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 1.2 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
111 @settitle @value{EDITION} User's Guide @value{PLATFORM}
112 @dircategory GNU Ada tools
114 * @value{EDITION} User's Guide: (gnat_ugn). @value{PLATFORM}
117 @include gcc-common.texi
119 @setchapternewpage odd
124 @title @value{EDITION} User's Guide
128 @titlefont{@i{@value{PLATFORM}}}
134 @subtitle GNAT, The GNU Ada Compiler
139 @vskip 0pt plus 1filll
146 @node Top, About This Guide, (dir), (dir)
147 @top @value{EDITION} User's Guide
150 @value{EDITION} User's Guide @value{PLATFORM}
153 GNAT, The GNU Ada Compiler@*
154 GCC version @value{version-GCC}@*
161 * Getting Started with GNAT::
162 * The GNAT Compilation Model::
163 * Compiling Using gcc::
164 * Binding Using gnatbind::
165 * Linking Using gnatlink::
166 * The GNAT Make Program gnatmake::
167 * Improving Performance::
168 * Renaming Files Using gnatchop::
169 * Configuration Pragmas::
170 * Handling Arbitrary File Naming Conventions Using gnatname::
171 * GNAT Project Manager::
172 * The Cross-Referencing Tools gnatxref and gnatfind::
173 * The GNAT Pretty-Printer gnatpp::
174 * The GNAT Metric Tool gnatmetric::
175 * File Name Krunching Using gnatkr::
176 * Preprocessing Using gnatprep::
178 * The GNAT Run-Time Library Builder gnatlbr::
180 * The GNAT Library Browser gnatls::
181 * Cleaning Up Using gnatclean::
183 * GNAT and Libraries::
184 * Using the GNU make Utility::
186 * Memory Management Issues::
187 * Stack Related Facilities::
188 * Verifying Properties Using gnatcheck::
189 * Creating Sample Bodies Using gnatstub::
190 * Generating Ada Bindings for C and C++ headers::
191 * Other Utility Programs::
192 * Running and Debugging Ada Programs::
194 * Code Coverage and Profiling::
197 * Compatibility with HP Ada::
199 * Platform-Specific Information for the Run-Time Libraries::
200 * Example of Binder Output File::
201 * Elaboration Order Handling in GNAT::
202 * Conditional Compilation::
204 * Compatibility and Porting Guide::
206 * Microsoft Windows Topics::
208 * GNU Free Documentation License::
211 --- The Detailed Node Listing ---
215 * What This Guide Contains::
216 * What You Should Know before Reading This Guide::
217 * Related Information::
220 Getting Started with GNAT
223 * Running a Simple Ada Program::
224 * Running a Program with Multiple Units::
225 * Using the gnatmake Utility::
227 * Editing with Emacs::
230 * Introduction to GPS::
233 The GNAT Compilation Model
235 * Source Representation::
236 * Foreign Language Representation::
237 * File Naming Rules::
238 * Using Other File Names::
239 * Alternative File Naming Schemes::
240 * Generating Object Files::
241 * Source Dependencies::
242 * The Ada Library Information Files::
243 * Binding an Ada Program::
244 * Mixed Language Programming::
246 * Building Mixed Ada & C++ Programs::
247 * Comparison between GNAT and C/C++ Compilation Models::
249 * Comparison between GNAT and Conventional Ada Library Models::
251 * Placement of temporary files::
254 Foreign Language Representation
257 * Other 8-Bit Codes::
258 * Wide Character Encodings::
260 Compiling Ada Programs With gcc
262 * Compiling Programs::
264 * Search Paths and the Run-Time Library (RTL)::
265 * Order of Compilation Issues::
270 * Output and Error Message Control::
271 * Warning Message Control::
272 * Debugging and Assertion Control::
273 * Validity Checking::
276 * Using gcc for Syntax Checking::
277 * Using gcc for Semantic Checking::
278 * Compiling Different Versions of Ada::
279 * Character Set Control::
280 * File Naming Control::
281 * Subprogram Inlining Control::
282 * Auxiliary Output Control::
283 * Debugging Control::
284 * Exception Handling Control::
285 * Units to Sources Mapping Files::
286 * Integrated Preprocessing::
291 Binding Ada Programs With gnatbind
294 * Switches for gnatbind::
295 * Command-Line Access::
296 * Search Paths for gnatbind::
297 * Examples of gnatbind Usage::
299 Switches for gnatbind
301 * Consistency-Checking Modes::
302 * Binder Error Message Control::
303 * Elaboration Control::
305 * Binding with Non-Ada Main Programs::
306 * Binding Programs with No Main Subprogram::
308 Linking Using gnatlink
311 * Switches for gnatlink::
313 The GNAT Make Program gnatmake
316 * Switches for gnatmake::
317 * Mode Switches for gnatmake::
318 * Notes on the Command Line::
319 * How gnatmake Works::
320 * Examples of gnatmake Usage::
322 Improving Performance
323 * Performance Considerations::
324 * Text_IO Suggestions::
325 * Reducing Size of Ada Executables with gnatelim::
326 * Reducing Size of Executables with unused subprogram/data elimination::
328 Performance Considerations
329 * Controlling Run-Time Checks::
330 * Use of Restrictions::
331 * Optimization Levels::
332 * Debugging Optimized Code::
333 * Inlining of Subprograms::
334 * Other Optimization Switches::
335 * Optimization and Strict Aliasing::
337 * Coverage Analysis::
340 Reducing Size of Ada Executables with gnatelim
343 * Correcting the List of Eliminate Pragmas::
344 * Making Your Executables Smaller::
345 * Summary of the gnatelim Usage Cycle::
347 Reducing Size of Executables with unused subprogram/data elimination
348 * About unused subprogram/data elimination::
349 * Compilation options::
351 Renaming Files Using gnatchop
353 * Handling Files with Multiple Units::
354 * Operating gnatchop in Compilation Mode::
355 * Command Line for gnatchop::
356 * Switches for gnatchop::
357 * Examples of gnatchop Usage::
359 Configuration Pragmas
361 * Handling of Configuration Pragmas::
362 * The Configuration Pragmas Files::
364 Handling Arbitrary File Naming Conventions Using gnatname
366 * Arbitrary File Naming Conventions::
368 * Switches for gnatname::
369 * Examples of gnatname Usage::
374 * Examples of Project Files::
375 * Project File Syntax::
376 * Objects and Sources in Project Files::
377 * Importing Projects::
378 * Project Extension::
379 * Project Hierarchy Extension::
380 * External References in Project Files::
381 * Packages in Project Files::
382 * Variables from Imported Projects::
385 * Stand-alone Library Projects::
386 * Switches Related to Project Files::
387 * Tools Supporting Project Files::
388 * An Extended Example::
389 * Project File Complete Syntax::
391 The Cross-Referencing Tools gnatxref and gnatfind
393 * gnatxref Switches::
394 * gnatfind Switches::
395 * Project Files for gnatxref and gnatfind::
396 * Regular Expressions in gnatfind and gnatxref::
397 * Examples of gnatxref Usage::
398 * Examples of gnatfind Usage::
400 The GNAT Pretty-Printer gnatpp
402 * Switches for gnatpp::
405 The GNAT Metrics Tool gnatmetric
407 * Switches for gnatmetric::
409 File Name Krunching Using gnatkr
414 * Examples of gnatkr Usage::
416 Preprocessing Using gnatprep
417 * Preprocessing Symbols::
419 * Switches for gnatprep::
420 * Form of Definitions File::
421 * Form of Input Text for gnatprep::
424 The GNAT Run-Time Library Builder gnatlbr
427 * Switches for gnatlbr::
428 * Examples of gnatlbr Usage::
431 The GNAT Library Browser gnatls
434 * Switches for gnatls::
435 * Examples of gnatls Usage::
437 Cleaning Up Using gnatclean
439 * Running gnatclean::
440 * Switches for gnatclean::
441 @c * Examples of gnatclean Usage::
447 * Introduction to Libraries in GNAT::
448 * General Ada Libraries::
449 * Stand-alone Ada Libraries::
450 * Rebuilding the GNAT Run-Time Library::
452 Using the GNU make Utility
454 * Using gnatmake in a Makefile::
455 * Automatically Creating a List of Directories::
456 * Generating the Command Line Switches::
457 * Overcoming Command Line Length Limits::
460 Memory Management Issues
462 * Some Useful Memory Pools::
463 * The GNAT Debug Pool Facility::
468 Stack Related Facilities
470 * Stack Overflow Checking::
471 * Static Stack Usage Analysis::
472 * Dynamic Stack Usage Analysis::
474 Some Useful Memory Pools
476 The GNAT Debug Pool Facility
482 * Switches for gnatmem::
483 * Example of gnatmem Usage::
486 Verifying Properties Using gnatcheck
488 * Format of the Report File::
489 * General gnatcheck Switches::
490 * gnatcheck Rule Options::
491 * Adding the Results of Compiler Checks to gnatcheck Output::
492 * Project-Wide Checks::
495 Sample Bodies Using gnatstub
498 * Switches for gnatstub::
500 Other Utility Programs
502 * Using Other Utility Programs with GNAT::
503 * The External Symbol Naming Scheme of GNAT::
504 * Converting Ada Files to html with gnathtml::
507 Code Coverage and Profiling
509 * Code Coverage of Ada Programs using gcov::
510 * Profiling an Ada Program using gprof::
513 Running and Debugging Ada Programs
515 * The GNAT Debugger GDB::
517 * Introduction to GDB Commands::
518 * Using Ada Expressions::
519 * Calling User-Defined Subprograms::
520 * Using the Next Command in a Function::
523 * Debugging Generic Units::
524 * GNAT Abnormal Termination or Failure to Terminate::
525 * Naming Conventions for GNAT Source Files::
526 * Getting Internal Debugging Information::
534 Compatibility with HP Ada
536 * Ada Language Compatibility::
537 * Differences in the Definition of Package System::
538 * Language-Related Features::
539 * The Package STANDARD::
540 * The Package SYSTEM::
541 * Tasking and Task-Related Features::
542 * Pragmas and Pragma-Related Features::
543 * Library of Predefined Units::
545 * Main Program Definition::
546 * Implementation-Defined Attributes::
547 * Compiler and Run-Time Interfacing::
548 * Program Compilation and Library Management::
550 * Implementation Limits::
551 * Tools and Utilities::
553 Language-Related Features
555 * Integer Types and Representations::
556 * Floating-Point Types and Representations::
557 * Pragmas Float_Representation and Long_Float::
558 * Fixed-Point Types and Representations::
559 * Record and Array Component Alignment::
561 * Other Representation Clauses::
563 Tasking and Task-Related Features
565 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
566 * Assigning Task IDs::
567 * Task IDs and Delays::
568 * Task-Related Pragmas::
569 * Scheduling and Task Priority::
571 * External Interrupts::
573 Pragmas and Pragma-Related Features
575 * Restrictions on the Pragma INLINE::
576 * Restrictions on the Pragma INTERFACE::
577 * Restrictions on the Pragma SYSTEM_NAME::
579 Library of Predefined Units
581 * Changes to DECLIB::
585 * Shared Libraries and Options Files::
589 Platform-Specific Information for the Run-Time Libraries
591 * Summary of Run-Time Configurations::
592 * Specifying a Run-Time Library::
593 * Choosing the Scheduling Policy::
594 * Solaris-Specific Considerations::
595 * Linux-Specific Considerations::
596 * AIX-Specific Considerations::
597 * Irix-Specific Considerations::
599 Example of Binder Output File
601 Elaboration Order Handling in GNAT
604 * Checking the Elaboration Order::
605 * Controlling the Elaboration Order::
606 * Controlling Elaboration in GNAT - Internal Calls::
607 * Controlling Elaboration in GNAT - External Calls::
608 * Default Behavior in GNAT - Ensuring Safety::
609 * Treatment of Pragma Elaborate::
610 * Elaboration Issues for Library Tasks::
611 * Mixing Elaboration Models::
612 * What to Do If the Default Elaboration Behavior Fails::
613 * Elaboration for Access-to-Subprogram Values::
614 * Summary of Procedures for Elaboration Control::
615 * Other Elaboration Order Considerations::
617 Conditional Compilation
618 * Use of Boolean Constants::
619 * Debugging - A Special Case::
620 * Conditionalizing Declarations::
621 * Use of Alternative Implementations::
626 * Basic Assembler Syntax::
627 * A Simple Example of Inline Assembler::
628 * Output Variables in Inline Assembler::
629 * Input Variables in Inline Assembler::
630 * Inlining Inline Assembler Code::
631 * Other Asm Functionality::
633 Compatibility and Porting Guide
635 * Compatibility with Ada 83::
636 * Compatibility between Ada 95 and Ada 2005::
637 * Implementation-dependent characteristics::
639 @c This brief section is only in the non-VMS version
640 @c The complete chapter on HP Ada issues is in the VMS version
641 * Compatibility with HP Ada 83::
643 * Compatibility with Other Ada Systems::
644 * Representation Clauses::
646 * Transitioning to 64-Bit GNAT for OpenVMS::
650 Microsoft Windows Topics
652 * Using GNAT on Windows::
653 * CONSOLE and WINDOWS subsystems::
655 * Mixed-Language Programming on Windows::
656 * Windows Calling Conventions::
657 * Introduction to Dynamic Link Libraries (DLLs)::
658 * Using DLLs with GNAT::
659 * Building DLLs with GNAT::
660 * GNAT and Windows Resources::
662 * Setting Stack Size from gnatlink::
663 * Setting Heap Size from gnatlink::
670 @node About This Guide
671 @unnumbered About This Guide
675 This guide describes the use of @value{EDITION},
676 a compiler and software development toolset for the full Ada
677 programming language, implemented on OpenVMS for HP's Alpha and
678 Integrity server (I64) platforms.
681 This guide describes the use of @value{EDITION},
682 a compiler and software development
683 toolset for the full Ada programming language.
685 It documents the features of the compiler and tools, and explains
686 how to use them to build Ada applications.
688 @value{EDITION} implements Ada 95 and Ada 2005, and it may also be invoked in
689 Ada 83 compatibility mode.
690 By default, @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
691 but you can override with a compiler switch
692 (@pxref{Compiling Different Versions of Ada})
693 to explicitly specify the language version.
694 Throughout this manual, references to ``Ada'' without a year suffix
695 apply to both the Ada 95 and Ada 2005 versions of the language.
699 For ease of exposition, ``@value{EDITION}'' will be referred to simply as
700 ``GNAT'' in the remainder of this document.
707 * What This Guide Contains::
708 * What You Should Know before Reading This Guide::
709 * Related Information::
713 @node What This Guide Contains
714 @unnumberedsec What This Guide Contains
717 This guide contains the following chapters:
721 @ref{Getting Started with GNAT}, describes how to get started compiling
722 and running Ada programs with the GNAT Ada programming environment.
724 @ref{The GNAT Compilation Model}, describes the compilation model used
728 @ref{Compiling Using gcc}, describes how to compile
729 Ada programs with @command{gcc}, the Ada compiler.
732 @ref{Binding Using gnatbind}, describes how to
733 perform binding of Ada programs with @code{gnatbind}, the GNAT binding
737 @ref{Linking Using gnatlink},
738 describes @command{gnatlink}, a
739 program that provides for linking using the GNAT run-time library to
740 construct a program. @command{gnatlink} can also incorporate foreign language
741 object units into the executable.
744 @ref{The GNAT Make Program gnatmake}, describes @command{gnatmake}, a
745 utility that automatically determines the set of sources
746 needed by an Ada compilation unit, and executes the necessary compilations
750 @ref{Improving Performance}, shows various techniques for making your
751 Ada program run faster or take less space.
752 It discusses the effect of the compiler's optimization switch and
753 also describes the @command{gnatelim} tool and unused subprogram/data
757 @ref{Renaming Files Using gnatchop}, describes
758 @code{gnatchop}, a utility that allows you to preprocess a file that
759 contains Ada source code, and split it into one or more new files, one
760 for each compilation unit.
763 @ref{Configuration Pragmas}, describes the configuration pragmas
767 @ref{Handling Arbitrary File Naming Conventions Using gnatname},
768 shows how to override the default GNAT file naming conventions,
769 either for an individual unit or globally.
772 @ref{GNAT Project Manager}, describes how to use project files
773 to organize large projects.
776 @ref{The Cross-Referencing Tools gnatxref and gnatfind}, discusses
777 @code{gnatxref} and @code{gnatfind}, two tools that provide an easy
778 way to navigate through sources.
781 @ref{The GNAT Pretty-Printer gnatpp}, shows how to produce a reformatted
782 version of an Ada source file with control over casing, indentation,
783 comment placement, and other elements of program presentation style.
786 @ref{The GNAT Metric Tool gnatmetric}, shows how to compute various
787 metrics for an Ada source file, such as the number of types and subprograms,
788 and assorted complexity measures.
791 @ref{File Name Krunching Using gnatkr}, describes the @code{gnatkr}
792 file name krunching utility, used to handle shortened
793 file names on operating systems with a limit on the length of names.
796 @ref{Preprocessing Using gnatprep}, describes @code{gnatprep}, a
797 preprocessor utility that allows a single source file to be used to
798 generate multiple or parameterized source files by means of macro
803 @ref{The GNAT Run-Time Library Builder gnatlbr}, describes @command{gnatlbr},
804 a tool for rebuilding the GNAT run time with user-supplied
805 configuration pragmas.
809 @ref{The GNAT Library Browser gnatls}, describes @code{gnatls}, a
810 utility that displays information about compiled units, including dependences
811 on the corresponding sources files, and consistency of compilations.
814 @ref{Cleaning Up Using gnatclean}, describes @code{gnatclean}, a utility
815 to delete files that are produced by the compiler, binder and linker.
819 @ref{GNAT and Libraries}, describes the process of creating and using
820 Libraries with GNAT. It also describes how to recompile the GNAT run-time
824 @ref{Using the GNU make Utility}, describes some techniques for using
825 the GNAT toolset in Makefiles.
829 @ref{Memory Management Issues}, describes some useful predefined storage pools
830 and in particular the GNAT Debug Pool facility, which helps detect incorrect
833 It also describes @command{gnatmem}, a utility that monitors dynamic
834 allocation and deallocation and helps detect ``memory leaks''.
838 @ref{Stack Related Facilities}, describes some useful tools associated with
839 stack checking and analysis.
842 @ref{Verifying Properties Using gnatcheck}, discusses @code{gnatcheck},
843 a utility that checks Ada code against a set of rules.
846 @ref{Creating Sample Bodies Using gnatstub}, discusses @code{gnatstub},
847 a utility that generates empty but compilable bodies for library units.
850 @ref{Generating Ada Bindings for C and C++ headers}, describes how to
851 generate automatically Ada bindings from C and C++ headers.
854 @ref{Other Utility Programs}, discusses several other GNAT utilities,
855 including @code{gnathtml}.
859 @ref{Code Coverage and Profiling}, describes how to perform a structural
860 coverage and profile the execution of Ada programs.
864 @ref{Running and Debugging Ada Programs}, describes how to run and debug
869 @ref{Compatibility with HP Ada}, details the compatibility of GNAT with
870 HP Ada 83 @footnote{``HP Ada'' refers to the legacy product originally
871 developed by Digital Equipment Corporation and currently supported by HP.}
872 for OpenVMS Alpha. This product was formerly known as DEC Ada,
875 historical compatibility reasons, the relevant libraries still use the
880 @ref{Platform-Specific Information for the Run-Time Libraries},
881 describes the various run-time
882 libraries supported by GNAT on various platforms and explains how to
883 choose a particular library.
886 @ref{Example of Binder Output File}, shows the source code for the binder
887 output file for a sample program.
890 @ref{Elaboration Order Handling in GNAT}, describes how GNAT helps
891 you deal with elaboration order issues.
894 @ref{Conditional Compilation}, describes how to model conditional compilation,
895 both with Ada in general and with GNAT facilities in particular.
898 @ref{Inline Assembler}, shows how to use the inline assembly facility
902 @ref{Compatibility and Porting Guide}, contains sections on compatibility
903 of GNAT with other Ada development environments (including Ada 83 systems),
904 to assist in porting code from those environments.
908 @ref{Microsoft Windows Topics}, presents information relevant to the
909 Microsoft Windows platform.
913 @c *************************************************
914 @node What You Should Know before Reading This Guide
915 @c *************************************************
916 @unnumberedsec What You Should Know before Reading This Guide
918 @cindex Ada 95 Language Reference Manual
919 @cindex Ada 2005 Language Reference Manual
921 This guide assumes a basic familiarity with the Ada 95 language, as
922 described in the International Standard ANSI/ISO/IEC-8652:1995, January
924 It does not require knowledge of the new features introduced by Ada 2005,
925 (officially known as ISO/IEC 8652:1995 with Technical Corrigendum 1
927 Both reference manuals are included in the GNAT documentation
930 @node Related Information
931 @unnumberedsec Related Information
934 For further information about related tools, refer to the following
939 @xref{Top, GNAT Reference Manual, About This Guide, gnat_rm, GNAT
940 Reference Manual}, which contains all reference material for the GNAT
941 implementation of Ada.
945 @cite{Using the GNAT Programming Studio}, which describes the GPS
946 Integrated Development Environment.
949 @cite{GNAT Programming Studio Tutorial}, which introduces the
950 main GPS features through examples.
954 @cite{Ada 95 Reference Manual}, which contains reference
955 material for the Ada 95 programming language.
958 @cite{Ada 2005 Reference Manual}, which contains reference
959 material for the Ada 2005 programming language.
962 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
964 in the GNU:[DOCS] directory,
966 for all details on the use of the GNU source-level debugger.
969 @xref{Top,, The extensible self-documenting text editor, emacs,
972 located in the GNU:[DOCS] directory if the EMACS kit is installed,
974 for full information on the extensible editor and programming
981 @unnumberedsec Conventions
983 @cindex Typographical conventions
986 Following are examples of the typographical and graphic conventions used
991 @code{Functions}, @command{utility program names}, @code{standard names},
995 @option{Option flags}
998 @file{File names}, @samp{button names}, and @samp{field names}.
1001 @code{Variables}, @env{environment variables}, and @var{metasyntactic
1008 @r{[}optional information or parameters@r{]}
1011 Examples are described by text
1013 and then shown this way.
1018 Commands that are entered by the user are preceded in this manual by the
1019 characters @w{``@code{$ }''} (dollar sign followed by space). If your system
1020 uses this sequence as a prompt, then the commands will appear exactly as
1021 you see them in the manual. If your system uses some other prompt, then
1022 the command will appear with the @code{$} replaced by whatever prompt
1023 character you are using.
1026 Full file names are shown with the ``@code{/}'' character
1027 as the directory separator; e.g., @file{parent-dir/subdir/myfile.adb}.
1028 If you are using GNAT on a Windows platform, please note that
1029 the ``@code{\}'' character should be used instead.
1032 @c ****************************
1033 @node Getting Started with GNAT
1034 @chapter Getting Started with GNAT
1037 This chapter describes some simple ways of using GNAT to build
1038 executable Ada programs.
1040 @ref{Running GNAT}, through @ref{Using the gnatmake Utility},
1041 show how to use the command line environment.
1042 @ref{Introduction to GPS}, provides a brief
1043 introduction to the GNAT Programming Studio, a visually-oriented
1044 Integrated Development Environment for GNAT.
1045 GPS offers a graphical ``look and feel'', support for development in
1046 other programming languages, comprehensive browsing features, and
1047 many other capabilities.
1048 For information on GPS please refer to
1049 @cite{Using the GNAT Programming Studio}.
1054 * Running a Simple Ada Program::
1055 * Running a Program with Multiple Units::
1056 * Using the gnatmake Utility::
1058 * Editing with Emacs::
1061 * Introduction to GPS::
1066 @section Running GNAT
1069 Three steps are needed to create an executable file from an Ada source
1074 The source file(s) must be compiled.
1076 The file(s) must be bound using the GNAT binder.
1078 All appropriate object files must be linked to produce an executable.
1082 All three steps are most commonly handled by using the @command{gnatmake}
1083 utility program that, given the name of the main program, automatically
1084 performs the necessary compilation, binding and linking steps.
1086 @node Running a Simple Ada Program
1087 @section Running a Simple Ada Program
1090 Any text editor may be used to prepare an Ada program.
1092 used, the optional Ada mode may be helpful in laying out the program.)
1094 program text is a normal text file. We will assume in our initial
1095 example that you have used your editor to prepare the following
1096 standard format text file:
1098 @smallexample @c ada
1100 with Ada.Text_IO; use Ada.Text_IO;
1103 Put_Line ("Hello WORLD!");
1109 This file should be named @file{hello.adb}.
1110 With the normal default file naming conventions, GNAT requires
1112 contain a single compilation unit whose file name is the
1114 with periods replaced by hyphens; the
1115 extension is @file{ads} for a
1116 spec and @file{adb} for a body.
1117 You can override this default file naming convention by use of the
1118 special pragma @code{Source_File_Name} (@pxref{Using Other File Names}).
1119 Alternatively, if you want to rename your files according to this default
1120 convention, which is probably more convenient if you will be using GNAT
1121 for all your compilations, then the @code{gnatchop} utility
1122 can be used to generate correctly-named source files
1123 (@pxref{Renaming Files Using gnatchop}).
1125 You can compile the program using the following command (@code{$} is used
1126 as the command prompt in the examples in this document):
1133 @command{gcc} is the command used to run the compiler. This compiler is
1134 capable of compiling programs in several languages, including Ada and
1135 C. It assumes that you have given it an Ada program if the file extension is
1136 either @file{.ads} or @file{.adb}, and it will then call
1137 the GNAT compiler to compile the specified file.
1140 The @option{-c} switch is required. It tells @command{gcc} to only do a
1141 compilation. (For C programs, @command{gcc} can also do linking, but this
1142 capability is not used directly for Ada programs, so the @option{-c}
1143 switch must always be present.)
1146 This compile command generates a file
1147 @file{hello.o}, which is the object
1148 file corresponding to your Ada program. It also generates
1149 an ``Ada Library Information'' file @file{hello.ali},
1150 which contains additional information used to check
1151 that an Ada program is consistent.
1152 To build an executable file,
1153 use @code{gnatbind} to bind the program
1154 and @command{gnatlink} to link it. The
1155 argument to both @code{gnatbind} and @command{gnatlink} is the name of the
1156 @file{ALI} file, but the default extension of @file{.ali} can
1157 be omitted. This means that in the most common case, the argument
1158 is simply the name of the main program:
1166 A simpler method of carrying out these steps is to use
1168 a master program that invokes all the required
1169 compilation, binding and linking tools in the correct order. In particular,
1170 @command{gnatmake} automatically recompiles any sources that have been
1171 modified since they were last compiled, or sources that depend
1172 on such modified sources, so that ``version skew'' is avoided.
