package, the file names must designate source files
14419 for main subprograms. For the @code{Binder} and @code{Linker} packages, the
14420 file names must designate @file{ALI} or source files for main subprograms.
14421 In each case just the file name without an explicit extension is acceptable.
14423 For each tool used in a program build (@command{gnatmake}, the compiler, the
14424 binder, and the linker), the corresponding package @dfn{contributes} a set of
14425 ^switches^switches^ for each file on which the tool is invoked, based on the
14426 ^switch^switch^-related attributes defined in the package.
14427 In particular, the ^switches^switches^
14428 that each of these packages contributes for a given file @var{f} comprise:
14432 the value of attribute @code{^Switches^Switches^ (@var{f})},
14433 if it is specified in the package for the given file,
14435 otherwise, the value of @code{^Default_Switches^Default_Switches^ ("Ada")},
14436 if it is specified in the package.
14440 If neither of these attributes is defined in the package, then the package does
14441 not contribute any ^switches^switches^ for the given file.
14443 When @command{gnatmake} is invoked on a file, the ^switches^switches^ comprise
14444 two sets, in the following order: those contributed for the file
14445 by the @code{Builder} package;
14446 and the switches passed on the command line.
14448 When @command{gnatmake} invokes a tool (compiler, binder, linker) on a file,
14449 the ^switches^switches^ passed to the tool comprise three sets,
14450 in the following order:
14454 the applicable ^switches^switches^ contributed for the file
14455 by the @code{Builder} package in the project file supplied on the command line;
14458 those contributed for the file by the package (in the relevant project file --
14459 see below) corresponding to the tool; and
14462 the applicable switches passed on the command line.
14466 The term @emph{applicable ^switches^switches^} reflects the fact that
14467 @command{gnatmake} ^switches^switches^ may or may not be passed to individual
14468 tools, depending on the individual ^switch^switch^.
14470 @command{gnatmake} may invoke the compiler on source files from different
14471 projects. The Project Manager will use the appropriate project file to
14472 determine the @code{Compiler} package for each source file being compiled.
14473 Likewise for the @code{Binder} and @code{Linker} packages.
14475 As an example, consider the following package in a project file:
14477 @smallexample @c projectfile
14480 package Compiler is
14481 for ^Default_Switches^Default_Switches^ ("Ada")
14483 for ^Switches^Switches^ ("a.adb")
14485 for ^Switches^Switches^ ("b.adb")
14487 "^-gnaty^-gnaty^");
14494 If @command{gnatmake} is invoked with this project file, and it needs to
14495 compile, say, the files @file{a.adb}, @file{b.adb}, and @file{c.adb}, then
14496 @file{a.adb} will be compiled with the ^switch^switch^
14497 @option{^-O1^-O1^},
14498 @file{b.adb} with ^switches^switches^
14500 and @option{^-gnaty^-gnaty^},
14501 and @file{c.adb} with @option{^-g^-g^}.
14503 The following example illustrates the ordering of the ^switches^switches^
14504 contributed by different packages:
14506 @smallexample @c projectfile
14510 for ^Switches^Switches^ ("main.adb")
14518 package Compiler is
14519 for ^Switches^Switches^ ("main.adb")
14527 If you issue the command:
14530 gnatmake ^-Pproj2^/PROJECT_FILE=PROJ2^ -O0 main
14534 then the compiler will be invoked on @file{main.adb} with the following
14535 sequence of ^switches^switches^
14538 ^-g -O1 -O2 -O0^-g -O1 -O2 -O0^
14541 with the last @option{^-O^-O^}
14542 ^switch^switch^ having precedence over the earlier ones;
14543 several other ^switches^switches^
14544 (such as @option{^-c^-c^}) are added implicitly.
14546 The ^switches^switches^
14548 and @option{^-O1^-O1^} are contributed by package
14549 @code{Builder}, @option{^-O2^-O2^} is contributed
14550 by the package @code{Compiler}
14551 and @option{^-O0^-O0^} comes from the command line.
14553 The @option{^-g^-g^}
14554 ^switch^switch^ will also be passed in the invocation of
14555 @command{Gnatlink.}
14557 A final example illustrates switch contributions from packages in different
14560 @smallexample @c projectfile
14563 for Source_Files use ("pack.ads", "pack.adb");
14564 package Compiler is
14565 for ^Default_Switches^Default_Switches^ ("Ada")
14566 use ("^-gnata^-gnata^");
14574 for Source_Files use ("foo_main.adb", "bar_main.adb");
14576 for ^Switches^Switches^ ("foo_main.adb")
14584 -- Ada source file:
14586 procedure Foo_Main is
14594 gnatmake ^-PProj4^/PROJECT_FILE=PROJ4^ foo_main.adb -cargs -gnato
14598 then the ^switches^switches^ passed to the compiler for @file{foo_main.adb} are
14599 @option{^-g^-g^} (contributed by the package @code{Proj4.Builder}) and
14600 @option{^-gnato^-gnato^} (passed on the command line).
14601 When the imported package @code{Pack} is compiled, the ^switches^switches^ used
14602 are @option{^-g^-g^} from @code{Proj4.Builder},
14603 @option{^-gnata^-gnata^} (contributed from package @code{Proj3.Compiler},
14604 and @option{^-gnato^-gnato^} from the command line.
14607 When using @command{gnatmake} with project files, some ^switches^switches^ or
14608 arguments may be expressed as relative paths. As the working directory where
14609 compilation occurs may change, these relative paths are converted to absolute
14610 paths. For the ^switches^switches^ found in a project file, the relative paths
14611 are relative to the project file directory, for the switches on the command
14612 line, they are relative to the directory where @command{gnatmake} is invoked.
14613 The ^switches^switches^ for which this occurs are:
14619 ^-aI^-aI^, as well as all arguments that are not switches (arguments to
14621 ^-o^-o^, object files specified in package @code{Linker} or after
14622 -largs on the command line). The exception to this rule is the ^switch^switch^
14623 ^--RTS=^--RTS=^ for which a relative path argument is never converted.
14625 @node Specifying Configuration Pragmas
14626 @subsubsection Specifying Configuration Pragmas
14628 When using @command{gnatmake} with project files, if there exists a file
14629 @file{gnat.adc} that contains configuration pragmas, this file will be
14632 Configuration pragmas can be defined by means of the following attributes in
14633 project files: @code{Global_Configuration_Pragmas} in package @code{Builder}
14634 and @code{Local_Configuration_Pragmas} in package @code{Compiler}.
14636 Both these attributes are single string attributes. Their values is the path
14637 name of a file containing configuration pragmas. If a path name is relative,
14638 then it is relative to the project directory of the project file where the
14639 attribute is defined.
14641 When compiling a source, the configuration pragmas used are, in order,
14642 those listed in the file designated by attribute
14643 @code{Global_Configuration_Pragmas} in package @code{Builder} of the main
14644 project file, if it is specified, and those listed in the file designated by
14645 attribute @code{Local_Configuration_Pragmas} in package @code{Compiler} of
14646 the project file of the source, if it exists.
14648 @node Project Files and Main Subprograms
14649 @subsubsection Project Files and Main Subprograms
14652 When using a project file, you can invoke @command{gnatmake}
14653 with one or several main subprograms, by specifying their source files on the
14657 gnatmake ^-P^/PROJECT_FILE=^prj main1 main2 main3
14661 Each of these needs to be a source file of the same project, except
14662 when the switch ^-u^/UNIQUE^ is used.
14665 When ^-u^/UNIQUE^ is not used, all the mains need to be sources of the
14666 same project, one of the project in the tree rooted at the project specified
14667 on the command line. The package @code{Builder} of this common project, the
14668 "main project" is the one that is considered by @command{gnatmake}.
14671 When ^-u^/UNIQUE^ is used, the specified source files may be in projects
14672 imported directly or indirectly by the project specified on the command line.
14673 Note that if such a source file is not part of the project specified on the
14674 command line, the ^switches^switches^ found in package @code{Builder} of the
14675 project specified on the command line, if any, that are transmitted
14676 to the compiler will still be used, not those found in the project file of
14680 When using a project file, you can also invoke @command{gnatmake} without
14681 explicitly specifying any main, and the effect depends on whether you have
14682 defined the @code{Main} attribute. This attribute has a string list value,
14683 where each element in the list is the name of a source file (the file
14684 extension is optional) that contains a unit that can be a main subprogram.
14686 If the @code{Main} attribute is defined in a project file as a non-empty
14687 string list and the switch @option{^-u^/UNIQUE^} is not used on the command
14688 line, then invoking @command{gnatmake} with this project file but without any
14689 main on the command line is equivalent to invoking @command{gnatmake} with all
14690 the file names in the @code{Main} attribute on the command line.
14693 @smallexample @c projectfile
14696 for Main use ("main1", "main2", "main3");
14702 With this project file, @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^"}
14704 @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^ main1 main2 main3"}.
14706 When the project attribute @code{Main} is not specified, or is specified
14707 as an empty string list, or when the switch @option{-u} is used on the command
14708 line, then invoking @command{gnatmake} with no main on the command line will
14709 result in all immediate sources of the project file being checked, and
14710 potentially recompiled. Depending on the presence of the switch @option{-u},
14711 sources from other project files on which the immediate sources of the main
14712 project file depend are also checked and potentially recompiled. In other
14713 words, the @option{-u} switch is applied to all of the immediate sources of the
14716 When no main is specified on the command line and attribute @code{Main} exists
14717 and includes several mains, or when several mains are specified on the
14718 command line, the default ^switches^switches^ in package @code{Builder} will
14719 be used for all mains, even if there are specific ^switches^switches^
14720 specified for one or several mains.
14722 But the ^switches^switches^ from package @code{Binder} or @code{Linker} will be
14723 the specific ^switches^switches^ for each main, if they are specified.
14725 @node Library Project Files
14726 @subsubsection Library Project Files
14729 When @command{gnatmake} is invoked with a main project file that is a library
14730 project file, it is not allowed to specify one or more mains on the command
14734 When a library project file is specified, switches ^-b^/ACTION=BIND^ and
14735 ^-l^/ACTION=LINK^ have special meanings.
14738 @item ^-b^/ACTION=BIND^ is only allowed for stand-alone libraries. It indicates
14739 to @command{gnatmake} that @command{gnatbind} should be invoked for the
14742 @item ^-l^/ACTION=LINK^ may be used for all library projects. It indicates
14743 to @command{gnatmake} that the binder generated file should be compiled
14744 (in the case of a stand-alone library) and that the library should be built.
14748 @node The GNAT Driver and Project Files
14749 @subsection The GNAT Driver and Project Files
14752 A number of GNAT tools, other than @command{^gnatmake^gnatmake^}
14753 can benefit from project files:
14754 (@command{^gnatbind^gnatbind^},
14755 @command{^gnatcheck^gnatcheck^},
14756 @command{^gnatclean^gnatclean^},
14757 @command{^gnatelim^gnatelim^},
14758 @command{^gnatfind^gnatfind^},
14759 @command{^gnatlink^gnatlink^},
14760 @command{^gnatls^gnatls^},
14761 @command{^gnatmetric^gnatmetric^},
14762 @command{^gnatpp^gnatpp^},
14763 @command{^gnatstub^gnatstub^},
14764 and @command{^gnatxref^gnatxref^}). However, none of these tools can be invoked
14765 directly with a project file switch (@option{^-P^/PROJECT_FILE=^}).
14766 They must be invoked through the @command{gnat} driver.
14768 The @command{gnat} driver is a wrapper that accepts a number of commands and
14769 calls the corresponding tool. It was designed initially for VMS platforms (to
14770 convert VMS qualifiers to Unix-style switches), but it is now available on all
14773 On non-VMS platforms, the @command{gnat} driver accepts the following commands
14774 (case insensitive):
14778 BIND to invoke @command{^gnatbind^gnatbind^}
14780 CHOP to invoke @command{^gnatchop^gnatchop^}
14782 CLEAN to invoke @command{^gnatclean^gnatclean^}
14784 COMP or COMPILE to invoke the compiler
14786 ELIM to invoke @command{^gnatelim^gnatelim^}
14788 FIND to invoke @command{^gnatfind^gnatfind^}
14790 KR or KRUNCH to invoke @command{^gnatkr^gnatkr^}
14792 LINK to invoke @command{^gnatlink^gnatlink^}
14794 LS or LIST to invoke @command{^gnatls^gnatls^}
14796 MAKE to invoke @command{^gnatmake^gnatmake^}
14798 NAME to invoke @command{^gnatname^gnatname^}
14800 PREP or PREPROCESS to invoke @command{^gnatprep^gnatprep^}
14802 PP or PRETTY to invoke @command{^gnatpp^gnatpp^}
14804 METRIC to invoke @command{^gnatmetric^gnatmetric^}
14806 STUB to invoke @command{^gnatstub^gnatstub^}
14808 XREF to invoke @command{^gnatxref^gnatxref^}
14812 (note that the compiler is invoked using the command
14813 @command{^gnatmake -f -u -c^gnatmake -f -u -c^}).
14816 On non-VMS platforms, between @command{gnat} and the command, two
14817 special switches may be used:
14821 @command{-v} to display the invocation of the tool.
14823 @command{-dn} to prevent the @command{gnat} driver from removing
14824 the temporary files it has created. These temporary files are
14825 configuration files and temporary file list files.
14829 The command may be followed by switches and arguments for the invoked
14833 gnat bind -C main.ali
14839 Switches may also be put in text files, one switch per line, and the text
14840 files may be specified with their path name preceded by '@@'.
14843 gnat bind @@args.txt main.ali
14847 In addition, for commands BIND, COMP or COMPILE, FIND, ELIM, LS or LIST, LINK,
14848 METRIC, PP or PRETTY, STUB and XREF, the project file related switches
14849 (@option{^-P^/PROJECT_FILE^},
14850 @option{^-X^/EXTERNAL_REFERENCE^} and
14851 @option{^-vP^/MESSAGES_PROJECT_FILE=^x}) may be used in addition to
14852 the switches of the invoking tool.
14855 When GNAT PP or GNAT PRETTY is used with a project file, but with no source
14856 specified on the command line, it invokes @command{^gnatpp^gnatpp^} with all
14857 the immediate sources of the specified project file.
14860 When GNAT METRIC is used with a project file, but with no source
14861 specified on the command line, it invokes @command{^gnatmetric^gnatmetric^}
14862 with all the immediate sources of the specified project file and with
14863 @option{^-d^/DIRECTORY^} with the parameter pointing to the object directory
14867 In addition, when GNAT PP, GNAT PRETTY or GNAT METRIC is used with
14868 a project file, no source is specified on the command line and
14869 switch ^-U^/ALL_PROJECTS^ is specified on the command line, then
14870 the underlying tool (^gnatpp^gnatpp^ or
14871 ^gnatmetric^gnatmetric^) is invoked for all sources of all projects,
14872 not only for the immediate sources of the main project.
14874 (-U stands for Universal or Union of the project files of the project tree)
14878 For each of the following commands, there is optionally a corresponding
14879 package in the main project.
14883 package @code{Binder} for command BIND (invoking @code{^gnatbind^gnatbind^})
14886 package @code{Check} for command CHECK (invoking
14887 @code{^gnatcheck^gnatcheck^})
14890 package @code{Compiler} for command COMP or COMPILE (invoking the compiler)
14893 package @code{Cross_Reference} for command XREF (invoking
14894 @code{^gnatxref^gnatxref^})
14897 package @code{Eliminate} for command ELIM (invoking
14898 @code{^gnatelim^gnatelim^})
14901 package @code{Finder} for command FIND (invoking @code{^gnatfind^gnatfind^})
14904 package @code{Gnatls} for command LS or LIST (invoking @code{^gnatls^gnatls^})
14907 package @code{Gnatstub} for command STUB
14908 (invoking @code{^gnatstub^gnatstub^})
14911 package @code{Linker} for command LINK (invoking @code{^gnatlink^gnatlink^})
14914 package @code{Check} for command CHECK
14915 (invoking @code{^gnatcheck^gnatcheck^})
14918 package @code{Metrics} for command METRIC
14919 (invoking @code{^gnatmetric^gnatmetric^})
14922 package @code{Pretty_Printer} for command PP or PRETTY
14923 (invoking @code{^gnatpp^gnatpp^})
14928 Package @code{Gnatls} has a unique attribute @code{^Switches^Switches^},
14929 a simple variable with a string list value. It contains ^switches^switches^
14930 for the invocation of @code{^gnatls^gnatls^}.
14932 @smallexample @c projectfile
14936 for ^Switches^Switches^
14945 All other packages have two attribute @code{^Switches^Switches^} and
14946 @code{^Default_Switches^Default_Switches^}.
14949 @code{^Switches^Switches^} is an associative array attribute, indexed by the
14950 source file name, that has a string list value: the ^switches^switches^ to be
14951 used when the tool corresponding to the package is invoked for the specific
14955 @code{^Default_Switches^Default_Switches^} is an associative array attribute,
14956 indexed by the programming language that has a string list value.
14957 @code{^Default_Switches^Default_Switches^ ("Ada")} contains the
14958 ^switches^switches^ for the invocation of the tool corresponding
14959 to the package, except if a specific @code{^Switches^Switches^} attribute
14960 is specified for the source file.
14962 @smallexample @c projectfile
14966 for Source_Dirs use ("./**");
14969 for ^Switches^Switches^ use
14976 package Compiler is
14977 for ^Default_Switches^Default_Switches^ ("Ada")
14978 use ("^-gnatv^-gnatv^",
14979 "^-gnatwa^-gnatwa^");
14985 for ^Default_Switches^Default_Switches^ ("Ada")
14993 for ^Default_Switches^Default_Switches^ ("Ada")
14995 for ^Switches^Switches^ ("main.adb")
15004 for ^Default_Switches^Default_Switches^ ("Ada")
15011 package Cross_Reference is
15012 for ^Default_Switches^Default_Switches^ ("Ada")
15017 end Cross_Reference;
15023 With the above project file, commands such as
15026 ^gnat comp -Pproj main^GNAT COMP /PROJECT_FILE=PROJ MAIN^
15027 ^gnat ls -Pproj main^GNAT LIST /PROJECT_FILE=PROJ MAIN^
15028 ^gnat xref -Pproj main^GNAT XREF /PROJECT_FILE=PROJ MAIN^
15029 ^gnat bind -Pproj main.ali^GNAT BIND /PROJECT_FILE=PROJ MAIN.ALI^
15030 ^gnat link -Pproj main.ali^GNAT LINK /PROJECT_FILE=PROJ MAIN.ALI^
15034 will set up the environment properly and invoke the tool with the switches
15035 found in the package corresponding to the tool:
15036 @code{^Default_Switches^Default_Switches^ ("Ada")} for all tools,
15037 except @code{^Switches^Switches^ ("main.adb")}
15038 for @code{^gnatlink^gnatlink^}.
15039 It is also possible to invoke some of the tools,
15040 (@code{^gnatcheck^gnatcheck^},
15041 @code{^gnatmetric^gnatmetric^},
15042 and @code{^gnatpp^gnatpp^})
15043 on a set of project units thanks to the combination of the switches
15044 @option{-P}, @option{-U} and possibly the main unit when one is interested
15045 in its closure. For instance,
15049 will compute the metrics for all the immediate units of project
15052 gnat metric -Pproj -U
15054 will compute the metrics for all the units of the closure of projects
15055 rooted at @code{proj}.
15057 gnat metric -Pproj -U main_unit
15059 will compute the metrics for the closure of units rooted at
15060 @code{main_unit}. This last possibility relies implicitly
15061 on @command{gnatbind}'s option @option{-R}. But if the argument files for the
15062 tool invoked by the the @command{gnat} driver are explicitly specified
15063 either directly or through the tool @option{-files} option, then the tool
15064 is called only for these explicitly specified files.
15066 @c **********************
15067 @node An Extended Example
15068 @section An Extended Example
15071 Suppose that we have two programs, @var{prog1} and @var{prog2},
15072 whose sources are in corresponding directories. We would like
15073 to build them with a single @command{gnatmake} command, and we want to place
15074 their object files into @file{build} subdirectories of the source directories.
15075 Furthermore, we want to have to have two separate subdirectories
15076 in @file{build} -- @file{release} and @file{debug} -- which will contain
15077 the object files compiled with different set of compilation flags.
15079 In other words, we have the following structure:
15096 Here are the project files that we must place in a directory @file{main}
15097 to maintain this structure:
15101 @item We create a @code{Common} project with a package @code{Compiler} that
15102 specifies the compilation ^switches^switches^:
15107 @b{project} Common @b{is}
15109 @b{for} Source_Dirs @b{use} (); -- No source files
15113 @b{type} Build_Type @b{is} ("release", "debug");
15114 Build : Build_Type := External ("BUILD", "debug");
15117 @b{package} Compiler @b{is}
15118 @b{case} Build @b{is}
15119 @b{when} "release" =>
15120 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15121 @b{use} ("^-O2^-O2^");
15122 @b{when} "debug" =>
15123 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15124 @b{use} ("^-g^-g^");
15132 @item We create separate projects for the two programs:
15139 @b{project} Prog1 @b{is}
15141 @b{for} Source_Dirs @b{use} ("prog1");
15142 @b{for} Object_Dir @b{use} "prog1/build/" & Common.Build;
15144 @b{package} Compiler @b{renames} Common.Compiler;
15155 @b{project} Prog2 @b{is}
15157 @b{for} Source_Dirs @b{use} ("prog2");
15158 @b{for} Object_Dir @b{use} "prog2/build/" & Common.Build;
15160 @b{package} Compiler @b{renames} Common.Compiler;
15166 @item We create a wrapping project @code{Main}:
15175 @b{project} Main @b{is}
15177 @b{package} Compiler @b{renames} Common.Compiler;
15183 @item Finally we need to create a dummy procedure that @code{with}s (either
15184 explicitly or implicitly) all the sources of our two programs.
15189 Now we can build the programs using the command
15192 gnatmake ^-P^/PROJECT_FILE=^main dummy
15196 for the Debug mode, or
15200 gnatmake -Pmain -XBUILD=release
15206 GNAT MAKE /PROJECT_FILE=main /EXTERNAL_REFERENCE=BUILD=release
15211 for the Release mode.
15213 @c ********************************
15214 @c * Project File Complete Syntax *
15215 @c ********************************
15217 @node Project File Complete Syntax
15218 @section Project File Complete Syntax
15222 context_clause project_declaration
15228 @b{with} path_name @{ , path_name @} ;
15233 project_declaration ::=
15234 simple_project_declaration | project_extension
15236 simple_project_declaration ::=
15237 @b{project} <project_>simple_name @b{is}
15238 @{declarative_item@}
15239 @b{end} <project_>simple_name;
15241 project_extension ::=
15242 @b{project} <project_>simple_name @b{extends} path_name @b{is}
15243 @{declarative_item@}
15244 @b{end} <project_>simple_name;
15246 declarative_item ::=
15247 package_declaration |
15248 typed_string_declaration |
15249 other_declarative_item
15251 package_declaration ::=
15252 package_spec | package_renaming
15255 @b{package} package_identifier @b{is}
15256 @{simple_declarative_item@}
15257 @b{end} package_identifier ;
15259 package_identifier ::=
15260 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
15261 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
15262 @code{^gnatls^gnatls^} | @code{IDE} | @code{Pretty_Printer}
15264 package_renaming ::==
15265 @b{package} package_identifier @b{renames}
15266 <project_>simple_name.package_identifier ;
15268 typed_string_declaration ::=
15269 @b{type} <typed_string_>_simple_name @b{is}
15270 ( string_literal @{, string_literal@} );
15272 other_declarative_item ::=
15273 attribute_declaration |
15274 typed_variable_declaration |
15275 variable_declaration |
15278 attribute_declaration ::=
15279 full_associative_array_declaration |
15280 @b{for} attribute_designator @b{use} expression ;
15282 full_associative_array_declaration ::=
15283 @b{for} <associative_array_attribute_>simple_name @b{use}
15284 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
15286 attribute_designator ::=
15287 <simple_attribute_>simple_name |
15288 <associative_array_attribute_>simple_name ( string_literal )
15290 typed_variable_declaration ::=
15291 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
15293 variable_declaration ::=
15294 <variable_>simple_name := expression;
15304 attribute_reference
15310 ( <string_>expression @{ , <string_>expression @} )
15313 @b{external} ( string_literal [, string_literal] )
15315 attribute_reference ::=
15316 attribute_prefix ' <simple_attribute_>simple_name [ ( literal_string ) ]
15318 attribute_prefix ::=
15320 <project_>simple_name | package_identifier |
15321 <project_>simple_name . package_identifier
15323 case_construction ::=
15324 @b{case} <typed_variable_>name @b{is}
15329 @b{when} discrete_choice_list =>
15330 @{case_construction | attribute_declaration@}
15332 discrete_choice_list ::=
15333 string_literal @{| string_literal@} |
15337 simple_name @{. simple_name@}
15340 identifier (same as Ada)
15344 @node The Cross-Referencing Tools gnatxref and gnatfind
15345 @chapter The Cross-Referencing Tools @code{gnatxref} and @code{gnatfind}
15350 The compiler generates cross-referencing information (unless
15351 you set the @samp{-gnatx} switch), which are saved in the @file{.ali} files.
15352 This information indicates where in the source each entity is declared and
15353 referenced. Note that entities in package Standard are not included, but
15354 entities in all other predefined units are included in the output.
15356 Before using any of these two tools, you need to compile successfully your
15357 application, so that GNAT gets a chance to generate the cross-referencing
15360 The two tools @code{gnatxref} and @code{gnatfind} take advantage of this
15361 information to provide the user with the capability to easily locate the
15362 declaration and references to an entity. These tools are quite similar,
15363 the difference being that @code{gnatfind} is intended for locating
15364 definitions and/or references to a specified entity or entities, whereas
15365 @code{gnatxref} is oriented to generating a full report of all
15368 To use these tools, you must not compile your application using the
15369 @option{-gnatx} switch on the @command{gnatmake} command line
15370 (@pxref{The GNAT Make Program gnatmake}). Otherwise, cross-referencing
15371 information will not be generated.
15373 Note: to invoke @code{gnatxref} or @code{gnatfind} with a project file,
15374 use the @code{gnat} driver (see @ref{The GNAT Driver and Project Files}).
15377 * Switches for gnatxref::
15378 * Switches for gnatfind::
15379 * Project Files for gnatxref and gnatfind::
15380 * Regular Expressions in gnatfind and gnatxref::
15381 * Examples of gnatxref Usage::
15382 * Examples of gnatfind Usage::
15385 @node Switches for gnatxref
15386 @section @code{gnatxref} Switches
15389 The command invocation for @code{gnatxref} is:
15391 @c $ gnatxref @ovar{switches} @var{sourcefile1} @r{[}@var{sourcefile2} @dots{}@r{]}
15392 @c Expanding @ovar macro inline (explanation in macro def comments)
15393 $ gnatxref @r{[}@var{switches}@r{]} @var{sourcefile1} @r{[}@var{sourcefile2} @dots{}@r{]}
15402 identifies the source files for which a report is to be generated. The
15403 ``with''ed units will be processed too. You must provide at least one file.
15405 These file names are considered to be regular expressions, so for instance
15406 specifying @file{source*.adb} is the same as giving every file in the current
15407 directory whose name starts with @file{source} and whose extension is
15410 You shouldn't specify any directory name, just base names. @command{gnatxref}
15411 and @command{gnatfind} will be able to locate these files by themselves using
15412 the source path. If you specify directories, no result is produced.
15417 The switches can be:
15421 @cindex @option{--version} @command{gnatxref}
15422 Display Copyright and version, then exit disregarding all other options.
15425 @cindex @option{--help} @command{gnatxref}
15426 If @option{--version} was not used, display usage, then exit disregarding
15429 @item ^-a^/ALL_FILES^
15430 @cindex @option{^-a^/ALL_FILES^} (@command{gnatxref})
15431 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15432 the read-only files found in the library search path. Otherwise, these files
15433 will be ignored. This option can be used to protect Gnat sources or your own
15434 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15435 much faster, and their output much smaller. Read-only here refers to access
15436 or permissions status in the file system for the current user.
15439 @cindex @option{-aIDIR} (@command{gnatxref})
15440 When looking for source files also look in directory DIR. The order in which
15441 source file search is undertaken is the same as for @command{gnatmake}.
15444 @cindex @option{-aODIR} (@command{gnatxref})
15445 When searching for library and object files, look in directory
15446 DIR. The order in which library files are searched is the same as for
15447 @command{gnatmake}.
15450 @cindex @option{-nostdinc} (@command{gnatxref})
15451 Do not look for sources in the system default directory.
15454 @cindex @option{-nostdlib} (@command{gnatxref})
15455 Do not look for library files in the system default directory.
15457 @item --RTS=@var{rts-path}
15458 @cindex @option{--RTS} (@command{gnatxref})
15459 Specifies the default location of the runtime library. Same meaning as the
15460 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15462 @item ^-d^/DERIVED_TYPES^
15463 @cindex @option{^-d^/DERIVED_TYPES^} (@command{gnatxref})
15464 If this switch is set @code{gnatxref} will output the parent type
15465 reference for each matching derived types.
15467 @item ^-f^/FULL_PATHNAME^
15468 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatxref})
15469 If this switch is set, the output file names will be preceded by their
15470 directory (if the file was found in the search path). If this switch is
15471 not set, the directory will not be printed.
15473 @item ^-g^/IGNORE_LOCALS^
15474 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatxref})
15475 If this switch is set, information is output only for library-level
15476 entities, ignoring local entities. The use of this switch may accelerate
15477 @code{gnatfind} and @code{gnatxref}.
15480 @cindex @option{-IDIR} (@command{gnatxref})
15481 Equivalent to @samp{-aODIR -aIDIR}.
15484 @cindex @option{-pFILE} (@command{gnatxref})
15485 Specify a project file to use @xref{Project Files}.
15486 If you need to use the @file{.gpr}
15487 project files, you should use gnatxref through the GNAT driver
15488 (@command{gnat xref -Pproject}).
15490 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15491 project file in the current directory.
15493 If a project file is either specified or found by the tools, then the content
15494 of the source directory and object directory lines are added as if they
15495 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^}
15496 and @samp{^-aO^OBJECT_SEARCH^}.
15498 Output only unused symbols. This may be really useful if you give your
15499 main compilation unit on the command line, as @code{gnatxref} will then
15500 display every unused entity and 'with'ed package.
15504 Instead of producing the default output, @code{gnatxref} will generate a
15505 @file{tags} file that can be used by vi. For examples how to use this
15506 feature, see @ref{Examples of gnatxref Usage}. The tags file is output
15507 to the standard output, thus you will have to redirect it to a file.
15513 All these switches may be in any order on the command line, and may even
15514 appear after the file names. They need not be separated by spaces, thus
15515 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15516 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15518 @node Switches for gnatfind
15519 @section @code{gnatfind} Switches
15522 The command line for @code{gnatfind} is:
15525 @c $ gnatfind @ovar{switches} @var{pattern}@r{[}:@var{sourcefile}@r{[}:@var{line}@r{[}:@var{column}@r{]]]}
15526 @c @r{[}@var{file1} @var{file2} @dots{}]
15527 @c Expanding @ovar macro inline (explanation in macro def comments)
15528 $ gnatfind @r{[}@var{switches}@r{]} @var{pattern}@r{[}:@var{sourcefile}@r{[}:@var{line}@r{[}:@var{column}@r{]]]}
15529 @r{[}@var{file1} @var{file2} @dots{}@r{]}
15537 An entity will be output only if it matches the regular expression found
15538 in @var{pattern}, see @ref{Regular Expressions in gnatfind and gnatxref}.
15540 Omitting the pattern is equivalent to specifying @samp{*}, which
15541 will match any entity. Note that if you do not provide a pattern, you
15542 have to provide both a sourcefile and a line.
15544 Entity names are given in Latin-1, with uppercase/lowercase equivalence
15545 for matching purposes. At the current time there is no support for
15546 8-bit codes other than Latin-1, or for wide characters in identifiers.
15549 @code{gnatfind} will look for references, bodies or declarations
15550 of symbols referenced in @file{@var{sourcefile}}, at line @var{line}
15551 and column @var{column}. See @ref{Examples of gnatfind Usage}
15552 for syntax examples.
15555 is a decimal integer identifying the line number containing
15556 the reference to the entity (or entities) to be located.
15559 is a decimal integer identifying the exact location on the
15560 line of the first character of the identifier for the
15561 entity reference. Columns are numbered from 1.
15563 @item file1 file2 @dots{}
15564 The search will be restricted to these source files. If none are given, then
15565 the search will be done for every library file in the search path.
15566 These file must appear only after the pattern or sourcefile.
15568 These file names are considered to be regular expressions, so for instance
15569 specifying @file{source*.adb} is the same as giving every file in the current
15570 directory whose name starts with @file{source} and whose extension is
15573 The location of the spec of the entity will always be displayed, even if it
15574 isn't in one of @file{@var{file1}}, @file{@var{file2}},@enddots{} The
15575 occurrences of the entity in the separate units of the ones given on the
15576 command line will also be displayed.
15578 Note that if you specify at least one file in this part, @code{gnatfind} may
15579 sometimes not be able to find the body of the subprograms.
15584 At least one of 'sourcefile' or 'pattern' has to be present on
15587 The following switches are available:
15591 @cindex @option{--version} @command{gnatfind}
15592 Display Copyright and version, then exit disregarding all other options.
15595 @cindex @option{--help} @command{gnatfind}
15596 If @option{--version} was not used, display usage, then exit disregarding
15599 @item ^-a^/ALL_FILES^
15600 @cindex @option{^-a^/ALL_FILES^} (@command{gnatfind})
15601 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15602 the read-only files found in the library search path. Otherwise, these files
15603 will be ignored. This option can be used to protect Gnat sources or your own
15604 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15605 much faster, and their output much smaller. Read-only here refers to access
15606 or permission status in the file system for the current user.
15609 @cindex @option{-aIDIR} (@command{gnatfind})
15610 When looking for source files also look in directory DIR. The order in which
15611 source file search is undertaken is the same as for @command{gnatmake}.
15614 @cindex @option{-aODIR} (@command{gnatfind})
15615 When searching for library and object files, look in directory
15616 DIR. The order in which library files are searched is the same as for
15617 @command{gnatmake}.
15620 @cindex @option{-nostdinc} (@command{gnatfind})
15621 Do not look for sources in the system default directory.
15624 @cindex @option{-nostdlib} (@command{gnatfind})
15625 Do not look for library files in the system default directory.
15627 @item --ext=@var{extension}
15628 @cindex @option{--ext} (@command{gnatfind})
15629 Specify an alternate ali file extension. The default is @code{ali} and other
15630 extensions (e.g. @code{sli} for SPARK library files) may be specified via this
15631 switch. Note that if this switch overrides the default, which means that only
15632 the new extension will be considered.
15634 @item --RTS=@var{rts-path}
15635 @cindex @option{--RTS} (@command{gnatfind})
15636 Specifies the default location of the runtime library. Same meaning as the
15637 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15639 @item ^-d^/DERIVED_TYPE_INFORMATION^
15640 @cindex @option{^-d^/DERIVED_TYPE_INFORMATION^} (@code{gnatfind})
15641 If this switch is set, then @code{gnatfind} will output the parent type
15642 reference for each matching derived types.
15644 @item ^-e^/EXPRESSIONS^
15645 @cindex @option{^-e^/EXPRESSIONS^} (@command{gnatfind})
15646 By default, @code{gnatfind} accept the simple regular expression set for
15647 @samp{pattern}. If this switch is set, then the pattern will be
15648 considered as full Unix-style regular expression.
15650 @item ^-f^/FULL_PATHNAME^
15651 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatfind})
15652 If this switch is set, the output file names will be preceded by their
15653 directory (if the file was found in the search path). If this switch is
15654 not set, the directory will not be printed.
15656 @item ^-g^/IGNORE_LOCALS^
15657 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatfind})
15658 If this switch is set, information is output only for library-level
15659 entities, ignoring local entities. The use of this switch may accelerate
15660 @code{gnatfind} and @code{gnatxref}.
15663 @cindex @option{-IDIR} (@command{gnatfind})
15664 Equivalent to @samp{-aODIR -aIDIR}.
15667 @cindex @option{-pFILE} (@command{gnatfind})
15668 Specify a project file (@pxref{Project Files}) to use.
15669 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15670 project file in the current directory.
15672 If a project file is either specified or found by the tools, then the content
15673 of the source directory and object directory lines are added as if they
15674 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^} and
15675 @samp{^-aO^/OBJECT_SEARCH^}.
15677 @item ^-r^/REFERENCES^
15678 @cindex @option{^-r^/REFERENCES^} (@command{gnatfind})
15679 By default, @code{gnatfind} will output only the information about the
15680 declaration, body or type completion of the entities. If this switch is
15681 set, the @code{gnatfind} will locate every reference to the entities in
15682 the files specified on the command line (or in every file in the search
15683 path if no file is given on the command line).
15685 @item ^-s^/PRINT_LINES^
15686 @cindex @option{^-s^/PRINT_LINES^} (@command{gnatfind})
15687 If this switch is set, then @code{gnatfind} will output the content
15688 of the Ada source file lines were the entity was found.
15690 @item ^-t^/TYPE_HIERARCHY^
15691 @cindex @option{^-t^/TYPE_HIERARCHY^} (@command{gnatfind})
15692 If this switch is set, then @code{gnatfind} will output the type hierarchy for
15693 the specified type. It act like -d option but recursively from parent
15694 type to parent type. When this switch is set it is not possible to
15695 specify more than one file.
15700 All these switches may be in any order on the command line, and may even
15701 appear after the file names. They need not be separated by spaces, thus
15702 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15703 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15705 As stated previously, gnatfind will search in every directory in the
15706 search path. You can force it to look only in the current directory if
15707 you specify @code{*} at the end of the command line.
15709 @node Project Files for gnatxref and gnatfind
15710 @section Project Files for @command{gnatxref} and @command{gnatfind}
15713 Project files allow a programmer to specify how to compile its
15714 application, where to find sources, etc. These files are used
15716 primarily by GPS, but they can also be used
15719 @code{gnatxref} and @code{gnatfind}.
15721 A project file name must end with @file{.gpr}. If a single one is
15722 present in the current directory, then @code{gnatxref} and @code{gnatfind} will
15723 extract the information from it. If multiple project files are found, none of
15724 them is read, and you have to use the @samp{-p} switch to specify the one
15727 The following lines can be included, even though most of them have default
15728 values which can be used in most cases.
15729 The lines can be entered in any order in the file.
15730 Except for @file{src_dir} and @file{obj_dir}, you can only have one instance of
15731 each line. If you have multiple instances, only the last one is taken into
15736 [default: @code{"^./^[]^"}]
15737 specifies a directory where to look for source files. Multiple @code{src_dir}
15738 lines can be specified and they will be searched in the order they
15742 [default: @code{"^./^[]^"}]
15743 specifies a directory where to look for object and library files. Multiple
15744 @code{obj_dir} lines can be specified, and they will be searched in the order
15747 @item comp_opt=SWITCHES
15748 [default: @code{""}]
15749 creates a variable which can be referred to subsequently by using
15750 the @code{$@{comp_opt@}} notation. This is intended to store the default
15751 switches given to @command{gnatmake} and @command{gcc}.
15753 @item bind_opt=SWITCHES
15754 [default: @code{""}]
15755 creates a variable which can be referred to subsequently by using
15756 the @samp{$@{bind_opt@}} notation. This is intended to store the default
15757 switches given to @command{gnatbind}.
15759 @item link_opt=SWITCHES
15760 [default: @code{""}]
15761 creates a variable which can be referred to subsequently by using
15762 the @samp{$@{link_opt@}} notation. This is intended to store the default
15763 switches given to @command{gnatlink}.
15765 @item main=EXECUTABLE
15766 [default: @code{""}]
15767 specifies the name of the executable for the application. This variable can
15768 be referred to in the following lines by using the @samp{$@{main@}} notation.
15771 @item comp_cmd=COMMAND
15772 [default: @code{"GNAT COMPILE /SEARCH=$@{src_dir@} /DEBUG /TRY_SEMANTICS"}]
15775 @item comp_cmd=COMMAND
15776 [default: @code{"gcc -c -I$@{src_dir@} -g -gnatq"}]
15778 specifies the command used to compile a single file in the application.
15781 @item make_cmd=COMMAND
15782 [default: @code{"GNAT MAKE $@{main@}
15783 /SOURCE_SEARCH=$@{src_dir@} /OBJECT_SEARCH=$@{obj_dir@}
15784 /DEBUG /TRY_SEMANTICS /COMPILER_QUALIFIERS $@{comp_opt@}
15785 /BINDER_QUALIFIERS $@{bind_opt@} /LINKER_QUALIFIERS $@{link_opt@}"}]
15788 @item make_cmd=COMMAND
15789 [default: @code{"gnatmake $@{main@} -aI$@{src_dir@}
15790 -aO$@{obj_dir@} -g -gnatq -cargs $@{comp_opt@}
15791 -bargs $@{bind_opt@} -largs $@{link_opt@}"}]
15793 specifies the command used to recompile the whole application.
15795 @item run_cmd=COMMAND
15796 [default: @code{"$@{main@}"}]
15797 specifies the command used to run the application.
15799 @item debug_cmd=COMMAND
15800 [default: @code{"gdb $@{main@}"}]
15801 specifies the command used to debug the application
15806 @command{gnatxref} and @command{gnatfind} only take into account the
15807 @code{src_dir} and @code{obj_dir} lines, and ignore the others.
15809 @node Regular Expressions in gnatfind and gnatxref
15810 @section Regular Expressions in @code{gnatfind} and @code{gnatxref}
15813 As specified in the section about @command{gnatfind}, the pattern can be a
15814 regular expression. Actually, there are to set of regular expressions
15815 which are recognized by the program:
15818 @item globbing patterns
15819 These are the most usual regular expression. They are the same that you
15820 generally used in a Unix shell command line, or in a DOS session.
15822 Here is a more formal grammar:
15829 term ::= elmt -- matches elmt
15830 term ::= elmt elmt -- concatenation (elmt then elmt)
15831 term ::= * -- any string of 0 or more characters
15832 term ::= ? -- matches any character
15833 term ::= [char @{char@}] -- matches any character listed
15834 term ::= [char - char] -- matches any character in range
15838 @item full regular expression
15839 The second set of regular expressions is much more powerful. This is the
15840 type of regular expressions recognized by utilities such a @file{grep}.
15842 The following is the form of a regular expression, expressed in Ada
15843 reference manual style BNF is as follows
15850 regexp ::= term @{| term@} -- alternation (term or term @dots{})
15852 term ::= item @{item@} -- concatenation (item then item)
15854 item ::= elmt -- match elmt
15855 item ::= elmt * -- zero or more elmt's
15856 item ::= elmt + -- one or more elmt's
15857 item ::= elmt ? -- matches elmt or nothing
15860 elmt ::= nschar -- matches given character
15861 elmt ::= [nschar @{nschar@}] -- matches any character listed
15862 elmt ::= [^^^ nschar @{nschar@}] -- matches any character not listed
15863 elmt ::= [char - char] -- matches chars in given range
15864 elmt ::= \ char -- matches given character
15865 elmt ::= . -- matches any single character
15866 elmt ::= ( regexp ) -- parens used for grouping
15868 char ::= any character, including special characters
15869 nschar ::= any character except ()[].*+?^^^
15873 Following are a few examples:
15877 will match any of the two strings @samp{abcde} and @samp{fghi},
15880 will match any string like @samp{abd}, @samp{abcd}, @samp{abccd},
15881 @samp{abcccd}, and so on,
15884 will match any string which has only lowercase characters in it (and at
15885 least one character.
15890 @node Examples of gnatxref Usage
15891 @section Examples of @code{gnatxref} Usage
15893 @subsection General Usage
15896 For the following examples, we will consider the following units:
15898 @smallexample @c ada
15904 3: procedure Foo (B : in Integer);
15911 1: package body Main is
15912 2: procedure Foo (B : in Integer) is
15923 2: procedure Print (B : Integer);
15932 The first thing to do is to recompile your application (for instance, in
15933 that case just by doing a @samp{gnatmake main}, so that GNAT generates
15934 the cross-referencing information.
15935 You can then issue any of the following commands:
15937 @item gnatxref main.adb
15938 @code{gnatxref} generates cross-reference information for main.adb
15939 and every unit 'with'ed by main.adb.
15941 The output would be:
15949 Decl: main.ads 3:20
15950 Body: main.adb 2:20
15951 Ref: main.adb 4:13 5:13 6:19
15954 Ref: main.adb 6:8 7:8
15964 Decl: main.ads 3:15
15965 Body: main.adb 2:15
15968 Body: main.adb 1:14
15971 Ref: main.adb 6:12 7:12
15975 that is the entity @code{Main} is declared in main.ads, line 2, column 9,
15976 its body is in main.adb, line 1, column 14 and is not referenced any where.
15978 The entity @code{Print} is declared in bar.ads, line 2, column 15 and it
15979 it referenced in main.adb, line 6 column 12 and line 7 column 12.
15981 @item gnatxref package1.adb package2.ads
15982 @code{gnatxref} will generates cross-reference information for
15983 package1.adb, package2.ads and any other package 'with'ed by any
15989 @subsection Using gnatxref with vi
15991 @code{gnatxref} can generate a tags file output, which can be used
15992 directly from @command{vi}. Note that the standard version of @command{vi}
15993 will not work properly with overloaded symbols. Consider using another
15994 free implementation of @command{vi}, such as @command{vim}.
15997 $ gnatxref -v gnatfind.adb > tags
16001 will generate the tags file for @code{gnatfind} itself (if the sources
16002 are in the search path!).
16004 From @command{vi}, you can then use the command @samp{:tag @var{entity}}
16005 (replacing @var{entity} by whatever you are looking for), and vi will
16006 display a new file with the corresponding declaration of entity.
16009 @node Examples of gnatfind Usage
16010 @section Examples of @code{gnatfind} Usage
16014 @item gnatfind ^-f^/FULL_PATHNAME^ xyz:main.adb
16015 Find declarations for all entities xyz referenced at least once in
16016 main.adb. The references are search in every library file in the search
16019 The directories will be printed as well (as the @samp{^-f^/FULL_PATHNAME^}
16022 The output will look like:
16024 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
16025 ^directory/^[directory]^main.adb:24:10: xyz <= body
16026 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
16030 that is to say, one of the entities xyz found in main.adb is declared at
16031 line 12 of main.ads (and its body is in main.adb), and another one is
16032 declared at line 45 of foo.ads
16034 @item gnatfind ^-fs^/FULL_PATHNAME/SOURCE_LINE^ xyz:main.adb
16035 This is the same command as the previous one, instead @code{gnatfind} will
16036 display the content of the Ada source file lines.
16038 The output will look like:
16041 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
16043 ^directory/^[directory]^main.adb:24:10: xyz <= body
16045 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
16050 This can make it easier to find exactly the location your are looking
16053 @item gnatfind ^-r^/REFERENCES^ "*x*":main.ads:123 foo.adb
16054 Find references to all entities containing an x that are
16055 referenced on line 123 of main.ads.
16056 The references will be searched only in main.ads and foo.adb.
16058 @item gnatfind main.ads:123
16059 Find declarations and bodies for all entities that are referenced on
16060 line 123 of main.ads.
16062 This is the same as @code{gnatfind "*":main.adb:123}.
16064 @item gnatfind ^mydir/^[mydir]^main.adb:123:45
16065 Find the declaration for the entity referenced at column 45 in
16066 line 123 of file main.adb in directory mydir. Note that it
16067 is usual to omit the identifier name when the column is given,
16068 since the column position identifies a unique reference.
16070 The column has to be the beginning of the identifier, and should not
16071 point to any character in the middle of the identifier.
16075 @c *********************************
16076 @node The GNAT Pretty-Printer gnatpp
16077 @chapter The GNAT Pretty-Printer @command{gnatpp}
16079 @cindex Pretty-Printer
16082 ^The @command{gnatpp} tool^GNAT PRETTY^ is an ASIS-based utility
16083 for source reformatting / pretty-printing.
16084 It takes an Ada source file as input and generates a reformatted
16086 You can specify various style directives via switches; e.g.,
16087 identifier case conventions, rules of indentation, and comment layout.
16089 To produce a reformatted file, @command{gnatpp} generates and uses the ASIS
16090 tree for the input source and thus requires the input to be syntactically and
16091 semantically legal.
16092 If this condition is not met, @command{gnatpp} will terminate with an
16093 error message; no output file will be generated.
16095 If the source files presented to @command{gnatpp} contain
16096 preprocessing directives, then the output file will
16097 correspond to the generated source after all
16098 preprocessing is carried out. There is no way
16099 using @command{gnatpp} to obtain pretty printed files that
16100 include the preprocessing directives.
16102 If the compilation unit
16103 contained in the input source depends semantically upon units located
16104 outside the current directory, you have to provide the source search path
16105 when invoking @command{gnatpp}, if these units are contained in files with
16106 names that do not follow the GNAT file naming rules, you have to provide
16107 the configuration file describing the corresponding naming scheme;
16108 see the description of the @command{gnatpp}
16109 switches below. Another possibility is to use a project file and to
16110 call @command{gnatpp} through the @command{gnat} driver
16112 The @command{gnatpp} command has the form
16115 @c $ gnatpp @ovar{switches} @var{filename}
16116 @c Expanding @ovar macro inline (explanation in macro def comments)
16117 $ gnatpp @r{[}@var{switches}@r{]} @var{filename} @r{[}-cargs @var{gcc_switches}@r{]}
16124 @var{switches} is an optional sequence of switches defining such properties as
16125 the formatting rules, the source search path, and the destination for the
16129 @var{filename} is the name (including the extension) of the source file to
16130 reformat; ``wildcards'' or several file names on the same gnatpp command are
16131 allowed. The file name may contain path information; it does not have to
16132 follow the GNAT file naming rules
16135 @samp{@var{gcc_switches}} is a list of switches for
16136 @command{gcc}. They will be passed on to all compiler invocations made by
16137 @command{gnatelim} to generate the ASIS trees. Here you can provide
16138 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
16139 use the @option{-gnatec} switch to set the configuration file etc.
16143 * Switches for gnatpp::
16144 * Formatting Rules::
16147 @node Switches for gnatpp
16148 @section Switches for @command{gnatpp}
16151 The following subsections describe the various switches accepted by
16152 @command{gnatpp}, organized by category.
16155 You specify a switch by supplying a name and generally also a value.
16156 In many cases the values for a switch with a given name are incompatible with
16158 (for example the switch that controls the casing of a reserved word may have
16159 exactly one value: upper case, lower case, or
16160 mixed case) and thus exactly one such switch can be in effect for an
16161 invocation of @command{gnatpp}.
16162 If more than one is supplied, the last one is used.
16163 However, some values for the same switch are mutually compatible.
16164 You may supply several such switches to @command{gnatpp}, but then
16165 each must be specified in full, with both the name and the value.
16166 Abbreviated forms (the name appearing once, followed by each value) are
16168 For example, to set
16169 the alignment of the assignment delimiter both in declarations and in
16170 assignment statements, you must write @option{-A2A3}
16171 (or @option{-A2 -A3}), but not @option{-A23}.
16175 In many cases the set of options for a given qualifier are incompatible with
16176 each other (for example the qualifier that controls the casing of a reserved
16177 word may have exactly one option, which specifies either upper case, lower
16178 case, or mixed case), and thus exactly one such option can be in effect for
16179 an invocation of @command{gnatpp}.
16180 If more than one is supplied, the last one is used.
16181 However, some qualifiers have options that are mutually compatible,
16182 and then you may then supply several such options when invoking
16186 In most cases, it is obvious whether or not the
16187 ^values for a switch with a given name^options for a given qualifier^
16188 are compatible with each other.
16189 When the semantics might not be evident, the summaries below explicitly
16190 indicate the effect.
16193 * Alignment Control::
16195 * Construct Layout Control::
16196 * General Text Layout Control::
16197 * Other Formatting Options::
16198 * Setting the Source Search Path::
16199 * Output File Control::
16200 * Other gnatpp Switches::
16203 @node Alignment Control
16204 @subsection Alignment Control
16205 @cindex Alignment control in @command{gnatpp}
16208 Programs can be easier to read if certain constructs are vertically aligned.
16209 By default all alignments are set ON.
16210 Through the @option{^-A0^/ALIGN=OFF^} switch you may reset the default to
16211 OFF, and then use one or more of the other
16212 ^@option{-A@var{n}} switches^@option{/ALIGN} options^
16213 to activate alignment for specific constructs.
16216 @cindex @option{^-A@var{n}^/ALIGN^} (@command{gnatpp})
16220 Set all alignments to ON
16223 @item ^-A0^/ALIGN=OFF^
16224 Set all alignments to OFF
16226 @item ^-A1^/ALIGN=COLONS^
16227 Align @code{:} in declarations
16229 @item ^-A2^/ALIGN=DECLARATIONS^
16230 Align @code{:=} in initializations in declarations
16232 @item ^-A3^/ALIGN=STATEMENTS^
16233 Align @code{:=} in assignment statements
16235 @item ^-A4^/ALIGN=ARROWS^
16236 Align @code{=>} in associations
16238 @item ^-A5^/ALIGN=COMPONENT_CLAUSES^
16239 Align @code{at} keywords in the component clauses in record
16240 representation clauses
16244 The @option{^-A^/ALIGN^} switches are mutually compatible; any combination
16247 @node Casing Control
16248 @subsection Casing Control
16249 @cindex Casing control in @command{gnatpp}
16252 @command{gnatpp} allows you to specify the casing for reserved words,
16253 pragma names, attribute designators and identifiers.
16254 For identifiers you may define a
16255 general rule for name casing but also override this rule
16256 via a set of dictionary files.
16258 Three types of casing are supported: lower case, upper case, and mixed case.
16259 Lower and upper case are self-explanatory (but since some letters in
16260 Latin1 and other GNAT-supported character sets
16261 exist only in lower-case form, an upper case conversion will have no
16263 ``Mixed case'' means that the first letter, and also each letter immediately
16264 following an underscore, are converted to their uppercase forms;
16265 all the other letters are converted to their lowercase forms.
16268 @cindex @option{^-a@var{x}^/ATTRIBUTE^} (@command{gnatpp})
16269 @item ^-aL^/ATTRIBUTE_CASING=LOWER_CASE^
16270 Attribute designators are lower case
16272 @item ^-aU^/ATTRIBUTE_CASING=UPPER_CASE^
16273 Attribute designators are upper case
16275 @item ^-aM^/ATTRIBUTE_CASING=MIXED_CASE^
16276 Attribute designators are mixed case (this is the default)
16278 @cindex @option{^-k@var{x}^/KEYWORD_CASING^} (@command{gnatpp})
16279 @item ^-kL^/KEYWORD_CASING=LOWER_CASE^
16280 Keywords (technically, these are known in Ada as @emph{reserved words}) are
16281 lower case (this is the default)
16283 @item ^-kU^/KEYWORD_CASING=UPPER_CASE^
16284 Keywords are upper case
16286 @cindex @option{^-n@var{x}^/NAME_CASING^} (@command{gnatpp})
16287 @item ^-nD^/NAME_CASING=AS_DECLARED^
16288 Name casing for defining occurrences are as they appear in the source file
16289 (this is the default)
16291 @item ^-nU^/NAME_CASING=UPPER_CASE^
16292 Names are in upper case
16294 @item ^-nL^/NAME_CASING=LOWER_CASE^
16295 Names are in lower case
16297 @item ^-nM^/NAME_CASING=MIXED_CASE^
16298 Names are in mixed case
16300 @cindex @option{^-p@var{x}^/PRAGMA_CASING^} (@command{gnatpp})
16301 @item ^-pL^/PRAGMA_CASING=LOWER_CASE^
16302 Pragma names are lower case
16304 @item ^-pU^/PRAGMA_CASING=UPPER_CASE^
16305 Pragma names are upper case
16307 @item ^-pM^/PRAGMA_CASING=MIXED_CASE^
16308 Pragma names are mixed case (this is the default)
16310 @item ^-D@var{file}^/DICTIONARY=@var{file}^
16311 @cindex @option{^-D^/DICTIONARY^} (@command{gnatpp})
16312 Use @var{file} as a @emph{dictionary file} that defines
16313 the casing for a set of specified names,
16314 thereby overriding the effect on these names by
16315 any explicit or implicit
16316 ^-n^/NAME_CASING^ switch.
16317 To supply more than one dictionary file,
16318 use ^several @option{-D} switches^a list of files as options^.
16321 @option{gnatpp} implicitly uses a @emph{default dictionary file}
16322 to define the casing for the Ada predefined names and
16323 the names declared in the GNAT libraries.
16325 @item ^-D-^/SPECIFIC_CASING^
16326 @cindex @option{^-D-^/SPECIFIC_CASING^} (@command{gnatpp})
16327 Do not use the default dictionary file;
16328 instead, use the casing
16329 defined by a @option{^-n^/NAME_CASING^} switch and any explicit
16334 The structure of a dictionary file, and details on the conventions
16335 used in the default dictionary file, are defined in @ref{Name Casing}.
16337 The @option{^-D-^/SPECIFIC_CASING^} and
16338 @option{^-D@var{file}^/DICTIONARY=@var{file}^} switches are mutually
16341 @node Construct Layout Control
16342 @subsection Construct Layout Control
16343 @cindex Layout control in @command{gnatpp}
16346 This group of @command{gnatpp} switches controls the layout of comments and
16347 complex syntactic constructs. See @ref{Formatting Comments} for details
16351 @cindex @option{^-c@var{n}^/COMMENTS_LAYOUT^} (@command{gnatpp})
16352 @item ^-c0^/COMMENTS_LAYOUT=UNTOUCHED^
16353 All the comments remain unchanged
16355 @item ^-c1^/COMMENTS_LAYOUT=DEFAULT^
16356 GNAT-style comment line indentation (this is the default).
16358 @item ^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^
16359 Reference-manual comment line indentation.
16361 @item ^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^
16362 GNAT-style comment beginning
16364 @item ^-c4^/COMMENTS_LAYOUT=REFORMAT^
16365 Reformat comment blocks
16367 @item ^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^
16368 Keep unchanged special form comments
16370 Reformat comment blocks
16372 @cindex @option{^-l@var{n}^/CONSTRUCT_LAYOUT^} (@command{gnatpp})
16373 @item ^-l1^/CONSTRUCT_LAYOUT=GNAT^
16374 GNAT-style layout (this is the default)
16376 @item ^-l2^/CONSTRUCT_LAYOUT=COMPACT^
16379 @item ^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^
16382 @cindex @option{^-N^/NOTABS^} (@command{gnatpp})
16384 All the VT characters are removed from the comment text. All the HT characters
16385 are expanded with the sequences of space characters to get to the next tab
16388 @cindex @option{^--no-separate-is^/NO_SEPARATE_IS^} (@command{gnatpp})
16389 @item ^--no-separate-is^/NO_SEPARATE_IS^
16390 Do not place the keyword @code{is} on a separate line in a subprogram body in
16391 case if the spec occupies more then one line.
16393 @cindex @option{^--separate-label^/SEPARATE_LABEL^} (@command{gnatpp})
16394 @item ^--separate-label^/SEPARATE_LABEL^
16395 Place statement label(s) on a separate line, with the following statement
16398 @cindex @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} (@command{gnatpp})
16399 @item ^--separate-loop-then^/SEPARATE_LOOP_THEN^
16400 Place the keyword @code{loop} in FOR and WHILE loop statements and the
16401 keyword @code{then} in IF statements on a separate line.
16403 @cindex @option{^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^} (@command{gnatpp})
16404 @item ^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^
16405 Do not place the keyword @code{loop} in FOR and WHILE loop statements and the
16406 keyword @code{then} in IF statements on a separate line. This option is
16407 incompatible with @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} option.
16409 @cindex @option{^--use-on-new-line^/USE_ON_NEW_LINE^} (@command{gnatpp})
16410 @item ^--use-on-new-line^/USE_ON_NEW_LINE^
16411 Start each USE clause in a context clause from a separate line.
16413 @cindex @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^} (@command{gnatpp})
16414 @item ^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^
16415 Use a separate line for a loop or block statement name, but do not use an extra
16416 indentation level for the statement itself.
16422 The @option{-c1} and @option{-c2} switches are incompatible.
16423 The @option{-c3} and @option{-c4} switches are compatible with each other and
16424 also with @option{-c1} and @option{-c2}. The @option{-c0} switch disables all
16425 the other comment formatting switches.
16427 The @option{-l1}, @option{-l2}, and @option{-l3} switches are incompatible.
16432 For the @option{/COMMENTS_LAYOUT} qualifier:
16435 The @option{DEFAULT} and @option{STANDARD_INDENT} options are incompatible.
16437 The @option{GNAT_BEGINNING} and @option{REFORMAT} options are compatible with
16438 each other and also with @option{DEFAULT} and @option{STANDARD_INDENT}.
16442 The @option{GNAT}, @option{COMPACT}, and @option{UNCOMPACT} options for the
16443 @option{/CONSTRUCT_LAYOUT} qualifier are incompatible.
16446 @node General Text Layout Control
16447 @subsection General Text Layout Control
16450 These switches allow control over line length and indentation.
16453 @item ^-M@var{nnn}^/LINE_LENGTH_MAX=@var{nnn}^
16454 @cindex @option{^-M^/LINE_LENGTH^} (@command{gnatpp})
16455 Maximum line length, @var{nnn} from 32@dots{}256, the default value is 79
16457 @item ^-i@var{nnn}^/INDENTATION_LEVEL=@var{nnn}^
16458 @cindex @option{^-i^/INDENTATION_LEVEL^} (@command{gnatpp})
16459 Indentation level, @var{nnn} from 1@dots{}9, the default value is 3
16461 @item ^-cl@var{nnn}^/CONTINUATION_INDENT=@var{nnn}^
16462 @cindex @option{^-cl^/CONTINUATION_INDENT^} (@command{gnatpp})
16463 Indentation level for continuation lines (relative to the line being
16464 continued), @var{nnn} from 1@dots{}9.
16466 value is one less then the (normal) indentation level, unless the
16467 indentation is set to 1 (in which case the default value for continuation
16468 line indentation is also 1)
16471 @node Other Formatting Options
16472 @subsection Other Formatting Options
16475 These switches control the inclusion of missing end/exit labels, and
16476 the indentation level in @b{case} statements.
16479 @item ^-e^/NO_MISSED_LABELS^
16480 @cindex @option{^-e^/NO_MISSED_LABELS^} (@command{gnatpp})
16481 Do not insert missing end/exit labels. An end label is the name of
16482 a construct that may optionally be repeated at the end of the
16483 construct's declaration;
16484 e.g., the names of packages, subprograms, and tasks.
16485 An exit label is the name of a loop that may appear as target
16486 of an exit statement within the loop.
16487 By default, @command{gnatpp} inserts these end/exit labels when
16488 they are absent from the original source. This option suppresses such
16489 insertion, so that the formatted source reflects the original.
16491 @item ^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^
16492 @cindex @option{^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^} (@command{gnatpp})
16493 Insert a Form Feed character after a pragma Page.
16495 @item ^-T@var{nnn}^/MAX_INDENT=@var{nnn}^
16496 @cindex @option{^-T^/MAX_INDENT^} (@command{gnatpp})
16497 Do not use an additional indentation level for @b{case} alternatives
16498 and variants if there are @var{nnn} or more (the default
16500 If @var{nnn} is 0, an additional indentation level is
16501 used for @b{case} alternatives and variants regardless of their number.
16504 @node Setting the Source Search Path
16505 @subsection Setting the Source Search Path
16508 To define the search path for the input source file, @command{gnatpp}
16509 uses the same switches as the GNAT compiler, with the same effects.
16512 @item ^-I^/SEARCH=^@var{dir}
16513 @cindex @option{^-I^/SEARCH^} (@code{gnatpp})
16514 The same as the corresponding gcc switch
16516 @item ^-I-^/NOCURRENT_DIRECTORY^
16517 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatpp})
16518 The same as the corresponding gcc switch
16520 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE^=@var{path}
16521 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@code{gnatpp})
16522 The same as the corresponding gcc switch
16524 @item ^--RTS^/RUNTIME_SYSTEM^=@var{path}
16525 @cindex @option{^--RTS^/RUNTIME_SYSTEM^} (@code{gnatpp})
16526 The same as the corresponding gcc switch
16530 @node Output File Control
16531 @subsection Output File Control
16534 By default the output is sent to the file whose name is obtained by appending
16535 the ^@file{.pp}^@file{$PP}^ suffix to the name of the input file
16536 (if the file with this name already exists, it is unconditionally overwritten).
16537 Thus if the input file is @file{^my_ada_proc.adb^MY_ADA_PROC.ADB^} then
16538 @command{gnatpp} will produce @file{^my_ada_proc.adb.pp^MY_ADA_PROC.ADB$PP^}
16540 The output may be redirected by the following switches:
16543 @item ^-pipe^/STANDARD_OUTPUT^
16544 @cindex @option{^-pipe^/STANDARD_OUTPUT^} (@code{gnatpp})
16545 Send the output to @code{Standard_Output}
16547 @item ^-o @var{output_file}^/OUTPUT=@var{output_file}^
16548 @cindex @option{^-o^/OUTPUT^} (@code{gnatpp})
16549 Write the output into @var{output_file}.
16550 If @var{output_file} already exists, @command{gnatpp} terminates without
16551 reading or processing the input file.
16553 @item ^-of ^/FORCED_OUTPUT=^@var{output_file}
16554 @cindex @option{^-of^/FORCED_OUTPUT^} (@code{gnatpp})
16555 Write the output into @var{output_file}, overwriting the existing file
16556 (if one is present).
16558 @item ^-r^/REPLACE^
16559 @cindex @option{^-r^/REPLACE^} (@code{gnatpp})
16560 Replace the input source file with the reformatted output, and copy the
16561 original input source into the file whose name is obtained by appending the
16562 ^@file{.npp}^@file{$NPP}^ suffix to the name of the input file.
16563 If a file with this name already exists, @command{gnatpp} terminates without
16564 reading or processing the input file.
16566 @item ^-rf^/OVERRIDING_REPLACE^
16567 @cindex @option{^-rf^/OVERRIDING_REPLACE^} (@code{gnatpp})
16568 Like @option{^-r^/REPLACE^} except that if the file with the specified name
16569 already exists, it is overwritten.
16571 @item ^-rnb^/REPLACE_NO_BACKUP^
16572 @cindex @option{^-rnb^/REPLACE_NO_BACKUP^} (@code{gnatpp})
16573 Replace the input source file with the reformatted output without
16574 creating any backup copy of the input source.
16576 @item ^--eol=@var{xxx}^/END_OF_LINE=@var{xxx}^
16577 @cindex @option{^--eol^/END_OF_LINE^} (@code{gnatpp})
16578 Specifies the format of the reformatted output file. The @var{xxx}
16579 ^string specified with the switch^option^ may be either
16581 @item ``@option{^dos^DOS^}'' MS DOS style, lines end with CR LF characters
16582 @item ``@option{^crlf^CRLF^}''
16583 the same as @option{^crlf^CRLF^}
16584 @item ``@option{^unix^UNIX^}'' UNIX style, lines end with LF character
16585 @item ``@option{^lf^LF^}''
16586 the same as @option{^unix^UNIX^}
16589 @item ^-W^/RESULT_ENCODING=^@var{e}
16590 @cindex @option{^-W^/RESULT_ENCODING=^} (@command{gnatpp})
16591 Specify the wide character encoding method used to write the code in the
16593 @var{e} is one of the following:
16601 Upper half encoding
16603 @item ^s^SHIFT_JIS^
16613 Brackets encoding (default value)
16619 Options @option{^-pipe^/STANDARD_OUTPUT^},
16620 @option{^-o^/OUTPUT^} and
16621 @option{^-of^/FORCED_OUTPUT^} are allowed only if the call to gnatpp
16622 contains only one file to reformat.
16624 @option{^--eol^/END_OF_LINE^}
16626 @option{^-W^/RESULT_ENCODING^}
16627 cannot be used together
16628 with @option{^-pipe^/STANDARD_OUTPUT^} option.
16630 @node Other gnatpp Switches
16631 @subsection Other @code{gnatpp} Switches
16634 The additional @command{gnatpp} switches are defined in this subsection.
16637 @item ^-files @var{filename}^/FILES=@var{filename}^
16638 @cindex @option{^-files^/FILES^} (@code{gnatpp})
16639 Take the argument source files from the specified file. This file should be an
16640 ordinary text file containing file names separated by spaces or
16641 line breaks. You can use this switch more than once in the same call to
16642 @command{gnatpp}. You also can combine this switch with an explicit list of
16645 @item ^-v^/VERBOSE^
16646 @cindex @option{^-v^/VERBOSE^} (@code{gnatpp})
16648 @command{gnatpp} generates version information and then
16649 a trace of the actions it takes to produce or obtain the ASIS tree.
16651 @item ^-w^/WARNINGS^
16652 @cindex @option{^-w^/WARNINGS^} (@code{gnatpp})
16654 @command{gnatpp} generates a warning whenever it cannot provide
16655 a required layout in the result source.
16658 @node Formatting Rules
16659 @section Formatting Rules
16662 The following subsections show how @command{gnatpp} treats ``white space'',
16663 comments, program layout, and name casing.
16664 They provide the detailed descriptions of the switches shown above.
16667 * White Space and Empty Lines::
16668 * Formatting Comments::
16669 * Construct Layout::
16673 @node White Space and Empty Lines
16674 @subsection White Space and Empty Lines
16677 @command{gnatpp} does not have an option to control space characters.
16678 It will add or remove spaces according to the style illustrated by the
16679 examples in the @cite{Ada Reference Manual}.
16681 The only format effectors
16682 (see @cite{Ada Reference Manual}, paragraph 2.1(13))
16683 that will appear in the output file are platform-specific line breaks,
16684 and also format effectors within (but not at the end of) comments.