1173 @cindex Version skew (avoided by @command{gnatmake})
1176 $ gnatmake hello.adb
1180 The result is an executable program called @file{hello}, which can be
1188 assuming that the current directory is on the search path
1189 for executable programs.
1192 and, if all has gone well, you will see
1199 appear in response to this command.
1201 @c ****************************************
1202 @node Running a Program with Multiple Units
1203 @section Running a Program with Multiple Units
1206 Consider a slightly more complicated example that has three files: a
1207 main program, and the spec and body of a package:
1209 @smallexample @c ada
1212 package Greetings is
1217 with Ada.Text_IO; use Ada.Text_IO;
1218 package body Greetings is
1221 Put_Line ("Hello WORLD!");
1224 procedure Goodbye is
1226 Put_Line ("Goodbye WORLD!");
1243 Following the one-unit-per-file rule, place this program in the
1244 following three separate files:
1248 spec of package @code{Greetings}
1251 body of package @code{Greetings}
1254 body of main program
1258 To build an executable version of
1259 this program, we could use four separate steps to compile, bind, and link
1260 the program, as follows:
1264 $ gcc -c greetings.adb
1270 Note that there is no required order of compilation when using GNAT.
1271 In particular it is perfectly fine to compile the main program first.
1272 Also, it is not necessary to compile package specs in the case where
1273 there is an accompanying body; you only need to compile the body. If you want
1274 to submit these files to the compiler for semantic checking and not code
1275 generation, then use the
1276 @option{-gnatc} switch:
1279 $ gcc -c greetings.ads -gnatc
1283 Although the compilation can be done in separate steps as in the
1284 above example, in practice it is almost always more convenient
1285 to use the @command{gnatmake} tool. All you need to know in this case
1286 is the name of the main program's source file. The effect of the above four
1287 commands can be achieved with a single one:
1290 $ gnatmake gmain.adb
1294 In the next section we discuss the advantages of using @command{gnatmake} in
1297 @c *****************************
1298 @node Using the gnatmake Utility
1299 @section Using the @command{gnatmake} Utility
1302 If you work on a program by compiling single components at a time using
1303 @command{gcc}, you typically keep track of the units you modify. In order to
1304 build a consistent system, you compile not only these units, but also any
1305 units that depend on the units you have modified.
1306 For example, in the preceding case,
1307 if you edit @file{gmain.adb}, you only need to recompile that file. But if
1308 you edit @file{greetings.ads}, you must recompile both
1309 @file{greetings.adb} and @file{gmain.adb}, because both files contain
1310 units that depend on @file{greetings.ads}.
1312 @code{gnatbind} will warn you if you forget one of these compilation
1313 steps, so that it is impossible to generate an inconsistent program as a
1314 result of forgetting to do a compilation. Nevertheless it is tedious and
1315 error-prone to keep track of dependencies among units.
1316 One approach to handle the dependency-bookkeeping is to use a
1317 makefile. However, makefiles present maintenance problems of their own:
1318 if the dependencies change as you change the program, you must make
1319 sure that the makefile is kept up-to-date manually, which is also an
1320 error-prone process.
1322 The @command{gnatmake} utility takes care of these details automatically.
1323 Invoke it using either one of the following forms:
1326 $ gnatmake gmain.adb
1327 $ gnatmake ^gmain^GMAIN^
1331 The argument is the name of the file containing the main program;
1332 you may omit the extension. @command{gnatmake}
1333 examines the environment, automatically recompiles any files that need
1334 recompiling, and binds and links the resulting set of object files,
1335 generating the executable file, @file{^gmain^GMAIN.EXE^}.
1336 In a large program, it
1337 can be extremely helpful to use @command{gnatmake}, because working out by hand
1338 what needs to be recompiled can be difficult.
1340 Note that @command{gnatmake}
1341 takes into account all the Ada rules that
1342 establish dependencies among units. These include dependencies that result
1343 from inlining subprogram bodies, and from
1344 generic instantiation. Unlike some other
1345 Ada make tools, @command{gnatmake} does not rely on the dependencies that were
1346 found by the compiler on a previous compilation, which may possibly
1347 be wrong when sources change. @command{gnatmake} determines the exact set of
1348 dependencies from scratch each time it is run.
1351 @node Editing with Emacs
1352 @section Editing with Emacs
1356 Emacs is an extensible self-documenting text editor that is available in a
1357 separate VMSINSTAL kit.
1359 Invoke Emacs by typing @kbd{Emacs} at the command prompt. To get started,
1360 click on the Emacs Help menu and run the Emacs Tutorial.
1361 In a character cell terminal, Emacs help is invoked with @kbd{Ctrl-h} (also
1362 written as @kbd{C-h}), and the tutorial by @kbd{C-h t}.
1364 Documentation on Emacs and other tools is available in Emacs under the
1365 pull-down menu button: @code{Help - Info}. After selecting @code{Info},
1366 use the middle mouse button to select a topic (e.g.@: Emacs).
1368 In a character cell terminal, do @kbd{C-h i} to invoke info, and then @kbd{m}
1369 (stands for menu) followed by the menu item desired, as in @kbd{m Emacs}, to
1370 get to the Emacs manual.
1371 Help on Emacs is also available by typing @kbd{HELP EMACS} at the DCL command
1374 The tutorial is highly recommended in order to learn the intricacies of Emacs,
1375 which is sufficiently extensible to provide for a complete programming
1376 environment and shell for the sophisticated user.
1380 @node Introduction to GPS
1381 @section Introduction to GPS
1382 @cindex GPS (GNAT Programming Studio)
1383 @cindex GNAT Programming Studio (GPS)
1385 Although the command line interface (@command{gnatmake}, etc.) alone
1386 is sufficient, a graphical Interactive Development
1387 Environment can make it easier for you to compose, navigate, and debug
1388 programs. This section describes the main features of GPS
1389 (``GNAT Programming Studio''), the GNAT graphical IDE.
1390 You will see how to use GPS to build and debug an executable, and
1391 you will also learn some of the basics of the GNAT ``project'' facility.
1393 GPS enables you to do much more than is presented here;
1394 e.g., you can produce a call graph, interface to a third-party
1395 Version Control System, and inspect the generated assembly language
1397 Indeed, GPS also supports languages other than Ada.
1398 Such additional information, and an explanation of all of the GPS menu
1399 items. may be found in the on-line help, which includes
1400 a user's guide and a tutorial (these are also accessible from the GNAT
1404 * Building a New Program with GPS::
1405 * Simple Debugging with GPS::
1408 @node Building a New Program with GPS
1409 @subsection Building a New Program with GPS
1411 GPS invokes the GNAT compilation tools using information
1412 contained in a @emph{project} (also known as a @emph{project file}):
1413 a collection of properties such
1414 as source directories, identities of main subprograms, tool switches, etc.,
1415 and their associated values.
1416 See @ref{GNAT Project Manager} for details.
1417 In order to run GPS, you will need to either create a new project
1418 or else open an existing one.
1420 This section will explain how you can use GPS to create a project,
1421 to associate Ada source files with a project, and to build and run
1425 @item @emph{Creating a project}
1427 Invoke GPS, either from the command line or the platform's IDE.
1428 After it starts, GPS will display a ``Welcome'' screen with three
1433 @code{Start with default project in directory}
1436 @code{Create new project with wizard}
1439 @code{Open existing project}
1443 Select @code{Create new project with wizard} and press @code{OK}.
1444 A new window will appear. In the text box labeled with
1445 @code{Enter the name of the project to create}, type @file{sample}
1446 as the project name.
1447 In the next box, browse to choose the directory in which you
1448 would like to create the project file.
1449 After selecting an appropriate directory, press @code{Forward}.
1451 A window will appear with the title
1452 @code{Version Control System Configuration}.
1453 Simply press @code{Forward}.
1455 A window will appear with the title
1456 @code{Please select the source directories for this project}.
1457 The directory that you specified for the project file will be selected
1458 by default as the one to use for sources; simply press @code{Forward}.
1460 A window will appear with the title
1461 @code{Please select the build directory for this project}.
1462 The directory that you specified for the project file will be selected
1463 by default for object files and executables;
1464 simply press @code{Forward}.
1466 A window will appear with the title
1467 @code{Please select the main units for this project}.
1468 You will supply this information later, after creating the source file.
1469 Simply press @code{Forward} for now.
1471 A window will appear with the title
1472 @code{Please select the switches to build the project}.
1473 Press @code{Apply}. This will create a project file named
1474 @file{sample.prj} in the directory that you had specified.
1476 @item @emph{Creating and saving the source file}
1478 After you create the new project, a GPS window will appear, which is
1479 partitioned into two main sections:
1483 A @emph{Workspace area}, initially greyed out, which you will use for
1484 creating and editing source files
1487 Directly below, a @emph{Messages area}, which initially displays a
1488 ``Welcome'' message.
1489 (If the Messages area is not visible, drag its border upward to expand it.)
1493 Select @code{File} on the menu bar, and then the @code{New} command.
1494 The Workspace area will become white, and you can now
1495 enter the source program explicitly.
1496 Type the following text
1498 @smallexample @c ada
1500 with Ada.Text_IO; use Ada.Text_IO;
1503 Put_Line("Hello from GPS!");
1509 Select @code{File}, then @code{Save As}, and enter the source file name
1511 The file will be saved in the same directory you specified as the
1512 location of the default project file.
1514 @item @emph{Updating the project file}
1516 You need to add the new source file to the project.
1518 the @code{Project} menu and then @code{Edit project properties}.
1519 Click the @code{Main files} tab on the left, and then the
1521 Choose @file{hello.adb} from the list, and press @code{Open}.
1522 The project settings window will reflect this action.
1525 @item @emph{Building and running the program}
1527 In the main GPS window, now choose the @code{Build} menu, then @code{Make},
1528 and select @file{hello.adb}.
1529 The Messages window will display the resulting invocations of @command{gcc},
1530 @command{gnatbind}, and @command{gnatlink}
1531 (reflecting the default switch settings from the
1532 project file that you created) and then a ``successful compilation/build''
1535 To run the program, choose the @code{Build} menu, then @code{Run}, and
1536 select @command{hello}.
1537 An @emph{Arguments Selection} window will appear.
1538 There are no command line arguments, so just click @code{OK}.
1540 The Messages window will now display the program's output (the string
1541 @code{Hello from GPS}), and at the bottom of the GPS window a status
1542 update is displayed (@code{Run: hello}).
1543 Close the GPS window (or select @code{File}, then @code{Exit}) to
1544 terminate this GPS session.
1547 @node Simple Debugging with GPS
1548 @subsection Simple Debugging with GPS
1550 This section illustrates basic debugging techniques (setting breakpoints,
1551 examining/modifying variables, single stepping).
1554 @item @emph{Opening a project}
1556 Start GPS and select @code{Open existing project}; browse to
1557 specify the project file @file{sample.prj} that you had created in the
1560 @item @emph{Creating a source file}
1562 Select @code{File}, then @code{New}, and type in the following program:
1564 @smallexample @c ada
1566 with Ada.Text_IO; use Ada.Text_IO;
1567 procedure Example is
1568 Line : String (1..80);
1571 Put_Line("Type a line of text at each prompt; an empty line to exit");
1575 Put_Line (Line (1..N) );
1583 Select @code{File}, then @code{Save as}, and enter the file name
1586 @item @emph{Updating the project file}
1588 Add @code{Example} as a new main unit for the project:
1591 Select @code{Project}, then @code{Edit Project Properties}.
1594 Select the @code{Main files} tab, click @code{Add}, then
1595 select the file @file{example.adb} from the list, and
1597 You will see the file name appear in the list of main units
1603 @item @emph{Building/running the executable}
1605 To build the executable
1606 select @code{Build}, then @code{Make}, and then choose @file{example.adb}.
1608 Run the program to see its effect (in the Messages area).
1609 Each line that you enter is displayed; an empty line will
1610 cause the loop to exit and the program to terminate.
1612 @item @emph{Debugging the program}
1614 Note that the @option{-g} switches to @command{gcc} and @command{gnatlink},
1615 which are required for debugging, are on by default when you create
1617 Thus unless you intentionally remove these settings, you will be able
1618 to debug any program that you develop using GPS.
1621 @item @emph{Initializing}
1623 Select @code{Debug}, then @code{Initialize}, then @file{example}
1625 @item @emph{Setting a breakpoint}
1627 After performing the initialization step, you will observe a small
1628 icon to the right of each line number.
1629 This serves as a toggle for breakpoints; clicking the icon will
1630 set a breakpoint at the corresponding line (the icon will change to
1631 a red circle with an ``x''), and clicking it again
1632 will remove the breakpoint / reset the icon.
1634 For purposes of this example, set a breakpoint at line 10 (the
1635 statement @code{Put_Line@ (Line@ (1..N));}
1637 @item @emph{Starting program execution}
1639 Select @code{Debug}, then @code{Run}. When the
1640 @code{Program Arguments} window appears, click @code{OK}.
1641 A console window will appear; enter some line of text,
1642 e.g.@: @code{abcde}, at the prompt.
1643 The program will pause execution when it gets to the
1644 breakpoint, and the corresponding line is highlighted.
1646 @item @emph{Examining a variable}
1648 Move the mouse over one of the occurrences of the variable @code{N}.
1649 You will see the value (5) displayed, in ``tool tip'' fashion.
1650 Right click on @code{N}, select @code{Debug}, then select @code{Display N}.
1651 You will see information about @code{N} appear in the @code{Debugger Data}
1652 pane, showing the value as 5.
1654 @item @emph{Assigning a new value to a variable}
1656 Right click on the @code{N} in the @code{Debugger Data} pane, and
1657 select @code{Set value of N}.
1658 When the input window appears, enter the value @code{4} and click
1660 This value does not automatically appear in the @code{Debugger Data}
1661 pane; to see it, right click again on the @code{N} in the
1662 @code{Debugger Data} pane and select @code{Update value}.
1663 The new value, 4, will appear in red.
1665 @item @emph{Single stepping}
1667 Select @code{Debug}, then @code{Next}.
1668 This will cause the next statement to be executed, in this case the
1669 call of @code{Put_Line} with the string slice.
1670 Notice in the console window that the displayed string is simply
1671 @code{abcd} and not @code{abcde} which you had entered.
1672 This is because the upper bound of the slice is now 4 rather than 5.
1674 @item @emph{Removing a breakpoint}
1676 Toggle the breakpoint icon at line 10.
1678 @item @emph{Resuming execution from a breakpoint}
1680 Select @code{Debug}, then @code{Continue}.
1681 The program will reach the next iteration of the loop, and
1682 wait for input after displaying the prompt.
1683 This time, just hit the @kbd{Enter} key.
1684 The value of @code{N} will be 0, and the program will terminate.
1685 The console window will disappear.
1690 @node The GNAT Compilation Model
1691 @chapter The GNAT Compilation Model
1692 @cindex GNAT compilation model
1693 @cindex Compilation model
1696 * Source Representation::
1697 * Foreign Language Representation::
1698 * File Naming Rules::
1699 * Using Other File Names::
1700 * Alternative File Naming Schemes::
1701 * Generating Object Files::
1702 * Source Dependencies::
1703 * The Ada Library Information Files::
1704 * Binding an Ada Program::
1705 * Mixed Language Programming::
1707 * Building Mixed Ada & C++ Programs::
1708 * Comparison between GNAT and C/C++ Compilation Models::
1710 * Comparison between GNAT and Conventional Ada Library Models::
1712 * Placement of temporary files::
1717 This chapter describes the compilation model used by GNAT. Although
1718 similar to that used by other languages, such as C and C++, this model
1719 is substantially different from the traditional Ada compilation models,
1720 which are based on a library. The model is initially described without
1721 reference to the library-based model. If you have not previously used an
1722 Ada compiler, you need only read the first part of this chapter. The
1723 last section describes and discusses the differences between the GNAT
1724 model and the traditional Ada compiler models. If you have used other
1725 Ada compilers, this section will help you to understand those
1726 differences, and the advantages of the GNAT model.
1728 @node Source Representation
1729 @section Source Representation
1733 Ada source programs are represented in standard text files, using
1734 Latin-1 coding. Latin-1 is an 8-bit code that includes the familiar
1735 7-bit ASCII set, plus additional characters used for
1736 representing foreign languages (@pxref{Foreign Language Representation}
1737 for support of non-USA character sets). The format effector characters
1738 are represented using their standard ASCII encodings, as follows:
1743 Vertical tab, @code{16#0B#}
1747 Horizontal tab, @code{16#09#}
1751 Carriage return, @code{16#0D#}
1755 Line feed, @code{16#0A#}
1759 Form feed, @code{16#0C#}
1763 Source files are in standard text file format. In addition, GNAT will
1764 recognize a wide variety of stream formats, in which the end of
1765 physical lines is marked by any of the following sequences:
1766 @code{LF}, @code{CR}, @code{CR-LF}, or @code{LF-CR}. This is useful
1767 in accommodating files that are imported from other operating systems.
1769 @cindex End of source file
1770 @cindex Source file, end
1772 The end of a source file is normally represented by the physical end of
1773 file. However, the control character @code{16#1A#} (@code{SUB}) is also
1774 recognized as signalling the end of the source file. Again, this is
1775 provided for compatibility with other operating systems where this
1776 code is used to represent the end of file.
1778 Each file contains a single Ada compilation unit, including any pragmas
1779 associated with the unit. For example, this means you must place a
1780 package declaration (a package @dfn{spec}) and the corresponding body in
1781 separate files. An Ada @dfn{compilation} (which is a sequence of
1782 compilation units) is represented using a sequence of files. Similarly,
1783 you will place each subunit or child unit in a separate file.
1785 @node Foreign Language Representation
1786 @section Foreign Language Representation
1789 GNAT supports the standard character sets defined in Ada as well as
1790 several other non-standard character sets for use in localized versions
1791 of the compiler (@pxref{Character Set Control}).
1794 * Other 8-Bit Codes::
1795 * Wide Character Encodings::
1803 The basic character set is Latin-1. This character set is defined by ISO
1804 standard 8859, part 1. The lower half (character codes @code{16#00#}
1805 @dots{} @code{16#7F#)} is identical to standard ASCII coding, but the upper half
1806 is used to represent additional characters. These include extended letters
1807 used by European languages, such as French accents, the vowels with umlauts
1808 used in German, and the extra letter A-ring used in Swedish.
1810 @findex Ada.Characters.Latin_1
1811 For a complete list of Latin-1 codes and their encodings, see the source
1812 file of library unit @code{Ada.Characters.Latin_1} in file
1813 @file{a-chlat1.ads}.
1814 You may use any of these extended characters freely in character or
1815 string literals. In addition, the extended characters that represent
1816 letters can be used in identifiers.
1818 @node Other 8-Bit Codes
1819 @subsection Other 8-Bit Codes
1822 GNAT also supports several other 8-bit coding schemes:
1825 @item ISO 8859-2 (Latin-2)
1828 Latin-2 letters allowed in identifiers, with uppercase and lowercase
1831 @item ISO 8859-3 (Latin-3)
1834 Latin-3 letters allowed in identifiers, with uppercase and lowercase
1837 @item ISO 8859-4 (Latin-4)
1840 Latin-4 letters allowed in identifiers, with uppercase and lowercase
1843 @item ISO 8859-5 (Cyrillic)
1846 ISO 8859-5 letters (Cyrillic) allowed in identifiers, with uppercase and
1847 lowercase equivalence.
1849 @item ISO 8859-15 (Latin-9)
1852 ISO 8859-15 (Latin-9) letters allowed in identifiers, with uppercase and
1853 lowercase equivalence
1855 @item IBM PC (code page 437)
1856 @cindex code page 437
1857 This code page is the normal default for PCs in the U.S. It corresponds
1858 to the original IBM PC character set. This set has some, but not all, of
1859 the extended Latin-1 letters, but these letters do not have the same
1860 encoding as Latin-1. In this mode, these letters are allowed in
1861 identifiers with uppercase and lowercase equivalence.
1863 @item IBM PC (code page 850)
1864 @cindex code page 850
1865 This code page is a modification of 437 extended to include all the
1866 Latin-1 letters, but still not with the usual Latin-1 encoding. In this
1867 mode, all these letters are allowed in identifiers with uppercase and
1868 lowercase equivalence.
1870 @item Full Upper 8-bit
1871 Any character in the range 80-FF allowed in identifiers, and all are
1872 considered distinct. In other words, there are no uppercase and lowercase
1873 equivalences in this range. This is useful in conjunction with
1874 certain encoding schemes used for some foreign character sets (e.g.,
1875 the typical method of representing Chinese characters on the PC).
1878 No upper-half characters in the range 80-FF are allowed in identifiers.
1879 This gives Ada 83 compatibility for identifier names.
1883 For precise data on the encodings permitted, and the uppercase and lowercase
1884 equivalences that are recognized, see the file @file{csets.adb} in
1885 the GNAT compiler sources. You will need to obtain a full source release
1886 of GNAT to obtain this file.
1888 @node Wide Character Encodings
1889 @subsection Wide Character Encodings
1892 GNAT allows wide character codes to appear in character and string
1893 literals, and also optionally in identifiers, by means of the following
1894 possible encoding schemes:
1899 In this encoding, a wide character is represented by the following five
1907 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1908 characters (using uppercase letters) of the wide character code. For
1909 example, ESC A345 is used to represent the wide character with code
1911 This scheme is compatible with use of the full Wide_Character set.
1913 @item Upper-Half Coding
1914 @cindex Upper-Half Coding
1915 The wide character with encoding @code{16#abcd#} where the upper bit is on
1916 (in other words, ``a'' is in the range 8-F) is represented as two bytes,
1917 @code{16#ab#} and @code{16#cd#}. The second byte cannot be a format control
1918 character, but is not required to be in the upper half. This method can
1919 be also used for shift-JIS or EUC, where the internal coding matches the
1922 @item Shift JIS Coding
1923 @cindex Shift JIS Coding
1924 A wide character is represented by a two-character sequence,
1926 @code{16#cd#}, with the restrictions described for upper-half encoding as
1927 described above. The internal character code is the corresponding JIS
1928 character according to the standard algorithm for Shift-JIS
1929 conversion. Only characters defined in the JIS code set table can be
1930 used with this encoding method.
1934 A wide character is represented by a two-character sequence
1936 @code{16#cd#}, with both characters being in the upper half. The internal
1937 character code is the corresponding JIS character according to the EUC
1938 encoding algorithm. Only characters defined in the JIS code set table
1939 can be used with this encoding method.
1942 A wide character is represented using
1943 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
1944 10646-1/Am.2. Depending on the character value, the representation
1945 is a one, two, or three byte sequence:
1950 16#0000#-16#007f#: 2#0@var{xxxxxxx}#
1951 16#0080#-16#07ff#: 2#110@var{xxxxx}# 2#10@var{xxxxxx}#
1952 16#0800#-16#ffff#: 2#1110@var{xxxx}# 2#10@var{xxxxxx}# 2#10@var{xxxxxx}#
1957 where the @var{xxx} bits correspond to the left-padded bits of the
1958 16-bit character value. Note that all lower half ASCII characters
1959 are represented as ASCII bytes and all upper half characters and
1960 other wide characters are represented as sequences of upper-half
1961 (The full UTF-8 scheme allows for encoding 31-bit characters as
1962 6-byte sequences, but in this implementation, all UTF-8 sequences
1963 of four or more bytes length will be treated as illegal).
1964 @item Brackets Coding
1965 In this encoding, a wide character is represented by the following eight
1973 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1974 characters (using uppercase letters) of the wide character code. For
1975 example, [``A345''] is used to represent the wide character with code
1976 @code{16#A345#}. It is also possible (though not required) to use the
1977 Brackets coding for upper half characters. For example, the code
1978 @code{16#A3#} can be represented as @code{[``A3'']}.
1980 This scheme is compatible with use of the full Wide_Character set,
1981 and is also the method used for wide character encoding in the standard
1982 ACVC (Ada Compiler Validation Capability) test suite distributions.
1987 Note: Some of these coding schemes do not permit the full use of the
1988 Ada character set. For example, neither Shift JIS, nor EUC allow the
1989 use of the upper half of the Latin-1 set.
1991 @node File Naming Rules
1992 @section File Naming Rules
1995 The default file name is determined by the name of the unit that the
1996 file contains. The name is formed by taking the full expanded name of
1997 the unit and replacing the separating dots with hyphens and using
1998 ^lowercase^uppercase^ for all letters.
2000 An exception arises if the file name generated by the above rules starts
2001 with one of the characters
2003 @samp{A}, @samp{G}, @samp{I}, or @samp{S},
2006 @samp{a}, @samp{g}, @samp{i}, or @samp{s},
2008 and the second character is a
2009 minus. In this case, the character ^tilde^dollar sign^ is used in place
2010 of the minus. The reason for this special rule is to avoid clashes with
2011 the standard names for child units of the packages System, Ada,
2012 Interfaces, and GNAT, which use the prefixes
2014 @samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},
2017 @samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},
2021 The file extension is @file{.ads} for a spec and
2022 @file{.adb} for a body. The following list shows some
2023 examples of these rules.
2030 @item arith_functions.ads
2031 Arith_Functions (package spec)
2032 @item arith_functions.adb
2033 Arith_Functions (package body)
2035 Func.Spec (child package spec)
2037 Func.Spec (child package body)
2039 Sub (subunit of Main)
2040 @item ^a~bad.adb^A$BAD.ADB^
2041 A.Bad (child package body)
2045 Following these rules can result in excessively long
2046 file names if corresponding
2047 unit names are long (for example, if child units or subunits are
2048 heavily nested). An option is available to shorten such long file names
2049 (called file name ``krunching''). This may be particularly useful when
2050 programs being developed with GNAT are to be used on operating systems
2051 with limited file name lengths. @xref{Using gnatkr}.
2053 Of course, no file shortening algorithm can guarantee uniqueness over
2054 all possible unit names; if file name krunching is used, it is your
2055 responsibility to ensure no name clashes occur. Alternatively you
2056 can specify the exact file names that you want used, as described
2057 in the next section. Finally, if your Ada programs are migrating from a
2058 compiler with a different naming convention, you can use the gnatchop
2059 utility to produce source files that follow the GNAT naming conventions.
2060 (For details @pxref{Renaming Files Using gnatchop}.)
2062 Note: in the case of @code{Windows NT/XP} or @code{OpenVMS} operating
2063 systems, case is not significant. So for example on @code{Windows XP}
2064 if the canonical name is @code{main-sub.adb}, you can use the file name
2065 @code{Main-Sub.adb} instead. However, case is significant for other
2066 operating systems, so for example, if you want to use other than
2067 canonically cased file names on a Unix system, you need to follow
2068 the procedures described in the next section.