16685 In particular, each horizontal tab character that is not inside
16686 a comment will be treated as a space and thus will appear in the
16687 output file as zero or more spaces depending on
16688 the reformatting of the line in which it appears.
16689 The only exception is a Form Feed character, which is inserted after a
16690 pragma @code{Page} when @option{-ff} is set.
16692 The output file will contain no lines with trailing ``white space'' (spaces,
16695 Empty lines in the original source are preserved
16696 only if they separate declarations or statements.
16697 In such contexts, a
16698 sequence of two or more empty lines is replaced by exactly one empty line.
16699 Note that a blank line will be removed if it separates two ``comment blocks''
16700 (a comment block is a sequence of whole-line comments).
16701 In order to preserve a visual separation between comment blocks, use an
16702 ``empty comment'' (a line comprising only hyphens) rather than an empty line.
16703 Likewise, if for some reason you wish to have a sequence of empty lines,
16704 use a sequence of empty comments instead.
16706 @node Formatting Comments
16707 @subsection Formatting Comments
16710 Comments in Ada code are of two kinds:
16713 a @emph{whole-line comment}, which appears by itself (possibly preceded by
16714 ``white space'') on a line
16717 an @emph{end-of-line comment}, which follows some other Ada lexical element
16722 The indentation of a whole-line comment is that of either
16723 the preceding or following line in
16724 the formatted source, depending on switch settings as will be described below.
16726 For an end-of-line comment, @command{gnatpp} leaves the same number of spaces
16727 between the end of the preceding Ada lexical element and the beginning
16728 of the comment as appear in the original source,
16729 unless either the comment has to be split to
16730 satisfy the line length limitation, or else the next line contains a
16731 whole line comment that is considered a continuation of this end-of-line
16732 comment (because it starts at the same position).
16734 cases, the start of the end-of-line comment is moved right to the nearest
16735 multiple of the indentation level.
16736 This may result in a ``line overflow'' (the right-shifted comment extending
16737 beyond the maximum line length), in which case the comment is split as
16740 There is a difference between @option{^-c1^/COMMENTS_LAYOUT=DEFAULT^}
16741 (GNAT-style comment line indentation)
16742 and @option{^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^}
16743 (reference-manual comment line indentation).
16744 With reference-manual style, a whole-line comment is indented as if it
16745 were a declaration or statement at the same place
16746 (i.e., according to the indentation of the preceding line(s)).
16747 With GNAT style, a whole-line comment that is immediately followed by an
16748 @b{if} or @b{case} statement alternative, a record variant, or the reserved
16749 word @b{begin}, is indented based on the construct that follows it.
16752 @smallexample @c ada
16764 Reference-manual indentation produces:
16766 @smallexample @c ada
16778 while GNAT-style indentation produces:
16780 @smallexample @c ada
16792 The @option{^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^} switch
16793 (GNAT style comment beginning) has the following
16798 For each whole-line comment that does not end with two hyphens,
16799 @command{gnatpp} inserts spaces if necessary after the starting two hyphens
16800 to ensure that there are at least two spaces between these hyphens and the
16801 first non-blank character of the comment.
16805 For an end-of-line comment, if in the original source the next line is a
16806 whole-line comment that starts at the same position
16807 as the end-of-line comment,
16808 then the whole-line comment (and all whole-line comments
16809 that follow it and that start at the same position)
16810 will start at this position in the output file.
16813 That is, if in the original source we have:
16815 @smallexample @c ada
16818 A := B + C; -- B must be in the range Low1..High1
16819 -- C must be in the range Low2..High2
16820 --B+C will be in the range Low1+Low2..High1+High2
16826 Then in the formatted source we get
16828 @smallexample @c ada
16831 A := B + C; -- B must be in the range Low1..High1
16832 -- C must be in the range Low2..High2
16833 -- B+C will be in the range Low1+Low2..High1+High2
16839 A comment that exceeds the line length limit will be split.
16841 @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} (reformat comment blocks) is set and
16842 the line belongs to a reformattable block, splitting the line generates a
16843 @command{gnatpp} warning.
16844 The @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} switch specifies that whole-line
16845 comments may be reformatted in typical
16846 word processor style (that is, moving words between lines and putting as
16847 many words in a line as possible).
16850 The @option{^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^} switch specifies, that comments
16851 that has a special format (that is, a character that is neither a letter nor digit
16852 not white space nor line break immediately following the leading @code{--} of
16853 the comment) should be without any change moved from the argument source
16854 into reformatted source. This switch allows to preserve comments that are used
16855 as a special marks in the code (e.g.@: SPARK annotation).
16857 @node Construct Layout
16858 @subsection Construct Layout
16861 In several cases the suggested layout in the Ada Reference Manual includes
16862 an extra level of indentation that many programmers prefer to avoid. The
16863 affected cases include:
16867 @item Record type declaration (RM 3.8)
16869 @item Record representation clause (RM 13.5.1)
16871 @item Loop statement in case if a loop has a statement identifier (RM 5.6)
16873 @item Block statement in case if a block has a statement identifier (RM 5.6)
16877 In compact mode (when GNAT style layout or compact layout is set),
16878 the pretty printer uses one level of indentation instead
16879 of two. This is achieved in the record definition and record representation
16880 clause cases by putting the @code{record} keyword on the same line as the
16881 start of the declaration or representation clause, and in the block and loop
16882 case by putting the block or loop header on the same line as the statement
16886 The difference between GNAT style @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^}
16887 and compact @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^}
16888 layout on the one hand, and uncompact layout
16889 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} on the other hand,
16890 can be illustrated by the following examples:
16894 @multitable @columnfractions .5 .5
16895 @item @i{GNAT style, compact layout} @tab @i{Uncompact layout}
16898 @smallexample @c ada
16905 @smallexample @c ada
16914 @smallexample @c ada
16916 a at 0 range 0 .. 31;
16917 b at 4 range 0 .. 31;
16921 @smallexample @c ada
16924 a at 0 range 0 .. 31;
16925 b at 4 range 0 .. 31;
16930 @smallexample @c ada
16938 @smallexample @c ada
16948 @smallexample @c ada
16949 Clear : for J in 1 .. 10 loop
16954 @smallexample @c ada
16956 for J in 1 .. 10 loop
16967 GNAT style, compact layout Uncompact layout
16969 type q is record type q is
16970 a : integer; record
16971 b : integer; a : integer;
16972 end record; b : integer;
16975 for q use record for q use
16976 a at 0 range 0 .. 31; record
16977 b at 4 range 0 .. 31; a at 0 range 0 .. 31;
16978 end record; b at 4 range 0 .. 31;
16981 Block : declare Block :
16982 A : Integer := 3; declare
16983 begin A : Integer := 3;
16985 end Block; Proc (A, A);
16988 Clear : for J in 1 .. 10 loop Clear :
16989 A (J) := 0; for J in 1 .. 10 loop
16990 end loop Clear; A (J) := 0;
16997 A further difference between GNAT style layout and compact layout is that
16998 GNAT style layout inserts empty lines as separation for
16999 compound statements, return statements and bodies.
17001 Note that the layout specified by
17002 @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^}
17003 for named block and loop statements overrides the layout defined by these
17004 constructs by @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^},
17005 @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^} or
17006 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} option.
17009 @subsection Name Casing
17012 @command{gnatpp} always converts the usage occurrence of a (simple) name to
17013 the same casing as the corresponding defining identifier.
17015 You control the casing for defining occurrences via the
17016 @option{^-n^/NAME_CASING^} switch.
17018 With @option{-nD} (``as declared'', which is the default),
17021 With @option{/NAME_CASING=AS_DECLARED}, which is the default,
17023 defining occurrences appear exactly as in the source file
17024 where they are declared.
17025 The other ^values for this switch^options for this qualifier^ ---
17026 @option{^-nU^UPPER_CASE^},
17027 @option{^-nL^LOWER_CASE^},
17028 @option{^-nM^MIXED_CASE^} ---
17030 ^upper, lower, or mixed case, respectively^the corresponding casing^.
17031 If @command{gnatpp} changes the casing of a defining
17032 occurrence, it analogously changes the casing of all the
17033 usage occurrences of this name.
17035 If the defining occurrence of a name is not in the source compilation unit
17036 currently being processed by @command{gnatpp}, the casing of each reference to
17037 this name is changed according to the value of the @option{^-n^/NAME_CASING^}
17038 switch (subject to the dictionary file mechanism described below).
17039 Thus @command{gnatpp} acts as though the @option{^-n^/NAME_CASING^} switch
17041 casing for the defining occurrence of the name.
17043 Some names may need to be spelled with casing conventions that are not
17044 covered by the upper-, lower-, and mixed-case transformations.
17045 You can arrange correct casing by placing such names in a
17046 @emph{dictionary file},
17047 and then supplying a @option{^-D^/DICTIONARY^} switch.
17048 The casing of names from dictionary files overrides
17049 any @option{^-n^/NAME_CASING^} switch.
17051 To handle the casing of Ada predefined names and the names from GNAT libraries,
17052 @command{gnatpp} assumes a default dictionary file.
17053 The name of each predefined entity is spelled with the same casing as is used
17054 for the entity in the @cite{Ada Reference Manual}.
17055 The name of each entity in the GNAT libraries is spelled with the same casing
17056 as is used in the declaration of that entity.
17058 The @w{@option{^-D-^/SPECIFIC_CASING^}} switch suppresses the use of the
17059 default dictionary file.
17060 Instead, the casing for predefined and GNAT-defined names will be established
17061 by the @option{^-n^/NAME_CASING^} switch or explicit dictionary files.
17062 For example, by default the names @code{Ada.Text_IO} and @code{GNAT.OS_Lib}
17063 will appear as just shown,
17064 even in the presence of a @option{^-nU^/NAME_CASING=UPPER_CASE^} switch.
17065 To ensure that even such names are rendered in uppercase,
17066 additionally supply the @w{@option{^-D-^/SPECIFIC_CASING^}} switch
17067 (or else, less conveniently, place these names in upper case in a dictionary
17070 A dictionary file is
17071 a plain text file; each line in this file can be either a blank line
17072 (containing only space characters and ASCII.HT characters), an Ada comment
17073 line, or the specification of exactly one @emph{casing schema}.
17075 A casing schema is a string that has the following syntax:
17079 @var{casing_schema} ::= @var{identifier} | *@var{simple_identifier}*
17081 @var{simple_identifier} ::= @var{letter}@{@var{letter_or_digit}@}
17086 (See @cite{Ada Reference Manual}, Section 2.3) for the definition of the
17087 @var{identifier} lexical element and the @var{letter_or_digit} category.)
17089 The casing schema string can be followed by white space and/or an Ada-style
17090 comment; any amount of white space is allowed before the string.
17092 If a dictionary file is passed as
17094 the value of a @option{-D@var{file}} switch
17097 an option to the @option{/DICTIONARY} qualifier
17100 simple name and every identifier, @command{gnatpp} checks if the dictionary
17101 defines the casing for the name or for some of its parts (the term ``subword''
17102 is used below to denote the part of a name which is delimited by ``_'' or by
17103 the beginning or end of the word and which does not contain any ``_'' inside):
17107 if the whole name is in the dictionary, @command{gnatpp} uses for this name
17108 the casing defined by the dictionary; no subwords are checked for this word
17111 for every subword @command{gnatpp} checks if the dictionary contains the
17112 corresponding string of the form @code{*@var{simple_identifier}*},
17113 and if it does, the casing of this @var{simple_identifier} is used
17117 if the whole name does not contain any ``_'' inside, and if for this name
17118 the dictionary contains two entries - one of the form @var{identifier},
17119 and another - of the form *@var{simple_identifier}*, then the first one
17120 is applied to define the casing of this name
17123 if more than one dictionary file is passed as @command{gnatpp} switches, each
17124 dictionary adds new casing exceptions and overrides all the existing casing
17125 exceptions set by the previous dictionaries
17128 when @command{gnatpp} checks if the word or subword is in the dictionary,
17129 this check is not case sensitive
17133 For example, suppose we have the following source to reformat:
17135 @smallexample @c ada
17138 name1 : integer := 1;
17139 name4_name3_name2 : integer := 2;
17140 name2_name3_name4 : Boolean;
17143 name2_name3_name4 := name4_name3_name2 > name1;
17149 And suppose we have two dictionaries:
17166 If @command{gnatpp} is called with the following switches:
17170 @command{gnatpp -nM -D dict1 -D dict2 test.adb}
17173 @command{gnatpp test.adb /NAME_CASING=MIXED_CASE /DICTIONARY=(dict1, dict2)}
17178 then we will get the following name casing in the @command{gnatpp} output:
17180 @smallexample @c ada
17183 NAME1 : Integer := 1;
17184 Name4_NAME3_Name2 : Integer := 2;
17185 Name2_NAME3_Name4 : Boolean;
17188 Name2_NAME3_Name4 := Name4_NAME3_Name2 > NAME1;
17193 @c *********************************
17194 @node The GNAT Metric Tool gnatmetric
17195 @chapter The GNAT Metric Tool @command{gnatmetric}
17197 @cindex Metric tool
17200 ^The @command{gnatmetric} tool^@command{GNAT METRIC}^ is an ASIS-based utility
17201 for computing various program metrics.
17202 It takes an Ada source file as input and generates a file containing the
17203 metrics data as output. Various switches control which
17204 metrics are computed and output.
17206 @command{gnatmetric} generates and uses the ASIS
17207 tree for the input source and thus requires the input to be syntactically and
17208 semantically legal.
17209 If this condition is not met, @command{gnatmetric} will generate
17210 an error message; no metric information for this file will be
17211 computed and reported.
17213 If the compilation unit contained in the input source depends semantically
17214 upon units in files located outside the current directory, you have to provide
17215 the source search path when invoking @command{gnatmetric}.
17216 If it depends semantically upon units that are contained
17217 in files with names that do not follow the GNAT file naming rules, you have to
17218 provide the configuration file describing the corresponding naming scheme (see
17219 the description of the @command{gnatmetric} switches below.)
17220 Alternatively, you may use a project file and invoke @command{gnatmetric}
17221 through the @command{gnat} driver.
17223 The @command{gnatmetric} command has the form
17226 @c $ gnatmetric @ovar{switches} @{@var{filename}@} @r{[}-cargs @var{gcc_switches}@r{]}
17227 @c Expanding @ovar macro inline (explanation in macro def comments)
17228 $ gnatmetric @r{[}@var{switches}@r{]} @{@var{filename}@} @r{[}-cargs @var{gcc_switches}@r{]}
17235 @var{switches} specify the metrics to compute and define the destination for
17239 Each @var{filename} is the name (including the extension) of a source
17240 file to process. ``Wildcards'' are allowed, and
17241 the file name may contain path information.
17242 If no @var{filename} is supplied, then the @var{switches} list must contain
17244 @option{-files} switch (@pxref{Other gnatmetric Switches}).
17245 Including both a @option{-files} switch and one or more
17246 @var{filename} arguments is permitted.
17249 @samp{@var{gcc_switches}} is a list of switches for
17250 @command{gcc}. They will be passed on to all compiler invocations made by
17251 @command{gnatmetric} to generate the ASIS trees. Here you can provide
17252 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
17253 and use the @option{-gnatec} switch to set the configuration file.
17257 * Switches for gnatmetric::
17260 @node Switches for gnatmetric
17261 @section Switches for @command{gnatmetric}
17264 The following subsections describe the various switches accepted by
17265 @command{gnatmetric}, organized by category.
17268 * Output Files Control::
17269 * Disable Metrics For Local Units::
17270 * Specifying a set of metrics to compute::
17271 * Other gnatmetric Switches::
17272 * Generate project-wide metrics::
17275 @node Output Files Control
17276 @subsection Output File Control
17277 @cindex Output file control in @command{gnatmetric}
17280 @command{gnatmetric} has two output formats. It can generate a
17281 textual (human-readable) form, and also XML. By default only textual
17282 output is generated.
17284 When generating the output in textual form, @command{gnatmetric} creates
17285 for each Ada source file a corresponding text file
17286 containing the computed metrics, except for the case when the set of metrics
17287 specified by gnatmetric parameters consists only of metrics that are computed
17288 for the whole set of analyzed sources, but not for each Ada source.
17289 By default, this file is placed in the same directory as where the source
17290 file is located, and its name is obtained
17291 by appending the ^@file{.metrix}^@file{$METRIX}^ suffix to the name of the
17294 All the output information generated in XML format is placed in a single
17295 file. By default this file is placed in the current directory and has the
17296 name ^@file{metrix.xml}^@file{METRIX$XML}^.
17298 Some of the computed metrics are summed over the units passed to
17299 @command{gnatmetric}; for example, the total number of lines of code.
17300 By default this information is sent to @file{stdout}, but a file
17301 can be specified with the @option{-og} switch.
17303 The following switches control the @command{gnatmetric} output:
17306 @cindex @option{^-x^/XML^} (@command{gnatmetric})
17308 Generate the XML output
17310 @cindex @option{^-xs^/XSD^} (@command{gnatmetric})
17312 Generate the XML output and the XML schema file that describes the structure
17313 of the XML metric report, this schema is assigned to the XML file. The schema
17314 file has the same name as the XML output file with @file{.xml} suffix replaced
17317 @cindex @option{^-nt^/NO_TEXT^} (@command{gnatmetric})
17318 @item ^-nt^/NO_TEXT^
17319 Do not generate the output in text form (implies @option{^-x^/XML^})
17321 @cindex @option{^-d^/DIRECTORY^} (@command{gnatmetric})
17322 @item ^-d @var{output_dir}^/DIRECTORY=@var{output_dir}^
17323 Put text files with detailed metrics into @var{output_dir}
17325 @cindex @option{^-o^/SUFFIX_DETAILS^} (@command{gnatmetric})
17326 @item ^-o @var{file_suffix}^/SUFFIX_DETAILS=@var{file_suffix}^
17327 Use @var{file_suffix}, instead of ^@file{.metrix}^@file{$METRIX}^
17328 in the name of the output file.
17330 @cindex @option{^-og^/GLOBAL_OUTPUT^} (@command{gnatmetric})
17331 @item ^-og @var{file_name}^/GLOBAL_OUTPUT=@var{file_name}^
17332 Put global metrics into @var{file_name}
17334 @cindex @option{^-ox^/XML_OUTPUT^} (@command{gnatmetric})
17335 @item ^-ox @var{file_name}^/XML_OUTPUT=@var{file_name}^
17336 Put the XML output into @var{file_name} (also implies @option{^-x^/XML^})
17338 @cindex @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} (@command{gnatmetric})
17339 @item ^-sfn^/SHORT_SOURCE_FILE_NAME^
17340 Use ``short'' source file names in the output. (The @command{gnatmetric}
17341 output includes the name(s) of the Ada source file(s) from which the metrics
17342 are computed. By default each name includes the absolute path. The
17343 @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} switch causes @command{gnatmetric}
17344 to exclude all directory information from the file names that are output.)
17348 @node Disable Metrics For Local Units
17349 @subsection Disable Metrics For Local Units
17350 @cindex Disable Metrics For Local Units in @command{gnatmetric}
17353 @command{gnatmetric} relies on the GNAT compilation model @minus{}
17355 unit per one source file. It computes line metrics for the whole source
17356 file, and it also computes syntax
17357 and complexity metrics for the file's outermost unit.
17359 By default, @command{gnatmetric} will also compute all metrics for certain
17360 kinds of locally declared program units:
17364 subprogram (and generic subprogram) bodies;
17367 package (and generic package) specs and bodies;
17370 task object and type specifications and bodies;
17373 protected object and type specifications and bodies.
17377 These kinds of entities will be referred to as
17378 @emph{eligible local program units}, or simply @emph{eligible local units},
17379 @cindex Eligible local unit (for @command{gnatmetric})
17380 in the discussion below.
17382 Note that a subprogram declaration, generic instantiation,
17383 or renaming declaration only receives metrics
17384 computation when it appear as the outermost entity
17387 Suppression of metrics computation for eligible local units can be
17388 obtained via the following switch:
17391 @cindex @option{^-n@var{x}^/SUPPRESS^} (@command{gnatmetric})
17392 @item ^-nolocal^/SUPPRESS=LOCAL_DETAILS^
17393 Do not compute detailed metrics for eligible local program units
17397 @node Specifying a set of metrics to compute
17398 @subsection Specifying a set of metrics to compute
17401 By default all the metrics are computed and reported. The switches
17402 described in this subsection allow you to control, on an individual
17403 basis, whether metrics are computed and
17404 reported. If at least one positive metric
17405 switch is specified (that is, a switch that defines that a given
17406 metric or set of metrics is to be computed), then only
17407 explicitly specified metrics are reported.
17410 * Line Metrics Control::
17411 * Syntax Metrics Control::
17412 * Complexity Metrics Control::
17413 * Object-Oriented Metrics Control::
17416 @node Line Metrics Control
17417 @subsubsection Line Metrics Control
17418 @cindex Line metrics control in @command{gnatmetric}
17421 For any (legal) source file, and for each of its
17422 eligible local program units, @command{gnatmetric} computes the following
17427 the total number of lines;
17430 the total number of code lines (i.e., non-blank lines that are not comments)
17433 the number of comment lines
17436 the number of code lines containing end-of-line comments;
17439 the comment percentage: the ratio between the number of lines that contain
17440 comments and the number of all non-blank lines, expressed as a percentage;
17443 the number of empty lines and lines containing only space characters and/or
17444 format effectors (blank lines)
17447 the average number of code lines in subprogram bodies, task bodies, entry
17448 bodies and statement sequences in package bodies (this metric is only computed
17449 across the whole set of the analyzed units)
17454 @command{gnatmetric} sums the values of the line metrics for all the
17455 files being processed and then generates the cumulative results. The tool
17456 also computes for all the files being processed the average number of code
17459 You can use the following switches to select the specific line metrics
17460 to be computed and reported.
17463 @cindex @option{^--lines@var{x}^/LINE_COUNT_METRICS^} (@command{gnatmetric})
17466 @cindex @option{--no-lines@var{x}}
17469 @item ^--lines-all^/LINE_COUNT_METRICS=ALL^
17470 Report all the line metrics
17472 @item ^--no-lines-all^/LINE_COUNT_METRICS=NONE^
17473 Do not report any of line metrics
17475 @item ^--lines^/LINE_COUNT_METRICS=ALL_LINES^
17476 Report the number of all lines
17478 @item ^--no-lines^/LINE_COUNT_METRICS=NOALL_LINES^
17479 Do not report the number of all lines
17481 @item ^--lines-code^/LINE_COUNT_METRICS=CODE_LINES^
17482 Report the number of code lines
17484 @item ^--no-lines-code^/LINE_COUNT_METRICS=NOCODE_LINES^
17485 Do not report the number of code lines
17487 @item ^--lines-comment^/LINE_COUNT_METRICS=COMMENT_LINES^
17488 Report the number of comment lines
17490 @item ^--no-lines-comment^/LINE_COUNT_METRICS=NOCOMMENT_LINES^
17491 Do not report the number of comment lines
17493 @item ^--lines-eol-comment^/LINE_COUNT_METRICS=CODE_COMMENT_LINES^
17494 Report the number of code lines containing
17495 end-of-line comments
17497 @item ^--no-lines-eol-comment^/LINE_COUNT_METRICS=NOCODE_COMMENT_LINES^
17498 Do not report the number of code lines containing
17499 end-of-line comments
17501 @item ^--lines-ratio^/LINE_COUNT_METRICS=COMMENT_PERCENTAGE^
17502 Report the comment percentage in the program text
17504 @item ^--no-lines-ratio^/LINE_COUNT_METRICS=NOCOMMENT_PERCENTAGE^
17505 Do not report the comment percentage in the program text
17507 @item ^--lines-blank^/LINE_COUNT_METRICS=BLANK_LINES^
17508 Report the number of blank lines
17510 @item ^--no-lines-blank^/LINE_COUNT_METRICS=NOBLANK_LINES^
17511 Do not report the number of blank lines
17513 @item ^--lines-average^/LINE_COUNT_METRICS=AVERAGE_BODY_LINES^
17514 Report the average number of code lines in subprogram bodies, task bodies,
17515 entry bodies and statement sequences in package bodies. The metric is computed
17516 and reported for the whole set of processed Ada sources only.
17518 @item ^--no-lines-average^/LINE_COUNT_METRICS=NOAVERAGE_BODY_LINES^
17519 Do not report the average number of code lines in subprogram bodies,
17520 task bodies, entry bodies and statement sequences in package bodies.
17524 @node Syntax Metrics Control
17525 @subsubsection Syntax Metrics Control
17526 @cindex Syntax metrics control in @command{gnatmetric}
17529 @command{gnatmetric} computes various syntactic metrics for the
17530 outermost unit and for each eligible local unit:
17533 @item LSLOC (``Logical Source Lines Of Code'')
17534 The total number of declarations and the total number of statements
17536 @item Maximal static nesting level of inner program units
17538 @cite{Ada Reference Manual}, 10.1(1), ``A program unit is either a
17539 package, a task unit, a protected unit, a
17540 protected entry, a generic unit, or an explicitly declared subprogram other
17541 than an enumeration literal.''
17543 @item Maximal nesting level of composite syntactic constructs
17544 This corresponds to the notion of the
17545 maximum nesting level in the GNAT built-in style checks
17546 (@pxref{Style Checking})
17550 For the outermost unit in the file, @command{gnatmetric} additionally computes
17551 the following metrics:
17554 @item Public subprograms
17555 This metric is computed for package specs. It is the
17556 number of subprograms and generic subprograms declared in the visible
17557 part (including the visible part of nested packages, protected objects, and
17560 @item All subprograms
17561 This metric is computed for bodies and subunits. The
17562 metric is equal to a total number of subprogram bodies in the compilation
17564 Neither generic instantiations nor renamings-as-a-body nor body stubs
17565 are counted. Any subprogram body is counted, independently of its nesting
17566 level and enclosing constructs. Generic bodies and bodies of protected
17567 subprograms are counted in the same way as ``usual'' subprogram bodies.
17570 This metric is computed for package specs and
17571 generic package declarations. It is the total number of types
17572 that can be referenced from outside this compilation unit, plus the
17573 number of types from all the visible parts of all the visible generic
17574 packages. Generic formal types are not counted. Only types, not subtypes,
17578 Along with the total number of public types, the following
17579 types are counted and reported separately:
17586 Root tagged types (abstract, non-abstract, private, non-private). Type
17587 extensions are @emph{not} counted
17590 Private types (including private extensions)
17601 This metric is computed for any compilation unit. It is equal to the total
17602 number of the declarations of different types given in the compilation unit.
17603 The private and the corresponding full type declaration are counted as one
17604 type declaration. Incomplete type declarations and generic formal types
17606 No distinction is made among different kinds of types (abstract,
17607 private etc.); the total number of types is computed and reported.
17612 By default, all the syntax metrics are computed and reported. You can use the
17613 following switches to select specific syntax metrics.
17617 @cindex @option{^--syntax@var{x}^/SYNTAX_METRICS^} (@command{gnatmetric})
17620 @cindex @option{--no-syntax@var{x}} (@command{gnatmetric})
17623 @item ^--syntax-all^/SYNTAX_METRICS=ALL^
17624 Report all the syntax metrics
17626 @item ^--no-syntax-all^/SYNTAX_METRICS=NONE^
17627 Do not report any of syntax metrics
17629 @item ^--declarations^/SYNTAX_METRICS=DECLARATIONS^
17630 Report the total number of declarations
17632 @item ^--no-declarations^/SYNTAX_METRICS=NODECLARATIONS^
17633 Do not report the total number of declarations
17635 @item ^--statements^/SYNTAX_METRICS=STATEMENTS^
17636 Report the total number of statements
17638 @item ^--no-statements^/SYNTAX_METRICS=NOSTATEMENTS^
17639 Do not report the total number of statements
17641 @item ^--public-subprograms^/SYNTAX_METRICS=PUBLIC_SUBPROGRAMS^
17642 Report the number of public subprograms in a compilation unit
17644 @item ^--no-public-subprograms^/SYNTAX_METRICS=NOPUBLIC_SUBPROGRAMS^
17645 Do not report the number of public subprograms in a compilation unit
17647 @item ^--all-subprograms^/SYNTAX_METRICS=ALL_SUBPROGRAMS^
17648 Report the number of all the subprograms in a compilation unit
17650 @item ^--no-all-subprograms^/SYNTAX_METRICS=NOALL_SUBPROGRAMS^
17651 Do not report the number of all the subprograms in a compilation unit
17653 @item ^--public-types^/SYNTAX_METRICS=PUBLIC_TYPES^
17654 Report the number of public types in a compilation unit
17656 @item ^--no-public-types^/SYNTAX_METRICS=NOPUBLIC_TYPES^
17657 Do not report the number of public types in a compilation unit
17659 @item ^--all-types^/SYNTAX_METRICS=ALL_TYPES^
17660 Report the number of all the types in a compilation unit
17662 @item ^--no-all-types^/SYNTAX_METRICS=NOALL_TYPES^
17663 Do not report the number of all the types in a compilation unit
17665 @item ^--unit-nesting^/SYNTAX_METRICS=UNIT_NESTING^
17666 Report the maximal program unit nesting level
17668 @item ^--no-unit-nesting^/SYNTAX_METRICS=UNIT_NESTING_OFF^
17669 Do not report the maximal program unit nesting level
17671 @item ^--construct-nesting^/SYNTAX_METRICS=CONSTRUCT_NESTING^
17672 Report the maximal construct nesting level
17674 @item ^--no-construct-nesting^/SYNTAX_METRICS=NOCONSTRUCT_NESTING^
17675 Do not report the maximal construct nesting level
17679 @node Complexity Metrics Control
17680 @subsubsection Complexity Metrics Control
17681 @cindex Complexity metrics control in @command{gnatmetric}
17684 For a program unit that is an executable body (a subprogram body (including
17685 generic bodies), task body, entry body or a package body containing
17686 its own statement sequence) @command{gnatmetric} computes the following
17687 complexity metrics:
17691 McCabe cyclomatic complexity;
17694 McCabe essential complexity;
17697 maximal loop nesting level
17702 The McCabe complexity metrics are defined
17703 in @url{http://www.mccabe.com/pdf/nist235r.pdf}
17705 According to McCabe, both control statements and short-circuit control forms
17706 should be taken into account when computing cyclomatic complexity. For each
17707 body, we compute three metric values:
17711 the complexity introduced by control
17712 statements only, without taking into account short-circuit forms,
17715 the complexity introduced by short-circuit control forms only, and
17719 cyclomatic complexity, which is the sum of these two values.
17723 When computing cyclomatic and essential complexity, @command{gnatmetric} skips
17724 the code in the exception handlers and in all the nested program units.
17726 By default, all the complexity metrics are computed and reported.
17727 For more fine-grained control you can use
17728 the following switches:
17731 @cindex @option{^-complexity@var{x}^/COMPLEXITY_METRICS^} (@command{gnatmetric})
17734 @cindex @option{--no-complexity@var{x}}
17737 @item ^--complexity-all^/COMPLEXITY_METRICS=ALL^
17738 Report all the complexity metrics
17740 @item ^--no-complexity-all^/COMPLEXITY_METRICS=NONE^
17741 Do not report any of complexity metrics
17743 @item ^--complexity-cyclomatic^/COMPLEXITY_METRICS=CYCLOMATIC^
17744 Report the McCabe Cyclomatic Complexity
17746 @item ^--no-complexity-cyclomatic^/COMPLEXITY_METRICS=NOCYCLOMATIC^
17747 Do not report the McCabe Cyclomatic Complexity
17749 @item ^--complexity-essential^/COMPLEXITY_METRICS=ESSENTIAL^
17750 Report the Essential Complexity
17752 @item ^--no-complexity-essential^/COMPLEXITY_METRICS=NOESSENTIAL^
17753 Do not report the Essential Complexity
17755 @item ^--loop-nesting^/COMPLEXITY_METRICS=LOOP_NESTING_ON^
17756 Report maximal loop nesting level
17758 @item ^--no-loop-nesting^/COMPLEXITY_METRICS=NOLOOP_NESTING^
17759 Do not report maximal loop nesting level
17761 @item ^--complexity-average^/COMPLEXITY_METRICS=AVERAGE_COMPLEXITY^
17762 Report the average McCabe Cyclomatic Complexity for all the subprogram bodies,
17763 task bodies, entry bodies and statement sequences in package bodies.
17764 The metric is computed and reported for whole set of processed Ada sources
17767 @item ^--no-complexity-average^/COMPLEXITY_METRICS=NOAVERAGE_COMPLEXITY^
17768 Do not report the average McCabe Cyclomatic Complexity for all the subprogram
17769 bodies, task bodies, entry bodies and statement sequences in package bodies
17771 @cindex @option{^-ne^/NO_EXITS_AS_GOTOS^} (@command{gnatmetric})
17772 @item ^-ne^/NO_EXITS_AS_GOTOS^
17773 Do not consider @code{exit} statements as @code{goto}s when
17774 computing Essential Complexity
17776 @item ^--extra-exit-points^/EXTRA_EXIT_POINTS^
17777 Report the extra exit points for subprogram bodies. As an exit point, this
17778 metric counts @code{return} statements and raise statements in case when the
17779 raised exception is not handled in the same body. In case of a function this
17780 metric subtracts 1 from the number of exit points, because a function body
17781 must contain at least one @code{return} statement.
17783 @item ^--no-extra-exit-points^/NOEXTRA_EXIT_POINTS^
17784 Do not report the extra exit points for subprogram bodies
17788 @node Object-Oriented Metrics Control
17789 @subsubsection Object-Oriented Metrics Control
17790 @cindex Object-Oriented metrics control in @command{gnatmetric}
17793 @cindex Coupling metrics (in in @command{gnatmetric})
17794 Coupling metrics are object-oriented metrics that measure the
17795 dependencies between a given class (or a group of classes) and the
17796 ``external world'' (that is, the other classes in the program). In this
17797 subsection the term ``class'' is used in its
17798 traditional object-oriented programming sense
17799 (an instantiable module that contains data and/or method members).
17800 A @emph{category} (of classes)
17801 is a group of closely related classes that are reused and/or
17804 A class @code{K}'s @emph{efferent coupling} is the number of classes
17805 that @code{K} depends upon.
17806 A category's efferent coupling is the number of classes outside the
17807 category that the classes inside the category depend upon.
17809 A class @code{K}'s @emph{afferent coupling} is the number of classes
17810 that depend upon @code{K}.
17811 A category's afferent coupling is the number of classes outside the
17812 category that depend on classes belonging to the category.
17814 Ada's implementation of the object-oriented paradigm does not use the
17815 traditional class notion, so the definition of the coupling
17816 metrics for Ada maps the class and class category notions
17817 onto Ada constructs.
17819 For the coupling metrics, several kinds of modules -- a library package,
17820 a library generic package, and a library generic package instantiation --
17821 that define a tagged type or an interface type are
17822 considered to be a class. A category consists of a library package (or
17823 a library generic package) that defines a tagged or an interface type,
17824 together with all its descendant (generic) packages that define tagged
17825 or interface types. For any package counted as a class,
17826 its body and subunits (if any) are considered
17827 together with its spec when counting the dependencies, and coupling
17828 metrics are reported for spec units only. For dependencies
17829 between classes, the Ada semantic dependencies are considered.
17830 For coupling metrics, only dependencies on units that are considered as
17831 classes, are considered.
17833 When computing coupling metrics, @command{gnatmetric} counts only
17834 dependencies between units that are arguments of the gnatmetric call.
17835 Coupling metrics are program-wide (or project-wide) metrics, so to
17836 get a valid result, you should call @command{gnatmetric} for
17837 the whole set of sources that make up your program. It can be done
17838 by calling @command{gnatmetric} from the GNAT driver with @option{-U}
17839 option (see See @ref{The GNAT Driver and Project Files} for details.
17841 By default, all the coupling metrics are disabled. You can use the following
17842 switches to specify the coupling metrics to be computed and reported:
17847 @cindex @option{--package@var{x}} (@command{gnatmetric})
17848 @cindex @option{--no-package@var{x}} (@command{gnatmetric})
17849 @cindex @option{--category@var{x}} (@command{gnatmetric})
17850 @cindex @option{--no-category@var{x}} (@command{gnatmetric})
17854 @cindex @option{/COUPLING_METRICS} (@command{gnatmetric})
17857 @item ^--coupling-all^/COUPLING_METRICS=ALL^
17858 Report all the coupling metrics
17860 @item ^--no-coupling-all^/COUPLING_METRICS=NONE^
17861 Do not report any of metrics
17863 @item ^--package-efferent-coupling^/COUPLING_METRICS=PACKAGE_EFFERENT^
17864 Report package efferent coupling
17866 @item ^--no-package-efferent-coupling^/COUPLING_METRICS=NOPACKAGE_EFFERENT^
17867 Do not report package efferent coupling
17869 @item ^--package-afferent-coupling^/COUPLING_METRICS=PACKAGE_AFFERENT^
17870 Report package afferent coupling
17872 @item ^--no-package-afferent-coupling^/COUPLING_METRICS=NOPACKAGE_AFFERENT^
17873 Do not report package afferent coupling
17875 @item ^--category-efferent-coupling^/COUPLING_METRICS=CATEGORY_EFFERENT^
17876 Report category efferent coupling
17878 @item ^--no-category-efferent-coupling^/COUPLING_METRICS=NOCATEGORY_EFFERENT^
17879 Do not report category efferent coupling
17881 @item ^--category-afferent-coupling^/COUPLING_METRICS=CATEGORY_AFFERENT^
17882 Report category afferent coupling
17884 @item ^--no-category-afferent-coupling^/COUPLING_METRICS=NOCATEGORY_AFFERENT^
17885 Do not report category afferent coupling
17889 @node Other gnatmetric Switches
17890 @subsection Other @code{gnatmetric} Switches
17893 Additional @command{gnatmetric} switches are as follows:
17896 @item ^-files @var{filename}^/FILES=@var{filename}^
17897 @cindex @option{^-files^/FILES^} (@code{gnatmetric})
17898 Take the argument source files from the specified file. This file should be an
17899 ordinary text file containing file names separated by spaces or
17900 line breaks. You can use this switch more than once in the same call to
17901 @command{gnatmetric}. You also can combine this switch with
17902 an explicit list of files.
17904 @item ^-v^/VERBOSE^
17905 @cindex @option{^-v^/VERBOSE^} (@code{gnatmetric})
17907 @command{gnatmetric} generates version information and then
17908 a trace of sources being processed.
17910 @item ^-dv^/DEBUG_OUTPUT^
17911 @cindex @option{^-dv^/DEBUG_OUTPUT^} (@code{gnatmetric})
17913 @command{gnatmetric} generates various messages useful to understand what
17914 happens during the metrics computation
17917 @cindex @option{^-q^/QUIET^} (@code{gnatmetric})
17921 @node Generate project-wide metrics
17922 @subsection Generate project-wide metrics
17924 In order to compute metrics on all units of a given project, you can use
17925 the @command{gnat} driver along with the @option{-P} option:
17931 If the project @code{proj} depends upon other projects, you can compute
17932 the metrics on the project closure using the @option{-U} option:
17934 gnat metric -Pproj -U
17938 Finally, if not all the units are relevant to a particular main
17939 program in the project closure, you can generate metrics for the set
17940 of units needed to create a given main program (unit closure) using
17941 the @option{-U} option followed by the name of the main unit:
17943 gnat metric -Pproj -U main
17947 @c ***********************************
17948 @node File Name Krunching Using gnatkr
17949 @chapter File Name Krunching Using @code{gnatkr}
17953 This chapter discusses the method used by the compiler to shorten
17954 the default file names chosen for Ada units so that they do not
17955 exceed the maximum length permitted. It also describes the
17956 @code{gnatkr} utility that can be used to determine the result of
17957 applying this shortening.
17961 * Krunching Method::
17962 * Examples of gnatkr Usage::
17966 @section About @code{gnatkr}
17969 The default file naming rule in GNAT
17970 is that the file name must be derived from
17971 the unit name. The exact default rule is as follows:
17974 Take the unit name and replace all dots by hyphens.
17976 If such a replacement occurs in the
17977 second character position of a name, and the first character is
17978 ^@samp{a}, @samp{g}, @samp{s}, or @samp{i}, ^@samp{A}, @samp{G}, @samp{S}, or @samp{I},^
17979 then replace the dot by the character
17980 ^@samp{~} (tilde)^@samp{$} (dollar sign)^
17981 instead of a minus.
17983 The reason for this exception is to avoid clashes
17984 with the standard names for children of System, Ada, Interfaces,
17985 and GNAT, which use the prefixes
17986 ^@samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},^@samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},^
17989 The @option{^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{nn}}
17990 switch of the compiler activates a ``krunching''
17991 circuit that limits file names to nn characters (where nn is a decimal
17992 integer). For example, using OpenVMS,
17993 where the maximum file name length is
17994 39, the value of nn is usually set to 39, but if you want to generate
17995 a set of files that would be usable if ported to a system with some
17996 different maximum file length, then a different value can be specified.
17997 The default value of 39 for OpenVMS need not be specified.
17999 The @code{gnatkr} utility can be used to determine the krunched name for
18000 a given file, when krunched to a specified maximum length.
18003 @section Using @code{gnatkr}
18006 The @code{gnatkr} command has the form
18010 @c $ gnatkr @var{name} @ovar{length}
18011 @c Expanding @ovar macro inline (explanation in macro def comments)
18012 $ gnatkr @var{name} @r{[}@var{length}@r{]}
18018 $ gnatkr @var{name} /COUNT=nn
18023 @var{name} is the uncrunched file name, derived from the name of the unit
18024 in the standard manner described in the previous section (i.e., in particular
18025 all dots are replaced by hyphens). The file name may or may not have an
18026 extension (defined as a suffix of the form period followed by arbitrary
18027 characters other than period). If an extension is present then it will
18028 be preserved in the output. For example, when krunching @file{hellofile.ads}
18029 to eight characters, the result will be hellofil.ads.
18031 Note: for compatibility with previous versions of @code{gnatkr} dots may
18032 appear in the name instead of hyphens, but the last dot will always be
18033 taken as the start of an extension. So if @code{gnatkr} is given an argument
18034 such as @file{Hello.World.adb} it will be treated exactly as if the first
18035 period had been a hyphen, and for example krunching to eight characters
18036 gives the result @file{hellworl.adb}.
18038 Note that the result is always all lower case (except on OpenVMS where it is
18039 all upper case). Characters of the other case are folded as required.
18041 @var{length} represents the length of the krunched name. The default
18042 when no argument is given is ^8^39^ characters. A length of zero stands for
18043 unlimited, in other words do not chop except for system files where the
18044 implied crunching length is always eight characters.
18047 The output is the krunched name. The output has an extension only if the
18048 original argument was a file name with an extension.
18050 @node Krunching Method
18051 @section Krunching Method
18054 The initial file name is determined by the name of the unit that the file
18055 contains. The name is formed by taking the full expanded name of the
18056 unit and replacing the separating dots with hyphens and
18057 using ^lowercase^uppercase^
18058 for all letters, except that a hyphen in the second character position is
18059 replaced by a ^tilde^dollar sign^ if the first character is
18060 ^@samp{a}, @samp{i}, @samp{g}, or @samp{s}^@samp{A}, @samp{I}, @samp{G}, or @samp{S}^.
18061 The extension is @code{.ads} for a
18062 spec and @code{.adb} for a body.
18063 Krunching does not affect the extension, but the file name is shortened to
18064 the specified length by following these rules:
18068 The name is divided into segments separated by hyphens, tildes or
18069 underscores and all hyphens, tildes, and underscores are
18070 eliminated. If this leaves the name short enough, we are done.
18073 If the name is too long, the longest segment is located (left-most
18074 if there are two of equal length), and shortened by dropping
18075 its last character. This is repeated until the name is short enough.
18077 As an example, consider the krunching of @*@file{our-strings-wide_fixed.adb}
18078 to fit the name into 8 characters as required by some operating systems.
18081 our-strings-wide_fixed 22
18082 our strings wide fixed 19
18083 our string wide fixed 18
18084 our strin wide fixed 17
18085 our stri wide fixed 16
18086 our stri wide fixe 15
18087 our str wide fixe 14
18088 our str wid fixe 13
18094 Final file name: oustwifi.adb
18098 The file names for all predefined units are always krunched to eight
18099 characters. The krunching of these predefined units uses the following
18100 special prefix replacements:
18104 replaced by @file{^a^A^-}
18107 replaced by @file{^g^G^-}
18110 replaced by @file{^i^I^-}
18113 replaced by @file{^s^S^-}
18116 These system files have a hyphen in the second character position. That
18117 is why normal user files replace such a character with a
18118 ^tilde^dollar sign^, to
18119 avoid confusion with system file names.
18121 As an example of this special rule, consider
18122 @*@file{ada-strings-wide_fixed.adb}, which gets krunched as follows:
18125 ada-strings-wide_fixed 22
18126 a- strings wide fixed 18
18127 a- string wide fixed 17
18128 a- strin wide fixed 16
18129 a- stri wide fixed 15
18130 a- stri wide fixe 14
18131 a- str wide fixe 13
18137 Final file name: a-stwifi.adb
18141 Of course no file shortening algorithm can guarantee uniqueness over all
18142 possible unit names, and if file name krunching is used then it is your
18143 responsibility to ensure that no name clashes occur. The utility
18144 program @code{gnatkr} is supplied for conveniently determining the
18145 krunched name of a file.
18147 @node Examples of gnatkr Usage
18148 @section Examples of @code{gnatkr} Usage
18155 $ gnatkr very_long_unit_name.ads --> velounna.ads
18156 $ gnatkr grandparent-parent-child.ads --> grparchi.ads
18157 $ gnatkr Grandparent.Parent.Child.ads --> grparchi.ads
18158 $ gnatkr grandparent-parent-child --> grparchi
18160 $ gnatkr very_long_unit_name.ads/count=6 --> vlunna.ads
18161 $ gnatkr very_long_unit_name.ads/count=0 --> very_long_unit_name.ads
18164 @node Preprocessing Using gnatprep
18165 @chapter Preprocessing Using @code{gnatprep}
18169 This chapter discusses how to use GNAT's @code{gnatprep} utility for simple
18171 Although designed for use with GNAT, @code{gnatprep} does not depend on any
18172 special GNAT features.
18173 For further discussion of conditional compilation in general, see
18174 @ref{Conditional Compilation}.
18177 * Preprocessing Symbols::
18179 * Switches for gnatprep::
18180 * Form of Definitions File::
18181 * Form of Input Text for gnatprep::
18184 @node Preprocessing Symbols
18185 @section Preprocessing Symbols
18188 Preprocessing symbols are defined in definition files and referred to in
18189 sources to be preprocessed. A Preprocessing symbol is an identifier, following
18190 normal Ada (case-insensitive) rules for its syntax, with the restriction that
18191 all characters need to be in the ASCII set (no accented letters).
18193 @node Using gnatprep
18194 @section Using @code{gnatprep}
18197 To call @code{gnatprep} use
18200 @c $ gnatprep @ovar{switches} @var{infile} @var{outfile} @ovar{deffile}
18201 @c Expanding @ovar macro inline (explanation in macro def comments)
18202 $ gnatprep @r{[}@var{switches}@r{]} @var{infile} @var{outfile} @r{[}@var{deffile}@r{]}
18209 is an optional sequence of switches as described in the next section.
18212 is the full name of the input file, which is an Ada source
18213 file containing preprocessor directives.
18216 is the full name of the output file, which is an Ada source
18217 in standard Ada form. When used with GNAT, this file name will
18218 normally have an ads or adb suffix.
18221 is the full name of a text file containing definitions of
18222 preprocessing symbols to be referenced by the preprocessor. This argument is
18223 optional, and can be replaced by the use of the @option{-D} switch.
18227 @node Switches for gnatprep
18228 @section Switches for @code{gnatprep}
18233 @item ^-b^/BLANK_LINES^
18234 @cindex @option{^-b^/BLANK_LINES^} (@command{gnatprep})
18235 Causes both preprocessor lines and the lines deleted by
18236 preprocessing to be replaced by blank lines in the output source file,
18237 preserving line numbers in the output file.
18239 @item ^-c^/COMMENTS^
18240 @cindex @option{^-c^/COMMENTS^} (@command{gnatprep})
18241 Causes both preprocessor lines and the lines deleted
18242 by preprocessing to be retained in the output source as comments marked
18243 with the special string @code{"--! "}. This option will result in line numbers
18244 being preserved in the output file.
18246 @item ^-C^/REPLACE_IN_COMMENTS^
18247 @cindex @option{^-C^/REPLACE_IN_COMMENTS^} (@command{gnatprep})
18248 Causes comments to be scanned. Normally comments are ignored by gnatprep.
18249 If this option is specified, then comments are scanned and any $symbol
18250 substitutions performed as in program text. This is particularly useful
18251 when structured comments are used (e.g., when writing programs in the
18252 SPARK dialect of Ada). Note that this switch is not available when
18253 doing integrated preprocessing (it would be useless in this context
18254 since comments are ignored by the compiler in any case).
18256 @item ^-Dsymbol=value^/ASSOCIATE="symbol=value"^
18257 @cindex @option{^-D^/ASSOCIATE^} (@command{gnatprep})
18258 Defines a new preprocessing symbol, associated with value. If no value is given
18259 on the command line, then symbol is considered to be @code{True}. This switch
18260 can be used in place of a definition file.
18264 @cindex @option{/REMOVE} (@command{gnatprep})
18265 This is the default setting which causes lines deleted by preprocessing
18266 to be entirely removed from the output file.
18269 @item ^-r^/REFERENCE^
18270 @cindex @option{^-r^/REFERENCE^} (@command{gnatprep})
18271 Causes a @code{Source_Reference} pragma to be generated that
18272 references the original input file, so that error messages will use
18273 the file name of this original file. The use of this switch implies
18274 that preprocessor lines are not to be removed from the file, so its
18275 use will force @option{^-b^/BLANK_LINES^} mode if
18276 @option{^-c^/COMMENTS^}
18277 has not been specified explicitly.
18279 Note that if the file to be preprocessed contains multiple units, then
18280 it will be necessary to @code{gnatchop} the output file from
18281 @code{gnatprep}. If a @code{Source_Reference} pragma is present
18282 in the preprocessed file, it will be respected by
18283 @code{gnatchop ^-r^/REFERENCE^}
18284 so that the final chopped files will correctly refer to the original
18285 input source file for @code{gnatprep}.
18287 @item ^-s^/SYMBOLS^
18288 @cindex @option{^-s^/SYMBOLS^} (@command{gnatprep})
18289 Causes a sorted list of symbol names and values to be
18290 listed on the standard output file.
18292 @item ^-u^/UNDEFINED^
18293 @cindex @option{^-u^/UNDEFINED^} (@command{gnatprep})
18294 Causes undefined symbols to be treated as having the value FALSE in the context
18295 of a preprocessor test. In the absence of this option, an undefined symbol in
18296 a @code{#if} or @code{#elsif} test will be treated as an error.
18302 Note: if neither @option{-b} nor @option{-c} is present,
18303 then preprocessor lines and
18304 deleted lines are completely removed from the output, unless -r is
18305 specified, in which case -b is assumed.
18308 @node Form of Definitions File
18309 @section Form of Definitions File
18312 The definitions file contains lines of the form
18319 where symbol is a preprocessing symbol, and value is one of the following:
18323 Empty, corresponding to a null substitution
18325 A string literal using normal Ada syntax
18327 Any sequence of characters from the set
18328 (letters, digits, period, underline).
18332 Comment lines may also appear in the definitions file, starting with
18333 the usual @code{--},
18334 and comments may be added to the definitions lines.
18336 @node Form of Input Text for gnatprep
18337 @section Form of Input Text for @code{gnatprep}
18340 The input text may contain preprocessor conditional inclusion lines,
18341 as well as general symbol substitution sequences.
18343 The preprocessor conditional inclusion commands have the form
18348 #if @i{expression} @r{[}then@r{]}
18350 #elsif @i{expression} @r{[}then@r{]}
18352 #elsif @i{expression} @r{[}then@r{]}
18363 In this example, @i{expression} is defined by the following grammar:
18365 @i{expression} ::= <symbol>
18366 @i{expression} ::= <symbol> = "<value>"
18367 @i{expression} ::= <symbol> = <symbol>
18368 @i{expression} ::= <symbol> 'Defined
18369 @i{expression} ::= not @i{expression}
18370 @i{expression} ::= @i{expression} and @i{expression}
18371 @i{expression} ::= @i{expression} or @i{expression}
18372 @i{expression} ::= @i{expression} and then @i{expression}
18373 @i{expression} ::= @i{expression} or else @i{expression}
18374 @i{expression} ::= ( @i{expression} )
18377 The following restriction exists: it is not allowed to have "and" or "or"
18378 following "not" in the same expression without parentheses. For example, this
18385 This should be one of the following:
18393 For the first test (@i{expression} ::= <symbol>) the symbol must have
18394 either the value true or false, that is to say the right-hand of the
18395 symbol definition must be one of the (case-insensitive) literals
18396 @code{True} or @code{False}. If the value is true, then the
18397 corresponding lines are included, and if the value is false, they are
18400 The test (@i{expression} ::= <symbol> @code{'Defined}) is true only if
18401 the symbol has been defined in the definition file or by a @option{-D}
18402 switch on the command line. Otherwise, the test is false.
18404 The equality tests are case insensitive, as are all the preprocessor lines.
18406 If the symbol referenced is not defined in the symbol definitions file,
18407 then the effect depends on whether or not switch @option{-u}
18408 is specified. If so, then the symbol is treated as if it had the value
18409 false and the test fails. If this switch is not specified, then
18410 it is an error to reference an undefined symbol. It is also an error to
18411 reference a symbol that is defined with a value other than @code{True}
18414 The use of the @code{not} operator inverts the sense of this logical test.
18415 The @code{not} operator cannot be combined with the @code{or} or @code{and}
18416 operators, without parentheses. For example, "if not X or Y then" is not
18417 allowed, but "if (not X) or Y then" and "if not (X or Y) then" are.
18419 The @code{then} keyword is optional as shown
18421 The @code{#} must be the first non-blank character on a line, but
18422 otherwise the format is free form. Spaces or tabs may appear between
18423 the @code{#} and the keyword. The keywords and the symbols are case
18424 insensitive as in normal Ada code. Comments may be used on a
18425 preprocessor line, but other than that, no other tokens may appear on a
18426 preprocessor line. Any number of @code{elsif} clauses can be present,
18427 including none at all. The @code{else} is optional, as in Ada.
18429 The @code{#} marking the start of a preprocessor line must be the first
18430 non-blank character on the line, i.e., it must be preceded only by
18431 spaces or horizontal tabs.
18433 Symbol substitution outside of preprocessor lines is obtained by using
18441 anywhere within a source line, except in a comment or within a
18442 string literal. The identifier
18443 following the @code{$} must match one of the symbols defined in the symbol
18444 definition file, and the result is to substitute the value of the
18445 symbol in place of @code{$symbol} in the output file.
18447 Note that although the substitution of strings within a string literal
18448 is not possible, it is possible to have a symbol whose defined value is
18449 a string literal. So instead of setting XYZ to @code{hello} and writing:
18452 Header : String := "$XYZ";
18456 you should set XYZ to @code{"hello"} and write:
18459 Header : String := $XYZ;
18463 and then the substitution will occur as desired.
18466 @node The GNAT Run-Time Library Builder gnatlbr
18467 @chapter The GNAT Run-Time Library Builder @code{gnatlbr}
18469 @cindex Library builder
18472 @code{gnatlbr} is a tool for rebuilding the GNAT run time with user
18473 supplied configuration pragmas.
18476 * Running gnatlbr::
18477 * Switches for gnatlbr::
18478 * Examples of gnatlbr Usage::
18481 @node Running gnatlbr
18482 @section Running @code{gnatlbr}
18485 The @code{gnatlbr} command has the form
18488 $ GNAT LIBRARY /@r{[}CREATE@r{|}SET@r{|}DELETE@r{]}=directory @r{[}/CONFIG=file@r{]}
18491 @node Switches for gnatlbr
18492 @section Switches for @code{gnatlbr}
18495 @code{gnatlbr} recognizes the following switches:
18499 @item /CREATE=directory
18500 @cindex @code{/CREATE} (@code{gnatlbr})
18501 Create the new run-time library in the specified directory.
18503 @item /SET=directory
18504 @cindex @code{/SET} (@code{gnatlbr})
18505 Make the library in the specified directory the current run-time library.
18507 @item /DELETE=directory
18508 @cindex @code{/DELETE} (@code{gnatlbr})
18509 Delete the run-time library in the specified directory.
18512 @cindex @code{/CONFIG} (@code{gnatlbr})
18513 With /CREATE: Use the configuration pragmas in the specified file when
18514 building the library.
18516 With /SET: Use the configuration pragmas in the specified file when
18521 @node Examples of gnatlbr Usage
18522 @section Example of @code{gnatlbr} Usage
18525 Contents of VAXFLOAT.ADC:
18526 pragma Float_Representation (VAX_Float);
18528 $ GNAT LIBRARY /CREATE=[.VAXFLOAT] /CONFIG=VAXFLOAT.ADC
18530 GNAT LIBRARY rebuilds the run-time library in directory [.VAXFLOAT]
18535 @node The GNAT Library Browser gnatls
18536 @chapter The GNAT Library Browser @code{gnatls}
18538 @cindex Library browser
18541 @code{gnatls} is a tool that outputs information about compiled
18542 units. It gives the relationship between objects, unit names and source
18543 files. It can also be used to check the source dependencies of a unit
18544 as well as various characteristics.
18546 Note: to invoke @code{gnatls} with a project file, use the @code{gnat}
18547 driver (see @ref{The GNAT Driver and Project Files}).
18551 * Switches for gnatls::
18552 * Examples of gnatls Usage::
18555 @node Running gnatls
18556 @section Running @code{gnatls}
18559 The @code{gnatls} command has the form
18562 $ gnatls switches @var{object_or_ali_file}
18566 The main argument is the list of object or @file{ali} files
18567 (@pxref{The Ada Library Information Files})
18568 for which information is requested.
18570 In normal mode, without additional option, @code{gnatls} produces a
18571 four-column listing. Each line represents information for a specific
18572 object. The first column gives the full path of the object, the second
18573 column gives the name of the principal unit in this object, the third
18574 column gives the status of the source and the fourth column gives the
18575 full path of the source representing this unit.
18576 Here is a simple example of use:
18580 ^./^[]^demo1.o demo1 DIF demo1.adb
18581 ^./^[]^demo2.o demo2 OK demo2.adb
18582 ^./^[]^hello.o h1 OK hello.adb
18583 ^./^[]^instr-child.o instr.child MOK instr-child.adb
18584 ^./^[]^instr.o instr OK instr.adb
18585 ^./^[]^tef.o tef DIF tef.adb
18586 ^./^[]^text_io_example.o text_io_example OK text_io_example.adb
18587 ^./^[]^tgef.o tgef DIF tgef.adb
18591 The first line can be interpreted as follows: the main unit which is
18593 object file @file{demo1.o} is demo1, whose main source is in
18594 @file{demo1.adb}. Furthermore, the version of the source used for the
18595 compilation of demo1 has been modified (DIF). Each source file has a status
18596 qualifier which can be:
18599 @item OK (unchanged)
18600 The version of the source file used for the compilation of the
18601 specified unit corresponds exactly to the actual source file.
18603 @item MOK (slightly modified)
18604 The version of the source file used for the compilation of the
18605 specified unit differs from the actual source file but not enough to
18606 require recompilation. If you use gnatmake with the qualifier
18607 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}, a file marked
18608 MOK will not be recompiled.
18610 @item DIF (modified)
18611 No version of the source found on the path corresponds to the source
18612 used to build this object.
18614 @item ??? (file not found)
18615 No source file was found for this unit.
18617 @item HID (hidden, unchanged version not first on PATH)
18618 The version of the source that corresponds exactly to the source used
18619 for compilation has been found on the path but it is hidden by another
18620 version of the same source that has been modified.
18624 @node Switches for gnatls
18625 @section Switches for @code{gnatls}
18628 @code{gnatls} recognizes the following switches:
18632 @cindex @option{--version} @command{gnatls}
18633 Display Copyright and version, then exit disregarding all other options.
18636 @cindex @option{--help} @command{gnatls}
18637 If @option{--version} was not used, display usage, then exit disregarding
18640 @item ^-a^/ALL_UNITS^
18641 @cindex @option{^-a^/ALL_UNITS^} (@code{gnatls})
18642 Consider all units, including those of the predefined Ada library.
18643 Especially useful with @option{^-d^/DEPENDENCIES^}.
18645 @item ^-d^/DEPENDENCIES^
18646 @cindex @option{^-d^/DEPENDENCIES^} (@code{gnatls})
18647 List sources from which specified units depend on.
18649 @item ^-h^/OUTPUT=OPTIONS^
18650 @cindex @option{^-h^/OUTPUT=OPTIONS^} (@code{gnatls})
18651 Output the list of options.
18653 @item ^-o^/OUTPUT=OBJECTS^
18654 @cindex @option{^-o^/OUTPUT=OBJECTS^} (@code{gnatls})
18655 Only output information about object files.
18657 @item ^-s^/OUTPUT=SOURCES^
18658 @cindex @option{^-s^/OUTPUT=SOURCES^} (@code{gnatls})
18659 Only output information about source files.
18661 @item ^-u^/OUTPUT=UNITS^
18662 @cindex @option{^-u^/OUTPUT=UNITS^} (@code{gnatls})
18663 Only output information about compilation units.
18665 @item ^-files^/FILES^=@var{file}
18666 @cindex @option{^-files^/FILES^} (@code{gnatls})
18667 Take as arguments the files listed in text file @var{file}.
18668 Text file @var{file} may contain empty lines that are ignored.
18669 Each nonempty line should contain the name of an existing file.
18670 Several such switches may be specified simultaneously.
18672 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18673 @itemx ^-aI^/SOURCE_SEARCH=^@var{dir}
18674 @itemx ^-I^/SEARCH=^@var{dir}
18675 @itemx ^-I-^/NOCURRENT_DIRECTORY^
18677 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatls})
18678 @cindex @option{^-aI^/SOURCE_SEARCH^} (@code{gnatls})
18679 @cindex @option{^-I^/SEARCH^} (@code{gnatls})
18680 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatls})
18681 Source path manipulation. Same meaning as the equivalent @command{gnatmake}
18682 flags (@pxref{Switches for gnatmake}).
18684 @item --RTS=@var{rts-path}
18685 @cindex @option{--RTS} (@code{gnatls})
18686 Specifies the default location of the runtime library. Same meaning as the
18687 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
18689 @item ^-v^/OUTPUT=VERBOSE^
18690 @cindex @option{^-v^/OUTPUT=VERBOSE^} (@code{gnatls})
18691 Verbose mode. Output the complete source, object and project paths. Do not use
18692 the default column layout but instead use long format giving as much as
18693 information possible on each requested units, including special
18694 characteristics such as:
18697 @item Preelaborable
18698 The unit is preelaborable in the Ada sense.
18701 No elaboration code has been produced by the compiler for this unit.
18704 The unit is pure in the Ada sense.
18706 @item Elaborate_Body
18707 The unit contains a pragma Elaborate_Body.
18710 The unit contains a pragma Remote_Types.
18712 @item Shared_Passive
18713 The unit contains a pragma Shared_Passive.
18716 This unit is part of the predefined environment and cannot be modified
18719 @item Remote_Call_Interface
18720 The unit contains a pragma Remote_Call_Interface.
18726 @node Examples of gnatls Usage
18727 @section Example of @code{gnatls} Usage
18731 Example of using the verbose switch. Note how the source and
18732 object paths are affected by the -I switch.
18735 $ gnatls -v -I.. demo1.o
18737 GNATLS 5.03w (20041123-34)
18738 Copyright 1997-2004 Free Software Foundation, Inc.
18740 Source Search Path:
18741 <Current_Directory>
18743 /home/comar/local/adainclude/
18745 Object Search Path:
18746 <Current_Directory>
18748 /home/comar/local/lib/gcc-lib/x86-linux/3.4.3/adalib/
18750 Project Search Path:
18751 <Current_Directory>
18752 /home/comar/local/lib/gnat/
18757 Kind => subprogram body
18758 Flags => No_Elab_Code
18759 Source => demo1.adb modified
18763 The following is an example of use of the dependency list.
18764 Note the use of the -s switch
18765 which gives a straight list of source files. This can be useful for
18766 building specialized scripts.
18769 $ gnatls -d demo2.o
18770 ./demo2.o demo2 OK demo2.adb
18776 $ gnatls -d -s -a demo1.o
18778 /home/comar/local/adainclude/ada.ads
18779 /home/comar/local/adainclude/a-finali.ads
18780 /home/comar/local/adainclude/a-filico.ads
18781 /home/comar/local/adainclude/a-stream.ads
18782 /home/comar/local/adainclude/a-tags.ads
18785 /home/comar/local/adainclude/gnat.ads
18786 /home/comar/local/adainclude/g-io.ads
18788 /home/comar/local/adainclude/system.ads
18789 /home/comar/local/adainclude/s-exctab.ads
18790 /home/comar/local/adainclude/s-finimp.ads
18791 /home/comar/local/adainclude/s-finroo.ads
18792 /home/comar/local/adainclude/s-secsta.ads
18793 /home/comar/local/adainclude/s-stalib.ads
18794 /home/comar/local/adainclude/s-stoele.ads
18795 /home/comar/local/adainclude/s-stratt.ads
18796 /home/comar/local/adainclude/s-tasoli.ads
18797 /home/comar/local/adainclude/s-unstyp.ads
18798 /home/comar/local/adainclude/unchconv.ads
18804 GNAT LIST /DEPENDENCIES /OUTPUT=SOURCES /ALL_UNITS DEMO1.ADB
18806 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]ada.ads
18807 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-finali.ads
18808 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-filico.ads
18809 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-stream.ads
18810 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-tags.ads
18814 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]gnat.ads
18815 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]g-io.ads
18817 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]system.ads
18818 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-exctab.ads
18819 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finimp.ads
18820 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finroo.ads
18821 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-secsta.ads
18822 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stalib.ads
18823 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stoele.ads
18824 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stratt.ads
18825 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-tasoli.ads
18826 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-unstyp.ads
18827 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]unchconv.ads
18831 @node Cleaning Up Using gnatclean
18832 @chapter Cleaning Up Using @code{gnatclean}
18834 @cindex Cleaning tool
18837 @code{gnatclean} is a tool that allows the deletion of files produced by the
18838 compiler, binder and linker, including ALI files, object files, tree files,
18839 expanded source files, library files, interface copy source files, binder
18840 generated files and executable files.
18843 * Running gnatclean::
18844 * Switches for gnatclean::
18845 @c * Examples of gnatclean Usage::
18848 @node Running gnatclean
18849 @section Running @code{gnatclean}
18852 The @code{gnatclean} command has the form:
18855 $ gnatclean switches @var{names}
18859 @var{names} is a list of source file names. Suffixes @code{.^ads^ADS^} and
18860 @code{^adb^ADB^} may be omitted. If a project file is specified using switch
18861 @code{^-P^/PROJECT_FILE=^}, then @var{names} may be completely omitted.
18864 In normal mode, @code{gnatclean} delete the files produced by the compiler and,
18865 if switch @code{^-c^/COMPILER_FILES_ONLY^} is not specified, by the binder and
18866 the linker. In informative-only mode, specified by switch
18867 @code{^-n^/NODELETE^}, the list of files that would have been deleted in
18868 normal mode is listed, but no file is actually deleted.
18870 @node Switches for gnatclean
18871 @section Switches for @code{gnatclean}
18874 @code{gnatclean} recognizes the following switches:
18878 @cindex @option{--version} @command{gnatclean}
18879 Display Copyright and version, then exit disregarding all other options.
18882 @cindex @option{--help} @command{gnatclean}
18883 If @option{--version} was not used, display usage, then exit disregarding
18886 @item ^-c^/COMPILER_FILES_ONLY^
18887 @cindex @option{^-c^/COMPILER_FILES_ONLY^} (@code{gnatclean})
18888 Only attempt to delete the files produced by the compiler, not those produced
18889 by the binder or the linker. The files that are not to be deleted are library
18890 files, interface copy files, binder generated files and executable files.
18892 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
18893 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@code{gnatclean})
18894 Indicate that ALI and object files should normally be found in directory
18897 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
18898 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@code{gnatclean})
18899 When using project files, if some errors or warnings are detected during
18900 parsing and verbose mode is not in effect (no use of switch
18901 ^-v^/VERBOSE^), then error lines start with the full path name of the project
18902 file, rather than its simple file name.
18905 @cindex @option{^-h^/HELP^} (@code{gnatclean})
18906 Output a message explaining the usage of @code{^gnatclean^gnatclean^}.
18908 @item ^-n^/NODELETE^
18909 @cindex @option{^-n^/NODELETE^} (@code{gnatclean})
18910 Informative-only mode. Do not delete any files. Output the list of the files
18911 that would have been deleted if this switch was not specified.
18913 @item ^-P^/PROJECT_FILE=^@var{project}
18914 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatclean})
18915 Use project file @var{project}. Only one such switch can be used.
18916 When cleaning a project file, the files produced by the compilation of the
18917 immediate sources or inherited sources of the project files are to be
18918 deleted. This is not depending on the presence or not of executable names
18919 on the command line.
18922 @cindex @option{^-q^/QUIET^} (@code{gnatclean})
18923 Quiet output. If there are no errors, do not output anything, except in
18924 verbose mode (switch ^-v^/VERBOSE^) or in informative-only mode
18925 (switch ^-n^/NODELETE^).