2070 @node Using Other File Names
2071 @section Using Other File Names
2075 In the previous section, we have described the default rules used by
2076 GNAT to determine the file name in which a given unit resides. It is
2077 often convenient to follow these default rules, and if you follow them,
2078 the compiler knows without being explicitly told where to find all
2081 However, in some cases, particularly when a program is imported from
2082 another Ada compiler environment, it may be more convenient for the
2083 programmer to specify which file names contain which units. GNAT allows
2084 arbitrary file names to be used by means of the Source_File_Name pragma.
2085 The form of this pragma is as shown in the following examples:
2086 @cindex Source_File_Name pragma
2088 @smallexample @c ada
2090 pragma Source_File_Name (My_Utilities.Stacks,
2091 Spec_File_Name => "myutilst_a.ada");
2092 pragma Source_File_name (My_Utilities.Stacks,
2093 Body_File_Name => "myutilst.ada");
2098 As shown in this example, the first argument for the pragma is the unit
2099 name (in this example a child unit). The second argument has the form
2100 of a named association. The identifier
2101 indicates whether the file name is for a spec or a body;
2102 the file name itself is given by a string literal.
2104 The source file name pragma is a configuration pragma, which means that
2105 normally it will be placed in the @file{gnat.adc}
2106 file used to hold configuration
2107 pragmas that apply to a complete compilation environment.
2108 For more details on how the @file{gnat.adc} file is created and used
2109 see @ref{Handling of Configuration Pragmas}.
2110 @cindex @file{gnat.adc}
2113 GNAT allows completely arbitrary file names to be specified using the
2114 source file name pragma. However, if the file name specified has an
2115 extension other than @file{.ads} or @file{.adb} it is necessary to use
2116 a special syntax when compiling the file. The name in this case must be
2117 preceded by the special sequence @option{-x} followed by a space and the name
2118 of the language, here @code{ada}, as in:
2121 $ gcc -c -x ada peculiar_file_name.sim
2126 @command{gnatmake} handles non-standard file names in the usual manner (the
2127 non-standard file name for the main program is simply used as the
2128 argument to gnatmake). Note that if the extension is also non-standard,
2129 then it must be included in the @command{gnatmake} command, it may not
2132 @node Alternative File Naming Schemes
2133 @section Alternative File Naming Schemes
2134 @cindex File naming schemes, alternative
2137 In the previous section, we described the use of the @code{Source_File_Name}
2138 pragma to allow arbitrary names to be assigned to individual source files.
2139 However, this approach requires one pragma for each file, and especially in
2140 large systems can result in very long @file{gnat.adc} files, and also create
2141 a maintenance problem.
2143 GNAT also provides a facility for specifying systematic file naming schemes
2144 other than the standard default naming scheme previously described. An
2145 alternative scheme for naming is specified by the use of
2146 @code{Source_File_Name} pragmas having the following format:
2147 @cindex Source_File_Name pragma
2149 @smallexample @c ada
2150 pragma Source_File_Name (
2151 Spec_File_Name => FILE_NAME_PATTERN
2152 @r{[},Casing => CASING_SPEC@r{]}
2153 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2155 pragma Source_File_Name (
2156 Body_File_Name => FILE_NAME_PATTERN
2157 @r{[},Casing => CASING_SPEC@r{]}
2158 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2160 pragma Source_File_Name (
2161 Subunit_File_Name => FILE_NAME_PATTERN
2162 @r{[},Casing => CASING_SPEC@r{]}
2163 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2165 FILE_NAME_PATTERN ::= STRING_LITERAL
2166 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
2170 The @code{FILE_NAME_PATTERN} string shows how the file name is constructed.
2171 It contains a single asterisk character, and the unit name is substituted
2172 systematically for this asterisk. The optional parameter
2173 @code{Casing} indicates
2174 whether the unit name is to be all upper-case letters, all lower-case letters,
2175 or mixed-case. If no
2176 @code{Casing} parameter is used, then the default is all
2177 ^lower-case^upper-case^.
2179 The optional @code{Dot_Replacement} string is used to replace any periods
2180 that occur in subunit or child unit names. If no @code{Dot_Replacement}
2181 argument is used then separating dots appear unchanged in the resulting
2183 Although the above syntax indicates that the
2184 @code{Casing} argument must appear
2185 before the @code{Dot_Replacement} argument, but it
2186 is also permissible to write these arguments in the opposite order.
2188 As indicated, it is possible to specify different naming schemes for
2189 bodies, specs, and subunits. Quite often the rule for subunits is the
2190 same as the rule for bodies, in which case, there is no need to give
2191 a separate @code{Subunit_File_Name} rule, and in this case the
2192 @code{Body_File_name} rule is used for subunits as well.
2194 The separate rule for subunits can also be used to implement the rather
2195 unusual case of a compilation environment (e.g.@: a single directory) which
2196 contains a subunit and a child unit with the same unit name. Although
2197 both units cannot appear in the same partition, the Ada Reference Manual
2198 allows (but does not require) the possibility of the two units coexisting
2199 in the same environment.
2201 The file name translation works in the following steps:
2206 If there is a specific @code{Source_File_Name} pragma for the given unit,
2207 then this is always used, and any general pattern rules are ignored.
2210 If there is a pattern type @code{Source_File_Name} pragma that applies to
2211 the unit, then the resulting file name will be used if the file exists. If
2212 more than one pattern matches, the latest one will be tried first, and the
2213 first attempt resulting in a reference to a file that exists will be used.
2216 If no pattern type @code{Source_File_Name} pragma that applies to the unit
2217 for which the corresponding file exists, then the standard GNAT default
2218 naming rules are used.
2223 As an example of the use of this mechanism, consider a commonly used scheme
2224 in which file names are all lower case, with separating periods copied
2225 unchanged to the resulting file name, and specs end with @file{.1.ada}, and
2226 bodies end with @file{.2.ada}. GNAT will follow this scheme if the following
2229 @smallexample @c ada
2230 pragma Source_File_Name
2231 (Spec_File_Name => "*.1.ada");
2232 pragma Source_File_Name
2233 (Body_File_Name => "*.2.ada");
2237 The default GNAT scheme is actually implemented by providing the following
2238 default pragmas internally:
2240 @smallexample @c ada
2241 pragma Source_File_Name
2242 (Spec_File_Name => "*.ads", Dot_Replacement => "-");
2243 pragma Source_File_Name
2244 (Body_File_Name => "*.adb", Dot_Replacement => "-");
2248 Our final example implements a scheme typically used with one of the
2249 Ada 83 compilers, where the separator character for subunits was ``__''
2250 (two underscores), specs were identified by adding @file{_.ADA}, bodies
2251 by adding @file{.ADA}, and subunits by
2252 adding @file{.SEP}. All file names were
2253 upper case. Child units were not present of course since this was an
2254 Ada 83 compiler, but it seems reasonable to extend this scheme to use
2255 the same double underscore separator for child units.
2257 @smallexample @c ada
2258 pragma Source_File_Name
2259 (Spec_File_Name => "*_.ADA",
2260 Dot_Replacement => "__",
2261 Casing = Uppercase);
2262 pragma Source_File_Name
2263 (Body_File_Name => "*.ADA",
2264 Dot_Replacement => "__",
2265 Casing = Uppercase);
2266 pragma Source_File_Name
2267 (Subunit_File_Name => "*.SEP",
2268 Dot_Replacement => "__",
2269 Casing = Uppercase);
2272 @node Generating Object Files
2273 @section Generating Object Files
2276 An Ada program consists of a set of source files, and the first step in
2277 compiling the program is to generate the corresponding object files.
2278 These are generated by compiling a subset of these source files.
2279 The files you need to compile are the following:
2283 If a package spec has no body, compile the package spec to produce the
2284 object file for the package.
2287 If a package has both a spec and a body, compile the body to produce the
2288 object file for the package. The source file for the package spec need
2289 not be compiled in this case because there is only one object file, which
2290 contains the code for both the spec and body of the package.
2293 For a subprogram, compile the subprogram body to produce the object file
2294 for the subprogram. The spec, if one is present, is as usual in a
2295 separate file, and need not be compiled.
2299 In the case of subunits, only compile the parent unit. A single object
2300 file is generated for the entire subunit tree, which includes all the
2304 Compile child units independently of their parent units
2305 (though, of course, the spec of all the ancestor unit must be present in order
2306 to compile a child unit).
2310 Compile generic units in the same manner as any other units. The object
2311 files in this case are small dummy files that contain at most the
2312 flag used for elaboration checking. This is because GNAT always handles generic
2313 instantiation by means of macro expansion. However, it is still necessary to
2314 compile generic units, for dependency checking and elaboration purposes.
2318 The preceding rules describe the set of files that must be compiled to
2319 generate the object files for a program. Each object file has the same
2320 name as the corresponding source file, except that the extension is
2323 You may wish to compile other files for the purpose of checking their
2324 syntactic and semantic correctness. For example, in the case where a
2325 package has a separate spec and body, you would not normally compile the
2326 spec. However, it is convenient in practice to compile the spec to make
2327 sure it is error-free before compiling clients of this spec, because such
2328 compilations will fail if there is an error in the spec.
2330 GNAT provides an option for compiling such files purely for the
2331 purposes of checking correctness; such compilations are not required as
2332 part of the process of building a program. To compile a file in this
2333 checking mode, use the @option{-gnatc} switch.
2335 @node Source Dependencies
2336 @section Source Dependencies
2339 A given object file clearly depends on the source file which is compiled
2340 to produce it. Here we are using @dfn{depends} in the sense of a typical
2341 @code{make} utility; in other words, an object file depends on a source
2342 file if changes to the source file require the object file to be
2344 In addition to this basic dependency, a given object may depend on
2345 additional source files as follows:
2349 If a file being compiled @code{with}'s a unit @var{X}, the object file
2350 depends on the file containing the spec of unit @var{X}. This includes
2351 files that are @code{with}'ed implicitly either because they are parents
2352 of @code{with}'ed child units or they are run-time units required by the
2353 language constructs used in a particular unit.
2356 If a file being compiled instantiates a library level generic unit, the
2357 object file depends on both the spec and body files for this generic
2361 If a file being compiled instantiates a generic unit defined within a
2362 package, the object file depends on the body file for the package as
2363 well as the spec file.
2367 @cindex @option{-gnatn} switch
2368 If a file being compiled contains a call to a subprogram for which
2369 pragma @code{Inline} applies and inlining is activated with the
2370 @option{-gnatn} switch, the object file depends on the file containing the
2371 body of this subprogram as well as on the file containing the spec. Note
2372 that for inlining to actually occur as a result of the use of this switch,
2373 it is necessary to compile in optimizing mode.
2375 @cindex @option{-gnatN} switch
2376 The use of @option{-gnatN} activates inlining optimization
2377 that is performed by the front end of the compiler. This inlining does
2378 not require that the code generation be optimized. Like @option{-gnatn},
2379 the use of this switch generates additional dependencies.
2381 When using a gcc-based back end (in practice this means using any version
2382 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
2383 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
2384 Historically front end inlining was more extensive than the gcc back end
2385 inlining, but that is no longer the case.
2388 If an object file @file{O} depends on the proper body of a subunit through
2389 inlining or instantiation, it depends on the parent unit of the subunit.
2390 This means that any modification of the parent unit or one of its subunits
2391 affects the compilation of @file{O}.
2394 The object file for a parent unit depends on all its subunit body files.
2397 The previous two rules meant that for purposes of computing dependencies and
2398 recompilation, a body and all its subunits are treated as an indivisible whole.
2401 These rules are applied transitively: if unit @code{A} @code{with}'s
2402 unit @code{B}, whose elaboration calls an inlined procedure in package
2403 @code{C}, the object file for unit @code{A} will depend on the body of
2404 @code{C}, in file @file{c.adb}.
2406 The set of dependent files described by these rules includes all the
2407 files on which the unit is semantically dependent, as dictated by the
2408 Ada language standard. However, it is a superset of what the
2409 standard describes, because it includes generic, inline, and subunit
2412 An object file must be recreated by recompiling the corresponding source
2413 file if any of the source files on which it depends are modified. For
2414 example, if the @code{make} utility is used to control compilation,
2415 the rule for an Ada object file must mention all the source files on
2416 which the object file depends, according to the above definition.
2417 The determination of the necessary
2418 recompilations is done automatically when one uses @command{gnatmake}.
2421 @node The Ada Library Information Files
2422 @section The Ada Library Information Files
2423 @cindex Ada Library Information files
2424 @cindex @file{ALI} files
2427 Each compilation actually generates two output files. The first of these
2428 is the normal object file that has a @file{.o} extension. The second is a
2429 text file containing full dependency information. It has the same
2430 name as the source file, but an @file{.ali} extension.
2431 This file is known as the Ada Library Information (@file{ALI}) file.
2432 The following information is contained in the @file{ALI} file.
2436 Version information (indicates which version of GNAT was used to compile
2437 the unit(s) in question)
2440 Main program information (including priority and time slice settings,
2441 as well as the wide character encoding used during compilation).
2444 List of arguments used in the @command{gcc} command for the compilation
2447 Attributes of the unit, including configuration pragmas used, an indication
2448 of whether the compilation was successful, exception model used etc.
2451 A list of relevant restrictions applying to the unit (used for consistency)
2455 Categorization information (e.g.@: use of pragma @code{Pure}).
2458 Information on all @code{with}'ed units, including presence of
2459 @code{Elaborate} or @code{Elaborate_All} pragmas.
2462 Information from any @code{Linker_Options} pragmas used in the unit
2465 Information on the use of @code{Body_Version} or @code{Version}
2466 attributes in the unit.
2469 Dependency information. This is a list of files, together with
2470 time stamp and checksum information. These are files on which
2471 the unit depends in the sense that recompilation is required
2472 if any of these units are modified.
2475 Cross-reference data. Contains information on all entities referenced
2476 in the unit. Used by tools like @code{gnatxref} and @code{gnatfind} to
2477 provide cross-reference information.
2482 For a full detailed description of the format of the @file{ALI} file,
2483 see the source of the body of unit @code{Lib.Writ}, contained in file
2484 @file{lib-writ.adb} in the GNAT compiler sources.
2486 @node Binding an Ada Program
2487 @section Binding an Ada Program
2490 When using languages such as C and C++, once the source files have been
2491 compiled the only remaining step in building an executable program
2492 is linking the object modules together. This means that it is possible to
2493 link an inconsistent version of a program, in which two units have
2494 included different versions of the same header.
2496 The rules of Ada do not permit such an inconsistent program to be built.
2497 For example, if two clients have different versions of the same package,
2498 it is illegal to build a program containing these two clients.
2499 These rules are enforced by the GNAT binder, which also determines an
2500 elaboration order consistent with the Ada rules.
2502 The GNAT binder is run after all the object files for a program have
2503 been created. It is given the name of the main program unit, and from
2504 this it determines the set of units required by the program, by reading the
2505 corresponding ALI files. It generates error messages if the program is
2506 inconsistent or if no valid order of elaboration exists.
2508 If no errors are detected, the binder produces a main program, in Ada by
2509 default, that contains calls to the elaboration procedures of those
2510 compilation unit that require them, followed by
2511 a call to the main program. This Ada program is compiled to generate the
2512 object file for the main program. The name of
2513 the Ada file is @file{b~@var{xxx}.adb} (with the corresponding spec
2514 @file{b~@var{xxx}.ads}) where @var{xxx} is the name of the
2517 Finally, the linker is used to build the resulting executable program,
2518 using the object from the main program from the bind step as well as the
2519 object files for the Ada units of the program.
2521 @node Mixed Language Programming
2522 @section Mixed Language Programming
2523 @cindex Mixed Language Programming
2526 This section describes how to develop a mixed-language program,
2527 specifically one that comprises units in both Ada and C.
2530 * Interfacing to C::
2531 * Calling Conventions::
2534 @node Interfacing to C
2535 @subsection Interfacing to C
2537 Interfacing Ada with a foreign language such as C involves using
2538 compiler directives to import and/or export entity definitions in each
2539 language---using @code{extern} statements in C, for instance, and the
2540 @code{Import}, @code{Export}, and @code{Convention} pragmas in Ada.
2541 A full treatment of these topics is provided in Appendix B, section 1
2542 of the Ada Reference Manual.
2544 There are two ways to build a program using GNAT that contains some Ada
2545 sources and some foreign language sources, depending on whether or not
2546 the main subprogram is written in Ada. Here is a source example with
2547 the main subprogram in Ada:
2553 void print_num (int num)
2555 printf ("num is %d.\n", num);
2561 /* num_from_Ada is declared in my_main.adb */
2562 extern int num_from_Ada;
2566 return num_from_Ada;
2570 @smallexample @c ada
2572 procedure My_Main is
2574 -- Declare then export an Integer entity called num_from_Ada
2575 My_Num : Integer := 10;
2576 pragma Export (C, My_Num, "num_from_Ada");
2578 -- Declare an Ada function spec for Get_Num, then use
2579 -- C function get_num for the implementation.
2580 function Get_Num return Integer;
2581 pragma Import (C, Get_Num, "get_num");
2583 -- Declare an Ada procedure spec for Print_Num, then use
2584 -- C function print_num for the implementation.
2585 procedure Print_Num (Num : Integer);
2586 pragma Import (C, Print_Num, "print_num");
2589 Print_Num (Get_Num);
2595 To build this example, first compile the foreign language files to
2596 generate object files:
2598 ^gcc -c file1.c^gcc -c FILE1.C^
2599 ^gcc -c file2.c^gcc -c FILE2.C^
2603 Then, compile the Ada units to produce a set of object files and ALI
2606 gnatmake ^-c^/ACTIONS=COMPILE^ my_main.adb
2610 Run the Ada binder on the Ada main program:
2612 gnatbind my_main.ali
2616 Link the Ada main program, the Ada objects and the other language
2619 gnatlink my_main.ali file1.o file2.o
2623 The last three steps can be grouped in a single command:
2625 gnatmake my_main.adb -largs file1.o file2.o
2628 @cindex Binder output file
2630 If the main program is in a language other than Ada, then you may have
2631 more than one entry point into the Ada subsystem. You must use a special
2632 binder option to generate callable routines that initialize and
2633 finalize the Ada units (@pxref{Binding with Non-Ada Main Programs}).
2634 Calls to the initialization and finalization routines must be inserted
2635 in the main program, or some other appropriate point in the code. The
2636 call to initialize the Ada units must occur before the first Ada
2637 subprogram is called, and the call to finalize the Ada units must occur
2638 after the last Ada subprogram returns. The binder will place the
2639 initialization and finalization subprograms into the
2640 @file{b~@var{xxx}.adb} file where they can be accessed by your C
2641 sources. To illustrate, we have the following example:
2645 extern void adainit (void);
2646 extern void adafinal (void);
2647 extern int add (int, int);
2648 extern int sub (int, int);
2650 int main (int argc, char *argv[])
2656 /* Should print "21 + 7 = 28" */
2657 printf ("%d + %d = %d\n", a, b, add (a, b));
2658 /* Should print "21 - 7 = 14" */
2659 printf ("%d - %d = %d\n", a, b, sub (a, b));
2665 @smallexample @c ada
2668 function Add (A, B : Integer) return Integer;
2669 pragma Export (C, Add, "add");
2673 package body Unit1 is
2674 function Add (A, B : Integer) return Integer is
2682 function Sub (A, B : Integer) return Integer;
2683 pragma Export (C, Sub, "sub");
2687 package body Unit2 is
2688 function Sub (A, B : Integer) return Integer is
2697 The build procedure for this application is similar to the last
2698 example's. First, compile the foreign language files to generate object
2701 ^gcc -c main.c^gcc -c main.c^
2705 Next, compile the Ada units to produce a set of object files and ALI
2708 gnatmake ^-c^/ACTIONS=COMPILE^ unit1.adb
2709 gnatmake ^-c^/ACTIONS=COMPILE^ unit2.adb
2713 Run the Ada binder on every generated ALI file. Make sure to use the
2714 @option{-n} option to specify a foreign main program:
2716 gnatbind ^-n^/NOMAIN^ unit1.ali unit2.ali
2720 Link the Ada main program, the Ada objects and the foreign language
2721 objects. You need only list the last ALI file here:
2723 gnatlink unit2.ali main.o -o exec_file
2726 This procedure yields a binary executable called @file{exec_file}.
2730 Depending on the circumstances (for example when your non-Ada main object
2731 does not provide symbol @code{main}), you may also need to instruct the
2732 GNAT linker not to include the standard startup objects by passing the
2733 @option{^-nostartfiles^/NOSTART_FILES^} switch to @command{gnatlink}.
2735 @node Calling Conventions
2736 @subsection Calling Conventions
2737 @cindex Foreign Languages
2738 @cindex Calling Conventions
2739 GNAT follows standard calling sequence conventions and will thus interface
2740 to any other language that also follows these conventions. The following
2741 Convention identifiers are recognized by GNAT:
2744 @cindex Interfacing to Ada
2745 @cindex Other Ada compilers
2746 @cindex Convention Ada
2748 This indicates that the standard Ada calling sequence will be
2749 used and all Ada data items may be passed without any limitations in the
2750 case where GNAT is used to generate both the caller and callee. It is also
2751 possible to mix GNAT generated code and code generated by another Ada
2752 compiler. In this case, the data types should be restricted to simple
2753 cases, including primitive types. Whether complex data types can be passed
2754 depends on the situation. Probably it is safe to pass simple arrays, such
2755 as arrays of integers or floats. Records may or may not work, depending
2756 on whether both compilers lay them out identically. Complex structures
2757 involving variant records, access parameters, tasks, or protected types,
2758 are unlikely to be able to be passed.
2760 Note that in the case of GNAT running
2761 on a platform that supports HP Ada 83, a higher degree of compatibility
2762 can be guaranteed, and in particular records are layed out in an identical
2763 manner in the two compilers. Note also that if output from two different
2764 compilers is mixed, the program is responsible for dealing with elaboration
2765 issues. Probably the safest approach is to write the main program in the
2766 version of Ada other than GNAT, so that it takes care of its own elaboration
2767 requirements, and then call the GNAT-generated adainit procedure to ensure
2768 elaboration of the GNAT components. Consult the documentation of the other
2769 Ada compiler for further details on elaboration.
2771 However, it is not possible to mix the tasking run time of GNAT and
2772 HP Ada 83, All the tasking operations must either be entirely within
2773 GNAT compiled sections of the program, or entirely within HP Ada 83
2774 compiled sections of the program.
2776 @cindex Interfacing to Assembly
2777 @cindex Convention Assembler
2779 Specifies assembler as the convention. In practice this has the
2780 same effect as convention Ada (but is not equivalent in the sense of being
2781 considered the same convention).
2783 @cindex Convention Asm
2786 Equivalent to Assembler.
2788 @cindex Interfacing to COBOL
2789 @cindex Convention COBOL
2792 Data will be passed according to the conventions described
2793 in section B.4 of the Ada Reference Manual.
2796 @cindex Interfacing to C
2797 @cindex Convention C
2799 Data will be passed according to the conventions described
2800 in section B.3 of the Ada Reference Manual.
2802 A note on interfacing to a C ``varargs'' function:
2803 @findex C varargs function
2804 @cindex Interfacing to C varargs function
2805 @cindex varargs function interfaces
2809 In C, @code{varargs} allows a function to take a variable number of
2810 arguments. There is no direct equivalent in this to Ada. One
2811 approach that can be used is to create a C wrapper for each
2812 different profile and then interface to this C wrapper. For
2813 example, to print an @code{int} value using @code{printf},
2814 create a C function @code{printfi} that takes two arguments, a
2815 pointer to a string and an int, and calls @code{printf}.
2816 Then in the Ada program, use pragma @code{Import} to
2817 interface to @code{printfi}.
2820 It may work on some platforms to directly interface to
2821 a @code{varargs} function by providing a specific Ada profile
2822 for a particular call. However, this does not work on
2823 all platforms, since there is no guarantee that the
2824 calling sequence for a two argument normal C function
2825 is the same as for calling a @code{varargs} C function with
2826 the same two arguments.
2829 @cindex Convention Default
2834 @cindex Convention External
2841 @cindex Interfacing to C++
2842 @cindex Convention C++
2843 @item C_Plus_Plus (or CPP)
2844 This stands for C++. For most purposes this is identical to C.
2845 See the separate description of the specialized GNAT pragmas relating to
2846 C++ interfacing for further details.
2850 @cindex Interfacing to Fortran
2851 @cindex Convention Fortran
2853 Data will be passed according to the conventions described
2854 in section B.5 of the Ada Reference Manual.
2857 This applies to an intrinsic operation, as defined in the Ada
2858 Reference Manual. If a pragma Import (Intrinsic) applies to a subprogram,
2859 this means that the body of the subprogram is provided by the compiler itself,
2860 usually by means of an efficient code sequence, and that the user does not
2861 supply an explicit body for it. In an application program, the pragma may
2862 be applied to the following sets of names:
2866 Rotate_Left, Rotate_Right, Shift_Left, Shift_Right,
2867 Shift_Right_Arithmetic. The corresponding subprogram declaration must have
2868 two formal parameters. The
2869 first one must be a signed integer type or a modular type with a binary
2870 modulus, and the second parameter must be of type Natural.
2871 The return type must be the same as the type of the first argument. The size
2872 of this type can only be 8, 16, 32, or 64.
2875 Binary arithmetic operators: ``+'', ``-'', ``*'', ``/''
2876 The corresponding operator declaration must have parameters and result type
2877 that have the same root numeric type (for example, all three are long_float
2878 types). This simplifies the definition of operations that use type checking
2879 to perform dimensional checks:
2881 @smallexample @c ada
2882 type Distance is new Long_Float;
2883 type Time is new Long_Float;
2884 type Velocity is new Long_Float;
2885 function "/" (D : Distance; T : Time)
2887 pragma Import (Intrinsic, "/");
2891 This common idiom is often programmed with a generic definition and an
2892 explicit body. The pragma makes it simpler to introduce such declarations.
2893 It incurs no overhead in compilation time or code size, because it is
2894 implemented as a single machine instruction.
2897 General subprogram entities, to bind an Ada subprogram declaration to
2898 a compiler builtin by name with back-ends where such interfaces are
2899 available. A typical example is the set of ``__builtin'' functions
2900 exposed by the GCC back-end, as in the following example:
2902 @smallexample @c ada
2903 function builtin_sqrt (F : Float) return Float;
2904 pragma Import (Intrinsic, builtin_sqrt, "__builtin_sqrtf");
2907 Most of the GCC builtins are accessible this way, and as for other
2908 import conventions (e.g. C), it is the user's responsibility to ensure
2909 that the Ada subprogram profile matches the underlying builtin
2917 @cindex Convention Stdcall
2919 This is relevant only to Windows XP/2000/NT implementations of GNAT,
2920 and specifies that the @code{Stdcall} calling sequence will be used,
2921 as defined by the NT API. Nevertheless, to ease building
2922 cross-platform bindings this convention will be handled as a @code{C} calling
2923 convention on non-Windows platforms.