18927 @item ^-r^/RECURSIVE^
18928 @cindex @option{^-r^/RECURSIVE^} (@code{gnatclean})
18929 When a project file is specified (using switch ^-P^/PROJECT_FILE=^),
18930 clean all imported and extended project files, recursively. If this switch
18931 is not specified, only the files related to the main project file are to be
18932 deleted. This switch has no effect if no project file is specified.
18934 @item ^-v^/VERBOSE^
18935 @cindex @option{^-v^/VERBOSE^} (@code{gnatclean})
18938 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
18939 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (@code{gnatclean})
18940 Indicates the verbosity of the parsing of GNAT project files.
18941 @xref{Switches Related to Project Files}.
18943 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
18944 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (@code{gnatclean})
18945 Indicates that external variable @var{name} has the value @var{value}.
18946 The Project Manager will use this value for occurrences of
18947 @code{external(name)} when parsing the project file.
18948 @xref{Switches Related to Project Files}.
18950 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18951 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatclean})
18952 When searching for ALI and object files, look in directory
18955 @item ^-I^/SEARCH=^@var{dir}
18956 @cindex @option{^-I^/SEARCH^} (@code{gnatclean})
18957 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}}.
18959 @item ^-I-^/NOCURRENT_DIRECTORY^
18960 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatclean})
18961 @cindex Source files, suppressing search
18962 Do not look for ALI or object files in the directory
18963 where @code{gnatclean} was invoked.
18967 @c @node Examples of gnatclean Usage
18968 @c @section Examples of @code{gnatclean} Usage
18971 @node GNAT and Libraries
18972 @chapter GNAT and Libraries
18973 @cindex Library, building, installing, using
18976 This chapter describes how to build and use libraries with GNAT, and also shows
18977 how to recompile the GNAT run-time library. You should be familiar with the
18978 Project Manager facility (@pxref{GNAT Project Manager}) before reading this
18982 * Introduction to Libraries in GNAT::
18983 * General Ada Libraries::
18984 * Stand-alone Ada Libraries::
18985 * Rebuilding the GNAT Run-Time Library::
18988 @node Introduction to Libraries in GNAT
18989 @section Introduction to Libraries in GNAT
18992 A library is, conceptually, a collection of objects which does not have its
18993 own main thread of execution, but rather provides certain services to the
18994 applications that use it. A library can be either statically linked with the
18995 application, in which case its code is directly included in the application,
18996 or, on platforms that support it, be dynamically linked, in which case
18997 its code is shared by all applications making use of this library.
18999 GNAT supports both types of libraries.
19000 In the static case, the compiled code can be provided in different ways. The
19001 simplest approach is to provide directly the set of objects resulting from
19002 compilation of the library source files. Alternatively, you can group the
19003 objects into an archive using whatever commands are provided by the operating
19004 system. For the latter case, the objects are grouped into a shared library.
19006 In the GNAT environment, a library has three types of components:
19012 @xref{The Ada Library Information Files}.
19014 Object files, an archive or a shared library.
19018 A GNAT library may expose all its source files, which is useful for
19019 documentation purposes. Alternatively, it may expose only the units needed by
19020 an external user to make use of the library. That is to say, the specs
19021 reflecting the library services along with all the units needed to compile
19022 those specs, which can include generic bodies or any body implementing an
19023 inlined routine. In the case of @emph{stand-alone libraries} those exposed
19024 units are called @emph{interface units} (@pxref{Stand-alone Ada Libraries}).
19026 All compilation units comprising an application, including those in a library,
19027 need to be elaborated in an order partially defined by Ada's semantics. GNAT
19028 computes the elaboration order from the @file{ALI} files and this is why they
19029 constitute a mandatory part of GNAT libraries.
19030 @emph{Stand-alone libraries} are the exception to this rule because a specific
19031 library elaboration routine is produced independently of the application(s)
19034 @node General Ada Libraries
19035 @section General Ada Libraries
19038 * Building a library::
19039 * Installing a library::
19040 * Using a library::
19043 @node Building a library
19044 @subsection Building a library
19047 The easiest way to build a library is to use the Project Manager,
19048 which supports a special type of project called a @emph{Library Project}
19049 (@pxref{Library Projects}).
19051 A project is considered a library project, when two project-level attributes
19052 are defined in it: @code{Library_Name} and @code{Library_Dir}. In order to
19053 control different aspects of library configuration, additional optional
19054 project-level attributes can be specified:
19057 This attribute controls whether the library is to be static or dynamic
19059 @item Library_Version
19060 This attribute specifies the library version; this value is used
19061 during dynamic linking of shared libraries to determine if the currently
19062 installed versions of the binaries are compatible.
19064 @item Library_Options
19066 These attributes specify additional low-level options to be used during
19067 library generation, and redefine the actual application used to generate
19072 The GNAT Project Manager takes full care of the library maintenance task,
19073 including recompilation of the source files for which objects do not exist
19074 or are not up to date, assembly of the library archive, and installation of
19075 the library (i.e., copying associated source, object and @file{ALI} files
19076 to the specified location).
19078 Here is a simple library project file:
19079 @smallexample @c ada
19081 for Source_Dirs use ("src1", "src2");
19082 for Object_Dir use "obj";
19083 for Library_Name use "mylib";
19084 for Library_Dir use "lib";
19085 for Library_Kind use "dynamic";
19090 and the compilation command to build and install the library:
19092 @smallexample @c ada
19093 $ gnatmake -Pmy_lib
19097 It is not entirely trivial to perform manually all the steps required to
19098 produce a library. We recommend that you use the GNAT Project Manager
19099 for this task. In special cases where this is not desired, the necessary
19100 steps are discussed below.
19102 There are various possibilities for compiling the units that make up the
19103 library: for example with a Makefile (@pxref{Using the GNU make Utility}) or
19104 with a conventional script. For simple libraries, it is also possible to create
19105 a dummy main program which depends upon all the packages that comprise the
19106 interface of the library. This dummy main program can then be given to
19107 @command{gnatmake}, which will ensure that all necessary objects are built.
19109 After this task is accomplished, you should follow the standard procedure
19110 of the underlying operating system to produce the static or shared library.
19112 Here is an example of such a dummy program:
19113 @smallexample @c ada
19115 with My_Lib.Service1;
19116 with My_Lib.Service2;
19117 with My_Lib.Service3;
19118 procedure My_Lib_Dummy is
19126 Here are the generic commands that will build an archive or a shared library.
19129 # compiling the library
19130 $ gnatmake -c my_lib_dummy.adb
19132 # we don't need the dummy object itself
19133 $ rm my_lib_dummy.o my_lib_dummy.ali
19135 # create an archive with the remaining objects
19136 $ ar rc libmy_lib.a *.o
19137 # some systems may require "ranlib" to be run as well
19139 # or create a shared library
19140 $ gcc -shared -o libmy_lib.so *.o
19141 # some systems may require the code to have been compiled with -fPIC
19143 # remove the object files that are now in the library
19146 # Make the ALI files read-only so that gnatmake will not try to
19147 # regenerate the objects that are in the library
19152 Please note that the library must have a name of the form @file{lib@var{xxx}.a}
19153 or @file{lib@var{xxx}.so} (or @file{lib@var{xxx}.dll} on Windows) in order to
19154 be accessed by the directive @option{-l@var{xxx}} at link time.
19156 @node Installing a library
19157 @subsection Installing a library
19158 @cindex @code{ADA_PROJECT_PATH}
19159 @cindex @code{GPR_PROJECT_PATH}
19162 If you use project files, library installation is part of the library build
19163 process. Thus no further action is needed in order to make use of the
19164 libraries that are built as part of the general application build. A usable
19165 version of the library is installed in the directory specified by the
19166 @code{Library_Dir} attribute of the library project file.
19168 You may want to install a library in a context different from where the library
19169 is built. This situation arises with third party suppliers, who may want
19170 to distribute a library in binary form where the user is not expected to be
19171 able to recompile the library. The simplest option in this case is to provide
19172 a project file slightly different from the one used to build the library, by
19173 using the @code{externally_built} attribute. For instance, the project
19174 file used to build the library in the previous section can be changed into the
19175 following one when the library is installed:
19177 @smallexample @c projectfile
19179 for Source_Dirs use ("src1", "src2");
19180 for Library_Name use "mylib";
19181 for Library_Dir use "lib";
19182 for Library_Kind use "dynamic";
19183 for Externally_Built use "true";
19188 This project file assumes that the directories @file{src1},
19189 @file{src2}, and @file{lib} exist in
19190 the directory containing the project file. The @code{externally_built}
19191 attribute makes it clear to the GNAT builder that it should not attempt to
19192 recompile any of the units from this library. It allows the library provider to
19193 restrict the source set to the minimum necessary for clients to make use of the
19194 library as described in the first section of this chapter. It is the
19195 responsibility of the library provider to install the necessary sources, ALI
19196 files and libraries in the directories mentioned in the project file. For
19197 convenience, the user's library project file should be installed in a location
19198 that will be searched automatically by the GNAT
19199 builder. These are the directories referenced in the @env{GPR_PROJECT_PATH}
19200 environment variable (@pxref{Importing Projects}), and also the default GNAT
19201 library location that can be queried with @command{gnatls -v} and is usually of
19202 the form $gnat_install_root/lib/gnat.
19204 When project files are not an option, it is also possible, but not recommended,
19205 to install the library so that the sources needed to use the library are on the
19206 Ada source path and the ALI files & libraries be on the Ada Object path (see
19207 @ref{Search Paths and the Run-Time Library (RTL)}. Alternatively, the system
19208 administrator can place general-purpose libraries in the default compiler
19209 paths, by specifying the libraries' location in the configuration files
19210 @file{ada_source_path} and @file{ada_object_path}. These configuration files
19211 must be located in the GNAT installation tree at the same place as the gcc spec
19212 file. The location of the gcc spec file can be determined as follows:
19218 The configuration files mentioned above have a simple format: each line
19219 must contain one unique directory name.
19220 Those names are added to the corresponding path
19221 in their order of appearance in the file. The names can be either absolute
19222 or relative; in the latter case, they are relative to where theses files
19225 The files @file{ada_source_path} and @file{ada_object_path} might not be
19227 GNAT installation, in which case, GNAT will look for its run-time library in
19228 the directories @file{adainclude} (for the sources) and @file{adalib} (for the
19229 objects and @file{ALI} files). When the files exist, the compiler does not
19230 look in @file{adainclude} and @file{adalib}, and thus the
19231 @file{ada_source_path} file
19232 must contain the location for the GNAT run-time sources (which can simply
19233 be @file{adainclude}). In the same way, the @file{ada_object_path} file must
19234 contain the location for the GNAT run-time objects (which can simply
19237 You can also specify a new default path to the run-time library at compilation
19238 time with the switch @option{--RTS=rts-path}. You can thus choose / change
19239 the run-time library you want your program to be compiled with. This switch is
19240 recognized by @command{gcc}, @command{gnatmake}, @command{gnatbind},
19241 @command{gnatls}, @command{gnatfind} and @command{gnatxref}.
19243 It is possible to install a library before or after the standard GNAT
19244 library, by reordering the lines in the configuration files. In general, a
19245 library must be installed before the GNAT library if it redefines
19248 @node Using a library
19249 @subsection Using a library
19251 @noindent Once again, the project facility greatly simplifies the use of
19252 libraries. In this context, using a library is just a matter of adding a
19253 @code{with} clause in the user project. For instance, to make use of the
19254 library @code{My_Lib} shown in examples in earlier sections, you can
19257 @smallexample @c projectfile
19264 Even if you have a third-party, non-Ada library, you can still use GNAT's
19265 Project Manager facility to provide a wrapper for it. For example, the
19266 following project, when @code{with}ed by your main project, will link with the
19267 third-party library @file{liba.a}:
19269 @smallexample @c projectfile
19272 for Externally_Built use "true";
19273 for Source_Files use ();
19274 for Library_Dir use "lib";
19275 for Library_Name use "a";
19276 for Library_Kind use "static";
19280 This is an alternative to the use of @code{pragma Linker_Options}. It is
19281 especially interesting in the context of systems with several interdependent
19282 static libraries where finding a proper linker order is not easy and best be
19283 left to the tools having visibility over project dependence information.
19286 In order to use an Ada library manually, you need to make sure that this
19287 library is on both your source and object path
19288 (see @ref{Search Paths and the Run-Time Library (RTL)}
19289 and @ref{Search Paths for gnatbind}). Furthermore, when the objects are grouped
19290 in an archive or a shared library, you need to specify the desired
19291 library at link time.
19293 For example, you can use the library @file{mylib} installed in
19294 @file{/dir/my_lib_src} and @file{/dir/my_lib_obj} with the following commands:
19297 $ gnatmake -aI/dir/my_lib_src -aO/dir/my_lib_obj my_appl \
19302 This can be expressed more simply:
19307 when the following conditions are met:
19310 @file{/dir/my_lib_src} has been added by the user to the environment
19311 variable @env{ADA_INCLUDE_PATH}, or by the administrator to the file
19312 @file{ada_source_path}
19314 @file{/dir/my_lib_obj} has been added by the user to the environment
19315 variable @env{ADA_OBJECTS_PATH}, or by the administrator to the file
19316 @file{ada_object_path}
19318 a pragma @code{Linker_Options} has been added to one of the sources.
19321 @smallexample @c ada
19322 pragma Linker_Options ("-lmy_lib");
19326 @node Stand-alone Ada Libraries
19327 @section Stand-alone Ada Libraries
19328 @cindex Stand-alone library, building, using
19331 * Introduction to Stand-alone Libraries::
19332 * Building a Stand-alone Library::
19333 * Creating a Stand-alone Library to be used in a non-Ada context::
19334 * Restrictions in Stand-alone Libraries::
19337 @node Introduction to Stand-alone Libraries
19338 @subsection Introduction to Stand-alone Libraries
19341 A Stand-alone Library (abbreviated ``SAL'') is a library that contains the
19343 elaborate the Ada units that are included in the library. In contrast with
19344 an ordinary library, which consists of all sources, objects and @file{ALI}
19346 library, a SAL may specify a restricted subset of compilation units
19347 to serve as a library interface. In this case, the fully
19348 self-sufficient set of files will normally consist of an objects
19349 archive, the sources of interface units' specs, and the @file{ALI}
19350 files of interface units.
19351 If an interface spec contains a generic unit or an inlined subprogram,
19353 source must also be provided; if the units that must be provided in the source
19354 form depend on other units, the source and @file{ALI} files of those must
19357 The main purpose of a SAL is to minimize the recompilation overhead of client
19358 applications when a new version of the library is installed. Specifically,
19359 if the interface sources have not changed, client applications do not need to
19360 be recompiled. If, furthermore, a SAL is provided in the shared form and its
19361 version, controlled by @code{Library_Version} attribute, is not changed,
19362 then the clients do not need to be relinked.
19364 SALs also allow the library providers to minimize the amount of library source
19365 text exposed to the clients. Such ``information hiding'' might be useful or
19366 necessary for various reasons.
19368 Stand-alone libraries are also well suited to be used in an executable whose
19369 main routine is not written in Ada.
19371 @node Building a Stand-alone Library
19372 @subsection Building a Stand-alone Library
19375 GNAT's Project facility provides a simple way of building and installing
19376 stand-alone libraries; see @ref{Stand-alone Library Projects}.
19377 To be a Stand-alone Library Project, in addition to the two attributes
19378 that make a project a Library Project (@code{Library_Name} and
19379 @code{Library_Dir}; see @ref{Library Projects}), the attribute
19380 @code{Library_Interface} must be defined. For example:
19382 @smallexample @c projectfile
19384 for Library_Dir use "lib_dir";
19385 for Library_Name use "dummy";
19386 for Library_Interface use ("int1", "int1.child");
19391 Attribute @code{Library_Interface} has a non-empty string list value,
19392 each string in the list designating a unit contained in an immediate source
19393 of the project file.
19395 When a Stand-alone Library is built, first the binder is invoked to build
19396 a package whose name depends on the library name
19397 (@file{^b~dummy.ads/b^B$DUMMY.ADS/B^} in the example above).
19398 This binder-generated package includes initialization and
19399 finalization procedures whose
19400 names depend on the library name (@code{dummyinit} and @code{dummyfinal}
19402 above). The object corresponding to this package is included in the library.
19404 You must ensure timely (e.g., prior to any use of interfaces in the SAL)
19405 calling of these procedures if a static SAL is built, or if a shared SAL
19407 with the project-level attribute @code{Library_Auto_Init} set to
19410 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
19411 (those that are listed in attribute @code{Library_Interface}) are copied to
19412 the Library Directory. As a consequence, only the Interface Units may be
19413 imported from Ada units outside of the library. If other units are imported,
19414 the binding phase will fail.
19416 The attribute @code{Library_Src_Dir} may be specified for a
19417 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
19418 single string value. Its value must be the path (absolute or relative to the
19419 project directory) of an existing directory. This directory cannot be the
19420 object directory or one of the source directories, but it can be the same as
19421 the library directory. The sources of the Interface
19422 Units of the library that are needed by an Ada client of the library will be
19423 copied to the designated directory, called the Interface Copy directory.
19424 These sources include the specs of the Interface Units, but they may also
19425 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
19426 are used, or when there is a generic unit in the spec. Before the sources
19427 are copied to the Interface Copy directory, an attempt is made to delete all
19428 files in the Interface Copy directory.
19430 Building stand-alone libraries by hand is somewhat tedious, but for those
19431 occasions when it is necessary here are the steps that you need to perform:
19434 Compile all library sources.
19437 Invoke the binder with the switch @option{-n} (No Ada main program),
19438 with all the @file{ALI} files of the interfaces, and
19439 with the switch @option{-L} to give specific names to the @code{init}
19440 and @code{final} procedures. For example:
19442 gnatbind -n int1.ali int2.ali -Lsal1
19446 Compile the binder generated file:
19452 Link the dynamic library with all the necessary object files,
19453 indicating to the linker the names of the @code{init} (and possibly
19454 @code{final}) procedures for automatic initialization (and finalization).
19455 The built library should be placed in a directory different from
19456 the object directory.
19459 Copy the @code{ALI} files of the interface to the library directory,
19460 add in this copy an indication that it is an interface to a SAL
19461 (i.e., add a word @option{SL} on the line in the @file{ALI} file that starts
19462 with letter ``P'') and make the modified copy of the @file{ALI} file
19467 Using SALs is not different from using other libraries
19468 (see @ref{Using a library}).
19470 @node Creating a Stand-alone Library to be used in a non-Ada context
19471 @subsection Creating a Stand-alone Library to be used in a non-Ada context
19474 It is easy to adapt the SAL build procedure discussed above for use of a SAL in
19477 The only extra step required is to ensure that library interface subprograms
19478 are compatible with the main program, by means of @code{pragma Export}
19479 or @code{pragma Convention}.
19481 Here is an example of simple library interface for use with C main program:
19483 @smallexample @c ada
19484 package My_Package is
19486 procedure Do_Something;
19487 pragma Export (C, Do_Something, "do_something");
19489 procedure Do_Something_Else;
19490 pragma Export (C, Do_Something_Else, "do_something_else");
19496 On the foreign language side, you must provide a ``foreign'' view of the
19497 library interface; remember that it should contain elaboration routines in
19498 addition to interface subprograms.
19500 The example below shows the content of @code{mylib_interface.h} (note
19501 that there is no rule for the naming of this file, any name can be used)
19503 /* the library elaboration procedure */
19504 extern void mylibinit (void);
19506 /* the library finalization procedure */
19507 extern void mylibfinal (void);
19509 /* the interface exported by the library */
19510 extern void do_something (void);
19511 extern void do_something_else (void);
19515 Libraries built as explained above can be used from any program, provided
19516 that the elaboration procedures (named @code{mylibinit} in the previous
19517 example) are called before the library services are used. Any number of
19518 libraries can be used simultaneously, as long as the elaboration
19519 procedure of each library is called.
19521 Below is an example of a C program that uses the @code{mylib} library.
19524 #include "mylib_interface.h"
19529 /* First, elaborate the library before using it */
19532 /* Main program, using the library exported entities */
19534 do_something_else ();
19536 /* Library finalization at the end of the program */
19543 Note that invoking any library finalization procedure generated by
19544 @code{gnatbind} shuts down the Ada run-time environment.
19546 finalization of all Ada libraries must be performed at the end of the program.
19547 No call to these libraries or to the Ada run-time library should be made
19548 after the finalization phase.
19550 @node Restrictions in Stand-alone Libraries
19551 @subsection Restrictions in Stand-alone Libraries
19554 The pragmas listed below should be used with caution inside libraries,
19555 as they can create incompatibilities with other Ada libraries:
19557 @item pragma @code{Locking_Policy}
19558 @item pragma @code{Queuing_Policy}
19559 @item pragma @code{Task_Dispatching_Policy}
19560 @item pragma @code{Unreserve_All_Interrupts}
19564 When using a library that contains such pragmas, the user must make sure
19565 that all libraries use the same pragmas with the same values. Otherwise,
19566 @code{Program_Error} will
19567 be raised during the elaboration of the conflicting
19568 libraries. The usage of these pragmas and its consequences for the user
19569 should therefore be well documented.
19571 Similarly, the traceback in the exception occurrence mechanism should be
19572 enabled or disabled in a consistent manner across all libraries.
19573 Otherwise, Program_Error will be raised during the elaboration of the
19574 conflicting libraries.
19576 If the @code{Version} or @code{Body_Version}
19577 attributes are used inside a library, then you need to
19578 perform a @code{gnatbind} step that specifies all @file{ALI} files in all
19579 libraries, so that version identifiers can be properly computed.
19580 In practice these attributes are rarely used, so this is unlikely
19581 to be a consideration.
19583 @node Rebuilding the GNAT Run-Time Library
19584 @section Rebuilding the GNAT Run-Time Library
19585 @cindex GNAT Run-Time Library, rebuilding
19586 @cindex Building the GNAT Run-Time Library
19587 @cindex Rebuilding the GNAT Run-Time Library
19588 @cindex Run-Time Library, rebuilding
19591 It may be useful to recompile the GNAT library in various contexts, the
19592 most important one being the use of partition-wide configuration pragmas
19593 such as @code{Normalize_Scalars}. A special Makefile called
19594 @code{Makefile.adalib} is provided to that effect and can be found in
19595 the directory containing the GNAT library. The location of this
19596 directory depends on the way the GNAT environment has been installed and can
19597 be determined by means of the command:
19604 The last entry in the object search path usually contains the
19605 gnat library. This Makefile contains its own documentation and in
19606 particular the set of instructions needed to rebuild a new library and
19609 @node Using the GNU make Utility
19610 @chapter Using the GNU @code{make} Utility
19614 This chapter offers some examples of makefiles that solve specific
19615 problems. It does not explain how to write a makefile (@pxref{Top,, GNU
19616 make, make, GNU @code{make}}), nor does it try to replace the
19617 @command{gnatmake} utility (@pxref{The GNAT Make Program gnatmake}).
19619 All the examples in this section are specific to the GNU version of
19620 make. Although @command{make} is a standard utility, and the basic language
19621 is the same, these examples use some advanced features found only in
19625 * Using gnatmake in a Makefile::
19626 * Automatically Creating a List of Directories::
19627 * Generating the Command Line Switches::
19628 * Overcoming Command Line Length Limits::
19631 @node Using gnatmake in a Makefile
19632 @section Using gnatmake in a Makefile
19637 Complex project organizations can be handled in a very powerful way by
19638 using GNU make combined with gnatmake. For instance, here is a Makefile
19639 which allows you to build each subsystem of a big project into a separate
19640 shared library. Such a makefile allows you to significantly reduce the link
19641 time of very big applications while maintaining full coherence at
19642 each step of the build process.
19644 The list of dependencies are handled automatically by
19645 @command{gnatmake}. The Makefile is simply used to call gnatmake in each of
19646 the appropriate directories.
19648 Note that you should also read the example on how to automatically
19649 create the list of directories
19650 (@pxref{Automatically Creating a List of Directories})
19651 which might help you in case your project has a lot of subdirectories.
19656 @font@heightrm=cmr8
19659 ## This Makefile is intended to be used with the following directory
19661 ## - The sources are split into a series of csc (computer software components)
19662 ## Each of these csc is put in its own directory.
19663 ## Their name are referenced by the directory names.
19664 ## They will be compiled into shared library (although this would also work
19665 ## with static libraries
19666 ## - The main program (and possibly other packages that do not belong to any
19667 ## csc is put in the top level directory (where the Makefile is).
19668 ## toplevel_dir __ first_csc (sources) __ lib (will contain the library)
19669 ## \_ second_csc (sources) __ lib (will contain the library)
19671 ## Although this Makefile is build for shared library, it is easy to modify
19672 ## to build partial link objects instead (modify the lines with -shared and
19675 ## With this makefile, you can change any file in the system or add any new
19676 ## file, and everything will be recompiled correctly (only the relevant shared
19677 ## objects will be recompiled, and the main program will be re-linked).
19679 # The list of computer software component for your project. This might be
19680 # generated automatically.
19683 # Name of the main program (no extension)
19686 # If we need to build objects with -fPIC, uncomment the following line
19689 # The following variable should give the directory containing libgnat.so
19690 # You can get this directory through 'gnatls -v'. This is usually the last
19691 # directory in the Object_Path.
19694 # The directories for the libraries
19695 # (This macro expands the list of CSC to the list of shared libraries, you
19696 # could simply use the expanded form:
19697 # LIB_DIR=aa/lib/libaa.so bb/lib/libbb.so cc/lib/libcc.so
19698 LIB_DIR=$@{foreach dir,$@{CSC_LIST@},$@{dir@}/lib/lib$@{dir@}.so@}
19700 $@{MAIN@}: objects $@{LIB_DIR@}
19701 gnatbind $@{MAIN@} $@{CSC_LIST:%=-aO%/lib@} -shared
19702 gnatlink $@{MAIN@} $@{CSC_LIST:%=-l%@}
19705 # recompile the sources
19706 gnatmake -c -i $@{MAIN@}.adb $@{NEED_FPIC@} $@{CSC_LIST:%=-I%@}
19708 # Note: In a future version of GNAT, the following commands will be simplified
19709 # by a new tool, gnatmlib
19711 mkdir -p $@{dir $@@ @}
19712 cd $@{dir $@@ @} && gcc -shared -o $@{notdir $@@ @} ../*.o -L$@{GLIB@} -lgnat
19713 cd $@{dir $@@ @} && cp -f ../*.ali .
19715 # The dependencies for the modules
19716 # Note that we have to force the expansion of *.o, since in some cases
19717 # make won't be able to do it itself.
19718 aa/lib/libaa.so: $@{wildcard aa/*.o@}
19719 bb/lib/libbb.so: $@{wildcard bb/*.o@}
19720 cc/lib/libcc.so: $@{wildcard cc/*.o@}
19722 # Make sure all of the shared libraries are in the path before starting the
19725 LD_LIBRARY_PATH=`pwd`/aa/lib:`pwd`/bb/lib:`pwd`/cc/lib ./$@{MAIN@}
19728 $@{RM@} -rf $@{CSC_LIST:%=%/lib@}
19729 $@{RM@} $@{CSC_LIST:%=%/*.ali@}
19730 $@{RM@} $@{CSC_LIST:%=%/*.o@}
19731 $@{RM@} *.o *.ali $@{MAIN@}
19734 @node Automatically Creating a List of Directories
19735 @section Automatically Creating a List of Directories
19738 In most makefiles, you will have to specify a list of directories, and
19739 store it in a variable. For small projects, it is often easier to
19740 specify each of them by hand, since you then have full control over what
19741 is the proper order for these directories, which ones should be
19744 However, in larger projects, which might involve hundreds of
19745 subdirectories, it might be more convenient to generate this list
19748 The example below presents two methods. The first one, although less
19749 general, gives you more control over the list. It involves wildcard
19750 characters, that are automatically expanded by @command{make}. Its
19751 shortcoming is that you need to explicitly specify some of the
19752 organization of your project, such as for instance the directory tree
19753 depth, whether some directories are found in a separate tree, @enddots{}
19755 The second method is the most general one. It requires an external
19756 program, called @command{find}, which is standard on all Unix systems. All
19757 the directories found under a given root directory will be added to the
19763 @font@heightrm=cmr8
19766 # The examples below are based on the following directory hierarchy:
19767 # All the directories can contain any number of files
19768 # ROOT_DIRECTORY -> a -> aa -> aaa
19771 # -> b -> ba -> baa
19774 # This Makefile creates a variable called DIRS, that can be reused any time
19775 # you need this list (see the other examples in this section)
19777 # The root of your project's directory hierarchy
19781 # First method: specify explicitly the list of directories
19782 # This allows you to specify any subset of all the directories you need.
19785 DIRS := a/aa/ a/ab/ b/ba/
19788 # Second method: use wildcards
19789 # Note that the argument(s) to wildcard below should end with a '/'.
19790 # Since wildcards also return file names, we have to filter them out
19791 # to avoid duplicate directory names.
19792 # We thus use make's @code{dir} and @code{sort} functions.
19793 # It sets DIRs to the following value (note that the directories aaa and baa
19794 # are not given, unless you change the arguments to wildcard).
19795 # DIRS= ./a/a/ ./b/ ./a/aa/ ./a/ab/ ./a/ac/ ./b/ba/ ./b/bb/ ./b/bc/
19798 DIRS := $@{sort $@{dir $@{wildcard $@{ROOT_DIRECTORY@}/*/
19799 $@{ROOT_DIRECTORY@}/*/*/@}@}@}
19802 # Third method: use an external program
19803 # This command is much faster if run on local disks, avoiding NFS slowdowns.
19804 # This is the most complete command: it sets DIRs to the following value:
19805 # DIRS= ./a ./a/aa ./a/aa/aaa ./a/ab ./a/ac ./b ./b/ba ./b/ba/baa ./b/bb ./b/bc
19808 DIRS := $@{shell find $@{ROOT_DIRECTORY@} -type d -print@}
19812 @node Generating the Command Line Switches
19813 @section Generating the Command Line Switches
19816 Once you have created the list of directories as explained in the
19817 previous section (@pxref{Automatically Creating a List of Directories}),
19818 you can easily generate the command line arguments to pass to gnatmake.
19820 For the sake of completeness, this example assumes that the source path
19821 is not the same as the object path, and that you have two separate lists
19825 # see "Automatically creating a list of directories" to create
19830 GNATMAKE_SWITCHES := $@{patsubst %,-aI%,$@{SOURCE_DIRS@}@}
19831 GNATMAKE_SWITCHES += $@{patsubst %,-aO%,$@{OBJECT_DIRS@}@}
19834 gnatmake $@{GNATMAKE_SWITCHES@} main_unit
19837 @node Overcoming Command Line Length Limits
19838 @section Overcoming Command Line Length Limits
19841 One problem that might be encountered on big projects is that many
19842 operating systems limit the length of the command line. It is thus hard to give
19843 gnatmake the list of source and object directories.
19845 This example shows how you can set up environment variables, which will
19846 make @command{gnatmake} behave exactly as if the directories had been
19847 specified on the command line, but have a much higher length limit (or
19848 even none on most systems).
19850 It assumes that you have created a list of directories in your Makefile,
19851 using one of the methods presented in
19852 @ref{Automatically Creating a List of Directories}.
19853 For the sake of completeness, we assume that the object
19854 path (where the ALI files are found) is different from the sources patch.
19856 Note a small trick in the Makefile below: for efficiency reasons, we
19857 create two temporary variables (SOURCE_LIST and OBJECT_LIST), that are
19858 expanded immediately by @code{make}. This way we overcome the standard
19859 make behavior which is to expand the variables only when they are
19862 On Windows, if you are using the standard Windows command shell, you must
19863 replace colons with semicolons in the assignments to these variables.
19868 @font@heightrm=cmr8
19871 # In this example, we create both ADA_INCLUDE_PATH and ADA_OBJECT_PATH.
19872 # This is the same thing as putting the -I arguments on the command line.
19873 # (the equivalent of using -aI on the command line would be to define
19874 # only ADA_INCLUDE_PATH, the equivalent of -aO is ADA_OBJECT_PATH).
19875 # You can of course have different values for these variables.
19877 # Note also that we need to keep the previous values of these variables, since
19878 # they might have been set before running 'make' to specify where the GNAT
19879 # library is installed.
19881 # see "Automatically creating a list of directories" to create these
19887 space:=$@{empty@} $@{empty@}
19888 SOURCE_LIST := $@{subst $@{space@},:,$@{SOURCE_DIRS@}@}
19889 OBJECT_LIST := $@{subst $@{space@},:,$@{OBJECT_DIRS@}@}
19890 ADA_INCLUDE_PATH += $@{SOURCE_LIST@}
19891 ADA_OBJECT_PATH += $@{OBJECT_LIST@}
19892 export ADA_INCLUDE_PATH
19893 export ADA_OBJECT_PATH
19900 @node Memory Management Issues
19901 @chapter Memory Management Issues
19904 This chapter describes some useful memory pools provided in the GNAT library
19905 and in particular the GNAT Debug Pool facility, which can be used to detect
19906 incorrect uses of access values (including ``dangling references'').
19908 It also describes the @command{gnatmem} tool, which can be used to track down
19913 * Some Useful Memory Pools::
19914 * The GNAT Debug Pool Facility::
19916 * The gnatmem Tool::
19920 @node Some Useful Memory Pools
19921 @section Some Useful Memory Pools
19922 @findex Memory Pool
19923 @cindex storage, pool
19926 The @code{System.Pool_Global} package offers the Unbounded_No_Reclaim_Pool
19927 storage pool. Allocations use the standard system call @code{malloc} while
19928 deallocations use the standard system call @code{free}. No reclamation is
19929 performed when the pool goes out of scope. For performance reasons, the
19930 standard default Ada allocators/deallocators do not use any explicit storage
19931 pools but if they did, they could use this storage pool without any change in
19932 behavior. That is why this storage pool is used when the user
19933 manages to make the default implicit allocator explicit as in this example:
19934 @smallexample @c ada
19935 type T1 is access Something;
19936 -- no Storage pool is defined for T2
19937 type T2 is access Something_Else;
19938 for T2'Storage_Pool use T1'Storage_Pool;
19939 -- the above is equivalent to
19940 for T2'Storage_Pool use System.Pool_Global.Global_Pool_Object;
19944 The @code{System.Pool_Local} package offers the Unbounded_Reclaim_Pool storage
19945 pool. The allocation strategy is similar to @code{Pool_Local}'s
19946 except that the all
19947 storage allocated with this pool is reclaimed when the pool object goes out of
19948 scope. This pool provides a explicit mechanism similar to the implicit one
19949 provided by several Ada 83 compilers for allocations performed through a local
19950 access type and whose purpose was to reclaim memory when exiting the
19951 scope of a given local access. As an example, the following program does not
19952 leak memory even though it does not perform explicit deallocation:
19954 @smallexample @c ada
19955 with System.Pool_Local;
19956 procedure Pooloc1 is
19957 procedure Internal is
19958 type A is access Integer;
19959 X : System.Pool_Local.Unbounded_Reclaim_Pool;
19960 for A'Storage_Pool use X;
19963 for I in 1 .. 50 loop
19968 for I in 1 .. 100 loop
19975 The @code{System.Pool_Size} package implements the Stack_Bounded_Pool used when
19976 @code{Storage_Size} is specified for an access type.
19977 The whole storage for the pool is
19978 allocated at once, usually on the stack at the point where the access type is
19979 elaborated. It is automatically reclaimed when exiting the scope where the
19980 access type is defined. This package is not intended to be used directly by the
19981 user and it is implicitly used for each such declaration:
19983 @smallexample @c ada
19984 type T1 is access Something;
19985 for T1'Storage_Size use 10_000;
19988 @node The GNAT Debug Pool Facility
19989 @section The GNAT Debug Pool Facility
19991 @cindex storage, pool, memory corruption
19994 The use of unchecked deallocation and unchecked conversion can easily
19995 lead to incorrect memory references. The problems generated by such
19996 references are usually difficult to tackle because the symptoms can be
19997 very remote from the origin of the problem. In such cases, it is
19998 very helpful to detect the problem as early as possible. This is the
19999 purpose of the Storage Pool provided by @code{GNAT.Debug_Pools}.
20001 In order to use the GNAT specific debugging pool, the user must
20002 associate a debug pool object with each of the access types that may be
20003 related to suspected memory problems. See Ada Reference Manual 13.11.
20004 @smallexample @c ada
20005 type Ptr is access Some_Type;
20006 Pool : GNAT.Debug_Pools.Debug_Pool;
20007 for Ptr'Storage_Pool use Pool;
20011 @code{GNAT.Debug_Pools} is derived from a GNAT-specific kind of
20012 pool: the @code{Checked_Pool}. Such pools, like standard Ada storage pools,
20013 allow the user to redefine allocation and deallocation strategies. They
20014 also provide a checkpoint for each dereference, through the use of
20015 the primitive operation @code{Dereference} which is implicitly called at
20016 each dereference of an access value.
20018 Once an access type has been associated with a debug pool, operations on
20019 values of the type may raise four distinct exceptions,
20020 which correspond to four potential kinds of memory corruption:
20023 @code{GNAT.Debug_Pools.Accessing_Not_Allocated_Storage}
20025 @code{GNAT.Debug_Pools.Accessing_Deallocated_Storage}
20027 @code{GNAT.Debug_Pools.Freeing_Not_Allocated_Storage}
20029 @code{GNAT.Debug_Pools.Freeing_Deallocated_Storage }
20033 For types associated with a Debug_Pool, dynamic allocation is performed using
20034 the standard GNAT allocation routine. References to all allocated chunks of
20035 memory are kept in an internal dictionary. Several deallocation strategies are
20036 provided, whereupon the user can choose to release the memory to the system,
20037 keep it allocated for further invalid access checks, or fill it with an easily
20038 recognizable pattern for debug sessions. The memory pattern is the old IBM
20039 hexadecimal convention: @code{16#DEADBEEF#}.
20041 See the documentation in the file g-debpoo.ads for more information on the
20042 various strategies.
20044 Upon each dereference, a check is made that the access value denotes a
20045 properly allocated memory location. Here is a complete example of use of
20046 @code{Debug_Pools}, that includes typical instances of memory corruption:
20047 @smallexample @c ada
20051 with Gnat.Io; use Gnat.Io;
20052 with Unchecked_Deallocation;
20053 with Unchecked_Conversion;
20054 with GNAT.Debug_Pools;
20055 with System.Storage_Elements;
20056 with Ada.Exceptions; use Ada.Exceptions;
20057 procedure Debug_Pool_Test is
20059 type T is access Integer;
20060 type U is access all T;
20062 P : GNAT.Debug_Pools.Debug_Pool;
20063 for T'Storage_Pool use P;
20065 procedure Free is new Unchecked_Deallocation (Integer, T);
20066 function UC is new Unchecked_Conversion (U, T);
20069 procedure Info is new GNAT.Debug_Pools.Print_Info(Put_Line);
20079 Put_Line (Integer'Image(B.all));
20081 when E : others => Put_Line ("raised: " & Exception_Name (E));
20086 when E : others => Put_Line ("raised: " & Exception_Name (E));
20090 Put_Line (Integer'Image(B.all));
20092 when E : others => Put_Line ("raised: " & Exception_Name (E));
20097 when E : others => Put_Line ("raised: " & Exception_Name (E));
20100 end Debug_Pool_Test;
20104 The debug pool mechanism provides the following precise diagnostics on the
20105 execution of this erroneous program:
20108 Total allocated bytes : 0
20109 Total deallocated bytes : 0
20110 Current Water Mark: 0
20114 Total allocated bytes : 8
20115 Total deallocated bytes : 0
20116 Current Water Mark: 8
20119 raised: GNAT.DEBUG_POOLS.ACCESSING_DEALLOCATED_STORAGE
20120 raised: GNAT.DEBUG_POOLS.FREEING_DEALLOCATED_STORAGE
20121 raised: GNAT.DEBUG_POOLS.ACCESSING_NOT_ALLOCATED_STORAGE
20122 raised: GNAT.DEBUG_POOLS.FREEING_NOT_ALLOCATED_STORAGE
20124 Total allocated bytes : 8
20125 Total deallocated bytes : 4
20126 Current Water Mark: 4
20131 @node The gnatmem Tool
20132 @section The @command{gnatmem} Tool
20136 The @code{gnatmem} utility monitors dynamic allocation and
20137 deallocation activity in a program, and displays information about
20138 incorrect deallocations and possible sources of memory leaks.
20139 It is designed to work in association with a static runtime library
20140 only and in this context provides three types of information:
20143 General information concerning memory management, such as the total
20144 number of allocations and deallocations, the amount of allocated
20145 memory and the high water mark, i.e.@: the largest amount of allocated
20146 memory in the course of program execution.
20149 Backtraces for all incorrect deallocations, that is to say deallocations
20150 which do not correspond to a valid allocation.
20153 Information on each allocation that is potentially the origin of a memory
20158 * Running gnatmem::
20159 * Switches for gnatmem::
20160 * Example of gnatmem Usage::
20163 @node Running gnatmem
20164 @subsection Running @code{gnatmem}
20167 @code{gnatmem} makes use of the output created by the special version of
20168 allocation and deallocation routines that record call information. This
20169 allows to obtain accurate dynamic memory usage history at a minimal cost to
20170 the execution speed. Note however, that @code{gnatmem} is not supported on
20171 all platforms (currently, it is supported on AIX, HP-UX, GNU/Linux,
20172 Solaris and Windows NT/2000/XP (x86).
20175 The @code{gnatmem} command has the form
20178 @c $ gnatmem @ovar{switches} user_program
20179 @c Expanding @ovar macro inline (explanation in macro def comments)
20180 $ gnatmem @r{[}@var{switches}@r{]} @var{user_program}
20184 The program must have been linked with the instrumented version of the
20185 allocation and deallocation routines. This is done by linking with the
20186 @file{libgmem.a} library. For correct symbolic backtrace information,
20187 the user program should be compiled with debugging options
20188 (see @ref{Switches for gcc}). For example to build @file{my_program}:
20191 $ gnatmake -g my_program -largs -lgmem
20195 As library @file{libgmem.a} contains an alternate body for package
20196 @code{System.Memory}, @file{s-memory.adb} should not be compiled and linked
20197 when an executable is linked with library @file{libgmem.a}. It is then not
20198 recommended to use @command{gnatmake} with switch @option{^-a^/ALL_FILES^}.
20201 When @file{my_program} is executed, the file @file{gmem.out} is produced.
20202 This file contains information about all allocations and deallocations
20203 performed by the program. It is produced by the instrumented allocations and
20204 deallocations routines and will be used by @code{gnatmem}.
20206 In order to produce symbolic backtrace information for allocations and
20207 deallocations performed by the GNAT run-time library, you need to use a
20208 version of that library that has been compiled with the @option{-g} switch
20209 (see @ref{Rebuilding the GNAT Run-Time Library}).
20211 Gnatmem must be supplied with the @file{gmem.out} file and the executable to
20212 examine. If the location of @file{gmem.out} file was not explicitly supplied by
20213 @option{-i} switch, gnatmem will assume that this file can be found in the
20214 current directory. For example, after you have executed @file{my_program},
20215 @file{gmem.out} can be analyzed by @code{gnatmem} using the command:
20218 $ gnatmem my_program
20222 This will produce the output with the following format:
20224 *************** debut cc
20226 $ gnatmem my_program
20230 Total number of allocations : 45
20231 Total number of deallocations : 6
20232 Final Water Mark (non freed mem) : 11.29 Kilobytes
20233 High Water Mark : 11.40 Kilobytes
20238 Allocation Root # 2
20239 -------------------
20240 Number of non freed allocations : 11
20241 Final Water Mark (non freed mem) : 1.16 Kilobytes
20242 High Water Mark : 1.27 Kilobytes
20244 my_program.adb:23 my_program.alloc
20250 The first block of output gives general information. In this case, the
20251 Ada construct ``@code{@b{new}}'' was executed 45 times, and only 6 calls to an
20252 Unchecked_Deallocation routine occurred.
20255 Subsequent paragraphs display information on all allocation roots.
20256 An allocation root is a specific point in the execution of the program
20257 that generates some dynamic allocation, such as a ``@code{@b{new}}''
20258 construct. This root is represented by an execution backtrace (or subprogram
20259 call stack). By default the backtrace depth for allocations roots is 1, so
20260 that a root corresponds exactly to a source location. The backtrace can
20261 be made deeper, to make the root more specific.
20263 @node Switches for gnatmem
20264 @subsection Switches for @code{gnatmem}
20267 @code{gnatmem} recognizes the following switches:
20272 @cindex @option{-q} (@code{gnatmem})
20273 Quiet. Gives the minimum output needed to identify the origin of the
20274 memory leaks. Omits statistical information.
20277 @cindex @var{N} (@code{gnatmem})
20278 N is an integer literal (usually between 1 and 10) which controls the
20279 depth of the backtraces defining allocation root. The default value for
20280 N is 1. The deeper the backtrace, the more precise the localization of
20281 the root. Note that the total number of roots can depend on this
20282 parameter. This parameter must be specified @emph{before} the name of the
20283 executable to be analyzed, to avoid ambiguity.
20286 @cindex @option{-b} (@code{gnatmem})
20287 This switch has the same effect as just depth parameter.
20289 @item -i @var{file}
20290 @cindex @option{-i} (@code{gnatmem})
20291 Do the @code{gnatmem} processing starting from @file{file}, rather than
20292 @file{gmem.out} in the current directory.
20295 @cindex @option{-m} (@code{gnatmem})
20296 This switch causes @code{gnatmem} to mask the allocation roots that have less
20297 than n leaks. The default value is 1. Specifying the value of 0 will allow to
20298 examine even the roots that didn't result in leaks.
20301 @cindex @option{-s} (@code{gnatmem})
20302 This switch causes @code{gnatmem} to sort the allocation roots according to the
20303 specified order of sort criteria, each identified by a single letter. The
20304 currently supported criteria are @code{n, h, w} standing respectively for
20305 number of unfreed allocations, high watermark, and final watermark
20306 corresponding to a specific root. The default order is @code{nwh}.
20310 @node Example of gnatmem Usage
20311 @subsection Example of @code{gnatmem} Usage
20314 The following example shows the use of @code{gnatmem}
20315 on a simple memory-leaking program.
20316 Suppose that we have the following Ada program:
20318 @smallexample @c ada
20321 with Unchecked_Deallocation;
20322 procedure Test_Gm is
20324 type T is array (1..1000) of Integer;
20325 type Ptr is access T;
20326 procedure Free is new Unchecked_Deallocation (T, Ptr);
20329 procedure My_Alloc is
20334 procedure My_DeAlloc is
20342 for I in 1 .. 5 loop
20343 for J in I .. 5 loop
20354 The program needs to be compiled with debugging option and linked with
20355 @code{gmem} library:
20358 $ gnatmake -g test_gm -largs -lgmem
20362 Then we execute the program as usual:
20369 Then @code{gnatmem} is invoked simply with
20375 which produces the following output (result may vary on different platforms):
20380 Total number of allocations : 18
20381 Total number of deallocations : 5
20382 Final Water Mark (non freed mem) : 53.00 Kilobytes
20383 High Water Mark : 56.90 Kilobytes
20385 Allocation Root # 1
20386 -------------------
20387 Number of non freed allocations : 11
20388 Final Water Mark (non freed mem) : 42.97 Kilobytes
20389 High Water Mark : 46.88 Kilobytes
20391 test_gm.adb:11 test_gm.my_alloc
20393 Allocation Root # 2
20394 -------------------
20395 Number of non freed allocations : 1
20396 Final Water Mark (non freed mem) : 10.02 Kilobytes
20397 High Water Mark : 10.02 Kilobytes
20399 s-secsta.adb:81 system.secondary_stack.ss_init
20401 Allocation Root # 3
20402 -------------------
20403 Number of non freed allocations : 1
20404 Final Water Mark (non freed mem) : 12 Bytes
20405 High Water Mark : 12 Bytes
20407 s-secsta.adb:181 system.secondary_stack.ss_init
20411 Note that the GNAT run time contains itself a certain number of
20412 allocations that have no corresponding deallocation,
20413 as shown here for root #2 and root
20414 #3. This is a normal behavior when the number of non-freed allocations
20415 is one, it allocates dynamic data structures that the run time needs for
20416 the complete lifetime of the program. Note also that there is only one
20417 allocation root in the user program with a single line back trace:
20418 test_gm.adb:11 test_gm.my_alloc, whereas a careful analysis of the
20419 program shows that 'My_Alloc' is called at 2 different points in the
20420 source (line 21 and line 24). If those two allocation roots need to be
20421 distinguished, the backtrace depth parameter can be used:
20424 $ gnatmem 3 test_gm
20428 which will give the following output:
20433 Total number of allocations : 18
20434 Total number of deallocations : 5
20435 Final Water Mark (non freed mem) : 53.00 Kilobytes
20436 High Water Mark : 56.90 Kilobytes
20438 Allocation Root # 1
20439 -------------------
20440 Number of non freed allocations : 10
20441 Final Water Mark (non freed mem) : 39.06 Kilobytes
20442 High Water Mark : 42.97 Kilobytes
20444 test_gm.adb:11 test_gm.my_alloc
20445 test_gm.adb:24 test_gm
20446 b_test_gm.c:52 main
20448 Allocation Root # 2
20449 -------------------
20450 Number of non freed allocations : 1
20451 Final Water Mark (non freed mem) : 10.02 Kilobytes
20452 High Water Mark : 10.02 Kilobytes
20454 s-secsta.adb:81 system.secondary_stack.ss_init
20455 s-secsta.adb:283 <system__secondary_stack___elabb>
20456 b_test_gm.c:33 adainit
20458 Allocation Root # 3
20459 -------------------
20460 Number of non freed allocations : 1
20461 Final Water Mark (non freed mem) : 3.91 Kilobytes
20462 High Water Mark : 3.91 Kilobytes
20464 test_gm.adb:11 test_gm.my_alloc
20465 test_gm.adb:21 test_gm
20466 b_test_gm.c:52 main
20468 Allocation Root # 4
20469 -------------------
20470 Number of non freed allocations : 1
20471 Final Water Mark (non freed mem) : 12 Bytes
20472 High Water Mark : 12 Bytes
20474 s-secsta.adb:181 system.secondary_stack.ss_init
20475 s-secsta.adb:283 <system__secondary_stack___elabb>
20476 b_test_gm.c:33 adainit
20480 The allocation root #1 of the first example has been split in 2 roots #1
20481 and #3 thanks to the more precise associated backtrace.
20485 @node Stack Related Facilities
20486 @chapter Stack Related Facilities
20489 This chapter describes some useful tools associated with stack
20490 checking and analysis. In
20491 particular, it deals with dynamic and static stack usage measurements.
20494 * Stack Overflow Checking::
20495 * Static Stack Usage Analysis::
20496 * Dynamic Stack Usage Analysis::
20499 @node Stack Overflow Checking
20500 @section Stack Overflow Checking
20501 @cindex Stack Overflow Checking
20502 @cindex -fstack-check
20505 For most operating systems, @command{gcc} does not perform stack overflow
20506 checking by default. This means that if the main environment task or
20507 some other task exceeds the available stack space, then unpredictable
20508 behavior will occur. Most native systems offer some level of protection by
20509 adding a guard page at the end of each task stack. This mechanism is usually
20510 not enough for dealing properly with stack overflow situations because
20511 a large local variable could ``jump'' above the guard page.
20512 Furthermore, when the
20513 guard page is hit, there may not be any space left on the stack for executing
20514 the exception propagation code. Enabling stack checking avoids
20517 To activate stack checking, compile all units with the gcc option
20518 @option{-fstack-check}. For example:
20521 gcc -c -fstack-check package1.adb
20525 Units compiled with this option will generate extra instructions to check
20526 that any use of the stack (for procedure calls or for declaring local
20527 variables in declare blocks) does not exceed the available stack space.
20528 If the space is exceeded, then a @code{Storage_Error} exception is raised.
20530 For declared tasks, the stack size is controlled by the size
20531 given in an applicable @code{Storage_Size} pragma or by the value specified
20532 at bind time with @option{-d} (@pxref{Switches for gnatbind}) or is set to
20533 the default size as defined in the GNAT runtime otherwise.
20535 For the environment task, the stack size depends on
20536 system defaults and is unknown to the compiler. Stack checking
20537 may still work correctly if a fixed
20538 size stack is allocated, but this cannot be guaranteed.
20540 To ensure that a clean exception is signalled for stack
20541 overflow, set the environment variable
20542 @env{GNAT_STACK_LIMIT} to indicate the maximum
20543 stack area that can be used, as in:
20544 @cindex GNAT_STACK_LIMIT
20547 SET GNAT_STACK_LIMIT 1600
20551 The limit is given in kilobytes, so the above declaration would
20552 set the stack limit of the environment task to 1.6 megabytes.
20553 Note that the only purpose of this usage is to limit the amount
20554 of stack used by the environment task. If it is necessary to
20555 increase the amount of stack for the environment task, then this
20556 is an operating systems issue, and must be addressed with the
20557 appropriate operating systems commands.
20560 To have a fixed size stack in the environment task, the stack must be put
20561 in the P0 address space and its size specified. Use these switches to
20565 gnatmake my_progs -largs "-Wl,--opt=STACK=4000,/p0image"
20569 The quotes are required to keep case. The number after @samp{STACK=} is the
20570 size of the environmental task stack in pagelets (512 bytes). In this example
20571 the stack size is about 2 megabytes.
20574 A consequence of the @option{/p0image} qualifier is also to makes RMS buffers
20575 be placed in P0 space. Refer to @cite{HP OpenVMS Linker Utility Manual} for
20576 more details about the @option{/p0image} qualifier and the @option{stack}
20580 @node Static Stack Usage Analysis
20581 @section Static Stack Usage Analysis
20582 @cindex Static Stack Usage Analysis
20583 @cindex -fstack-usage
20586 A unit compiled with @option{-fstack-usage} will generate an extra file
20588 the maximum amount of stack used, on a per-function basis.
20589 The file has the same
20590 basename as the target object file with a @file{.su} extension.
20591 Each line of this file is made up of three fields:
20595 The name of the function.
20599 One or more qualifiers: @code{static}, @code{dynamic}, @code{bounded}.
20602 The second field corresponds to the size of the known part of the function
20605 The qualifier @code{static} means that the function frame size
20607 It usually means that all local variables have a static size.
20608 In this case, the second field is a reliable measure of the function stack
20611 The qualifier @code{dynamic} means that the function frame size is not static.
20612 It happens mainly when some local variables have a dynamic size. When this
20613 qualifier appears alone, the second field is not a reliable measure
20614 of the function stack analysis. When it is qualified with @code{bounded}, it
20615 means that the second field is a reliable maximum of the function stack
20618 @node Dynamic Stack Usage Analysis
20619 @section Dynamic Stack Usage Analysis
20622 It is possible to measure the maximum amount of stack used by a task, by
20623 adding a switch to @command{gnatbind}, as:
20626 $ gnatbind -u0 file
20630 With this option, at each task termination, its stack usage is output on
20632 It is not always convenient to output the stack usage when the program
20633 is still running. Hence, it is possible to delay this output until program
20634 termination. for a given number of tasks specified as the argument of the
20635 @option{-u} option. For instance:
20638 $ gnatbind -u100 file
20642 will buffer the stack usage information of the first 100 tasks to terminate and
20643 output this info at program termination. Results are displayed in four
20647 Index | Task Name | Stack Size | Stack Usage [Value +/- Variation]
20654 is a number associated with each task.
20657 is the name of the task analyzed.
20660 is the maximum size for the stack.
20663 is the measure done by the stack analyzer. In order to prevent overflow, the stack
20664 is not entirely analyzed, and it's not possible to know exactly how
20665 much has actually been used. The report thus contains the theoretical stack usage
20666 (Value) and the possible variation (Variation) around this value.
20671 The environment task stack, e.g., the stack that contains the main unit, is
20672 only processed when the environment variable GNAT_STACK_LIMIT is set.
20675 @c *********************************
20677 @c *********************************
20678 @node Verifying Properties Using gnatcheck
20679 @chapter Verifying Properties Using @command{gnatcheck}
20681 @cindex @command{gnatcheck}
20684 The @command{gnatcheck} tool is an ASIS-based utility that checks properties
20685 of Ada source files according to a given set of semantic rules.
20688 In order to check compliance with a given rule, @command{gnatcheck} has to
20689 semantically analyze the Ada sources.
20690 Therefore, checks can only be performed on
20691 legal Ada units. Moreover, when a unit depends semantically upon units located
20692 outside the current directory, the source search path has to be provided when
20693 calling @command{gnatcheck}, either through a specified project file or
20694 through @command{gnatcheck} switches as described below.
20696 A number of rules are predefined in @command{gnatcheck} and are described
20697 later in this chapter.
20698 You can also add new rules, by modifying the @command{gnatcheck} code and
20699 rebuilding the tool. In order to add a simple rule making some local checks,
20700 a small amount of straightforward ASIS-based programming is usually needed.
20702 Project support for @command{gnatcheck} is provided by the GNAT
20703 driver (see @ref{The GNAT Driver and Project Files}).
20705 Invoking @command{gnatcheck} on the command line has the form:
20708 @c $ gnatcheck @ovar{switches} @{@var{filename}@}
20709 @c @r{[}^-files^/FILES^=@{@var{arg_list_filename}@}@r{]}
20710 @c @r{[}-cargs @var{gcc_switches}@r{]} -rules @var{rule_options}
20711 @c Expanding @ovar macro inline (explanation in macro def comments)
20712 $ gnatcheck @r{[}@var{switches}@r{]} @{@var{filename}@}
20713 @r{[}^-files^/FILES^=@{@var{arg_list_filename}@}@r{]}
20714 @r{[}-cargs @var{gcc_switches}@r{]} -rules @var{rule_options}
20721 @var{switches} specify the general tool options
20724 Each @var{filename} is the name (including the extension) of a source
20725 file to process. ``Wildcards'' are allowed, and
20726 the file name may contain path information.
20729 Each @var{arg_list_filename} is the name (including the extension) of a text
20730 file containing the names of the source files to process, separated by spaces
20734 @var{gcc_switches} is a list of switches for
20735 @command{gcc}. They will be passed on to all compiler invocations made by
20736 @command{gnatcheck} to generate the ASIS trees. Here you can provide
20737 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
20738 and use the @option{-gnatec} switch to set the configuration file.
20741 @var{rule_options} is a list of options for controlling a set of
20742 rules to be checked by @command{gnatcheck} (@pxref{gnatcheck Rule Options}).
20746 Either a @file{@var{filename}} or an @file{@var{arg_list_filename}} must be
20750 * Format of the Report File::
20751 * General gnatcheck Switches::
20752 * gnatcheck Rule Options::
20753 * Adding the Results of Compiler Checks to gnatcheck Output::
20754 * Project-Wide Checks::
20756 * Predefined Rules::
20757 * Example of gnatcheck Usage::
20760 @node Format of the Report File
20761 @section Format of the Report File
20762 @cindex Report file (for @code{gnatcheck})
20765 The @command{gnatcheck} tool outputs on @file{stdout} all messages concerning
20767 It also creates a text file that
20768 contains the complete report of the last gnatcheck run. By default this file
20769 is named named @file{^gnatcheck.out^GNATCHECK.OUT^} and it is located in the
20770 current directory; the @option{^-o^/OUTPUT^} option can be used to change the
20771 name and/or location of the report file. This report contains:
20773 @item date and time of @command{gnatcheck} run, the version of
20774 the tool that has generated this report and the full parameters
20775 of the @command{gnatcheck} invocation;
20776 @item list of enabled rules;
20777 @item total number of detected violations;
20778 @item list of source files where rule violations have been detected;
20779 @item list of source files where no violations have been detected.
20782 @node General gnatcheck Switches
20783 @section General @command{gnatcheck} Switches
20786 The following switches control the general @command{gnatcheck} behavior
20790 @cindex @option{^-a^/ALL^} (@command{gnatcheck})
20792 Process all units including those with read-only ALI files such as
20793 those from the GNAT Run-Time library.
20797 @cindex @option{-d} (@command{gnatcheck})
20802 @cindex @option{-dd} (@command{gnatcheck})
20804 Progress indicator mode (for use in GPS).
20807 @cindex @option{^-h^/HELP^} (@command{gnatcheck})
20809 List the predefined and user-defined rules. For more details see
20810 @ref{Predefined Rules}.
20812 @cindex @option{^-l^/LOCS^} (@command{gnatcheck})
20814 Use full source locations references in the report file. For a construct from
20815 a generic instantiation a full source location is a chain from the location
20816 of this construct in the generic unit to the place where this unit is
20819 @cindex @option{^-log^/LOG^} (@command{gnatcheck})
20821 Duplicate all the output sent to @file{stderr} into a log file. The log file
20822 is named @file{gnatcheck.log} and is located in the current directory.
20824 @cindex @option{^-m^/DIAGNOSTIC_LIMIT^} (@command{gnatcheck})
20825 @item ^-m@i{nnnn}^/DIAGNOSTIC_LIMIT=@i{nnnn}^
20826 Maximum number of diagnostics to be sent to @file{stdout}, where @i{nnnn} is in
20827 the range 0@dots{}1000;
20828 the default value is 500. Zero means that there is no limitation on
20829 the number of diagnostic messages to be output.
20831 @cindex @option{^-q^/QUIET^} (@command{gnatcheck})
20833 Quiet mode. All the diagnostics about rule violations are placed in the
20834 @command{gnatcheck} report file only, without duplication on @file{stdout}.
20836 @cindex @option{^-s^/SHORT^} (@command{gnatcheck})
20838 Short format of the report file (no version information, no list of applied
20839 rules, no list of checked sources is included)
20841 @cindex @option{^--include-file=@var{file}^/INCLUDE_FILE=@var{file}^} (@command{gnatcheck})
20842 @item ^--include-file^/INCLUDE_FILE^
20843 Append the content of the specified text file to the report file
20845 @cindex @option{^-t^/TIME^} (@command{gnatcheck})
20847 Print out execution time.
20849 @cindex @option{^-v^/VERBOSE^} (@command{gnatcheck})
20850 @item ^-v^/VERBOSE^
20851 Verbose mode; @command{gnatcheck} generates version information and then
20852 a trace of sources being processed.
20854 @cindex @option{^-o ^/OUTPUT^} (@command{gnatcheck})
20855 @item ^-o ^/OUTPUT=^@var{report_file}
20856 Set name of report file file to @var{report_file} .
20860 @node gnatcheck Rule Options
20861 @section @command{gnatcheck} Rule Options
20864 The following options control the processing performed by
20865 @command{gnatcheck}.
20868 @cindex @option{+ALL} (@command{gnatcheck})
20870 Turn all the rule checks ON.
20872 @cindex @option{-ALL} (@command{gnatcheck})
20874 Turn all the rule checks OFF.
20876 @cindex @option{+R} (@command{gnatcheck})
20877 @item +R@var{rule_id}@r{[}:@var{param}@r{]}
20878 Turn on the check for a specified rule with the specified parameter, if any.
20879 @var{rule_id} must be the identifier of one of the currently implemented rules
20880 (use @option{^-h^/HELP^} for the list of implemented rules). Rule identifiers
20881 are not case-sensitive. The @var{param} item must
20882 be a string representing a valid parameter(s) for the specified rule.
20883 If it contains any space characters then this string must be enclosed in
20886 @cindex @option{-R} (@command{gnatcheck})
20887 @item -R@var{rule_id}@r{[}:@var{param}@r{]}
20888 Turn off the check for a specified rule with the specified parameter, if any.
20890 @cindex @option{-from} (@command{gnatcheck})
20891 @item -from=@var{rule_option_filename}
20892 Read the rule options from the text file @var{rule_option_filename}, referred
20893 to as a ``coding standard file'' below.
20898 The default behavior is that all the rule checks are disabled.
20900 A coding standard file is a text file that contains a set of rule options
20902 @cindex Coding standard file (for @code{gnatcheck})
20903 The file may contain empty lines and Ada-style comments (comment
20904 lines and end-of-line comments). There can be several rule options on a
20905 single line (separated by a space).
20907 A coding standard file may reference other coding standard files by including
20908 more @option{-from=@var{rule_option_filename}}
20909 options, each such option being replaced with the content of the
20910 corresponding coding standard file during processing. In case a
20911 cycle is detected (that is, @file{@var{rule_file_1}} reads rule options
20912 from @file{@var{rule_file_2}}, and @file{@var{rule_file_2}} reads
20913 (directly or indirectly) rule options from @file{@var{rule_file_1}}),
20914 processing fails with an error message.
20917 @node Adding the Results of Compiler Checks to gnatcheck Output
20918 @section Adding the Results of Compiler Checks to @command{gnatcheck} Output
20921 The @command{gnatcheck} tool can include in the generated diagnostic messages
20923 the report file the results of the checks performed by the compiler. Though
20924 disabled by default, this effect may be obtained by using @option{+R} with
20925 the following rule identifiers and parameters:
20929 To record restrictions violations (which are performed by the compiler if the
20930 pragma @code{Restrictions} or @code{Restriction_Warnings} are given),
20931 use the @code{Restrictions} rule
20932 with the same parameters as pragma
20933 @code{Restrictions} or @code{Restriction_Warnings}.
20936 To record compiler style checks (@pxref{Style Checking}), use the
20937 @code{Style_Checks} rule.
20938 This rule takes a parameter in one of the following forms:
20942 which enables the standard style checks corresponding to the @option{-gnatyy}
20943 GNAT style check option, or
20946 a string with the same
20947 structure and semantics as the @code{string_LITERAL} parameter of the
20948 GNAT pragma @code{Style_Checks}
20949 (for further information about this pragma,
20950 @pxref{Pragma Style_Checks,,, gnat_rm, GNAT Reference Manual}).
20955 @code{+RStyle_Checks:O} rule option activates
20956 the compiler style check that corresponds to
20957 @code{-gnatyO} style check option.
20960 To record compiler warnings (@pxref{Warning Message Control}), use the
20961 @code{Warnings} rule with a parameter that is a valid
20962 @i{static_string_expression} argument of the GNAT pragma @code{Warnings}
20963 (for further information about this pragma,
20964 @pxref{Pragma Warnings,,,gnat_rm, GNAT Reference Manual}).
20965 Note that in case of gnatcheck
20966 's' parameter, that corresponds to the GNAT @option{-gnatws} option, disables
20967 all the specific warnings, but not suppresses the warning mode,
20968 and 'e' parameter, corresponding to @option{-gnatwe} that means
20969 "treat warnings as errors", does not have any effect.
20973 To disable a specific restriction check, use @code{-RStyle_Checks} gnatcheck
20974 option with the corresponding restriction name as a parameter. @code{-R} is
20975 not available for @code{Style_Checks} and @code{Warnings} options, to disable
20976 warnings and style checks, use the corresponding warning and style options.
20978 @node Project-Wide Checks
20979 @section Project-Wide Checks
20980 @cindex Project-wide checks (for @command{gnatcheck})
20983 In order to perform checks on all units of a given project, you can use
20984 the GNAT driver along with the @option{-P} option:
20986 gnat check -Pproj -rules -from=my_rules
20990 If the project @code{proj} depends upon other projects, you can perform
20991 checks on the project closure using the @option{-U} option:
20993 gnat check -Pproj -U -rules -from=my_rules
20997 Finally, if not all the units are relevant to a particular main
20998 program in the project closure, you can perform checks for the set
20999 of units needed to create a given main program (unit closure) using
21000 the @option{-U} option followed by the name of the main unit:
21002 gnat check -Pproj -U main -rules -from=my_rules
21006 @node Rule exemption
21007 @section Rule exemption
21008 @cindex Rule exemption (for @command{gnatcheck})
21011 One of the most useful applications of @command{gnatcheck} is to
21012 automate the enforcement of project-specific coding standards,
21013 for example in safety-critical systems where particular features
21014 must be restricted in order to simplify the certification effort.
21015 However, it may sometimes be appropriate to violate a coding standard rule,
21016 and in such cases the rationale for the violation should be provided
21017 in the source program itself so that the individuals
21018 reviewing or maintaining the program can immediately understand the intent.
21020 The @command{gnatcheck} tool supports this practice with the notion of
21021 a ``rule exemption'' covering a specific source code section. Normally
21022 rule violation messages are issued both on @file{stderr}
21023 and in a report file. In contrast, exempted violations are not listed on
21024 @file{stderr}; thus users invoking @command{gnatcheck} interactively
21025 (e.g. in its GPS interface) do not need to pay attention to known and
21026 justified violations. However, exempted violations along with their
21027 justification are documented in a special section of the report file that
21028 @command{gnatcheck} generates.
21031 * Using pragma Annotate to Control Rule Exemption::
21032 * gnatcheck Annotations Rules::
21035 @node Using pragma Annotate to Control Rule Exemption
21036 @subsection Using pragma @code{Annotate} to Control Rule Exemption
21037 @cindex Using pragma Annotate to control rule exemption
21040 Rule exemption is controlled by pragma @code{Annotate} when its first
21041 argument is ``gnatcheck''. The syntax of @command{gnatcheck}'s
21042 exemption control annotations is as follows:
21044 @smallexample @c ada
21046 pragma Annotate (gnatcheck, @i{exemption_control}, @i{Rule_Name}, [@i{justification}]);
21048 @i{exemption_control} ::= Exempt_On | Exempt_Off
21050 @i{Rule_Name} ::= string_literal
21052 @i{justification} ::= string_literal
21057 When a @command{gnatcheck} annotation has more then four arguments,
21058 @command{gnatcheck} issues a warning and ignores the additional arguments.
21059 If the additional arguments do not follow the syntax above,
21060 @command{gnatcheck} emits a warning and ignores the annotation.
21062 The @i{@code{Rule_Name}} argument should be the name of some existing
21063 @command{gnatcheck} rule.
21064 Otherwise a warning message is generated and the pragma is
21065 ignored. If @code{Rule_Name} denotes a rule that is not activated by the given
21066 @command{gnatcheck} call, the pragma is ignored and no warning is issued.
21068 A source code section where an exemption is active for a given rule is
21069 delimited by an @code{exempt_on} and @code{exempt_off} annotation pair:
21071 @smallexample @c ada
21072 pragma Annotate (gnatcheck, Exempt_On, Rule_Name, "justification");
21073 -- source code section
21074 pragma Annotate (gnatcheck, Exempt_Off, Rule_Name);
21078 @node gnatcheck Annotations Rules
21079 @subsection @command{gnatcheck} Annotations Rules
21080 @cindex @command{gnatcheck} annotations rules
21085 An ``Exempt_Off'' annotation can only appear after a corresponding
21086 ``Exempt_On'' annotation.
21089 Exempted source code sections are only based on the source location of the
21090 annotations. Any source construct between the two
21091 annotations is part of the exempted source code section.
21094 Exempted source code sections for different rules are independent. They can
21095 be nested or intersect with one another without limitation.
21096 Creating nested or intersecting source code sections for the same rule is
21100 Malformed exempted source code sections are reported by a warning, and
21101 the corresponding rule exemptions are ignored.
21104 When an exempted source code section does not contain at least one violation
21105 of the exempted rule, a warning is emitted on @file{stderr}.
21108 If an ``Exempt_On'' annotation pragma does not have a matching
21109 ``Exempt_Off'' annotation pragma in the same compilation unit, then the
21110 exemption for the given rule is ignored and a warning is issued.
21114 @node Predefined Rules
21115 @section Predefined Rules
21116 @cindex Predefined rules (for @command{gnatcheck})
21119 @c (Jan 2007) Since the global rules are still under development and are not
21120 @c documented, there is no point in explaining the difference between
21121 @c global and local rules
21123 A rule in @command{gnatcheck} is either local or global.
21124 A @emph{local rule} is a rule that applies to a well-defined section
21125 of a program and that can be checked by analyzing only this section.
21126 A @emph{global rule} requires analysis of some global properties of the
21127 whole program (mostly related to the program call graph).
21128 As of @value{NOW}, the implementation of global rules should be
21129 considered to be at a preliminary stage. You can use the
21130 @option{+GLOBAL} option to enable all the global rules, and the
21131 @option{-GLOBAL} rule option to disable all the global rules.
21133 All the global rules in the list below are
21134 so indicated by marking them ``GLOBAL''.
21135 This +GLOBAL and -GLOBAL options are not
21136 included in the list of gnatcheck options above, because at the moment they
21137 are considered as a temporary debug options.
21139 @command{gnatcheck} performs rule checks for generic
21140 instances only for global rules. This limitation may be relaxed in a later
21145 The predefined rules implemented in @command{gnatcheck}
21146 are described in a companion document,
21147 @cite{GNATcheck Reference Manual -- Predefined Rules}.
21148 The rule identifier is
21149 used as a parameter of @command{gnatcheck}'s @option{+R} or @option{-R}
21153 @node Example of gnatcheck Usage
21154 @section Example of @command{gnatcheck} Usage
21157 Here is a simple example. Suppose that in the current directory we have a
21158 project file named @file{gnatcheck_example.gpr} with the following content:
21160 @smallexample @c projectfile
21161 project Gnatcheck_Example is
21163 for Source_Dirs use ("src");
21164 for Object_Dir use "obj";
21165 for Main use ("main.adb");
21168 for Default_Switches ("ada") use ("-rules", "-from=coding_standard");
21171 end Gnatcheck_Example;
21175 And the file named @file{coding_standard} is also located in the current
21176 directory and has the following content:
21179 -----------------------------------------------------
21180 -- This is a sample gnatcheck coding standard file --
21181 -----------------------------------------------------
21183 -- First, turning on rules, that are directly implemented in gnatcheck
21184 +RAbstract_Type_Declarations
21187 +RFloat_Equality_Checks
21188 +REXIT_Statements_With_No_Loop_Name
21190 -- Then, activating compiler checks of interest:
21192 -- This style check checks if a unit name is present on END keyword that
21193 -- is the end of the unit declaration
21197 And the subdirectory @file{src} contains the following Ada sources:
21201 @smallexample @c ada
21203 type T is abstract tagged private;
21204 procedure P (X : T) is abstract;
21207 type My_Float is digits 8;
21208 function Is_Equal (L, R : My_Float) return Boolean;
21211 type T is abstract tagged null record;
21218 @smallexample @c ada
21219 package body Pack is
21220 package body Inner is
21221 function Is_Equal (L, R : My_Float) return Boolean is
21230 and @file{main.adb}
21232 @smallexample @c ada
21233 with Pack; use Pack;
21237 (gnatcheck, Exempt_On, "Anonymous_Arrays", "this one is fine");
21238 Float_Array : array (1 .. 10) of Inner.My_Float;
21239 pragma Annotate (gnatcheck, Exempt_Off, "Anonymous_Arrays");
21241 Another_Float_Array : array (1 .. 10) of Inner.My_Float;
21245 B : Boolean := False;
21248 for J in Float_Array'Range loop
21249 if Is_Equal (Float_Array (J), Another_Float_Array (J)) then
21258 And suppose we call @command{gnatcheck} from the current directory using
21259 the @command{gnat} driver:
21262 gnat check -Pgnatcheck_example.gpr
21266 As a result, @command{gnatcheck} is called to check all the files from the
21267 project @file{gnatcheck_example.gpr} using the coding standard defined by
21268 the file @file{coding_standard}. As the result, the @command{gnatcheck}
21269 report file named @file{gnatcheck.out} will be created in the current
21270 directory, and it will have the following content:
21273 RULE CHECKING REPORT
21277 Date and time of execution: 2009.10.28 14:17
21278 Tool version: GNATCHECK (built with ASIS 2.0.R for GNAT Pro 6.3.0w (20091016))
21281 gnatcheck -files=.../GNAT-TEMP-000004.TMP -cargs -gnatec=.../GNAT-TEMP-000003.TMP -rules -from=coding_standard
21283 Coding standard (applied rules):
21284 Abstract_Type_Declarations
21286 EXIT_Statements_With_No_Loop_Name
21287 Float_Equality_Checks
21290 Compiler style checks: -gnatye
21292 Number of coding standard violations: 6
21293 Number of exempted coding standard violations: 1
21295 2. DETECTED RULE VIOLATIONS
21297 2.1. NON-EXEMPTED VIOLATIONS
21299 Source files with non-exempted violations
21304 List of violations grouped by files, and ordered by increasing source location:
21306 pack.ads:2:4: declaration of abstract type
21307 pack.ads:5:4: declaration of local package
21308 pack.ads:10:30: declaration of abstract type
21309 pack.ads:11:1: (style) "end Pack" required
21310 pack.adb:5:19: use of equality operation for float values
21311 pack.adb:6:7: (style) "end Is_Equal" required
21312 main.adb:9:26: anonymous array type
21313 main.adb:19:10: exit statement with no loop name
21315 2.2. EXEMPTED VIOLATIONS
21317 Source files with exempted violations
21320 List of violations grouped by files, and ordered by increasing source location:
21322 main.adb:6:18: anonymous array type
21325 2.3. SOURCE FILES WITH NO VIOLATION
21327 No files without violations
21333 @c *********************************
21334 @node Creating Sample Bodies Using gnatstub
21335 @chapter Creating Sample Bodies Using @command{gnatstub}
21339 @command{gnatstub} creates body stubs, that is, empty but compilable bodies
21340 for library unit declarations.
21342 Note: to invoke @code{gnatstub} with a project file, use the @code{gnat}
21343 driver (see @ref{The GNAT Driver and Project Files}).
21345 To create a body stub, @command{gnatstub} has to compile the library
21346 unit declaration. Therefore, bodies can be created only for legal
21347 library units. Moreover, if a library unit depends semantically upon
21348 units located outside the current directory, you have to provide
21349 the source search path when calling @command{gnatstub}, see the description
21350 of @command{gnatstub} switches below.
21352 By default, all the program unit body stubs generated by @code{gnatstub}
21353 raise the predefined @code{Program_Error} exception, which will catch
21354 accidental calls of generated stubs. This behavior can be changed with
21355 option @option{^--no-exception^/NO_EXCEPTION^} (see below).
21358 * Running gnatstub::
21359 * Switches for gnatstub::
21362 @node Running gnatstub
21363 @section Running @command{gnatstub}
21366 @command{gnatstub} has the command-line interface of the form
21369 @c $ gnatstub @ovar{switches} @var{filename} @ovar{directory}
21370 @c Expanding @ovar macro inline (explanation in macro def comments)
21371 $ gnatstub @r{[}@var{switches}@r{]} @var{filename} @r{[}@var{directory}@r{]} @r{[}-cargs @var{gcc_switches}@r{]}
21378 is the name of the source file that contains a library unit declaration
21379 for which a body must be created. The file name may contain the path
21381 The file name does not have to follow the GNAT file name conventions. If the
21383 does not follow GNAT file naming conventions, the name of the body file must
21385 explicitly as the value of the @option{^-o^/BODY=^@var{body-name}} option.
21386 If the file name follows the GNAT file naming
21387 conventions and the name of the body file is not provided,
21390 of the body file from the argument file name by replacing the @file{.ads}
21392 with the @file{.adb} suffix.
21395 indicates the directory in which the body stub is to be placed (the default
21399 @item @samp{@var{gcc_switches}} is a list of switches for
21400 @command{gcc}. They will be passed on to all compiler invocations made by
21401 @command{gnatelim} to generate the ASIS trees. Here you can provide
21402 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
21403 use the @option{-gnatec} switch to set the configuration file etc.
21406 is an optional sequence of switches as described in the next section
21409 @node Switches for gnatstub
21410 @section Switches for @command{gnatstub}
21416 @cindex @option{^-f^/FULL^} (@command{gnatstub})
21417 If the destination directory already contains a file with the name of the
21419 for the argument spec file, replace it with the generated body stub.
21421 @item ^-hs^/HEADER=SPEC^
21422 @cindex @option{^-hs^/HEADER=SPEC^} (@command{gnatstub})
21423 Put the comment header (i.e., all the comments preceding the
21424 compilation unit) from the source of the library unit declaration
21425 into the body stub.
21427 @item ^-hg^/HEADER=GENERAL^
21428 @cindex @option{^-hg^/HEADER=GENERAL^} (@command{gnatstub})
21429 Put a sample comment header into the body stub.
21431 @item ^--header-file=@var{filename}^/FROM_HEADER_FILE=@var{filename}^
21432 @cindex @option{^--header-file^/FROM_HEADER_FILE=^} (@command{gnatstub})
21433 Use the content of the file as the comment header for a generated body stub.
21437 @cindex @option{-IDIR} (@command{gnatstub})
21439 @cindex @option{-I-} (@command{gnatstub})
21442 @item /NOCURRENT_DIRECTORY
21443 @cindex @option{/NOCURRENT_DIRECTORY} (@command{gnatstub})
21445 ^These switches have ^This switch has^ the same meaning as in calls to
21447 ^They define ^It defines ^ the source search path in the call to
21448 @command{gcc} issued
21449 by @command{gnatstub} to compile an argument source file.
21451 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE=^@var{PATH}
21452 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@command{gnatstub})
21453 This switch has the same meaning as in calls to @command{gcc}.
21454 It defines the additional configuration file to be passed to the call to
21455 @command{gcc} issued
21456 by @command{gnatstub} to compile an argument source file.
21458 @item ^-gnatyM^/MAX_LINE_LENGTH=^@var{n}
21459 @cindex @option{^-gnatyM^/MAX_LINE_LENGTH^} (@command{gnatstub})
21460 (@var{n} is a non-negative integer). Set the maximum line length in the
21461 body stub to @var{n}; the default is 79. The maximum value that can be
21462 specified is 32767. Note that in the special case of configuration
21463 pragma files, the maximum is always 32767 regardless of whether or
21464 not this switch appears.
21466 @item ^-gnaty^/STYLE_CHECKS=^@var{n}
21467 @cindex @option{^-gnaty^/STYLE_CHECKS=^} (@command{gnatstub})
21468 (@var{n} is a non-negative integer from 1 to 9). Set the indentation level in
21469 the generated body sample to @var{n}.
21470 The default indentation is 3.
21472 @item ^-gnatyo^/ORDERED_SUBPROGRAMS^
21473 @cindex @option{^-gnato^/ORDERED_SUBPROGRAMS^} (@command{gnatstub})
21474 Order local bodies alphabetically. (By default local bodies are ordered
21475 in the same way as the corresponding local specs in the argument spec file.)
21477 @item ^-i^/INDENTATION=^@var{n}
21478 @cindex @option{^-i^/INDENTATION^} (@command{gnatstub})
21479 Same as @option{^-gnaty^/STYLE_CHECKS=^@var{n}}
21481 @item ^-k^/TREE_FILE=SAVE^
21482 @cindex @option{^-k^/TREE_FILE=SAVE^} (@command{gnatstub})
21483 Do not remove the tree file (i.e., the snapshot of the compiler internal
21484 structures used by @command{gnatstub}) after creating the body stub.
21486 @item ^-l^/LINE_LENGTH=^@var{n}
21487 @cindex @option{^-l^/LINE_LENGTH^} (@command{gnatstub})
21488 Same as @option{^-gnatyM^/MAX_LINE_LENGTH=^@var{n}}
21490 @item ^--no-exception^/NO_EXCEPTION^
21491 @cindex @option{^--no-exception^/NO_EXCEPTION^} (@command{gnatstub})
21492 Avoind raising PROGRAM_ERROR in the generated bodies of program unit stubs.
21493 This is not always possible for function stubs.
21495 @item ^--no-local-header^/NO_LOCAL_HEADER^
21496 @cindex @option{^--no-local-header^/NO_LOCAL_HEADER^} (@command{gnatstub})
21497 Do not place local comment header with unit name before body stub for a
21500 @item ^-o ^/BODY=^@var{body-name}
21501 @cindex @option{^-o^/BODY^} (@command{gnatstub})
21502 Body file name. This should be set if the argument file name does not
21504 the GNAT file naming
21505 conventions. If this switch is omitted the default name for the body will be
21507 from the argument file name according to the GNAT file naming conventions.
21510 @cindex @option{^-q^/QUIET^} (@command{gnatstub})
21511 Quiet mode: do not generate a confirmation when a body is
21512 successfully created, and do not generate a message when a body is not
21516 @item ^-r^/TREE_FILE=REUSE^
21517 @cindex @option{^-r^/TREE_FILE=REUSE^} (@command{gnatstub})
21518 Reuse the tree file (if it exists) instead of creating it. Instead of
21519 creating the tree file for the library unit declaration, @command{gnatstub}
21520 tries to find it in the current directory and use it for creating
21521 a body. If the tree file is not found, no body is created. This option
21522 also implies @option{^-k^/SAVE^}, whether or not
21523 the latter is set explicitly.
21525 @item ^-t^/TREE_FILE=OVERWRITE^
21526 @cindex @option{^-t^/TREE_FILE=OVERWRITE^} (@command{gnatstub})
21527 Overwrite the existing tree file. If the current directory already
21528 contains the file which, according to the GNAT file naming rules should
21529 be considered as a tree file for the argument source file,
21531 will refuse to create the tree file needed to create a sample body
21532 unless this option is set.
21534 @item ^-v^/VERBOSE^
21535 @cindex @option{^-v^/VERBOSE^} (@command{gnatstub})
21536 Verbose mode: generate version information.
21540 @c *********************************
21541 @node Generating Ada Bindings for C and C++ headers
21542 @chapter Generating Ada Bindings for C and C++ headers
21546 GNAT now comes with a binding generator for C and C++ headers which is
21547 intended to do 95% of the tedious work of generating Ada specs from C
21548 or C++ header files.
21550 Note that this capability is not intended to generate 100% correct Ada specs,
21551 and will is some cases require manual adjustments, although it can often
21552 be used out of the box in practice.
21554 Some of the known limitations include:
21557 @item only very simple character constant macros are translated into Ada
21558 constants. Function macros (macros with arguments) are partially translated
21559 as comments, to be completed manually if needed.
21560 @item some extensions (e.g. vector types) are not supported
21561 @item pointers to pointers or complex structures are mapped to System.Address
21564 The code generated is using the Ada 2005 syntax, which makes it
21565 easier to interface with other languages than previous versions of Ada.
21568 * Running the binding generator::
21569 * Generating bindings for C++ headers::
21573 @node Running the binding generator
21574 @section Running the binding generator
21577 The binding generator is part of the @command{gcc} compiler and can be
21578 invoked via the @option{-fdump-ada-spec} switch, which will generate Ada
21579 spec files for the header files specified on the command line, and all
21580 header files needed by these files transitivitely. For example:
21583 $ g++ -c -fdump-ada-spec -C /usr/include/time.h
21584 $ gcc -c -gnat05 *.ads
21587 will generate, under GNU/Linux, the following files: @file{time_h.ads},
21588 @file{bits_time_h.ads}, @file{stddef_h.ads}, @file{bits_types_h.ads} which
21589 correspond to the files @file{/usr/include/time.h},
21590 @file{/usr/include/bits/time.h}, etc@dots{}, and will then compile in Ada 2005
21591 mode these Ada specs.
21593 The @code{-C} switch tells @command{gcc} to extract comments from headers,
21594 and will attempt to generate corresponding Ada comments.
21596 If you want to generate a single Ada file and not the transitive closure, you
21597 can use instead the @option{-fdump-ada-spec-slim} switch.
21599 Note that we recommend when possible to use the @command{g++} driver to
21600 generate bindings, even for most C headers, since this will in general
21601 generate better Ada specs. For generating bindings for C++ headers, it is
21602 mandatory to use the @command{g++} command, or @command{gcc -x c++} which
21603 is equivalent in this case. If @command{g++} cannot work on your C headers
21604 because of incompatibilities between C and C++, then you can fallback to
21605 @command{gcc} instead.
21607 For an example of better bindings generated from the C++ front-end,
21608 the name of the parameters (when available) are actually ignored by the C
21609 front-end. Consider the following C header:
21612 extern void foo (int variable);
21615 with the C front-end, @code{variable} is ignored, and the above is handled as:
21618 extern void foo (int);
21621 generating a generic:
21624 procedure foo (param1 : int);
21627 with the C++ front-end, the name is available, and we generate:
21630 procedure foo (variable : int);
21633 In some cases, the generated bindings will be more complete or more meaningful
21634 when defining some macros, which you can do via the @option{-D} switch. This
21635 is for example the case with @file{Xlib.h} under GNU/Linux:
21638 g++ -c -fdump-ada-spec -DXLIB_ILLEGAL_ACCESS -C /usr/include/X11/Xlib.h
21641 The above will generate more complete bindings than a straight call without
21642 the @option{-DXLIB_ILLEGAL_ACCESS} switch.
21644 In other cases, it is not possible to parse a header file in a stand alone
21645 manner, because other include files need to be included first. In this
21646 case, the solution is to create a small header file including the needed
21647 @code{#include} and possible @code{#define} directives. For example, to
21648 generate Ada bindings for @file{readline/readline.h}, you need to first
21649 include @file{stdio.h}, so you can create a file with the following two
21650 lines in e.g. @file{readline1.h}:
21654 #include <readline/readline.h>
21657 and then generate Ada bindings from this file:
21660 $ g++ -c -fdump-ada-spec readline1.h
21663 @node Generating bindings for C++ headers
21664 @section Generating bindings for C++ headers
21667 Generating bindings for C++ headers is done using the same options, always
21668 with the @command{g++} compiler.
21670 In this mode, C++ classes will be mapped to Ada tagged types, constructors
21671 will be mapped using the @code{CPP_Constructor} pragma, and when possible,
21672 multiple inheritance of abstract classes will be mapped to Ada interfaces
21673 (@xref{Interfacing to C++,,,gnat_rm, GNAT Reference Manual}, for additional
21674 information on interfacing to C++).
21676 For example, given the following C++ header file:
21683 virtual int Number_Of_Teeth () = 0;
21688 virtual void Set_Owner (char* Name) = 0;
21694 virtual void Set_Age (int New_Age);
21697 class Dog : Animal, Carnivore, Domestic @{
21702 virtual int Number_Of_Teeth ();
21703 virtual void Set_Owner (char* Name);
21711 The corresponding Ada code is generated:
21713 @smallexample @c ada
21716 package Class_Carnivore is
21717 type Carnivore is limited interface;
21718 pragma Import (CPP, Carnivore);
21720 function Number_Of_Teeth (this : access Carnivore) return int is abstract;
21722 use Class_Carnivore;
21724 package Class_Domestic is
21725 type Domestic is limited interface;
21726 pragma Import (CPP, Domestic);
21728 procedure Set_Owner
21729 (this : access Domestic;
21730 Name : Interfaces.C.Strings.chars_ptr) is abstract;
21732 use Class_Domestic;
21734 package Class_Animal is
21735 type Animal is tagged limited record
21736 Age_Count : aliased int;
21738 pragma Import (CPP, Animal);
21740 procedure Set_Age (this : access Animal; New_Age : int);
21741 pragma Import (CPP, Set_Age, "_ZN6Animal7Set_AgeEi");
21745 package Class_Dog is
21746 type Dog is new Animal and Carnivore and Domestic with record
21747 Tooth_Count : aliased int;
21748 Owner : Interfaces.C.Strings.chars_ptr;
21750 pragma Import (CPP, Dog);
21752 function Number_Of_Teeth (this : access Dog) return int;
21753 pragma Import (CPP, Number_Of_Teeth, "_ZN3Dog15Number_Of_TeethEv");
21755 procedure Set_Owner
21756 (this : access Dog; Name : Interfaces.C.Strings.chars_ptr);
21757 pragma Import (CPP, Set_Owner, "_ZN3Dog9Set_OwnerEPc");
21759 function New_Dog return Dog;
21760 pragma CPP_Constructor (New_Dog);
21761 pragma Import (CPP, New_Dog, "_ZN3DogC1Ev");
21772 @item -fdump-ada-spec
21773 @cindex @option{-fdump-ada-spec} (@command{gcc})
21774 Generate Ada spec files for the given header files transitively (including
21775 all header files that these headers depend upon).
21777 @item -fdump-ada-spec-slim
21778 @cindex @option{-fdump-ada-spec-slim} (@command{gcc})
21779 Generate Ada spec files for the header files specified on the command line
21783 @cindex @option{-C} (@command{gcc})
21784 Extract comments from headers and generate Ada comments in the Ada spec files.
21787 @node Other Utility Programs
21788 @chapter Other Utility Programs
21791 This chapter discusses some other utility programs available in the Ada
21795 * Using Other Utility Programs with GNAT::
21796 * The External Symbol Naming Scheme of GNAT::
21797 * Converting Ada Files to html with gnathtml::
21798 * Installing gnathtml::
21805 @node Using Other Utility Programs with GNAT
21806 @section Using Other Utility Programs with GNAT
21809 The object files generated by GNAT are in standard system format and in
21810 particular the debugging information uses this format. This means
21811 programs generated by GNAT can be used with existing utilities that
21812 depend on these formats.
21815 In general, any utility program that works with C will also often work with
21816 Ada programs generated by GNAT. This includes software utilities such as
21817 gprof (a profiling program), @code{gdb} (the FSF debugger), and utilities such
21821 @node The External Symbol Naming Scheme of GNAT
21822 @section The External Symbol Naming Scheme of GNAT
21825 In order to interpret the output from GNAT, when using tools that are
21826 originally intended for use with other languages, it is useful to
21827 understand the conventions used to generate link names from the Ada
21830 All link names are in all lowercase letters. With the exception of library
21831 procedure names, the mechanism used is simply to use the full expanded
21832 Ada name with dots replaced by double underscores. For example, suppose
21833 we have the following package spec:
21835 @smallexample @c ada
21846 The variable @code{MN} has a full expanded Ada name of @code{QRS.MN}, so
21847 the corresponding link name is @code{qrs__mn}.
21849 Of course if a @code{pragma Export} is used this may be overridden:
21851 @smallexample @c ada
21856 pragma Export (Var1, C, External_Name => "var1_name");
21858 pragma Export (Var2, C, Link_Name => "var2_link_name");
21865 In this case, the link name for @var{Var1} is whatever link name the
21866 C compiler would assign for the C function @var{var1_name}. This typically
21867 would be either @var{var1_name} or @var{_var1_name}, depending on operating
21868 system conventions, but other possibilities exist. The link name for
21869 @var{Var2} is @var{var2_link_name}, and this is not operating system
21873 One exception occurs for library level procedures. A potential ambiguity
21874 arises between the required name @code{_main} for the C main program,
21875 and the name we would otherwise assign to an Ada library level procedure
21876 called @code{Main} (which might well not be the main program).
21878 To avoid this ambiguity, we attach the prefix @code{_ada_} to such
21879 names. So if we have a library level procedure such as
21881 @smallexample @c ada
21884 procedure Hello (S : String);
21890 the external name of this procedure will be @var{_ada_hello}.
21893 @node Converting Ada Files to html with gnathtml
21894 @section Converting Ada Files to HTML with @code{gnathtml}
21897 This @code{Perl} script allows Ada source files to be browsed using
21898 standard Web browsers. For installation procedure, see the section
21899 @xref{Installing gnathtml}.
21901 Ada reserved keywords are highlighted in a bold font and Ada comments in
21902 a blue font. Unless your program was compiled with the gcc @option{-gnatx}
21903 switch to suppress the generation of cross-referencing information, user
21904 defined variables and types will appear in a different color; you will
21905 be able to click on any identifier and go to its declaration.
21907 The command line is as follow:
21909 @c $ perl gnathtml.pl @ovar{^switches^options^} @var{ada-files}
21910 @c Expanding @ovar macro inline (explanation in macro def comments)
21911 $ perl gnathtml.pl @r{[}@var{^switches^options^}@r{]} @var{ada-files}
21915 You can pass it as many Ada files as you want. @code{gnathtml} will generate
21916 an html file for every ada file, and a global file called @file{index.htm}.
21917 This file is an index of every identifier defined in the files.
21919 The available ^switches^options^ are the following ones:
21923 @cindex @option{-83} (@code{gnathtml})
21924 Only the Ada 83 subset of keywords will be highlighted.
21926 @item -cc @var{color}
21927 @cindex @option{-cc} (@code{gnathtml})
21928 This option allows you to change the color used for comments. The default
21929 value is green. The color argument can be any name accepted by html.
21932 @cindex @option{-d} (@code{gnathtml})
21933 If the Ada files depend on some other files (for instance through
21934 @code{with} clauses, the latter files will also be converted to html.
21935 Only the files in the user project will be converted to html, not the files
21936 in the run-time library itself.
21939 @cindex @option{-D} (@code{gnathtml})
21940 This command is the same as @option{-d} above, but @command{gnathtml} will
21941 also look for files in the run-time library, and generate html files for them.
21943 @item -ext @var{extension}
21944 @cindex @option{-ext} (@code{gnathtml})
21945 This option allows you to change the extension of the generated HTML files.
21946 If you do not specify an extension, it will default to @file{htm}.
21949 @cindex @option{-f} (@code{gnathtml})
21950 By default, gnathtml will generate html links only for global entities
21951 ('with'ed units, global variables and types,@dots{}). If you specify
21952 @option{-f} on the command line, then links will be generated for local
21955 @item -l @var{number}
21956 @cindex @option{-l} (@code{gnathtml})
21957 If this ^switch^option^ is provided and @var{number} is not 0, then
21958 @code{gnathtml} will number the html files every @var{number} line.
21961 @cindex @option{-I} (@code{gnathtml})
21962 Specify a directory to search for library files (@file{.ALI} files) and
21963 source files. You can provide several -I switches on the command line,
21964 and the directories will be parsed in the order of the command line.
21967 @cindex @option{-o} (@code{gnathtml})
21968 Specify the output directory for html files. By default, gnathtml will
21969 saved the generated html files in a subdirectory named @file{html/}.
21971 @item -p @var{file}
21972 @cindex @option{-p} (@code{gnathtml})
21973 If you are using Emacs and the most recent Emacs Ada mode, which provides
21974 a full Integrated Development Environment for compiling, checking,
21975 running and debugging applications, you may use @file{.gpr} files
21976 to give the directories where Emacs can find sources and object files.
21978 Using this ^switch^option^, you can tell gnathtml to use these files.
21979 This allows you to get an html version of your application, even if it
21980 is spread over multiple directories.
21982 @item -sc @var{color}
21983 @cindex @option{-sc} (@code{gnathtml})
21984 This ^switch^option^ allows you to change the color used for symbol
21986 The default value is red. The color argument can be any name accepted by html.
21988 @item -t @var{file}
21989 @cindex @option{-t} (@code{gnathtml})
21990 This ^switch^option^ provides the name of a file. This file contains a list of
21991 file names to be converted, and the effect is exactly as though they had
21992 appeared explicitly on the command line. This
21993 is the recommended way to work around the command line length limit on some
21998 @node Installing gnathtml
21999 @section Installing @code{gnathtml}
22002 @code{Perl} needs to be installed on your machine to run this script.
22003 @code{Perl} is freely available for almost every architecture and
22004 Operating System via the Internet.
22006 On Unix systems, you may want to modify the first line of the script
22007 @code{gnathtml}, to explicitly tell the Operating system where Perl
22008 is. The syntax of this line is:
22010 #!full_path_name_to_perl
22014 Alternatively, you may run the script using the following command line:
22017 @c $ perl gnathtml.pl @ovar{switches} @var{files}
22018 @c Expanding @ovar macro inline (explanation in macro def comments)
22019 $ perl gnathtml.pl @r{[}@var{switches}@r{]} @var{files}
22028 The GNAT distribution provides an Ada 95 template for the HP Language
22029 Sensitive Editor (LSE), a component of DECset. In order to
22030 access it, invoke LSE with the qualifier /ENVIRONMENT=GNU:[LIB]ADA95.ENV.
22037 GNAT supports The HP Performance Coverage Analyzer (PCA), a component
22038 of DECset. To use it proceed as outlined under ``HELP PCA'', except for running
22039 the collection phase with the /DEBUG qualifier.
22042 $ GNAT MAKE /DEBUG <PROGRAM_NAME>
22043 $ DEFINE LIB$DEBUG PCA$COLLECTOR
22044 $ RUN/DEBUG <PROGRAM_NAME>
22050 @c ******************************
22051 @node Code Coverage and Profiling
22052 @chapter Code Coverage and Profiling
22053 @cindex Code Coverage
22057 This chapter describes how to use @code{gcov} - coverage testing tool - and
22058 @code{gprof} - profiler tool - on your Ada programs.
22061 * Code Coverage of Ada Programs using gcov::
22062 * Profiling an Ada Program using gprof::
22065 @node Code Coverage of Ada Programs using gcov
22066 @section Code Coverage of Ada Programs using gcov
22068 @cindex -fprofile-arcs
22069 @cindex -ftest-coverage
22071 @cindex Code Coverage
22074 @code{gcov} is a test coverage program: it analyzes the execution of a given
22075 program on selected tests, to help you determine the portions of the program
22076 that are still untested.
22078 @code{gcov} is part of the GCC suite, and is described in detail in the GCC
22079 User's Guide. You can refer to this documentation for a more complete
22082 This chapter provides a quick startup guide, and
22083 details some Gnat-specific features.
22086 * Quick startup guide::
22090 @node Quick startup guide
22091 @subsection Quick startup guide
22093 In order to perform coverage analysis of a program using @code{gcov}, 3
22098 Code instrumentation during the compilation process
22100 Execution of the instrumented program
22102 Execution of the @code{gcov} tool to generate the result.
22105 The code instrumentation needed by gcov is created at the object level:
22106 The source code is not modified in any way, because the instrumentation code is
22107 inserted by gcc during the compilation process. To compile your code with code
22108 coverage activated, you need to recompile your whole project using the
22110 @code{-fprofile-arcs} and @code{-ftest-coverage}, and link it using
22111 @code{-fprofile-arcs}.
22114 $ gnatmake -P my_project.gpr -f -cargs -fprofile-arcs -ftest-coverage \
22115 -largs -fprofile-arcs
22118 This compilation process will create @file{.gcno} files together with
22119 the usual object files.
22121 Once the program is compiled with coverage instrumentation, you can
22122 run it as many times as needed - on portions of a test suite for
22123 example. The first execution will produce @file{.gcda} files at the
22124 same location as the @file{.gcno} files. The following executions
22125 will update those files, so that a cumulative result of the covered
22126 portions of the program is generated.
22128 Finally, you need to call the @code{gcov} tool. The different options of
22129 @code{gcov} are available in the GCC User's Guide, section 'Invoking gcov'.
22131 This will create annotated source files with a @file{.gcov} extension:
22132 @file{my_main.adb} file will be analysed in @file{my_main.adb.gcov}.
22134 @node Gnat specifics
22135 @subsection Gnat specifics
22137 Because Ada semantics, portions of the source code may be shared among
22138 several object files. This is the case for example when generics are
22139 involved, when inlining is active or when declarations generate initialisation
22140 calls. In order to take
22141 into account this shared code, you need to call @code{gcov} on all
22142 source files of the tested program at once.
22144 The list of source files might exceed the system's maximum command line
22145 length. In order to bypass this limitation, a new mechanism has been
22146 implemented in @code{gcov}: you can now list all your project's files into a
22147 text file, and provide this file to gcov as a parameter, preceded by a @@
22148 (e.g. @samp{gcov @@mysrclist.txt}).
22150 Note that on AIX compiling a static library with @code{-fprofile-arcs} is
22151 not supported as there can be unresolved symbols during the final link.
22153 @node Profiling an Ada Program using gprof
22154 @section Profiling an Ada Program using gprof
22160 This section is not meant to be an exhaustive documentation of @code{gprof}.
22161 Full documentation for it can be found in the GNU Profiler User's Guide
22162 documentation that is part of this GNAT distribution.
22164 Profiling a program helps determine the parts of a program that are executed
22165 most often, and are therefore the most time-consuming.
22167 @code{gprof} is the standard GNU profiling tool; it has been enhanced to
22168 better handle Ada programs and multitasking.
22169 It is currently supported on the following platforms
22174 solaris sparc/sparc64/x86
22180 In order to profile a program using @code{gprof}, 3 steps are needed:
22184 Code instrumentation, requiring a full recompilation of the project with the
22187 Execution of the program under the analysis conditions, i.e. with the desired
22190 Analysis of the results using the @code{gprof} tool.
22194 The following sections detail the different steps, and indicate how
22195 to interpret the results:
22197 * Compilation for profiling::
22198 * Program execution::
22200 * Interpretation of profiling results::
22203 @node Compilation for profiling
22204 @subsection Compilation for profiling
22208 In order to profile a program the first step is to tell the compiler
22209 to generate the necessary profiling information. The compiler switch to be used
22210 is @code{-pg}, which must be added to other compilation switches. This
22211 switch needs to be specified both during compilation and link stages, and can
22212 be specified once when using gnatmake:
22215 gnatmake -f -pg -P my_project
22219 Note that only the objects that were compiled with the @samp{-pg} switch will be
22220 profiled; if you need to profile your whole project, use the
22221 @samp{-f} gnatmake switch to force full recompilation.
22223 @node Program execution
22224 @subsection Program execution
22227 Once the program has been compiled for profiling, you can run it as usual.
22229 The only constraint imposed by profiling is that the program must terminate
22230 normally. An interrupted program (via a Ctrl-C, kill, etc.) will not be
22233 Once the program completes execution, a data file called @file{gmon.out} is
22234 generated in the directory where the program was launched from. If this file
22235 already exists, it will be overwritten.
22237 @node Running gprof
22238 @subsection Running gprof
22241 The @code{gprof} tool is called as follow:
22244 gprof my_prog gmon.out
22255 The complete form of the gprof command line is the following:
22258 gprof [^switches^options^] [executable [data-file]]
22262 @code{gprof} supports numerous ^switch^options^. The order of these
22263 ^switch^options^ does not matter. The full list of options can be found in
22264 the GNU Profiler User's Guide documentation that comes with this documentation.
22266 The following is the subset of those switches that is most relevant:
22270 @item --demangle[=@var{style}]
22271 @itemx --no-demangle
22272 @cindex @option{--demangle} (@code{gprof})
22273 These options control whether symbol names should be demangled when
22274 printing output. The default is to demangle C++ symbols. The
22275 @code{--no-demangle} option may be used to turn off demangling. Different
22276 compilers have different mangling styles. The optional demangling style
22277 argument can be used to choose an appropriate demangling style for your
22278 compiler, in particular Ada symbols generated by GNAT can be demangled using
22279 @code{--demangle=gnat}.
22281 @item -e @var{function_name}
22282 @cindex @option{-e} (@code{gprof})
22283 The @samp{-e @var{function}} option tells @code{gprof} not to print
22284 information about the function @var{function_name} (and its
22285 children@dots{}) in the call graph. The function will still be listed
22286 as a child of any functions that call it, but its index number will be
22287 shown as @samp{[not printed]}. More than one @samp{-e} option may be
22288 given; only one @var{function_name} may be indicated with each @samp{-e}
22291 @item -E @var{function_name}
22292 @cindex @option{-E} (@code{gprof})
22293 The @code{-E @var{function}} option works like the @code{-e} option, but
22294 execution time spent in the function (and children who were not called from
22295 anywhere else), will not be used to compute the percentages-of-time for
22296 the call graph. More than one @samp{-E} option may be given; only one
22297 @var{function_name} may be indicated with each @samp{-E} option.
22299 @item -f @var{function_name}
22300 @cindex @option{-f} (@code{gprof})
22301 The @samp{-f @var{function}} option causes @code{gprof} to limit the
22302 call graph to the function @var{function_name} and its children (and
22303 their children@dots{}). More than one @samp{-f} option may be given;
22304 only one @var{function_name} may be indicated with each @samp{-f}
22307 @item -F @var{function_name}
22308 @cindex @option{-F} (@code{gprof})
22309 The @samp{-F @var{function}} option works like the @code{-f} option, but
22310 only time spent in the function and its children (and their
22311 children@dots{}) will be used to determine total-time and
22312 percentages-of-time for the call graph. More than one @samp{-F} option
22313 may be given; only one @var{function_name} may be indicated with each
22314 @samp{-F} option. The @samp{-F} option overrides the @samp{-E} option.
22318 @node Interpretation of profiling results
22319 @subsection Interpretation of profiling results
22323 The results of the profiling analysis are represented by two arrays: the
22324 'flat profile' and the 'call graph'. Full documentation of those outputs
22325 can be found in the GNU Profiler User's Guide.
22327 The flat profile shows the time spent in each function of the program, and how
22328 many time it has been called. This allows you to locate easily the most
22329 time-consuming functions.
22331 The call graph shows, for each subprogram, the subprograms that call it,
22332 and the subprograms that it calls. It also provides an estimate of the time
22333 spent in each of those callers/called subprograms.
22336 @c ******************************
22337 @node Running and Debugging Ada Programs
22338 @chapter Running and Debugging Ada Programs
22342 This chapter discusses how to debug Ada programs.
22344 It applies to GNAT on the Alpha OpenVMS platform;
22345 for I64 OpenVMS please refer to the @cite{OpenVMS Debugger Manual},
22346 since HP has implemented Ada support in the OpenVMS debugger on I64.
22349 An incorrect Ada program may be handled in three ways by the GNAT compiler:
22353 The illegality may be a violation of the static semantics of Ada. In
22354 that case GNAT diagnoses the constructs in the program that are illegal.
22355 It is then a straightforward matter for the user to modify those parts of
22359 The illegality may be a violation of the dynamic semantics of Ada. In
22360 that case the program compiles and executes, but may generate incorrect
22361 results, or may terminate abnormally with some exception.
22364 When presented with a program that contains convoluted errors, GNAT
22365 itself may terminate abnormally without providing full diagnostics on
22366 the incorrect user program.
22370 * The GNAT Debugger GDB::
22372 * Introduction to GDB Commands::
22373 * Using Ada Expressions::
22374 * Calling User-Defined Subprograms::
22375 * Using the Next Command in a Function::
22378 * Debugging Generic Units::
22379 * Remote Debugging using gdbserver::
22380 * GNAT Abnormal Termination or Failure to Terminate::
22381 * Naming Conventions for GNAT Source Files::
22382 * Getting Internal Debugging Information::
22383 * Stack Traceback::
22389 @node The GNAT Debugger GDB
22390 @section The GNAT Debugger GDB
22393 @code{GDB} is a general purpose, platform-independent debugger that
22394 can be used to debug mixed-language programs compiled with @command{gcc},
22395 and in particular is capable of debugging Ada programs compiled with
22396 GNAT. The latest versions of @code{GDB} are Ada-aware and can handle
22397 complex Ada data structures.
22399 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
22401 located in the GNU:[DOCS] directory,
22403 for full details on the usage of @code{GDB}, including a section on
22404 its usage on programs. This manual should be consulted for full
22405 details. The section that follows is a brief introduction to the
22406 philosophy and use of @code{GDB}.
22408 When GNAT programs are compiled, the compiler optionally writes debugging
22409 information into the generated object file, including information on
22410 line numbers, and on declared types and variables. This information is
22411 separate from the generated code. It makes the object files considerably
22412 larger, but it does not add to the size of the actual executable that
22413 will be loaded into memory, and has no impact on run-time performance. The
22414 generation of debug information is triggered by the use of the
22415 ^-g^/DEBUG^ switch in the @command{gcc} or @command{gnatmake} command
22416 used to carry out the compilations. It is important to emphasize that
22417 the use of these options does not change the generated code.
22419 The debugging information is written in standard system formats that
22420 are used by many tools, including debuggers and profilers. The format
22421 of the information is typically designed to describe C types and
22422 semantics, but GNAT implements a translation scheme which allows full
22423 details about Ada types and variables to be encoded into these
22424 standard C formats. Details of this encoding scheme may be found in
22425 the file exp_dbug.ads in the GNAT source distribution. However, the
22426 details of this encoding are, in general, of no interest to a user,
22427 since @code{GDB} automatically performs the necessary decoding.
22429 When a program is bound and linked, the debugging information is
22430 collected from the object files, and stored in the executable image of
22431 the program. Again, this process significantly increases the size of
22432 the generated executable file, but it does not increase the size of
22433 the executable program itself. Furthermore, if this program is run in
22434 the normal manner, it runs exactly as if the debug information were
22435 not present, and takes no more actual memory.
22437 However, if the program is run under control of @code{GDB}, the
22438 debugger is activated. The image of the program is loaded, at which
22439 point it is ready to run. If a run command is given, then the program
22440 will run exactly as it would have if @code{GDB} were not present. This
22441 is a crucial part of the @code{GDB} design philosophy. @code{GDB} is
22442 entirely non-intrusive until a breakpoint is encountered. If no
22443 breakpoint is ever hit, the program will run exactly as it would if no
22444 debugger were present. When a breakpoint is hit, @code{GDB} accesses
22445 the debugging information and can respond to user commands to inspect
22446 variables, and more generally to report on the state of execution.
22450 @section Running GDB
22453 This section describes how to initiate the debugger.
22454 @c The above sentence is really just filler, but it was otherwise
22455 @c clumsy to get the first paragraph nonindented given the conditional
22456 @c nature of the description
22459 The debugger can be launched from a @code{GPS} menu or
22460 directly from the command line. The description below covers the latter use.
22461 All the commands shown can be used in the @code{GPS} debug console window,
22462 but there are usually more GUI-based ways to achieve the same effect.
22465 The command to run @code{GDB} is
22468 $ ^gdb program^GDB PROGRAM^
22472 where @code{^program^PROGRAM^} is the name of the executable file. This
22473 activates the debugger and results in a prompt for debugger commands.
22474 The simplest command is simply @code{run}, which causes the program to run
22475 exactly as if the debugger were not present. The following section
22476 describes some of the additional commands that can be given to @code{GDB}.
22478 @c *******************************
22479 @node Introduction to GDB Commands
22480 @section Introduction to GDB Commands
22483 @code{GDB} contains a large repertoire of commands. @xref{Top,,
22484 Debugging with GDB, gdb, Debugging with GDB},
22486 located in the GNU:[DOCS] directory,
22488 for extensive documentation on the use
22489 of these commands, together with examples of their use. Furthermore,
22490 the command @command{help} invoked from within GDB activates a simple help
22491 facility which summarizes the available commands and their options.
22492 In this section we summarize a few of the most commonly
22493 used commands to give an idea of what @code{GDB} is about. You should create
22494 a simple program with debugging information and experiment with the use of
22495 these @code{GDB} commands on the program as you read through the
22499 @item set args @var{arguments}
22500 The @var{arguments} list above is a list of arguments to be passed to
22501 the program on a subsequent run command, just as though the arguments
22502 had been entered on a normal invocation of the program. The @code{set args}
22503 command is not needed if the program does not require arguments.
22506 The @code{run} command causes execution of the program to start from
22507 the beginning. If the program is already running, that is to say if
22508 you are currently positioned at a breakpoint, then a prompt will ask
22509 for confirmation that you want to abandon the current execution and
22512 @item breakpoint @var{location}
22513 The breakpoint command sets a breakpoint, that is to say a point at which
22514 execution will halt and @code{GDB} will await further
22515 commands. @var{location} is
22516 either a line number within a file, given in the format @code{file:linenumber},
22517 or it is the name of a subprogram. If you request that a breakpoint be set on
22518 a subprogram that is overloaded, a prompt will ask you to specify on which of
22519 those subprograms you want to breakpoint. You can also
22520 specify that all of them should be breakpointed. If the program is run
22521 and execution encounters the breakpoint, then the program
22522 stops and @code{GDB} signals that the breakpoint was encountered by
22523 printing the line of code before which the program is halted.
22525 @item catch exception @var{name}
22526 This command causes the program execution to stop whenever exception
22527 @var{name} is raised. If @var{name} is omitted, then the execution is
22528 suspended when any exception is raised.
22530 @item print @var{expression}
22531 This will print the value of the given expression. Most simple
22532 Ada expression formats are properly handled by @code{GDB}, so the expression
22533 can contain function calls, variables, operators, and attribute references.
22536 Continues execution following a breakpoint, until the next breakpoint or the
22537 termination of the program.
22540 Executes a single line after a breakpoint. If the next statement
22541 is a subprogram call, execution continues into (the first statement of)
22542 the called subprogram.
22545 Executes a single line. If this line is a subprogram call, executes and
22546 returns from the call.
22549 Lists a few lines around the current source location. In practice, it
22550 is usually more convenient to have a separate edit window open with the
22551 relevant source file displayed. Successive applications of this command
22552 print subsequent lines. The command can be given an argument which is a
22553 line number, in which case it displays a few lines around the specified one.
22556 Displays a backtrace of the call chain. This command is typically
22557 used after a breakpoint has occurred, to examine the sequence of calls that
22558 leads to the current breakpoint. The display includes one line for each
22559 activation record (frame) corresponding to an active subprogram.
22562 At a breakpoint, @code{GDB} can display the values of variables local
22563 to the current frame. The command @code{up} can be used to
22564 examine the contents of other active frames, by moving the focus up
22565 the stack, that is to say from callee to caller, one frame at a time.
22568 Moves the focus of @code{GDB} down from the frame currently being
22569 examined to the frame of its callee (the reverse of the previous command),
22571 @item frame @var{n}
22572 Inspect the frame with the given number. The value 0 denotes the frame
22573 of the current breakpoint, that is to say the top of the call stack.
22578 The above list is a very short introduction to the commands that
22579 @code{GDB} provides. Important additional capabilities, including conditional
22580 breakpoints, the ability to execute command sequences on a breakpoint,
22581 the ability to debug at the machine instruction level and many other
22582 features are described in detail in @ref{Top,, Debugging with GDB, gdb,
22583 Debugging with GDB}. Note that most commands can be abbreviated
22584 (for example, c for continue, bt for backtrace).
22586 @node Using Ada Expressions
22587 @section Using Ada Expressions
22588 @cindex Ada expressions
22591 @code{GDB} supports a fairly large subset of Ada expression syntax, with some
22592 extensions. The philosophy behind the design of this subset is
22596 That @code{GDB} should provide basic literals and access to operations for
22597 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
22598 leaving more sophisticated computations to subprograms written into the
22599 program (which therefore may be called from @code{GDB}).
22602 That type safety and strict adherence to Ada language restrictions
22603 are not particularly important to the @code{GDB} user.
22606 That brevity is important to the @code{GDB} user.
22610 Thus, for brevity, the debugger acts as if there were
22611 implicit @code{with} and @code{use} clauses in effect for all user-written
22612 packages, thus making it unnecessary to fully qualify most names with
22613 their packages, regardless of context. Where this causes ambiguity,
22614 @code{GDB} asks the user's intent.
22616 For details on the supported Ada syntax, see @ref{Top,, Debugging with
22617 GDB, gdb, Debugging with GDB}.
22619 @node Calling User-Defined Subprograms
22620 @section Calling User-Defined Subprograms
22623 An important capability of @code{GDB} is the ability to call user-defined
22624 subprograms while debugging. This is achieved simply by entering
22625 a subprogram call statement in the form:
22628 call subprogram-name (parameters)
22632 The keyword @code{call} can be omitted in the normal case where the
22633 @code{subprogram-name} does not coincide with any of the predefined
22634 @code{GDB} commands.
22636 The effect is to invoke the given subprogram, passing it the
22637 list of parameters that is supplied. The parameters can be expressions and
22638 can include variables from the program being debugged. The
22639 subprogram must be defined
22640 at the library level within your program, and @code{GDB} will call the
22641 subprogram within the environment of your program execution (which
22642 means that the subprogram is free to access or even modify variables
22643 within your program).
22645 The most important use of this facility is in allowing the inclusion of
22646 debugging routines that are tailored to particular data structures
22647 in your program. Such debugging routines can be written to provide a suitably
22648 high-level description of an abstract type, rather than a low-level dump
22649 of its physical layout. After all, the standard
22650 @code{GDB print} command only knows the physical layout of your
22651 types, not their abstract meaning. Debugging routines can provide information
22652 at the desired semantic level and are thus enormously useful.
22654 For example, when debugging GNAT itself, it is crucial to have access to
22655 the contents of the tree nodes used to represent the program internally.
22656 But tree nodes are represented simply by an integer value (which in turn
22657 is an index into a table of nodes).
22658 Using the @code{print} command on a tree node would simply print this integer
22659 value, which is not very useful. But the PN routine (defined in file
22660 treepr.adb in the GNAT sources) takes a tree node as input, and displays
22661 a useful high level representation of the tree node, which includes the
22662 syntactic category of the node, its position in the source, the integers
22663 that denote descendant nodes and parent node, as well as varied
22664 semantic information. To study this example in more detail, you might want to
22665 look at the body of the PN procedure in the stated file.
22667 @node Using the Next Command in a Function
22668 @section Using the Next Command in a Function
22671 When you use the @code{next} command in a function, the current source
22672 location will advance to the next statement as usual. A special case
22673 arises in the case of a @code{return} statement.
22675 Part of the code for a return statement is the ``epilog'' of the function.
22676 This is the code that returns to the caller. There is only one copy of
22677 this epilog code, and it is typically associated with the last return
22678 statement in the function if there is more than one return. In some
22679 implementations, this epilog is associated with the first statement
22682 The result is that if you use the @code{next} command from a return
22683 statement that is not the last return statement of the function you
22684 may see a strange apparent jump to the last return statement or to
22685 the start of the function. You should simply ignore this odd jump.
22686 The value returned is always that from the first return statement
22687 that was stepped through.
22689 @node Ada Exceptions
22690 @section Stopping when Ada Exceptions are Raised
22694 You can set catchpoints that stop the program execution when your program
22695 raises selected exceptions.
22698 @item catch exception
22699 Set a catchpoint that stops execution whenever (any task in the) program
22700 raises any exception.
22702 @item catch exception @var{name}
22703 Set a catchpoint that stops execution whenever (any task in the) program
22704 raises the exception @var{name}.
22706 @item catch exception unhandled
22707 Set a catchpoint that stops executino whenever (any task in the) program
22708 raises an exception for which there is no handler.
22710 @item info exceptions
22711 @itemx info exceptions @var{regexp}
22712 The @code{info exceptions} command permits the user to examine all defined
22713 exceptions within Ada programs. With a regular expression, @var{regexp}, as
22714 argument, prints out only those exceptions whose name matches @var{regexp}.
22722 @code{GDB} allows the following task-related commands:
22726 This command shows a list of current Ada tasks, as in the following example:
22733 ID TID P-ID Thread Pri State Name
22734 1 8088000 0 807e000 15 Child Activation Wait main_task
22735 2 80a4000 1 80ae000 15 Accept/Select Wait b
22736 3 809a800 1 80a4800 15 Child Activation Wait a
22737 * 4 80ae800 3 80b8000 15 Running c
22741 In this listing, the asterisk before the first task indicates it to be the
22742 currently running task. The first column lists the task ID that is used
22743 to refer to tasks in the following commands.
22745 @item break @var{linespec} task @var{taskid}
22746 @itemx break @var{linespec} task @var{taskid} if @dots{}
22747 @cindex Breakpoints and tasks
22748 These commands are like the @code{break @dots{} thread @dots{}}.
22749 @var{linespec} specifies source lines.
22751 Use the qualifier @samp{task @var{taskid}} with a breakpoint command
22752 to specify that you only want @code{GDB} to stop the program when a
22753 particular Ada task reaches this breakpoint. @var{taskid} is one of the
22754 numeric task identifiers assigned by @code{GDB}, shown in the first
22755 column of the @samp{info tasks} display.
22757 If you do not specify @samp{task @var{taskid}} when you set a
22758 breakpoint, the breakpoint applies to @emph{all} tasks of your
22761 You can use the @code{task} qualifier on conditional breakpoints as
22762 well; in this case, place @samp{task @var{taskid}} before the
22763 breakpoint condition (before the @code{if}).
22765 @item task @var{taskno}
22766 @cindex Task switching
22768 This command allows to switch to the task referred by @var{taskno}. In
22769 particular, This allows to browse the backtrace of the specified
22770 task. It is advised to switch back to the original task before
22771 continuing execution otherwise the scheduling of the program may be
22776 For more detailed information on the tasking support,
22777 see @ref{Top,, Debugging with GDB, gdb, Debugging with GDB}.
22779 @node Debugging Generic Units
22780 @section Debugging Generic Units
22781 @cindex Debugging Generic Units
22785 GNAT always uses code expansion for generic instantiation. This means that
22786 each time an instantiation occurs, a complete copy of the original code is
22787 made, with appropriate substitutions of formals by actuals.
22789 It is not possible to refer to the original generic entities in
22790 @code{GDB}, but it is always possible to debug a particular instance of
22791 a generic, by using the appropriate expanded names. For example, if we have
22793 @smallexample @c ada
22798 generic package k is
22799 procedure kp (v1 : in out integer);
22803 procedure kp (v1 : in out integer) is
22809 package k1 is new k;
22810 package k2 is new k;
22812 var : integer := 1;
22825 Then to break on a call to procedure kp in the k2 instance, simply
22829 (gdb) break g.k2.kp
22833 When the breakpoint occurs, you can step through the code of the
22834 instance in the normal manner and examine the values of local variables, as for
22837 @node Remote Debugging using gdbserver
22838 @section Remote Debugging using gdbserver
22839 @cindex Remote Debugging using gdbserver
22842 On platforms where gdbserver is supported, it is possible to use this tool
22843 to debug your application remotely. This can be useful in situations
22844 where the program needs to be run on a target host that is different
22845 from the host used for development, particularly when the target has
22846 a limited amount of resources (either CPU and/or memory).
22848 To do so, start your program using gdbserver on the target machine.
22849 gdbserver then automatically suspends the execution of your program
22850 at its entry point, waiting for a debugger to connect to it. The
22851 following commands starts an application and tells gdbserver to
22852 wait for a connection with the debugger on localhost port 4444.
22855 $ gdbserver localhost:4444 program
22856 Process program created; pid = 5685
22857 Listening on port 4444
22860 Once gdbserver has started listening, we can tell the debugger to establish
22861 a connection with this gdbserver, and then start the same debugging session
22862 as if the program was being debugged on the same host, directly under
22863 the control of GDB.
22867 (gdb) target remote targethost:4444
22868 Remote debugging using targethost:4444
22869 0x00007f29936d0af0 in ?? () from /lib64/ld-linux-x86-64.so.
22871 Breakpoint 1 at 0x401f0c: file foo.adb, line 3.
22875 Breakpoint 1, foo () at foo.adb:4
22879 It is also possible to use gdbserver to attach to an already running
22880 program, in which case the execution of that program is simply suspended
22881 until the connection between the debugger and gdbserver is established.
22883 For more information on how to use gdbserver, @ref{Top, Server, Using
22884 the gdbserver Program, gdb, Debugging with GDB}. GNAT Pro provides support
22885 for gdbserver on x86-linux, x86-windows and x86_64-linux.
22887 @node GNAT Abnormal Termination or Failure to Terminate
22888 @section GNAT Abnormal Termination or Failure to Terminate
22889 @cindex GNAT Abnormal Termination or Failure to Terminate
22892 When presented with programs that contain serious errors in syntax
22894 GNAT may on rare occasions experience problems in operation, such
22896 segmentation fault or illegal memory access, raising an internal
22897 exception, terminating abnormally, or failing to terminate at all.
22898 In such cases, you can activate
22899 various features of GNAT that can help you pinpoint the construct in your
22900 program that is the likely source of the problem.
22902 The following strategies are presented in increasing order of
22903 difficulty, corresponding to your experience in using GNAT and your
22904 familiarity with compiler internals.
22908 Run @command{gcc} with the @option{-gnatf}. This first
22909 switch causes all errors on a given line to be reported. In its absence,
22910 only the first error on a line is displayed.
22912 The @option{-gnatdO} switch causes errors to be displayed as soon as they
22913 are encountered, rather than after compilation is terminated. If GNAT
22914 terminates prematurely or goes into an infinite loop, the last error
22915 message displayed may help to pinpoint the culprit.
22918 Run @command{gcc} with the @option{^-v (verbose)^/VERBOSE^} switch. In this
22919 mode, @command{gcc} produces ongoing information about the progress of the
22920 compilation and provides the name of each procedure as code is
22921 generated. This switch allows you to find which Ada procedure was being
22922 compiled when it encountered a code generation problem.
22925 @cindex @option{-gnatdc} switch
22926 Run @command{gcc} with the @option{-gnatdc} switch. This is a GNAT specific
22927 switch that does for the front-end what @option{^-v^VERBOSE^} does
22928 for the back end. The system prints the name of each unit,
22929 either a compilation unit or nested unit, as it is being analyzed.
22931 Finally, you can start
22932 @code{gdb} directly on the @code{gnat1} executable. @code{gnat1} is the
22933 front-end of GNAT, and can be run independently (normally it is just
22934 called from @command{gcc}). You can use @code{gdb} on @code{gnat1} as you
22935 would on a C program (but @pxref{The GNAT Debugger GDB} for caveats). The
22936 @code{where} command is the first line of attack; the variable
22937 @code{lineno} (seen by @code{print lineno}), used by the second phase of
22938 @code{gnat1} and by the @command{gcc} backend, indicates the source line at
22939 which the execution stopped, and @code{input_file name} indicates the name of
22943 @node Naming Conventions for GNAT Source Files
22944 @section Naming Conventions for GNAT Source Files
22947 In order to examine the workings of the GNAT system, the following
22948 brief description of its organization may be helpful:
22952 Files with prefix @file{^sc^SC^} contain the lexical scanner.
22955 All files prefixed with @file{^par^PAR^} are components of the parser. The
22956 numbers correspond to chapters of the Ada Reference Manual. For example,
22957 parsing of select statements can be found in @file{par-ch9.adb}.
22960 All files prefixed with @file{^sem^SEM^} perform semantic analysis. The
22961 numbers correspond to chapters of the Ada standard. For example, all
22962 issues involving context clauses can be found in @file{sem_ch10.adb}. In
22963 addition, some features of the language require sufficient special processing
22964 to justify their own semantic files: sem_aggr for aggregates, sem_disp for
22965 dynamic dispatching, etc.
22968 All files prefixed with @file{^exp^EXP^} perform normalization and
22969 expansion of the intermediate representation (abstract syntax tree, or AST).
22970 these files use the same numbering scheme as the parser and semantics files.
22971 For example, the construction of record initialization procedures is done in
22972 @file{exp_ch3.adb}.
22975 The files prefixed with @file{^bind^BIND^} implement the binder, which
22976 verifies the consistency of the compilation, determines an order of
22977 elaboration, and generates the bind file.
22980 The files @file{atree.ads} and @file{atree.adb} detail the low-level
22981 data structures used by the front-end.
22984 The files @file{sinfo.ads} and @file{sinfo.adb} detail the structure of
22985 the abstract syntax tree as produced by the parser.
22988 The files @file{einfo.ads} and @file{einfo.adb} detail the attributes of
22989 all entities, computed during semantic analysis.
22992 Library management issues are dealt with in files with prefix
22998 Ada files with the prefix @file{^a-^A-^} are children of @code{Ada}, as
22999 defined in Annex A.
23004 Files with prefix @file{^i-^I-^} are children of @code{Interfaces}, as
23005 defined in Annex B.
23009 Files with prefix @file{^s-^S-^} are children of @code{System}. This includes
23010 both language-defined children and GNAT run-time routines.
23014 Files with prefix @file{^g-^G-^} are children of @code{GNAT}. These are useful
23015 general-purpose packages, fully documented in their specs. All
23016 the other @file{.c} files are modifications of common @command{gcc} files.
23019 @node Getting Internal Debugging Information
23020 @section Getting Internal Debugging Information
23023 Most compilers have internal debugging switches and modes. GNAT
23024 does also, except GNAT internal debugging switches and modes are not
23025 secret. A summary and full description of all the compiler and binder
23026 debug flags are in the file @file{debug.adb}. You must obtain the
23027 sources of the compiler to see the full detailed effects of these flags.
23029 The switches that print the source of the program (reconstructed from
23030 the internal tree) are of general interest for user programs, as are the
23032 the full internal tree, and the entity table (the symbol table
23033 information). The reconstructed source provides a readable version of the
23034 program after the front-end has completed analysis and expansion,
23035 and is useful when studying the performance of specific constructs.
23036 For example, constraint checks are indicated, complex aggregates
23037 are replaced with loops and assignments, and tasking primitives
23038 are replaced with run-time calls.
23040 @node Stack Traceback
23041 @section Stack Traceback
23043 @cindex stack traceback
23044 @cindex stack unwinding
23047 Traceback is a mechanism to display the sequence of subprogram calls that
23048 leads to a specified execution point in a program. Often (but not always)
23049 the execution point is an instruction at which an exception has been raised.
23050 This mechanism is also known as @i{stack unwinding} because it obtains
23051 its information by scanning the run-time stack and recovering the activation
23052 records of all active subprograms. Stack unwinding is one of the most
23053 important tools for program debugging.
23055 The first entry stored in traceback corresponds to the deepest calling level,
23056 that is to say the subprogram currently executing the instruction
23057 from which we want to obtain the traceback.
23059 Note that there is no runtime performance penalty when stack traceback
23060 is enabled, and no exception is raised during program execution.
23063 * Non-Symbolic Traceback::
23064 * Symbolic Traceback::
23067 @node Non-Symbolic Traceback
23068 @subsection Non-Symbolic Traceback
23069 @cindex traceback, non-symbolic
23072 Note: this feature is not supported on all platforms. See
23073 @file{GNAT.Traceback spec in g-traceb.ads} for a complete list of supported
23077 * Tracebacks From an Unhandled Exception::
23078 * Tracebacks From Exception Occurrences (non-symbolic)::
23079 * Tracebacks From Anywhere in a Program (non-symbolic)::
23082 @node Tracebacks From an Unhandled Exception
23083 @subsubsection Tracebacks From an Unhandled Exception
23086 A runtime non-symbolic traceback is a list of addresses of call instructions.
23087 To enable this feature you must use the @option{-E}
23088 @code{gnatbind}'s option. With this option a stack traceback is stored as part
23089 of exception information. You can retrieve this information using the
23090 @code{addr2line} tool.
23092 Here is a simple example:
23094 @smallexample @c ada
23100 raise Constraint_Error;
23115 $ gnatmake stb -bargs -E
23118 Execution terminated by unhandled exception
23119 Exception name: CONSTRAINT_ERROR
23121 Call stack traceback locations:
23122 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
23126 As we see the traceback lists a sequence of addresses for the unhandled
23127 exception @code{CONSTRAINT_ERROR} raised in procedure P1. It is easy to
23128 guess that this exception come from procedure P1. To translate these
23129 addresses into the source lines where the calls appear, the
23130 @code{addr2line} tool, described below, is invaluable. The use of this tool
23131 requires the program to be compiled with debug information.
23134 $ gnatmake -g stb -bargs -E
23137 Execution terminated by unhandled exception
23138 Exception name: CONSTRAINT_ERROR
23140 Call stack traceback locations:
23141 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
23143 $ addr2line --exe=stb 0x401373 0x40138b 0x40139c 0x401335 0x4011c4
23144 0x4011f1 0x77e892a4
23146 00401373 at d:/stb/stb.adb:5
23147 0040138B at d:/stb/stb.adb:10
23148 0040139C at d:/stb/stb.adb:14
23149 00401335 at d:/stb/b~stb.adb:104
23150 004011C4 at /build/@dots{}/crt1.c:200
23151 004011F1 at /build/@dots{}/crt1.c:222
23152 77E892A4 in ?? at ??:0
23156 The @code{addr2line} tool has several other useful options:
23160 to get the function name corresponding to any location
23162 @item --demangle=gnat
23163 to use the gnat decoding mode for the function names. Note that
23164 for binutils version 2.9.x the option is simply @option{--demangle}.
23168 $ addr2line --exe=stb --functions --demangle=gnat 0x401373 0x40138b
23169 0x40139c 0x401335 0x4011c4 0x4011f1
23171 00401373 in stb.p1 at d:/stb/stb.adb:5
23172 0040138B in stb.p2 at d:/stb/stb.adb:10
23173 0040139C in stb at d:/stb/stb.adb:14
23174 00401335 in main at d:/stb/b~stb.adb:104
23175 004011C4 in <__mingw_CRTStartup> at /build/@dots{}/crt1.c:200
23176 004011F1 in <mainCRTStartup> at /build/@dots{}/crt1.c:222
23180 From this traceback we can see that the exception was raised in
23181 @file{stb.adb} at line 5, which was reached from a procedure call in
23182 @file{stb.adb} at line 10, and so on. The @file{b~std.adb} is the binder file,
23183 which contains the call to the main program.
23184 @xref{Running gnatbind}. The remaining entries are assorted runtime routines,
23185 and the output will vary from platform to platform.
23187 It is also possible to use @code{GDB} with these traceback addresses to debug
23188 the program. For example, we can break at a given code location, as reported
23189 in the stack traceback:
23195 Furthermore, this feature is not implemented inside Windows DLL. Only
23196 the non-symbolic traceback is reported in this case.
23199 (gdb) break *0x401373
23200 Breakpoint 1 at 0x401373: file stb.adb, line 5.
23204 It is important to note that the stack traceback addresses
23205 do not change when debug information is included. This is particularly useful
23206 because it makes it possible to release software without debug information (to
23207 minimize object size), get a field report that includes a stack traceback
23208 whenever an internal bug occurs, and then be able to retrieve the sequence
23209 of calls with the same program compiled with debug information.
23211 @node Tracebacks From Exception Occurrences (non-symbolic)
23212 @subsubsection Tracebacks From Exception Occurrences
23215 Non-symbolic tracebacks are obtained by using the @option{-E} binder argument.
23216 The stack traceback is attached to the exception information string, and can
23217 be retrieved in an exception handler within the Ada program, by means of the
23218 Ada facilities defined in @code{Ada.Exceptions}. Here is a simple example:
23220 @smallexample @c ada
23222 with Ada.Exceptions;
23227 use Ada.Exceptions;
23235 Text_IO.Put_Line (Exception_Information (E));
23249 This program will output:
23254 Exception name: CONSTRAINT_ERROR
23255 Message: stb.adb:12
23256 Call stack traceback locations:
23257 0x4015e4 0x401633 0x401644 0x401461 0x4011c4 0x4011f1 0x77e892a4
23260 @node Tracebacks From Anywhere in a Program (non-symbolic)
23261 @subsubsection Tracebacks From Anywhere in a Program
23264 It is also possible to retrieve a stack traceback from anywhere in a
23265 program. For this you need to
23266 use the @code{GNAT.Traceback} API. This package includes a procedure called
23267 @code{Call_Chain} that computes a complete stack traceback, as well as useful
23268 display procedures described below. It is not necessary to use the
23269 @option{-E gnatbind} option in this case, because the stack traceback mechanism
23270 is invoked explicitly.
23273 In the following example we compute a traceback at a specific location in
23274 the program, and we display it using @code{GNAT.Debug_Utilities.Image} to
23275 convert addresses to strings:
23277 @smallexample @c ada
23279 with GNAT.Traceback;
23280 with GNAT.Debug_Utilities;
23286 use GNAT.Traceback;
23289 TB : Tracebacks_Array (1 .. 10);
23290 -- We are asking for a maximum of 10 stack frames.
23292 -- Len will receive the actual number of stack frames returned.
23294 Call_Chain (TB, Len);
23296 Text_IO.Put ("In STB.P1 : ");
23298 for K in 1 .. Len loop
23299 Text_IO.Put (Debug_Utilities.Image (TB (K)));
23320 In STB.P1 : 16#0040_F1E4# 16#0040_14F2# 16#0040_170B# 16#0040_171C#
23321 16#0040_1461# 16#0040_11C4# 16#0040_11F1# 16#77E8_92A4#
23325 You can then get further information by invoking the @code{addr2line}
23326 tool as described earlier (note that the hexadecimal addresses
23327 need to be specified in C format, with a leading ``0x'').
23329 @node Symbolic Traceback
23330 @subsection Symbolic Traceback
23331 @cindex traceback, symbolic
23334 A symbolic traceback is a stack traceback in which procedure names are
23335 associated with each code location.