2926 @cindex Convention DLL
2928 This is equivalent to @code{Stdcall}.
2931 @cindex Convention Win32
2933 This is equivalent to @code{Stdcall}.
2937 @cindex Convention Stubbed
2939 This is a special convention that indicates that the compiler
2940 should provide a stub body that raises @code{Program_Error}.
2944 GNAT additionally provides a useful pragma @code{Convention_Identifier}
2945 that can be used to parametrize conventions and allow additional synonyms
2946 to be specified. For example if you have legacy code in which the convention
2947 identifier Fortran77 was used for Fortran, you can use the configuration
2950 @smallexample @c ada
2951 pragma Convention_Identifier (Fortran77, Fortran);
2955 And from now on the identifier Fortran77 may be used as a convention
2956 identifier (for example in an @code{Import} pragma) with the same
2960 @node Building Mixed Ada & C++ Programs
2961 @section Building Mixed Ada and C++ Programs
2964 A programmer inexperienced with mixed-language development may find that
2965 building an application containing both Ada and C++ code can be a
2966 challenge. This section gives a few
2967 hints that should make this task easier. The first section addresses
2968 the differences between interfacing with C and interfacing with C++.
2970 looks into the delicate problem of linking the complete application from
2971 its Ada and C++ parts. The last section gives some hints on how the GNAT
2972 run-time library can be adapted in order to allow inter-language dispatching
2973 with a new C++ compiler.
2976 * Interfacing to C++::
2977 * Linking a Mixed C++ & Ada Program::
2978 * A Simple Example::
2979 * Interfacing with C++ constructors::
2980 * Interfacing with C++ at the Class Level::
2983 @node Interfacing to C++
2984 @subsection Interfacing to C++
2987 GNAT supports interfacing with the G++ compiler (or any C++ compiler
2988 generating code that is compatible with the G++ Application Binary
2989 Interface ---see http://www.codesourcery.com/archives/cxx-abi).
2992 Interfacing can be done at 3 levels: simple data, subprograms, and
2993 classes. In the first two cases, GNAT offers a specific @code{Convention
2994 C_Plus_Plus} (or @code{CPP}) that behaves exactly like @code{Convention C}.
2995 Usually, C++ mangles the names of subprograms. To generate proper mangled
2996 names automatically, see @ref{Generating Ada Bindings for C and C++ headers}).
2997 This problem can also be addressed manually in two ways:
3001 by modifying the C++ code in order to force a C convention using
3002 the @code{extern "C"} syntax.
3005 by figuring out the mangled name (using e.g. @command{nm}) and using it as the
3006 Link_Name argument of the pragma import.
3010 Interfacing at the class level can be achieved by using the GNAT specific
3011 pragmas such as @code{CPP_Constructor}. @xref{Interfacing to C++,,,
3012 gnat_rm, GNAT Reference Manual}, for additional information.
3014 @node Linking a Mixed C++ & Ada Program
3015 @subsection Linking a Mixed C++ & Ada Program
3018 Usually the linker of the C++ development system must be used to link
3019 mixed applications because most C++ systems will resolve elaboration
3020 issues (such as calling constructors on global class instances)
3021 transparently during the link phase. GNAT has been adapted to ease the
3022 use of a foreign linker for the last phase. Three cases can be
3027 Using GNAT and G++ (GNU C++ compiler) from the same GCC installation:
3028 The C++ linker can simply be called by using the C++ specific driver
3031 Note that if the C++ code uses inline functions, you will need to
3032 compile your C++ code with the @code{-fkeep-inline-functions} switch in
3033 order to provide an existing function implementation that the Ada code can
3037 $ g++ -c -fkeep-inline-functions file1.C
3038 $ g++ -c -fkeep-inline-functions file2.C
3039 $ gnatmake ada_unit -largs file1.o file2.o --LINK=g++
3043 Using GNAT and G++ from two different GCC installations: If both
3044 compilers are on the @env{PATH}, the previous method may be used. It is
3045 important to note that environment variables such as
3046 @env{C_INCLUDE_PATH}, @env{GCC_EXEC_PREFIX}, @env{BINUTILS_ROOT}, and
3047 @env{GCC_ROOT} will affect both compilers
3048 at the same time and may make one of the two compilers operate
3049 improperly if set during invocation of the wrong compiler. It is also
3050 very important that the linker uses the proper @file{libgcc.a} GCC
3051 library -- that is, the one from the C++ compiler installation. The
3052 implicit link command as suggested in the @command{gnatmake} command
3053 from the former example can be replaced by an explicit link command with
3054 the full-verbosity option in order to verify which library is used:
3057 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=c++
3059 If there is a problem due to interfering environment variables, it can
3060 be worked around by using an intermediate script. The following example
3061 shows the proper script to use when GNAT has not been installed at its
3062 default location and g++ has been installed at its default location:
3070 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=./my_script
3074 Using a non-GNU C++ compiler: The commands previously described can be
3075 used to insure that the C++ linker is used. Nonetheless, you need to add
3076 a few more parameters to the link command line, depending on the exception
3079 If the @code{setjmp/longjmp} exception mechanism is used, only the paths
3080 to the libgcc libraries are required:
3085 CC $* `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a`
3086 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3089 Where CC is the name of the non-GNU C++ compiler.
3091 If the @code{zero cost} exception mechanism is used, and the platform
3092 supports automatic registration of exception tables (e.g.@: Solaris or IRIX),
3093 paths to more objects are required:
3098 CC `gcc -print-file-name=crtbegin.o` $* \
3099 `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a` \
3100 `gcc -print-file-name=crtend.o`
3101 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3104 If the @code{zero cost} exception mechanism is used, and the platform
3105 doesn't support automatic registration of exception tables (e.g.@: HP-UX,
3106 Tru64 or AIX), the simple approach described above will not work and
3107 a pre-linking phase using GNAT will be necessary.
3111 Another alternative is to use the @command{gprbuild} multi-language builder
3112 which has a large knowledge base and knows how to link Ada and C++ code
3113 together automatically in most cases.
3115 @node A Simple Example
3116 @subsection A Simple Example
3118 The following example, provided as part of the GNAT examples, shows how
3119 to achieve procedural interfacing between Ada and C++ in both
3120 directions. The C++ class A has two methods. The first method is exported
3121 to Ada by the means of an extern C wrapper function. The second method
3122 calls an Ada subprogram. On the Ada side, The C++ calls are modelled by
3123 a limited record with a layout comparable to the C++ class. The Ada
3124 subprogram, in turn, calls the C++ method. So, starting from the C++
3125 main program, the process passes back and forth between the two
3129 Here are the compilation commands:
3131 $ gnatmake -c simple_cpp_interface
3134 $ gnatbind -n simple_cpp_interface
3135 $ gnatlink simple_cpp_interface -o cpp_main --LINK=g++
3136 -lstdc++ ex7.o cpp_main.o
3140 Here are the corresponding sources:
3148 void adainit (void);
3149 void adafinal (void);
3150 void method1 (A *t);
3172 class A : public Origin @{
3174 void method1 (void);
3175 void method2 (int v);
3185 extern "C" @{ void ada_method2 (A *t, int v);@}
3187 void A::method1 (void)
3190 printf ("in A::method1, a_value = %d \n",a_value);
3194 void A::method2 (int v)
3196 ada_method2 (this, v);
3197 printf ("in A::method2, a_value = %d \n",a_value);
3204 printf ("in A::A, a_value = %d \n",a_value);
3208 @smallexample @c ada
3210 package body Simple_Cpp_Interface is
3212 procedure Ada_Method2 (This : in out A; V : Integer) is
3218 end Simple_Cpp_Interface;
3221 package Simple_Cpp_Interface is
3224 Vptr : System.Address;
3228 pragma Convention (C, A);
3230 procedure Method1 (This : in out A);
3231 pragma Import (C, Method1);
3233 procedure Ada_Method2 (This : in out A; V : Integer);
3234 pragma Export (C, Ada_Method2);
3236 end Simple_Cpp_Interface;
3239 @node Interfacing with C++ constructors
3240 @subsection Interfacing with C++ constructors
3243 In order to interface with C++ constructors GNAT provides the
3244 @code{pragma CPP_Constructor} (@xref{Interfacing to C++,,,
3245 gnat_rm, GNAT Reference Manual}, for additional information).
3246 In this section we present some common uses of C++ constructors
3247 in mixed-languages programs in GNAT.
3249 Let us assume that we need to interface with the following
3257 @b{virtual} int Get_Value ();
3258 Root(); // Default constructor
3259 Root(int v); // 1st non-default constructor
3260 Root(int v, int w); // 2nd non-default constructor
3264 For this purpose we can write the following package spec (further
3265 information on how to build this spec is available in
3266 @ref{Interfacing with C++ at the Class Level} and
3267 @ref{Generating Ada Bindings for C and C++ headers}).
3269 @smallexample @c ada
3270 with Interfaces.C; use Interfaces.C;
3272 type Root is tagged limited record
3276 pragma Import (CPP, Root);
3278 function Get_Value (Obj : Root) return int;
3279 pragma Import (CPP, Get_Value);
3281 function Constructor return Root'Class;
3282 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ev");
3284 function Constructor (v : Integer) return Root'Class;
3285 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ei");
3287 function Constructor (v, w : Integer) return Root'Class;
3288 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Eii");
3292 On the Ada side the constructor is represented by a function (whose
3293 name is arbitrary) that returns the classwide type corresponding to
3294 the imported C++ class. Although the constructor is described as a
3295 function, it is typically a procedure with an extra implicit argument
3296 (the object being initialized) at the implementation level. GNAT
3297 issues the appropriate call, whatever it is, to get the object
3298 properly initialized.
3300 Constructors can only appear in the following contexts:
3304 On the right side of an initialization of an object of type @var{T}.
3306 On the right side of an initialization of a record component of type @var{T}.
3308 In an Ada 2005 limited aggregate.
3310 In an Ada 2005 nested limited aggregate.
3312 In an Ada 2005 limited aggregate that initializes an object built in
3313 place by an extended return statement.
3317 In a declaration of an object whose type is a class imported from C++,
3318 either the default C++ constructor is implicitly called by GNAT, or
3319 else the required C++ constructor must be explicitly called in the
3320 expression that initializes the object. For example:
3322 @smallexample @c ada
3324 Obj2 : Root := Constructor;
3325 Obj3 : Root := Constructor (v => 10);
3326 Obj4 : Root := Constructor (30, 40);
3329 The first two declarations are equivalent: in both cases the default C++
3330 constructor is invoked (in the former case the call to the constructor is
3331 implicit, and in the latter case the call is explicit in the object
3332 declaration). @code{Obj3} is initialized by the C++ non-default constructor
3333 that takes an integer argument, and @code{Obj4} is initialized by the
3334 non-default C++ constructor that takes two integers.
3336 Let us derive the imported C++ class in the Ada side. For example:
3338 @smallexample @c ada
3339 type DT is new Root with record
3340 C_Value : Natural := 2009;
3344 In this case the components DT inherited from the C++ side must be
3345 initialized by a C++ constructor, and the additional Ada components
3346 of type DT are initialized by GNAT. The initialization of such an
3347 object is done either by default, or by means of a function returning
3348 an aggregate of type DT, or by means of an extension aggregate.
3350 @smallexample @c ada
3352 Obj6 : DT := Function_Returning_DT (50);
3353 Obj7 : DT := (Constructor (30,40) with C_Value => 50);
3356 The declaration of @code{Obj5} invokes the default constructors: the
3357 C++ default constructor of the parent type takes care of the initialization
3358 of the components inherited from Root, and GNAT takes care of the default
3359 initialization of the additional Ada components of type DT (that is,
3360 @code{C_Value} is initialized to value 2009). The order of invocation of
3361 the constructors is consistent with the order of elaboration required by
3362 Ada and C++. That is, the constructor of the parent type is always called
3363 before the constructor of the derived type.
3365 Let us now consider a record that has components whose type is imported
3366 from C++. For example:
3368 @smallexample @c ada
3369 type Rec1 is limited record
3370 Data1 : Root := Constructor (10);
3371 Value : Natural := 1000;
3374 type Rec2 (D : Integer := 20) is limited record
3376 Data2 : Root := Constructor (D, 30);
3380 The initialization of an object of type @code{Rec2} will call the
3381 non-default C++ constructors specified for the imported components.
3384 @smallexample @c ada
3388 Using Ada 2005 we can use limited aggregates to initialize an object
3389 invoking C++ constructors that differ from those specified in the type
3390 declarations. For example:
3392 @smallexample @c ada
3393 Obj9 : Rec2 := (Rec => (Data1 => Constructor (15, 16),
3398 The above declaration uses an Ada 2005 limited aggregate to
3399 initialize @code{Obj9}, and the C++ constructor that has two integer
3400 arguments is invoked to initialize the @code{Data1} component instead
3401 of the constructor specified in the declaration of type @code{Rec1}. In
3402 Ada 2005 the box in the aggregate indicates that unspecified components
3403 are initialized using the expression (if any) available in the component
3404 declaration. That is, in this case discriminant @code{D} is initialized
3405 to value @code{20}, @code{Value} is initialized to value 1000, and the
3406 non-default C++ constructor that handles two integers takes care of
3407 initializing component @code{Data2} with values @code{20,30}.
3409 In Ada 2005 we can use the extended return statement to build the Ada
3410 equivalent to C++ non-default constructors. For example:
3412 @smallexample @c ada
3413 function Constructor (V : Integer) return Rec2 is
3415 return Obj : Rec2 := (Rec => (Data1 => Constructor (V, 20),
3418 -- Further actions required for construction of
3419 -- objects of type Rec2
3425 In this example the extended return statement construct is used to
3426 build in place the returned object whose components are initialized
3427 by means of a limited aggregate. Any further action associated with
3428 the constructor can be placed inside the construct.
3430 @node Interfacing with C++ at the Class Level
3431 @subsection Interfacing with C++ at the Class Level
3433 In this section we demonstrate the GNAT features for interfacing with
3434 C++ by means of an example making use of Ada 2005 abstract interface
3435 types. This example consists of a classification of animals; classes
3436 have been used to model our main classification of animals, and
3437 interfaces provide support for the management of secondary
3438 classifications. We first demonstrate a case in which the types and
3439 constructors are defined on the C++ side and imported from the Ada
3440 side, and latter the reverse case.
3442 The root of our derivation will be the @code{Animal} class, with a
3443 single private attribute (the @code{Age} of the animal) and two public
3444 primitives to set and get the value of this attribute.
3449 @b{virtual} void Set_Age (int New_Age);
3450 @b{virtual} int Age ();
3456 Abstract interface types are defined in C++ by means of classes with pure
3457 virtual functions and no data members. In our example we will use two
3458 interfaces that provide support for the common management of @code{Carnivore}
3459 and @code{Domestic} animals:
3462 @b{class} Carnivore @{
3464 @b{virtual} int Number_Of_Teeth () = 0;
3467 @b{class} Domestic @{
3469 @b{virtual void} Set_Owner (char* Name) = 0;
3473 Using these declarations, we can now say that a @code{Dog} is an animal that is
3474 both Carnivore and Domestic, that is:
3477 @b{class} Dog : Animal, Carnivore, Domestic @{
3479 @b{virtual} int Number_Of_Teeth ();
3480 @b{virtual} void Set_Owner (char* Name);
3482 Dog(); // Constructor
3489 In the following examples we will assume that the previous declarations are
3490 located in a file named @code{animals.h}. The following package demonstrates
3491 how to import these C++ declarations from the Ada side:
3493 @smallexample @c ada
3494 with Interfaces.C.Strings; use Interfaces.C.Strings;
3496 type Carnivore is interface;
3497 pragma Convention (C_Plus_Plus, Carnivore);
3498 function Number_Of_Teeth (X : Carnivore)
3499 return Natural is abstract;
3501 type Domestic is interface;
3502 pragma Convention (C_Plus_Plus, Set_Owner);
3504 (X : in out Domestic;
3505 Name : Chars_Ptr) is abstract;
3507 type Animal is tagged record
3510 pragma Import (C_Plus_Plus, Animal);
3512 procedure Set_Age (X : in out Animal; Age : Integer);
3513 pragma Import (C_Plus_Plus, Set_Age);
3515 function Age (X : Animal) return Integer;
3516 pragma Import (C_Plus_Plus, Age);
3518 type Dog is new Animal and Carnivore and Domestic with record
3519 Tooth_Count : Natural;
3520 Owner : String (1 .. 30);
3522 pragma Import (C_Plus_Plus, Dog);
3524 function Number_Of_Teeth (A : Dog) return Integer;
3525 pragma Import (C_Plus_Plus, Number_Of_Teeth);
3527 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3528 pragma Import (C_Plus_Plus, Set_Owner);
3530 function New_Dog return Dog'Class;
3531 pragma CPP_Constructor (New_Dog);
3532 pragma Import (CPP, New_Dog, "_ZN3DogC2Ev");
3536 Thanks to the compatibility between GNAT run-time structures and the C++ ABI,
3537 interfacing with these C++ classes is easy. The only requirement is that all
3538 the primitives and components must be declared exactly in the same order in
3541 Regarding the abstract interfaces, we must indicate to the GNAT compiler by
3542 means of a @code{pragma Convention (C_Plus_Plus)}, the convention used to pass
3543 the arguments to the called primitives will be the same as for C++. For the
3544 imported classes we use @code{pragma Import} with convention @code{C_Plus_Plus}
3545 to indicate that they have been defined on the C++ side; this is required
3546 because the dispatch table associated with these tagged types will be built
3547 in the C++ side and therefore will not contain the predefined Ada primitives
3548 which Ada would otherwise expect.
3550 As the reader can see there is no need to indicate the C++ mangled names
3551 associated with each subprogram because it is assumed that all the calls to
3552 these primitives will be dispatching calls. The only exception is the
3553 constructor, which must be registered with the compiler by means of
3554 @code{pragma CPP_Constructor} and needs to provide its associated C++
3555 mangled name because the Ada compiler generates direct calls to it.
3557 With the above packages we can now declare objects of type Dog on the Ada side
3558 and dispatch calls to the corresponding subprograms on the C++ side. We can
3559 also extend the tagged type Dog with further fields and primitives, and
3560 override some of its C++ primitives on the Ada side. For example, here we have
3561 a type derivation defined on the Ada side that inherits all the dispatching
3562 primitives of the ancestor from the C++ side.
3565 @b{with} Animals; @b{use} Animals;
3566 @b{package} Vaccinated_Animals @b{is}
3567 @b{type} Vaccinated_Dog @b{is new} Dog @b{with null record};
3568 @b{function} Vaccination_Expired (A : Vaccinated_Dog) @b{return} Boolean;
3569 @b{end} Vaccinated_Animals;
3572 It is important to note that, because of the ABI compatibility, the programmer
3573 does not need to add any further information to indicate either the object
3574 layout or the dispatch table entry associated with each dispatching operation.
3576 Now let us define all the types and constructors on the Ada side and export
3577 them to C++, using the same hierarchy of our previous example:
3579 @smallexample @c ada
3580 with Interfaces.C.Strings;
3581 use Interfaces.C.Strings;
3583 type Carnivore is interface;
3584 pragma Convention (C_Plus_Plus, Carnivore);
3585 function Number_Of_Teeth (X : Carnivore)
3586 return Natural is abstract;
3588 type Domestic is interface;
3589 pragma Convention (C_Plus_Plus, Set_Owner);
3591 (X : in out Domestic;
3592 Name : Chars_Ptr) is abstract;
3594 type Animal is tagged record
3597 pragma Convention (C_Plus_Plus, Animal);
3599 procedure Set_Age (X : in out Animal; Age : Integer);
3600 pragma Export (C_Plus_Plus, Set_Age);
3602 function Age (X : Animal) return Integer;
3603 pragma Export (C_Plus_Plus, Age);
3605 type Dog is new Animal and Carnivore and Domestic with record
3606 Tooth_Count : Natural;
3607 Owner : String (1 .. 30);
3609 pragma Convention (C_Plus_Plus, Dog);
3611 function Number_Of_Teeth (A : Dog) return Integer;
3612 pragma Export (C_Plus_Plus, Number_Of_Teeth);
3614 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3615 pragma Export (C_Plus_Plus, Set_Owner);
3617 function New_Dog return Dog'Class;
3618 pragma Export (C_Plus_Plus, New_Dog);
3622 Compared with our previous example the only difference is the use of
3623 @code{pragma Export} to indicate to the GNAT compiler that the primitives will
3624 be available to C++. Thanks to the ABI compatibility, on the C++ side there is
3625 nothing else to be done; as explained above, the only requirement is that all
3626 the primitives and components are declared in exactly the same order.
3628 For completeness, let us see a brief C++ main program that uses the
3629 declarations available in @code{animals.h} (presented in our first example) to
3630 import and use the declarations from the Ada side, properly initializing and
3631 finalizing the Ada run-time system along the way:
3634 @b{#include} "animals.h"
3635 @b{#include} <iostream>
3636 @b{using namespace} std;
3638 void Check_Carnivore (Carnivore *obj) @{@dots{}@}
3639 void Check_Domestic (Domestic *obj) @{@dots{}@}
3640 void Check_Animal (Animal *obj) @{@dots{}@}
3641 void Check_Dog (Dog *obj) @{@dots{}@}
3644 void adainit (void);
3645 void adafinal (void);
3651 Dog *obj = new_dog(); // Ada constructor
3652 Check_Carnivore (obj); // Check secondary DT
3653 Check_Domestic (obj); // Check secondary DT
3654 Check_Animal (obj); // Check primary DT
3655 Check_Dog (obj); // Check primary DT
3660 adainit (); test(); adafinal ();
3665 @node Comparison between GNAT and C/C++ Compilation Models
3666 @section Comparison between GNAT and C/C++ Compilation Models
3669 The GNAT model of compilation is close to the C and C++ models. You can
3670 think of Ada specs as corresponding to header files in C. As in C, you
3671 don't need to compile specs; they are compiled when they are used. The
3672 Ada @code{with} is similar in effect to the @code{#include} of a C
3675 One notable difference is that, in Ada, you may compile specs separately
3676 to check them for semantic and syntactic accuracy. This is not always
3677 possible with C headers because they are fragments of programs that have
3678 less specific syntactic or semantic rules.
3680 The other major difference is the requirement for running the binder,
3681 which performs two important functions. First, it checks for
3682 consistency. In C or C++, the only defense against assembling
3683 inconsistent programs lies outside the compiler, in a makefile, for
3684 example. The binder satisfies the Ada requirement that it be impossible
3685 to construct an inconsistent program when the compiler is used in normal
3688 @cindex Elaboration order control
3689 The other important function of the binder is to deal with elaboration
3690 issues. There are also elaboration issues in C++ that are handled
3691 automatically. This automatic handling has the advantage of being
3692 simpler to use, but the C++ programmer has no control over elaboration.
3693 Where @code{gnatbind} might complain there was no valid order of
3694 elaboration, a C++ compiler would simply construct a program that
3695 malfunctioned at run time.
3698 @node Comparison between GNAT and Conventional Ada Library Models
3699 @section Comparison between GNAT and Conventional Ada Library Models
3702 This section is intended for Ada programmers who have
3703 used an Ada compiler implementing the traditional Ada library
3704 model, as described in the Ada Reference Manual.
3706 @cindex GNAT library
3707 In GNAT, there is no ``library'' in the normal sense. Instead, the set of
3708 source files themselves acts as the library. Compiling Ada programs does
3709 not generate any centralized information, but rather an object file and
3710 a ALI file, which are of interest only to the binder and linker.
3711 In a traditional system, the compiler reads information not only from
3712 the source file being compiled, but also from the centralized library.
3713 This means that the effect of a compilation depends on what has been
3714 previously compiled. In particular:
3718 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3719 to the version of the unit most recently compiled into the library.
3722 Inlining is effective only if the necessary body has already been
3723 compiled into the library.
3726 Compiling a unit may obsolete other units in the library.
3730 In GNAT, compiling one unit never affects the compilation of any other
3731 units because the compiler reads only source files. Only changes to source
3732 files can affect the results of a compilation. In particular:
3736 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3737 to the source version of the unit that is currently accessible to the
3742 Inlining requires the appropriate source files for the package or
3743 subprogram bodies to be available to the compiler. Inlining is always
3744 effective, independent of the order in which units are complied.
3747 Compiling a unit never affects any other compilations. The editing of
3748 sources may cause previous compilations to be out of date if they
3749 depended on the source file being modified.
3753 The most important result of these differences is that order of compilation
3754 is never significant in GNAT. There is no situation in which one is
3755 required to do one compilation before another. What shows up as order of
3756 compilation requirements in the traditional Ada library becomes, in
3757 GNAT, simple source dependencies; in other words, there is only a set
3758 of rules saying what source files must be present when a file is
3762 @node Placement of temporary files
3763 @section Placement of temporary files
3764 @cindex Temporary files (user control over placement)
3767 GNAT creates temporary files in the directory designated by the environment
3768 variable @env{TMPDIR}.
3769 (See the HP @emph{C RTL Reference Manual} on the function @code{getenv()}
3770 for detailed information on how environment variables are resolved.
3771 For most users the easiest way to make use of this feature is to simply
3772 define @env{TMPDIR} as a job level logical name).
3773 For example, if you wish to use a Ramdisk (assuming DECRAM is installed)
3774 for compiler temporary files, then you can include something like the
3775 following command in your @file{LOGIN.COM} file:
3778 $ define/job TMPDIR "/disk$scratchram/000000/temp/"
3782 If @env{TMPDIR} is not defined, then GNAT uses the directory designated by
3783 @env{TMP}; if @env{TMP} is not defined, then GNAT uses the directory
3784 designated by @env{TEMP}.
3785 If none of these environment variables are defined then GNAT uses the
3786 directory designated by the logical name @code{SYS$SCRATCH:}
3787 (by default the user's home directory). If all else fails
3788 GNAT uses the current directory for temporary files.
3791 @c *************************
3792 @node Compiling Using gcc
3793 @chapter Compiling Using @command{gcc}
3796 This chapter discusses how to compile Ada programs using the @command{gcc}
3797 command. It also describes the set of switches
3798 that can be used to control the behavior of the compiler.
3800 * Compiling Programs::
3801 * Switches for gcc::
3802 * Search Paths and the Run-Time Library (RTL)::
3803 * Order of Compilation Issues::
3807 @node Compiling Programs
3808 @section Compiling Programs
3811 The first step in creating an executable program is to compile the units
3812 of the program using the @command{gcc} command. You must compile the
3817 the body file (@file{.adb}) for a library level subprogram or generic
3821 the spec file (@file{.ads}) for a library level package or generic
3822 package that has no body
3825 the body file (@file{.adb}) for a library level package
3826 or generic package that has a body
3831 You need @emph{not} compile the following files
3836 the spec of a library unit which has a body
3843 because they are compiled as part of compiling related units. GNAT
3845 when the corresponding body is compiled, and subunits when the parent is
3848 @cindex cannot generate code
3849 If you attempt to compile any of these files, you will get one of the
3850 following error messages (where @var{fff} is the name of the file you compiled):
3853 cannot generate code for file @var{fff} (package spec)
3854 to check package spec, use -gnatc
3856 cannot generate code for file @var{fff} (missing subunits)
3857 to check parent unit, use -gnatc
3859 cannot generate code for file @var{fff} (subprogram spec)
3860 to check subprogram spec, use -gnatc
3862 cannot generate code for file @var{fff} (subunit)
3863 to check subunit, use -gnatc
3867 As indicated by the above error messages, if you want to submit
3868 one of these files to the compiler to check for correct semantics
3869 without generating code, then use the @option{-gnatc} switch.
3871 The basic command for compiling a file containing an Ada unit is
3874 $ gcc -c @ovar{switches} @file{file name}
3878 where @var{file name} is the name of the Ada file (usually
3880 @file{.ads} for a spec or @file{.adb} for a body).
3883 @option{-c} switch to tell @command{gcc} to compile, but not link, the file.
3885 The result of a successful compilation is an object file, which has the
3886 same name as the source file but an extension of @file{.o} and an Ada
3887 Library Information (ALI) file, which also has the same name as the
3888 source file, but with @file{.ali} as the extension. GNAT creates these
3889 two output files in the current directory, but you may specify a source
3890 file in any directory using an absolute or relative path specification
3891 containing the directory information.
3894 @command{gcc} is actually a driver program that looks at the extensions of
3895 the file arguments and loads the appropriate compiler. For example, the
3896 GNU C compiler is @file{cc1}, and the Ada compiler is @file{gnat1}.
3897 These programs are in directories known to the driver program (in some
3898 configurations via environment variables you set), but need not be in
3899 your path. The @command{gcc} driver also calls the assembler and any other
3900 utilities needed to complete the generation of the required object
3903 It is possible to supply several file names on the same @command{gcc}
3904 command. This causes @command{gcc} to call the appropriate compiler for
3905 each file. For example, the following command lists three separate
3906 files to be compiled:
3909 $ gcc -c x.adb y.adb z.c
3913 calls @code{gnat1} (the Ada compiler) twice to compile @file{x.adb} and
3914 @file{y.adb}, and @code{cc1} (the C compiler) once to compile @file{z.c}.
3915 The compiler generates three object files @file{x.o}, @file{y.o} and
3916 @file{z.o} and the two ALI files @file{x.ali} and @file{y.ali} from the
3917 Ada compilations. Any switches apply to all the files ^listed,^listed.^
3920 @option{-gnat@var{x}} switches, which apply only to Ada compilations.
3923 @node Switches for gcc
3924 @section Switches for @command{gcc}
3927 The @command{gcc} command accepts switches that control the
3928 compilation process. These switches are fully described in this section.
3929 First we briefly list all the switches, in alphabetical order, then we
3930 describe the switches in more detail in functionally grouped sections.
3932 More switches exist for GCC than those documented here, especially
3933 for specific targets. However, their use is not recommended as
3934 they may change code generation in ways that are incompatible with
3935 the Ada run-time library, or can cause inconsistencies between
3939 * Output and Error Message Control::
3940 * Warning Message Control::
3941 * Debugging and Assertion Control::
3942 * Validity Checking::
3945 * Using gcc for Syntax Checking::
3946 * Using gcc for Semantic Checking::
3947 * Compiling Different Versions of Ada::
3948 * Character Set Control::
3949 * File Naming Control::
3950 * Subprogram Inlining Control::
3951 * Auxiliary Output Control::
3952 * Debugging Control::
3953 * Exception Handling Control::
3954 * Units to Sources Mapping Files::
3955 * Integrated Preprocessing::
3956 * Code Generation Control::
3965 @cindex @option{-b} (@command{gcc})
3966 @item -b @var{target}
3967 Compile your program to run on @var{target}, which is the name of a
3968 system configuration. You must have a GNAT cross-compiler built if
3969 @var{target} is not the same as your host system.
3972 @cindex @option{-B} (@command{gcc})
3973 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
3974 from @var{dir} instead of the default location. Only use this switch
3975 when multiple versions of the GNAT compiler are available.
3976 @xref{Directory Options,, Options for Directory Search, gcc, Using the
3977 GNU Compiler Collection (GCC)}, for further details. You would normally
3978 use the @option{-b} or @option{-V} switch instead.
3981 @cindex @option{-c} (@command{gcc})
3982 Compile. Always use this switch when compiling Ada programs.
3984 Note: for some other languages when using @command{gcc}, notably in
3985 the case of C and C++, it is possible to use
3986 use @command{gcc} without a @option{-c} switch to
3987 compile and link in one step. In the case of GNAT, you
3988 cannot use this approach, because the binder must be run
3989 and @command{gcc} cannot be used to run the GNAT binder.
3993 @cindex @option{-fno-inline} (@command{gcc})
3994 Suppresses all back-end inlining, even if other optimization or inlining
3996 This includes suppression of inlining that results
3997 from the use of the pragma @code{Inline_Always}.
3998 Any occurrences of pragma @code{Inline} or @code{Inline_Always}
3999 are ignored, and @option{-gnatn} and @option{-gnatN} have no
4000 effect if this switch is present.
4002 @item -fno-inline-functions
4003 @cindex @option{-fno-inline-functions} (@command{gcc})
4004 Suppresses automatic inlining of simple subprograms, which is enabled
4005 if @option{-O3} is used.
4007 @item -fno-inline-small-functions
4008 @cindex @option{-fno-inline-small-functions} (@command{gcc})
4009 Suppresses automatic inlining of small subprograms, which is enabled
4010 if @option{-O2} is used.
4012 @item -fno-inline-functions-called-once
4013 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
4014 Suppresses inlining of subprograms local to the unit and called once
4015 from within it, which is enabled if @option{-O1} is used.
4018 @cindex @option{-fno-ivopts} (@command{gcc})
4019 Suppresses high-level loop induction variable optimizations, which are
4020 enabled if @option{-O1} is used. These optimizations are generally
4021 profitable but, for some specific cases of loops with numerous uses
4022 of the iteration variable that follow a common pattern, they may end
4023 up destroying the regularity that could be exploited at a lower level
4024 and thus producing inferior code.
4026 @item -fno-strict-aliasing
4027 @cindex @option{-fno-strict-aliasing} (@command{gcc})
4028 Causes the compiler to avoid assumptions regarding non-aliasing
4029 of objects of different types. See
4030 @ref{Optimization and Strict Aliasing} for details.
4033 @cindex @option{-fstack-check} (@command{gcc})
4034 Activates stack checking.
4035 See @ref{Stack Overflow Checking} for details.
4038 @cindex @option{-fstack-usage} (@command{gcc})
4039 Makes the compiler output stack usage information for the program, on a
4040 per-function basis. See @ref{Static Stack Usage Analysis} for details.
4042 @item -fcallgraph-info@r{[}=su@r{]}
4043 @cindex @option{-fcallgraph-info} (@command{gcc})
4044 Makes the compiler output callgraph information for the program, on a
4045 per-file basis. The information is generated in the VCG format. It can
4046 be decorated with stack-usage per-node information.
4049 @cindex @option{^-g^/DEBUG^} (@command{gcc})
4050 Generate debugging information. This information is stored in the object
4051 file and copied from there to the final executable file by the linker,
4052 where it can be read by the debugger. You must use the
4053 @option{^-g^/DEBUG^} switch if you plan on using the debugger.
4056 @cindex @option{-gnat83} (@command{gcc})
4057 Enforce Ada 83 restrictions.
4060 @cindex @option{-gnat95} (@command{gcc})
4061 Enforce Ada 95 restrictions.
4064 @cindex @option{-gnat05} (@command{gcc})
4065 Allow full Ada 2005 features.
4068 @cindex @option{-gnata} (@command{gcc})
4069 Assertions enabled. @code{Pragma Assert} and @code{pragma Debug} to be
4070 activated. Note that these pragmas can also be controlled using the
4071 configuration pragmas @code{Assertion_Policy} and @code{Debug_Policy}.
4072 It also activates pragmas @code{Check}, @code{Precondition}, and
4073 @code{Postcondition}. Note that these pragmas can also be controlled
4074 using the configuration pragma @code{Check_Policy}.
4077 @cindex @option{-gnatA} (@command{gcc})
4078 Avoid processing @file{gnat.adc}. If a @file{gnat.adc} file is present,
4082 @cindex @option{-gnatb} (@command{gcc})
4083 Generate brief messages to @file{stderr} even if verbose mode set.
4086 @cindex @option{-gnatB} (@command{gcc})
4087 Assume no invalid (bad) values except for 'Valid attribute use.
4090 @cindex @option{-gnatc} (@command{gcc})
4091 Check syntax and semantics only (no code generation attempted).
4094 @cindex @option{-gnatd} (@command{gcc})
4095 Specify debug options for the compiler. The string of characters after
4096 the @option{-gnatd} specify the specific debug options. The possible
4097 characters are 0-9, a-z, A-Z, optionally preceded by a dot. See
4098 compiler source file @file{debug.adb} for details of the implemented
4099 debug options. Certain debug options are relevant to applications
4100 programmers, and these are documented at appropriate points in this
4105 @cindex @option{-gnatD[nn]} (@command{gcc})
4108 @item /XDEBUG /LXDEBUG=nnn
4110 Create expanded source files for source level debugging. This switch
4111 also suppress generation of cross-reference information
4112 (see @option{-gnatx}).
4114 @item -gnatec=@var{path}
4115 @cindex @option{-gnatec} (@command{gcc})
4116 Specify a configuration pragma file
4118 (the equal sign is optional)
4120 (@pxref{The Configuration Pragmas Files}).
4122 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=@var{value}@r{]}
4123 @cindex @option{-gnateD} (@command{gcc})
4124 Defines a symbol, associated with @var{value}, for preprocessing.
4125 (@pxref{Integrated Preprocessing}).
4128 @cindex @option{-gnatef} (@command{gcc})
4129 Display full source path name in brief error messages.
4132 @cindex @option{-gnateG} (@command{gcc})
4133 Save result of preprocessing in a text file.
4135 @item -gnatem=@var{path}
4136 @cindex @option{-gnatem} (@command{gcc})
4137 Specify a mapping file
4139 (the equal sign is optional)
4141 (@pxref{Units to Sources Mapping Files}).
4143 @item -gnatep=@var{file}
4144 @cindex @option{-gnatep} (@command{gcc})
4145 Specify a preprocessing data file
4147 (the equal sign is optional)
4149 (@pxref{Integrated Preprocessing}).
4152 @cindex @option{-gnatE} (@command{gcc})
4153 Full dynamic elaboration checks.
4156 @cindex @option{-gnatf} (@command{gcc})
4157 Full errors. Multiple errors per line, all undefined references, do not
4158 attempt to suppress cascaded errors.
4161 @cindex @option{-gnatF} (@command{gcc})
4162 Externals names are folded to all uppercase.
4164 @item ^-gnatg^/GNAT_INTERNAL^
4165 @cindex @option{^-gnatg^/GNAT_INTERNAL^} (@command{gcc})
4166 Internal GNAT implementation mode. This should not be used for
4167 applications programs, it is intended only for use by the compiler
4168 and its run-time library. For documentation, see the GNAT sources.
4169 Note that @option{^-gnatg^/GNAT_INTERNAL^} implies
4170 @option{^-gnatwae^/WARNINGS=ALL,ERRORS^} and
4171 @option{^-gnatyg^/STYLE_CHECKS=GNAT^}
4172 so that all standard warnings and all standard style options are turned on.
4173 All warnings and style error messages are treated as errors.
4177 @cindex @option{-gnatG[nn]} (@command{gcc})
4180 @item /EXPAND_SOURCE, /LEXPAND_SOURCE=nnn
4182 List generated expanded code in source form.
4184 @item ^-gnath^/HELP^
4185 @cindex @option{^-gnath^/HELP^} (@command{gcc})
4186 Output usage information. The output is written to @file{stdout}.
4188 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
4189 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
4190 Identifier character set
4192 (@var{c}=1/2/3/4/8/9/p/f/n/w).
4194 For details of the possible selections for @var{c},
4195 see @ref{Character Set Control}.
4197 @item ^-gnatI^/IGNORE_REP_CLAUSES^
4198 @cindex @option{^-gnatI^IGNORE_REP_CLAUSES^} (@command{gcc})
4199 Ignore representation clauses. When this switch is used,
4200 representation clauses are treated as comments. This is useful
4201 when initially porting code where you want to ignore rep clause
4202 problems, and also for compiling foreign code (particularly
4203 for use with ASIS). The representation clauses that are ignored
4204 are: enumeration_representation_clause, record_representation_clause,
4205 and attribute_definition_clause for the following attributes:
4206 Address, Alignment, Bit_Order, Component_Size, Machine_Radix,
4207 Object_Size, Size, Small, Stream_Size, and Value_Size.
4208 Note that this option should be used only for compiling -- the
4209 code is likely to malfunction at run time.
4212 @cindex @option{-gnatjnn} (@command{gcc})
4213 Reformat error messages to fit on nn character lines
4215 @item -gnatk=@var{n}
4216 @cindex @option{-gnatk} (@command{gcc})
4217 Limit file names to @var{n} (1-999) characters ^(@code{k} = krunch)^^.
4220 @cindex @option{-gnatl} (@command{gcc})
4221 Output full source listing with embedded error messages.
4224 @cindex @option{-gnatL} (@command{gcc})
4225 Used in conjunction with -gnatG or -gnatD to intersperse original
4226 source lines (as comment lines with line numbers) in the expanded
4229 @item -gnatm=@var{n}
4230 @cindex @option{-gnatm} (@command{gcc})
4231 Limit number of detected error or warning messages to @var{n}
4232 where @var{n} is in the range 1..999999. The default setting if
4233 no switch is given is 9999. If the number of warnings reaches this
4234 limit, then a message is output and further warnings are suppressed,
4235 but the compilation is continued. If the number of error messages
4236 reaches this limit, then a message is output and the compilation
4237 is abandoned. The equal sign here is optional. A value of zero
4238 means that no limit applies.
4241 @cindex @option{-gnatn} (@command{gcc})
4242 Activate inlining for subprograms for which
4243 pragma @code{inline} is specified. This inlining is performed
4244 by the GCC back-end.
4247 @cindex @option{-gnatN} (@command{gcc})
4248 Activate front end inlining for subprograms for which
4249 pragma @code{Inline} is specified. This inlining is performed
4250 by the front end and will be visible in the
4251 @option{-gnatG} output.
4253 When using a gcc-based back end (in practice this means using any version
4254 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
4255 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
4256 Historically front end inlining was more extensive than the gcc back end
4257 inlining, but that is no longer the case.
4260 @cindex @option{-gnato} (@command{gcc})
4261 Enable numeric overflow checking (which is not normally enabled by
4262 default). Note that division by zero is a separate check that is not
4263 controlled by this switch (division by zero checking is on by default).
4266 @cindex @option{-gnatp} (@command{gcc})
4267 Suppress all checks. See @ref{Run-Time Checks} for details.
4270 @cindex @option{-gnatP} (@command{gcc})
4271 Enable polling. This is required on some systems (notably Windows NT) to
4272 obtain asynchronous abort and asynchronous transfer of control capability.
4273 @xref{Pragma Polling,,, gnat_rm, GNAT Reference Manual}, for full
4277 @cindex @option{-gnatq} (@command{gcc})
4278 Don't quit. Try semantics, even if parse errors.
4281 @cindex @option{-gnatQ} (@command{gcc})
4282 Don't quit. Generate @file{ALI} and tree files even if illegalities.
4285 @cindex @option{-gnatr} (@command{gcc})
4286 Treat pragma Restrictions as Restriction_Warnings.
4288 @item ^-gnatR@r{[}0@r{/}1@r{/}2@r{/}3@r{[}s@r{]]}^/REPRESENTATION_INFO^
4289 @cindex @option{-gnatR} (@command{gcc})
4290 Output representation information for declared types and objects.
4293 @cindex @option{-gnats} (@command{gcc})
4297 @cindex @option{-gnatS} (@command{gcc})
4298 Print package Standard.
4301 @cindex @option{-gnatt} (@command{gcc})
4302 Generate tree output file.
4304 @item ^-gnatT^/TABLE_MULTIPLIER=^@var{nnn}
4305 @cindex @option{^-gnatT^/TABLE_MULTIPLIER^} (@command{gcc})
4306 All compiler tables start at @var{nnn} times usual starting size.
4309 @cindex @option{-gnatu} (@command{gcc})
4310 List units for this compilation.
4313 @cindex @option{-gnatU} (@command{gcc})
4314 Tag all error messages with the unique string ``error:''
4317 @cindex @option{-gnatv} (@command{gcc})
4318 Verbose mode. Full error output with source lines to @file{stdout}.
4321 @cindex @option{-gnatV} (@command{gcc})
4322 Control level of validity checking. See separate section describing
4325 @item ^-gnatw@var{xxx}^/WARNINGS=(@var{option}@r{[},@dots{}@r{]})^
4326 @cindex @option{^-gnatw^/WARNINGS^} (@command{gcc})
4328 ^@var{xxx} is a string of option letters that^the list of options^ denotes
4329 the exact warnings that
4330 are enabled or disabled (@pxref{Warning Message Control}).
4332 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
4333 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
4334 Wide character encoding method
4336 (@var{e}=n/h/u/s/e/8).
4339 (@var{e}=@code{BRACKETS, NONE, HEX, UPPER, SHIFT_JIS, EUC, UTF8})
4343 @cindex @option{-gnatx} (@command{gcc})
4344 Suppress generation of cross-reference information.
4346 @item ^-gnaty^/STYLE_CHECKS=(option,option@dots{})^
4347 @cindex @option{^-gnaty^/STYLE_CHECKS^} (@command{gcc})
4348 Enable built-in style checks (@pxref{Style Checking}).
4350 @item ^-gnatz^/DISTRIBUTION_STUBS=^@var{m}
4351 @cindex @option{^-gnatz^/DISTRIBUTION_STUBS^} (@command{gcc})
4352 Distribution stub generation and compilation
4354 (@var{m}=r/c for receiver/caller stubs).
4357 (@var{m}=@code{RECEIVER} or @code{CALLER} to specify the type of stubs
4358 to be generated and compiled).
4361 @item ^-I^/SEARCH=^@var{dir}
4362 @cindex @option{^-I^/SEARCH^} (@command{gcc})
4364 Direct GNAT to search the @var{dir} directory for source files needed by
4365 the current compilation
4366 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4368 @item ^-I-^/NOCURRENT_DIRECTORY^
4369 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gcc})
4371 Except for the source file named in the command line, do not look for source
4372 files in the directory containing the source file named in the command line
4373 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4377 @cindex @option{-mbig-switch} (@command{gcc})
4378 @cindex @code{case} statement (effect of @option{-mbig-switch} option)
4379 This standard gcc switch causes the compiler to use larger offsets in its
4380 jump table representation for @code{case} statements.
4381 This may result in less efficient code, but is sometimes necessary
4382 (for example on HP-UX targets)
4383 @cindex HP-UX and @option{-mbig-switch} option
4384 in order to compile large and/or nested @code{case} statements.
4387 @cindex @option{-o} (@command{gcc})
4388 This switch is used in @command{gcc} to redirect the generated object file
4389 and its associated ALI file. Beware of this switch with GNAT, because it may
4390 cause the object file and ALI file to have different names which in turn
4391 may confuse the binder and the linker.
4395 @cindex @option{-nostdinc} (@command{gcc})
4396 Inhibit the search of the default location for the GNAT Run Time
4397 Library (RTL) source files.
4400 @cindex @option{-nostdlib} (@command{gcc})
4401 Inhibit the search of the default location for the GNAT Run Time
4402 Library (RTL) ALI files.
4406 @cindex @option{-O} (@command{gcc})
4407 @var{n} controls the optimization level.
4411 No optimization, the default setting if no @option{-O} appears
4414 Normal optimization, the default if you specify @option{-O} without
4415 an operand. A good compromise between code quality and compilation
4419 Extensive optimization, may improve execution time, possibly at the cost of
4420 substantially increased compilation time.
4423 Same as @option{-O2}, and also includes inline expansion for small subprograms
4427 Optimize space usage
4431 See also @ref{Optimization Levels}.
4436 @cindex @option{/NOOPTIMIZE} (@code{GNAT COMPILE})
4437 Equivalent to @option{/OPTIMIZE=NONE}.
4438 This is the default behavior in the absence of an @option{/OPTIMIZE}
4441 @item /OPTIMIZE@r{[}=(keyword@r{[},@dots{}@r{]})@r{]}
4442 @cindex @option{/OPTIMIZE} (@code{GNAT COMPILE})
4443 Selects the level of optimization for your program. The supported
4444 keywords are as follows:
4447 Perform most optimizations, including those that
4449 This is the default if the @option{/OPTIMIZE} qualifier is supplied
4450 without keyword options.
4453 Do not do any optimizations. Same as @code{/NOOPTIMIZE}.
4456 Perform some optimizations, but omit ones that are costly.
4459 Same as @code{SOME}.
4462 Full optimization as in @option{/OPTIMIZE=ALL}, and also attempts
4463 automatic inlining of small subprograms within a unit
4466 Try to unroll loops. This keyword may be specified together with
4467 any keyword above other than @code{NONE}. Loop unrolling
4468 usually, but not always, improves the performance of programs.
4471 Optimize space usage
4475 See also @ref{Optimization Levels}.
4479 @item -pass-exit-codes
4480 @cindex @option{-pass-exit-codes} (@command{gcc})
4481 Catch exit codes from the compiler and use the most meaningful as
4485 @item --RTS=@var{rts-path}
4486 @cindex @option{--RTS} (@command{gcc})
4487 Specifies the default location of the runtime library. Same meaning as the
4488 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
4491 @cindex @option{^-S^/ASM^} (@command{gcc})
4492 ^Used in place of @option{-c} to^Used to^
4493 cause the assembler source file to be
4494 generated, using @file{^.s^.S^} as the extension,
4495 instead of the object file.
4496 This may be useful if you need to examine the generated assembly code.
4498 @item ^-fverbose-asm^/VERBOSE_ASM^
4499 @cindex @option{^-fverbose-asm^/VERBOSE_ASM^} (@command{gcc})
4500 ^Used in conjunction with @option{-S}^Used in place of @option{/ASM}^
4501 to cause the generated assembly code file to be annotated with variable
4502 names, making it significantly easier to follow.
4505 @cindex @option{^-v^/VERBOSE^} (@command{gcc})
4506 Show commands generated by the @command{gcc} driver. Normally used only for
4507 debugging purposes or if you need to be sure what version of the
4508 compiler you are executing.
4512 @cindex @option{-V} (@command{gcc})
4513 Execute @var{ver} version of the compiler. This is the @command{gcc}
4514 version, not the GNAT version.
4517 @item ^-w^/NO_BACK_END_WARNINGS^
4518 @cindex @option{-w} (@command{gcc})
4519 Turn off warnings generated by the back end of the compiler. Use of
4520 this switch also causes the default for front end warnings to be set
4521 to suppress (as though @option{-gnatws} had appeared at the start of
4527 @c Combining qualifiers does not work on VMS
4528 You may combine a sequence of GNAT switches into a single switch. For
4529 example, the combined switch
4531 @cindex Combining GNAT switches
4537 is equivalent to specifying the following sequence of switches:
4540 -gnato -gnatf -gnati3
4545 The following restrictions apply to the combination of switches
4550 The switch @option{-gnatc} if combined with other switches must come
4551 first in the string.
4554 The switch @option{-gnats} if combined with other switches must come
4555 first in the string.
4559 @option{^-gnatz^/DISTRIBUTION_STUBS^}, @option{-gnatzc}, and @option{-gnatzr}
4560 may not be combined with any other switches.
4564 Once a ``y'' appears in the string (that is a use of the @option{-gnaty}
4565 switch), then all further characters in the switch are interpreted
4566 as style modifiers (see description of @option{-gnaty}).
4569 Once a ``d'' appears in the string (that is a use of the @option{-gnatd}
4570 switch), then all further characters in the switch are interpreted
4571 as debug flags (see description of @option{-gnatd}).
4574 Once a ``w'' appears in the string (that is a use of the @option{-gnatw}
4575 switch), then all further characters in the switch are interpreted
4576 as warning mode modifiers (see description of @option{-gnatw}).
4579 Once a ``V'' appears in the string (that is a use of the @option{-gnatV}
4580 switch), then all further characters in the switch are interpreted
4581 as validity checking options (see description of @option{-gnatV}).
4585 @node Output and Error Message Control
4586 @subsection Output and Error Message Control
4590 The standard default format for error messages is called ``brief format''.
4591 Brief format messages are written to @file{stderr} (the standard error
4592 file) and have the following form:
4595 e.adb:3:04: Incorrect spelling of keyword "function"
4596 e.adb:4:20: ";" should be "is"
4600 The first integer after the file name is the line number in the file,
4601 and the second integer is the column number within the line.
4603 @code{GPS} can parse the error messages
4604 and point to the referenced character.
4606 The following switches provide control over the error message
4612 @cindex @option{-gnatv} (@command{gcc})
4615 The v stands for verbose.
4617 The effect of this setting is to write long-format error
4618 messages to @file{stdout} (the standard output file.
4619 The same program compiled with the
4620 @option{-gnatv} switch would generate:
4624 3. funcion X (Q : Integer)
4626 >>> Incorrect spelling of keyword "function"
4629 >>> ";" should be "is"
4634 The vertical bar indicates the location of the error, and the @samp{>>>}
4635 prefix can be used to search for error messages. When this switch is
4636 used the only source lines output are those with errors.
4639 @cindex @option{-gnatl} (@command{gcc})
4641 The @code{l} stands for list.
4643 This switch causes a full listing of
4644 the file to be generated. In the case where a body is
4645 compiled, the corresponding spec is also listed, along
4646 with any subunits. Typical output from compiling a package
4647 body @file{p.adb} might look like:
4649 @smallexample @c ada
4653 1. package body p is
4655 3. procedure a is separate;
4666 2. pragma Elaborate_Body
4690 When you specify the @option{-gnatv} or @option{-gnatl} switches and
4691 standard output is redirected, a brief summary is written to
4692 @file{stderr} (standard error) giving the number of error messages and
4693 warning messages generated.