23338 Note that this feature is not supported on all platforms. See
23339 @file{GNAT.Traceback.Symbolic spec in g-trasym.ads} for a complete
23340 list of currently supported platforms.
23343 Note that the symbolic traceback requires that the program be compiled
23344 with debug information. If it is not compiled with debug information
23345 only the non-symbolic information will be valid.
23348 * Tracebacks From Exception Occurrences (symbolic)::
23349 * Tracebacks From Anywhere in a Program (symbolic)::
23352 @node Tracebacks From Exception Occurrences (symbolic)
23353 @subsubsection Tracebacks From Exception Occurrences
23355 @smallexample @c ada
23357 with GNAT.Traceback.Symbolic;
23363 raise Constraint_Error;
23380 Ada.Text_IO.Put_Line (GNAT.Traceback.Symbolic.Symbolic_Traceback (E));
23385 $ gnatmake -g .\stb -bargs -E -largs -lgnat -laddr2line -lintl
23388 0040149F in stb.p1 at stb.adb:8
23389 004014B7 in stb.p2 at stb.adb:13
23390 004014CF in stb.p3 at stb.adb:18
23391 004015DD in ada.stb at stb.adb:22
23392 00401461 in main at b~stb.adb:168
23393 004011C4 in __mingw_CRTStartup at crt1.c:200
23394 004011F1 in mainCRTStartup at crt1.c:222
23395 77E892A4 in ?? at ??:0
23399 In the above example the ``.\'' syntax in the @command{gnatmake} command
23400 is currently required by @command{addr2line} for files that are in
23401 the current working directory.
23402 Moreover, the exact sequence of linker options may vary from platform
23404 The above @option{-largs} section is for Windows platforms. By contrast,
23405 under Unix there is no need for the @option{-largs} section.
23406 Differences across platforms are due to details of linker implementation.
23408 @node Tracebacks From Anywhere in a Program (symbolic)
23409 @subsubsection Tracebacks From Anywhere in a Program
23412 It is possible to get a symbolic stack traceback
23413 from anywhere in a program, just as for non-symbolic tracebacks.
23414 The first step is to obtain a non-symbolic
23415 traceback, and then call @code{Symbolic_Traceback} to compute the symbolic
23416 information. Here is an example:
23418 @smallexample @c ada
23420 with GNAT.Traceback;
23421 with GNAT.Traceback.Symbolic;
23426 use GNAT.Traceback;
23427 use GNAT.Traceback.Symbolic;
23430 TB : Tracebacks_Array (1 .. 10);
23431 -- We are asking for a maximum of 10 stack frames.
23433 -- Len will receive the actual number of stack frames returned.
23435 Call_Chain (TB, Len);
23436 Text_IO.Put_Line (Symbolic_Traceback (TB (1 .. Len)));
23449 @c ******************************
23451 @node Compatibility with HP Ada
23452 @chapter Compatibility with HP Ada
23453 @cindex Compatibility
23458 @cindex Compatibility between GNAT and HP Ada
23459 This chapter compares HP Ada (formerly known as ``DEC Ada'')
23460 for OpenVMS Alpha and GNAT for OpenVMS for Alpha and for I64.
23461 GNAT is highly compatible
23462 with HP Ada, and it should generally be straightforward to port code
23463 from the HP Ada environment to GNAT. However, there are a few language
23464 and implementation differences of which the user must be aware. These
23465 differences are discussed in this chapter. In
23466 addition, the operating environment and command structure for the
23467 compiler are different, and these differences are also discussed.
23469 For further details on these and other compatibility issues,
23470 see Appendix E of the HP publication
23471 @cite{HP Ada, Technical Overview and Comparison on HP Platforms}.
23473 Except where otherwise indicated, the description of GNAT for OpenVMS
23474 applies to both the Alpha and I64 platforms.
23476 For information on porting Ada code from GNAT on Alpha OpenVMS to GNAT on
23477 I64 OpenVMS, see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
23479 The discussion in this chapter addresses specifically the implementation
23480 of Ada 83 for HP OpenVMS Alpha Systems. In cases where the implementation
23481 of HP Ada differs between OpenVMS Alpha Systems and OpenVMS VAX Systems,
23482 GNAT always follows the Alpha implementation.
23484 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
23485 attributes are recognized, although only a subset of them can sensibly
23486 be implemented. The description of pragmas in
23487 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
23488 indicates whether or not they are applicable to non-VMS systems.
23491 * Ada Language Compatibility::
23492 * Differences in the Definition of Package System::
23493 * Language-Related Features::
23494 * The Package STANDARD::
23495 * The Package SYSTEM::
23496 * Tasking and Task-Related Features::
23497 * Pragmas and Pragma-Related Features::
23498 * Library of Predefined Units::
23500 * Main Program Definition::
23501 * Implementation-Defined Attributes::
23502 * Compiler and Run-Time Interfacing::
23503 * Program Compilation and Library Management::
23505 * Implementation Limits::
23506 * Tools and Utilities::
23509 @node Ada Language Compatibility
23510 @section Ada Language Compatibility
23513 GNAT handles Ada 95 and Ada 2005 as well as Ada 83, whereas HP Ada is only
23514 for Ada 83. Ada 95 and Ada 2005 are almost completely upwards compatible
23515 with Ada 83, and therefore Ada 83 programs will compile
23516 and run under GNAT with
23517 no changes or only minor changes. The @cite{Annotated Ada Reference Manual}
23518 provides details on specific incompatibilities.
23520 GNAT provides the switch @option{/83} on the @command{GNAT COMPILE} command,
23521 as well as the pragma @code{ADA_83}, to force the compiler to
23522 operate in Ada 83 mode. This mode does not guarantee complete
23523 conformance to Ada 83, but in practice is sufficient to
23524 eliminate most sources of incompatibilities.
23525 In particular, it eliminates the recognition of the
23526 additional Ada 95 and Ada 2005 keywords, so that their use as identifiers
23527 in Ada 83 programs is legal, and handles the cases of packages
23528 with optional bodies, and generics that instantiate unconstrained
23529 types without the use of @code{(<>)}.
23531 @node Differences in the Definition of Package System
23532 @section Differences in the Definition of Package @code{System}
23535 An Ada compiler is allowed to add
23536 implementation-dependent declarations to package @code{System}.
23538 GNAT does not take advantage of this permission, and the version of
23539 @code{System} provided by GNAT exactly matches that defined in the Ada
23542 However, HP Ada adds an extensive set of declarations to package
23544 as fully documented in the HP Ada manuals. To minimize changes required
23545 for programs that make use of these extensions, GNAT provides the pragma
23546 @code{Extend_System} for extending the definition of package System. By using:
23547 @cindex pragma @code{Extend_System}
23548 @cindex @code{Extend_System} pragma
23550 @smallexample @c ada
23553 pragma Extend_System (Aux_DEC);
23559 the set of definitions in @code{System} is extended to include those in
23560 package @code{System.Aux_DEC}.
23561 @cindex @code{System.Aux_DEC} package
23562 @cindex @code{Aux_DEC} package (child of @code{System})
23563 These definitions are incorporated directly into package @code{System},
23564 as though they had been declared there. For a
23565 list of the declarations added, see the spec of this package,
23566 which can be found in the file @file{s-auxdec.ads} in the GNAT library.
23567 @cindex @file{s-auxdec.ads} file
23568 The pragma @code{Extend_System} is a configuration pragma, which means that
23569 it can be placed in the file @file{gnat.adc}, so that it will automatically
23570 apply to all subsequent compilations. See @ref{Configuration Pragmas},
23571 for further details.
23573 An alternative approach that avoids the use of the non-standard
23574 @code{Extend_System} pragma is to add a context clause to the unit that
23575 references these facilities:
23577 @smallexample @c ada
23579 with System.Aux_DEC;
23580 use System.Aux_DEC;
23585 The effect is not quite semantically identical to incorporating
23586 the declarations directly into package @code{System},
23587 but most programs will not notice a difference
23588 unless they use prefix notation (e.g.@: @code{System.Integer_8})
23589 to reference the entities directly in package @code{System}.
23590 For units containing such references,
23591 the prefixes must either be removed, or the pragma @code{Extend_System}
23594 @node Language-Related Features
23595 @section Language-Related Features
23598 The following sections highlight differences in types,
23599 representations of types, operations, alignment, and
23603 * Integer Types and Representations::
23604 * Floating-Point Types and Representations::
23605 * Pragmas Float_Representation and Long_Float::
23606 * Fixed-Point Types and Representations::
23607 * Record and Array Component Alignment::
23608 * Address Clauses::
23609 * Other Representation Clauses::
23612 @node Integer Types and Representations
23613 @subsection Integer Types and Representations
23616 The set of predefined integer types is identical in HP Ada and GNAT.
23617 Furthermore the representation of these integer types is also identical,
23618 including the capability of size clauses forcing biased representation.
23621 HP Ada for OpenVMS Alpha systems has defined the
23622 following additional integer types in package @code{System}:
23639 @code{LARGEST_INTEGER}
23643 In GNAT, the first four of these types may be obtained from the
23644 standard Ada package @code{Interfaces}.
23645 Alternatively, by use of the pragma @code{Extend_System}, identical
23646 declarations can be referenced directly in package @code{System}.
23647 On both GNAT and HP Ada, the maximum integer size is 64 bits.
23649 @node Floating-Point Types and Representations
23650 @subsection Floating-Point Types and Representations
23651 @cindex Floating-Point types
23654 The set of predefined floating-point types is identical in HP Ada and GNAT.
23655 Furthermore the representation of these floating-point
23656 types is also identical. One important difference is that the default
23657 representation for HP Ada is @code{VAX_Float}, but the default representation
23660 Specific types may be declared to be @code{VAX_Float} or IEEE, using the
23661 pragma @code{Float_Representation} as described in the HP Ada
23663 For example, the declarations:
23665 @smallexample @c ada
23667 type F_Float is digits 6;
23668 pragma Float_Representation (VAX_Float, F_Float);
23673 declares a type @code{F_Float} that will be represented in @code{VAX_Float}
23675 This set of declarations actually appears in @code{System.Aux_DEC},
23677 the full set of additional floating-point declarations provided in
23678 the HP Ada version of package @code{System}.
23679 This and similar declarations may be accessed in a user program
23680 by using pragma @code{Extend_System}. The use of this
23681 pragma, and the related pragma @code{Long_Float} is described in further
23682 detail in the following section.
23684 @node Pragmas Float_Representation and Long_Float
23685 @subsection Pragmas @code{Float_Representation} and @code{Long_Float}
23688 HP Ada provides the pragma @code{Float_Representation}, which
23689 acts as a program library switch to allow control over
23690 the internal representation chosen for the predefined
23691 floating-point types declared in the package @code{Standard}.
23692 The format of this pragma is as follows:
23694 @smallexample @c ada
23696 pragma Float_Representation(VAX_Float | IEEE_Float);
23701 This pragma controls the representation of floating-point
23706 @code{VAX_Float} specifies that floating-point
23707 types are represented by default with the VAX system hardware types
23708 @code{F-floating}, @code{D-floating}, @code{G-floating}.
23709 Note that the @code{H-floating}
23710 type was available only on VAX systems, and is not available
23711 in either HP Ada or GNAT.
23714 @code{IEEE_Float} specifies that floating-point
23715 types are represented by default with the IEEE single and
23716 double floating-point types.
23720 GNAT provides an identical implementation of the pragma
23721 @code{Float_Representation}, except that it functions as a
23722 configuration pragma. Note that the
23723 notion of configuration pragma corresponds closely to the
23724 HP Ada notion of a program library switch.
23726 When no pragma is used in GNAT, the default is @code{IEEE_Float},
23728 from HP Ada 83, where the default is @code{VAX_Float}. In addition, the
23729 predefined libraries in GNAT are built using @code{IEEE_Float}, so it is not
23730 advisable to change the format of numbers passed to standard library
23731 routines, and if necessary explicit type conversions may be needed.
23733 The use of @code{IEEE_Float} is recommended in GNAT since it is more
23734 efficient, and (given that it conforms to an international standard)
23735 potentially more portable.
23736 The situation in which @code{VAX_Float} may be useful is in interfacing
23737 to existing code and data that expect the use of @code{VAX_Float}.
23738 In such a situation use the predefined @code{VAX_Float}
23739 types in package @code{System}, as extended by
23740 @code{Extend_System}. For example, use @code{System.F_Float}
23741 to specify the 32-bit @code{F-Float} format.
23744 On OpenVMS systems, HP Ada provides the pragma @code{Long_Float}
23745 to allow control over the internal representation chosen
23746 for the predefined type @code{Long_Float} and for floating-point
23747 type declarations with digits specified in the range 7 .. 15.
23748 The format of this pragma is as follows:
23750 @smallexample @c ada
23752 pragma Long_Float (D_FLOAT | G_FLOAT);
23756 @node Fixed-Point Types and Representations
23757 @subsection Fixed-Point Types and Representations
23760 On HP Ada for OpenVMS Alpha systems, rounding is
23761 away from zero for both positive and negative numbers.
23762 Therefore, @code{+0.5} rounds to @code{1},
23763 and @code{-0.5} rounds to @code{-1}.
23765 On GNAT the results of operations
23766 on fixed-point types are in accordance with the Ada
23767 rules. In particular, results of operations on decimal
23768 fixed-point types are truncated.
23770 @node Record and Array Component Alignment
23771 @subsection Record and Array Component Alignment
23774 On HP Ada for OpenVMS Alpha, all non-composite components
23775 are aligned on natural boundaries. For example, 1-byte
23776 components are aligned on byte boundaries, 2-byte
23777 components on 2-byte boundaries, 4-byte components on 4-byte
23778 byte boundaries, and so on. The OpenVMS Alpha hardware
23779 runs more efficiently with naturally aligned data.
23781 On GNAT, alignment rules are compatible
23782 with HP Ada for OpenVMS Alpha.
23784 @node Address Clauses
23785 @subsection Address Clauses
23788 In HP Ada and GNAT, address clauses are supported for
23789 objects and imported subprograms.
23790 The predefined type @code{System.Address} is a private type
23791 in both compilers on Alpha OpenVMS, with the same representation
23792 (it is simply a machine pointer). Addition, subtraction, and comparison
23793 operations are available in the standard Ada package
23794 @code{System.Storage_Elements}, or in package @code{System}
23795 if it is extended to include @code{System.Aux_DEC} using a
23796 pragma @code{Extend_System} as previously described.
23798 Note that code that @code{with}'s both this extended package @code{System}
23799 and the package @code{System.Storage_Elements} should not @code{use}
23800 both packages, or ambiguities will result. In general it is better
23801 not to mix these two sets of facilities. The Ada package was
23802 designed specifically to provide the kind of features that HP Ada
23803 adds directly to package @code{System}.
23805 The type @code{System.Address} is a 64-bit integer type in GNAT for
23806 I64 OpenVMS. For more information,
23807 see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
23809 GNAT is compatible with HP Ada in its handling of address
23810 clauses, except for some limitations in
23811 the form of address clauses for composite objects with
23812 initialization. Such address clauses are easily replaced
23813 by the use of an explicitly-defined constant as described
23814 in the Ada Reference Manual (13.1(22)). For example, the sequence
23817 @smallexample @c ada
23819 X, Y : Integer := Init_Func;
23820 Q : String (X .. Y) := "abc";
23822 for Q'Address use Compute_Address;
23827 will be rejected by GNAT, since the address cannot be computed at the time
23828 that @code{Q} is declared. To achieve the intended effect, write instead:
23830 @smallexample @c ada
23833 X, Y : Integer := Init_Func;
23834 Q_Address : constant Address := Compute_Address;
23835 Q : String (X .. Y) := "abc";
23837 for Q'Address use Q_Address;
23843 which will be accepted by GNAT (and other Ada compilers), and is also
23844 compatible with Ada 83. A fuller description of the restrictions
23845 on address specifications is found in @ref{Top, GNAT Reference Manual,
23846 About This Guide, gnat_rm, GNAT Reference Manual}.
23848 @node Other Representation Clauses
23849 @subsection Other Representation Clauses
23852 GNAT implements in a compatible manner all the representation
23853 clauses supported by HP Ada. In addition, GNAT
23854 implements the representation clause forms that were introduced in Ada 95,
23855 including @code{COMPONENT_SIZE} and @code{SIZE} clauses for objects.
23857 @node The Package STANDARD
23858 @section The Package @code{STANDARD}
23861 The package @code{STANDARD}, as implemented by HP Ada, is fully
23862 described in the @cite{Ada Reference Manual} and in the
23863 @cite{HP Ada Language Reference Manual}. As implemented by GNAT, the
23864 package @code{STANDARD} is described in the @cite{Ada Reference Manual}.
23866 In addition, HP Ada supports the Latin-1 character set in
23867 the type @code{CHARACTER}. GNAT supports the Latin-1 character set
23868 in the type @code{CHARACTER} and also Unicode (ISO 10646 BMP) in
23869 the type @code{WIDE_CHARACTER}.
23871 The floating-point types supported by GNAT are those
23872 supported by HP Ada, but the defaults are different, and are controlled by
23873 pragmas. See @ref{Floating-Point Types and Representations}, for details.
23875 @node The Package SYSTEM
23876 @section The Package @code{SYSTEM}
23879 HP Ada provides a specific version of the package
23880 @code{SYSTEM} for each platform on which the language is implemented.
23881 For the complete spec of the package @code{SYSTEM}, see
23882 Appendix F of the @cite{HP Ada Language Reference Manual}.
23884 On HP Ada, the package @code{SYSTEM} includes the following conversion
23887 @item @code{TO_ADDRESS(INTEGER)}
23889 @item @code{TO_ADDRESS(UNSIGNED_LONGWORD)}
23891 @item @code{TO_ADDRESS(}@i{universal_integer}@code{)}
23893 @item @code{TO_INTEGER(ADDRESS)}
23895 @item @code{TO_UNSIGNED_LONGWORD(ADDRESS)}
23897 @item Function @code{IMPORT_VALUE return UNSIGNED_LONGWORD} and the
23898 functions @code{IMPORT_ADDRESS} and @code{IMPORT_LARGEST_VALUE}
23902 By default, GNAT supplies a version of @code{SYSTEM} that matches
23903 the definition given in the @cite{Ada Reference Manual}.
23905 is a subset of the HP system definitions, which is as
23906 close as possible to the original definitions. The only difference
23907 is that the definition of @code{SYSTEM_NAME} is different:
23909 @smallexample @c ada
23911 type Name is (SYSTEM_NAME_GNAT);
23912 System_Name : constant Name := SYSTEM_NAME_GNAT;
23917 Also, GNAT adds the Ada declarations for
23918 @code{BIT_ORDER} and @code{DEFAULT_BIT_ORDER}.
23920 However, the use of the following pragma causes GNAT
23921 to extend the definition of package @code{SYSTEM} so that it
23922 encompasses the full set of HP-specific extensions,
23923 including the functions listed above:
23925 @smallexample @c ada
23927 pragma Extend_System (Aux_DEC);
23932 The pragma @code{Extend_System} is a configuration pragma that
23933 is most conveniently placed in the @file{gnat.adc} file. @xref{Pragma
23934 Extend_System,,, gnat_rm, GNAT Reference Manual}, for further details.
23936 HP Ada does not allow the recompilation of the package
23937 @code{SYSTEM}. Instead HP Ada provides several pragmas
23938 (@code{SYSTEM_NAME}, @code{STORAGE_UNIT}, and @code{MEMORY_SIZE})
23939 to modify values in the package @code{SYSTEM}.
23940 On OpenVMS Alpha systems, the pragma
23941 @code{SYSTEM_NAME} takes the enumeration literal @code{OPENVMS_AXP} as
23942 its single argument.
23944 GNAT does permit the recompilation of package @code{SYSTEM} using
23945 the special switch @option{-gnatg}, and this switch can be used if
23946 it is necessary to modify the definitions in @code{SYSTEM}. GNAT does
23947 not permit the specification of @code{SYSTEM_NAME}, @code{STORAGE_UNIT}
23948 or @code{MEMORY_SIZE} by any other means.
23950 On GNAT systems, the pragma @code{SYSTEM_NAME} takes the
23951 enumeration literal @code{SYSTEM_NAME_GNAT}.
23953 The definitions provided by the use of
23955 @smallexample @c ada
23956 pragma Extend_System (AUX_Dec);
23960 are virtually identical to those provided by the HP Ada 83 package
23961 @code{SYSTEM}. One important difference is that the name of the
23963 function for type @code{UNSIGNED_LONGWORD} is changed to
23964 @code{TO_ADDRESS_LONG}.
23965 @xref{Address Clauses,,, gnat_rm, GNAT Reference Manual}, for a
23966 discussion of why this change was necessary.
23969 The version of @code{TO_ADDRESS} taking a @i{universal_integer} argument
23971 an extension to Ada 83 not strictly compatible with the reference manual.
23972 GNAT, in order to be exactly compatible with the standard,
23973 does not provide this capability. In HP Ada 83, the
23974 point of this definition is to deal with a call like:
23976 @smallexample @c ada
23977 TO_ADDRESS (16#12777#);
23981 Normally, according to Ada 83 semantics, one would expect this to be
23982 ambiguous, since it matches both the @code{INTEGER} and
23983 @code{UNSIGNED_LONGWORD} forms of @code{TO_ADDRESS}.
23984 However, in HP Ada 83, there is no ambiguity, since the
23985 definition using @i{universal_integer} takes precedence.
23987 In GNAT, since the version with @i{universal_integer} cannot be supplied,
23989 not possible to be 100% compatible. Since there are many programs using
23990 numeric constants for the argument to @code{TO_ADDRESS}, the decision in
23992 to change the name of the function in the @code{UNSIGNED_LONGWORD} case,
23993 so the declarations provided in the GNAT version of @code{AUX_Dec} are:
23995 @smallexample @c ada
23996 function To_Address (X : Integer) return Address;
23997 pragma Pure_Function (To_Address);
23999 function To_Address_Long (X : Unsigned_Longword) return Address;
24000 pragma Pure_Function (To_Address_Long);
24004 This means that programs using @code{TO_ADDRESS} for
24005 @code{UNSIGNED_LONGWORD} must change the name to @code{TO_ADDRESS_LONG}.
24007 @node Tasking and Task-Related Features
24008 @section Tasking and Task-Related Features
24011 This section compares the treatment of tasking in GNAT
24012 and in HP Ada for OpenVMS Alpha.
24013 The GNAT description applies to both Alpha and I64 OpenVMS.
24014 For detailed information on tasking in
24015 HP Ada, see the @cite{HP Ada Language Reference Manual} and the
24016 relevant run-time reference manual.
24019 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
24020 * Assigning Task IDs::
24021 * Task IDs and Delays::
24022 * Task-Related Pragmas::
24023 * Scheduling and Task Priority::
24025 * External Interrupts::
24028 @node Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
24029 @subsection Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
24032 On OpenVMS Alpha systems, each Ada task (except a passive
24033 task) is implemented as a single stream of execution
24034 that is created and managed by the kernel. On these
24035 systems, HP Ada tasking support is based on DECthreads,
24036 an implementation of the POSIX standard for threads.
24038 Also, on OpenVMS Alpha systems, HP Ada tasks and foreign
24039 code that calls DECthreads routines can be used together.
24040 The interaction between Ada tasks and DECthreads routines
24041 can have some benefits. For example when on OpenVMS Alpha,
24042 HP Ada can call C code that is already threaded.
24044 GNAT uses the facilities of DECthreads,
24045 and Ada tasks are mapped to threads.
24047 @node Assigning Task IDs
24048 @subsection Assigning Task IDs
24051 The HP Ada Run-Time Library always assigns @code{%TASK 1} to
24052 the environment task that executes the main program. On
24053 OpenVMS Alpha systems, @code{%TASK 0} is often used for tasks
24054 that have been created but are not yet activated.
24056 On OpenVMS Alpha systems, task IDs are assigned at
24057 activation. On GNAT systems, task IDs are also assigned at
24058 task creation but do not have the same form or values as
24059 task ID values in HP Ada. There is no null task, and the
24060 environment task does not have a specific task ID value.
24062 @node Task IDs and Delays
24063 @subsection Task IDs and Delays
24066 On OpenVMS Alpha systems, tasking delays are implemented
24067 using Timer System Services. The Task ID is used for the
24068 identification of the timer request (the @code{REQIDT} parameter).
24069 If Timers are used in the application take care not to use
24070 @code{0} for the identification, because cancelling such a timer
24071 will cancel all timers and may lead to unpredictable results.
24073 @node Task-Related Pragmas
24074 @subsection Task-Related Pragmas
24077 Ada supplies the pragma @code{TASK_STORAGE}, which allows
24078 specification of the size of the guard area for a task
24079 stack. (The guard area forms an area of memory that has no
24080 read or write access and thus helps in the detection of
24081 stack overflow.) On OpenVMS Alpha systems, if the pragma
24082 @code{TASK_STORAGE} specifies a value of zero, a minimal guard
24083 area is created. In the absence of a pragma @code{TASK_STORAGE},
24084 a default guard area is created.
24086 GNAT supplies the following task-related pragmas:
24089 @item @code{TASK_INFO}
24091 This pragma appears within a task definition and
24092 applies to the task in which it appears. The argument
24093 must be of type @code{SYSTEM.TASK_INFO.TASK_INFO_TYPE}.
24095 @item @code{TASK_STORAGE}
24097 GNAT implements pragma @code{TASK_STORAGE} in the same way as HP Ada.
24098 Both HP Ada and GNAT supply the pragmas @code{PASSIVE},
24099 @code{SUPPRESS}, and @code{VOLATILE}.
24101 @node Scheduling and Task Priority
24102 @subsection Scheduling and Task Priority
24105 HP Ada implements the Ada language requirement that
24106 when two tasks are eligible for execution and they have
24107 different priorities, the lower priority task does not
24108 execute while the higher priority task is waiting. The HP
24109 Ada Run-Time Library keeps a task running until either the
24110 task is suspended or a higher priority task becomes ready.
24112 On OpenVMS Alpha systems, the default strategy is round-
24113 robin with preemption. Tasks of equal priority take turns
24114 at the processor. A task is run for a certain period of
24115 time and then placed at the tail of the ready queue for
24116 its priority level.
24118 HP Ada provides the implementation-defined pragma @code{TIME_SLICE},
24119 which can be used to enable or disable round-robin
24120 scheduling of tasks with the same priority.
24121 See the relevant HP Ada run-time reference manual for
24122 information on using the pragmas to control HP Ada task
24125 GNAT follows the scheduling rules of Annex D (Real-Time
24126 Annex) of the @cite{Ada Reference Manual}. In general, this
24127 scheduling strategy is fully compatible with HP Ada
24128 although it provides some additional constraints (as
24129 fully documented in Annex D).
24130 GNAT implements time slicing control in a manner compatible with
24131 HP Ada 83, by means of the pragma @code{Time_Slice}, whose semantics
24132 are identical to the HP Ada 83 pragma of the same name.
24133 Note that it is not possible to mix GNAT tasking and
24134 HP Ada 83 tasking in the same program, since the two run-time
24135 libraries are not compatible.
24137 @node The Task Stack
24138 @subsection The Task Stack
24141 In HP Ada, a task stack is allocated each time a
24142 non-passive task is activated. As soon as the task is
24143 terminated, the storage for the task stack is deallocated.
24144 If you specify a size of zero (bytes) with @code{T'STORAGE_SIZE},
24145 a default stack size is used. Also, regardless of the size
24146 specified, some additional space is allocated for task
24147 management purposes. On OpenVMS Alpha systems, at least
24148 one page is allocated.
24150 GNAT handles task stacks in a similar manner. In accordance with
24151 the Ada rules, it provides the pragma @code{STORAGE_SIZE} as
24152 an alternative method for controlling the task stack size.
24153 The specification of the attribute @code{T'STORAGE_SIZE} is also
24154 supported in a manner compatible with HP Ada.
24156 @node External Interrupts
24157 @subsection External Interrupts
24160 On HP Ada, external interrupts can be associated with task entries.
24161 GNAT is compatible with HP Ada in its handling of external interrupts.
24163 @node Pragmas and Pragma-Related Features
24164 @section Pragmas and Pragma-Related Features
24167 Both HP Ada and GNAT supply all language-defined pragmas
24168 as specified by the Ada 83 standard. GNAT also supplies all
24169 language-defined pragmas introduced by Ada 95 and Ada 2005.
24170 In addition, GNAT implements the implementation-defined pragmas
24174 @item @code{AST_ENTRY}
24176 @item @code{COMMON_OBJECT}
24178 @item @code{COMPONENT_ALIGNMENT}
24180 @item @code{EXPORT_EXCEPTION}
24182 @item @code{EXPORT_FUNCTION}
24184 @item @code{EXPORT_OBJECT}
24186 @item @code{EXPORT_PROCEDURE}
24188 @item @code{EXPORT_VALUED_PROCEDURE}
24190 @item @code{FLOAT_REPRESENTATION}
24194 @item @code{IMPORT_EXCEPTION}
24196 @item @code{IMPORT_FUNCTION}
24198 @item @code{IMPORT_OBJECT}
24200 @item @code{IMPORT_PROCEDURE}
24202 @item @code{IMPORT_VALUED_PROCEDURE}
24204 @item @code{INLINE_GENERIC}
24206 @item @code{INTERFACE_NAME}
24208 @item @code{LONG_FLOAT}
24210 @item @code{MAIN_STORAGE}
24212 @item @code{PASSIVE}
24214 @item @code{PSECT_OBJECT}
24216 @item @code{SHARE_GENERIC}
24218 @item @code{SUPPRESS_ALL}
24220 @item @code{TASK_STORAGE}
24222 @item @code{TIME_SLICE}
24228 These pragmas are all fully implemented, with the exception of @code{TITLE},
24229 @code{PASSIVE}, and @code{SHARE_GENERIC}, which are
24230 recognized, but which have no
24231 effect in GNAT. The effect of @code{PASSIVE} may be obtained by the
24232 use of Ada protected objects. In GNAT, all generics are inlined.
24234 Unlike HP Ada, the GNAT ``@code{EXPORT_}@i{subprogram}'' pragmas require
24235 a separate subprogram specification which must appear before the
24238 GNAT also supplies a number of implementation-defined pragmas as follows:
24240 @item @code{ABORT_DEFER}
24242 @item @code{ADA_83}
24244 @item @code{ADA_95}
24246 @item @code{ADA_05}
24248 @item @code{ANNOTATE}
24250 @item @code{ASSERT}
24252 @item @code{C_PASS_BY_COPY}
24254 @item @code{CPP_CLASS}
24256 @item @code{CPP_CONSTRUCTOR}
24258 @item @code{CPP_DESTRUCTOR}
24262 @item @code{EXTEND_SYSTEM}
24264 @item @code{LINKER_ALIAS}
24266 @item @code{LINKER_SECTION}
24268 @item @code{MACHINE_ATTRIBUTE}
24270 @item @code{NO_RETURN}
24272 @item @code{PURE_FUNCTION}
24274 @item @code{SOURCE_FILE_NAME}
24276 @item @code{SOURCE_REFERENCE}
24278 @item @code{TASK_INFO}
24280 @item @code{UNCHECKED_UNION}
24282 @item @code{UNIMPLEMENTED_UNIT}
24284 @item @code{UNIVERSAL_DATA}
24286 @item @code{UNSUPPRESS}
24288 @item @code{WARNINGS}
24290 @item @code{WEAK_EXTERNAL}
24294 For full details on these GNAT implementation-defined pragmas,
24295 see @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference
24299 * Restrictions on the Pragma INLINE::
24300 * Restrictions on the Pragma INTERFACE::
24301 * Restrictions on the Pragma SYSTEM_NAME::
24304 @node Restrictions on the Pragma INLINE
24305 @subsection Restrictions on Pragma @code{INLINE}
24308 HP Ada enforces the following restrictions on the pragma @code{INLINE}:
24310 @item Parameters cannot have a task type.
24312 @item Function results cannot be task types, unconstrained
24313 array types, or unconstrained types with discriminants.
24315 @item Bodies cannot declare the following:
24317 @item Subprogram body or stub (imported subprogram is allowed)
24321 @item Generic declarations
24323 @item Instantiations
24327 @item Access types (types derived from access types allowed)
24329 @item Array or record types
24331 @item Dependent tasks
24333 @item Direct recursive calls of subprogram or containing
24334 subprogram, directly or via a renaming
24340 In GNAT, the only restriction on pragma @code{INLINE} is that the
24341 body must occur before the call if both are in the same
24342 unit, and the size must be appropriately small. There are
24343 no other specific restrictions which cause subprograms to
24344 be incapable of being inlined.
24346 @node Restrictions on the Pragma INTERFACE
24347 @subsection Restrictions on Pragma @code{INTERFACE}
24350 The following restrictions on pragma @code{INTERFACE}
24351 are enforced by both HP Ada and GNAT:
24353 @item Languages accepted: Ada, Bliss, C, Fortran, Default.
24354 Default is the default on OpenVMS Alpha systems.
24356 @item Parameter passing: Language specifies default
24357 mechanisms but can be overridden with an @code{EXPORT} pragma.
24360 @item Ada: Use internal Ada rules.
24362 @item Bliss, C: Parameters must be mode @code{in}; cannot be
24363 record or task type. Result cannot be a string, an
24364 array, or a record.
24366 @item Fortran: Parameters cannot have a task type. Result cannot
24367 be a string, an array, or a record.
24372 GNAT is entirely upwards compatible with HP Ada, and in addition allows
24373 record parameters for all languages.
24375 @node Restrictions on the Pragma SYSTEM_NAME
24376 @subsection Restrictions on Pragma @code{SYSTEM_NAME}
24379 For HP Ada for OpenVMS Alpha, the enumeration literal
24380 for the type @code{NAME} is @code{OPENVMS_AXP}.
24381 In GNAT, the enumeration
24382 literal for the type @code{NAME} is @code{SYSTEM_NAME_GNAT}.
24384 @node Library of Predefined Units
24385 @section Library of Predefined Units
24388 A library of predefined units is provided as part of the
24389 HP Ada and GNAT implementations. HP Ada does not provide
24390 the package @code{MACHINE_CODE} but instead recommends importing
24393 The GNAT versions of the HP Ada Run-Time Library (@code{ADA$PREDEFINED:})
24394 units are taken from the OpenVMS Alpha version, not the OpenVMS VAX
24396 The HP Ada Predefined Library units are modified to remove post-Ada 83
24397 incompatibilities and to make them interoperable with GNAT
24398 (@pxref{Changes to DECLIB}, for details).
24399 The units are located in the @file{DECLIB} directory.
24401 The GNAT RTL is contained in
24402 the @file{ADALIB} directory, and
24403 the default search path is set up to find @code{DECLIB} units in preference
24404 to @code{ADALIB} units with the same name (@code{TEXT_IO},
24405 @code{SEQUENTIAL_IO}, and @code{DIRECT_IO}, for example).
24408 * Changes to DECLIB::
24411 @node Changes to DECLIB
24412 @subsection Changes to @code{DECLIB}
24415 The changes made to the HP Ada predefined library for GNAT and post-Ada 83
24416 compatibility are minor and include the following:
24419 @item Adjusting the location of pragmas and record representation
24420 clauses to obey Ada 95 (and thus Ada 2005) rules
24422 @item Adding the proper notation to generic formal parameters
24423 that take unconstrained types in instantiation
24425 @item Adding pragma @code{ELABORATE_BODY} to package specs
24426 that have package bodies not otherwise allowed
24428 @item Replacing occurrences of the identifier ``@code{PROTECTED}'' by
24429 ``@code{PROTECTD}''.
24430 Currently these are found only in the @code{STARLET} package spec.
24432 @item Changing @code{SYSTEM.ADDRESS} to @code{SYSTEM.SHORT_ADDRESS}
24433 where the address size is constrained to 32 bits.
24437 None of the above changes is visible to users.
24443 On OpenVMS Alpha, HP Ada provides the following strongly-typed bindings:
24446 @item Command Language Interpreter (CLI interface)
24448 @item DECtalk Run-Time Library (DTK interface)
24450 @item Librarian utility routines (LBR interface)
24452 @item General Purpose Run-Time Library (LIB interface)
24454 @item Math Run-Time Library (MTH interface)
24456 @item National Character Set Run-Time Library (NCS interface)
24458 @item Compiled Code Support Run-Time Library (OTS interface)
24460 @item Parallel Processing Run-Time Library (PPL interface)
24462 @item Screen Management Run-Time Library (SMG interface)
24464 @item Sort Run-Time Library (SOR interface)
24466 @item String Run-Time Library (STR interface)
24468 @item STARLET System Library
24471 @item X Window System Version 11R4 and 11R5 (X, XLIB interface)
24473 @item X Windows Toolkit (XT interface)
24475 @item X/Motif Version 1.1.3 and 1.2 (XM interface)
24479 GNAT provides implementations of these HP bindings in the @code{DECLIB}
24480 directory, on both the Alpha and I64 OpenVMS platforms.
24482 The X/Motif bindings used to build @code{DECLIB} are whatever versions are
24484 HP Ada @file{ADA$PREDEFINED} directory with extension @file{.ADC}.
24485 A pragma @code{Linker_Options} has been added to packages @code{Xm},
24486 @code{Xt}, and @code{X_Lib}
24487 causing the default X/Motif sharable image libraries to be linked in. This
24488 is done via options files named @file{xm.opt}, @file{xt.opt}, and
24489 @file{x_lib.opt} (also located in the @file{DECLIB} directory).
24491 It may be necessary to edit these options files to update or correct the
24492 library names if, for example, the newer X/Motif bindings from
24493 @file{ADA$EXAMPLES}
24494 had been (previous to installing GNAT) copied and renamed to supersede the
24495 default @file{ADA$PREDEFINED} versions.
24498 * Shared Libraries and Options Files::
24499 * Interfaces to C::
24502 @node Shared Libraries and Options Files
24503 @subsection Shared Libraries and Options Files
24506 When using the HP Ada
24507 predefined X and Motif bindings, the linking with their sharable images is
24508 done automatically by @command{GNAT LINK}.
24509 When using other X and Motif bindings, you need
24510 to add the corresponding sharable images to the command line for
24511 @code{GNAT LINK}. When linking with shared libraries, or with
24512 @file{.OPT} files, you must
24513 also add them to the command line for @command{GNAT LINK}.
24515 A shared library to be used with GNAT is built in the same way as other
24516 libraries under VMS. The VMS Link command can be used in standard fashion.
24518 @node Interfaces to C
24519 @subsection Interfaces to C
24523 provides the following Ada types and operations:
24526 @item C types package (@code{C_TYPES})
24528 @item C strings (@code{C_TYPES.NULL_TERMINATED})
24530 @item Other_types (@code{SHORT_INT})
24534 Interfacing to C with GNAT, you can use the above approach
24535 described for HP Ada or the facilities of Annex B of
24536 the @cite{Ada Reference Manual} (packages @code{INTERFACES.C},
24537 @code{INTERFACES.C.STRINGS} and @code{INTERFACES.C.POINTERS}). For more
24538 information, see @ref{Interfacing to C,,, gnat_rm, GNAT Reference Manual}.
24540 The @option{-gnatF} qualifier forces default and explicit
24541 @code{External_Name} parameters in pragmas @code{Import} and @code{Export}
24542 to be uppercased for compatibility with the default behavior
24543 of HP C. The qualifier has no effect on @code{Link_Name} parameters.
24545 @node Main Program Definition
24546 @section Main Program Definition
24549 The following section discusses differences in the
24550 definition of main programs on HP Ada and GNAT.
24551 On HP Ada, main programs are defined to meet the
24552 following conditions:
24554 @item Procedure with no formal parameters (returns @code{0} upon
24557 @item Procedure with no formal parameters (returns @code{42} when
24558 an unhandled exception is raised)
24560 @item Function with no formal parameters whose returned value
24561 is of a discrete type
24563 @item Procedure with one @code{out} formal of a discrete type for
24564 which a specification of pragma @code{EXPORT_VALUED_PROCEDURE} is given.
24569 When declared with the pragma @code{EXPORT_VALUED_PROCEDURE},
24570 a main function or main procedure returns a discrete
24571 value whose size is less than 64 bits (32 on VAX systems),
24572 the value is zero- or sign-extended as appropriate.
24573 On GNAT, main programs are defined as follows:
24575 @item Must be a non-generic, parameterless subprogram that
24576 is either a procedure or function returning an Ada
24577 @code{STANDARD.INTEGER} (the predefined type)
24579 @item Cannot be a generic subprogram or an instantiation of a
24583 @node Implementation-Defined Attributes
24584 @section Implementation-Defined Attributes
24587 GNAT provides all HP Ada implementation-defined
24590 @node Compiler and Run-Time Interfacing
24591 @section Compiler and Run-Time Interfacing
24594 HP Ada provides the following qualifiers to pass options to the linker
24597 @item @option{/WAIT} and @option{/SUBMIT}
24599 @item @option{/COMMAND}
24601 @item @option{/@r{[}NO@r{]}MAP}
24603 @item @option{/OUTPUT=@var{file-spec}}
24605 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
24609 To pass options to the linker, GNAT provides the following
24613 @item @option{/EXECUTABLE=@var{exec-name}}
24615 @item @option{/VERBOSE}
24617 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
24621 For more information on these switches, see
24622 @ref{Switches for gnatlink}.
24623 In HP Ada, the command-line switch @option{/OPTIMIZE} is available
24624 to control optimization. HP Ada also supplies the
24627 @item @code{OPTIMIZE}
24629 @item @code{INLINE}
24631 @item @code{INLINE_GENERIC}
24633 @item @code{SUPPRESS_ALL}
24635 @item @code{PASSIVE}
24639 In GNAT, optimization is controlled strictly by command
24640 line parameters, as described in the corresponding section of this guide.
24641 The HP pragmas for control of optimization are
24642 recognized but ignored.
24644 Note that in GNAT, the default is optimization off, whereas in HP Ada
24645 the default is that optimization is turned on.
24647 @node Program Compilation and Library Management
24648 @section Program Compilation and Library Management
24651 HP Ada and GNAT provide a comparable set of commands to
24652 build programs. HP Ada also provides a program library,
24653 which is a concept that does not exist on GNAT. Instead,
24654 GNAT provides directories of sources that are compiled as
24657 The following table summarizes
24658 the HP Ada commands and provides
24659 equivalent GNAT commands. In this table, some GNAT
24660 equivalents reflect the fact that GNAT does not use the
24661 concept of a program library. Instead, it uses a model
24662 in which collections of source and object files are used
24663 in a manner consistent with other languages like C and
24664 Fortran. Therefore, standard system file commands are used
24665 to manipulate these elements. Those GNAT commands are marked with
24667 Note that, unlike HP Ada, none of the GNAT commands accepts wild cards.
24670 @multitable @columnfractions .35 .65
24672 @item @emph{HP Ada Command}
24673 @tab @emph{GNAT Equivalent / Description}
24675 @item @command{ADA}
24676 @tab @command{GNAT COMPILE}@*
24677 Invokes the compiler to compile one or more Ada source files.
24679 @item @command{ACS ATTACH}@*
24680 @tab [No equivalent]@*
24681 Switches control of terminal from current process running the program
24684 @item @command{ACS CHECK}
24685 @tab @command{GNAT MAKE /DEPENDENCY_LIST}@*
24686 Forms the execution closure of one
24687 or more compiled units and checks completeness and currency.
24689 @item @command{ACS COMPILE}
24690 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
24691 Forms the execution closure of one or
24692 more specified units, checks completeness and currency,
24693 identifies units that have revised source files, compiles same,
24694 and recompiles units that are or will become obsolete.
24695 Also completes incomplete generic instantiations.
24697 @item @command{ACS COPY FOREIGN}
24699 Copies a foreign object file into the program library as a
24702 @item @command{ACS COPY UNIT}
24704 Copies a compiled unit from one program library to another.
24706 @item @command{ACS CREATE LIBRARY}
24707 @tab Create /directory (*)@*
24708 Creates a program library.
24710 @item @command{ACS CREATE SUBLIBRARY}
24711 @tab Create /directory (*)@*
24712 Creates a program sublibrary.
24714 @item @command{ACS DELETE LIBRARY}
24716 Deletes a program library and its contents.
24718 @item @command{ACS DELETE SUBLIBRARY}
24720 Deletes a program sublibrary and its contents.
24722 @item @command{ACS DELETE UNIT}
24723 @tab Delete file (*)@*
24724 On OpenVMS systems, deletes one or more compiled units from
24725 the current program library.
24727 @item @command{ACS DIRECTORY}
24728 @tab Directory (*)@*
24729 On OpenVMS systems, lists units contained in the current
24732 @item @command{ACS ENTER FOREIGN}
24734 Allows the import of a foreign body as an Ada library
24735 spec and enters a reference to a pointer.
24737 @item @command{ACS ENTER UNIT}
24739 Enters a reference (pointer) from the current program library to
24740 a unit compiled into another program library.
24742 @item @command{ACS EXIT}
24743 @tab [No equivalent]@*
24744 Exits from the program library manager.
24746 @item @command{ACS EXPORT}
24748 Creates an object file that contains system-specific object code
24749 for one or more units. With GNAT, object files can simply be copied
24750 into the desired directory.
24752 @item @command{ACS EXTRACT SOURCE}
24754 Allows access to the copied source file for each Ada compilation unit
24756 @item @command{ACS HELP}
24757 @tab @command{HELP GNAT}@*
24758 Provides online help.
24760 @item @command{ACS LINK}
24761 @tab @command{GNAT LINK}@*
24762 Links an object file containing Ada units into an executable file.
24764 @item @command{ACS LOAD}
24766 Loads (partially compiles) Ada units into the program library.
24767 Allows loading a program from a collection of files into a library
24768 without knowing the relationship among units.
24770 @item @command{ACS MERGE}
24772 Merges into the current program library, one or more units from
24773 another library where they were modified.
24775 @item @command{ACS RECOMPILE}
24776 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
24777 Recompiles from external or copied source files any obsolete
24778 unit in the closure. Also, completes any incomplete generic
24781 @item @command{ACS REENTER}
24782 @tab @command{GNAT MAKE}@*
24783 Reenters current references to units compiled after last entered
24784 with the @command{ACS ENTER UNIT} command.
24786 @item @command{ACS SET LIBRARY}
24787 @tab Set default (*)@*
24788 Defines a program library to be the compilation context as well
24789 as the target library for compiler output and commands in general.
24791 @item @command{ACS SET PRAGMA}
24792 @tab Edit @file{gnat.adc} (*)@*
24793 Redefines specified values of the library characteristics
24794 @code{LONG_ FLOAT}, @code{MEMORY_SIZE}, @code{SYSTEM_NAME},
24795 and @code{Float_Representation}.
24797 @item @command{ACS SET SOURCE}
24798 @tab Define @code{ADA_INCLUDE_PATH} path (*)@*
24799 Defines the source file search list for the @command{ACS COMPILE} command.
24801 @item @command{ACS SHOW LIBRARY}
24802 @tab Directory (*)@*
24803 Lists information about one or more program libraries.
24805 @item @command{ACS SHOW PROGRAM}
24806 @tab [No equivalent]@*
24807 Lists information about the execution closure of one or
24808 more units in the program library.
24810 @item @command{ACS SHOW SOURCE}
24811 @tab Show logical @code{ADA_INCLUDE_PATH}@*
24812 Shows the source file search used when compiling units.
24814 @item @command{ACS SHOW VERSION}
24815 @tab Compile with @option{VERBOSE} option
24816 Displays the version number of the compiler and program library
24819 @item @command{ACS SPAWN}
24820 @tab [No equivalent]@*
24821 Creates a subprocess of the current process (same as @command{DCL SPAWN}
24824 @item @command{ACS VERIFY}
24825 @tab [No equivalent]@*
24826 Performs a series of consistency checks on a program library to
24827 determine whether the library structure and library files are in
24834 @section Input-Output
24837 On OpenVMS Alpha systems, HP Ada uses OpenVMS Record
24838 Management Services (RMS) to perform operations on
24842 HP Ada and GNAT predefine an identical set of input-
24843 output packages. To make the use of the
24844 generic @code{TEXT_IO} operations more convenient, HP Ada
24845 provides predefined library packages that instantiate the
24846 integer and floating-point operations for the predefined
24847 integer and floating-point types as shown in the following table.
24849 @multitable @columnfractions .45 .55
24850 @item @emph{Package Name} @tab Instantiation
24852 @item @code{INTEGER_TEXT_IO}
24853 @tab @code{INTEGER_IO(INTEGER)}
24855 @item @code{SHORT_INTEGER_TEXT_IO}
24856 @tab @code{INTEGER_IO(SHORT_INTEGER)}
24858 @item @code{SHORT_SHORT_INTEGER_TEXT_IO}
24859 @tab @code{INTEGER_IO(SHORT_SHORT_INTEGER)}
24861 @item @code{FLOAT_TEXT_IO}
24862 @tab @code{FLOAT_IO(FLOAT)}
24864 @item @code{LONG_FLOAT_TEXT_IO}
24865 @tab @code{FLOAT_IO(LONG_FLOAT)}
24869 The HP Ada predefined packages and their operations
24870 are implemented using OpenVMS Alpha files and input-output
24871 facilities. HP Ada supports asynchronous input-output on OpenVMS Alpha.
24872 Familiarity with the following is recommended:
24874 @item RMS file organizations and access methods
24876 @item OpenVMS file specifications and directories
24878 @item OpenVMS File Definition Language (FDL)
24882 GNAT provides I/O facilities that are completely
24883 compatible with HP Ada. The distribution includes the
24884 standard HP Ada versions of all I/O packages, operating
24885 in a manner compatible with HP Ada. In particular, the
24886 following packages are by default the HP Ada (Ada 83)
24887 versions of these packages rather than the renamings
24888 suggested in Annex J of the Ada Reference Manual:
24890 @item @code{TEXT_IO}
24892 @item @code{SEQUENTIAL_IO}
24894 @item @code{DIRECT_IO}
24898 The use of the standard child package syntax (for
24899 example, @code{ADA.TEXT_IO}) retrieves the post-Ada 83 versions of these
24901 GNAT provides HP-compatible predefined instantiations
24902 of the @code{TEXT_IO} packages, and also
24903 provides the standard predefined instantiations required
24904 by the @cite{Ada Reference Manual}.
24906 For further information on how GNAT interfaces to the file
24907 system or how I/O is implemented in programs written in
24908 mixed languages, see @ref{Implementation of the Standard I/O,,,
24909 gnat_rm, GNAT Reference Manual}.
24910 This chapter covers the following:
24912 @item Standard I/O packages
24914 @item @code{FORM} strings
24916 @item @code{ADA.DIRECT_IO}
24918 @item @code{ADA.SEQUENTIAL_IO}
24920 @item @code{ADA.TEXT_IO}
24922 @item Stream pointer positioning
24924 @item Reading and writing non-regular files
24926 @item @code{GET_IMMEDIATE}
24928 @item Treating @code{TEXT_IO} files as streams
24935 @node Implementation Limits
24936 @section Implementation Limits
24939 The following table lists implementation limits for HP Ada
24941 @multitable @columnfractions .60 .20 .20
24943 @item @emph{Compilation Parameter}
24948 @item In a subprogram or entry declaration, maximum number of
24949 formal parameters that are of an unconstrained record type
24954 @item Maximum identifier length (number of characters)
24959 @item Maximum number of characters in a source line
24964 @item Maximum collection size (number of bytes)
24969 @item Maximum number of discriminants for a record type
24974 @item Maximum number of formal parameters in an entry or
24975 subprogram declaration
24980 @item Maximum number of dimensions in an array type
24985 @item Maximum number of library units and subunits in a compilation.
24990 @item Maximum number of library units and subunits in an execution.
24995 @item Maximum number of objects declared with the pragma @code{COMMON_OBJECT}
24996 or @code{PSECT_OBJECT}
25001 @item Maximum number of enumeration literals in an enumeration type
25007 @item Maximum number of lines in a source file
25012 @item Maximum number of bits in any object
25017 @item Maximum size of the static portion of a stack frame (approximate)
25022 @node Tools and Utilities
25023 @section Tools and Utilities
25026 The following table lists some of the OpenVMS development tools
25027 available for HP Ada, and the corresponding tools for
25028 use with @value{EDITION} on Alpha and I64 platforms.
25029 Aside from the debugger, all the OpenVMS tools identified are part
25030 of the DECset package.
25033 @c Specify table in TeX since Texinfo does a poor job
25037 \settabs\+Language-Sensitive Editor\quad
25038 &Product with HP Ada\quad
25041 &\it Product with HP Ada
25042 & \it Product with GNAT Pro\cr
25044 \+Code Management System
25048 \+Language-Sensitive Editor
25050 & emacs or HP LSE (Alpha)\cr
25060 & OpenVMS Debug (I64)\cr
25062 \+Source Code Analyzer /
25079 \+Coverage Analyzer
25083 \+Module Management
25085 & Not applicable\cr
25095 @c This is the Texinfo version of the table. It renders poorly in pdf, hence
25096 @c the TeX version above for the printed version
25098 @c @multitable @columnfractions .3 .4 .4
25099 @multitable {Source Code Analyzer /}{Tool with HP Ada}{Tool with GNAT Pro}
25101 @tab @i{Tool with HP Ada}
25102 @tab @i{Tool with @value{EDITION}}
25103 @item Code Management@*System
25106 @item Language-Sensitive@*Editor
25108 @tab emacs or HP LSE (Alpha)
25117 @tab OpenVMS Debug (I64)
25118 @item Source Code Analyzer /@*Cross Referencer
25122 @tab HP Digital Test@*Manager (DTM)
25124 @item Performance and@*Coverage Analyzer
25127 @item Module Management@*System
25129 @tab Not applicable
25136 @c **************************************
25137 @node Platform-Specific Information for the Run-Time Libraries
25138 @appendix Platform-Specific Information for the Run-Time Libraries
25139 @cindex Tasking and threads libraries
25140 @cindex Threads libraries and tasking
25141 @cindex Run-time libraries (platform-specific information)
25144 The GNAT run-time implementation may vary with respect to both the
25145 underlying threads library and the exception handling scheme.
25146 For threads support, one or more of the following are supplied:
25148 @item @b{native threads library}, a binding to the thread package from
25149 the underlying operating system
25151 @item @b{pthreads library} (Sparc Solaris only), a binding to the Solaris
25152 POSIX thread package
25156 For exception handling, either or both of two models are supplied:
25158 @item @b{Zero-Cost Exceptions} (``ZCX''),@footnote{
25159 Most programs should experience a substantial speed improvement by
25160 being compiled with a ZCX run-time.
25161 This is especially true for
25162 tasking applications or applications with many exception handlers.}
25163 @cindex Zero-Cost Exceptions
25164 @cindex ZCX (Zero-Cost Exceptions)
25165 which uses binder-generated tables that
25166 are interrogated at run time to locate a handler
25168 @item @b{setjmp / longjmp} (``SJLJ''),
25169 @cindex setjmp/longjmp Exception Model
25170 @cindex SJLJ (setjmp/longjmp Exception Model)
25171 which uses dynamically-set data to establish
25172 the set of handlers
25176 This appendix summarizes which combinations of threads and exception support
25177 are supplied on various GNAT platforms.
25178 It then shows how to select a particular library either
25179 permanently or temporarily,
25180 explains the properties of (and tradeoffs among) the various threads
25181 libraries, and provides some additional
25182 information about several specific platforms.
25185 * Summary of Run-Time Configurations::
25186 * Specifying a Run-Time Library::
25187 * Choosing the Scheduling Policy::
25188 * Solaris-Specific Considerations::
25189 * Linux-Specific Considerations::
25190 * AIX-Specific Considerations::
25191 * Irix-Specific Considerations::
25192 * RTX-Specific Considerations::
25193 * HP-UX-Specific Considerations::
25196 @node Summary of Run-Time Configurations
25197 @section Summary of Run-Time Configurations
25199 @multitable @columnfractions .30 .70
25200 @item @b{alpha-openvms}
25201 @item @code{@ @ }@i{rts-native (default)}
25202 @item @code{@ @ @ @ }Tasking @tab native VMS threads
25203 @item @code{@ @ @ @ }Exceptions @tab ZCX
25205 @item @b{alpha-tru64}
25206 @item @code{@ @ }@i{rts-native (default)}
25207 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
25208 @item @code{@ @ @ @ }Exceptions @tab ZCX
25210 @item @code{@ @ }@i{rts-sjlj}
25211 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
25212 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25214 @item @b{ia64-hp_linux}
25215 @item @code{@ @ }@i{rts-native (default)}
25216 @item @code{@ @ @ @ }Tasking @tab pthread library
25217 @item @code{@ @ @ @ }Exceptions @tab ZCX
25219 @item @b{ia64-hpux}
25220 @item @code{@ @ }@i{rts-native (default)}
25221 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
25222 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25224 @item @b{ia64-openvms}
25225 @item @code{@ @ }@i{rts-native (default)}
25226 @item @code{@ @ @ @ }Tasking @tab native VMS threads
25227 @item @code{@ @ @ @ }Exceptions @tab ZCX
25229 @item @b{ia64-sgi_linux}
25230 @item @code{@ @ }@i{rts-native (default)}
25231 @item @code{@ @ @ @ }Tasking @tab pthread library
25232 @item @code{@ @ @ @ }Exceptions @tab ZCX
25234 @item @b{mips-irix}
25235 @item @code{@ @ }@i{rts-native (default)}
25236 @item @code{@ @ @ @ }Tasking @tab native IRIX threads
25237 @item @code{@ @ @ @ }Exceptions @tab ZCX
25240 @item @code{@ @ }@i{rts-native (default)}
25241 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
25242 @item @code{@ @ @ @ }Exceptions @tab ZCX
25244 @item @code{@ @ }@i{rts-sjlj}
25245 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
25246 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25249 @item @code{@ @ }@i{rts-native (default)}
25250 @item @code{@ @ @ @ }Tasking @tab native AIX threads
25251 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25253 @item @b{ppc-darwin}
25254 @item @code{@ @ }@i{rts-native (default)}
25255 @item @code{@ @ @ @ }Tasking @tab native MacOS threads
25256 @item @code{@ @ @ @ }Exceptions @tab ZCX
25258 @item @b{sparc-solaris} @tab
25259 @item @code{@ @ }@i{rts-native (default)}
25260 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
25261 @item @code{@ @ @ @ }Exceptions @tab ZCX
25263 @item @code{@ @ }@i{rts-pthread}
25264 @item @code{@ @ @ @ }Tasking @tab pthread library
25265 @item @code{@ @ @ @ }Exceptions @tab ZCX
25267 @item @code{@ @ }@i{rts-sjlj}
25268 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
25269 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25271 @item @b{sparc64-solaris} @tab
25272 @item @code{@ @ }@i{rts-native (default)}
25273 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
25274 @item @code{@ @ @ @ }Exceptions @tab ZCX
25276 @item @b{x86-linux}
25277 @item @code{@ @ }@i{rts-native (default)}
25278 @item @code{@ @ @ @ }Tasking @tab pthread library
25279 @item @code{@ @ @ @ }Exceptions @tab ZCX
25281 @item @code{@ @ }@i{rts-sjlj}
25282 @item @code{@ @ @ @ }Tasking @tab pthread library
25283 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25286 @item @code{@ @ }@i{rts-native (default)}
25287 @item @code{@ @ @ @ }Tasking @tab native LynxOS threads
25288 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25290 @item @b{x86-solaris}
25291 @item @code{@ @ }@i{rts-native (default)}
25292 @item @code{@ @ @ @ }Tasking @tab native Solaris threads
25293 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25295 @item @b{x86-windows}
25296 @item @code{@ @ }@i{rts-native (default)}
25297 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
25298 @item @code{@ @ @ @ }Exceptions @tab ZCX
25300 @item @code{@ @ }@i{rts-sjlj (default)}
25301 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
25302 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25304 @item @b{x86-windows-rtx}
25305 @item @code{@ @ }@i{rts-rtx-rtss (default)}
25306 @item @code{@ @ @ @ }Tasking @tab RTX real-time subsystem RTSS threads (kernel mode)
25307 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25309 @item @code{@ @ }@i{rts-rtx-w32}
25310 @item @code{@ @ @ @ }Tasking @tab RTX Win32 threads (user mode)
25311 @item @code{@ @ @ @ }Exceptions @tab ZCX
25313 @item @b{x86_64-linux}
25314 @item @code{@ @ }@i{rts-native (default)}
25315 @item @code{@ @ @ @ }Tasking @tab pthread library
25316 @item @code{@ @ @ @ }Exceptions @tab ZCX
25318 @item @code{@ @ }@i{rts-sjlj}
25319 @item @code{@ @ @ @ }Tasking @tab pthread library
25320 @item @code{@ @ @ @ }Exceptions @tab SJLJ
25324 @node Specifying a Run-Time Library
25325 @section Specifying a Run-Time Library
25328 The @file{adainclude} subdirectory containing the sources of the GNAT
25329 run-time library, and the @file{adalib} subdirectory containing the
25330 @file{ALI} files and the static and/or shared GNAT library, are located
25331 in the gcc target-dependent area:
25334 target=$prefix/lib/gcc/gcc-@i{dumpmachine}/gcc-@i{dumpversion}/
25338 As indicated above, on some platforms several run-time libraries are supplied.
25339 These libraries are installed in the target dependent area and
25340 contain a complete source and binary subdirectory. The detailed description
25341 below explains the differences between the different libraries in terms of
25342 their thread support.
25344 The default run-time library (when GNAT is installed) is @emph{rts-native}.
25345 This default run time is selected by the means of soft links.
25346 For example on x86-linux:
25352 +--- adainclude----------+
25354 +--- adalib-----------+ |
25356 +--- rts-native | |
25358 | +--- adainclude <---+
25360 | +--- adalib <----+
25371 If the @i{rts-sjlj} library is to be selected on a permanent basis,
25372 these soft links can be modified with the following commands:
25376 $ rm -f adainclude adalib
25377 $ ln -s rts-sjlj/adainclude adainclude
25378 $ ln -s rts-sjlj/adalib adalib
25382 Alternatively, you can specify @file{rts-sjlj/adainclude} in the file
25383 @file{$target/ada_source_path} and @file{rts-sjlj/adalib} in
25384 @file{$target/ada_object_path}.
25386 Selecting another run-time library temporarily can be
25387 achieved by using the @option{--RTS} switch, e.g., @option{--RTS=sjlj}
25388 @cindex @option{--RTS} option
25390 @node Choosing the Scheduling Policy
25391 @section Choosing the Scheduling Policy
25394 When using a POSIX threads implementation, you have a choice of several
25395 scheduling policies: @code{SCHED_FIFO},
25396 @cindex @code{SCHED_FIFO} scheduling policy
25398 @cindex @code{SCHED_RR} scheduling policy
25399 and @code{SCHED_OTHER}.
25400 @cindex @code{SCHED_OTHER} scheduling policy
25401 Typically, the default is @code{SCHED_OTHER}, while using @code{SCHED_FIFO}
25402 or @code{SCHED_RR} requires special (e.g., root) privileges.
25404 By default, GNAT uses the @code{SCHED_OTHER} policy. To specify
25406 @cindex @code{SCHED_FIFO} scheduling policy
25407 you can use one of the following:
25411 @code{pragma Time_Slice (0.0)}
25412 @cindex pragma Time_Slice
25414 the corresponding binder option @option{-T0}
25415 @cindex @option{-T0} option
25417 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
25418 @cindex pragma Task_Dispatching_Policy
25422 To specify @code{SCHED_RR},
25423 @cindex @code{SCHED_RR} scheduling policy
25424 you should use @code{pragma Time_Slice} with a
25425 value greater than @code{0.0}, or else use the corresponding @option{-T}
25428 @node Solaris-Specific Considerations
25429 @section Solaris-Specific Considerations
25430 @cindex Solaris Sparc threads libraries
25433 This section addresses some topics related to the various threads libraries
25437 * Solaris Threads Issues::
25440 @node Solaris Threads Issues
25441 @subsection Solaris Threads Issues
25444 GNAT under Solaris/Sparc 32 bits comes with an alternate tasking run-time
25445 library based on POSIX threads --- @emph{rts-pthread}.
25446 @cindex rts-pthread threads library
25447 This run-time library has the advantage of being mostly shared across all
25448 POSIX-compliant thread implementations, and it also provides under
25449 @w{Solaris 8} the @code{PTHREAD_PRIO_INHERIT}
25450 @cindex @code{PTHREAD_PRIO_INHERIT} policy (under rts-pthread)
25451 and @code{PTHREAD_PRIO_PROTECT}
25452 @cindex @code{PTHREAD_PRIO_PROTECT} policy (under rts-pthread)
25453 semantics that can be selected using the predefined pragma
25454 @code{Locking_Policy}
25455 @cindex pragma Locking_Policy (under rts-pthread)
25457 @code{Inheritance_Locking} and @code{Ceiling_Locking} as the policy.
25458 @cindex @code{Inheritance_Locking} (under rts-pthread)
25459 @cindex @code{Ceiling_Locking} (under rts-pthread)
25461 As explained above, the native run-time library is based on the Solaris thread
25462 library (@code{libthread}) and is the default library.
25464 When the Solaris threads library is used (this is the default), programs
25465 compiled with GNAT can automatically take advantage of
25466 and can thus execute on multiple processors.
25467 The user can alternatively specify a processor on which the program should run
25468 to emulate a single-processor system. The multiprocessor / uniprocessor choice
25470 setting the environment variable @env{GNAT_PROCESSOR}
25471 @cindex @env{GNAT_PROCESSOR} environment variable (on Sparc Solaris)
25472 to one of the following:
25476 Use the default configuration (run the program on all
25477 available processors) - this is the same as having @code{GNAT_PROCESSOR}
25481 Let the run-time implementation choose one processor and run the program on
25484 @item 0 .. Last_Proc
25485 Run the program on the specified processor.
25486 @code{Last_Proc} is equal to @code{_SC_NPROCESSORS_CONF - 1}
25487 (where @code{_SC_NPROCESSORS_CONF} is a system variable).
25490 @node Linux-Specific Considerations
25491 @section Linux-Specific Considerations
25492 @cindex Linux threads libraries
25495 On GNU/Linux without NPTL support (usually system with GNU C Library
25496 older than 2.3), the signal model is not POSIX compliant, which means
25497 that to send a signal to the process, you need to send the signal to all
25498 threads, e.g.@: by using @code{killpg()}.
25500 @node AIX-Specific Considerations
25501 @section AIX-Specific Considerations
25502 @cindex AIX resolver library
25505 On AIX, the resolver library initializes some internal structure on
25506 the first call to @code{get*by*} functions, which are used to implement
25507 @code{GNAT.Sockets.Get_Host_By_Name} and
25508 @code{GNAT.Sockets.Get_Host_By_Address}.
25509 If such initialization occurs within an Ada task, and the stack size for
25510 the task is the default size, a stack overflow may occur.
25512 To avoid this overflow, the user should either ensure that the first call
25513 to @code{GNAT.Sockets.Get_Host_By_Name} or
25514 @code{GNAT.Sockets.Get_Host_By_Addrss}
25515 occurs in the environment task, or use @code{pragma Storage_Size} to
25516 specify a sufficiently large size for the stack of the task that contains
25519 @node Irix-Specific Considerations
25520 @section Irix-Specific Considerations
25521 @cindex Irix libraries
25524 The GCC support libraries coming with the Irix compiler have moved to
25525 their canonical place with respect to the general Irix ABI related
25526 conventions. Running applications built with the default shared GNAT
25527 run-time now requires the LD_LIBRARY_PATH environment variable to
25528 include this location. A possible way to achieve this is to issue the
25529 following command line on a bash prompt:
25533 $ LD_LIBRARY_PATH=$LD_LIBRARY_PATH:`dirname \`gcc --print-file-name=libgcc_s.so\``
25537 @node RTX-Specific Considerations
25538 @section RTX-Specific Considerations
25539 @cindex RTX libraries
25542 The Real-time Extension (RTX) to Windows is based on the Windows Win32
25543 API. Applications can be built to work in two different modes:
25547 Windows executables that run in Ring 3 to utilize memory protection
25548 (@emph{rts-rtx-w32}).
25551 Real-time subsystem (RTSS) executables that run in Ring 0, where
25552 performance can be optimized with RTSS applications taking precedent
25553 over all Windows applications (@emph{rts-rtx-rtss}).
25557 @node HP-UX-Specific Considerations
25558 @section HP-UX-Specific Considerations
25559 @cindex HP-UX Scheduling
25562 On HP-UX, appropriate privileges are required to change the scheduling
25563 parameters of a task. The calling process must have appropriate
25564 privileges or be a member of a group having @code{PRIV_RTSCHED} access to
25565 successfully change the scheduling parameters.
25567 By default, GNAT uses the @code{SCHED_HPUX} policy. To have access to the
25568 priority range 0-31 either the @code{FIFO_Within_Priorities} or the
25569 @code{Round_Robin_Within_Priorities} scheduling policies need to be set.
25571 To specify the @code{FIFO_Within_Priorities} scheduling policy you can use
25572 one of the following:
25576 @code{pragma Time_Slice (0.0)}
25577 @cindex pragma Time_Slice
25579 the corresponding binder option @option{-T0}
25580 @cindex @option{-T0} option
25582 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
25583 @cindex pragma Task_Dispatching_Policy
25587 To specify the @code{Round_Robin_Within_Priorities}, scheduling policy
25588 you should use @code{pragma Time_Slice} with a
25589 value greater than @code{0.0}, or use the corresponding @option{-T}
25590 binder option, or set the @code{pragma Task_Dispatching_Policy
25591 (Round_Robin_Within_Priorities)}.
25593 @c *******************************
25594 @node Example of Binder Output File
25595 @appendix Example of Binder Output File
25598 This Appendix displays the source code for @command{gnatbind}'s output
25599 file generated for a simple ``Hello World'' program.