4695 @item -^gnatl^OUTPUT_FILE^=file
4696 @cindex @option{^-gnatl^OUTPUT_FILE^=fname} (@command{gcc})
4697 This has the same effect as @option{-gnatl} except that the output is
4698 written to a file instead of to standard output. If the given name
4699 @file{fname} does not start with a period, then it is the full name
4700 of the file to be written. If @file{fname} is an extension, it is
4701 appended to the name of the file being compiled. For example, if
4702 file @file{xyz.adb} is compiled with @option{^-gnatl^OUTPUT_FILE^=.lst},
4703 then the output is written to file ^xyz.adb.lst^xyz.adb_lst^.
4706 @cindex @option{-gnatU} (@command{gcc})
4707 This switch forces all error messages to be preceded by the unique
4708 string ``error:''. This means that error messages take a few more
4709 characters in space, but allows easy searching for and identification
4713 @cindex @option{-gnatb} (@command{gcc})
4715 The @code{b} stands for brief.
4717 This switch causes GNAT to generate the
4718 brief format error messages to @file{stderr} (the standard error
4719 file) as well as the verbose
4720 format message or full listing (which as usual is written to
4721 @file{stdout} (the standard output file).
4723 @item -gnatm=@var{n}
4724 @cindex @option{-gnatm} (@command{gcc})
4726 The @code{m} stands for maximum.
4728 @var{n} is a decimal integer in the
4729 range of 1 to 999999 and limits the number of error or warning
4730 messages to be generated. For example, using
4731 @option{-gnatm2} might yield
4734 e.adb:3:04: Incorrect spelling of keyword "function"
4735 e.adb:5:35: missing ".."
4736 fatal error: maximum number of errors detected
4737 compilation abandoned
4741 The default setting if
4742 no switch is given is 9999. If the number of warnings reaches this
4743 limit, then a message is output and further warnings are suppressed,
4744 but the compilation is continued. If the number of error messages
4745 reaches this limit, then a message is output and the compilation
4746 is abandoned. A value of zero means that no limit applies.
4749 Note that the equal sign is optional, so the switches
4750 @option{-gnatm2} and @option{-gnatm=2} are equivalent.
4753 @cindex @option{-gnatf} (@command{gcc})
4754 @cindex Error messages, suppressing
4756 The @code{f} stands for full.
4758 Normally, the compiler suppresses error messages that are likely to be
4759 redundant. This switch causes all error
4760 messages to be generated. In particular, in the case of
4761 references to undefined variables. If a given variable is referenced
4762 several times, the normal format of messages is
4764 e.adb:7:07: "V" is undefined (more references follow)
4768 where the parenthetical comment warns that there are additional
4769 references to the variable @code{V}. Compiling the same program with the
4770 @option{-gnatf} switch yields
4773 e.adb:7:07: "V" is undefined
4774 e.adb:8:07: "V" is undefined
4775 e.adb:8:12: "V" is undefined
4776 e.adb:8:16: "V" is undefined
4777 e.adb:9:07: "V" is undefined
4778 e.adb:9:12: "V" is undefined
4782 The @option{-gnatf} switch also generates additional information for
4783 some error messages. Some examples are:
4787 Full details on entities not available in high integrity mode
4789 Details on possibly non-portable unchecked conversion
4791 List possible interpretations for ambiguous calls
4793 Additional details on incorrect parameters
4797 @cindex @option{-gnatjnn} (@command{gcc})
4798 In normal operation mode (or if @option{-gnatj0} is used, then error messages
4799 with continuation lines are treated as though the continuation lines were
4800 separate messages (and so a warning with two continuation lines counts as
4801 three warnings, and is listed as three separate messages).
4803 If the @option{-gnatjnn} switch is used with a positive value for nn, then
4804 messages are output in a different manner. A message and all its continuation
4805 lines are treated as a unit, and count as only one warning or message in the
4806 statistics totals. Furthermore, the message is reformatted so that no line
4807 is longer than nn characters.
4810 @cindex @option{-gnatq} (@command{gcc})
4812 The @code{q} stands for quit (really ``don't quit'').
4814 In normal operation mode, the compiler first parses the program and
4815 determines if there are any syntax errors. If there are, appropriate
4816 error messages are generated and compilation is immediately terminated.
4818 GNAT to continue with semantic analysis even if syntax errors have been
4819 found. This may enable the detection of more errors in a single run. On
4820 the other hand, the semantic analyzer is more likely to encounter some
4821 internal fatal error when given a syntactically invalid tree.
4824 @cindex @option{-gnatQ} (@command{gcc})
4825 In normal operation mode, the @file{ALI} file is not generated if any
4826 illegalities are detected in the program. The use of @option{-gnatQ} forces
4827 generation of the @file{ALI} file. This file is marked as being in
4828 error, so it cannot be used for binding purposes, but it does contain
4829 reasonably complete cross-reference information, and thus may be useful
4830 for use by tools (e.g., semantic browsing tools or integrated development
4831 environments) that are driven from the @file{ALI} file. This switch
4832 implies @option{-gnatq}, since the semantic phase must be run to get a
4833 meaningful ALI file.
4835 In addition, if @option{-gnatt} is also specified, then the tree file is
4836 generated even if there are illegalities. It may be useful in this case
4837 to also specify @option{-gnatq} to ensure that full semantic processing
4838 occurs. The resulting tree file can be processed by ASIS, for the purpose
4839 of providing partial information about illegal units, but if the error
4840 causes the tree to be badly malformed, then ASIS may crash during the
4843 When @option{-gnatQ} is used and the generated @file{ALI} file is marked as
4844 being in error, @command{gnatmake} will attempt to recompile the source when it
4845 finds such an @file{ALI} file, including with switch @option{-gnatc}.
4847 Note that @option{-gnatQ} has no effect if @option{-gnats} is specified,
4848 since ALI files are never generated if @option{-gnats} is set.
4852 @node Warning Message Control
4853 @subsection Warning Message Control
4854 @cindex Warning messages
4856 In addition to error messages, which correspond to illegalities as defined
4857 in the Ada Reference Manual, the compiler detects two kinds of warning
4860 First, the compiler considers some constructs suspicious and generates a
4861 warning message to alert you to a possible error. Second, if the
4862 compiler detects a situation that is sure to raise an exception at
4863 run time, it generates a warning message. The following shows an example
4864 of warning messages:
4866 e.adb:4:24: warning: creation of object may raise Storage_Error
4867 e.adb:10:17: warning: static value out of range
4868 e.adb:10:17: warning: "Constraint_Error" will be raised at run time
4872 GNAT considers a large number of situations as appropriate
4873 for the generation of warning messages. As always, warnings are not
4874 definite indications of errors. For example, if you do an out-of-range
4875 assignment with the deliberate intention of raising a
4876 @code{Constraint_Error} exception, then the warning that may be
4877 issued does not indicate an error. Some of the situations for which GNAT
4878 issues warnings (at least some of the time) are given in the following
4879 list. This list is not complete, and new warnings are often added to
4880 subsequent versions of GNAT. The list is intended to give a general idea
4881 of the kinds of warnings that are generated.
4885 Possible infinitely recursive calls
4888 Out-of-range values being assigned
4891 Possible order of elaboration problems
4894 Assertions (pragma Assert) that are sure to fail
4900 Address clauses with possibly unaligned values, or where an attempt is
4901 made to overlay a smaller variable with a larger one.
4904 Fixed-point type declarations with a null range
4907 Direct_IO or Sequential_IO instantiated with a type that has access values
4910 Variables that are never assigned a value
4913 Variables that are referenced before being initialized
4916 Task entries with no corresponding @code{accept} statement
4919 Duplicate accepts for the same task entry in a @code{select}
4922 Objects that take too much storage
4925 Unchecked conversion between types of differing sizes
4928 Missing @code{return} statement along some execution path in a function
4931 Incorrect (unrecognized) pragmas
4934 Incorrect external names
4937 Allocation from empty storage pool
4940 Potentially blocking operation in protected type
4943 Suspicious parenthesization of expressions
4946 Mismatching bounds in an aggregate
4949 Attempt to return local value by reference
4952 Premature instantiation of a generic body
4955 Attempt to pack aliased components
4958 Out of bounds array subscripts
4961 Wrong length on string assignment
4964 Violations of style rules if style checking is enabled
4967 Unused @code{with} clauses
4970 @code{Bit_Order} usage that does not have any effect
4973 @code{Standard.Duration} used to resolve universal fixed expression
4976 Dereference of possibly null value
4979 Declaration that is likely to cause storage error
4982 Internal GNAT unit @code{with}'ed by application unit
4985 Values known to be out of range at compile time
4988 Unreferenced labels and variables
4991 Address overlays that could clobber memory
4994 Unexpected initialization when address clause present
4997 Bad alignment for address clause
5000 Useless type conversions
5003 Redundant assignment statements and other redundant constructs
5006 Useless exception handlers
5009 Accidental hiding of name by child unit
5012 Access before elaboration detected at compile time
5015 A range in a @code{for} loop that is known to be null or might be null
5020 The following section lists compiler switches that are available
5021 to control the handling of warning messages. It is also possible
5022 to exercise much finer control over what warnings are issued and
5023 suppressed using the GNAT pragma Warnings, @xref{Pragma Warnings,,,
5024 gnat_rm, GNAT Reference manual}.
5029 @emph{Activate all optional errors.}
5030 @cindex @option{-gnatwa} (@command{gcc})
5031 This switch activates most optional warning messages, see remaining list
5032 in this section for details on optional warning messages that can be
5033 individually controlled. The warnings that are not turned on by this
5035 @option{-gnatwd} (implicit dereferencing),
5036 @option{-gnatwh} (hiding),
5037 @option{-gnatwl} (elaboration warnings),
5038 @option{-gnatw.o} (warn on values set by out parameters ignored)
5039 and @option{-gnatwt} (tracking of deleted conditional code).
5040 All other optional warnings are turned on.
5043 @emph{Suppress all optional errors.}
5044 @cindex @option{-gnatwA} (@command{gcc})
5045 This switch suppresses all optional warning messages, see remaining list
5046 in this section for details on optional warning messages that can be
5047 individually controlled.
5050 @emph{Activate warnings on failing assertions.}
5051 @cindex @option{-gnatw.a} (@command{gcc})
5052 @cindex Assert failures
5053 This switch activates warnings for assertions where the compiler can tell at
5054 compile time that the assertion will fail. Note that this warning is given
5055 even if assertions are disabled. The default is that such warnings are
5059 @emph{Suppress warnings on failing assertions.}
5060 @cindex @option{-gnatw.A} (@command{gcc})
5061 @cindex Assert failures
5062 This switch suppresses warnings for assertions where the compiler can tell at
5063 compile time that the assertion will fail.
5066 @emph{Activate warnings on bad fixed values.}
5067 @cindex @option{-gnatwb} (@command{gcc})
5068 @cindex Bad fixed values
5069 @cindex Fixed-point Small value
5071 This switch activates warnings for static fixed-point expressions whose
5072 value is not an exact multiple of Small. Such values are implementation
5073 dependent, since an implementation is free to choose either of the multiples
5074 that surround the value. GNAT always chooses the closer one, but this is not
5075 required behavior, and it is better to specify a value that is an exact
5076 multiple, ensuring predictable execution. The default is that such warnings
5080 @emph{Suppress warnings on bad fixed values.}
5081 @cindex @option{-gnatwB} (@command{gcc})
5082 This switch suppresses warnings for static fixed-point expressions whose
5083 value is not an exact multiple of Small.
5086 @emph{Activate warnings on biased representation.}
5087 @cindex @option{-gnatw.b} (@command{gcc})
5088 @cindex Biased representation
5089 This switch activates warnings when a size clause, value size clause, component
5090 clause, or component size clause forces the use of biased representation for an
5091 integer type (e.g. representing a range of 10..11 in a single bit by using 0/1
5092 to represent 10/11). The default is that such warnings are generated.
5095 @emph{Suppress warnings on biased representation.}
5096 @cindex @option{-gnatwB} (@command{gcc})
5097 This switch suppresses warnings for representation clauses that force the use
5098 of biased representation.
5101 @emph{Activate warnings on conditionals.}
5102 @cindex @option{-gnatwc} (@command{gcc})
5103 @cindex Conditionals, constant
5104 This switch activates warnings for conditional expressions used in
5105 tests that are known to be True or False at compile time. The default
5106 is that such warnings are not generated.
5107 Note that this warning does
5108 not get issued for the use of boolean variables or constants whose
5109 values are known at compile time, since this is a standard technique
5110 for conditional compilation in Ada, and this would generate too many
5111 false positive warnings.
5113 This warning option also activates a special test for comparisons using
5114 the operators ``>='' and`` <=''.
5115 If the compiler can tell that only the equality condition is possible,
5116 then it will warn that the ``>'' or ``<'' part of the test
5117 is useless and that the operator could be replaced by ``=''.
5118 An example would be comparing a @code{Natural} variable <= 0.
5120 This warning option also generates warnings if
5121 one or both tests is optimized away in a membership test for integer
5122 values if the result can be determined at compile time. Range tests on
5123 enumeration types are not included, since it is common for such tests
5124 to include an end point.
5126 This warning can also be turned on using @option{-gnatwa}.
5129 @emph{Suppress warnings on conditionals.}
5130 @cindex @option{-gnatwC} (@command{gcc})
5131 This switch suppresses warnings for conditional expressions used in
5132 tests that are known to be True or False at compile time.
5135 @emph{Activate warnings on missing component clauses.}
5136 @cindex @option{-gnatw.c} (@command{gcc})
5137 @cindex Component clause, missing
5138 This switch activates warnings for record components where a record
5139 representation clause is present and has component clauses for the
5140 majority, but not all, of the components. A warning is given for each
5141 component for which no component clause is present.
5143 This warning can also be turned on using @option{-gnatwa}.
5146 @emph{Suppress warnings on missing component clauses.}
5147 @cindex @option{-gnatwC} (@command{gcc})
5148 This switch suppresses warnings for record components that are
5149 missing a component clause in the situation described above.
5152 @emph{Activate warnings on implicit dereferencing.}
5153 @cindex @option{-gnatwd} (@command{gcc})
5154 If this switch is set, then the use of a prefix of an access type
5155 in an indexed component, slice, or selected component without an
5156 explicit @code{.all} will generate a warning. With this warning
5157 enabled, access checks occur only at points where an explicit
5158 @code{.all} appears in the source code (assuming no warnings are
5159 generated as a result of this switch). The default is that such
5160 warnings are not generated.
5161 Note that @option{-gnatwa} does not affect the setting of
5162 this warning option.
5165 @emph{Suppress warnings on implicit dereferencing.}
5166 @cindex @option{-gnatwD} (@command{gcc})
5167 @cindex Implicit dereferencing
5168 @cindex Dereferencing, implicit
5169 This switch suppresses warnings for implicit dereferences in
5170 indexed components, slices, and selected components.
5173 @emph{Treat warnings as errors.}
5174 @cindex @option{-gnatwe} (@command{gcc})
5175 @cindex Warnings, treat as error
5176 This switch causes warning messages to be treated as errors.
5177 The warning string still appears, but the warning messages are counted
5178 as errors, and prevent the generation of an object file.
5181 @emph{Activate every optional warning}
5182 @cindex @option{-gnatw.e} (@command{gcc})
5183 @cindex Warnings, activate every optional warning
5184 This switch activates all optional warnings, including those which
5185 are not activated by @code{-gnatwa}.
5188 @emph{Activate warnings on unreferenced formals.}
5189 @cindex @option{-gnatwf} (@command{gcc})
5190 @cindex Formals, unreferenced
5191 This switch causes a warning to be generated if a formal parameter
5192 is not referenced in the body of the subprogram. This warning can
5193 also be turned on using @option{-gnatwa} or @option{-gnatwu}. The
5194 default is that these warnings are not generated.
5197 @emph{Suppress warnings on unreferenced formals.}
5198 @cindex @option{-gnatwF} (@command{gcc})
5199 This switch suppresses warnings for unreferenced formal
5200 parameters. Note that the
5201 combination @option{-gnatwu} followed by @option{-gnatwF} has the
5202 effect of warning on unreferenced entities other than subprogram
5206 @emph{Activate warnings on unrecognized pragmas.}
5207 @cindex @option{-gnatwg} (@command{gcc})
5208 @cindex Pragmas, unrecognized
5209 This switch causes a warning to be generated if an unrecognized
5210 pragma is encountered. Apart from issuing this warning, the
5211 pragma is ignored and has no effect. This warning can
5212 also be turned on using @option{-gnatwa}. The default
5213 is that such warnings are issued (satisfying the Ada Reference
5214 Manual requirement that such warnings appear).
5217 @emph{Suppress warnings on unrecognized pragmas.}
5218 @cindex @option{-gnatwG} (@command{gcc})
5219 This switch suppresses warnings for unrecognized pragmas.
5222 @emph{Activate warnings on hiding.}
5223 @cindex @option{-gnatwh} (@command{gcc})
5224 @cindex Hiding of Declarations
5225 This switch activates warnings on hiding declarations.
5226 A declaration is considered hiding
5227 if it is for a non-overloadable entity, and it declares an entity with the
5228 same name as some other entity that is directly or use-visible. The default
5229 is that such warnings are not generated.
5230 Note that @option{-gnatwa} does not affect the setting of this warning option.
5233 @emph{Suppress warnings on hiding.}
5234 @cindex @option{-gnatwH} (@command{gcc})
5235 This switch suppresses warnings on hiding declarations.
5238 @emph{Activate warnings on implementation units.}
5239 @cindex @option{-gnatwi} (@command{gcc})
5240 This switch activates warnings for a @code{with} of an internal GNAT
5241 implementation unit, defined as any unit from the @code{Ada},
5242 @code{Interfaces}, @code{GNAT},
5243 ^^@code{DEC},^ or @code{System}
5244 hierarchies that is not
5245 documented in either the Ada Reference Manual or the GNAT
5246 Programmer's Reference Manual. Such units are intended only
5247 for internal implementation purposes and should not be @code{with}'ed
5248 by user programs. The default is that such warnings are generated
5249 This warning can also be turned on using @option{-gnatwa}.
5252 @emph{Disable warnings on implementation units.}
5253 @cindex @option{-gnatwI} (@command{gcc})
5254 This switch disables warnings for a @code{with} of an internal GNAT
5255 implementation unit.
5258 @emph{Activate warnings on obsolescent features (Annex J).}
5259 @cindex @option{-gnatwj} (@command{gcc})
5260 @cindex Features, obsolescent
5261 @cindex Obsolescent features
5262 If this warning option is activated, then warnings are generated for
5263 calls to subprograms marked with @code{pragma Obsolescent} and
5264 for use of features in Annex J of the Ada Reference Manual. In the
5265 case of Annex J, not all features are flagged. In particular use
5266 of the renamed packages (like @code{Text_IO}) and use of package
5267 @code{ASCII} are not flagged, since these are very common and
5268 would generate many annoying positive warnings. The default is that
5269 such warnings are not generated. This warning is also turned on by
5270 the use of @option{-gnatwa}.
5272 In addition to the above cases, warnings are also generated for
5273 GNAT features that have been provided in past versions but which
5274 have been superseded (typically by features in the new Ada standard).
5275 For example, @code{pragma Ravenscar} will be flagged since its
5276 function is replaced by @code{pragma Profile(Ravenscar)}.
5278 Note that this warning option functions differently from the
5279 restriction @code{No_Obsolescent_Features} in two respects.
5280 First, the restriction applies only to annex J features.
5281 Second, the restriction does flag uses of package @code{ASCII}.
5284 @emph{Suppress warnings on obsolescent features (Annex J).}
5285 @cindex @option{-gnatwJ} (@command{gcc})
5286 This switch disables warnings on use of obsolescent features.
5289 @emph{Activate warnings on variables that could be constants.}
5290 @cindex @option{-gnatwk} (@command{gcc})
5291 This switch activates warnings for variables that are initialized but
5292 never modified, and then could be declared constants. The default is that
5293 such warnings are not given.
5294 This warning can also be turned on using @option{-gnatwa}.
5297 @emph{Suppress warnings on variables that could be constants.}
5298 @cindex @option{-gnatwK} (@command{gcc})
5299 This switch disables warnings on variables that could be declared constants.
5302 @emph{Activate warnings for elaboration pragmas.}
5303 @cindex @option{-gnatwl} (@command{gcc})
5304 @cindex Elaboration, warnings
5305 This switch activates warnings on missing
5306 @code{Elaborate_All} and @code{Elaborate} pragmas.
5307 See the section in this guide on elaboration checking for details on
5308 when such pragmas should be used. In dynamic elaboration mode, this switch
5309 generations warnings about the need to add elaboration pragmas. Note however,
5310 that if you blindly follow these warnings, and add @code{Elaborate_All}
5311 warnings wherever they are recommended, you basically end up with the
5312 equivalent of the static elaboration model, which may not be what you want for
5313 legacy code for which the static model does not work.
5315 For the static model, the messages generated are labeled "info:" (for
5316 information messages). They are not warnings to add elaboration pragmas,
5317 merely informational messages showing what implicit elaboration pragmas
5318 have been added, for use in analyzing elaboration circularity problems.
5320 Warnings are also generated if you
5321 are using the static mode of elaboration, and a @code{pragma Elaborate}
5322 is encountered. The default is that such warnings
5324 This warning is not automatically turned on by the use of @option{-gnatwa}.
5327 @emph{Suppress warnings for elaboration pragmas.}
5328 @cindex @option{-gnatwL} (@command{gcc})
5329 This switch suppresses warnings on missing Elaborate and Elaborate_All pragmas.
5330 See the section in this guide on elaboration checking for details on
5331 when such pragmas should be used.
5334 @emph{Activate warnings on modified but unreferenced variables.}
5335 @cindex @option{-gnatwm} (@command{gcc})
5336 This switch activates warnings for variables that are assigned (using
5337 an initialization value or with one or more assignment statements) but
5338 whose value is never read. The warning is suppressed for volatile
5339 variables and also for variables that are renamings of other variables
5340 or for which an address clause is given.
5341 This warning can also be turned on using @option{-gnatwa}.
5342 The default is that these warnings are not given.
5345 @emph{Disable warnings on modified but unreferenced variables.}
5346 @cindex @option{-gnatwM} (@command{gcc})
5347 This switch disables warnings for variables that are assigned or
5348 initialized, but never read.
5351 @emph{Set normal warnings mode.}
5352 @cindex @option{-gnatwn} (@command{gcc})
5353 This switch sets normal warning mode, in which enabled warnings are
5354 issued and treated as warnings rather than errors. This is the default
5355 mode. the switch @option{-gnatwn} can be used to cancel the effect of
5356 an explicit @option{-gnatws} or
5357 @option{-gnatwe}. It also cancels the effect of the
5358 implicit @option{-gnatwe} that is activated by the
5359 use of @option{-gnatg}.
5362 @emph{Activate warnings on address clause overlays.}
5363 @cindex @option{-gnatwo} (@command{gcc})
5364 @cindex Address Clauses, warnings
5365 This switch activates warnings for possibly unintended initialization
5366 effects of defining address clauses that cause one variable to overlap
5367 another. The default is that such warnings are generated.
5368 This warning can also be turned on using @option{-gnatwa}.
5371 @emph{Suppress warnings on address clause overlays.}
5372 @cindex @option{-gnatwO} (@command{gcc})
5373 This switch suppresses warnings on possibly unintended initialization
5374 effects of defining address clauses that cause one variable to overlap
5378 @emph{Activate warnings on modified but unreferenced out parameters.}
5379 @cindex @option{-gnatw.o} (@command{gcc})
5380 This switch activates warnings for variables that are modified by using
5381 them as actuals for a call to a procedure with an out mode formal, where
5382 the resulting assigned value is never read. It is applicable in the case
5383 where there is more than one out mode formal. If there is only one out
5384 mode formal, the warning is issued by default (controlled by -gnatwu).
5385 The warning is suppressed for volatile
5386 variables and also for variables that are renamings of other variables
5387 or for which an address clause is given.
5388 The default is that these warnings are not given. Note that this warning
5389 is not included in -gnatwa, it must be activated explicitly.
5392 @emph{Disable warnings on modified but unreferenced out parameters.}
5393 @cindex @option{-gnatw.O} (@command{gcc})
5394 This switch suppresses warnings for variables that are modified by using
5395 them as actuals for a call to a procedure with an out mode formal, where
5396 the resulting assigned value is never read.
5399 @emph{Activate warnings on ineffective pragma Inlines.}
5400 @cindex @option{-gnatwp} (@command{gcc})
5401 @cindex Inlining, warnings
5402 This switch activates warnings for failure of front end inlining
5403 (activated by @option{-gnatN}) to inline a particular call. There are
5404 many reasons for not being able to inline a call, including most
5405 commonly that the call is too complex to inline. The default is
5406 that such warnings are not given.
5407 This warning can also be turned on using @option{-gnatwa}.
5408 Warnings on ineffective inlining by the gcc back-end can be activated
5409 separately, using the gcc switch -Winline.
5412 @emph{Suppress warnings on ineffective pragma Inlines.}
5413 @cindex @option{-gnatwP} (@command{gcc})
5414 This switch suppresses warnings on ineffective pragma Inlines. If the
5415 inlining mechanism cannot inline a call, it will simply ignore the
5419 @emph{Activate warnings on parameter ordering.}
5420 @cindex @option{-gnatw.p} (@command{gcc})
5421 @cindex Parameter order, warnings
5422 This switch activates warnings for cases of suspicious parameter
5423 ordering when the list of arguments are all simple identifiers that
5424 match the names of the formals, but are in a different order. The
5425 warning is suppressed if any use of named parameter notation is used,
5426 so this is the appropriate way to suppress a false positive (and
5427 serves to emphasize that the "misordering" is deliberate). The
5429 that such warnings are not given.
5430 This warning can also be turned on using @option{-gnatwa}.
5433 @emph{Suppress warnings on parameter ordering.}
5434 @cindex @option{-gnatw.P} (@command{gcc})
5435 This switch suppresses warnings on cases of suspicious parameter
5439 @emph{Activate warnings on questionable missing parentheses.}
5440 @cindex @option{-gnatwq} (@command{gcc})
5441 @cindex Parentheses, warnings
5442 This switch activates warnings for cases where parentheses are not used and
5443 the result is potential ambiguity from a readers point of view. For example
5444 (not a > b) when a and b are modular means ((not a) > b) and very likely the
5445 programmer intended (not (a > b)). Similarly (-x mod 5) means (-(x mod 5)) and
5446 quite likely ((-x) mod 5) was intended. In such situations it seems best to
5447 follow the rule of always parenthesizing to make the association clear, and
5448 this warning switch warns if such parentheses are not present. The default
5449 is that these warnings are given.