25600 Comments have been added for clarification purposes.
25602 @smallexample @c adanocomment
25606 -- The package is called Ada_Main unless this name is actually used
25607 -- as a unit name in the partition, in which case some other unique
25611 package ada_main is
25613 Elab_Final_Code : Integer;
25614 pragma Import (C, Elab_Final_Code, "__gnat_inside_elab_final_code");
25616 -- The main program saves the parameters (argument count,
25617 -- argument values, environment pointer) in global variables
25618 -- for later access by other units including
25619 -- Ada.Command_Line.
25621 gnat_argc : Integer;
25622 gnat_argv : System.Address;
25623 gnat_envp : System.Address;
25625 -- The actual variables are stored in a library routine. This
25626 -- is useful for some shared library situations, where there
25627 -- are problems if variables are not in the library.
25629 pragma Import (C, gnat_argc);
25630 pragma Import (C, gnat_argv);
25631 pragma Import (C, gnat_envp);
25633 -- The exit status is similarly an external location
25635 gnat_exit_status : Integer;
25636 pragma Import (C, gnat_exit_status);
25638 GNAT_Version : constant String :=
25639 "GNAT Version: 6.0.0w (20061115)";
25640 pragma Export (C, GNAT_Version, "__gnat_version");
25642 -- This is the generated adafinal routine that performs
25643 -- finalization at the end of execution. In the case where
25644 -- Ada is the main program, this main program makes a call
25645 -- to adafinal at program termination.
25647 procedure adafinal;
25648 pragma Export (C, adafinal, "adafinal");
25650 -- This is the generated adainit routine that performs
25651 -- initialization at the start of execution. In the case
25652 -- where Ada is the main program, this main program makes
25653 -- a call to adainit at program startup.
25656 pragma Export (C, adainit, "adainit");
25658 -- This routine is called at the start of execution. It is
25659 -- a dummy routine that is used by the debugger to breakpoint
25660 -- at the start of execution.
25662 procedure Break_Start;
25663 pragma Import (C, Break_Start, "__gnat_break_start");
25665 -- This is the actual generated main program (it would be
25666 -- suppressed if the no main program switch were used). As
25667 -- required by standard system conventions, this program has
25668 -- the external name main.
25672 argv : System.Address;
25673 envp : System.Address)
25675 pragma Export (C, main, "main");
25677 -- The following set of constants give the version
25678 -- identification values for every unit in the bound
25679 -- partition. This identification is computed from all
25680 -- dependent semantic units, and corresponds to the
25681 -- string that would be returned by use of the
25682 -- Body_Version or Version attributes.
25684 type Version_32 is mod 2 ** 32;
25685 u00001 : constant Version_32 := 16#7880BEB3#;
25686 u00002 : constant Version_32 := 16#0D24CBD0#;
25687 u00003 : constant Version_32 := 16#3283DBEB#;
25688 u00004 : constant Version_32 := 16#2359F9ED#;
25689 u00005 : constant Version_32 := 16#664FB847#;
25690 u00006 : constant Version_32 := 16#68E803DF#;
25691 u00007 : constant Version_32 := 16#5572E604#;
25692 u00008 : constant Version_32 := 16#46B173D8#;
25693 u00009 : constant Version_32 := 16#156A40CF#;
25694 u00010 : constant Version_32 := 16#033DABE0#;
25695 u00011 : constant Version_32 := 16#6AB38FEA#;
25696 u00012 : constant Version_32 := 16#22B6217D#;
25697 u00013 : constant Version_32 := 16#68A22947#;
25698 u00014 : constant Version_32 := 16#18CC4A56#;
25699 u00015 : constant Version_32 := 16#08258E1B#;
25700 u00016 : constant Version_32 := 16#367D5222#;
25701 u00017 : constant Version_32 := 16#20C9ECA4#;
25702 u00018 : constant Version_32 := 16#50D32CB6#;
25703 u00019 : constant Version_32 := 16#39A8BB77#;
25704 u00020 : constant Version_32 := 16#5CF8FA2B#;
25705 u00021 : constant Version_32 := 16#2F1EB794#;
25706 u00022 : constant Version_32 := 16#31AB6444#;
25707 u00023 : constant Version_32 := 16#1574B6E9#;
25708 u00024 : constant Version_32 := 16#5109C189#;
25709 u00025 : constant Version_32 := 16#56D770CD#;
25710 u00026 : constant Version_32 := 16#02F9DE3D#;
25711 u00027 : constant Version_32 := 16#08AB6B2C#;
25712 u00028 : constant Version_32 := 16#3FA37670#;
25713 u00029 : constant Version_32 := 16#476457A0#;
25714 u00030 : constant Version_32 := 16#731E1B6E#;
25715 u00031 : constant Version_32 := 16#23C2E789#;
25716 u00032 : constant Version_32 := 16#0F1BD6A1#;
25717 u00033 : constant Version_32 := 16#7C25DE96#;
25718 u00034 : constant Version_32 := 16#39ADFFA2#;
25719 u00035 : constant Version_32 := 16#571DE3E7#;
25720 u00036 : constant Version_32 := 16#5EB646AB#;
25721 u00037 : constant Version_32 := 16#4249379B#;
25722 u00038 : constant Version_32 := 16#0357E00A#;
25723 u00039 : constant Version_32 := 16#3784FB72#;
25724 u00040 : constant Version_32 := 16#2E723019#;
25725 u00041 : constant Version_32 := 16#623358EA#;
25726 u00042 : constant Version_32 := 16#107F9465#;
25727 u00043 : constant Version_32 := 16#6843F68A#;
25728 u00044 : constant Version_32 := 16#63305874#;
25729 u00045 : constant Version_32 := 16#31E56CE1#;
25730 u00046 : constant Version_32 := 16#02917970#;
25731 u00047 : constant Version_32 := 16#6CCBA70E#;
25732 u00048 : constant Version_32 := 16#41CD4204#;
25733 u00049 : constant Version_32 := 16#572E3F58#;
25734 u00050 : constant Version_32 := 16#20729FF5#;
25735 u00051 : constant Version_32 := 16#1D4F93E8#;
25736 u00052 : constant Version_32 := 16#30B2EC3D#;
25737 u00053 : constant Version_32 := 16#34054F96#;
25738 u00054 : constant Version_32 := 16#5A199860#;
25739 u00055 : constant Version_32 := 16#0E7F912B#;
25740 u00056 : constant Version_32 := 16#5760634A#;
25741 u00057 : constant Version_32 := 16#5D851835#;
25743 -- The following Export pragmas export the version numbers
25744 -- with symbolic names ending in B (for body) or S
25745 -- (for spec) so that they can be located in a link. The
25746 -- information provided here is sufficient to track down
25747 -- the exact versions of units used in a given build.
25749 pragma Export (C, u00001, "helloB");
25750 pragma Export (C, u00002, "system__standard_libraryB");
25751 pragma Export (C, u00003, "system__standard_libraryS");
25752 pragma Export (C, u00004, "adaS");
25753 pragma Export (C, u00005, "ada__text_ioB");
25754 pragma Export (C, u00006, "ada__text_ioS");
25755 pragma Export (C, u00007, "ada__exceptionsB");
25756 pragma Export (C, u00008, "ada__exceptionsS");
25757 pragma Export (C, u00009, "gnatS");
25758 pragma Export (C, u00010, "gnat__heap_sort_aB");
25759 pragma Export (C, u00011, "gnat__heap_sort_aS");
25760 pragma Export (C, u00012, "systemS");
25761 pragma Export (C, u00013, "system__exception_tableB");
25762 pragma Export (C, u00014, "system__exception_tableS");
25763 pragma Export (C, u00015, "gnat__htableB");
25764 pragma Export (C, u00016, "gnat__htableS");
25765 pragma Export (C, u00017, "system__exceptionsS");
25766 pragma Export (C, u00018, "system__machine_state_operationsB");
25767 pragma Export (C, u00019, "system__machine_state_operationsS");
25768 pragma Export (C, u00020, "system__machine_codeS");
25769 pragma Export (C, u00021, "system__storage_elementsB");
25770 pragma Export (C, u00022, "system__storage_elementsS");
25771 pragma Export (C, u00023, "system__secondary_stackB");
25772 pragma Export (C, u00024, "system__secondary_stackS");
25773 pragma Export (C, u00025, "system__parametersB");
25774 pragma Export (C, u00026, "system__parametersS");
25775 pragma Export (C, u00027, "system__soft_linksB");
25776 pragma Export (C, u00028, "system__soft_linksS");
25777 pragma Export (C, u00029, "system__stack_checkingB");
25778 pragma Export (C, u00030, "system__stack_checkingS");
25779 pragma Export (C, u00031, "system__tracebackB");
25780 pragma Export (C, u00032, "system__tracebackS");
25781 pragma Export (C, u00033, "ada__streamsS");
25782 pragma Export (C, u00034, "ada__tagsB");
25783 pragma Export (C, u00035, "ada__tagsS");
25784 pragma Export (C, u00036, "system__string_opsB");
25785 pragma Export (C, u00037, "system__string_opsS");
25786 pragma Export (C, u00038, "interfacesS");
25787 pragma Export (C, u00039, "interfaces__c_streamsB");
25788 pragma Export (C, u00040, "interfaces__c_streamsS");
25789 pragma Export (C, u00041, "system__file_ioB");
25790 pragma Export (C, u00042, "system__file_ioS");
25791 pragma Export (C, u00043, "ada__finalizationB");
25792 pragma Export (C, u00044, "ada__finalizationS");
25793 pragma Export (C, u00045, "system__finalization_rootB");
25794 pragma Export (C, u00046, "system__finalization_rootS");
25795 pragma Export (C, u00047, "system__finalization_implementationB");
25796 pragma Export (C, u00048, "system__finalization_implementationS");
25797 pragma Export (C, u00049, "system__string_ops_concat_3B");
25798 pragma Export (C, u00050, "system__string_ops_concat_3S");
25799 pragma Export (C, u00051, "system__stream_attributesB");
25800 pragma Export (C, u00052, "system__stream_attributesS");
25801 pragma Export (C, u00053, "ada__io_exceptionsS");
25802 pragma Export (C, u00054, "system__unsigned_typesS");
25803 pragma Export (C, u00055, "system__file_control_blockS");
25804 pragma Export (C, u00056, "ada__finalization__list_controllerB");
25805 pragma Export (C, u00057, "ada__finalization__list_controllerS");
25807 -- BEGIN ELABORATION ORDER
25810 -- gnat.heap_sort_a (spec)
25811 -- gnat.heap_sort_a (body)
25812 -- gnat.htable (spec)
25813 -- gnat.htable (body)
25814 -- interfaces (spec)
25816 -- system.machine_code (spec)
25817 -- system.parameters (spec)
25818 -- system.parameters (body)
25819 -- interfaces.c_streams (spec)
25820 -- interfaces.c_streams (body)
25821 -- system.standard_library (spec)
25822 -- ada.exceptions (spec)
25823 -- system.exception_table (spec)
25824 -- system.exception_table (body)
25825 -- ada.io_exceptions (spec)
25826 -- system.exceptions (spec)
25827 -- system.storage_elements (spec)
25828 -- system.storage_elements (body)
25829 -- system.machine_state_operations (spec)
25830 -- system.machine_state_operations (body)
25831 -- system.secondary_stack (spec)
25832 -- system.stack_checking (spec)
25833 -- system.soft_links (spec)
25834 -- system.soft_links (body)
25835 -- system.stack_checking (body)
25836 -- system.secondary_stack (body)
25837 -- system.standard_library (body)
25838 -- system.string_ops (spec)
25839 -- system.string_ops (body)
25842 -- ada.streams (spec)
25843 -- system.finalization_root (spec)
25844 -- system.finalization_root (body)
25845 -- system.string_ops_concat_3 (spec)
25846 -- system.string_ops_concat_3 (body)
25847 -- system.traceback (spec)
25848 -- system.traceback (body)
25849 -- ada.exceptions (body)
25850 -- system.unsigned_types (spec)
25851 -- system.stream_attributes (spec)
25852 -- system.stream_attributes (body)
25853 -- system.finalization_implementation (spec)
25854 -- system.finalization_implementation (body)
25855 -- ada.finalization (spec)
25856 -- ada.finalization (body)
25857 -- ada.finalization.list_controller (spec)
25858 -- ada.finalization.list_controller (body)
25859 -- system.file_control_block (spec)
25860 -- system.file_io (spec)
25861 -- system.file_io (body)
25862 -- ada.text_io (spec)
25863 -- ada.text_io (body)
25865 -- END ELABORATION ORDER
25869 -- The following source file name pragmas allow the generated file
25870 -- names to be unique for different main programs. They are needed
25871 -- since the package name will always be Ada_Main.
25873 pragma Source_File_Name (ada_main, Spec_File_Name => "b~hello.ads");
25874 pragma Source_File_Name (ada_main, Body_File_Name => "b~hello.adb");
25876 -- Generated package body for Ada_Main starts here
25878 package body ada_main is
25880 -- The actual finalization is performed by calling the
25881 -- library routine in System.Standard_Library.Adafinal
25883 procedure Do_Finalize;
25884 pragma Import (C, Do_Finalize, "system__standard_library__adafinal");
25891 procedure adainit is
25893 -- These booleans are set to True once the associated unit has
25894 -- been elaborated. It is also used to avoid elaborating the
25895 -- same unit twice.
25898 pragma Import (Ada, E040, "interfaces__c_streams_E");
25901 pragma Import (Ada, E008, "ada__exceptions_E");
25904 pragma Import (Ada, E014, "system__exception_table_E");
25907 pragma Import (Ada, E053, "ada__io_exceptions_E");
25910 pragma Import (Ada, E017, "system__exceptions_E");
25913 pragma Import (Ada, E024, "system__secondary_stack_E");
25916 pragma Import (Ada, E030, "system__stack_checking_E");
25919 pragma Import (Ada, E028, "system__soft_links_E");
25922 pragma Import (Ada, E035, "ada__tags_E");
25925 pragma Import (Ada, E033, "ada__streams_E");
25928 pragma Import (Ada, E046, "system__finalization_root_E");
25931 pragma Import (Ada, E048, "system__finalization_implementation_E");
25934 pragma Import (Ada, E044, "ada__finalization_E");
25937 pragma Import (Ada, E057, "ada__finalization__list_controller_E");
25940 pragma Import (Ada, E055, "system__file_control_block_E");
25943 pragma Import (Ada, E042, "system__file_io_E");
25946 pragma Import (Ada, E006, "ada__text_io_E");
25948 -- Set_Globals is a library routine that stores away the
25949 -- value of the indicated set of global values in global
25950 -- variables within the library.
25952 procedure Set_Globals
25953 (Main_Priority : Integer;
25954 Time_Slice_Value : Integer;
25955 WC_Encoding : Character;
25956 Locking_Policy : Character;
25957 Queuing_Policy : Character;
25958 Task_Dispatching_Policy : Character;
25959 Adafinal : System.Address;
25960 Unreserve_All_Interrupts : Integer;
25961 Exception_Tracebacks : Integer);
25962 @findex __gnat_set_globals
25963 pragma Import (C, Set_Globals, "__gnat_set_globals");
25965 -- SDP_Table_Build is a library routine used to build the
25966 -- exception tables. See unit Ada.Exceptions in files
25967 -- a-except.ads/adb for full details of how zero cost
25968 -- exception handling works. This procedure, the call to
25969 -- it, and the two following tables are all omitted if the
25970 -- build is in longjmp/setjmp exception mode.
25972 @findex SDP_Table_Build
25973 @findex Zero Cost Exceptions
25974 procedure SDP_Table_Build
25975 (SDP_Addresses : System.Address;
25976 SDP_Count : Natural;
25977 Elab_Addresses : System.Address;
25978 Elab_Addr_Count : Natural);
25979 pragma Import (C, SDP_Table_Build, "__gnat_SDP_Table_Build");
25981 -- Table of Unit_Exception_Table addresses. Used for zero
25982 -- cost exception handling to build the top level table.
25984 ST : aliased constant array (1 .. 23) of System.Address := (
25986 Ada.Text_Io'UET_Address,
25987 Ada.Exceptions'UET_Address,
25988 Gnat.Heap_Sort_A'UET_Address,
25989 System.Exception_Table'UET_Address,
25990 System.Machine_State_Operations'UET_Address,
25991 System.Secondary_Stack'UET_Address,
25992 System.Parameters'UET_Address,
25993 System.Soft_Links'UET_Address,
25994 System.Stack_Checking'UET_Address,
25995 System.Traceback'UET_Address,
25996 Ada.Streams'UET_Address,
25997 Ada.Tags'UET_Address,
25998 System.String_Ops'UET_Address,
25999 Interfaces.C_Streams'UET_Address,
26000 System.File_Io'UET_Address,
26001 Ada.Finalization'UET_Address,
26002 System.Finalization_Root'UET_Address,
26003 System.Finalization_Implementation'UET_Address,
26004 System.String_Ops_Concat_3'UET_Address,
26005 System.Stream_Attributes'UET_Address,
26006 System.File_Control_Block'UET_Address,
26007 Ada.Finalization.List_Controller'UET_Address);
26009 -- Table of addresses of elaboration routines. Used for
26010 -- zero cost exception handling to make sure these
26011 -- addresses are included in the top level procedure
26014 EA : aliased constant array (1 .. 23) of System.Address := (
26015 adainit'Code_Address,
26016 Do_Finalize'Code_Address,
26017 Ada.Exceptions'Elab_Spec'Address,
26018 System.Exceptions'Elab_Spec'Address,
26019 Interfaces.C_Streams'Elab_Spec'Address,
26020 System.Exception_Table'Elab_Body'Address,
26021 Ada.Io_Exceptions'Elab_Spec'Address,
26022 System.Stack_Checking'Elab_Spec'Address,
26023 System.Soft_Links'Elab_Body'Address,
26024 System.Secondary_Stack'Elab_Body'Address,
26025 Ada.Tags'Elab_Spec'Address,
26026 Ada.Tags'Elab_Body'Address,
26027 Ada.Streams'Elab_Spec'Address,
26028 System.Finalization_Root'Elab_Spec'Address,
26029 Ada.Exceptions'Elab_Body'Address,
26030 System.Finalization_Implementation'Elab_Spec'Address,
26031 System.Finalization_Implementation'Elab_Body'Address,
26032 Ada.Finalization'Elab_Spec'Address,
26033 Ada.Finalization.List_Controller'Elab_Spec'Address,
26034 System.File_Control_Block'Elab_Spec'Address,
26035 System.File_Io'Elab_Body'Address,
26036 Ada.Text_Io'Elab_Spec'Address,
26037 Ada.Text_Io'Elab_Body'Address);
26039 -- Start of processing for adainit
26043 -- Call SDP_Table_Build to build the top level procedure
26044 -- table for zero cost exception handling (omitted in
26045 -- longjmp/setjmp mode).
26047 SDP_Table_Build (ST'Address, 23, EA'Address, 23);
26049 -- Call Set_Globals to record various information for
26050 -- this partition. The values are derived by the binder
26051 -- from information stored in the ali files by the compiler.
26053 @findex __gnat_set_globals
26055 (Main_Priority => -1,
26056 -- Priority of main program, -1 if no pragma Priority used
26058 Time_Slice_Value => -1,
26059 -- Time slice from Time_Slice pragma, -1 if none used
26061 WC_Encoding => 'b',
26062 -- Wide_Character encoding used, default is brackets
26064 Locking_Policy => ' ',
26065 -- Locking_Policy used, default of space means not
26066 -- specified, otherwise it is the first character of
26067 -- the policy name.
26069 Queuing_Policy => ' ',
26070 -- Queuing_Policy used, default of space means not
26071 -- specified, otherwise it is the first character of
26072 -- the policy name.
26074 Task_Dispatching_Policy => ' ',
26075 -- Task_Dispatching_Policy used, default of space means
26076 -- not specified, otherwise first character of the
26079 Adafinal => System.Null_Address,
26080 -- Address of Adafinal routine, not used anymore
26082 Unreserve_All_Interrupts => 0,
26083 -- Set true if pragma Unreserve_All_Interrupts was used
26085 Exception_Tracebacks => 0);
26086 -- Indicates if exception tracebacks are enabled
26088 Elab_Final_Code := 1;
26090 -- Now we have the elaboration calls for all units in the partition.
26091 -- The Elab_Spec and Elab_Body attributes generate references to the
26092 -- implicit elaboration procedures generated by the compiler for
26093 -- each unit that requires elaboration.
26096 Interfaces.C_Streams'Elab_Spec;
26100 Ada.Exceptions'Elab_Spec;
26103 System.Exception_Table'Elab_Body;
26107 Ada.Io_Exceptions'Elab_Spec;
26111 System.Exceptions'Elab_Spec;
26115 System.Stack_Checking'Elab_Spec;
26118 System.Soft_Links'Elab_Body;
26123 System.Secondary_Stack'Elab_Body;
26127 Ada.Tags'Elab_Spec;
26130 Ada.Tags'Elab_Body;
26134 Ada.Streams'Elab_Spec;
26138 System.Finalization_Root'Elab_Spec;
26142 Ada.Exceptions'Elab_Body;
26146 System.Finalization_Implementation'Elab_Spec;
26149 System.Finalization_Implementation'Elab_Body;
26153 Ada.Finalization'Elab_Spec;
26157 Ada.Finalization.List_Controller'Elab_Spec;
26161 System.File_Control_Block'Elab_Spec;
26165 System.File_Io'Elab_Body;
26169 Ada.Text_Io'Elab_Spec;
26172 Ada.Text_Io'Elab_Body;
26176 Elab_Final_Code := 0;
26184 procedure adafinal is
26193 -- main is actually a function, as in the ANSI C standard,
26194 -- defined to return the exit status. The three parameters
26195 -- are the argument count, argument values and environment
26198 @findex Main Program
26201 argv : System.Address;
26202 envp : System.Address)
26205 -- The initialize routine performs low level system
26206 -- initialization using a standard library routine which
26207 -- sets up signal handling and performs any other
26208 -- required setup. The routine can be found in file
26211 @findex __gnat_initialize
26212 procedure initialize;
26213 pragma Import (C, initialize, "__gnat_initialize");
26215 -- The finalize routine performs low level system
26216 -- finalization using a standard library routine. The
26217 -- routine is found in file a-final.c and in the standard
26218 -- distribution is a dummy routine that does nothing, so
26219 -- really this is a hook for special user finalization.
26221 @findex __gnat_finalize
26222 procedure finalize;
26223 pragma Import (C, finalize, "__gnat_finalize");
26225 -- We get to the main program of the partition by using
26226 -- pragma Import because if we try to with the unit and
26227 -- call it Ada style, then not only do we waste time
26228 -- recompiling it, but also, we don't really know the right
26229 -- switches (e.g.@: identifier character set) to be used
26232 procedure Ada_Main_Program;
26233 pragma Import (Ada, Ada_Main_Program, "_ada_hello");
26235 -- Start of processing for main
26238 -- Save global variables
26244 -- Call low level system initialization
26248 -- Call our generated Ada initialization routine
26252 -- This is the point at which we want the debugger to get
26257 -- Now we call the main program of the partition
26261 -- Perform Ada finalization
26265 -- Perform low level system finalization
26269 -- Return the proper exit status
26270 return (gnat_exit_status);
26273 -- This section is entirely comments, so it has no effect on the
26274 -- compilation of the Ada_Main package. It provides the list of
26275 -- object files and linker options, as well as some standard
26276 -- libraries needed for the link. The gnatlink utility parses
26277 -- this b~hello.adb file to read these comment lines to generate
26278 -- the appropriate command line arguments for the call to the
26279 -- system linker. The BEGIN/END lines are used for sentinels for
26280 -- this parsing operation.
26282 -- The exact file names will of course depend on the environment,
26283 -- host/target and location of files on the host system.
26285 @findex Object file list
26286 -- BEGIN Object file/option list
26289 -- -L/usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/
26290 -- /usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/libgnat.a
26291 -- END Object file/option list
26297 The Ada code in the above example is exactly what is generated by the
26298 binder. We have added comments to more clearly indicate the function
26299 of each part of the generated @code{Ada_Main} package.
26301 The code is standard Ada in all respects, and can be processed by any
26302 tools that handle Ada. In particular, it is possible to use the debugger
26303 in Ada mode to debug the generated @code{Ada_Main} package. For example,
26304 suppose that for reasons that you do not understand, your program is crashing
26305 during elaboration of the body of @code{Ada.Text_IO}. To locate this bug,
26306 you can place a breakpoint on the call:
26308 @smallexample @c ada
26309 Ada.Text_Io'Elab_Body;
26313 and trace the elaboration routine for this package to find out where
26314 the problem might be (more usually of course you would be debugging
26315 elaboration code in your own application).
26317 @node Elaboration Order Handling in GNAT
26318 @appendix Elaboration Order Handling in GNAT
26319 @cindex Order of elaboration
26320 @cindex Elaboration control
26323 * Elaboration Code::
26324 * Checking the Elaboration Order::
26325 * Controlling the Elaboration Order::
26326 * Controlling Elaboration in GNAT - Internal Calls::
26327 * Controlling Elaboration in GNAT - External Calls::
26328 * Default Behavior in GNAT - Ensuring Safety::
26329 * Treatment of Pragma Elaborate::
26330 * Elaboration Issues for Library Tasks::
26331 * Mixing Elaboration Models::
26332 * What to Do If the Default Elaboration Behavior Fails::
26333 * Elaboration for Access-to-Subprogram Values::
26334 * Summary of Procedures for Elaboration Control::
26335 * Other Elaboration Order Considerations::
26339 This chapter describes the handling of elaboration code in Ada and
26340 in GNAT, and discusses how the order of elaboration of program units can
26341 be controlled in GNAT, either automatically or with explicit programming
26344 @node Elaboration Code
26345 @section Elaboration Code
26348 Ada provides rather general mechanisms for executing code at elaboration
26349 time, that is to say before the main program starts executing. Such code arises
26353 @item Initializers for variables.
26354 Variables declared at the library level, in package specs or bodies, can
26355 require initialization that is performed at elaboration time, as in:
26356 @smallexample @c ada
26358 Sqrt_Half : Float := Sqrt (0.5);
26362 @item Package initialization code
26363 Code in a @code{BEGIN-END} section at the outer level of a package body is
26364 executed as part of the package body elaboration code.
26366 @item Library level task allocators
26367 Tasks that are declared using task allocators at the library level
26368 start executing immediately and hence can execute at elaboration time.
26372 Subprogram calls are possible in any of these contexts, which means that
26373 any arbitrary part of the program may be executed as part of the elaboration
26374 code. It is even possible to write a program which does all its work at
26375 elaboration time, with a null main program, although stylistically this
26376 would usually be considered an inappropriate way to structure
26379 An important concern arises in the context of elaboration code:
26380 we have to be sure that it is executed in an appropriate order. What we
26381 have is a series of elaboration code sections, potentially one section
26382 for each unit in the program. It is important that these execute
26383 in the correct order. Correctness here means that, taking the above
26384 example of the declaration of @code{Sqrt_Half},
26385 if some other piece of
26386 elaboration code references @code{Sqrt_Half},
26387 then it must run after the
26388 section of elaboration code that contains the declaration of
26391 There would never be any order of elaboration problem if we made a rule
26392 that whenever you @code{with} a unit, you must elaborate both the spec and body
26393 of that unit before elaborating the unit doing the @code{with}'ing:
26395 @smallexample @c ada
26399 package Unit_2 is @dots{}
26405 would require that both the body and spec of @code{Unit_1} be elaborated
26406 before the spec of @code{Unit_2}. However, a rule like that would be far too
26407 restrictive. In particular, it would make it impossible to have routines
26408 in separate packages that were mutually recursive.
26410 You might think that a clever enough compiler could look at the actual
26411 elaboration code and determine an appropriate correct order of elaboration,
26412 but in the general case, this is not possible. Consider the following
26415 In the body of @code{Unit_1}, we have a procedure @code{Func_1}
26417 the variable @code{Sqrt_1}, which is declared in the elaboration code
26418 of the body of @code{Unit_1}:
26420 @smallexample @c ada
26422 Sqrt_1 : Float := Sqrt (0.1);
26427 The elaboration code of the body of @code{Unit_1} also contains:
26429 @smallexample @c ada
26432 if expression_1 = 1 then
26433 Q := Unit_2.Func_2;
26440 @code{Unit_2} is exactly parallel,
26441 it has a procedure @code{Func_2} that references
26442 the variable @code{Sqrt_2}, which is declared in the elaboration code of
26443 the body @code{Unit_2}:
26445 @smallexample @c ada
26447 Sqrt_2 : Float := Sqrt (0.1);
26452 The elaboration code of the body of @code{Unit_2} also contains:
26454 @smallexample @c ada
26457 if expression_2 = 2 then
26458 Q := Unit_1.Func_1;
26465 Now the question is, which of the following orders of elaboration is
26490 If you carefully analyze the flow here, you will see that you cannot tell
26491 at compile time the answer to this question.
26492 If @code{expression_1} is not equal to 1,
26493 and @code{expression_2} is not equal to 2,
26494 then either order is acceptable, because neither of the function calls is
26495 executed. If both tests evaluate to true, then neither order is acceptable
26496 and in fact there is no correct order.
26498 If one of the two expressions is true, and the other is false, then one
26499 of the above orders is correct, and the other is incorrect. For example,
26500 if @code{expression_1} /= 1 and @code{expression_2} = 2,
26501 then the call to @code{Func_1}
26502 will occur, but not the call to @code{Func_2.}
26503 This means that it is essential
26504 to elaborate the body of @code{Unit_1} before
26505 the body of @code{Unit_2}, so the first
26506 order of elaboration is correct and the second is wrong.
26508 By making @code{expression_1} and @code{expression_2}
26509 depend on input data, or perhaps
26510 the time of day, we can make it impossible for the compiler or binder
26511 to figure out which of these expressions will be true, and hence it
26512 is impossible to guarantee a safe order of elaboration at run time.
26514 @node Checking the Elaboration Order
26515 @section Checking the Elaboration Order
26518 In some languages that involve the same kind of elaboration problems,
26519 e.g.@: Java and C++, the programmer is expected to worry about these
26520 ordering problems himself, and it is common to
26521 write a program in which an incorrect elaboration order gives
26522 surprising results, because it references variables before they
26524 Ada is designed to be a safe language, and a programmer-beware approach is
26525 clearly not sufficient. Consequently, the language provides three lines
26529 @item Standard rules
26530 Some standard rules restrict the possible choice of elaboration
26531 order. In particular, if you @code{with} a unit, then its spec is always
26532 elaborated before the unit doing the @code{with}. Similarly, a parent
26533 spec is always elaborated before the child spec, and finally
26534 a spec is always elaborated before its corresponding body.
26536 @item Dynamic elaboration checks
26537 @cindex Elaboration checks
26538 @cindex Checks, elaboration
26539 Dynamic checks are made at run time, so that if some entity is accessed
26540 before it is elaborated (typically by means of a subprogram call)
26541 then the exception (@code{Program_Error}) is raised.
26543 @item Elaboration control
26544 Facilities are provided for the programmer to specify the desired order
26548 Let's look at these facilities in more detail. First, the rules for
26549 dynamic checking. One possible rule would be simply to say that the
26550 exception is raised if you access a variable which has not yet been
26551 elaborated. The trouble with this approach is that it could require
26552 expensive checks on every variable reference. Instead Ada has two
26553 rules which are a little more restrictive, but easier to check, and
26557 @item Restrictions on calls
26558 A subprogram can only be called at elaboration time if its body
26559 has been elaborated. The rules for elaboration given above guarantee
26560 that the spec of the subprogram has been elaborated before the
26561 call, but not the body. If this rule is violated, then the
26562 exception @code{Program_Error} is raised.
26564 @item Restrictions on instantiations
26565 A generic unit can only be instantiated if the body of the generic
26566 unit has been elaborated. Again, the rules for elaboration given above
26567 guarantee that the spec of the generic unit has been elaborated
26568 before the instantiation, but not the body. If this rule is
26569 violated, then the exception @code{Program_Error} is raised.
26573 The idea is that if the body has been elaborated, then any variables
26574 it references must have been elaborated; by checking for the body being
26575 elaborated we guarantee that none of its references causes any
26576 trouble. As we noted above, this is a little too restrictive, because a
26577 subprogram that has no non-local references in its body may in fact be safe
26578 to call. However, it really would be unsafe to rely on this, because
26579 it would mean that the caller was aware of details of the implementation
26580 in the body. This goes against the basic tenets of Ada.
26582 A plausible implementation can be described as follows.
26583 A Boolean variable is associated with each subprogram
26584 and each generic unit. This variable is initialized to False, and is set to
26585 True at the point body is elaborated. Every call or instantiation checks the
26586 variable, and raises @code{Program_Error} if the variable is False.
26588 Note that one might think that it would be good enough to have one Boolean
26589 variable for each package, but that would not deal with cases of trying
26590 to call a body in the same package as the call
26591 that has not been elaborated yet.
26592 Of course a compiler may be able to do enough analysis to optimize away
26593 some of the Boolean variables as unnecessary, and @code{GNAT} indeed
26594 does such optimizations, but still the easiest conceptual model is to
26595 think of there being one variable per subprogram.
26597 @node Controlling the Elaboration Order
26598 @section Controlling the Elaboration Order
26601 In the previous section we discussed the rules in Ada which ensure
26602 that @code{Program_Error} is raised if an incorrect elaboration order is
26603 chosen. This prevents erroneous executions, but we need mechanisms to
26604 specify a correct execution and avoid the exception altogether.
26605 To achieve this, Ada provides a number of features for controlling
26606 the order of elaboration. We discuss these features in this section.
26608 First, there are several ways of indicating to the compiler that a given
26609 unit has no elaboration problems:
26612 @item packages that do not require a body
26613 A library package that does not require a body does not permit
26614 a body (this rule was introduced in Ada 95).
26615 Thus if we have a such a package, as in:
26617 @smallexample @c ada
26620 package Definitions is
26622 type m is new integer;
26624 type a is array (1 .. 10) of m;
26625 type b is array (1 .. 20) of m;
26633 A package that @code{with}'s @code{Definitions} may safely instantiate
26634 @code{Definitions.Subp} because the compiler can determine that there
26635 definitely is no package body to worry about in this case
26638 @cindex pragma Pure
26640 Places sufficient restrictions on a unit to guarantee that
26641 no call to any subprogram in the unit can result in an
26642 elaboration problem. This means that the compiler does not need
26643 to worry about the point of elaboration of such units, and in
26644 particular, does not need to check any calls to any subprograms
26647 @item pragma Preelaborate
26648 @findex Preelaborate
26649 @cindex pragma Preelaborate
26650 This pragma places slightly less stringent restrictions on a unit than
26652 but these restrictions are still sufficient to ensure that there
26653 are no elaboration problems with any calls to the unit.
26655 @item pragma Elaborate_Body
26656 @findex Elaborate_Body
26657 @cindex pragma Elaborate_Body
26658 This pragma requires that the body of a unit be elaborated immediately
26659 after its spec. Suppose a unit @code{A} has such a pragma,
26660 and unit @code{B} does
26661 a @code{with} of unit @code{A}. Recall that the standard rules require
26662 the spec of unit @code{A}
26663 to be elaborated before the @code{with}'ing unit; given the pragma in
26664 @code{A}, we also know that the body of @code{A}
26665 will be elaborated before @code{B}, so
26666 that calls to @code{A} are safe and do not need a check.
26671 unlike pragma @code{Pure} and pragma @code{Preelaborate},
26673 @code{Elaborate_Body} does not guarantee that the program is
26674 free of elaboration problems, because it may not be possible
26675 to satisfy the requested elaboration order.
26676 Let's go back to the example with @code{Unit_1} and @code{Unit_2}.
26678 marks @code{Unit_1} as @code{Elaborate_Body},
26679 and not @code{Unit_2,} then the order of
26680 elaboration will be:
26692 Now that means that the call to @code{Func_1} in @code{Unit_2}
26693 need not be checked,
26694 it must be safe. But the call to @code{Func_2} in
26695 @code{Unit_1} may still fail if
26696 @code{Expression_1} is equal to 1,
26697 and the programmer must still take
26698 responsibility for this not being the case.
26700 If all units carry a pragma @code{Elaborate_Body}, then all problems are
26701 eliminated, except for calls entirely within a body, which are
26702 in any case fully under programmer control. However, using the pragma
26703 everywhere is not always possible.
26704 In particular, for our @code{Unit_1}/@code{Unit_2} example, if
26705 we marked both of them as having pragma @code{Elaborate_Body}, then
26706 clearly there would be no possible elaboration order.
26708 The above pragmas allow a server to guarantee safe use by clients, and
26709 clearly this is the preferable approach. Consequently a good rule
26710 is to mark units as @code{Pure} or @code{Preelaborate} if possible,
26711 and if this is not possible,
26712 mark them as @code{Elaborate_Body} if possible.
26713 As we have seen, there are situations where neither of these
26714 three pragmas can be used.
26715 So we also provide methods for clients to control the
26716 order of elaboration of the servers on which they depend:
26719 @item pragma Elaborate (unit)
26721 @cindex pragma Elaborate
26722 This pragma is placed in the context clause, after a @code{with} clause,
26723 and it requires that the body of the named unit be elaborated before
26724 the unit in which the pragma occurs. The idea is to use this pragma
26725 if the current unit calls at elaboration time, directly or indirectly,
26726 some subprogram in the named unit.
26728 @item pragma Elaborate_All (unit)
26729 @findex Elaborate_All
26730 @cindex pragma Elaborate_All
26731 This is a stronger version of the Elaborate pragma. Consider the
26735 Unit A @code{with}'s unit B and calls B.Func in elab code
26736 Unit B @code{with}'s unit C, and B.Func calls C.Func
26740 Now if we put a pragma @code{Elaborate (B)}
26741 in unit @code{A}, this ensures that the
26742 body of @code{B} is elaborated before the call, but not the
26743 body of @code{C}, so
26744 the call to @code{C.Func} could still cause @code{Program_Error} to
26747 The effect of a pragma @code{Elaborate_All} is stronger, it requires
26748 not only that the body of the named unit be elaborated before the
26749 unit doing the @code{with}, but also the bodies of all units that the
26750 named unit uses, following @code{with} links transitively. For example,
26751 if we put a pragma @code{Elaborate_All (B)} in unit @code{A},
26753 not only that the body of @code{B} be elaborated before @code{A},
26755 body of @code{C}, because @code{B} @code{with}'s @code{C}.
26759 We are now in a position to give a usage rule in Ada for avoiding
26760 elaboration problems, at least if dynamic dispatching and access to
26761 subprogram values are not used. We will handle these cases separately
26764 The rule is simple. If a unit has elaboration code that can directly or
26765 indirectly make a call to a subprogram in a @code{with}'ed unit, or instantiate
26766 a generic package in a @code{with}'ed unit,
26767 then if the @code{with}'ed unit does not have
26768 pragma @code{Pure} or @code{Preelaborate}, then the client should have
26769 a pragma @code{Elaborate_All}
26770 for the @code{with}'ed unit. By following this rule a client is
26771 assured that calls can be made without risk of an exception.
26773 For generic subprogram instantiations, the rule can be relaxed to
26774 require only a pragma @code{Elaborate} since elaborating the body
26775 of a subprogram cannot cause any transitive elaboration (we are
26776 not calling the subprogram in this case, just elaborating its
26779 If this rule is not followed, then a program may be in one of four
26783 @item No order exists
26784 No order of elaboration exists which follows the rules, taking into
26785 account any @code{Elaborate}, @code{Elaborate_All},
26786 or @code{Elaborate_Body} pragmas. In
26787 this case, an Ada compiler must diagnose the situation at bind
26788 time, and refuse to build an executable program.
26790 @item One or more orders exist, all incorrect
26791 One or more acceptable elaboration orders exist, and all of them
26792 generate an elaboration order problem. In this case, the binder
26793 can build an executable program, but @code{Program_Error} will be raised
26794 when the program is run.
26796 @item Several orders exist, some right, some incorrect
26797 One or more acceptable elaboration orders exists, and some of them
26798 work, and some do not. The programmer has not controlled
26799 the order of elaboration, so the binder may or may not pick one of
26800 the correct orders, and the program may or may not raise an
26801 exception when it is run. This is the worst case, because it means
26802 that the program may fail when moved to another compiler, or even
26803 another version of the same compiler.
26805 @item One or more orders exists, all correct
26806 One ore more acceptable elaboration orders exist, and all of them
26807 work. In this case the program runs successfully. This state of
26808 affairs can be guaranteed by following the rule we gave above, but
26809 may be true even if the rule is not followed.
26813 Note that one additional advantage of following our rules on the use
26814 of @code{Elaborate} and @code{Elaborate_All}
26815 is that the program continues to stay in the ideal (all orders OK) state
26816 even if maintenance
26817 changes some bodies of some units. Conversely, if a program that does
26818 not follow this rule happens to be safe at some point, this state of affairs
26819 may deteriorate silently as a result of maintenance changes.
26821 You may have noticed that the above discussion did not mention
26822 the use of @code{Elaborate_Body}. This was a deliberate omission. If you
26823 @code{with} an @code{Elaborate_Body} unit, it still may be the case that
26824 code in the body makes calls to some other unit, so it is still necessary
26825 to use @code{Elaborate_All} on such units.
26827 @node Controlling Elaboration in GNAT - Internal Calls
26828 @section Controlling Elaboration in GNAT - Internal Calls
26831 In the case of internal calls, i.e., calls within a single package, the
26832 programmer has full control over the order of elaboration, and it is up
26833 to the programmer to elaborate declarations in an appropriate order. For
26836 @smallexample @c ada
26839 function One return Float;
26843 function One return Float is
26852 will obviously raise @code{Program_Error} at run time, because function
26853 One will be called before its body is elaborated. In this case GNAT will
26854 generate a warning that the call will raise @code{Program_Error}:
26860 2. function One return Float;
26862 4. Q : Float := One;
26864 >>> warning: cannot call "One" before body is elaborated
26865 >>> warning: Program_Error will be raised at run time
26868 6. function One return Float is
26881 Note that in this particular case, it is likely that the call is safe, because
26882 the function @code{One} does not access any global variables.
26883 Nevertheless in Ada, we do not want the validity of the check to depend on
26884 the contents of the body (think about the separate compilation case), so this
26885 is still wrong, as we discussed in the previous sections.
26887 The error is easily corrected by rearranging the declarations so that the
26888 body of @code{One} appears before the declaration containing the call
26889 (note that in Ada 95 and Ada 2005,
26890 declarations can appear in any order, so there is no restriction that
26891 would prevent this reordering, and if we write:
26893 @smallexample @c ada
26896 function One return Float;
26898 function One return Float is
26909 then all is well, no warning is generated, and no
26910 @code{Program_Error} exception
26912 Things are more complicated when a chain of subprograms is executed:
26914 @smallexample @c ada
26917 function A return Integer;
26918 function B return Integer;
26919 function C return Integer;
26921 function B return Integer is begin return A; end;
26922 function C return Integer is begin return B; end;
26926 function A return Integer is begin return 1; end;
26932 Now the call to @code{C}
26933 at elaboration time in the declaration of @code{X} is correct, because
26934 the body of @code{C} is already elaborated,
26935 and the call to @code{B} within the body of
26936 @code{C} is correct, but the call
26937 to @code{A} within the body of @code{B} is incorrect, because the body
26938 of @code{A} has not been elaborated, so @code{Program_Error}
26939 will be raised on the call to @code{A}.
26940 In this case GNAT will generate a
26941 warning that @code{Program_Error} may be
26942 raised at the point of the call. Let's look at the warning:
26948 2. function A return Integer;
26949 3. function B return Integer;
26950 4. function C return Integer;
26952 6. function B return Integer is begin return A; end;
26954 >>> warning: call to "A" before body is elaborated may
26955 raise Program_Error
26956 >>> warning: "B" called at line 7
26957 >>> warning: "C" called at line 9
26959 7. function C return Integer is begin return B; end;
26961 9. X : Integer := C;
26963 11. function A return Integer is begin return 1; end;
26973 Note that the message here says ``may raise'', instead of the direct case,
26974 where the message says ``will be raised''. That's because whether
26976 actually called depends in general on run-time flow of control.
26977 For example, if the body of @code{B} said
26979 @smallexample @c ada
26982 function B return Integer is
26984 if some-condition-depending-on-input-data then
26995 then we could not know until run time whether the incorrect call to A would
26996 actually occur, so @code{Program_Error} might
26997 or might not be raised. It is possible for a compiler to
26998 do a better job of analyzing bodies, to
26999 determine whether or not @code{Program_Error}
27000 might be raised, but it certainly
27001 couldn't do a perfect job (that would require solving the halting problem
27002 and is provably impossible), and because this is a warning anyway, it does
27003 not seem worth the effort to do the analysis. Cases in which it
27004 would be relevant are rare.
27006 In practice, warnings of either of the forms given
27007 above will usually correspond to
27008 real errors, and should be examined carefully and eliminated.
27009 In the rare case where a warning is bogus, it can be suppressed by any of
27010 the following methods:
27014 Compile with the @option{-gnatws} switch set
27017 Suppress @code{Elaboration_Check} for the called subprogram
27020 Use pragma @code{Warnings_Off} to turn warnings off for the call
27024 For the internal elaboration check case,
27025 GNAT by default generates the
27026 necessary run-time checks to ensure
27027 that @code{Program_Error} is raised if any
27028 call fails an elaboration check. Of course this can only happen if a
27029 warning has been issued as described above. The use of pragma
27030 @code{Suppress (Elaboration_Check)} may (but is not guaranteed to) suppress
27031 some of these checks, meaning that it may be possible (but is not
27032 guaranteed) for a program to be able to call a subprogram whose body
27033 is not yet elaborated, without raising a @code{Program_Error} exception.
27035 @node Controlling Elaboration in GNAT - External Calls
27036 @section Controlling Elaboration in GNAT - External Calls
27039 The previous section discussed the case in which the execution of a
27040 particular thread of elaboration code occurred entirely within a
27041 single unit. This is the easy case to handle, because a programmer
27042 has direct and total control over the order of elaboration, and
27043 furthermore, checks need only be generated in cases which are rare
27044 and which the compiler can easily detect.
27045 The situation is more complex when separate compilation is taken into account.
27046 Consider the following:
27048 @smallexample @c ada
27052 function Sqrt (Arg : Float) return Float;
27055 package body Math is
27056 function Sqrt (Arg : Float) return Float is
27065 X : Float := Math.Sqrt (0.5);
27078 where @code{Main} is the main program. When this program is executed, the
27079 elaboration code must first be executed, and one of the jobs of the
27080 binder is to determine the order in which the units of a program are
27081 to be elaborated. In this case we have four units: the spec and body
27083 the spec of @code{Stuff} and the body of @code{Main}).
27084 In what order should the four separate sections of elaboration code
27087 There are some restrictions in the order of elaboration that the binder
27088 can choose. In particular, if unit U has a @code{with}
27089 for a package @code{X}, then you
27090 are assured that the spec of @code{X}
27091 is elaborated before U , but you are
27092 not assured that the body of @code{X}
27093 is elaborated before U.
27094 This means that in the above case, the binder is allowed to choose the
27105 but that's not good, because now the call to @code{Math.Sqrt}
27106 that happens during
27107 the elaboration of the @code{Stuff}
27108 spec happens before the body of @code{Math.Sqrt} is
27109 elaborated, and hence causes @code{Program_Error} exception to be raised.
27110 At first glance, one might say that the binder is misbehaving, because
27111 obviously you want to elaborate the body of something you @code{with}
27113 that is not a general rule that can be followed in all cases. Consider
27115 @smallexample @c ada
27118 package X is @dots{}
27120 package Y is @dots{}
27123 package body Y is @dots{}
27126 package body X is @dots{}
27132 This is a common arrangement, and, apart from the order of elaboration
27133 problems that might arise in connection with elaboration code, this works fine.
27134 A rule that says that you must first elaborate the body of anything you
27135 @code{with} cannot work in this case:
27136 the body of @code{X} @code{with}'s @code{Y},
27137 which means you would have to
27138 elaborate the body of @code{Y} first, but that @code{with}'s @code{X},
27140 you have to elaborate the body of @code{X} first, but @dots{} and we have a
27141 loop that cannot be broken.
27143 It is true that the binder can in many cases guess an order of elaboration
27144 that is unlikely to cause a @code{Program_Error}
27145 exception to be raised, and it tries to do so (in the
27146 above example of @code{Math/Stuff/Spec}, the GNAT binder will
27148 elaborate the body of @code{Math} right after its spec, so all will be well).
27150 However, a program that blindly relies on the binder to be helpful can
27151 get into trouble, as we discussed in the previous sections, so
27153 provides a number of facilities for assisting the programmer in
27154 developing programs that are robust with respect to elaboration order.
27156 @node Default Behavior in GNAT - Ensuring Safety
27157 @section Default Behavior in GNAT - Ensuring Safety
27160 The default behavior in GNAT ensures elaboration safety. In its
27161 default mode GNAT implements the
27162 rule we previously described as the right approach. Let's restate it:
27166 @emph{If a unit has elaboration code that can directly or indirectly make a
27167 call to a subprogram in a @code{with}'ed unit, or instantiate a generic
27168 package in a @code{with}'ed unit, then if the @code{with}'ed unit
27169 does not have pragma @code{Pure} or
27170 @code{Preelaborate}, then the client should have an
27171 @code{Elaborate_All} pragma for the @code{with}'ed unit.}
27173 @emph{In the case of instantiating a generic subprogram, it is always
27174 sufficient to have only an @code{Elaborate} pragma for the
27175 @code{with}'ed unit.}
27179 By following this rule a client is assured that calls and instantiations
27180 can be made without risk of an exception.
27182 In this mode GNAT traces all calls that are potentially made from
27183 elaboration code, and puts in any missing implicit @code{Elaborate}
27184 and @code{Elaborate_All} pragmas.
27185 The advantage of this approach is that no elaboration problems
27186 are possible if the binder can find an elaboration order that is
27187 consistent with these implicit @code{Elaborate} and
27188 @code{Elaborate_All} pragmas. The
27189 disadvantage of this approach is that no such order may exist.
27191 If the binder does not generate any diagnostics, then it means that it has
27192 found an elaboration order that is guaranteed to be safe. However, the binder
27193 may still be relying on implicitly generated @code{Elaborate} and
27194 @code{Elaborate_All} pragmas so portability to other compilers than GNAT is not
27197 If it is important to guarantee portability, then the compilations should
27200 (warn on elaboration problems) switch. This will cause warning messages
27201 to be generated indicating the missing @code{Elaborate} and
27202 @code{Elaborate_All} pragmas.
27203 Consider the following source program:
27205 @smallexample @c ada
27210 m : integer := k.r;
27217 where it is clear that there
27218 should be a pragma @code{Elaborate_All}
27219 for unit @code{k}. An implicit pragma will be generated, and it is
27220 likely that the binder will be able to honor it. However, if you want
27221 to port this program to some other Ada compiler than GNAT.
27222 it is safer to include the pragma explicitly in the source. If this
27223 unit is compiled with the
27225 switch, then the compiler outputs a warning:
27232 3. m : integer := k.r;
27234 >>> warning: call to "r" may raise Program_Error
27235 >>> warning: missing pragma Elaborate_All for "k"
27243 and these warnings can be used as a guide for supplying manually
27244 the missing pragmas. It is usually a bad idea to use this warning
27245 option during development. That's because it will warn you when
27246 you need to put in a pragma, but cannot warn you when it is time
27247 to take it out. So the use of pragma @code{Elaborate_All} may lead to
27248 unnecessary dependencies and even false circularities.
27250 This default mode is more restrictive than the Ada Reference
27251 Manual, and it is possible to construct programs which will compile
27252 using the dynamic model described there, but will run into a
27253 circularity using the safer static model we have described.
27255 Of course any Ada compiler must be able to operate in a mode
27256 consistent with the requirements of the Ada Reference Manual,
27257 and in particular must have the capability of implementing the
27258 standard dynamic model of elaboration with run-time checks.
27260 In GNAT, this standard mode can be achieved either by the use of
27261 the @option{-gnatE} switch on the compiler (@command{gcc} or
27262 @command{gnatmake}) command, or by the use of the configuration pragma:
27264 @smallexample @c ada
27265 pragma Elaboration_Checks (DYNAMIC);
27269 Either approach will cause the unit affected to be compiled using the
27270 standard dynamic run-time elaboration checks described in the Ada
27271 Reference Manual. The static model is generally preferable, since it
27272 is clearly safer to rely on compile and link time checks rather than
27273 run-time checks. However, in the case of legacy code, it may be
27274 difficult to meet the requirements of the static model. This
27275 issue is further discussed in
27276 @ref{What to Do If the Default Elaboration Behavior Fails}.
27278 Note that the static model provides a strict subset of the allowed
27279 behavior and programs of the Ada Reference Manual, so if you do
27280 adhere to the static model and no circularities exist,
27281 then you are assured that your program will
27282 work using the dynamic model, providing that you remove any
27283 pragma Elaborate statements from the source.
27285 @node Treatment of Pragma Elaborate
27286 @section Treatment of Pragma Elaborate
27287 @cindex Pragma Elaborate
27290 The use of @code{pragma Elaborate}
27291 should generally be avoided in Ada 95 and Ada 2005 programs,
27292 since there is no guarantee that transitive calls
27293 will be properly handled. Indeed at one point, this pragma was placed
27294 in Annex J (Obsolescent Features), on the grounds that it is never useful.
27296 Now that's a bit restrictive. In practice, the case in which
27297 @code{pragma Elaborate} is useful is when the caller knows that there
27298 are no transitive calls, or that the called unit contains all necessary
27299 transitive @code{pragma Elaborate} statements, and legacy code often
27300 contains such uses.
27302 Strictly speaking the static mode in GNAT should ignore such pragmas,
27303 since there is no assurance at compile time that the necessary safety
27304 conditions are met. In practice, this would cause GNAT to be incompatible
27305 with correctly written Ada 83 code that had all necessary
27306 @code{pragma Elaborate} statements in place. Consequently, we made the
27307 decision that GNAT in its default mode will believe that if it encounters
27308 a @code{pragma Elaborate} then the programmer knows what they are doing,
27309 and it will trust that no elaboration errors can occur.
27311 The result of this decision is two-fold. First to be safe using the
27312 static mode, you should remove all @code{pragma Elaborate} statements.
27313 Second, when fixing circularities in existing code, you can selectively
27314 use @code{pragma Elaborate} statements to convince the static mode of
27315 GNAT that it need not generate an implicit @code{pragma Elaborate_All}
27318 When using the static mode with @option{-gnatwl}, any use of
27319 @code{pragma Elaborate} will generate a warning about possible
27322 @node Elaboration Issues for Library Tasks
27323 @section Elaboration Issues for Library Tasks
27324 @cindex Library tasks, elaboration issues
27325 @cindex Elaboration of library tasks
27328 In this section we examine special elaboration issues that arise for
27329 programs that declare library level tasks.
27331 Generally the model of execution of an Ada program is that all units are
27332 elaborated, and then execution of the program starts. However, the
27333 declaration of library tasks definitely does not fit this model. The
27334 reason for this is that library tasks start as soon as they are declared
27335 (more precisely, as soon as the statement part of the enclosing package
27336 body is reached), that is to say before elaboration
27337 of the program is complete. This means that if such a task calls a
27338 subprogram, or an entry in another task, the callee may or may not be
27339 elaborated yet, and in the standard
27340 Reference Manual model of dynamic elaboration checks, you can even
27341 get timing dependent Program_Error exceptions, since there can be
27342 a race between the elaboration code and the task code.
27344 The static model of elaboration in GNAT seeks to avoid all such
27345 dynamic behavior, by being conservative, and the conservative
27346 approach in this particular case is to assume that all the code
27347 in a task body is potentially executed at elaboration time if
27348 a task is declared at the library level.
27350 This can definitely result in unexpected circularities. Consider
27351 the following example
27353 @smallexample @c ada
27359 type My_Int is new Integer;
27361 function Ident (M : My_Int) return My_Int;
27365 package body Decls is
27366 task body Lib_Task is
27372 function Ident (M : My_Int) return My_Int is
27380 procedure Put_Val (Arg : Decls.My_Int);
27384 package body Utils is
27385 procedure Put_Val (Arg : Decls.My_Int) is
27387 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
27394 Decls.Lib_Task.Start;
27399 If the above example is compiled in the default static elaboration
27400 mode, then a circularity occurs. The circularity comes from the call
27401 @code{Utils.Put_Val} in the task body of @code{Decls.Lib_Task}. Since
27402 this call occurs in elaboration code, we need an implicit pragma
27403 @code{Elaborate_All} for @code{Utils}. This means that not only must
27404 the spec and body of @code{Utils} be elaborated before the body
27405 of @code{Decls}, but also the spec and body of any unit that is
27406 @code{with'ed} by the body of @code{Utils} must also be elaborated before
27407 the body of @code{Decls}. This is the transitive implication of
27408 pragma @code{Elaborate_All} and it makes sense, because in general
27409 the body of @code{Put_Val} might have a call to something in a
27410 @code{with'ed} unit.
27412 In this case, the body of Utils (actually its spec) @code{with's}
27413 @code{Decls}. Unfortunately this means that the body of @code{Decls}
27414 must be elaborated before itself, in case there is a call from the
27415 body of @code{Utils}.
27417 Here is the exact chain of events we are worrying about:
27421 In the body of @code{Decls} a call is made from within the body of a library
27422 task to a subprogram in the package @code{Utils}. Since this call may
27423 occur at elaboration time (given that the task is activated at elaboration
27424 time), we have to assume the worst, i.e., that the
27425 call does happen at elaboration time.
27428 This means that the body and spec of @code{Util} must be elaborated before
27429 the body of @code{Decls} so that this call does not cause an access before
27433 Within the body of @code{Util}, specifically within the body of
27434 @code{Util.Put_Val} there may be calls to any unit @code{with}'ed
27438 One such @code{with}'ed package is package @code{Decls}, so there
27439 might be a call to a subprogram in @code{Decls} in @code{Put_Val}.
27440 In fact there is such a call in this example, but we would have to
27441 assume that there was such a call even if it were not there, since
27442 we are not supposed to write the body of @code{Decls} knowing what
27443 is in the body of @code{Utils}; certainly in the case of the
27444 static elaboration model, the compiler does not know what is in
27445 other bodies and must assume the worst.
27448 This means that the spec and body of @code{Decls} must also be
27449 elaborated before we elaborate the unit containing the call, but
27450 that unit is @code{Decls}! This means that the body of @code{Decls}
27451 must be elaborated before itself, and that's a circularity.
27455 Indeed, if you add an explicit pragma @code{Elaborate_All} for @code{Utils} in
27456 the body of @code{Decls} you will get a true Ada Reference Manual
27457 circularity that makes the program illegal.
27459 In practice, we have found that problems with the static model of
27460 elaboration in existing code often arise from library tasks, so
27461 we must address this particular situation.
27463 Note that if we compile and run the program above, using the dynamic model of
27464 elaboration (that is to say use the @option{-gnatE} switch),
27465 then it compiles, binds,
27466 links, and runs, printing the expected result of 2. Therefore in some sense
27467 the circularity here is only apparent, and we need to capture
27468 the properties of this program that distinguish it from other library-level
27469 tasks that have real elaboration problems.
27471 We have four possible answers to this question:
27476 Use the dynamic model of elaboration.
27478 If we use the @option{-gnatE} switch, then as noted above, the program works.
27479 Why is this? If we examine the task body, it is apparent that the task cannot
27481 @code{accept} statement until after elaboration has been completed, because
27482 the corresponding entry call comes from the main program, not earlier.
27483 This is why the dynamic model works here. But that's really giving
27484 up on a precise analysis, and we prefer to take this approach only if we cannot
27486 problem in any other manner. So let us examine two ways to reorganize
27487 the program to avoid the potential elaboration problem.
27490 Split library tasks into separate packages.
27492 Write separate packages, so that library tasks are isolated from
27493 other declarations as much as possible. Let us look at a variation on
27496 @smallexample @c ada
27504 package body Decls1 is
27505 task body Lib_Task is
27513 type My_Int is new Integer;
27514 function Ident (M : My_Int) return My_Int;
27518 package body Decls2 is
27519 function Ident (M : My_Int) return My_Int is
27527 procedure Put_Val (Arg : Decls2.My_Int);
27531 package body Utils is
27532 procedure Put_Val (Arg : Decls2.My_Int) is
27534 Text_IO.Put_Line (Decls2.My_Int'Image (Decls2.Ident (Arg)));
27541 Decls1.Lib_Task.Start;
27546 All we have done is to split @code{Decls} into two packages, one
27547 containing the library task, and one containing everything else. Now
27548 there is no cycle, and the program compiles, binds, links and executes
27549 using the default static model of elaboration.
27552 Declare separate task types.
27554 A significant part of the problem arises because of the use of the
27555 single task declaration form. This means that the elaboration of
27556 the task type, and the elaboration of the task itself (i.e.@: the
27557 creation of the task) happen at the same time. A good rule
27558 of style in Ada is to always create explicit task types. By
27559 following the additional step of placing task objects in separate
27560 packages from the task type declaration, many elaboration problems
27561 are avoided. Here is another modified example of the example program:
27563 @smallexample @c ada
27565 task type Lib_Task_Type is
27569 type My_Int is new Integer;
27571 function Ident (M : My_Int) return My_Int;
27575 package body Decls is
27576 task body Lib_Task_Type is
27582 function Ident (M : My_Int) return My_Int is
27590 procedure Put_Val (Arg : Decls.My_Int);
27594 package body Utils is
27595 procedure Put_Val (Arg : Decls.My_Int) is
27597 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
27603 Lib_Task : Decls.Lib_Task_Type;
27609 Declst.Lib_Task.Start;
27614 What we have done here is to replace the @code{task} declaration in
27615 package @code{Decls} with a @code{task type} declaration. Then we
27616 introduce a separate package @code{Declst} to contain the actual
27617 task object. This separates the elaboration issues for
27618 the @code{task type}
27619 declaration, which causes no trouble, from the elaboration issues
27620 of the task object, which is also unproblematic, since it is now independent
27621 of the elaboration of @code{Utils}.
27622 This separation of concerns also corresponds to
27623 a generally sound engineering principle of separating declarations
27624 from instances. This version of the program also compiles, binds, links,
27625 and executes, generating the expected output.
27628 Use No_Entry_Calls_In_Elaboration_Code restriction.
27629 @cindex No_Entry_Calls_In_Elaboration_Code
27631 The previous two approaches described how a program can be restructured
27632 to avoid the special problems caused by library task bodies. in practice,
27633 however, such restructuring may be difficult to apply to existing legacy code,
27634 so we must consider solutions that do not require massive rewriting.
27636 Let us consider more carefully why our original sample program works
27637 under the dynamic model of elaboration. The reason is that the code
27638 in the task body blocks immediately on the @code{accept}
27639 statement. Now of course there is nothing to prohibit elaboration
27640 code from making entry calls (for example from another library level task),
27641 so we cannot tell in isolation that
27642 the task will not execute the accept statement during elaboration.
27644 However, in practice it is very unusual to see elaboration code
27645 make any entry calls, and the pattern of tasks starting
27646 at elaboration time and then immediately blocking on @code{accept} or
27647 @code{select} statements is very common. What this means is that
27648 the compiler is being too pessimistic when it analyzes the
27649 whole package body as though it might be executed at elaboration
27652 If we know that the elaboration code contains no entry calls, (a very safe
27653 assumption most of the time, that could almost be made the default
27654 behavior), then we can compile all units of the program under control
27655 of the following configuration pragma:
27658 pragma Restrictions (No_Entry_Calls_In_Elaboration_Code);
27662 This pragma can be placed in the @file{gnat.adc} file in the usual
27663 manner. If we take our original unmodified program and compile it
27664 in the presence of a @file{gnat.adc} containing the above pragma,
27665 then once again, we can compile, bind, link, and execute, obtaining
27666 the expected result. In the presence of this pragma, the compiler does
27667 not trace calls in a task body, that appear after the first @code{accept}
27668 or @code{select} statement, and therefore does not report a potential
27669 circularity in the original program.
27671 The compiler will check to the extent it can that the above
27672 restriction is not violated, but it is not always possible to do a
27673 complete check at compile time, so it is important to use this
27674 pragma only if the stated restriction is in fact met, that is to say
27675 no task receives an entry call before elaboration of all units is completed.
27679 @node Mixing Elaboration Models
27680 @section Mixing Elaboration Models
27682 So far, we have assumed that the entire program is either compiled
27683 using the dynamic model or static model, ensuring consistency. It
27684 is possible to mix the two models, but rules have to be followed
27685 if this mixing is done to ensure that elaboration checks are not
27688 The basic rule is that @emph{a unit compiled with the static model cannot
27689 be @code{with'ed} by a unit compiled with the dynamic model}. The
27690 reason for this is that in the static model, a unit assumes that
27691 its clients guarantee to use (the equivalent of) pragma
27692 @code{Elaborate_All} so that no elaboration checks are required
27693 in inner subprograms, and this assumption is violated if the
27694 client is compiled with dynamic checks.
27696 The precise rule is as follows. A unit that is compiled with dynamic
27697 checks can only @code{with} a unit that meets at least one of the
27698 following criteria:
27703 The @code{with'ed} unit is itself compiled with dynamic elaboration
27704 checks (that is with the @option{-gnatE} switch.
27707 The @code{with'ed} unit is an internal GNAT implementation unit from
27708 the System, Interfaces, Ada, or GNAT hierarchies.
27711 The @code{with'ed} unit has pragma Preelaborate or pragma Pure.
27714 The @code{with'ing} unit (that is the client) has an explicit pragma
27715 @code{Elaborate_All} for the @code{with'ed} unit.
27720 If this rule is violated, that is if a unit with dynamic elaboration
27721 checks @code{with's} a unit that does not meet one of the above four
27722 criteria, then the binder (@code{gnatbind}) will issue a warning
27723 similar to that in the following example:
27726 warning: "x.ads" has dynamic elaboration checks and with's
27727 warning: "y.ads" which has static elaboration checks
27731 These warnings indicate that the rule has been violated, and that as a result
27732 elaboration checks may be missed in the resulting executable file.
27733 This warning may be suppressed using the @option{-ws} binder switch
27734 in the usual manner.
27736 One useful application of this mixing rule is in the case of a subsystem
27737 which does not itself @code{with} units from the remainder of the
27738 application. In this case, the entire subsystem can be compiled with
27739 dynamic checks to resolve a circularity in the subsystem, while
27740 allowing the main application that uses this subsystem to be compiled
27741 using the more reliable default static model.
27743 @node What to Do If the Default Elaboration Behavior Fails
27744 @section What to Do If the Default Elaboration Behavior Fails
27747 If the binder cannot find an acceptable order, it outputs detailed
27748 diagnostics. For example:
27754 error: elaboration circularity detected
27755 info: "proc (body)" must be elaborated before "pack (body)"
27756 info: reason: Elaborate_All probably needed in unit "pack (body)"
27757 info: recompile "pack (body)" with -gnatwl
27758 info: for full details
27759 info: "proc (body)"
27760 info: is needed by its spec:
27761 info: "proc (spec)"
27762 info: which is withed by:
27763 info: "pack (body)"
27764 info: "pack (body)" must be elaborated before "proc (body)"
27765 info: reason: pragma Elaborate in unit "proc (body)"
27771 In this case we have a cycle that the binder cannot break. On the one
27772 hand, there is an explicit pragma Elaborate in @code{proc} for
27773 @code{pack}. This means that the body of @code{pack} must be elaborated
27774 before the body of @code{proc}. On the other hand, there is elaboration
27775 code in @code{pack} that calls a subprogram in @code{proc}. This means
27776 that for maximum safety, there should really be a pragma
27777 Elaborate_All in @code{pack} for @code{proc} which would require that
27778 the body of @code{proc} be elaborated before the body of
27779 @code{pack}. Clearly both requirements cannot be satisfied.
27780 Faced with a circularity of this kind, you have three different options.
27783 @item Fix the program
27784 The most desirable option from the point of view of long-term maintenance
27785 is to rearrange the program so that the elaboration problems are avoided.
27786 One useful technique is to place the elaboration code into separate
27787 child packages. Another is to move some of the initialization code to
27788 explicitly called subprograms, where the program controls the order
27789 of initialization explicitly. Although this is the most desirable option,
27790 it may be impractical and involve too much modification, especially in
27791 the case of complex legacy code.
27793 @item Perform dynamic checks
27794 If the compilations are done using the
27796 (dynamic elaboration check) switch, then GNAT behaves in a quite different
27797 manner. Dynamic checks are generated for all calls that could possibly result
27798 in raising an exception. With this switch, the compiler does not generate
27799 implicit @code{Elaborate} or @code{Elaborate_All} pragmas. The behavior then is
27800 exactly as specified in the @cite{Ada Reference Manual}.
27801 The binder will generate
27802 an executable program that may or may not raise @code{Program_Error}, and then
27803 it is the programmer's job to ensure that it does not raise an exception. Note
27804 that it is important to compile all units with the switch, it cannot be used
27807 @item Suppress checks
27808 The drawback of dynamic checks is that they generate a
27809 significant overhead at run time, both in space and time. If you
27810 are absolutely sure that your program cannot raise any elaboration
27811 exceptions, and you still want to use the dynamic elaboration model,
27812 then you can use the configuration pragma
27813 @code{Suppress (Elaboration_Check)} to suppress all such checks. For
27814 example this pragma could be placed in the @file{gnat.adc} file.