5450 This warning can also be turned on using @option{-gnatwa}.
5453 @emph{Suppress warnings on questionable missing parentheses.}
5454 @cindex @option{-gnatwQ} (@command{gcc})
5455 This switch suppresses warnings for cases where the association is not
5456 clear and the use of parentheses is preferred.
5459 @emph{Activate warnings on redundant constructs.}
5460 @cindex @option{-gnatwr} (@command{gcc})
5461 This switch activates warnings for redundant constructs. The following
5462 is the current list of constructs regarded as redundant:
5466 Assignment of an item to itself.
5468 Type conversion that converts an expression to its own type.
5470 Use of the attribute @code{Base} where @code{typ'Base} is the same
5473 Use of pragma @code{Pack} when all components are placed by a record
5474 representation clause.
5476 Exception handler containing only a reraise statement (raise with no
5477 operand) which has no effect.
5479 Use of the operator abs on an operand that is known at compile time
5482 Comparison of boolean expressions to an explicit True value.
5485 This warning can also be turned on using @option{-gnatwa}.
5486 The default is that warnings for redundant constructs are not given.
5489 @emph{Suppress warnings on redundant constructs.}
5490 @cindex @option{-gnatwR} (@command{gcc})
5491 This switch suppresses warnings for redundant constructs.
5494 @emph{Activate warnings for object renaming function.}
5495 @cindex @option{-gnatw.r} (@command{gcc})
5496 This switch activates warnings for an object renaming that renames a
5497 function call, which is equivalent to a constant declaration (as
5498 opposed to renaming the function itself). The default is that these
5499 warnings are given. This warning can also be turned on using
5503 @emph{Suppress warnings for object renaming function.}
5504 @cindex @option{-gnatwT} (@command{gcc})
5505 This switch suppresses warnings for object renaming function.
5508 @emph{Suppress all warnings.}
5509 @cindex @option{-gnatws} (@command{gcc})
5510 This switch completely suppresses the
5511 output of all warning messages from the GNAT front end.
5512 Note that it does not suppress warnings from the @command{gcc} back end.
5513 To suppress these back end warnings as well, use the switch @option{-w}
5514 in addition to @option{-gnatws}.
5517 @emph{Activate warnings for tracking of deleted conditional code.}
5518 @cindex @option{-gnatwt} (@command{gcc})
5519 @cindex Deactivated code, warnings
5520 @cindex Deleted code, warnings
5521 This switch activates warnings for tracking of code in conditionals (IF and
5522 CASE statements) that is detected to be dead code which cannot be executed, and
5523 which is removed by the front end. This warning is off by default, and is not
5524 turned on by @option{-gnatwa}, it has to be turned on explicitly. This may be
5525 useful for detecting deactivated code in certified applications.
5528 @emph{Suppress warnings for tracking of deleted conditional code.}
5529 @cindex @option{-gnatwT} (@command{gcc})
5530 This switch suppresses warnings for tracking of deleted conditional code.
5533 @emph{Activate warnings on unused entities.}
5534 @cindex @option{-gnatwu} (@command{gcc})
5535 This switch activates warnings to be generated for entities that
5536 are declared but not referenced, and for units that are @code{with}'ed
5538 referenced. In the case of packages, a warning is also generated if
5539 no entities in the package are referenced. This means that if the package
5540 is referenced but the only references are in @code{use}
5541 clauses or @code{renames}
5542 declarations, a warning is still generated. A warning is also generated
5543 for a generic package that is @code{with}'ed but never instantiated.
5544 In the case where a package or subprogram body is compiled, and there
5545 is a @code{with} on the corresponding spec
5546 that is only referenced in the body,
5547 a warning is also generated, noting that the
5548 @code{with} can be moved to the body. The default is that
5549 such warnings are not generated.
5550 This switch also activates warnings on unreferenced formals
5551 (it includes the effect of @option{-gnatwf}).
5552 This warning can also be turned on using @option{-gnatwa}.
5555 @emph{Suppress warnings on unused entities.}
5556 @cindex @option{-gnatwU} (@command{gcc})
5557 This switch suppresses warnings for unused entities and packages.
5558 It also turns off warnings on unreferenced formals (and thus includes
5559 the effect of @option{-gnatwF}).
5562 @emph{Activate warnings on unassigned variables.}
5563 @cindex @option{-gnatwv} (@command{gcc})
5564 @cindex Unassigned variable warnings
5565 This switch activates warnings for access to variables which
5566 may not be properly initialized. The default is that
5567 such warnings are generated.
5568 This warning can also be turned on using @option{-gnatwa}.
5571 @emph{Suppress warnings on unassigned variables.}
5572 @cindex @option{-gnatwV} (@command{gcc})
5573 This switch suppresses warnings for access to variables which
5574 may not be properly initialized.
5575 For variables of a composite type, the warning can also be suppressed in
5576 Ada 2005 by using a default initialization with a box. For example, if
5577 Table is an array of records whose components are only partially uninitialized,
5578 then the following code:
5580 @smallexample @c ada
5581 Tab : Table := (others => <>);
5584 will suppress warnings on subsequent statements that access components
5588 @emph{Activate warnings on wrong low bound assumption.}
5589 @cindex @option{-gnatww} (@command{gcc})
5590 @cindex String indexing warnings
5591 This switch activates warnings for indexing an unconstrained string parameter
5592 with a literal or S'Length. This is a case where the code is assuming that the
5593 low bound is one, which is in general not true (for example when a slice is
5594 passed). The default is that such warnings are generated.
5595 This warning can also be turned on using @option{-gnatwa}.
5598 @emph{Suppress warnings on wrong low bound assumption.}
5599 @cindex @option{-gnatwW} (@command{gcc})
5600 This switch suppresses warnings for indexing an unconstrained string parameter
5601 with a literal or S'Length. Note that this warning can also be suppressed
5602 in a particular case by adding an
5603 assertion that the lower bound is 1,
5604 as shown in the following example.
5606 @smallexample @c ada
5607 procedure K (S : String) is
5608 pragma Assert (S'First = 1);
5613 @emph{Activate warnings on unnecessary Warnings Off pragmas}
5614 @cindex @option{-gnatw.w} (@command{gcc})
5615 @cindex Warnings Off control
5616 This switch activates warnings for use of @code{pragma Warnings (Off, entity}
5617 where either the pragma is entirely useless (because it suppresses no
5618 warnings), or it could be replaced by @code{pragma Unreferenced} or
5619 @code{pragma Unmodified}.The default is that these warnings are not given.
5620 Note that this warning is not included in -gnatwa, it must be
5621 activated explicitly.
5624 @emph{Suppress warnings on unnecessary Warnings Off pragmas}
5625 @cindex @option{-gnatw.W} (@command{gcc})
5626 This switch suppresses warnings for use of @code{pragma Warnings (Off, entity}.
5629 @emph{Activate warnings on Export/Import pragmas.}
5630 @cindex @option{-gnatwx} (@command{gcc})
5631 @cindex Export/Import pragma warnings
5632 This switch activates warnings on Export/Import pragmas when
5633 the compiler detects a possible conflict between the Ada and
5634 foreign language calling sequences. For example, the use of
5635 default parameters in a convention C procedure is dubious
5636 because the C compiler cannot supply the proper default, so
5637 a warning is issued. The default is that such warnings are
5639 This warning can also be turned on using @option{-gnatwa}.
5642 @emph{Suppress warnings on Export/Import pragmas.}
5643 @cindex @option{-gnatwX} (@command{gcc})
5644 This switch suppresses warnings on Export/Import pragmas.
5645 The sense of this is that you are telling the compiler that
5646 you know what you are doing in writing the pragma, and it
5647 should not complain at you.
5650 @emph{Activate warnings for No_Exception_Propagation mode.}
5651 @cindex @option{-gnatwm} (@command{gcc})
5652 This switch activates warnings for exception usage when pragma Restrictions
5653 (No_Exception_Propagation) is in effect. Warnings are given for implicit or
5654 explicit exception raises which are not covered by a local handler, and for
5655 exception handlers which do not cover a local raise. The default is that these
5656 warnings are not given.
5659 @emph{Disable warnings for No_Exception_Propagation mode.}
5660 This switch disables warnings for exception usage when pragma Restrictions
5661 (No_Exception_Propagation) is in effect.
5664 @emph{Activate warnings for Ada 2005 compatibility issues.}
5665 @cindex @option{-gnatwy} (@command{gcc})
5666 @cindex Ada 2005 compatibility issues warnings
5667 For the most part Ada 2005 is upwards compatible with Ada 95,
5668 but there are some exceptions (for example the fact that
5669 @code{interface} is now a reserved word in Ada 2005). This
5670 switch activates several warnings to help in identifying
5671 and correcting such incompatibilities. The default is that
5672 these warnings are generated. Note that at one point Ada 2005
5673 was called Ada 0Y, hence the choice of character.
5674 This warning can also be turned on using @option{-gnatwa}.
5677 @emph{Disable warnings for Ada 2005 compatibility issues.}
5678 @cindex @option{-gnatwY} (@command{gcc})
5679 @cindex Ada 2005 compatibility issues warnings
5680 This switch suppresses several warnings intended to help in identifying
5681 incompatibilities between Ada 95 and Ada 2005.
5684 @emph{Activate warnings on unchecked conversions.}
5685 @cindex @option{-gnatwz} (@command{gcc})
5686 @cindex Unchecked_Conversion warnings
5687 This switch activates warnings for unchecked conversions
5688 where the types are known at compile time to have different
5690 is that such warnings are generated. Warnings are also
5691 generated for subprogram pointers with different conventions,
5692 and, on VMS only, for data pointers with different conventions.
5693 This warning can also be turned on using @option{-gnatwa}.
5696 @emph{Suppress warnings on unchecked conversions.}
5697 @cindex @option{-gnatwZ} (@command{gcc})
5698 This switch suppresses warnings for unchecked conversions
5699 where the types are known at compile time to have different
5700 sizes or conventions.
5702 @item ^-Wunused^WARNINGS=UNUSED^
5703 @cindex @option{-Wunused}
5704 The warnings controlled by the @option{-gnatw} switch are generated by
5705 the front end of the compiler. The @option{GCC} back end can provide
5706 additional warnings and they are controlled by the @option{-W} switch.
5707 For example, @option{^-Wunused^WARNINGS=UNUSED^} activates back end
5708 warnings for entities that are declared but not referenced.
5710 @item ^-Wuninitialized^WARNINGS=UNINITIALIZED^
5711 @cindex @option{-Wuninitialized}
5712 Similarly, @option{^-Wuninitialized^WARNINGS=UNINITIALIZED^} activates
5713 the back end warning for uninitialized variables. This switch must be
5714 used in conjunction with an optimization level greater than zero.
5716 @item ^-Wall^/ALL_BACK_END_WARNINGS^
5717 @cindex @option{-Wall}
5718 This switch enables all the above warnings from the @option{GCC} back end.
5719 The code generator detects a number of warning situations that are missed
5720 by the @option{GNAT} front end, and this switch can be used to activate them.
5721 The use of this switch also sets the default front end warning mode to
5722 @option{-gnatwa}, that is, most front end warnings activated as well.
5724 @item ^-w^/NO_BACK_END_WARNINGS^
5726 Conversely, this switch suppresses warnings from the @option{GCC} back end.
5727 The use of this switch also sets the default front end warning mode to
5728 @option{-gnatws}, that is, front end warnings suppressed as well.
5734 A string of warning parameters can be used in the same parameter. For example:
5741 will turn on all optional warnings except for elaboration pragma warnings,
5742 and also specify that warnings should be treated as errors.
5744 When no switch @option{^-gnatw^/WARNINGS^} is used, this is equivalent to:
5769 @node Debugging and Assertion Control
5770 @subsection Debugging and Assertion Control
5774 @cindex @option{-gnata} (@command{gcc})
5780 The pragmas @code{Assert} and @code{Debug} normally have no effect and
5781 are ignored. This switch, where @samp{a} stands for assert, causes
5782 @code{Assert} and @code{Debug} pragmas to be activated.
5784 The pragmas have the form:
5788 @b{pragma} Assert (@var{Boolean-expression} @r{[},
5789 @var{static-string-expression}@r{]})
5790 @b{pragma} Debug (@var{procedure call})
5795 The @code{Assert} pragma causes @var{Boolean-expression} to be tested.
5796 If the result is @code{True}, the pragma has no effect (other than
5797 possible side effects from evaluating the expression). If the result is
5798 @code{False}, the exception @code{Assert_Failure} declared in the package
5799 @code{System.Assertions} is
5800 raised (passing @var{static-string-expression}, if present, as the
5801 message associated with the exception). If no string expression is
5802 given the default is a string giving the file name and line number
5805 The @code{Debug} pragma causes @var{procedure} to be called. Note that
5806 @code{pragma Debug} may appear within a declaration sequence, allowing
5807 debugging procedures to be called between declarations.
5810 @item /DEBUG@r{[}=debug-level@r{]}
5812 Specifies how much debugging information is to be included in
5813 the resulting object file where 'debug-level' is one of the following:
5816 Include both debugger symbol records and traceback
5818 This is the default setting.
5820 Include both debugger symbol records and traceback in
5823 Excludes both debugger symbol records and traceback
5824 the object file. Same as /NODEBUG.
5826 Includes only debugger symbol records in the object
5827 file. Note that this doesn't include traceback information.
5832 @node Validity Checking
5833 @subsection Validity Checking
5834 @findex Validity Checking
5837 The Ada Reference Manual has specific requirements for checking
5838 for invalid values. In particular, RM 13.9.1 requires that the
5839 evaluation of invalid values (for example from unchecked conversions),
5840 not result in erroneous execution. In GNAT, the result of such an
5841 evaluation in normal default mode is to either use the value
5842 unmodified, or to raise Constraint_Error in those cases where use
5843 of the unmodified value would cause erroneous execution. The cases
5844 where unmodified values might lead to erroneous execution are case
5845 statements (where a wild jump might result from an invalid value),
5846 and subscripts on the left hand side (where memory corruption could
5847 occur as a result of an invalid value).
5849 The @option{-gnatB} switch tells the compiler to assume that all
5850 values are valid (that is, within their declared subtype range)
5851 except in the context of a use of the Valid attribute. This means
5852 the compiler can generate more efficient code, since the range
5853 of values is better known at compile time.
5855 The @option{-gnatV^@var{x}^^} switch allows more control over the validity
5858 The @code{x} argument is a string of letters that
5859 indicate validity checks that are performed or not performed in addition
5860 to the default checks described above.
5863 The options allowed for this qualifier
5864 indicate validity checks that are performed or not performed in addition
5865 to the default checks described above.
5871 @emph{All validity checks.}
5872 @cindex @option{-gnatVa} (@command{gcc})
5873 All validity checks are turned on.
5875 That is, @option{-gnatVa} is
5876 equivalent to @option{gnatVcdfimorst}.
5880 @emph{Validity checks for copies.}
5881 @cindex @option{-gnatVc} (@command{gcc})
5882 The right hand side of assignments, and the initializing values of
5883 object declarations are validity checked.
5886 @emph{Default (RM) validity checks.}
5887 @cindex @option{-gnatVd} (@command{gcc})
5888 Some validity checks are done by default following normal Ada semantics
5890 A check is done in case statements that the expression is within the range
5891 of the subtype. If it is not, Constraint_Error is raised.
5892 For assignments to array components, a check is done that the expression used
5893 as index is within the range. If it is not, Constraint_Error is raised.
5894 Both these validity checks may be turned off using switch @option{-gnatVD}.
5895 They are turned on by default. If @option{-gnatVD} is specified, a subsequent
5896 switch @option{-gnatVd} will leave the checks turned on.
5897 Switch @option{-gnatVD} should be used only if you are sure that all such
5898 expressions have valid values. If you use this switch and invalid values
5899 are present, then the program is erroneous, and wild jumps or memory
5900 overwriting may occur.
5903 @emph{Validity checks for elementary components.}
5904 @cindex @option{-gnatVe} (@command{gcc})
5905 In the absence of this switch, assignments to record or array components are
5906 not validity checked, even if validity checks for assignments generally
5907 (@option{-gnatVc}) are turned on. In Ada, assignment of composite values do not
5908 require valid data, but assignment of individual components does. So for
5909 example, there is a difference between copying the elements of an array with a
5910 slice assignment, compared to assigning element by element in a loop. This
5911 switch allows you to turn off validity checking for components, even when they
5912 are assigned component by component.
5915 @emph{Validity checks for floating-point values.}
5916 @cindex @option{-gnatVf} (@command{gcc})
5917 In the absence of this switch, validity checking occurs only for discrete
5918 values. If @option{-gnatVf} is specified, then validity checking also applies
5919 for floating-point values, and NaNs and infinities are considered invalid,
5920 as well as out of range values for constrained types. Note that this means
5921 that standard IEEE infinity mode is not allowed. The exact contexts
5922 in which floating-point values are checked depends on the setting of other
5923 options. For example,
5924 @option{^-gnatVif^VALIDITY_CHECKING=(IN_PARAMS,FLOATS)^} or
5925 @option{^-gnatVfi^VALIDITY_CHECKING=(FLOATS,IN_PARAMS)^}
5926 (the order does not matter) specifies that floating-point parameters of mode
5927 @code{in} should be validity checked.
5930 @emph{Validity checks for @code{in} mode parameters}
5931 @cindex @option{-gnatVi} (@command{gcc})
5932 Arguments for parameters of mode @code{in} are validity checked in function
5933 and procedure calls at the point of call.
5936 @emph{Validity checks for @code{in out} mode parameters.}
5937 @cindex @option{-gnatVm} (@command{gcc})
5938 Arguments for parameters of mode @code{in out} are validity checked in
5939 procedure calls at the point of call. The @code{'m'} here stands for
5940 modify, since this concerns parameters that can be modified by the call.
5941 Note that there is no specific option to test @code{out} parameters,
5942 but any reference within the subprogram will be tested in the usual
5943 manner, and if an invalid value is copied back, any reference to it
5944 will be subject to validity checking.
5947 @emph{No validity checks.}
5948 @cindex @option{-gnatVn} (@command{gcc})
5949 This switch turns off all validity checking, including the default checking
5950 for case statements and left hand side subscripts. Note that the use of
5951 the switch @option{-gnatp} suppresses all run-time checks, including
5952 validity checks, and thus implies @option{-gnatVn}. When this switch
5953 is used, it cancels any other @option{-gnatV} previously issued.
5956 @emph{Validity checks for operator and attribute operands.}
5957 @cindex @option{-gnatVo} (@command{gcc})
5958 Arguments for predefined operators and attributes are validity checked.
5959 This includes all operators in package @code{Standard},
5960 the shift operators defined as intrinsic in package @code{Interfaces}
5961 and operands for attributes such as @code{Pos}. Checks are also made
5962 on individual component values for composite comparisons, and on the
5963 expressions in type conversions and qualified expressions. Checks are
5964 also made on explicit ranges using @samp{..} (e.g.@: slices, loops etc).
5967 @emph{Validity checks for parameters.}
5968 @cindex @option{-gnatVp} (@command{gcc})
5969 This controls the treatment of parameters within a subprogram (as opposed
5970 to @option{-gnatVi} and @option{-gnatVm} which control validity testing
5971 of parameters on a call. If either of these call options is used, then
5972 normally an assumption is made within a subprogram that the input arguments
5973 have been validity checking at the point of call, and do not need checking
5974 again within a subprogram). If @option{-gnatVp} is set, then this assumption
5975 is not made, and parameters are not assumed to be valid, so their validity
5976 will be checked (or rechecked) within the subprogram.
5979 @emph{Validity checks for function returns.}
5980 @cindex @option{-gnatVr} (@command{gcc})
5981 The expression in @code{return} statements in functions is validity
5985 @emph{Validity checks for subscripts.}
5986 @cindex @option{-gnatVs} (@command{gcc})
5987 All subscripts expressions are checked for validity, whether they appear
5988 on the right side or left side (in default mode only left side subscripts
5989 are validity checked).
5992 @emph{Validity checks for tests.}
5993 @cindex @option{-gnatVt} (@command{gcc})
5994 Expressions used as conditions in @code{if}, @code{while} or @code{exit}
5995 statements are checked, as well as guard expressions in entry calls.
6000 The @option{-gnatV} switch may be followed by
6001 ^a string of letters^a list of options^
6002 to turn on a series of validity checking options.
6004 @option{^-gnatVcr^/VALIDITY_CHECKING=(COPIES, RETURNS)^}
6005 specifies that in addition to the default validity checking, copies and
6006 function return expressions are to be validity checked.
6007 In order to make it easier
6008 to specify the desired combination of effects,
6010 the upper case letters @code{CDFIMORST} may
6011 be used to turn off the corresponding lower case option.
6014 the prefix @code{NO} on an option turns off the corresponding validity
6017 @item @code{NOCOPIES}
6018 @item @code{NODEFAULT}
6019 @item @code{NOFLOATS}
6020 @item @code{NOIN_PARAMS}
6021 @item @code{NOMOD_PARAMS}
6022 @item @code{NOOPERANDS}
6023 @item @code{NORETURNS}
6024 @item @code{NOSUBSCRIPTS}
6025 @item @code{NOTESTS}
6029 @option{^-gnatVaM^/VALIDITY_CHECKING=(ALL, NOMOD_PARAMS)^}
6030 turns on all validity checking options except for
6031 checking of @code{@b{in out}} procedure arguments.
6033 The specification of additional validity checking generates extra code (and
6034 in the case of @option{-gnatVa} the code expansion can be substantial).
6035 However, these additional checks can be very useful in detecting
6036 uninitialized variables, incorrect use of unchecked conversion, and other
6037 errors leading to invalid values. The use of pragma @code{Initialize_Scalars}
6038 is useful in conjunction with the extra validity checking, since this
6039 ensures that wherever possible uninitialized variables have invalid values.
6041 See also the pragma @code{Validity_Checks} which allows modification of
6042 the validity checking mode at the program source level, and also allows for
6043 temporary disabling of validity checks.
6045 @node Style Checking
6046 @subsection Style Checking
6047 @findex Style checking
6050 The @option{-gnaty^x^(option,option,@dots{})^} switch
6051 @cindex @option{-gnaty} (@command{gcc})
6052 causes the compiler to
6053 enforce specified style rules. A limited set of style rules has been used
6054 in writing the GNAT sources themselves. This switch allows user programs
6055 to activate all or some of these checks. If the source program fails a
6056 specified style check, an appropriate warning message is given, preceded by
6057 the character sequence ``(style)''.
6059 @code{(option,option,@dots{})} is a sequence of keywords
6062 The string @var{x} is a sequence of letters or digits
6064 indicating the particular style
6065 checks to be performed. The following checks are defined:
6070 @emph{Specify indentation level.}
6071 If a digit from 1-9 appears
6072 ^in the string after @option{-gnaty}^as an option for /STYLE_CHECKS^
6073 then proper indentation is checked, with the digit indicating the
6074 indentation level required. A value of zero turns off this style check.
6075 The general style of required indentation is as specified by
6076 the examples in the Ada Reference Manual. Full line comments must be
6077 aligned with the @code{--} starting on a column that is a multiple of
6078 the alignment level, or they may be aligned the same way as the following
6079 non-blank line (this is useful when full line comments appear in the middle
6083 @emph{Check attribute casing.}
6084 Attribute names, including the case of keywords such as @code{digits}
6085 used as attributes names, must be written in mixed case, that is, the
6086 initial letter and any letter following an underscore must be uppercase.
6087 All other letters must be lowercase.
6089 @item ^A^ARRAY_INDEXES^
6090 @emph{Use of array index numbers in array attributes.}
6091 When using the array attributes First, Last, Range,
6092 or Length, the index number must be omitted for one-dimensional arrays
6093 and is required for multi-dimensional arrays.
6096 @emph{Blanks not allowed at statement end.}
6097 Trailing blanks are not allowed at the end of statements. The purpose of this
6098 rule, together with h (no horizontal tabs), is to enforce a canonical format
6099 for the use of blanks to separate source tokens.
6102 @emph{Check comments.}
6103 Comments must meet the following set of rules:
6108 The ``@code{--}'' that starts the column must either start in column one,
6109 or else at least one blank must precede this sequence.
6112 Comments that follow other tokens on a line must have at least one blank
6113 following the ``@code{--}'' at the start of the comment.
6116 Full line comments must have two blanks following the ``@code{--}'' that
6117 starts the comment, with the following exceptions.
6120 A line consisting only of the ``@code{--}'' characters, possibly preceded
6121 by blanks is permitted.
6124 A comment starting with ``@code{--x}'' where @code{x} is a special character
6126 This allows proper processing of the output generated by specialized tools
6127 including @command{gnatprep} (where ``@code{--!}'' is used) and the SPARK
6129 language (where ``@code{--#}'' is used). For the purposes of this rule, a
6130 special character is defined as being in one of the ASCII ranges
6131 @code{16#21#@dots{}16#2F#} or @code{16#3A#@dots{}16#3F#}.
6132 Note that this usage is not permitted
6133 in GNAT implementation units (i.e., when @option{-gnatg} is used).
6136 A line consisting entirely of minus signs, possibly preceded by blanks, is
6137 permitted. This allows the construction of box comments where lines of minus
6138 signs are used to form the top and bottom of the box.
6141 A comment that starts and ends with ``@code{--}'' is permitted as long as at
6142 least one blank follows the initial ``@code{--}''. Together with the preceding
6143 rule, this allows the construction of box comments, as shown in the following
6146 ---------------------------
6147 -- This is a box comment --
6148 -- with two text lines. --
6149 ---------------------------
6153 @item ^d^DOS_LINE_ENDINGS^
6154 @emph{Check no DOS line terminators present.}
6155 All lines must be terminated by a single ASCII.LF
6156 character (in particular the DOS line terminator sequence CR/LF is not
6160 @emph{Check end/exit labels.}
6161 Optional labels on @code{end} statements ending subprograms and on
6162 @code{exit} statements exiting named loops, are required to be present.
6165 @emph{No form feeds or vertical tabs.}
6166 Neither form feeds nor vertical tab characters are permitted
6170 @emph{GNAT style mode}
6171 The set of style check switches is set to match that used by the GNAT sources.
6172 This may be useful when developing code that is eventually intended to be
6173 incorporated into GNAT. For further details, see GNAT sources.
6176 @emph{No horizontal tabs.}
6177 Horizontal tab characters are not permitted in the source text.
6178 Together with the b (no blanks at end of line) check, this
6179 enforces a canonical form for the use of blanks to separate
6183 @emph{Check if-then layout.}
6184 The keyword @code{then} must appear either on the same
6185 line as corresponding @code{if}, or on a line on its own, lined
6186 up under the @code{if} with at least one non-blank line in between
6187 containing all or part of the condition to be tested.