27816 @item Suppress checks selectively
27817 When you know that certain calls or instantiations in elaboration code cannot
27818 possibly lead to an elaboration error, and the binder nevertheless complains
27819 about implicit @code{Elaborate} and @code{Elaborate_All} pragmas that lead to
27820 elaboration circularities, it is possible to remove those warnings locally and
27821 obtain a program that will bind. Clearly this can be unsafe, and it is the
27822 responsibility of the programmer to make sure that the resulting program has no
27823 elaboration anomalies. The pragma @code{Suppress (Elaboration_Check)} can be
27824 used with different granularity to suppress warnings and break elaboration
27829 Place the pragma that names the called subprogram in the declarative part
27830 that contains the call.
27833 Place the pragma in the declarative part, without naming an entity. This
27834 disables warnings on all calls in the corresponding declarative region.
27837 Place the pragma in the package spec that declares the called subprogram,
27838 and name the subprogram. This disables warnings on all elaboration calls to
27842 Place the pragma in the package spec that declares the called subprogram,
27843 without naming any entity. This disables warnings on all elaboration calls to
27844 all subprograms declared in this spec.
27846 @item Use Pragma Elaborate
27847 As previously described in section @xref{Treatment of Pragma Elaborate},
27848 GNAT in static mode assumes that a @code{pragma} Elaborate indicates correctly
27849 that no elaboration checks are required on calls to the designated unit.
27850 There may be cases in which the caller knows that no transitive calls
27851 can occur, so that a @code{pragma Elaborate} will be sufficient in a
27852 case where @code{pragma Elaborate_All} would cause a circularity.
27856 These five cases are listed in order of decreasing safety, and therefore
27857 require increasing programmer care in their application. Consider the
27860 @smallexample @c adanocomment
27862 function F1 return Integer;
27867 function F2 return Integer;
27868 function Pure (x : integer) return integer;
27869 -- pragma Suppress (Elaboration_Check, On => Pure); -- (3)
27870 -- pragma Suppress (Elaboration_Check); -- (4)
27874 package body Pack1 is
27875 function F1 return Integer is
27879 Val : integer := Pack2.Pure (11); -- Elab. call (1)
27882 -- pragma Suppress(Elaboration_Check, Pack2.F2); -- (1)
27883 -- pragma Suppress(Elaboration_Check); -- (2)
27885 X1 := Pack2.F2 + 1; -- Elab. call (2)
27890 package body Pack2 is
27891 function F2 return Integer is
27895 function Pure (x : integer) return integer is
27897 return x ** 3 - 3 * x;
27901 with Pack1, Ada.Text_IO;
27904 Ada.Text_IO.Put_Line(Pack1.X1'Img); -- 101
27907 In the absence of any pragmas, an attempt to bind this program produces
27908 the following diagnostics:
27914 error: elaboration circularity detected
27915 info: "pack1 (body)" must be elaborated before "pack1 (body)"
27916 info: reason: Elaborate_All probably needed in unit "pack1 (body)"
27917 info: recompile "pack1 (body)" with -gnatwl for full details
27918 info: "pack1 (body)"
27919 info: must be elaborated along with its spec:
27920 info: "pack1 (spec)"
27921 info: which is withed by:
27922 info: "pack2 (body)"
27923 info: which must be elaborated along with its spec:
27924 info: "pack2 (spec)"
27925 info: which is withed by:
27926 info: "pack1 (body)"
27929 The sources of the circularity are the two calls to @code{Pack2.Pure} and
27930 @code{Pack2.F2} in the body of @code{Pack1}. We can see that the call to
27931 F2 is safe, even though F2 calls F1, because the call appears after the
27932 elaboration of the body of F1. Therefore the pragma (1) is safe, and will
27933 remove the warning on the call. It is also possible to use pragma (2)
27934 because there are no other potentially unsafe calls in the block.
27937 The call to @code{Pure} is safe because this function does not depend on the
27938 state of @code{Pack2}. Therefore any call to this function is safe, and it
27939 is correct to place pragma (3) in the corresponding package spec.
27942 Finally, we could place pragma (4) in the spec of @code{Pack2} to disable
27943 warnings on all calls to functions declared therein. Note that this is not
27944 necessarily safe, and requires more detailed examination of the subprogram
27945 bodies involved. In particular, a call to @code{F2} requires that @code{F1}
27946 be already elaborated.
27950 It is hard to generalize on which of these four approaches should be
27951 taken. Obviously if it is possible to fix the program so that the default
27952 treatment works, this is preferable, but this may not always be practical.
27953 It is certainly simple enough to use
27955 but the danger in this case is that, even if the GNAT binder
27956 finds a correct elaboration order, it may not always do so,
27957 and certainly a binder from another Ada compiler might not. A
27958 combination of testing and analysis (for which the warnings generated
27961 switch can be useful) must be used to ensure that the program is free
27962 of errors. One switch that is useful in this testing is the
27963 @option{^-p (pessimistic elaboration order)^/PESSIMISTIC_ELABORATION_ORDER^}
27966 Normally the binder tries to find an order that has the best chance
27967 of avoiding elaboration problems. However, if this switch is used, the binder
27968 plays a devil's advocate role, and tries to choose the order that
27969 has the best chance of failing. If your program works even with this
27970 switch, then it has a better chance of being error free, but this is still
27973 For an example of this approach in action, consider the C-tests (executable
27974 tests) from the ACVC suite. If these are compiled and run with the default
27975 treatment, then all but one of them succeed without generating any error
27976 diagnostics from the binder. However, there is one test that fails, and
27977 this is not surprising, because the whole point of this test is to ensure
27978 that the compiler can handle cases where it is impossible to determine
27979 a correct order statically, and it checks that an exception is indeed
27980 raised at run time.
27982 This one test must be compiled and run using the
27984 switch, and then it passes. Alternatively, the entire suite can
27985 be run using this switch. It is never wrong to run with the dynamic
27986 elaboration switch if your code is correct, and we assume that the
27987 C-tests are indeed correct (it is less efficient, but efficiency is
27988 not a factor in running the ACVC tests.)
27990 @node Elaboration for Access-to-Subprogram Values
27991 @section Elaboration for Access-to-Subprogram Values
27992 @cindex Access-to-subprogram
27995 Access-to-subprogram types (introduced in Ada 95) complicate
27996 the handling of elaboration. The trouble is that it becomes
27997 impossible to tell at compile time which procedure
27998 is being called. This means that it is not possible for the binder
27999 to analyze the elaboration requirements in this case.
28001 If at the point at which the access value is created
28002 (i.e., the evaluation of @code{P'Access} for a subprogram @code{P}),
28003 the body of the subprogram is
28004 known to have been elaborated, then the access value is safe, and its use
28005 does not require a check. This may be achieved by appropriate arrangement
28006 of the order of declarations if the subprogram is in the current unit,
28007 or, if the subprogram is in another unit, by using pragma
28008 @code{Pure}, @code{Preelaborate}, or @code{Elaborate_Body}
28009 on the referenced unit.
28011 If the referenced body is not known to have been elaborated at the point
28012 the access value is created, then any use of the access value must do a
28013 dynamic check, and this dynamic check will fail and raise a
28014 @code{Program_Error} exception if the body has not been elaborated yet.
28015 GNAT will generate the necessary checks, and in addition, if the
28017 switch is set, will generate warnings that such checks are required.
28019 The use of dynamic dispatching for tagged types similarly generates
28020 a requirement for dynamic checks, and premature calls to any primitive
28021 operation of a tagged type before the body of the operation has been
28022 elaborated, will result in the raising of @code{Program_Error}.
28024 @node Summary of Procedures for Elaboration Control
28025 @section Summary of Procedures for Elaboration Control
28026 @cindex Elaboration control
28029 First, compile your program with the default options, using none of
28030 the special elaboration control switches. If the binder successfully
28031 binds your program, then you can be confident that, apart from issues
28032 raised by the use of access-to-subprogram types and dynamic dispatching,
28033 the program is free of elaboration errors. If it is important that the
28034 program be portable, then use the
28036 switch to generate warnings about missing @code{Elaborate} or
28037 @code{Elaborate_All} pragmas, and supply the missing pragmas.
28039 If the program fails to bind using the default static elaboration
28040 handling, then you can fix the program to eliminate the binder
28041 message, or recompile the entire program with the
28042 @option{-gnatE} switch to generate dynamic elaboration checks,
28043 and, if you are sure there really are no elaboration problems,
28044 use a global pragma @code{Suppress (Elaboration_Check)}.
28046 @node Other Elaboration Order Considerations
28047 @section Other Elaboration Order Considerations
28049 This section has been entirely concerned with the issue of finding a valid
28050 elaboration order, as defined by the Ada Reference Manual. In a case
28051 where several elaboration orders are valid, the task is to find one
28052 of the possible valid elaboration orders (and the static model in GNAT
28053 will ensure that this is achieved).
28055 The purpose of the elaboration rules in the Ada Reference Manual is to
28056 make sure that no entity is accessed before it has been elaborated. For
28057 a subprogram, this means that the spec and body must have been elaborated
28058 before the subprogram is called. For an object, this means that the object
28059 must have been elaborated before its value is read or written. A violation
28060 of either of these two requirements is an access before elaboration order,
28061 and this section has been all about avoiding such errors.
28063 In the case where more than one order of elaboration is possible, in the
28064 sense that access before elaboration errors are avoided, then any one of
28065 the orders is ``correct'' in the sense that it meets the requirements of
28066 the Ada Reference Manual, and no such error occurs.
28068 However, it may be the case for a given program, that there are
28069 constraints on the order of elaboration that come not from consideration
28070 of avoiding elaboration errors, but rather from extra-lingual logic
28071 requirements. Consider this example:
28073 @smallexample @c ada
28074 with Init_Constants;
28075 package Constants is
28080 package Init_Constants is
28081 procedure P; -- require a body
28082 end Init_Constants;
28085 package body Init_Constants is
28086 procedure P is begin null; end;
28090 end Init_Constants;
28094 Z : Integer := Constants.X + Constants.Y;
28098 with Text_IO; use Text_IO;
28101 Put_Line (Calc.Z'Img);
28106 In this example, there is more than one valid order of elaboration. For
28107 example both the following are correct orders:
28110 Init_Constants spec
28113 Init_Constants body
28118 Init_Constants spec
28119 Init_Constants body
28126 There is no language rule to prefer one or the other, both are correct
28127 from an order of elaboration point of view. But the programmatic effects
28128 of the two orders are very different. In the first, the elaboration routine
28129 of @code{Calc} initializes @code{Z} to zero, and then the main program
28130 runs with this value of zero. But in the second order, the elaboration
28131 routine of @code{Calc} runs after the body of Init_Constants has set
28132 @code{X} and @code{Y} and thus @code{Z} is set to 7 before @code{Main}
28135 One could perhaps by applying pretty clever non-artificial intelligence
28136 to the situation guess that it is more likely that the second order of
28137 elaboration is the one desired, but there is no formal linguistic reason
28138 to prefer one over the other. In fact in this particular case, GNAT will
28139 prefer the second order, because of the rule that bodies are elaborated
28140 as soon as possible, but it's just luck that this is what was wanted
28141 (if indeed the second order was preferred).
28143 If the program cares about the order of elaboration routines in a case like
28144 this, it is important to specify the order required. In this particular
28145 case, that could have been achieved by adding to the spec of Calc:
28147 @smallexample @c ada
28148 pragma Elaborate_All (Constants);
28152 which requires that the body (if any) and spec of @code{Constants},
28153 as well as the body and spec of any unit @code{with}'ed by
28154 @code{Constants} be elaborated before @code{Calc} is elaborated.
28156 Clearly no automatic method can always guess which alternative you require,
28157 and if you are working with legacy code that had constraints of this kind
28158 which were not properly specified by adding @code{Elaborate} or
28159 @code{Elaborate_All} pragmas, then indeed it is possible that two different
28160 compilers can choose different orders.
28162 However, GNAT does attempt to diagnose the common situation where there
28163 are uninitialized variables in the visible part of a package spec, and the
28164 corresponding package body has an elaboration block that directly or
28165 indirectly initialized one or more of these variables. This is the situation
28166 in which a pragma Elaborate_Body is usually desirable, and GNAT will generate
28167 a warning that suggests this addition if it detects this situation.
28169 The @code{gnatbind}
28170 @option{^-p^/PESSIMISTIC_ELABORATION^} switch may be useful in smoking
28171 out problems. This switch causes bodies to be elaborated as late as possible
28172 instead of as early as possible. In the example above, it would have forced
28173 the choice of the first elaboration order. If you get different results
28174 when using this switch, and particularly if one set of results is right,
28175 and one is wrong as far as you are concerned, it shows that you have some
28176 missing @code{Elaborate} pragmas. For the example above, we have the
28180 gnatmake -f -q main
28183 gnatmake -f -q main -bargs -p
28189 It is of course quite unlikely that both these results are correct, so
28190 it is up to you in a case like this to investigate the source of the
28191 difference, by looking at the two elaboration orders that are chosen,
28192 and figuring out which is correct, and then adding the necessary
28193 @code{Elaborate} or @code{Elaborate_All} pragmas to ensure the desired order.
28197 @c *******************************
28198 @node Conditional Compilation
28199 @appendix Conditional Compilation
28200 @c *******************************
28201 @cindex Conditional compilation
28204 It is often necessary to arrange for a single source program
28205 to serve multiple purposes, where it is compiled in different
28206 ways to achieve these different goals. Some examples of the
28207 need for this feature are
28210 @item Adapting a program to a different hardware environment
28211 @item Adapting a program to a different target architecture
28212 @item Turning debugging features on and off
28213 @item Arranging for a program to compile with different compilers
28217 In C, or C++, the typical approach would be to use the preprocessor
28218 that is defined as part of the language. The Ada language does not
28219 contain such a feature. This is not an oversight, but rather a very
28220 deliberate design decision, based on the experience that overuse of
28221 the preprocessing features in C and C++ can result in programs that
28222 are extremely difficult to maintain. For example, if we have ten
28223 switches that can be on or off, this means that there are a thousand
28224 separate programs, any one of which might not even be syntactically
28225 correct, and even if syntactically correct, the resulting program
28226 might not work correctly. Testing all combinations can quickly become
28229 Nevertheless, the need to tailor programs certainly exists, and in
28230 this Appendix we will discuss how this can
28231 be achieved using Ada in general, and GNAT in particular.
28234 * Use of Boolean Constants::
28235 * Debugging - A Special Case::
28236 * Conditionalizing Declarations::
28237 * Use of Alternative Implementations::
28241 @node Use of Boolean Constants
28242 @section Use of Boolean Constants
28245 In the case where the difference is simply which code
28246 sequence is executed, the cleanest solution is to use Boolean
28247 constants to control which code is executed.
28249 @smallexample @c ada
28251 FP_Initialize_Required : constant Boolean := True;
28253 if FP_Initialize_Required then
28260 Not only will the code inside the @code{if} statement not be executed if
28261 the constant Boolean is @code{False}, but it will also be completely
28262 deleted from the program.
28263 However, the code is only deleted after the @code{if} statement
28264 has been checked for syntactic and semantic correctness.
28265 (In contrast, with preprocessors the code is deleted before the
28266 compiler ever gets to see it, so it is not checked until the switch
28268 @cindex Preprocessors (contrasted with conditional compilation)
28270 Typically the Boolean constants will be in a separate package,
28273 @smallexample @c ada
28276 FP_Initialize_Required : constant Boolean := True;
28277 Reset_Available : constant Boolean := False;
28284 The @code{Config} package exists in multiple forms for the various targets,
28285 with an appropriate script selecting the version of @code{Config} needed.
28286 Then any other unit requiring conditional compilation can do a @code{with}
28287 of @code{Config} to make the constants visible.
28290 @node Debugging - A Special Case
28291 @section Debugging - A Special Case
28294 A common use of conditional code is to execute statements (for example
28295 dynamic checks, or output of intermediate results) under control of a
28296 debug switch, so that the debugging behavior can be turned on and off.
28297 This can be done using a Boolean constant to control whether the code
28300 @smallexample @c ada
28303 Put_Line ("got to the first stage!");
28311 @smallexample @c ada
28313 if Debugging and then Temperature > 999.0 then
28314 raise Temperature_Crazy;
28320 Since this is a common case, there are special features to deal with
28321 this in a convenient manner. For the case of tests, Ada 2005 has added
28322 a pragma @code{Assert} that can be used for such tests. This pragma is modeled
28323 @cindex pragma @code{Assert}
28324 on the @code{Assert} pragma that has always been available in GNAT, so this
28325 feature may be used with GNAT even if you are not using Ada 2005 features.
28326 The use of pragma @code{Assert} is described in
28327 @ref{Pragma Assert,,, gnat_rm, GNAT Reference Manual}, but as an
28328 example, the last test could be written:
28330 @smallexample @c ada
28331 pragma Assert (Temperature <= 999.0, "Temperature Crazy");
28337 @smallexample @c ada
28338 pragma Assert (Temperature <= 999.0);
28342 In both cases, if assertions are active and the temperature is excessive,
28343 the exception @code{Assert_Failure} will be raised, with the given string in
28344 the first case or a string indicating the location of the pragma in the second
28345 case used as the exception message.
28347 You can turn assertions on and off by using the @code{Assertion_Policy}
28349 @cindex pragma @code{Assertion_Policy}
28350 This is an Ada 2005 pragma which is implemented in all modes by
28351 GNAT, but only in the latest versions of GNAT which include Ada 2005
28352 capability. Alternatively, you can use the @option{-gnata} switch
28353 @cindex @option{-gnata} switch
28354 to enable assertions from the command line (this is recognized by all versions
28357 For the example above with the @code{Put_Line}, the GNAT-specific pragma
28358 @code{Debug} can be used:
28359 @cindex pragma @code{Debug}
28361 @smallexample @c ada
28362 pragma Debug (Put_Line ("got to the first stage!"));
28366 If debug pragmas are enabled, the argument, which must be of the form of
28367 a procedure call, is executed (in this case, @code{Put_Line} will be called).
28368 Only one call can be present, but of course a special debugging procedure
28369 containing any code you like can be included in the program and then
28370 called in a pragma @code{Debug} argument as needed.
28372 One advantage of pragma @code{Debug} over the @code{if Debugging then}
28373 construct is that pragma @code{Debug} can appear in declarative contexts,
28374 such as at the very beginning of a procedure, before local declarations have
28377 Debug pragmas are enabled using either the @option{-gnata} switch that also
28378 controls assertions, or with a separate Debug_Policy pragma.
28379 @cindex pragma @code{Debug_Policy}
28380 The latter pragma is new in the Ada 2005 versions of GNAT (but it can be used
28381 in Ada 95 and Ada 83 programs as well), and is analogous to
28382 pragma @code{Assertion_Policy} to control assertions.
28384 @code{Assertion_Policy} and @code{Debug_Policy} are configuration pragmas,
28385 and thus they can appear in @file{gnat.adc} if you are not using a
28386 project file, or in the file designated to contain configuration pragmas
28388 They then apply to all subsequent compilations. In practice the use of
28389 the @option{-gnata} switch is often the most convenient method of controlling
28390 the status of these pragmas.
28392 Note that a pragma is not a statement, so in contexts where a statement
28393 sequence is required, you can't just write a pragma on its own. You have
28394 to add a @code{null} statement.
28396 @smallexample @c ada
28399 @dots{} -- some statements
28401 pragma Assert (Num_Cases < 10);
28408 @node Conditionalizing Declarations
28409 @section Conditionalizing Declarations
28412 In some cases, it may be necessary to conditionalize declarations to meet
28413 different requirements. For example we might want a bit string whose length
28414 is set to meet some hardware message requirement.
28416 In some cases, it may be possible to do this using declare blocks controlled
28417 by conditional constants:
28419 @smallexample @c ada
28421 if Small_Machine then
28423 X : Bit_String (1 .. 10);
28429 X : Large_Bit_String (1 .. 1000);
28438 Note that in this approach, both declarations are analyzed by the
28439 compiler so this can only be used where both declarations are legal,
28440 even though one of them will not be used.
28442 Another approach is to define integer constants, e.g.@: @code{Bits_Per_Word}, or
28443 Boolean constants, e.g.@: @code{Little_Endian}, and then write declarations
28444 that are parameterized by these constants. For example
28446 @smallexample @c ada
28449 Field1 at 0 range Boolean'Pos (Little_Endian) * 10 .. Bits_Per_Word;
28455 If @code{Bits_Per_Word} is set to 32, this generates either
28457 @smallexample @c ada
28460 Field1 at 0 range 0 .. 32;
28466 for the big endian case, or
28468 @smallexample @c ada
28471 Field1 at 0 range 10 .. 32;
28477 for the little endian case. Since a powerful subset of Ada expression
28478 notation is usable for creating static constants, clever use of this
28479 feature can often solve quite difficult problems in conditionalizing
28480 compilation (note incidentally that in Ada 95, the little endian
28481 constant was introduced as @code{System.Default_Bit_Order}, so you do not
28482 need to define this one yourself).
28485 @node Use of Alternative Implementations
28486 @section Use of Alternative Implementations
28489 In some cases, none of the approaches described above are adequate. This
28490 can occur for example if the set of declarations required is radically
28491 different for two different configurations.
28493 In this situation, the official Ada way of dealing with conditionalizing
28494 such code is to write separate units for the different cases. As long as
28495 this does not result in excessive duplication of code, this can be done
28496 without creating maintenance problems. The approach is to share common
28497 code as far as possible, and then isolate the code and declarations
28498 that are different. Subunits are often a convenient method for breaking
28499 out a piece of a unit that is to be conditionalized, with separate files
28500 for different versions of the subunit for different targets, where the
28501 build script selects the right one to give to the compiler.
28502 @cindex Subunits (and conditional compilation)
28504 As an example, consider a situation where a new feature in Ada 2005
28505 allows something to be done in a really nice way. But your code must be able
28506 to compile with an Ada 95 compiler. Conceptually you want to say:
28508 @smallexample @c ada
28511 @dots{} neat Ada 2005 code
28513 @dots{} not quite as neat Ada 95 code
28519 where @code{Ada_2005} is a Boolean constant.
28521 But this won't work when @code{Ada_2005} is set to @code{False},
28522 since the @code{then} clause will be illegal for an Ada 95 compiler.
28523 (Recall that although such unreachable code would eventually be deleted
28524 by the compiler, it still needs to be legal. If it uses features
28525 introduced in Ada 2005, it will be illegal in Ada 95.)
28527 So instead we write
28529 @smallexample @c ada
28530 procedure Insert is separate;
28534 Then we have two files for the subunit @code{Insert}, with the two sets of
28536 If the package containing this is called @code{File_Queries}, then we might
28540 @item @file{file_queries-insert-2005.adb}
28541 @item @file{file_queries-insert-95.adb}
28545 and the build script renames the appropriate file to
28548 file_queries-insert.adb
28552 and then carries out the compilation.
28554 This can also be done with project files' naming schemes. For example:
28556 @smallexample @c project
28557 For Body ("File_Queries.Insert") use "file_queries-insert-2005.ada";
28561 Note also that with project files it is desirable to use a different extension
28562 than @file{ads} / @file{adb} for alternative versions. Otherwise a naming
28563 conflict may arise through another commonly used feature: to declare as part
28564 of the project a set of directories containing all the sources obeying the
28565 default naming scheme.
28567 The use of alternative units is certainly feasible in all situations,
28568 and for example the Ada part of the GNAT run-time is conditionalized
28569 based on the target architecture using this approach. As a specific example,
28570 consider the implementation of the AST feature in VMS. There is one
28578 which is the same for all architectures, and three bodies:
28582 used for all non-VMS operating systems
28583 @item s-asthan-vms-alpha.adb
28584 used for VMS on the Alpha
28585 @item s-asthan-vms-ia64.adb
28586 used for VMS on the ia64
28590 The dummy version @file{s-asthan.adb} simply raises exceptions noting that
28591 this operating system feature is not available, and the two remaining
28592 versions interface with the corresponding versions of VMS to provide
28593 VMS-compatible AST handling. The GNAT build script knows the architecture
28594 and operating system, and automatically selects the right version,
28595 renaming it if necessary to @file{s-asthan.adb} before the run-time build.
28597 Another style for arranging alternative implementations is through Ada's
28598 access-to-subprogram facility.
28599 In case some functionality is to be conditionally included,
28600 you can declare an access-to-procedure variable @code{Ref} that is initialized
28601 to designate a ``do nothing'' procedure, and then invoke @code{Ref.all}
28603 In some library package, set @code{Ref} to @code{Proc'Access} for some
28604 procedure @code{Proc} that performs the relevant processing.
28605 The initialization only occurs if the library package is included in the
28607 The same idea can also be implemented using tagged types and dispatching
28611 @node Preprocessing
28612 @section Preprocessing
28613 @cindex Preprocessing
28616 Although it is quite possible to conditionalize code without the use of
28617 C-style preprocessing, as described earlier in this section, it is
28618 nevertheless convenient in some cases to use the C approach. Moreover,
28619 older Ada compilers have often provided some preprocessing capability,
28620 so legacy code may depend on this approach, even though it is not
28623 To accommodate such use, GNAT provides a preprocessor (modeled to a large
28624 extent on the various preprocessors that have been used
28625 with legacy code on other compilers, to enable easier transition).
28627 The preprocessor may be used in two separate modes. It can be used quite
28628 separately from the compiler, to generate a separate output source file
28629 that is then fed to the compiler as a separate step. This is the
28630 @code{gnatprep} utility, whose use is fully described in
28631 @ref{Preprocessing Using gnatprep}.
28632 @cindex @code{gnatprep}
28634 The preprocessing language allows such constructs as
28638 #if DEBUG or PRIORITY > 4 then
28639 bunch of declarations
28641 completely different bunch of declarations
28647 The values of the symbols @code{DEBUG} and @code{PRIORITY} can be
28648 defined either on the command line or in a separate file.
28650 The other way of running the preprocessor is even closer to the C style and
28651 often more convenient. In this approach the preprocessing is integrated into
28652 the compilation process. The compiler is fed the preprocessor input which
28653 includes @code{#if} lines etc, and then the compiler carries out the
28654 preprocessing internally and processes the resulting output.
28655 For more details on this approach, see @ref{Integrated Preprocessing}.
28658 @c *******************************
28659 @node Inline Assembler
28660 @appendix Inline Assembler
28661 @c *******************************
28664 If you need to write low-level software that interacts directly
28665 with the hardware, Ada provides two ways to incorporate assembly
28666 language code into your program. First, you can import and invoke
28667 external routines written in assembly language, an Ada feature fully
28668 supported by GNAT@. However, for small sections of code it may be simpler
28669 or more efficient to include assembly language statements directly
28670 in your Ada source program, using the facilities of the implementation-defined
28671 package @code{System.Machine_Code}, which incorporates the gcc
28672 Inline Assembler. The Inline Assembler approach offers a number of advantages,
28673 including the following:
28676 @item No need to use non-Ada tools
28677 @item Consistent interface over different targets
28678 @item Automatic usage of the proper calling conventions
28679 @item Access to Ada constants and variables
28680 @item Definition of intrinsic routines
28681 @item Possibility of inlining a subprogram comprising assembler code
28682 @item Code optimizer can take Inline Assembler code into account
28685 This chapter presents a series of examples to show you how to use
28686 the Inline Assembler. Although it focuses on the Intel x86,
28687 the general approach applies also to other processors.
28688 It is assumed that you are familiar with Ada
28689 and with assembly language programming.
28692 * Basic Assembler Syntax::
28693 * A Simple Example of Inline Assembler::
28694 * Output Variables in Inline Assembler::
28695 * Input Variables in Inline Assembler::
28696 * Inlining Inline Assembler Code::
28697 * Other Asm Functionality::
28700 @c ---------------------------------------------------------------------------
28701 @node Basic Assembler Syntax
28702 @section Basic Assembler Syntax
28705 The assembler used by GNAT and gcc is based not on the Intel assembly
28706 language, but rather on a language that descends from the AT&T Unix
28707 assembler @emph{as} (and which is often referred to as ``AT&T syntax'').
28708 The following table summarizes the main features of @emph{as} syntax
28709 and points out the differences from the Intel conventions.
28710 See the gcc @emph{as} and @emph{gas} (an @emph{as} macro
28711 pre-processor) documentation for further information.
28714 @item Register names
28715 gcc / @emph{as}: Prefix with ``%''; for example @code{%eax}
28717 Intel: No extra punctuation; for example @code{eax}
28719 @item Immediate operand
28720 gcc / @emph{as}: Prefix with ``$''; for example @code{$4}
28722 Intel: No extra punctuation; for example @code{4}
28725 gcc / @emph{as}: Prefix with ``$''; for example @code{$loc}
28727 Intel: No extra punctuation; for example @code{loc}
28729 @item Memory contents
28730 gcc / @emph{as}: No extra punctuation; for example @code{loc}
28732 Intel: Square brackets; for example @code{[loc]}
28734 @item Register contents
28735 gcc / @emph{as}: Parentheses; for example @code{(%eax)}
28737 Intel: Square brackets; for example @code{[eax]}
28739 @item Hexadecimal numbers
28740 gcc / @emph{as}: Leading ``0x'' (C language syntax); for example @code{0xA0}
28742 Intel: Trailing ``h''; for example @code{A0h}
28745 gcc / @emph{as}: Explicit in op code; for example @code{movw} to move
28748 Intel: Implicit, deduced by assembler; for example @code{mov}
28750 @item Instruction repetition
28751 gcc / @emph{as}: Split into two lines; for example
28757 Intel: Keep on one line; for example @code{rep stosl}
28759 @item Order of operands
28760 gcc / @emph{as}: Source first; for example @code{movw $4, %eax}
28762 Intel: Destination first; for example @code{mov eax, 4}
28765 @c ---------------------------------------------------------------------------
28766 @node A Simple Example of Inline Assembler
28767 @section A Simple Example of Inline Assembler
28770 The following example will generate a single assembly language statement,
28771 @code{nop}, which does nothing. Despite its lack of run-time effect,
28772 the example will be useful in illustrating the basics of
28773 the Inline Assembler facility.
28775 @smallexample @c ada
28777 with System.Machine_Code; use System.Machine_Code;
28778 procedure Nothing is
28785 @code{Asm} is a procedure declared in package @code{System.Machine_Code};
28786 here it takes one parameter, a @emph{template string} that must be a static
28787 expression and that will form the generated instruction.
28788 @code{Asm} may be regarded as a compile-time procedure that parses
28789 the template string and additional parameters (none here),
28790 from which it generates a sequence of assembly language instructions.
28792 The examples in this chapter will illustrate several of the forms
28793 for invoking @code{Asm}; a complete specification of the syntax
28794 is found in @ref{Machine Code Insertions,,, gnat_rm, GNAT Reference
28797 Under the standard GNAT conventions, the @code{Nothing} procedure
28798 should be in a file named @file{nothing.adb}.
28799 You can build the executable in the usual way:
28803 However, the interesting aspect of this example is not its run-time behavior
28804 but rather the generated assembly code.
28805 To see this output, invoke the compiler as follows:
28807 gcc -c -S -fomit-frame-pointer -gnatp @file{nothing.adb}
28809 where the options are:
28813 compile only (no bind or link)
28815 generate assembler listing
28816 @item -fomit-frame-pointer
28817 do not set up separate stack frames
28819 do not add runtime checks
28822 This gives a human-readable assembler version of the code. The resulting
28823 file will have the same name as the Ada source file, but with a @code{.s}
28824 extension. In our example, the file @file{nothing.s} has the following
28829 .file "nothing.adb"
28831 ___gnu_compiled_ada:
28834 .globl __ada_nothing
28846 The assembly code you included is clearly indicated by
28847 the compiler, between the @code{#APP} and @code{#NO_APP}
28848 delimiters. The character before the 'APP' and 'NOAPP'
28849 can differ on different targets. For example, GNU/Linux uses '#APP' while
28850 on NT you will see '/APP'.
28852 If you make a mistake in your assembler code (such as using the
28853 wrong size modifier, or using a wrong operand for the instruction) GNAT
28854 will report this error in a temporary file, which will be deleted when
28855 the compilation is finished. Generating an assembler file will help
28856 in such cases, since you can assemble this file separately using the
28857 @emph{as} assembler that comes with gcc.
28859 Assembling the file using the command
28862 as @file{nothing.s}
28865 will give you error messages whose lines correspond to the assembler
28866 input file, so you can easily find and correct any mistakes you made.
28867 If there are no errors, @emph{as} will generate an object file
28868 @file{nothing.out}.
28870 @c ---------------------------------------------------------------------------
28871 @node Output Variables in Inline Assembler
28872 @section Output Variables in Inline Assembler
28875 The examples in this section, showing how to access the processor flags,
28876 illustrate how to specify the destination operands for assembly language
28879 @smallexample @c ada
28881 with Interfaces; use Interfaces;
28882 with Ada.Text_IO; use Ada.Text_IO;
28883 with System.Machine_Code; use System.Machine_Code;
28884 procedure Get_Flags is
28885 Flags : Unsigned_32;
28888 Asm ("pushfl" & LF & HT & -- push flags on stack
28889 "popl %%eax" & LF & HT & -- load eax with flags
28890 "movl %%eax, %0", -- store flags in variable
28891 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28892 Put_Line ("Flags register:" & Flags'Img);
28897 In order to have a nicely aligned assembly listing, we have separated
28898 multiple assembler statements in the Asm template string with linefeed
28899 (ASCII.LF) and horizontal tab (ASCII.HT) characters.
28900 The resulting section of the assembly output file is:
28907 movl %eax, -40(%ebp)
28912 It would have been legal to write the Asm invocation as:
28915 Asm ("pushfl popl %%eax movl %%eax, %0")
28918 but in the generated assembler file, this would come out as:
28922 pushfl popl %eax movl %eax, -40(%ebp)
28926 which is not so convenient for the human reader.
28928 We use Ada comments
28929 at the end of each line to explain what the assembler instructions
28930 actually do. This is a useful convention.
28932 When writing Inline Assembler instructions, you need to precede each register
28933 and variable name with a percent sign. Since the assembler already requires
28934 a percent sign at the beginning of a register name, you need two consecutive
28935 percent signs for such names in the Asm template string, thus @code{%%eax}.
28936 In the generated assembly code, one of the percent signs will be stripped off.
28938 Names such as @code{%0}, @code{%1}, @code{%2}, etc., denote input or output
28939 variables: operands you later define using @code{Input} or @code{Output}
28940 parameters to @code{Asm}.
28941 An output variable is illustrated in
28942 the third statement in the Asm template string:
28946 The intent is to store the contents of the eax register in a variable that can
28947 be accessed in Ada. Simply writing @code{movl %%eax, Flags} would not
28948 necessarily work, since the compiler might optimize by using a register
28949 to hold Flags, and the expansion of the @code{movl} instruction would not be
28950 aware of this optimization. The solution is not to store the result directly
28951 but rather to advise the compiler to choose the correct operand form;
28952 that is the purpose of the @code{%0} output variable.
28954 Information about the output variable is supplied in the @code{Outputs}
28955 parameter to @code{Asm}:
28957 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28960 The output is defined by the @code{Asm_Output} attribute of the target type;
28961 the general format is
28963 Type'Asm_Output (constraint_string, variable_name)
28966 The constraint string directs the compiler how
28967 to store/access the associated variable. In the example
28969 Unsigned_32'Asm_Output ("=m", Flags);
28971 the @code{"m"} (memory) constraint tells the compiler that the variable
28972 @code{Flags} should be stored in a memory variable, thus preventing
28973 the optimizer from keeping it in a register. In contrast,
28975 Unsigned_32'Asm_Output ("=r", Flags);
28977 uses the @code{"r"} (register) constraint, telling the compiler to
28978 store the variable in a register.
28980 If the constraint is preceded by the equal character (@strong{=}), it tells
28981 the compiler that the variable will be used to store data into it.
28983 In the @code{Get_Flags} example, we used the @code{"g"} (global) constraint,
28984 allowing the optimizer to choose whatever it deems best.
28986 There are a fairly large number of constraints, but the ones that are
28987 most useful (for the Intel x86 processor) are the following:
28993 global (i.e.@: can be stored anywhere)
29011 use one of eax, ebx, ecx or edx
29013 use one of eax, ebx, ecx, edx, esi or edi
29016 The full set of constraints is described in the gcc and @emph{as}
29017 documentation; note that it is possible to combine certain constraints
29018 in one constraint string.
29020 You specify the association of an output variable with an assembler operand
29021 through the @code{%}@emph{n} notation, where @emph{n} is a non-negative
29023 @smallexample @c ada
29025 Asm ("pushfl" & LF & HT & -- push flags on stack
29026 "popl %%eax" & LF & HT & -- load eax with flags
29027 "movl %%eax, %0", -- store flags in variable
29028 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
29032 @code{%0} will be replaced in the expanded code by the appropriate operand,
29034 the compiler decided for the @code{Flags} variable.
29036 In general, you may have any number of output variables:
29039 Count the operands starting at 0; thus @code{%0}, @code{%1}, etc.
29041 Specify the @code{Outputs} parameter as a parenthesized comma-separated list
29042 of @code{Asm_Output} attributes
29046 @smallexample @c ada
29048 Asm ("movl %%eax, %0" & LF & HT &
29049 "movl %%ebx, %1" & LF & HT &
29051 Outputs => (Unsigned_32'Asm_Output ("=g", Var_A), -- %0 = Var_A
29052 Unsigned_32'Asm_Output ("=g", Var_B), -- %1 = Var_B
29053 Unsigned_32'Asm_Output ("=g", Var_C))); -- %2 = Var_C
29057 where @code{Var_A}, @code{Var_B}, and @code{Var_C} are variables
29058 in the Ada program.
29060 As a variation on the @code{Get_Flags} example, we can use the constraints
29061 string to direct the compiler to store the eax register into the @code{Flags}
29062 variable, instead of including the store instruction explicitly in the
29063 @code{Asm} template string:
29065 @smallexample @c ada
29067 with Interfaces; use Interfaces;
29068 with Ada.Text_IO; use Ada.Text_IO;
29069 with System.Machine_Code; use System.Machine_Code;
29070 procedure Get_Flags_2 is
29071 Flags : Unsigned_32;
29074 Asm ("pushfl" & LF & HT & -- push flags on stack
29075 "popl %%eax", -- save flags in eax
29076 Outputs => Unsigned_32'Asm_Output ("=a", Flags));
29077 Put_Line ("Flags register:" & Flags'Img);
29083 The @code{"a"} constraint tells the compiler that the @code{Flags}
29084 variable will come from the eax register. Here is the resulting code:
29092 movl %eax,-40(%ebp)
29097 The compiler generated the store of eax into Flags after
29098 expanding the assembler code.
29100 Actually, there was no need to pop the flags into the eax register;
29101 more simply, we could just pop the flags directly into the program variable:
29103 @smallexample @c ada
29105 with Interfaces; use Interfaces;
29106 with Ada.Text_IO; use Ada.Text_IO;
29107 with System.Machine_Code; use System.Machine_Code;
29108 procedure Get_Flags_3 is
29109 Flags : Unsigned_32;
29112 Asm ("pushfl" & LF & HT & -- push flags on stack
29113 "pop %0", -- save flags in Flags
29114 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
29115 Put_Line ("Flags register:" & Flags'Img);
29120 @c ---------------------------------------------------------------------------
29121 @node Input Variables in Inline Assembler
29122 @section Input Variables in Inline Assembler
29125 The example in this section illustrates how to specify the source operands
29126 for assembly language statements.
29127 The program simply increments its input value by 1:
29129 @smallexample @c ada
29131 with Interfaces; use Interfaces;
29132 with Ada.Text_IO; use Ada.Text_IO;
29133 with System.Machine_Code; use System.Machine_Code;
29134 procedure Increment is
29136 function Incr (Value : Unsigned_32) return Unsigned_32 is
29137 Result : Unsigned_32;
29140 Inputs => Unsigned_32'Asm_Input ("a", Value),
29141 Outputs => Unsigned_32'Asm_Output ("=a", Result));
29145 Value : Unsigned_32;
29149 Put_Line ("Value before is" & Value'Img);
29150 Value := Incr (Value);
29151 Put_Line ("Value after is" & Value'Img);
29156 The @code{Outputs} parameter to @code{Asm} specifies
29157 that the result will be in the eax register and that it is to be stored
29158 in the @code{Result} variable.
29160 The @code{Inputs} parameter looks much like the @code{Outputs} parameter,
29161 but with an @code{Asm_Input} attribute.
29162 The @code{"="} constraint, indicating an output value, is not present.
29164 You can have multiple input variables, in the same way that you can have more
29165 than one output variable.
29167 The parameter count (%0, %1) etc, now starts at the first input
29168 statement, and continues with the output statements.
29169 When both parameters use the same variable, the
29170 compiler will treat them as the same %n operand, which is the case here.
29172 Just as the @code{Outputs} parameter causes the register to be stored into the
29173 target variable after execution of the assembler statements, so does the
29174 @code{Inputs} parameter cause its variable to be loaded into the register
29175 before execution of the assembler statements.
29177 Thus the effect of the @code{Asm} invocation is:
29179 @item load the 32-bit value of @code{Value} into eax
29180 @item execute the @code{incl %eax} instruction
29181 @item store the contents of eax into the @code{Result} variable
29184 The resulting assembler file (with @option{-O2} optimization) contains:
29187 _increment__incr.1:
29200 @c ---------------------------------------------------------------------------
29201 @node Inlining Inline Assembler Code
29202 @section Inlining Inline Assembler Code
29205 For a short subprogram such as the @code{Incr} function in the previous
29206 section, the overhead of the call and return (creating / deleting the stack
29207 frame) can be significant, compared to the amount of code in the subprogram
29208 body. A solution is to apply Ada's @code{Inline} pragma to the subprogram,
29209 which directs the compiler to expand invocations of the subprogram at the
29210 point(s) of call, instead of setting up a stack frame for out-of-line calls.
29211 Here is the resulting program:
29213 @smallexample @c ada
29215 with Interfaces; use Interfaces;
29216 with Ada.Text_IO; use Ada.Text_IO;
29217 with System.Machine_Code; use System.Machine_Code;
29218 procedure Increment_2 is
29220 function Incr (Value : Unsigned_32) return Unsigned_32 is
29221 Result : Unsigned_32;
29224 Inputs => Unsigned_32'Asm_Input ("a", Value),
29225 Outputs => Unsigned_32'Asm_Output ("=a", Result));
29228 pragma Inline (Increment);
29230 Value : Unsigned_32;
29234 Put_Line ("Value before is" & Value'Img);
29235 Value := Increment (Value);
29236 Put_Line ("Value after is" & Value'Img);
29241 Compile the program with both optimization (@option{-O2}) and inlining
29242 (@option{-gnatn}) enabled.
29244 The @code{Incr} function is still compiled as usual, but at the
29245 point in @code{Increment} where our function used to be called:
29250 call _increment__incr.1
29255 the code for the function body directly appears:
29268 thus saving the overhead of stack frame setup and an out-of-line call.
29270 @c ---------------------------------------------------------------------------
29271 @node Other Asm Functionality
29272 @section Other @code{Asm} Functionality
29275 This section describes two important parameters to the @code{Asm}
29276 procedure: @code{Clobber}, which identifies register usage;
29277 and @code{Volatile}, which inhibits unwanted optimizations.
29280 * The Clobber Parameter::
29281 * The Volatile Parameter::
29284 @c ---------------------------------------------------------------------------
29285 @node The Clobber Parameter
29286 @subsection The @code{Clobber} Parameter
29289 One of the dangers of intermixing assembly language and a compiled language
29290 such as Ada is that the compiler needs to be aware of which registers are
29291 being used by the assembly code. In some cases, such as the earlier examples,
29292 the constraint string is sufficient to indicate register usage (e.g.,
29294 the eax register). But more generally, the compiler needs an explicit
29295 identification of the registers that are used by the Inline Assembly
29298 Using a register that the compiler doesn't know about
29299 could be a side effect of an instruction (like @code{mull}
29300 storing its result in both eax and edx).
29301 It can also arise from explicit register usage in your
29302 assembly code; for example:
29305 Asm ("movl %0, %%ebx" & LF & HT &
29307 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
29308 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out));
29312 where the compiler (since it does not analyze the @code{Asm} template string)
29313 does not know you are using the ebx register.
29315 In such cases you need to supply the @code{Clobber} parameter to @code{Asm},
29316 to identify the registers that will be used by your assembly code:
29320 Asm ("movl %0, %%ebx" & LF & HT &
29322 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
29323 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
29328 The Clobber parameter is a static string expression specifying the
29329 register(s) you are using. Note that register names are @emph{not} prefixed
29330 by a percent sign. Also, if more than one register is used then their names
29331 are separated by commas; e.g., @code{"eax, ebx"}
29333 The @code{Clobber} parameter has several additional uses:
29335 @item Use ``register'' name @code{cc} to indicate that flags might have changed
29336 @item Use ``register'' name @code{memory} if you changed a memory location
29339 @c ---------------------------------------------------------------------------
29340 @node The Volatile Parameter
29341 @subsection The @code{Volatile} Parameter
29342 @cindex Volatile parameter
29345 Compiler optimizations in the presence of Inline Assembler may sometimes have
29346 unwanted effects. For example, when an @code{Asm} invocation with an input
29347 variable is inside a loop, the compiler might move the loading of the input
29348 variable outside the loop, regarding it as a one-time initialization.
29350 If this effect is not desired, you can disable such optimizations by setting
29351 the @code{Volatile} parameter to @code{True}; for example:
29353 @smallexample @c ada
29355 Asm ("movl %0, %%ebx" & LF & HT &
29357 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
29358 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
29364 By default, @code{Volatile} is set to @code{False} unless there is no
29365 @code{Outputs} parameter.
29367 Although setting @code{Volatile} to @code{True} prevents unwanted
29368 optimizations, it will also disable other optimizations that might be
29369 important for efficiency. In general, you should set @code{Volatile}
29370 to @code{True} only if the compiler's optimizations have created
29372 @c END OF INLINE ASSEMBLER CHAPTER
29373 @c ===============================
29375 @c ***********************************
29376 @c * Compatibility and Porting Guide *
29377 @c ***********************************
29378 @node Compatibility and Porting Guide
29379 @appendix Compatibility and Porting Guide
29382 This chapter describes the compatibility issues that may arise between
29383 GNAT and other Ada compilation systems (including those for Ada 83),
29384 and shows how GNAT can expedite porting
29385 applications developed in other Ada environments.
29388 * Compatibility with Ada 83::
29389 * Compatibility between Ada 95 and Ada 2005::
29390 * Implementation-dependent characteristics::
29391 * Compatibility with Other Ada Systems::
29392 * Representation Clauses::
29394 @c Brief section is only in non-VMS version
29395 @c Full chapter is in VMS version
29396 * Compatibility with HP Ada 83::
29399 * Transitioning to 64-Bit GNAT for OpenVMS::
29403 @node Compatibility with Ada 83
29404 @section Compatibility with Ada 83
29405 @cindex Compatibility (between Ada 83 and Ada 95 / Ada 2005)
29408 Ada 95 and Ada 2005 are highly upwards compatible with Ada 83. In
29409 particular, the design intention was that the difficulties associated
29410 with moving from Ada 83 to Ada 95 or Ada 2005 should be no greater than those
29411 that occur when moving from one Ada 83 system to another.
29413 However, there are a number of points at which there are minor
29414 incompatibilities. The @cite{Ada 95 Annotated Reference Manual} contains
29415 full details of these issues,
29416 and should be consulted for a complete treatment.
29418 following subsections treat the most likely issues to be encountered.
29421 * Legal Ada 83 programs that are illegal in Ada 95::
29422 * More deterministic semantics::
29423 * Changed semantics::
29424 * Other language compatibility issues::
29427 @node Legal Ada 83 programs that are illegal in Ada 95
29428 @subsection Legal Ada 83 programs that are illegal in Ada 95
29430 Some legal Ada 83 programs are illegal (i.e., they will fail to compile) in
29431 Ada 95 and thus also in Ada 2005:
29434 @item Character literals
29435 Some uses of character literals are ambiguous. Since Ada 95 has introduced
29436 @code{Wide_Character} as a new predefined character type, some uses of
29437 character literals that were legal in Ada 83 are illegal in Ada 95.
29439 @smallexample @c ada
29440 for Char in 'A' .. 'Z' loop @dots{} end loop;
29444 The problem is that @code{'A'} and @code{'Z'} could be from either
29445 @code{Character} or @code{Wide_Character}. The simplest correction
29446 is to make the type explicit; e.g.:
29447 @smallexample @c ada
29448 for Char in Character range 'A' .. 'Z' loop @dots{} end loop;
29451 @item New reserved words
29452 The identifiers @code{abstract}, @code{aliased}, @code{protected},
29453 @code{requeue}, @code{tagged}, and @code{until} are reserved in Ada 95.
29454 Existing Ada 83 code using any of these identifiers must be edited to
29455 use some alternative name.
29457 @item Freezing rules
29458 The rules in Ada 95 are slightly different with regard to the point at
29459 which entities are frozen, and representation pragmas and clauses are
29460 not permitted past the freeze point. This shows up most typically in
29461 the form of an error message complaining that a representation item
29462 appears too late, and the appropriate corrective action is to move
29463 the item nearer to the declaration of the entity to which it refers.
29465 A particular case is that representation pragmas
29468 extended HP Ada 83 compatibility pragmas such as @code{Export_Procedure})
29470 cannot be applied to a subprogram body. If necessary, a separate subprogram
29471 declaration must be introduced to which the pragma can be applied.
29473 @item Optional bodies for library packages
29474 In Ada 83, a package that did not require a package body was nevertheless
29475 allowed to have one. This lead to certain surprises in compiling large
29476 systems (situations in which the body could be unexpectedly ignored by the
29477 binder). In Ada 95, if a package does not require a body then it is not
29478 permitted to have a body. To fix this problem, simply remove a redundant
29479 body if it is empty, or, if it is non-empty, introduce a dummy declaration
29480 into the spec that makes the body required. One approach is to add a private
29481 part to the package declaration (if necessary), and define a parameterless
29482 procedure called @code{Requires_Body}, which must then be given a dummy
29483 procedure body in the package body, which then becomes required.
29484 Another approach (assuming that this does not introduce elaboration
29485 circularities) is to add an @code{Elaborate_Body} pragma to the package spec,
29486 since one effect of this pragma is to require the presence of a package body.
29488 @item @code{Numeric_Error} is now the same as @code{Constraint_Error}
29489 In Ada 95, the exception @code{Numeric_Error} is a renaming of
29490 @code{Constraint_Error}.
29491 This means that it is illegal to have separate exception handlers for
29492 the two exceptions. The fix is simply to remove the handler for the
29493 @code{Numeric_Error} case (since even in Ada 83, a compiler was free to raise
29494 @code{Constraint_Error} in place of @code{Numeric_Error} in all cases).
29496 @item Indefinite subtypes in generics
29497 In Ada 83, it was permissible to pass an indefinite type (e.g.@: @code{String})
29498 as the actual for a generic formal private type, but then the instantiation
29499 would be illegal if there were any instances of declarations of variables
29500 of this type in the generic body. In Ada 95, to avoid this clear violation
29501 of the methodological principle known as the ``contract model'',
29502 the generic declaration explicitly indicates whether
29503 or not such instantiations are permitted. If a generic formal parameter
29504 has explicit unknown discriminants, indicated by using @code{(<>)} after the
29505 type name, then it can be instantiated with indefinite types, but no
29506 stand-alone variables can be declared of this type. Any attempt to declare
29507 such a variable will result in an illegality at the time the generic is
29508 declared. If the @code{(<>)} notation is not used, then it is illegal
29509 to instantiate the generic with an indefinite type.
29510 This is the potential incompatibility issue when porting Ada 83 code to Ada 95.
29511 It will show up as a compile time error, and
29512 the fix is usually simply to add the @code{(<>)} to the generic declaration.
29515 @node More deterministic semantics
29516 @subsection More deterministic semantics
29520 Conversions from real types to integer types round away from 0. In Ada 83
29521 the conversion Integer(2.5) could deliver either 2 or 3 as its value. This
29522 implementation freedom was intended to support unbiased rounding in
29523 statistical applications, but in practice it interfered with portability.
29524 In Ada 95 the conversion semantics are unambiguous, and rounding away from 0
29525 is required. Numeric code may be affected by this change in semantics.
29526 Note, though, that this issue is no worse than already existed in Ada 83
29527 when porting code from one vendor to another.
29530 The Real-Time Annex introduces a set of policies that define the behavior of
29531 features that were implementation dependent in Ada 83, such as the order in
29532 which open select branches are executed.
29535 @node Changed semantics
29536 @subsection Changed semantics
29539 The worst kind of incompatibility is one where a program that is legal in
29540 Ada 83 is also legal in Ada 95 but can have an effect in Ada 95 that was not
29541 possible in Ada 83. Fortunately this is extremely rare, but the one
29542 situation that you should be alert to is the change in the predefined type
29543 @code{Character} from 7-bit ASCII to 8-bit Latin-1.
29546 @item Range of type @code{Character}
29547 The range of @code{Standard.Character} is now the full 256 characters
29548 of Latin-1, whereas in most Ada 83 implementations it was restricted
29549 to 128 characters. Although some of the effects of
29550 this change will be manifest in compile-time rejection of legal
29551 Ada 83 programs it is possible for a working Ada 83 program to have
29552 a different effect in Ada 95, one that was not permitted in Ada 83.
29553 As an example, the expression
29554 @code{Character'Pos(Character'Last)} returned @code{127} in Ada 83 and now
29555 delivers @code{255} as its value.
29556 In general, you should look at the logic of any
29557 character-processing Ada 83 program and see whether it needs to be adapted
29558 to work correctly with Latin-1. Note that the predefined Ada 95 API has a
29559 character handling package that may be relevant if code needs to be adapted
29560 to account for the additional Latin-1 elements.
29561 The desirable fix is to
29562 modify the program to accommodate the full character set, but in some cases
29563 it may be convenient to define a subtype or derived type of Character that
29564 covers only the restricted range.
29568 @node Other language compatibility issues
29569 @subsection Other language compatibility issues
29572 @item @option{-gnat83} switch
29573 All implementations of GNAT provide a switch that causes GNAT to operate
29574 in Ada 83 mode. In this mode, some but not all compatibility problems
29575 of the type described above are handled automatically. For example, the
29576 new reserved words introduced in Ada 95 and Ada 2005 are treated simply
29577 as identifiers as in Ada 83.
29579 in practice, it is usually advisable to make the necessary modifications
29580 to the program to remove the need for using this switch.
29581 See @ref{Compiling Different Versions of Ada}.
29583 @item Support for removed Ada 83 pragmas and attributes
29584 A number of pragmas and attributes from Ada 83 were removed from Ada 95,
29585 generally because they were replaced by other mechanisms. Ada 95 and Ada 2005
29586 compilers are allowed, but not required, to implement these missing
29587 elements. In contrast with some other compilers, GNAT implements all
29588 such pragmas and attributes, eliminating this compatibility concern. These
29589 include @code{pragma Interface} and the floating point type attributes
29590 (@code{Emax}, @code{Mantissa}, etc.), among other items.
29594 @node Compatibility between Ada 95 and Ada 2005
29595 @section Compatibility between Ada 95 and Ada 2005
29596 @cindex Compatibility between Ada 95 and Ada 2005
29599 Although Ada 2005 was designed to be upwards compatible with Ada 95, there are
29600 a number of incompatibilities. Several are enumerated below;
29601 for a complete description please see the
29602 Annotated Ada 2005 Reference Manual, or section 9.1.1 in
29603 @cite{Rationale for Ada 2005}.
29606 @item New reserved words.
29607 The words @code{interface}, @code{overriding} and @code{synchronized} are
29608 reserved in Ada 2005.
29609 A pre-Ada 2005 program that uses any of these as an identifier will be
29612 @item New declarations in predefined packages.
29613 A number of packages in the predefined environment contain new declarations:
29614 @code{Ada.Exceptions}, @code{Ada.Real_Time}, @code{Ada.Strings},
29615 @code{Ada.Strings.Fixed}, @code{Ada.Strings.Bounded},
29616 @code{Ada.Strings.Unbounded}, @code{Ada.Strings.Wide_Fixed},
29617 @code{Ada.Strings.Wide_Bounded}, @code{Ada.Strings.Wide_Unbounded},
29618 @code{Ada.Tags}, @code{Ada.Text_IO}, and @code{Interfaces.C}.
29619 If an Ada 95 program does a @code{with} and @code{use} of any of these
29620 packages, the new declarations may cause name clashes.
29622 @item Access parameters.
29623 A nondispatching subprogram with an access parameter cannot be renamed
29624 as a dispatching operation. This was permitted in Ada 95.
29626 @item Access types, discriminants, and constraints.
29627 Rule changes in this area have led to some incompatibilities; for example,
29628 constrained subtypes of some access types are not permitted in Ada 2005.
29630 @item Aggregates for limited types.
29631 The allowance of aggregates for limited types in Ada 2005 raises the
29632 possibility of ambiguities in legal Ada 95 programs, since additional types
29633 now need to be considered in expression resolution.
29635 @item Fixed-point multiplication and division.
29636 Certain expressions involving ``*'' or ``/'' for a fixed-point type, which
29637 were legal in Ada 95 and invoked the predefined versions of these operations,
29639 The ambiguity may be resolved either by applying a type conversion to the
29640 expression, or by explicitly invoking the operation from package
29643 @item Return-by-reference types.
29644 The Ada 95 return-by-reference mechanism has been removed. Instead, the user
29645 can declare a function returning a value from an anonymous access type.
29649 @node Implementation-dependent characteristics
29650 @section Implementation-dependent characteristics
29652 Although the Ada language defines the semantics of each construct as
29653 precisely as practical, in some situations (for example for reasons of
29654 efficiency, or where the effect is heavily dependent on the host or target
29655 platform) the implementation is allowed some freedom. In porting Ada 83
29656 code to GNAT, you need to be aware of whether / how the existing code
29657 exercised such implementation dependencies. Such characteristics fall into
29658 several categories, and GNAT offers specific support in assisting the
29659 transition from certain Ada 83 compilers.
29662 * Implementation-defined pragmas::
29663 * Implementation-defined attributes::
29665 * Elaboration order::
29666 * Target-specific aspects::
29669 @node Implementation-defined pragmas
29670 @subsection Implementation-defined pragmas
29673 Ada compilers are allowed to supplement the language-defined pragmas, and
29674 these are a potential source of non-portability. All GNAT-defined pragmas
29675 are described in @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT
29676 Reference Manual}, and these include several that are specifically
29677 intended to correspond to other vendors' Ada 83 pragmas.
29678 For migrating from VADS, the pragma @code{Use_VADS_Size} may be useful.
29679 For compatibility with HP Ada 83, GNAT supplies the pragmas
29680 @code{Extend_System}, @code{Ident}, @code{Inline_Generic},
29681 @code{Interface_Name}, @code{Passive}, @code{Suppress_All},
29682 and @code{Volatile}.
29683 Other relevant pragmas include @code{External} and @code{Link_With}.
29684 Some vendor-specific
29685 Ada 83 pragmas (@code{Share_Generic}, @code{Subtitle}, and @code{Title}) are
29687 avoiding compiler rejection of units that contain such pragmas; they are not
29688 relevant in a GNAT context and hence are not otherwise implemented.
29690 @node Implementation-defined attributes
29691 @subsection Implementation-defined attributes
29693 Analogous to pragmas, the set of attributes may be extended by an
29694 implementation. All GNAT-defined attributes are described in
29695 @ref{Implementation Defined Attributes,,, gnat_rm, GNAT Reference
29696 Manual}, and these include several that are specifically intended
29697 to correspond to other vendors' Ada 83 attributes. For migrating from VADS,
29698 the attribute @code{VADS_Size} may be useful. For compatibility with HP
29699 Ada 83, GNAT supplies the attributes @code{Bit}, @code{Machine_Size} and
29703 @subsection Libraries
29705 Vendors may supply libraries to supplement the standard Ada API. If Ada 83
29706 code uses vendor-specific libraries then there are several ways to manage
29707 this in Ada 95 or Ada 2005:
29710 If the source code for the libraries (specs and bodies) are
29711 available, then the libraries can be migrated in the same way as the
29714 If the source code for the specs but not the bodies are
29715 available, then you can reimplement the bodies.
29717 Some features introduced by Ada 95 obviate the need for library support. For
29718 example most Ada 83 vendors supplied a package for unsigned integers. The
29719 Ada 95 modular type feature is the preferred way to handle this need, so
29720 instead of migrating or reimplementing the unsigned integer package it may
29721 be preferable to retrofit the application using modular types.
29724 @node Elaboration order
29725 @subsection Elaboration order
29727 The implementation can choose any elaboration order consistent with the unit
29728 dependency relationship. This freedom means that some orders can result in
29729 Program_Error being raised due to an ``Access Before Elaboration'': an attempt
29730 to invoke a subprogram its body has been elaborated, or to instantiate a
29731 generic before the generic body has been elaborated. By default GNAT
29732 attempts to choose a safe order (one that will not encounter access before
29733 elaboration problems) by implicitly inserting @code{Elaborate} or
29734 @code{Elaborate_All} pragmas where
29735 needed. However, this can lead to the creation of elaboration circularities
29736 and a resulting rejection of the program by gnatbind. This issue is
29737 thoroughly described in @ref{Elaboration Order Handling in GNAT}.
29738 In brief, there are several
29739 ways to deal with this situation:
29743 Modify the program to eliminate the circularities, e.g.@: by moving
29744 elaboration-time code into explicitly-invoked procedures
29746 Constrain the elaboration order by including explicit @code{Elaborate_Body} or
29747 @code{Elaborate} pragmas, and then inhibit the generation of implicit
29748 @code{Elaborate_All}
29749 pragmas either globally (as an effect of the @option{-gnatE} switch) or locally
29750 (by selectively suppressing elaboration checks via pragma
29751 @code{Suppress(Elaboration_Check)} when it is safe to do so).
29754 @node Target-specific aspects
29755 @subsection Target-specific aspects
29757 Low-level applications need to deal with machine addresses, data
29758 representations, interfacing with assembler code, and similar issues. If
29759 such an Ada 83 application is being ported to different target hardware (for
29760 example where the byte endianness has changed) then you will need to
29761 carefully examine the program logic; the porting effort will heavily depend
29762 on the robustness of the original design. Moreover, Ada 95 (and thus
29763 Ada 2005) are sometimes
29764 incompatible with typical Ada 83 compiler practices regarding implicit
29765 packing, the meaning of the Size attribute, and the size of access values.
29766 GNAT's approach to these issues is described in @ref{Representation Clauses}.
29768 @node Compatibility with Other Ada Systems
29769 @section Compatibility with Other Ada Systems
29772 If programs avoid the use of implementation dependent and
29773 implementation defined features, as documented in the @cite{Ada
29774 Reference Manual}, there should be a high degree of portability between
29775 GNAT and other Ada systems. The following are specific items which
29776 have proved troublesome in moving Ada 95 programs from GNAT to other Ada 95
29777 compilers, but do not affect porting code to GNAT@.
29778 (As of @value{NOW}, GNAT is the only compiler available for Ada 2005;
29779 the following issues may or may not arise for Ada 2005 programs
29780 when other compilers appear.)
29783 @item Ada 83 Pragmas and Attributes
29784 Ada 95 compilers are allowed, but not required, to implement the missing
29785 Ada 83 pragmas and attributes that are no longer defined in Ada 95.
29786 GNAT implements all such pragmas and attributes, eliminating this as
29787 a compatibility concern, but some other Ada 95 compilers reject these
29788 pragmas and attributes.
29790 @item Specialized Needs Annexes
29791 GNAT implements the full set of special needs annexes. At the
29792 current time, it is the only Ada 95 compiler to do so. This means that
29793 programs making use of these features may not be portable to other Ada
29794 95 compilation systems.
29796 @item Representation Clauses
29797 Some other Ada 95 compilers implement only the minimal set of
29798 representation clauses required by the Ada 95 reference manual. GNAT goes
29799 far beyond this minimal set, as described in the next section.
29802 @node Representation Clauses
29803 @section Representation Clauses
29806 The Ada 83 reference manual was quite vague in describing both the minimal
29807 required implementation of representation clauses, and also their precise
29808 effects. Ada 95 (and thus also Ada 2005) are much more explicit, but the
29809 minimal set of capabilities required is still quite limited.
29811 GNAT implements the full required set of capabilities in
29812 Ada 95 and Ada 2005, but also goes much further, and in particular
29813 an effort has been made to be compatible with existing Ada 83 usage to the
29814 greatest extent possible.
29816 A few cases exist in which Ada 83 compiler behavior is incompatible with
29817 the requirements in Ada 95 (and thus also Ada 2005). These are instances of
29818 intentional or accidental dependence on specific implementation dependent
29819 characteristics of these Ada 83 compilers. The following is a list of
29820 the cases most likely to arise in existing Ada 83 code.
29823 @item Implicit Packing
29824 Some Ada 83 compilers allowed a Size specification to cause implicit
29825 packing of an array or record. This could cause expensive implicit
29826 conversions for change of representation in the presence of derived
29827 types, and the Ada design intends to avoid this possibility.
29828 Subsequent AI's were issued to make it clear that such implicit
29829 change of representation in response to a Size clause is inadvisable,
29830 and this recommendation is represented explicitly in the Ada 95 (and Ada 2005)
29831 Reference Manuals as implementation advice that is followed by GNAT@.
29832 The problem will show up as an error
29833 message rejecting the size clause. The fix is simply to provide
29834 the explicit pragma @code{Pack}, or for more fine tuned control, provide
29835 a Component_Size clause.
29837 @item Meaning of Size Attribute
29838 The Size attribute in Ada 95 (and Ada 2005) for discrete types is defined as
29839 the minimal number of bits required to hold values of the type. For example,
29840 on a 32-bit machine, the size of @code{Natural} will typically be 31 and not
29841 32 (since no sign bit is required). Some Ada 83 compilers gave 31, and
29842 some 32 in this situation. This problem will usually show up as a compile
29843 time error, but not always. It is a good idea to check all uses of the
29844 'Size attribute when porting Ada 83 code. The GNAT specific attribute
29845 Object_Size can provide a useful way of duplicating the behavior of
29846 some Ada 83 compiler systems.
29848 @item Size of Access Types
29849 A common assumption in Ada 83 code is that an access type is in fact a pointer,
29850 and that therefore it will be the same size as a System.Address value. This
29851 assumption is true for GNAT in most cases with one exception. For the case of
29852 a pointer to an unconstrained array type (where the bounds may vary from one
29853 value of the access type to another), the default is to use a ``fat pointer'',
29854 which is represented as two separate pointers, one to the bounds, and one to
29855 the array. This representation has a number of advantages, including improved
29856 efficiency. However, it may cause some difficulties in porting existing Ada 83
29857 code which makes the assumption that, for example, pointers fit in 32 bits on
29858 a machine with 32-bit addressing.
29860 To get around this problem, GNAT also permits the use of ``thin pointers'' for
29861 access types in this case (where the designated type is an unconstrained array
29862 type). These thin pointers are indeed the same size as a System.Address value.
29863 To specify a thin pointer, use a size clause for the type, for example:
29865 @smallexample @c ada
29866 type X is access all String;
29867 for X'Size use Standard'Address_Size;
29871 which will cause the type X to be represented using a single pointer.
29872 When using this representation, the bounds are right behind the array.
29873 This representation is slightly less efficient, and does not allow quite
29874 such flexibility in the use of foreign pointers or in using the
29875 Unrestricted_Access attribute to create pointers to non-aliased objects.
29876 But for any standard portable use of the access type it will work in
29877 a functionally correct manner and allow porting of existing code.
29878 Note that another way of forcing a thin pointer representation
29879 is to use a component size clause for the element size in an array,
29880 or a record representation clause for an access field in a record.
29884 @c This brief section is only in the non-VMS version
29885 @c The complete chapter on HP Ada is in the VMS version
29886 @node Compatibility with HP Ada 83
29887 @section Compatibility with HP Ada 83
29890 The VMS version of GNAT fully implements all the pragmas and attributes
29891 provided by HP Ada 83, as well as providing the standard HP Ada 83
29892 libraries, including Starlet. In addition, data layouts and parameter
29893 passing conventions are highly compatible. This means that porting
29894 existing HP Ada 83 code to GNAT in VMS systems should be easier than
29895 most other porting efforts. The following are some of the most
29896 significant differences between GNAT and HP Ada 83.
29899 @item Default floating-point representation
29900 In GNAT, the default floating-point format is IEEE, whereas in HP Ada 83,
29901 it is VMS format. GNAT does implement the necessary pragmas
29902 (Long_Float, Float_Representation) for changing this default.
29905 The package System in GNAT exactly corresponds to the definition in the
29906 Ada 95 reference manual, which means that it excludes many of the
29907 HP Ada 83 extensions. However, a separate package Aux_DEC is provided
29908 that contains the additional definitions, and a special pragma,
29909 Extend_System allows this package to be treated transparently as an
29910 extension of package System.
29913 The definitions provided by Aux_DEC are exactly compatible with those
29914 in the HP Ada 83 version of System, with one exception.
29915 HP Ada provides the following declarations:
29917 @smallexample @c ada
29918 TO_ADDRESS (INTEGER)
29919 TO_ADDRESS (UNSIGNED_LONGWORD)
29920 TO_ADDRESS (@i{universal_integer})
29924 The version of TO_ADDRESS taking a @i{universal integer} argument is in fact
29925 an extension to Ada 83 not strictly compatible with the reference manual.
29926 In GNAT, we are constrained to be exactly compatible with the standard,
29927 and this means we cannot provide this capability. In HP Ada 83, the
29928 point of this definition is to deal with a call like:
29930 @smallexample @c ada
29931 TO_ADDRESS (16#12777#);
29935 Normally, according to the Ada 83 standard, one would expect this to be
29936 ambiguous, since it matches both the INTEGER and UNSIGNED_LONGWORD forms
29937 of TO_ADDRESS@. However, in HP Ada 83, there is no ambiguity, since the
29938 definition using @i{universal_integer} takes precedence.
29940 In GNAT, since the version with @i{universal_integer} cannot be supplied, it
29941 is not possible to be 100% compatible. Since there are many programs using
29942 numeric constants for the argument to TO_ADDRESS, the decision in GNAT was
29943 to change the name of the function in the UNSIGNED_LONGWORD case, so the
29944 declarations provided in the GNAT version of AUX_Dec are:
29946 @smallexample @c ada
29947 function To_Address (X : Integer) return Address;
29948 pragma Pure_Function (To_Address);
29950 function To_Address_Long (X : Unsigned_Longword)
29952 pragma Pure_Function (To_Address_Long);
29956 This means that programs using TO_ADDRESS for UNSIGNED_LONGWORD must
29957 change the name to TO_ADDRESS_LONG@.
29959 @item Task_Id values
29960 The Task_Id values assigned will be different in the two systems, and GNAT
29961 does not provide a specified value for the Task_Id of the environment task,
29962 which in GNAT is treated like any other declared task.
29966 For full details on these and other less significant compatibility issues,
29967 see appendix E of the HP publication entitled @cite{HP Ada, Technical
29968 Overview and Comparison on HP Platforms}.
29970 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
29971 attributes are recognized, although only a subset of them can sensibly
29972 be implemented. The description of pragmas in @ref{Implementation
29973 Defined Pragmas,,, gnat_rm, GNAT Reference Manual}
29974 indicates whether or not they are applicable to non-VMS systems.
29978 @node Transitioning to 64-Bit GNAT for OpenVMS
29979 @section Transitioning to 64-Bit @value{EDITION} for OpenVMS
29982 This section is meant to assist users of pre-2006 @value{EDITION}
29983 for Alpha OpenVMS who are transitioning to 64-bit @value{EDITION},
29984 the version of the GNAT technology supplied in 2006 and later for
29985 OpenVMS on both Alpha and I64.
29988 * Introduction to transitioning::
29989 * Migration of 32 bit code::
29990 * Taking advantage of 64 bit addressing::
29991 * Technical details::
29994 @node Introduction to transitioning
29995 @subsection Introduction
29998 64-bit @value{EDITION} for Open VMS has been designed to meet
30003 Providing a full conforming implementation of Ada 95 and Ada 2005
30006 Allowing maximum backward compatibility, thus easing migration of existing
30010 Supplying a path for exploiting the full 64-bit address range
30014 Ada's strong typing semantics has made it
30015 impractical to have different 32-bit and 64-bit modes. As soon as
30016 one object could possibly be outside the 32-bit address space, this
30017 would make it necessary for the @code{System.Address} type to be 64 bits.
30018 In particular, this would cause inconsistencies if 32-bit code is
30019 called from 64-bit code that raises an exception.
30021 This issue has been resolved by always using 64-bit addressing
30022 at the system level, but allowing for automatic conversions between
30023 32-bit and 64-bit addresses where required. Thus users who
30024 do not currently require 64-bit addressing capabilities, can
30025 recompile their code with only minimal changes (and indeed
30026 if the code is written in portable Ada, with no assumptions about
30027 the size of the @code{Address} type, then no changes at all are necessary).
30029 this approach provides a simple, gradual upgrade path to future
30030 use of larger memories than available for 32-bit systems.
30031 Also, newly written applications or libraries will by default
30032 be fully compatible with future systems exploiting 64-bit
30033 addressing capabilities.
30035 @ref{Migration of 32 bit code}, will focus on porting applications
30036 that do not require more than 2 GB of
30037 addressable memory. This code will be referred to as
30038 @emph{32-bit code}.
30039 For applications intending to exploit the full 64-bit address space,
30040 @ref{Taking advantage of 64 bit addressing},
30041 will consider further changes that may be required.
30042 Such code will be referred to below as @emph{64-bit code}.
30044 @node Migration of 32 bit code
30045 @subsection Migration of 32-bit code
30050 * Unchecked conversions::
30051 * Predefined constants::
30052 * Interfacing with C::
30053 * Experience with source compatibility::
30056 @node Address types
30057 @subsubsection Address types
30060 To solve the problem of mixing 64-bit and 32-bit addressing,
30061 while maintaining maximum backward compatibility, the following
30062 approach has been taken:
30066 @code{System.Address} always has a size of 64 bits
30069 @code{System.Short_Address} is a 32-bit subtype of @code{System.Address}
30073 Since @code{System.Short_Address} is a subtype of @code{System.Address},
30074 a @code{Short_Address}
30075 may be used where an @code{Address} is required, and vice versa, without
30076 needing explicit type conversions.
30077 By virtue of the Open VMS parameter passing conventions,
30079 and exported subprograms that have 32-bit address parameters are
30080 compatible with those that have 64-bit address parameters.
30081 (See @ref{Making code 64 bit clean} for details.)
30083 The areas that may need attention are those where record types have
30084 been defined that contain components of the type @code{System.Address}, and
30085 where objects of this type are passed to code expecting a record layout with
30088 Different compilers on different platforms cannot be
30089 expected to represent the same type in the same way,
30090 since alignment constraints
30091 and other system-dependent properties affect the compiler's decision.
30092 For that reason, Ada code
30093 generally uses representation clauses to specify the expected
30094 layout where required.
30096 If such a representation clause uses 32 bits for a component having
30097 the type @code{System.Address}, 64-bit @value{EDITION} for OpenVMS
30098 will detect that error and produce a specific diagnostic message.
30099 The developer should then determine whether the representation
30100 should be 64 bits or not and make either of two changes:
30101 change the size to 64 bits and leave the type as @code{System.Address}, or
30102 leave the size as 32 bits and change the type to @code{System.Short_Address}.
30103 Since @code{Short_Address} is a subtype of @code{Address}, no changes are
30104 required in any code setting or accessing the field; the compiler will
30105 automatically perform any needed conversions between address
30109 @subsubsection Access types
30112 By default, objects designated by access values are always
30113 allocated in the 32-bit
30114 address space. Thus legacy code will never contain
30115 any objects that are not addressable with 32-bit addresses, and
30116 the compiler will never raise exceptions as result of mixing
30117 32-bit and 64-bit addresses.
30119 However, the access values themselves are represented in 64 bits, for optimum
30120 performance and future compatibility with 64-bit code. As was
30121 the case with @code{System.Address}, the compiler will give an error message
30122 if an object or record component has a representation clause that
30123 requires the access value to fit in 32 bits. In such a situation,
30124 an explicit size clause for the access type, specifying 32 bits,
30125 will have the desired effect.
30127 General access types (declared with @code{access all}) can never be
30128 32 bits, as values of such types must be able to refer to any object
30129 of the designated type,
30130 including objects residing outside the 32-bit address range.
30131 Existing Ada 83 code will not contain such type definitions,
30132 however, since general access types were introduced in Ada 95.
30134 @node Unchecked conversions
30135 @subsubsection Unchecked conversions
30138 In the case of an @code{Unchecked_Conversion} where the source type is a
30139 64-bit access type or the type @code{System.Address}, and the target
30140 type is a 32-bit type, the compiler will generate a warning.
30141 Even though the generated code will still perform the required
30142 conversions, it is highly recommended in these cases to use
30143 respectively a 32-bit access type or @code{System.Short_Address}
30144 as the source type.
30146 @node Predefined constants
30147 @subsubsection Predefined constants
30150 The following table shows the correspondence between pre-2006 versions of
30151 @value{EDITION} on Alpha OpenVMS (``Old'') and 64-bit @value{EDITION}
30154 @multitable {@code{System.Short_Memory_Size}} {2**32} {2**64}
30155 @item @b{Constant} @tab @b{Old} @tab @b{New}
30156 @item @code{System.Word_Size} @tab 32 @tab 64
30157 @item @code{System.Memory_Size} @tab 2**32 @tab 2**64
30158 @item @code{System.Short_Memory_Size} @tab 2**32 @tab 2**32
30159 @item @code{System.Address_Size} @tab 32 @tab 64
30163 If you need to refer to the specific
30164 memory size of a 32-bit implementation, instead of the
30165 actual memory size, use @code{System.Short_Memory_Size}
30166 rather than @code{System.Memory_Size}.
30167 Similarly, references to @code{System.Address_Size} may need
30168 to be replaced by @code{System.Short_Address'Size}.
30169 The program @command{gnatfind} may be useful for locating
30170 references to the above constants, so that you can verify that they
30173 @node Interfacing with C
30174 @subsubsection Interfacing with C
30177 In order to minimize the impact of the transition to 64-bit addresses on
30178 legacy programs, some fundamental types in the @code{Interfaces.C}
30179 package hierarchy continue to be represented in 32 bits.
30180 These types are: @code{ptrdiff_t}, @code{size_t}, and @code{chars_ptr}.
30181 This eases integration with the default HP C layout choices, for example
30182 as found in the system routines in @code{DECC$SHR.EXE}.
30183 Because of this implementation choice, the type fully compatible with
30184 @code{chars_ptr} is now @code{Short_Address} and not @code{Address}.
30185 Depending on the context the compiler will issue a
30186 warning or an error when type @code{Address} is used, alerting the user to a
30187 potential problem. Otherwise 32-bit programs that use
30188 @code{Interfaces.C} should normally not require code modifications
30190 The other issue arising with C interfacing concerns pragma @code{Convention}.
30191 For VMS 64-bit systems, there is an issue of the appropriate default size
30192 of C convention pointers in the absence of an explicit size clause. The HP
30193 C compiler can choose either 32 or 64 bits depending on compiler options.
30194 GNAT chooses 32-bits rather than 64-bits in the default case where no size
30195 clause is given. This proves a better choice for porting 32-bit legacy
30196 applications. In order to have a 64-bit representation, it is necessary to
30197 specify a size representation clause. For example:
30199 @smallexample @c ada
30200 type int_star is access Interfaces.C.int;
30201 pragma Convention(C, int_star);
30202 for int_star'Size use 64; -- Necessary to get 64 and not 32 bits
30205 @node Experience with source compatibility
30206 @subsubsection Experience with source compatibility
30209 The Security Server and STARLET on I64 provide an interesting ``test case''
30210 for source compatibility issues, since it is in such system code
30211 where assumptions about @code{Address} size might be expected to occur.
30212 Indeed, there were a small number of occasions in the Security Server
30213 file @file{jibdef.ads}
30214 where a representation clause for a record type specified
30215 32 bits for a component of type @code{Address}.
30216 All of these errors were detected by the compiler.
30217 The repair was obvious and immediate; to simply replace @code{Address} by
30218 @code{Short_Address}.
30220 In the case of STARLET, there were several record types that should
30221 have had representation clauses but did not. In these record types
30222 there was an implicit assumption that an @code{Address} value occupied
30224 These compiled without error, but their usage resulted in run-time error
30225 returns from STARLET system calls.
30226 Future GNAT technology enhancements may include a tool that detects and flags
30227 these sorts of potential source code porting problems.
30229 @c ****************************************
30230 @node Taking advantage of 64 bit addressing
30231 @subsection Taking advantage of 64-bit addressing
30234 * Making code 64 bit clean::
30235 * Allocating memory from the 64 bit storage pool::
30236 * Restrictions on use of 64 bit objects::
30237 * Using 64 bit storage pools by default::
30238 * General access types::
30239 * STARLET and other predefined libraries::
30242 @node Making code 64 bit clean
30243 @subsubsection Making code 64-bit clean
30246 In order to prevent problems that may occur when (parts of) a
30247 system start using memory outside the 32-bit address range,
30248 we recommend some additional guidelines:
30252 For imported subprograms that take parameters of the
30253 type @code{System.Address}, ensure that these subprograms can
30254 indeed handle 64-bit addresses. If not, or when in doubt,
30255 change the subprogram declaration to specify
30256 @code{System.Short_Address} instead.
30259 Resolve all warnings related to size mismatches in
30260 unchecked conversions. Failing to do so causes
30261 erroneous execution if the source object is outside
30262 the 32-bit address space.
30265 (optional) Explicitly use the 32-bit storage pool
30266 for access types used in a 32-bit context, or use
30267 generic access types where possible
30268 (@pxref{Restrictions on use of 64 bit objects}).
30272 If these rules are followed, the compiler will automatically insert
30273 any necessary checks to ensure that no addresses or access values
30274 passed to 32-bit code ever refer to objects outside the 32-bit
30276 Any attempt to do this will raise @code{Constraint_Error}.
30278 @node Allocating memory from the 64 bit storage pool
30279 @subsubsection Allocating memory from the 64-bit storage pool
30282 For any access type @code{T} that potentially requires memory allocations
30283 beyond the 32-bit address space,
30284 use the following representation clause:
30286 @smallexample @c ada
30287 for T'Storage_Pool use System.Pool_64;
30290 @node Restrictions on use of 64 bit objects
30291 @subsubsection Restrictions on use of 64-bit objects
30294 Taking the address of an object allocated from a 64-bit storage pool,
30295 and then passing this address to a subprogram expecting
30296 @code{System.Short_Address},
30297 or assigning it to a variable of type @code{Short_Address}, will cause
30298 @code{Constraint_Error} to be raised. In case the code is not 64-bit clean
30299 (@pxref{Making code 64 bit clean}), or checks are suppressed,
30300 no exception is raised and execution
30301 will become erroneous.
30303 @node Using 64 bit storage pools by default
30304 @subsubsection Using 64-bit storage pools by default
30307 In some cases it may be desirable to have the compiler allocate
30308 from 64-bit storage pools by default. This may be the case for
30309 libraries that are 64-bit clean, but may be used in both 32-bit
30310 and 64-bit contexts. For these cases the following configuration
30311 pragma may be specified:
30313 @smallexample @c ada
30314 pragma Pool_64_Default;
30318 Any code compiled in the context of this pragma will by default
30319 use the @code{System.Pool_64} storage pool. This default may be overridden
30320 for a specific access type @code{T} by the representation clause:
30322 @smallexample @c ada
30323 for T'Storage_Pool use System.Pool_32;
30327 Any object whose address may be passed to a subprogram with a
30328 @code{Short_Address} argument, or assigned to a variable of type
30329 @code{Short_Address}, needs to be allocated from this pool.
30331 @node General access types
30332 @subsubsection General access types
30335 Objects designated by access values from a
30336 general access type (declared with @code{access all}) are never allocated
30337 from a 64-bit storage pool. Code that uses general access types will
30338 accept objects allocated in either 32-bit or 64-bit address spaces,
30339 but never allocate objects outside the 32-bit address space.
30340 Using general access types ensures maximum compatibility with both
30341 32-bit and 64-bit code.
30343 @node STARLET and other predefined libraries
30344 @subsubsection STARLET and other predefined libraries
30347 All code that comes as part of GNAT is 64-bit clean, but the
30348 restrictions given in @ref{Restrictions on use of 64 bit objects},
30349 still apply. Look at the package
30350 specs to see in which contexts objects allocated
30351 in 64-bit address space are acceptable.
30353 @node Technical details
30354 @subsection Technical details
30357 64-bit @value{EDITION} for Open VMS takes advantage of the freedom given in the
30358 Ada standard with respect to the type of @code{System.Address}. Previous
30359 versions of GNAT Pro have defined this type as private and implemented it as a
30362 In order to allow defining @code{System.Short_Address} as a proper subtype,
30363 and to match the implicit sign extension in parameter passing,
30364 in 64-bit @value{EDITION} for Open VMS, @code{System.Address} is defined as a
30365 visible (i.e., non-private) integer type.
30366 Standard operations on the type, such as the binary operators ``+'', ``-'',
30367 etc., that take @code{Address} operands and return an @code{Address} result,
30368 have been hidden by declaring these
30369 @code{abstract}, a feature introduced in Ada 95 that helps avoid the potential
30370 ambiguities that would otherwise result from overloading.
30371 (Note that, although @code{Address} is a visible integer type,
30372 good programming practice dictates against exploiting the type's
30373 integer properties such as literals, since this will compromise
30376 Defining @code{Address} as a visible integer type helps achieve
30377 maximum compatibility for existing Ada code,
30378 without sacrificing the capabilities of the 64-bit architecture.
30381 @c ************************************************
30383 @node Microsoft Windows Topics
30384 @appendix Microsoft Windows Topics
30390 This chapter describes topics that are specific to the Microsoft Windows
30391 platforms (NT, 2000, and XP Professional).
30394 * Using GNAT on Windows::
30395 * Using a network installation of GNAT::
30396 * CONSOLE and WINDOWS subsystems::
30397 * Temporary Files::
30398 * Mixed-Language Programming on Windows::
30399 * Windows Calling Conventions::
30400 * Introduction to Dynamic Link Libraries (DLLs)::
30401 * Using DLLs with GNAT::
30402 * Building DLLs with GNAT::
30403 * Building DLLs with GNAT Project files::
30404 * Building DLLs with gnatdll::
30405 * GNAT and Windows Resources::
30406 * Debugging a DLL::
30407 * Setting Stack Size from gnatlink::
30408 * Setting Heap Size from gnatlink::
30411 @node Using GNAT on Windows
30412 @section Using GNAT on Windows
30415 One of the strengths of the GNAT technology is that its tool set
30416 (@command{gcc}, @command{gnatbind}, @command{gnatlink}, @command{gnatmake}, the
30417 @code{gdb} debugger, etc.) is used in the same way regardless of the
30420 On Windows this tool set is complemented by a number of Microsoft-specific
30421 tools that have been provided to facilitate interoperability with Windows
30422 when this is required. With these tools:
30427 You can build applications using the @code{CONSOLE} or @code{WINDOWS}
30431 You can use any Dynamically Linked Library (DLL) in your Ada code (both
30432 relocatable and non-relocatable DLLs are supported).
30435 You can build Ada DLLs for use in other applications. These applications
30436 can be written in a language other than Ada (e.g., C, C++, etc). Again both
30437 relocatable and non-relocatable Ada DLLs are supported.
30440 You can include Windows resources in your Ada application.
30443 You can use or create COM/DCOM objects.
30447 Immediately below are listed all known general GNAT-for-Windows restrictions.
30448 Other restrictions about specific features like Windows Resources and DLLs
30449 are listed in separate sections below.
30454 It is not possible to use @code{GetLastError} and @code{SetLastError}
30455 when tasking, protected records, or exceptions are used. In these
30456 cases, in order to implement Ada semantics, the GNAT run-time system
30457 calls certain Win32 routines that set the last error variable to 0 upon
30458 success. It should be possible to use @code{GetLastError} and
30459 @code{SetLastError} when tasking, protected record, and exception
30460 features are not used, but it is not guaranteed to work.
30463 It is not possible to link against Microsoft libraries except for
30464 import libraries. The library must be built to be compatible with
30465 @file{MSVCRT.LIB} (/MD Microsoft compiler option), @file{LIBC.LIB} and
30466 @file{LIBCMT.LIB} (/ML or /MT Microsoft compiler options) are known to
30467 not be compatible with the GNAT runtime. Even if the library is
30468 compatible with @file{MSVCRT.LIB} it is not guaranteed to work.
30471 When the compilation environment is located on FAT32 drives, users may
30472 experience recompilations of the source files that have not changed if
30473 Daylight Saving Time (DST) state has changed since the last time files
30474 were compiled. NTFS drives do not have this problem.
30477 No components of the GNAT toolset use any entries in the Windows
30478 registry. The only entries that can be created are file associations and
30479 PATH settings, provided the user has chosen to create them at installation
30480 time, as well as some minimal book-keeping information needed to correctly
30481 uninstall or integrate different GNAT products.
30484 @node Using a network installation of GNAT
30485 @section Using a network installation of GNAT
30488 Make sure the system on which GNAT is installed is accessible from the
30489 current machine, i.e., the install location is shared over the network.
30490 Shared resources are accessed on Windows by means of UNC paths, which
30491 have the format @code{\\server\sharename\path}
30493 In order to use such a network installation, simply add the UNC path of the
30494 @file{bin} directory of your GNAT installation in front of your PATH. For
30495 example, if GNAT is installed in @file{\GNAT} directory of a share location
30496 called @file{c-drive} on a machine @file{LOKI}, the following command will
30499 @code{@ @ @ path \\loki\c-drive\gnat\bin;%path%}
30501 Be aware that every compilation using the network installation results in the
30502 transfer of large amounts of data across the network and will likely cause
30503 serious performance penalty.
30505 @node CONSOLE and WINDOWS subsystems
30506 @section CONSOLE and WINDOWS subsystems
30507 @cindex CONSOLE Subsystem
30508 @cindex WINDOWS Subsystem
30512 There are two main subsystems under Windows. The @code{CONSOLE} subsystem
30513 (which is the default subsystem) will always create a console when
30514 launching the application. This is not something desirable when the
30515 application has a Windows GUI. To get rid of this console the
30516 application must be using the @code{WINDOWS} subsystem. To do so
30517 the @option{-mwindows} linker option must be specified.
30520 $ gnatmake winprog -largs -mwindows
30523 @node Temporary Files
30524 @section Temporary Files
30525 @cindex Temporary files
30528 It is possible to control where temporary files gets created by setting
30529 the @env{TMP} environment variable. The file will be created:
30532 @item Under the directory pointed to by the @env{TMP} environment variable if
30533 this directory exists.
30535 @item Under @file{c:\temp}, if the @env{TMP} environment variable is not
30536 set (or not pointing to a directory) and if this directory exists.
30538 @item Under the current working directory otherwise.
30542 This allows you to determine exactly where the temporary
30543 file will be created. This is particularly useful in networked
30544 environments where you may not have write access to some
30547 @node Mixed-Language Programming on Windows
30548 @section Mixed-Language Programming on Windows
30551 Developing pure Ada applications on Windows is no different than on
30552 other GNAT-supported platforms. However, when developing or porting an
30553 application that contains a mix of Ada and C/C++, the choice of your
30554 Windows C/C++ development environment conditions your overall
30555 interoperability strategy.
30557 If you use @command{gcc} to compile the non-Ada part of your application,
30558 there are no Windows-specific restrictions that affect the overall
30559 interoperability with your Ada code. If you plan to use
30560 Microsoft tools (e.g.@: Microsoft Visual C/C++), you should be aware of
30561 the following limitations:
30565 You cannot link your Ada code with an object or library generated with
30566 Microsoft tools if these use the @code{.tls} section (Thread Local
30567 Storage section) since the GNAT linker does not yet support this section.
30570 You cannot link your Ada code with an object or library generated with
30571 Microsoft tools if these use I/O routines other than those provided in
30572 the Microsoft DLL: @code{msvcrt.dll}. This is because the GNAT run time
30573 uses the services of @code{msvcrt.dll} for its I/Os. Use of other I/O
30574 libraries can cause a conflict with @code{msvcrt.dll} services. For
30575 instance Visual C++ I/O stream routines conflict with those in
30580 If you do want to use the Microsoft tools for your non-Ada code and hit one
30581 of the above limitations, you have two choices:
30585 Encapsulate your non-Ada code in a DLL to be linked with your Ada
30586 application. In this case, use the Microsoft or whatever environment to
30587 build the DLL and use GNAT to build your executable
30588 (@pxref{Using DLLs with GNAT}).
30591 Or you can encapsulate your Ada code in a DLL to be linked with the
30592 other part of your application. In this case, use GNAT to build the DLL
30593 (@pxref{Building DLLs with GNAT}) and use the Microsoft or whatever
30594 environment to build your executable.
30597 @node Windows Calling Conventions
30598 @section Windows Calling Conventions
30603 * C Calling Convention::
30604 * Stdcall Calling Convention::
30605 * Win32 Calling Convention::
30606 * DLL Calling Convention::
30610 When a subprogram @code{F} (caller) calls a subprogram @code{G}
30611 (callee), there are several ways to push @code{G}'s parameters on the
30612 stack and there are several possible scenarios to clean up the stack
30613 upon @code{G}'s return. A calling convention is an agreed upon software
30614 protocol whereby the responsibilities between the caller (@code{F}) and
30615 the callee (@code{G}) are clearly defined. Several calling conventions
30616 are available for Windows:
30620 @code{C} (Microsoft defined)
30623 @code{Stdcall} (Microsoft defined)
30626 @code{Win32} (GNAT specific)
30629 @code{DLL} (GNAT specific)
30632 @node C Calling Convention
30633 @subsection @code{C} Calling Convention
30636 This is the default calling convention used when interfacing to C/C++
30637 routines compiled with either @command{gcc} or Microsoft Visual C++.
30639 In the @code{C} calling convention subprogram parameters are pushed on the
30640 stack by the caller from right to left. The caller itself is in charge of
30641 cleaning up the stack after the call. In addition, the name of a routine
30642 with @code{C} calling convention is mangled by adding a leading underscore.
30644 The name to use on the Ada side when importing (or exporting) a routine
30645 with @code{C} calling convention is the name of the routine. For
30646 instance the C function:
30649 int get_val (long);
30653 should be imported from Ada as follows:
30655 @smallexample @c ada
30657 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
30658 pragma Import (C, Get_Val, External_Name => "get_val");
30663 Note that in this particular case the @code{External_Name} parameter could
30664 have been omitted since, when missing, this parameter is taken to be the
30665 name of the Ada entity in lower case. When the @code{Link_Name} parameter
30666 is missing, as in the above example, this parameter is set to be the
30667 @code{External_Name} with a leading underscore.
30669 When importing a variable defined in C, you should always use the @code{C}
30670 calling convention unless the object containing the variable is part of a
30671 DLL (in which case you should use the @code{Stdcall} calling
30672 convention, @pxref{Stdcall Calling Convention}).
30674 @node Stdcall Calling Convention
30675 @subsection @code{Stdcall} Calling Convention
30678 This convention, which was the calling convention used for Pascal
30679 programs, is used by Microsoft for all the routines in the Win32 API for
30680 efficiency reasons. It must be used to import any routine for which this
30681 convention was specified.
30683 In the @code{Stdcall} calling convention subprogram parameters are pushed
30684 on the stack by the caller from right to left. The callee (and not the
30685 caller) is in charge of cleaning the stack on routine exit. In addition,
30686 the name of a routine with @code{Stdcall} calling convention is mangled by
30687 adding a leading underscore (as for the @code{C} calling convention) and a
30688 trailing @code{@@}@code{@var{nn}}, where @var{nn} is the overall size (in
30689 bytes) of the parameters passed to the routine.
30691 The name to use on the Ada side when importing a C routine with a
30692 @code{Stdcall} calling convention is the name of the C routine. The leading
30693 underscore and trailing @code{@@}@code{@var{nn}} are added automatically by
30694 the compiler. For instance the Win32 function:
30697 @b{APIENTRY} int get_val (long);
30701 should be imported from Ada as follows:
30703 @smallexample @c ada
30705 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
30706 pragma Import (Stdcall, Get_Val);
30707 -- On the x86 a long is 4 bytes, so the Link_Name is "_get_val@@4"
30712 As for the @code{C} calling convention, when the @code{External_Name}
30713 parameter is missing, it is taken to be the name of the Ada entity in lower
30714 case. If instead of writing the above import pragma you write:
30716 @smallexample @c ada
30718 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
30719 pragma Import (Stdcall, Get_Val, External_Name => "retrieve_val");
30724 then the imported routine is @code{_retrieve_val@@4}. However, if instead
30725 of specifying the @code{External_Name} parameter you specify the
30726 @code{Link_Name} as in the following example:
30728 @smallexample @c ada
30730 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
30731 pragma Import (Stdcall, Get_Val, Link_Name => "retrieve_val");
30736 then the imported routine is @code{retrieve_val}, that is, there is no
30737 decoration at all. No leading underscore and no Stdcall suffix
30738 @code{@@}@code{@var{nn}}.
30741 This is especially important as in some special cases a DLL's entry
30742 point name lacks a trailing @code{@@}@code{@var{nn}} while the exported
30743 name generated for a call has it.
30746 It is also possible to import variables defined in a DLL by using an
30747 import pragma for a variable. As an example, if a DLL contains a
30748 variable defined as:
30755 then, to access this variable from Ada you should write:
30757 @smallexample @c ada
30759 My_Var : Interfaces.C.int;
30760 pragma Import (Stdcall, My_Var);
30765 Note that to ease building cross-platform bindings this convention
30766 will be handled as a @code{C} calling convention on non-Windows platforms.
30768 @node Win32 Calling Convention
30769 @subsection @code{Win32} Calling Convention
30772 This convention, which is GNAT-specific is fully equivalent to the
30773 @code{Stdcall} calling convention described above.
30775 @node DLL Calling Convention
30776 @subsection @code{DLL} Calling Convention
30779 This convention, which is GNAT-specific is fully equivalent to the
30780 @code{Stdcall} calling convention described above.
30782 @node Introduction to Dynamic Link Libraries (DLLs)
30783 @section Introduction to Dynamic Link Libraries (DLLs)
30787 A Dynamically Linked Library (DLL) is a library that can be shared by
30788 several applications running under Windows. A DLL can contain any number of
30789 routines and variables.
30791 One advantage of DLLs is that you can change and enhance them without
30792 forcing all the applications that depend on them to be relinked or
30793 recompiled. However, you should be aware than all calls to DLL routines are
30794 slower since, as you will understand below, such calls are indirect.
30796 To illustrate the remainder of this section, suppose that an application
30797 wants to use the services of a DLL @file{API.dll}. To use the services
30798 provided by @file{API.dll} you must statically link against the DLL or
30799 an import library which contains a jump table with an entry for each
30800 routine and variable exported by the DLL. In the Microsoft world this
30801 import library is called @file{API.lib}. When using GNAT this import
30802 library is called either @file{libAPI.dll.a}, @file{libapi.dll.a},
30803 @file{libAPI.a} or @file{libapi.a} (names are case insensitive).
30805 After you have linked your application with the DLL or the import library
30806 and you run your application, here is what happens:
30810 Your application is loaded into memory.
30813 The DLL @file{API.dll} is mapped into the address space of your
30814 application. This means that:
30818 The DLL will use the stack of the calling thread.
30821 The DLL will use the virtual address space of the calling process.
30824 The DLL will allocate memory from the virtual address space of the calling
30828 Handles (pointers) can be safely exchanged between routines in the DLL
30829 routines and routines in the application using the DLL.
30833 The entries in the jump table (from the import library @file{libAPI.dll.a}
30834 or @file{API.lib} or automatically created when linking against a DLL)
30835 which is part of your application are initialized with the addresses
30836 of the routines and variables in @file{API.dll}.
30839 If present in @file{API.dll}, routines @code{DllMain} or
30840 @code{DllMainCRTStartup} are invoked. These routines typically contain
30841 the initialization code needed for the well-being of the routines and
30842 variables exported by the DLL.
30846 There is an additional point which is worth mentioning. In the Windows
30847 world there are two kind of DLLs: relocatable and non-relocatable
30848 DLLs. Non-relocatable DLLs can only be loaded at a very specific address
30849 in the target application address space. If the addresses of two
30850 non-relocatable DLLs overlap and these happen to be used by the same
30851 application, a conflict will occur and the application will run
30852 incorrectly. Hence, when possible, it is always preferable to use and
30853 build relocatable DLLs. Both relocatable and non-relocatable DLLs are
30854 supported by GNAT. Note that the @option{-s} linker option (see GNU Linker
30855 User's Guide) removes the debugging symbols from the DLL but the DLL can
30856 still be relocated.
30858 As a side note, an interesting difference between Microsoft DLLs and
30859 Unix shared libraries, is the fact that on most Unix systems all public
30860 routines are exported by default in a Unix shared library, while under
30861 Windows it is possible (but not required) to list exported routines in
30862 a definition file (@pxref{The Definition File}).
30864 @node Using DLLs with GNAT
30865 @section Using DLLs with GNAT
30868 * Creating an Ada Spec for the DLL Services::
30869 * Creating an Import Library::
30873 To use the services of a DLL, say @file{API.dll}, in your Ada application
30878 The Ada spec for the routines and/or variables you want to access in
30879 @file{API.dll}. If not available this Ada spec must be built from the C/C++
30880 header files provided with the DLL.
30883 The import library (@file{libAPI.dll.a} or @file{API.lib}). As previously
30884 mentioned an import library is a statically linked library containing the
30885 import table which will be filled at load time to point to the actual
30886 @file{API.dll} routines. Sometimes you don't have an import library for the
30887 DLL you want to use. The following sections will explain how to build
30888 one. Note that this is optional.
30891 The actual DLL, @file{API.dll}.
30895 Once you have all the above, to compile an Ada application that uses the
30896 services of @file{API.dll} and whose main subprogram is @code{My_Ada_App},
30897 you simply issue the command
30900 $ gnatmake my_ada_app -largs -lAPI
30904 The argument @option{-largs -lAPI} at the end of the @command{gnatmake} command
30905 tells the GNAT linker to look first for a library named @file{API.lib}
30906 (Microsoft-style name) and if not found for a libraries named
30907 @file{libAPI.dll.a}, @file{API.dll.a} or @file{libAPI.a}.
30908 (GNAT-style name). Note that if the Ada package spec for @file{API.dll}
30909 contains the following pragma
30911 @smallexample @c ada
30912 pragma Linker_Options ("-lAPI");
30916 you do not have to add @option{-largs -lAPI} at the end of the
30917 @command{gnatmake} command.
30919 If any one of the items above is missing you will have to create it
30920 yourself. The following sections explain how to do so using as an
30921 example a fictitious DLL called @file{API.dll}.
30923 @node Creating an Ada Spec for the DLL Services
30924 @subsection Creating an Ada Spec for the DLL Services
30927 A DLL typically comes with a C/C++ header file which provides the
30928 definitions of the routines and variables exported by the DLL. The Ada
30929 equivalent of this header file is a package spec that contains definitions
30930 for the imported entities. If the DLL you intend to use does not come with
30931 an Ada spec you have to generate one such spec yourself. For example if
30932 the header file of @file{API.dll} is a file @file{api.h} containing the
30933 following two definitions:
30945 then the equivalent Ada spec could be:
30947 @smallexample @c ada
30950 with Interfaces.C.Strings;
30955 function Get (Str : C.Strings.Chars_Ptr) return C.int;
30958 pragma Import (C, Get);
30959 pragma Import (DLL, Some_Var);
30966 Note that a variable is
30967 @strong{always imported with a Stdcall convention}. A function
30968 can have @code{C} or @code{Stdcall} convention.
30969 (@pxref{Windows Calling Conventions}).
30971 @node Creating an Import Library
30972 @subsection Creating an Import Library
30973 @cindex Import library
30976 * The Definition File::
30977 * GNAT-Style Import Library::
30978 * Microsoft-Style Import Library::
30982 If a Microsoft-style import library @file{API.lib} or a GNAT-style
30983 import library @file{libAPI.dll.a} or @file{libAPI.a} is available
30984 with @file{API.dll} you can skip this section. You can also skip this
30985 section if @file{API.dll} or @file{libAPI.dll} is built with GNU tools
30986 as in this case it is possible to link directly against the
30987 DLL. Otherwise read on.
30989 @node The Definition File
30990 @subsubsection The Definition File
30991 @cindex Definition file
30995 As previously mentioned, and unlike Unix systems, the list of symbols
30996 that are exported from a DLL must be provided explicitly in Windows.
30997 The main goal of a definition file is precisely that: list the symbols
30998 exported by a DLL. A definition file (usually a file with a @code{.def}
30999 suffix) has the following structure:
31004 @r{[}LIBRARY @var{name}@r{]}
31005 @r{[}DESCRIPTION @var{string}@r{]}
31015 @item LIBRARY @var{name}
31016 This section, which is optional, gives the name of the DLL.
31018 @item DESCRIPTION @var{string}
31019 This section, which is optional, gives a description string that will be
31020 embedded in the import library.
31023 This section gives the list of exported symbols (procedures, functions or
31024 variables). For instance in the case of @file{API.dll} the @code{EXPORTS}
31025 section of @file{API.def} looks like:
31039 Note that you must specify the correct suffix (@code{@@}@code{@var{nn}})
31040 (@pxref{Windows Calling Conventions}) for a Stdcall
31041 calling convention function in the exported symbols list.
31044 There can actually be other sections in a definition file, but these
31045 sections are not relevant to the discussion at hand.
31047 @node GNAT-Style Import Library
31048 @subsubsection GNAT-Style Import Library
31051 To create a static import library from @file{API.dll} with the GNAT tools
31052 you should proceed as follows:
31056 Create the definition file @file{API.def} (@pxref{The Definition File}).
31057 For that use the @code{dll2def} tool as follows:
31060 $ dll2def API.dll > API.def
31064 @code{dll2def} is a very simple tool: it takes as input a DLL and prints
31065 to standard output the list of entry points in the DLL. Note that if
31066 some routines in the DLL have the @code{Stdcall} convention
31067 (@pxref{Windows Calling Conventions}) with stripped @code{@@}@var{nn}
31068 suffix then you'll have to edit @file{api.def} to add it, and specify
31069 @option{-k} to @command{gnatdll} when creating the import library.
31072 Here are some hints to find the right @code{@@}@var{nn} suffix.
31076 If you have the Microsoft import library (.lib), it is possible to get
31077 the right symbols by using Microsoft @code{dumpbin} tool (see the
31078 corresponding Microsoft documentation for further details).
31081 $ dumpbin /exports api.lib
31085 If you have a message about a missing symbol at link time the compiler
31086 tells you what symbol is expected. You just have to go back to the
31087 definition file and add the right suffix.
31091 Build the import library @code{libAPI.dll.a}, using @code{gnatdll}
31092 (@pxref{Using gnatdll}) as follows:
31095 $ gnatdll -e API.def -d API.dll
31099 @code{gnatdll} takes as input a definition file @file{API.def} and the
31100 name of the DLL containing the services listed in the definition file
31101 @file{API.dll}. The name of the static import library generated is
31102 computed from the name of the definition file as follows: if the
31103 definition file name is @var{xyz}@code{.def}, the import library name will
31104 be @code{lib}@var{xyz}@code{.a}. Note that in the previous example option
31105 @option{-e} could have been removed because the name of the definition
31106 file (before the ``@code{.def}'' suffix) is the same as the name of the
31107 DLL (@pxref{Using gnatdll} for more information about @code{gnatdll}).
31110 @node Microsoft-Style Import Library
31111 @subsubsection Microsoft-Style Import Library
31114 With GNAT you can either use a GNAT-style or Microsoft-style import
31115 library. A Microsoft import library is needed only if you plan to make an
31116 Ada DLL available to applications developed with Microsoft
31117 tools (@pxref{Mixed-Language Programming on Windows}).
31119 To create a Microsoft-style import library for @file{API.dll} you
31120 should proceed as follows:
31124 Create the definition file @file{API.def} from the DLL. For this use either
31125 the @code{dll2def} tool as described above or the Microsoft @code{dumpbin}
31126 tool (see the corresponding Microsoft documentation for further details).
31129 Build the actual import library using Microsoft's @code{lib} utility:
31132 $ lib -machine:IX86 -def:API.def -out:API.lib
31136 If you use the above command the definition file @file{API.def} must
31137 contain a line giving the name of the DLL:
31144 See the Microsoft documentation for further details about the usage of
31148 @node Building DLLs with GNAT
31149 @section Building DLLs with GNAT
31150 @cindex DLLs, building
31153 This section explain how to build DLLs using the GNAT built-in DLL
31154 support. With the following procedure it is straight forward to build
31155 and use DLLs with GNAT.
31159 @item building object files
31161 The first step is to build all objects files that are to be included
31162 into the DLL. This is done by using the standard @command{gnatmake} tool.
31164 @item building the DLL
31166 To build the DLL you must use @command{gcc}'s @option{-shared}
31167 option. It is quite simple to use this method:
31170 $ gcc -shared -o api.dll obj1.o obj2.o @dots{}
31173 It is important to note that in this case all symbols found in the
31174 object files are automatically exported. It is possible to restrict
31175 the set of symbols to export by passing to @command{gcc} a definition
31176 file, @pxref{The Definition File}. For example:
31179 $ gcc -shared -o api.dll api.def obj1.o obj2.o @dots{}
31182 If you use a definition file you must export the elaboration procedures
31183 for every package that required one. Elaboration procedures are named
31184 using the package name followed by "_E".
31186 @item preparing DLL to be used
31188 For the DLL to be used by client programs the bodies must be hidden
31189 from it and the .ali set with read-only attribute. This is very important
31190 otherwise GNAT will recompile all packages and will not actually use
31191 the code in the DLL. For example:
31195 $ copy *.ads *.ali api.dll apilib
31196 $ attrib +R apilib\*.ali
31201 At this point it is possible to use the DLL by directly linking
31202 against it. Note that you must use the GNAT shared runtime when using
31203 GNAT shared libraries. This is achieved by using @option{-shared} binder's
31207 $ gnatmake main -Iapilib -bargs -shared -largs -Lapilib -lAPI
31210 @node Building DLLs with GNAT Project files
31211 @section Building DLLs with GNAT Project files
31212 @cindex DLLs, building
31215 There is nothing specific to Windows in the build process.
31216 @pxref{Library Projects}.
31219 Due to a system limitation, it is not possible under Windows to create threads
31220 when inside the @code{DllMain} routine which is used for auto-initialization
31221 of shared libraries, so it is not possible to have library level tasks in SALs.
31223 @node Building DLLs with gnatdll
31224 @section Building DLLs with gnatdll
31225 @cindex DLLs, building
31228 * Limitations When Using Ada DLLs from Ada::
31229 * Exporting Ada Entities::
31230 * Ada DLLs and Elaboration::
31231 * Ada DLLs and Finalization::
31232 * Creating a Spec for Ada DLLs::
31233 * Creating the Definition File::
31238 Note that it is preferred to use the built-in GNAT DLL support
31239 (@pxref{Building DLLs with GNAT}) or GNAT Project files
31240 (@pxref{Building DLLs with GNAT Project files}) to build DLLs.
31242 This section explains how to build DLLs containing Ada code using
31243 @code{gnatdll}. These DLLs will be referred to as Ada DLLs in the
31244 remainder of this section.
31246 The steps required to build an Ada DLL that is to be used by Ada as well as
31247 non-Ada applications are as follows:
31251 You need to mark each Ada @i{entity} exported by the DLL with a @code{C} or
31252 @code{Stdcall} calling convention to avoid any Ada name mangling for the
31253 entities exported by the DLL (@pxref{Exporting Ada Entities}). You can
31254 skip this step if you plan to use the Ada DLL only from Ada applications.
31257 Your Ada code must export an initialization routine which calls the routine
31258 @code{adainit} generated by @command{gnatbind} to perform the elaboration of
31259 the Ada code in the DLL (@pxref{Ada DLLs and Elaboration}). The initialization
31260 routine exported by the Ada DLL must be invoked by the clients of the DLL
31261 to initialize the DLL.
31264 When useful, the DLL should also export a finalization routine which calls
31265 routine @code{adafinal} generated by @command{gnatbind} to perform the
31266 finalization of the Ada code in the DLL (@pxref{Ada DLLs and Finalization}).
31267 The finalization routine exported by the Ada DLL must be invoked by the
31268 clients of the DLL when the DLL services are no further needed.
31271 You must provide a spec for the services exported by the Ada DLL in each
31272 of the programming languages to which you plan to make the DLL available.
31275 You must provide a definition file listing the exported entities
31276 (@pxref{The Definition File}).
31279 Finally you must use @code{gnatdll} to produce the DLL and the import
31280 library (@pxref{Using gnatdll}).
31284 Note that a relocatable DLL stripped using the @code{strip}
31285 binutils tool will not be relocatable anymore. To build a DLL without
31286 debug information pass @code{-largs -s} to @code{gnatdll}. This
31287 restriction does not apply to a DLL built using a Library Project.
31288 @pxref{Library Projects}.
31290 @node Limitations When Using Ada DLLs from Ada
31291 @subsection Limitations When Using Ada DLLs from Ada
31294 When using Ada DLLs from Ada applications there is a limitation users
31295 should be aware of. Because on Windows the GNAT run time is not in a DLL of
31296 its own, each Ada DLL includes a part of the GNAT run time. Specifically,
31297 each Ada DLL includes the services of the GNAT run time that are necessary
31298 to the Ada code inside the DLL. As a result, when an Ada program uses an
31299 Ada DLL there are two independent GNAT run times: one in the Ada DLL and
31300 one in the main program.
31302 It is therefore not possible to exchange GNAT run-time objects between the
31303 Ada DLL and the main Ada program. Example of GNAT run-time objects are file
31304 handles (e.g.@: @code{Text_IO.File_Type}), tasks types, protected objects
31307 It is completely safe to exchange plain elementary, array or record types,
31308 Windows object handles, etc.
31310 @node Exporting Ada Entities
31311 @subsection Exporting Ada Entities
31312 @cindex Export table
31315 Building a DLL is a way to encapsulate a set of services usable from any
31316 application. As a result, the Ada entities exported by a DLL should be
31317 exported with the @code{C} or @code{Stdcall} calling conventions to avoid
31318 any Ada name mangling. As an example here is an Ada package
31319 @code{API}, spec and body, exporting two procedures, a function, and a
31322 @smallexample @c ada
31325 with Interfaces.C; use Interfaces;
31327 Count : C.int := 0;
31328 function Factorial (Val : C.int) return C.int;
31330 procedure Initialize_API;
31331 procedure Finalize_API;
31332 -- Initialization & Finalization routines. More in the next section.
31334 pragma Export (C, Initialize_API);
31335 pragma Export (C, Finalize_API);
31336 pragma Export (C, Count);
31337 pragma Export (C, Factorial);
31343 @smallexample @c ada
31346 package body API is
31347 function Factorial (Val : C.int) return C.int is
31350 Count := Count + 1;
31351 for K in 1 .. Val loop
31357 procedure Initialize_API is
31359 pragma Import (C, Adainit);
31362 end Initialize_API;
31364 procedure Finalize_API is
31365 procedure Adafinal;
31366 pragma Import (C, Adafinal);
31376 If the Ada DLL you are building will only be used by Ada applications
31377 you do not have to export Ada entities with a @code{C} or @code{Stdcall}
31378 convention. As an example, the previous package could be written as
31381 @smallexample @c ada
31385 Count : Integer := 0;
31386 function Factorial (Val : Integer) return Integer;
31388 procedure Initialize_API;
31389 procedure Finalize_API;
31390 -- Initialization and Finalization routines.
31396 @smallexample @c ada
31399 package body API is
31400 function Factorial (Val : Integer) return Integer is
31401 Fact : Integer := 1;
31403 Count := Count + 1;
31404 for K in 1 .. Val loop
31411 -- The remainder of this package body is unchanged.
31418 Note that if you do not export the Ada entities with a @code{C} or
31419 @code{Stdcall} convention you will have to provide the mangled Ada names
31420 in the definition file of the Ada DLL
31421 (@pxref{Creating the Definition File}).
31423 @node Ada DLLs and Elaboration
31424 @subsection Ada DLLs and Elaboration
31425 @cindex DLLs and elaboration
31428 The DLL that you are building contains your Ada code as well as all the
31429 routines in the Ada library that are needed by it. The first thing a
31430 user of your DLL must do is elaborate the Ada code
31431 (@pxref{Elaboration Order Handling in GNAT}).
31433 To achieve this you must export an initialization routine
31434 (@code{Initialize_API} in the previous example), which must be invoked
31435 before using any of the DLL services. This elaboration routine must call
31436 the Ada elaboration routine @code{adainit} generated by the GNAT binder
31437 (@pxref{Binding with Non-Ada Main Programs}). See the body of
31438 @code{Initialize_Api} for an example. Note that the GNAT binder is
31439 automatically invoked during the DLL build process by the @code{gnatdll}
31440 tool (@pxref{Using gnatdll}).
31442 When a DLL is loaded, Windows systematically invokes a routine called
31443 @code{DllMain}. It would therefore be possible to call @code{adainit}
31444 directly from @code{DllMain} without having to provide an explicit
31445 initialization routine. Unfortunately, it is not possible to call
31446 @code{adainit} from the @code{DllMain} if your program has library level
31447 tasks because access to the @code{DllMain} entry point is serialized by
31448 the system (that is, only a single thread can execute ``through'' it at a
31449 time), which means that the GNAT run time will deadlock waiting for the
31450 newly created task to complete its initialization.
31452 @node Ada DLLs and Finalization
31453 @subsection Ada DLLs and Finalization
31454 @cindex DLLs and finalization
31457 When the services of an Ada DLL are no longer needed, the client code should
31458 invoke the DLL finalization routine, if available. The DLL finalization
31459 routine is in charge of releasing all resources acquired by the DLL. In the
31460 case of the Ada code contained in the DLL, this is achieved by calling
31461 routine @code{adafinal} generated by the GNAT binder
31462 (@pxref{Binding with Non-Ada Main Programs}).
31463 See the body of @code{Finalize_Api} for an
31464 example. As already pointed out the GNAT binder is automatically invoked
31465 during the DLL build process by the @code{gnatdll} tool
31466 (@pxref{Using gnatdll}).
31468 @node Creating a Spec for Ada DLLs
31469 @subsection Creating a Spec for Ada DLLs
31472 To use the services exported by the Ada DLL from another programming
31473 language (e.g.@: C), you have to translate the specs of the exported Ada
31474 entities in that language. For instance in the case of @code{API.dll},
31475 the corresponding C header file could look like:
31480 extern int *_imp__count;
31481 #define count (*_imp__count)
31482 int factorial (int);
31488 It is important to understand that when building an Ada DLL to be used by
31489 other Ada applications, you need two different specs for the packages
31490 contained in the DLL: one for building the DLL and the other for using
31491 the DLL. This is because the @code{DLL} calling convention is needed to
31492 use a variable defined in a DLL, but when building the DLL, the variable
31493 must have either the @code{Ada} or @code{C} calling convention. As an
31494 example consider a DLL comprising the following package @code{API}:
31496 @smallexample @c ada
31500 Count : Integer := 0;
31502 -- Remainder of the package omitted.
31509 After producing a DLL containing package @code{API}, the spec that
31510 must be used to import @code{API.Count} from Ada code outside of the
31513 @smallexample @c ada
31518 pragma Import (DLL, Count);
31524 @node Creating the Definition File
31525 @subsection Creating the Definition File
31528 The definition file is the last file needed to build the DLL. It lists
31529 the exported symbols. As an example, the definition file for a DLL
31530 containing only package @code{API} (where all the entities are exported
31531 with a @code{C} calling convention) is:
31546 If the @code{C} calling convention is missing from package @code{API},
31547 then the definition file contains the mangled Ada names of the above
31548 entities, which in this case are:
31557 api__initialize_api
31562 @node Using gnatdll
31563 @subsection Using @code{gnatdll}
31567 * gnatdll Example::
31568 * gnatdll behind the Scenes::
31573 @code{gnatdll} is a tool to automate the DLL build process once all the Ada
31574 and non-Ada sources that make up your DLL have been compiled.
31575 @code{gnatdll} is actually in charge of two distinct tasks: build the
31576 static import library for the DLL and the actual DLL. The form of the
31577 @code{gnatdll} command is
31581 @c $ gnatdll @ovar{switches} @var{list-of-files} @r{[}-largs @var{opts}@r{]}
31582 @c Expanding @ovar macro inline (explanation in macro def comments)
31583 $ gnatdll @r{[}@var{switches}@r{]} @var{list-of-files} @r{[}-largs @var{opts}@r{]}
31588 where @var{list-of-files} is a list of ALI and object files. The object
31589 file list must be the exact list of objects corresponding to the non-Ada
31590 sources whose services are to be included in the DLL. The ALI file list
31591 must be the exact list of ALI files for the corresponding Ada sources
31592 whose services are to be included in the DLL. If @var{list-of-files} is
31593 missing, only the static import library is generated.
31596 You may specify any of the following switches to @code{gnatdll}:
31599 @c @item -a@ovar{address}
31600 @c Expanding @ovar macro inline (explanation in macro def comments)
31601 @item -a@r{[}@var{address}@r{]}
31602 @cindex @option{-a} (@code{gnatdll})
31603 Build a non-relocatable DLL at @var{address}. If @var{address} is not
31604 specified the default address @var{0x11000000} will be used. By default,
31605 when this switch is missing, @code{gnatdll} builds relocatable DLL. We
31606 advise the reader to build relocatable DLL.
31608 @item -b @var{address}
31609 @cindex @option{-b} (@code{gnatdll})
31610 Set the relocatable DLL base address. By default the address is
31613 @item -bargs @var{opts}
31614 @cindex @option{-bargs} (@code{gnatdll})
31615 Binder options. Pass @var{opts} to the binder.
31617 @item -d @var{dllfile}
31618 @cindex @option{-d} (@code{gnatdll})
31619 @var{dllfile} is the name of the DLL. This switch must be present for
31620 @code{gnatdll} to do anything. The name of the generated import library is
31621 obtained algorithmically from @var{dllfile} as shown in the following
31622 example: if @var{dllfile} is @code{xyz.dll}, the import library name is
31623 @code{libxyz.dll.a}. The name of the definition file to use (if not specified
31624 by option @option{-e}) is obtained algorithmically from @var{dllfile}
31625 as shown in the following example:
31626 if @var{dllfile} is @code{xyz.dll}, the definition
31627 file used is @code{xyz.def}.
31629 @item -e @var{deffile}
31630 @cindex @option{-e} (@code{gnatdll})
31631 @var{deffile} is the name of the definition file.
31634 @cindex @option{-g} (@code{gnatdll})
31635 Generate debugging information. This information is stored in the object
31636 file and copied from there to the final DLL file by the linker,
31637 where it can be read by the debugger. You must use the
31638 @option{-g} switch if you plan on using the debugger or the symbolic
31642 @cindex @option{-h} (@code{gnatdll})
31643 Help mode. Displays @code{gnatdll} switch usage information.
31646 @cindex @option{-I} (@code{gnatdll})
31647 Direct @code{gnatdll} to search the @var{dir} directory for source and
31648 object files needed to build the DLL.
31649 (@pxref{Search Paths and the Run-Time Library (RTL)}).
31652 @cindex @option{-k} (@code{gnatdll})
31653 Removes the @code{@@}@var{nn} suffix from the import library's exported
31654 names, but keeps them for the link names. You must specify this
31655 option if you want to use a @code{Stdcall} function in a DLL for which
31656 the @code{@@}@var{nn} suffix has been removed. This is the case for most
31657 of the Windows NT DLL for example. This option has no effect when
31658 @option{-n} option is specified.
31660 @item -l @var{file}
31661 @cindex @option{-l} (@code{gnatdll})
31662 The list of ALI and object files used to build the DLL are listed in
31663 @var{file}, instead of being given in the command line. Each line in
31664 @var{file} contains the name of an ALI or object file.
31667 @cindex @option{-n} (@code{gnatdll})
31668 No Import. Do not create the import library.
31671 @cindex @option{-q} (@code{gnatdll})
31672 Quiet mode. Do not display unnecessary messages.
31675 @cindex @option{-v} (@code{gnatdll})
31676 Verbose mode. Display extra information.
31678 @item -largs @var{opts}
31679 @cindex @option{-largs} (@code{gnatdll})
31680 Linker options. Pass @var{opts} to the linker.
31683 @node gnatdll Example
31684 @subsubsection @code{gnatdll} Example
31687 As an example the command to build a relocatable DLL from @file{api.adb}
31688 once @file{api.adb} has been compiled and @file{api.def} created is
31691 $ gnatdll -d api.dll api.ali
31695 The above command creates two files: @file{libapi.dll.a} (the import
31696 library) and @file{api.dll} (the actual DLL). If you want to create
31697 only the DLL, just type:
31700 $ gnatdll -d api.dll -n api.ali
31704 Alternatively if you want to create just the import library, type:
31707 $ gnatdll -d api.dll
31710 @node gnatdll behind the Scenes
31711 @subsubsection @code{gnatdll} behind the Scenes
31714 This section details the steps involved in creating a DLL. @code{gnatdll}
31715 does these steps for you. Unless you are interested in understanding what
31716 goes on behind the scenes, you should skip this section.
31718 We use the previous example of a DLL containing the Ada package @code{API},
31719 to illustrate the steps necessary to build a DLL. The starting point is a
31720 set of objects that will make up the DLL and the corresponding ALI
31721 files. In the case of this example this means that @file{api.o} and
31722 @file{api.ali} are available. To build a relocatable DLL, @code{gnatdll} does
31727 @code{gnatdll} builds the base file (@file{api.base}). A base file gives
31728 the information necessary to generate relocation information for the
31734 $ gnatlink api -o api.jnk -mdll -Wl,--base-file,api.base
31739 In addition to the base file, the @command{gnatlink} command generates an
31740 output file @file{api.jnk} which can be discarded. The @option{-mdll} switch
31741 asks @command{gnatlink} to generate the routines @code{DllMain} and
31742 @code{DllMainCRTStartup} that are called by the Windows loader when the DLL
31743 is loaded into memory.
31746 @code{gnatdll} uses @code{dlltool} (@pxref{Using dlltool}) to build the
31747 export table (@file{api.exp}). The export table contains the relocation
31748 information in a form which can be used during the final link to ensure
31749 that the Windows loader is able to place the DLL anywhere in memory.
31753 $ dlltool --dllname api.dll --def api.def --base-file api.base \
31754 --output-exp api.exp
31759 @code{gnatdll} builds the base file using the new export table. Note that
31760 @command{gnatbind} must be called once again since the binder generated file
31761 has been deleted during the previous call to @command{gnatlink}.
31766 $ gnatlink api -o api.jnk api.exp -mdll
31767 -Wl,--base-file,api.base
31772 @code{gnatdll} builds the new export table using the new base file and
31773 generates the DLL import library @file{libAPI.dll.a}.
31777 $ dlltool --dllname api.dll --def api.def --base-file api.base \
31778 --output-exp api.exp --output-lib libAPI.a
31783 Finally @code{gnatdll} builds the relocatable DLL using the final export
31789 $ gnatlink api api.exp -o api.dll -mdll
31794 @node Using dlltool
31795 @subsubsection Using @code{dlltool}
31798 @code{dlltool} is the low-level tool used by @code{gnatdll} to build
31799 DLLs and static import libraries. This section summarizes the most
31800 common @code{dlltool} switches. The form of the @code{dlltool} command
31804 @c $ dlltool @ovar{switches}
31805 @c Expanding @ovar macro inline (explanation in macro def comments)
31806 $ dlltool @r{[}@var{switches}@r{]}
31810 @code{dlltool} switches include:
31813 @item --base-file @var{basefile}
31814 @cindex @option{--base-file} (@command{dlltool})
31815 Read the base file @var{basefile} generated by the linker. This switch
31816 is used to create a relocatable DLL.
31818 @item --def @var{deffile}
31819 @cindex @option{--def} (@command{dlltool})
31820 Read the definition file.
31822 @item --dllname @var{name}
31823 @cindex @option{--dllname} (@command{dlltool})
31824 Gives the name of the DLL. This switch is used to embed the name of the
31825 DLL in the static import library generated by @code{dlltool} with switch
31826 @option{--output-lib}.
31829 @cindex @option{-k} (@command{dlltool})
31830 Kill @code{@@}@var{nn} from exported names
31831 (@pxref{Windows Calling Conventions}
31832 for a discussion about @code{Stdcall}-style symbols.
31835 @cindex @option{--help} (@command{dlltool})
31836 Prints the @code{dlltool} switches with a concise description.
31838 @item --output-exp @var{exportfile}
31839 @cindex @option{--output-exp} (@command{dlltool})
31840 Generate an export file @var{exportfile}. The export file contains the
31841 export table (list of symbols in the DLL) and is used to create the DLL.
31843 @item --output-lib @var{libfile}
31844 @cindex @option{--output-lib} (@command{dlltool})
31845 Generate a static import library @var{libfile}.
31848 @cindex @option{-v} (@command{dlltool})
31851 @item --as @var{assembler-name}
31852 @cindex @option{--as} (@command{dlltool})
31853 Use @var{assembler-name} as the assembler. The default is @code{as}.
31856 @node GNAT and Windows Resources
31857 @section GNAT and Windows Resources
31858 @cindex Resources, windows
31861 * Building Resources::
31862 * Compiling Resources::
31863 * Using Resources::
31867 Resources are an easy way to add Windows specific objects to your
31868 application. The objects that can be added as resources include:
31897 This section explains how to build, compile and use resources.
31899 @node Building Resources
31900 @subsection Building Resources
31901 @cindex Resources, building
31904 A resource file is an ASCII file. By convention resource files have an
31905 @file{.rc} extension.
31906 The easiest way to build a resource file is to use Microsoft tools
31907 such as @code{imagedit.exe} to build bitmaps, icons and cursors and
31908 @code{dlgedit.exe} to build dialogs.
31909 It is always possible to build an @file{.rc} file yourself by writing a
31912 It is not our objective to explain how to write a resource file. A
31913 complete description of the resource script language can be found in the
31914 Microsoft documentation.
31916 @node Compiling Resources
31917 @subsection Compiling Resources
31920 @cindex Resources, compiling
31923 This section describes how to build a GNAT-compatible (COFF) object file
31924 containing the resources. This is done using the Resource Compiler
31925 @code{windres} as follows:
31928 $ windres -i myres.rc -o myres.o
31932 By default @code{windres} will run @command{gcc} to preprocess the @file{.rc}
31933 file. You can specify an alternate preprocessor (usually named
31934 @file{cpp.exe}) using the @code{windres} @option{--preprocessor}
31935 parameter. A list of all possible options may be obtained by entering
31936 the command @code{windres} @option{--help}.
31938 It is also possible to use the Microsoft resource compiler @code{rc.exe}
31939 to produce a @file{.res} file (binary resource file). See the
31940 corresponding Microsoft documentation for further details. In this case
31941 you need to use @code{windres} to translate the @file{.res} file to a
31942 GNAT-compatible object file as follows:
31945 $ windres -i myres.res -o myres.o
31948 @node Using Resources
31949 @subsection Using Resources
31950 @cindex Resources, using
31953 To include the resource file in your program just add the
31954 GNAT-compatible object file for the resource(s) to the linker
31955 arguments. With @command{gnatmake} this is done by using the @option{-largs}
31959 $ gnatmake myprog -largs myres.o
31962 @node Debugging a DLL
31963 @section Debugging a DLL
31964 @cindex DLL debugging
31967 * Program and DLL Both Built with GCC/GNAT::
31968 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
31972 Debugging a DLL is similar to debugging a standard program. But
31973 we have to deal with two different executable parts: the DLL and the
31974 program that uses it. We have the following four possibilities:
31978 The program and the DLL are built with @code{GCC/GNAT}.
31980 The program is built with foreign tools and the DLL is built with
31983 The program is built with @code{GCC/GNAT} and the DLL is built with
31989 In this section we address only cases one and two above.
31990 There is no point in trying to debug
31991 a DLL with @code{GNU/GDB}, if there is no GDB-compatible debugging
31992 information in it. To do so you must use a debugger compatible with the
31993 tools suite used to build the DLL.
31995 @node Program and DLL Both Built with GCC/GNAT
31996 @subsection Program and DLL Both Built with GCC/GNAT
31999 This is the simplest case. Both the DLL and the program have @code{GDB}
32000 compatible debugging information. It is then possible to break anywhere in
32001 the process. Let's suppose here that the main procedure is named
32002 @code{ada_main} and that in the DLL there is an entry point named
32006 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) and
32007 program must have been built with the debugging information (see GNAT -g
32008 switch). Here are the step-by-step instructions for debugging it:
32011 @item Launch @code{GDB} on the main program.
32017 @item Start the program and stop at the beginning of the main procedure
32024 This step is required to be able to set a breakpoint inside the DLL. As long
32025 as the program is not run, the DLL is not loaded. This has the
32026 consequence that the DLL debugging information is also not loaded, so it is not
32027 possible to set a breakpoint in the DLL.
32029 @item Set a breakpoint inside the DLL
32032 (gdb) break ada_dll
32039 At this stage a breakpoint is set inside the DLL. From there on
32040 you can use the standard approach to debug the whole program
32041 (@pxref{Running and Debugging Ada Programs}).
32044 @c This used to work, probably because the DLLs were non-relocatable
32045 @c keep this section around until the problem is sorted out.
32047 To break on the @code{DllMain} routine it is not possible to follow
32048 the procedure above. At the time the program stop on @code{ada_main}
32049 the @code{DllMain} routine as already been called. Either you can use
32050 the procedure below @pxref{Debugging the DLL Directly} or this procedure:
32053 @item Launch @code{GDB} on the main program.
32059 @item Load DLL symbols
32062 (gdb) add-sym api.dll
32065 @item Set a breakpoint inside the DLL
32068 (gdb) break ada_dll.adb:45
32071 Note that at this point it is not possible to break using the routine symbol
32072 directly as the program is not yet running. The solution is to break
32073 on the proper line (break in @file{ada_dll.adb} line 45).
32075 @item Start the program
32084 @node Program Built with Foreign Tools and DLL Built with GCC/GNAT
32085 @subsection Program Built with Foreign Tools and DLL Built with GCC/GNAT
32088 * Debugging the DLL Directly::
32089 * Attaching to a Running Process::
32093 In this case things are slightly more complex because it is not possible to
32094 start the main program and then break at the beginning to load the DLL and the
32095 associated DLL debugging information. It is not possible to break at the
32096 beginning of the program because there is no @code{GDB} debugging information,
32097 and therefore there is no direct way of getting initial control. This
32098 section addresses this issue by describing some methods that can be used
32099 to break somewhere in the DLL to debug it.
32102 First suppose that the main procedure is named @code{main} (this is for
32103 example some C code built with Microsoft Visual C) and that there is a
32104 DLL named @code{test.dll} containing an Ada entry point named
32108 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) must have
32109 been built with debugging information (see GNAT -g option).
32111 @node Debugging the DLL Directly
32112 @subsubsection Debugging the DLL Directly
32116 Find out the executable starting address
32119 $ objdump --file-header main.exe
32122 The starting address is reported on the last line. For example:
32125 main.exe: file format pei-i386
32126 architecture: i386, flags 0x0000010a:
32127 EXEC_P, HAS_DEBUG, D_PAGED
32128 start address 0x00401010
32132 Launch the debugger on the executable.
32139 Set a breakpoint at the starting address, and launch the program.
32142 $ (gdb) break *0x00401010
32146 The program will stop at the given address.
32149 Set a breakpoint on a DLL subroutine.
32152 (gdb) break ada_dll.adb:45
32155 Or if you want to break using a symbol on the DLL, you need first to
32156 select the Ada language (language used by the DLL).
32159 (gdb) set language ada
32160 (gdb) break ada_dll
32164 Continue the program.
32171 This will run the program until it reaches the breakpoint that has been
32172 set. From that point you can use the standard way to debug a program
32173 as described in (@pxref{Running and Debugging Ada Programs}).
32178 It is also possible to debug the DLL by attaching to a running process.
32180 @node Attaching to a Running Process
32181 @subsubsection Attaching to a Running Process
32182 @cindex DLL debugging, attach to process
32185 With @code{GDB} it is always possible to debug a running process by
32186 attaching to it. It is possible to debug a DLL this way. The limitation
32187 of this approach is that the DLL must run long enough to perform the
32188 attach operation. It may be useful for instance to insert a time wasting
32189 loop in the code of the DLL to meet this criterion.
32193 @item Launch the main program @file{main.exe}.
32199 @item Use the Windows @i{Task Manager} to find the process ID. Let's say
32200 that the process PID for @file{main.exe} is 208.
32208 @item Attach to the running process to be debugged.
32214 @item Load the process debugging information.
32217 (gdb) symbol-file main.exe
32220 @item Break somewhere in the DLL.
32223 (gdb) break ada_dll
32226 @item Continue process execution.
32235 This last step will resume the process execution, and stop at
32236 the breakpoint we have set. From there you can use the standard
32237 approach to debug a program as described in
32238 (@pxref{Running and Debugging Ada Programs}).
32240 @node Setting Stack Size from gnatlink
32241 @section Setting Stack Size from @command{gnatlink}
32244 It is possible to specify the program stack size at link time. On modern
32245 versions of Windows, starting with XP, this is mostly useful to set the size of
32246 the main stack (environment task). The other task stacks are set with pragma
32247 Storage_Size or with the @command{gnatbind -d} command.
32249 Since older versions of Windows (2000, NT4, etc.) do not allow setting the
32250 reserve size of individual tasks, the link-time stack size applies to all
32251 tasks, and pragma Storage_Size has no effect.
32252 In particular, Stack Overflow checks are made against this
32253 link-time specified size.
32255 This setting can be done with
32256 @command{gnatlink} using either:
32260 @item using @option{-Xlinker} linker option
32263 $ gnatlink hello -Xlinker --stack=0x10000,0x1000
32266 This sets the stack reserve size to 0x10000 bytes and the stack commit
32267 size to 0x1000 bytes.
32269 @item using @option{-Wl} linker option
32272 $ gnatlink hello -Wl,--stack=0x1000000
32275 This sets the stack reserve size to 0x1000000 bytes. Note that with
32276 @option{-Wl} option it is not possible to set the stack commit size
32277 because the coma is a separator for this option.
32281 @node Setting Heap Size from gnatlink
32282 @section Setting Heap Size from @command{gnatlink}
32285 Under Windows systems, it is possible to specify the program heap size from
32286 @command{gnatlink} using either:
32290 @item using @option{-Xlinker} linker option
32293 $ gnatlink hello -Xlinker --heap=0x10000,0x1000
32296 This sets the heap reserve size to 0x10000 bytes and the heap commit
32297 size to 0x1000 bytes.
32299 @item using @option{-Wl} linker option
32302 $ gnatlink hello -Wl,--heap=0x1000000
32305 This sets the heap reserve size to 0x1000000 bytes. Note that with
32306 @option{-Wl} option it is not possible to set the heap commit size
32307 because the coma is a separator for this option.
32313 @c **********************************
32314 @c * GNU Free Documentation License *
32315 @c **********************************
32317 @c GNU Free Documentation License
32319 @node Index,,GNU Free Documentation License, Top
32325 @c Put table of contents at end, otherwise it precedes the "title page" in
32326 @c the .txt version
32327 @c Edit the pdf file to move the contents to the beginning, after the title