6190 @emph{check mode IN keywords}
6191 Mode @code{in} (the default mode) is not
6192 allowed to be given explicitly. @code{in out} is fine,
6193 but not @code{in} on its own.
6196 @emph{Check keyword casing.}
6197 All keywords must be in lower case (with the exception of keywords
6198 such as @code{digits} used as attribute names to which this check
6202 @emph{Check layout.}
6203 Layout of statement and declaration constructs must follow the
6204 recommendations in the Ada Reference Manual, as indicated by the
6205 form of the syntax rules. For example an @code{else} keyword must
6206 be lined up with the corresponding @code{if} keyword.
6208 There are two respects in which the style rule enforced by this check
6209 option are more liberal than those in the Ada Reference Manual. First
6210 in the case of record declarations, it is permissible to put the
6211 @code{record} keyword on the same line as the @code{type} keyword, and
6212 then the @code{end} in @code{end record} must line up under @code{type}.
6213 This is also permitted when the type declaration is split on two lines.
6214 For example, any of the following three layouts is acceptable:
6216 @smallexample @c ada
6239 Second, in the case of a block statement, a permitted alternative
6240 is to put the block label on the same line as the @code{declare} or
6241 @code{begin} keyword, and then line the @code{end} keyword up under
6242 the block label. For example both the following are permitted:
6244 @smallexample @c ada
6262 The same alternative format is allowed for loops. For example, both of
6263 the following are permitted:
6265 @smallexample @c ada
6267 Clear : while J < 10 loop
6278 @item ^Lnnn^MAX_NESTING=nnn^
6279 @emph{Set maximum nesting level}
6280 The maximum level of nesting of constructs (including subprograms, loops,
6281 blocks, packages, and conditionals) may not exceed the given value
6282 @option{nnn}. A value of zero disconnects this style check.
6284 @item ^m^LINE_LENGTH^
6285 @emph{Check maximum line length.}
6286 The length of source lines must not exceed 79 characters, including
6287 any trailing blanks. The value of 79 allows convenient display on an
6288 80 character wide device or window, allowing for possible special
6289 treatment of 80 character lines. Note that this count is of
6290 characters in the source text. This means that a tab character counts
6291 as one character in this count but a wide character sequence counts as
6292 a single character (however many bytes are needed in the encoding).
6294 @item ^Mnnn^MAX_LENGTH=nnn^
6295 @emph{Set maximum line length.}
6296 The length of lines must not exceed the
6297 given value @option{nnn}. The maximum value that can be specified is 32767.
6299 @item ^n^STANDARD_CASING^
6300 @emph{Check casing of entities in Standard.}
6301 Any identifier from Standard must be cased
6302 to match the presentation in the Ada Reference Manual (for example,
6303 @code{Integer} and @code{ASCII.NUL}).
6306 @emph{Turn off all style checks}
6307 All style check options are turned off.
6309 @item ^o^ORDERED_SUBPROGRAMS^
6310 @emph{Check order of subprogram bodies.}
6311 All subprogram bodies in a given scope
6312 (e.g.@: a package body) must be in alphabetical order. The ordering
6313 rule uses normal Ada rules for comparing strings, ignoring casing
6314 of letters, except that if there is a trailing numeric suffix, then
6315 the value of this suffix is used in the ordering (e.g.@: Junk2 comes
6318 @item ^O^OVERRIDING_INDICATORS^
6319 @emph{Check that overriding subprograms are explicitly marked as such.}
6320 The declaration of a primitive operation of a type extension that overrides
6321 an inherited operation must carry an overriding indicator.
6324 @emph{Check pragma casing.}
6325 Pragma names must be written in mixed case, that is, the
6326 initial letter and any letter following an underscore must be uppercase.
6327 All other letters must be lowercase.
6329 @item ^r^REFERENCES^
6330 @emph{Check references.}
6331 All identifier references must be cased in the same way as the
6332 corresponding declaration. No specific casing style is imposed on
6333 identifiers. The only requirement is for consistency of references
6336 @item ^S^STATEMENTS_AFTER_THEN_ELSE^
6337 @emph{Check no statements after THEN/ELSE.}
6338 No statements are allowed
6339 on the same line as a THEN or ELSE keyword following the
6340 keyword in an IF statement. OR ELSE and AND THEN are not affected,
6341 and a special exception allows a pragma to appear after ELSE.
6344 @emph{Check separate specs.}
6345 Separate declarations (``specs'') are required for subprograms (a
6346 body is not allowed to serve as its own declaration). The only
6347 exception is that parameterless library level procedures are
6348 not required to have a separate declaration. This exception covers
6349 the most frequent form of main program procedures.
6352 @emph{Check token spacing.}
6353 The following token spacing rules are enforced:
6358 The keywords @code{@b{abs}} and @code{@b{not}} must be followed by a space.
6361 The token @code{=>} must be surrounded by spaces.
6364 The token @code{<>} must be preceded by a space or a left parenthesis.
6367 Binary operators other than @code{**} must be surrounded by spaces.
6368 There is no restriction on the layout of the @code{**} binary operator.
6371 Colon must be surrounded by spaces.
6374 Colon-equal (assignment, initialization) must be surrounded by spaces.
6377 Comma must be the first non-blank character on the line, or be
6378 immediately preceded by a non-blank character, and must be followed
6382 If the token preceding a left parenthesis ends with a letter or digit, then
6383 a space must separate the two tokens.
6386 A right parenthesis must either be the first non-blank character on
6387 a line, or it must be preceded by a non-blank character.
6390 A semicolon must not be preceded by a space, and must not be followed by
6391 a non-blank character.
6394 A unary plus or minus may not be followed by a space.
6397 A vertical bar must be surrounded by spaces.
6400 @item ^u^UNNECESSARY_BLANK_LINES^
6401 @emph{Check unnecessary blank lines.}
6402 Unnecessary blank lines are not allowed. A blank line is considered
6403 unnecessary if it appears at the end of the file, or if more than
6404 one blank line occurs in sequence.
6406 @item ^x^XTRA_PARENS^
6407 @emph{Check extra parentheses.}
6408 Unnecessary extra level of parentheses (C-style) are not allowed
6409 around conditions in @code{if} statements, @code{while} statements and
6410 @code{exit} statements.
6412 @item ^y^ALL_BUILTIN^
6413 @emph{Set all standard style check options}
6414 This is equivalent to @code{gnaty3aAbcefhiklmnprst}, that is all checking
6415 options enabled with the exception of @option{-gnatyo}, @option{-gnatyI},
6416 @option{-gnatyS}, @option{-gnatyLnnn},
6417 @option{-gnatyd}, @option{-gnatyu}, and @option{-gnatyx}.
6421 @emph{Remove style check options}
6422 This causes any subsequent options in the string to act as canceling the
6423 corresponding style check option. To cancel maximum nesting level control,
6424 use @option{L} parameter witout any integer value after that, because any
6425 digit following @option{-} in the parameter string of the @option{-gnaty}
6426 option will be threated as canceling indentation check. The same is true
6427 for @option{M} parameter. @option{y} and @option{N} parameters are not
6428 allowed after @option{-}.
6431 This causes any subsequent options in the string to enable the corresponding
6432 style check option. That is, it cancels the effect of a previous ^-^REMOVE^,
6438 @emph{Removing style check options}
6439 If the name of a style check is preceded by @option{NO} then the corresponding
6440 style check is turned off. For example @option{NOCOMMENTS} turns off style
6441 checking for comments.
6446 In the above rules, appearing in column one is always permitted, that is,
6447 counts as meeting either a requirement for a required preceding space,
6448 or as meeting a requirement for no preceding space.
6450 Appearing at the end of a line is also always permitted, that is, counts
6451 as meeting either a requirement for a following space, or as meeting
6452 a requirement for no following space.
6455 If any of these style rules is violated, a message is generated giving
6456 details on the violation. The initial characters of such messages are
6457 always ``@code{(style)}''. Note that these messages are treated as warning
6458 messages, so they normally do not prevent the generation of an object
6459 file. The @option{-gnatwe} switch can be used to treat warning messages,
6460 including style messages, as fatal errors.
6464 @option{-gnaty} on its own (that is not
6465 followed by any letters or digits), then the effect is equivalent
6466 to the use of @option{-gnatyy}, as described above, that is all
6467 built-in standard style check options are enabled.
6471 /STYLE_CHECKS=ALL_BUILTIN enables all checking options with
6472 the exception of ORDERED_SUBPROGRAMS, UNNECESSARY_BLANK_LINES,
6473 XTRA_PARENS, and DOS_LINE_ENDINGS. In addition
6485 clears any previously set style checks.
6487 @node Run-Time Checks
6488 @subsection Run-Time Checks
6489 @cindex Division by zero
6490 @cindex Access before elaboration
6491 @cindex Checks, division by zero
6492 @cindex Checks, access before elaboration
6493 @cindex Checks, stack overflow checking
6496 By default, the following checks are suppressed: integer overflow
6497 checks, stack overflow checks, and checks for access before
6498 elaboration on subprogram calls. All other checks, including range
6499 checks and array bounds checks, are turned on by default. The
6500 following @command{gcc} switches refine this default behavior.
6505 @cindex @option{-gnatp} (@command{gcc})
6506 @cindex Suppressing checks
6507 @cindex Checks, suppressing
6509 Suppress all run-time checks as though @code{pragma Suppress (All_checks)}
6510 had been present in the source. Validity checks are also suppressed (in
6511 other words @option{-gnatp} also implies @option{-gnatVn}.
6512 Use this switch to improve the performance
6513 of the code at the expense of safety in the presence of invalid data or
6516 Note that when checks are suppressed, the compiler is allowed, but not
6517 required, to omit the checking code. If the run-time cost of the
6518 checking code is zero or near-zero, the compiler will generate it even
6519 if checks are suppressed. In particular, if the compiler can prove
6520 that a certain check will necessarily fail, it will generate code to
6521 do an unconditional ``raise'', even if checks are suppressed. The
6522 compiler warns in this case.
6524 Of course, run-time checks are omitted whenever the compiler can prove
6525 that they will not fail, whether or not checks are suppressed.
6527 Note that if you suppress a check that would have failed, program
6528 execution is erroneous, which means the behavior is totally
6529 unpredictable. The program might crash, or print wrong answers, or
6530 do anything else. It might even do exactly what you wanted it to do
6531 (and then it might start failing mysteriously next week or next
6532 year). The compiler will generate code based on the assumption that
6533 the condition being checked is true, which can result in disaster if
6534 that assumption is wrong.
6537 @cindex @option{-gnato} (@command{gcc})
6538 @cindex Overflow checks
6539 @cindex Check, overflow
6540 Enables overflow checking for integer operations.
6541 This causes GNAT to generate slower and larger executable
6542 programs by adding code to check for overflow (resulting in raising
6543 @code{Constraint_Error} as required by standard Ada
6544 semantics). These overflow checks correspond to situations in which
6545 the true value of the result of an operation may be outside the base
6546 range of the result type. The following example shows the distinction:
6548 @smallexample @c ada
6549 X1 : Integer := "Integer'Last";
6550 X2 : Integer range 1 .. 5 := "5";
6551 X3 : Integer := "Integer'Last";
6552 X4 : Integer range 1 .. 5 := "5";
6553 F : Float := "2.0E+20";
6562 Note that if explicit values are assigned at compile time, the
6563 compiler may be able to detect overflow at compile time, in which case
6564 no actual run-time checking code is required, and Constraint_Error
6565 will be raised unconditionally, with or without
6566 @option{-gnato}. That's why the assigned values in the above fragment
6567 are in quotes, the meaning is "assign a value not known to the
6568 compiler that happens to be equal to ...". The remaining discussion
6569 assumes that the compiler cannot detect the values at compile time.
6571 Here the first addition results in a value that is outside the base range
6572 of Integer, and hence requires an overflow check for detection of the
6573 constraint error. Thus the first assignment to @code{X1} raises a
6574 @code{Constraint_Error} exception only if @option{-gnato} is set.
6576 The second increment operation results in a violation of the explicit
6577 range constraint; such range checks are performed by default, and are
6578 unaffected by @option{-gnato}.
6580 The two conversions of @code{F} both result in values that are outside
6581 the base range of type @code{Integer} and thus will raise
6582 @code{Constraint_Error} exceptions only if @option{-gnato} is used.
6583 The fact that the result of the second conversion is assigned to
6584 variable @code{X4} with a restricted range is irrelevant, since the problem
6585 is in the conversion, not the assignment.
6587 Basically the rule is that in the default mode (@option{-gnato} not
6588 used), the generated code assures that all integer variables stay
6589 within their declared ranges, or within the base range if there is
6590 no declared range. This prevents any serious problems like indexes
6591 out of range for array operations.
6593 What is not checked in default mode is an overflow that results in
6594 an in-range, but incorrect value. In the above example, the assignments
6595 to @code{X1}, @code{X2}, @code{X3} all give results that are within the
6596 range of the target variable, but the result is wrong in the sense that
6597 it is too large to be represented correctly. Typically the assignment
6598 to @code{X1} will result in wrap around to the largest negative number.
6599 The conversions of @code{F} will result in some @code{Integer} value
6600 and if that integer value is out of the @code{X4} range then the
6601 subsequent assignment would generate an exception.
6603 @findex Machine_Overflows
6604 Note that the @option{-gnato} switch does not affect the code generated
6605 for any floating-point operations; it applies only to integer
6607 For floating-point, GNAT has the @code{Machine_Overflows}
6608 attribute set to @code{False} and the normal mode of operation is to
6609 generate IEEE NaN and infinite values on overflow or invalid operations
6610 (such as dividing 0.0 by 0.0).
6612 The reason that we distinguish overflow checking from other kinds of
6613 range constraint checking is that a failure of an overflow check, unlike
6614 for example the failure of a range check, can result in an incorrect
6615 value, but cannot cause random memory destruction (like an out of range
6616 subscript), or a wild jump (from an out of range case value). Overflow
6617 checking is also quite expensive in time and space, since in general it
6618 requires the use of double length arithmetic.
6620 Note again that @option{-gnato} is off by default, so overflow checking is
6621 not performed in default mode. This means that out of the box, with the
6622 default settings, GNAT does not do all the checks expected from the
6623 language description in the Ada Reference Manual. If you want all constraint
6624 checks to be performed, as described in this Manual, then you must
6625 explicitly use the -gnato switch either on the @command{gnatmake} or
6626 @command{gcc} command.
6629 @cindex @option{-gnatE} (@command{gcc})
6630 @cindex Elaboration checks
6631 @cindex Check, elaboration
6632 Enables dynamic checks for access-before-elaboration
6633 on subprogram calls and generic instantiations.
6634 Note that @option{-gnatE} is not necessary for safety, because in the
6635 default mode, GNAT ensures statically that the checks would not fail.
6636 For full details of the effect and use of this switch,
6637 @xref{Compiling Using gcc}.
6640 @cindex @option{-fstack-check} (@command{gcc})
6641 @cindex Stack Overflow Checking
6642 @cindex Checks, stack overflow checking
6643 Activates stack overflow checking. For full details of the effect and use of
6644 this switch see @ref{Stack Overflow Checking}.
6649 The setting of these switches only controls the default setting of the
6650 checks. You may modify them using either @code{Suppress} (to remove
6651 checks) or @code{Unsuppress} (to add back suppressed checks) pragmas in
6654 @node Using gcc for Syntax Checking
6655 @subsection Using @command{gcc} for Syntax Checking
6658 @cindex @option{-gnats} (@command{gcc})
6662 The @code{s} stands for ``syntax''.
6665 Run GNAT in syntax checking only mode. For
6666 example, the command
6669 $ gcc -c -gnats x.adb
6673 compiles file @file{x.adb} in syntax-check-only mode. You can check a
6674 series of files in a single command
6676 , and can use wild cards to specify such a group of files.
6677 Note that you must specify the @option{-c} (compile
6678 only) flag in addition to the @option{-gnats} flag.
6681 You may use other switches in conjunction with @option{-gnats}. In
6682 particular, @option{-gnatl} and @option{-gnatv} are useful to control the
6683 format of any generated error messages.
6685 When the source file is empty or contains only empty lines and/or comments,
6686 the output is a warning:
6689 $ gcc -c -gnats -x ada toto.txt
6690 toto.txt:1:01: warning: empty file, contains no compilation units
6694 Otherwise, the output is simply the error messages, if any. No object file or
6695 ALI file is generated by a syntax-only compilation. Also, no units other
6696 than the one specified are accessed. For example, if a unit @code{X}
6697 @code{with}'s a unit @code{Y}, compiling unit @code{X} in syntax
6698 check only mode does not access the source file containing unit
6701 @cindex Multiple units, syntax checking
6702 Normally, GNAT allows only a single unit in a source file. However, this
6703 restriction does not apply in syntax-check-only mode, and it is possible
6704 to check a file containing multiple compilation units concatenated
6705 together. This is primarily used by the @code{gnatchop} utility
6706 (@pxref{Renaming Files Using gnatchop}).
6709 @node Using gcc for Semantic Checking
6710 @subsection Using @command{gcc} for Semantic Checking
6713 @cindex @option{-gnatc} (@command{gcc})
6717 The @code{c} stands for ``check''.
6719 Causes the compiler to operate in semantic check mode,
6720 with full checking for all illegalities specified in the
6721 Ada Reference Manual, but without generation of any object code
6722 (no object file is generated).
6724 Because dependent files must be accessed, you must follow the GNAT
6725 semantic restrictions on file structuring to operate in this mode:
6729 The needed source files must be accessible
6730 (@pxref{Search Paths and the Run-Time Library (RTL)}).
6733 Each file must contain only one compilation unit.
6736 The file name and unit name must match (@pxref{File Naming Rules}).
6739 The output consists of error messages as appropriate. No object file is
6740 generated. An @file{ALI} file is generated for use in the context of
6741 cross-reference tools, but this file is marked as not being suitable
6742 for binding (since no object file is generated).
6743 The checking corresponds exactly to the notion of
6744 legality in the Ada Reference Manual.
6746 Any unit can be compiled in semantics-checking-only mode, including
6747 units that would not normally be compiled (subunits,
6748 and specifications where a separate body is present).
6751 @node Compiling Different Versions of Ada
6752 @subsection Compiling Different Versions of Ada
6755 The switches described in this section allow you to explicitly specify
6756 the version of the Ada language that your programs are written in.
6757 By default @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
6758 but you can also specify @value{NONDEFAULTLANGUAGEVERSION} or
6759 indicate Ada 83 compatibility mode.
6762 @cindex Compatibility with Ada 83
6764 @item -gnat83 (Ada 83 Compatibility Mode)
6765 @cindex @option{-gnat83} (@command{gcc})
6766 @cindex ACVC, Ada 83 tests
6770 Although GNAT is primarily an Ada 95 / Ada 2005 compiler, this switch
6771 specifies that the program is to be compiled in Ada 83 mode. With
6772 @option{-gnat83}, GNAT rejects most post-Ada 83 extensions and applies Ada 83
6773 semantics where this can be done easily.
6774 It is not possible to guarantee this switch does a perfect
6775 job; some subtle tests, such as are
6776 found in earlier ACVC tests (and that have been removed from the ACATS suite
6777 for Ada 95), might not compile correctly.
6778 Nevertheless, this switch may be useful in some circumstances, for example
6779 where, due to contractual reasons, existing code needs to be maintained
6780 using only Ada 83 features.
6782 With few exceptions (most notably the need to use @code{<>} on
6783 @cindex Generic formal parameters
6784 unconstrained generic formal parameters, the use of the new Ada 95 / Ada 2005
6785 reserved words, and the use of packages
6786 with optional bodies), it is not necessary to specify the
6787 @option{-gnat83} switch when compiling Ada 83 programs, because, with rare
6788 exceptions, Ada 95 and Ada 2005 are upwardly compatible with Ada 83. Thus
6789 a correct Ada 83 program is usually also a correct program
6790 in these later versions of the language standard.
6791 For further information, please refer to @ref{Compatibility and Porting Guide}.
6793 @item -gnat95 (Ada 95 mode)
6794 @cindex @option{-gnat95} (@command{gcc})
6798 This switch directs the compiler to implement the Ada 95 version of the
6800 Since Ada 95 is almost completely upwards
6801 compatible with Ada 83, Ada 83 programs may generally be compiled using
6802 this switch (see the description of the @option{-gnat83} switch for further
6803 information about Ada 83 mode).
6804 If an Ada 2005 program is compiled in Ada 95 mode,
6805 uses of the new Ada 2005 features will cause error
6806 messages or warnings.
6808 This switch also can be used to cancel the effect of a previous
6809 @option{-gnat83} or @option{-gnat05} switch earlier in the command line.
6811 @item -gnat05 (Ada 2005 mode)
6812 @cindex @option{-gnat05} (@command{gcc})
6813 @cindex Ada 2005 mode
6816 This switch directs the compiler to implement the Ada 2005 version of the
6818 Since Ada 2005 is almost completely upwards
6819 compatible with Ada 95 (and thus also with Ada 83), Ada 83 and Ada 95 programs
6820 may generally be compiled using this switch (see the description of the
6821 @option{-gnat83} and @option{-gnat95} switches for further
6824 For information about the approved ``Ada Issues'' that have been incorporated
6825 into Ada 2005, see @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs}.
6826 Included with GNAT releases is a file @file{features-ada0y} that describes
6827 the set of implemented Ada 2005 features.
6831 @node Character Set Control
6832 @subsection Character Set Control
6834 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
6835 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
6838 Normally GNAT recognizes the Latin-1 character set in source program
6839 identifiers, as described in the Ada Reference Manual.
6841 GNAT to recognize alternate character sets in identifiers. @var{c} is a
6842 single character ^^or word^ indicating the character set, as follows:
6846 ISO 8859-1 (Latin-1) identifiers
6849 ISO 8859-2 (Latin-2) letters allowed in identifiers
6852 ISO 8859-3 (Latin-3) letters allowed in identifiers
6855 ISO 8859-4 (Latin-4) letters allowed in identifiers
6858 ISO 8859-5 (Cyrillic) letters allowed in identifiers
6861 ISO 8859-15 (Latin-9) letters allowed in identifiers
6864 IBM PC letters (code page 437) allowed in identifiers
6867 IBM PC letters (code page 850) allowed in identifiers
6869 @item ^f^FULL_UPPER^
6870 Full upper-half codes allowed in identifiers
6873 No upper-half codes allowed in identifiers
6876 Wide-character codes (that is, codes greater than 255)
6877 allowed in identifiers
6880 @xref{Foreign Language Representation}, for full details on the
6881 implementation of these character sets.
6883 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
6884 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
6885 Specify the method of encoding for wide characters.
6886 @var{e} is one of the following:
6891 Hex encoding (brackets coding also recognized)
6894 Upper half encoding (brackets encoding also recognized)
6897 Shift/JIS encoding (brackets encoding also recognized)
6900 EUC encoding (brackets encoding also recognized)
6903 UTF-8 encoding (brackets encoding also recognized)
6906 Brackets encoding only (default value)
6908 For full details on these encoding
6909 methods see @ref{Wide Character Encodings}.
6910 Note that brackets coding is always accepted, even if one of the other
6911 options is specified, so for example @option{-gnatW8} specifies that both
6912 brackets and UTF-8 encodings will be recognized. The units that are
6913 with'ed directly or indirectly will be scanned using the specified
6914 representation scheme, and so if one of the non-brackets scheme is
6915 used, it must be used consistently throughout the program. However,
6916 since brackets encoding is always recognized, it may be conveniently
6917 used in standard libraries, allowing these libraries to be used with
6918 any of the available coding schemes.
6921 If no @option{-gnatW?} parameter is present, then the default
6922 representation is normally Brackets encoding only. However, if the
6923 first three characters of the file are 16#EF# 16#BB# 16#BF# (the standard
6924 byte order mark or BOM for UTF-8), then these three characters are
6925 skipped and the default representation for the file is set to UTF-8.
6927 Note that the wide character representation that is specified (explicitly
6928 or by default) for the main program also acts as the default encoding used
6929 for Wide_Text_IO files if not specifically overridden by a WCEM form
6933 @node File Naming Control
6934 @subsection File Naming Control
6937 @item ^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{n}
6938 @cindex @option{-gnatk} (@command{gcc})
6939 Activates file name ``krunching''. @var{n}, a decimal integer in the range
6940 1-999, indicates the maximum allowable length of a file name (not
6941 including the @file{.ads} or @file{.adb} extension). The default is not
6942 to enable file name krunching.
6944 For the source file naming rules, @xref{File Naming Rules}.
6947 @node Subprogram Inlining Control
6948 @subsection Subprogram Inlining Control
6953 @cindex @option{-gnatn} (@command{gcc})
6955 The @code{n} here is intended to suggest the first syllable of the
6958 GNAT recognizes and processes @code{Inline} pragmas. However, for the
6959 inlining to actually occur, optimization must be enabled. To enable
6960 inlining of subprograms specified by pragma @code{Inline},
6961 you must also specify this switch.
6962 In the absence of this switch, GNAT does not attempt
6963 inlining and does not need to access the bodies of
6964 subprograms for which @code{pragma Inline} is specified if they are not
6965 in the current unit.
6967 If you specify this switch the compiler will access these bodies,
6968 creating an extra source dependency for the resulting object file, and
6969 where possible, the call will be inlined.
6970 For further details on when inlining is possible
6971 see @ref{Inlining of Subprograms}.
6974 @cindex @option{-gnatN} (@command{gcc})
6975 This switch activates front-end inlining which also
6976 generates additional dependencies.
6978 When using a gcc-based back end (in practice this means using any version
6979 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
6980 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
6981 Historically front end inlining was more extensive than the gcc back end
6982 inlining, but that is no longer the case.
6985 @node Auxiliary Output Control
6986 @subsection Auxiliary Output Control
6990 @cindex @option{-gnatt} (@command{gcc})
6991 @cindex Writing internal trees
6992 @cindex Internal trees, writing to file
6993 Causes GNAT to write the internal tree for a unit to a file (with the
6994 extension @file{.adt}.
6995 This not normally required, but is used by separate analysis tools.
6997 these tools do the necessary compilations automatically, so you should
6998 not have to specify this switch in normal operation.
7001 @cindex @option{-gnatu} (@command{gcc})
7002 Print a list of units required by this compilation on @file{stdout}.
7003 The listing includes all units on which the unit being compiled depends
7004 either directly or indirectly.
7007 @item -pass-exit-codes
7008 @cindex @option{-pass-exit-codes} (@command{gcc})
7009 If this switch is not used, the exit code returned by @command{gcc} when
7010 compiling multiple files indicates whether all source files have
7011 been successfully used to generate object files or not.
7013 When @option{-pass-exit-codes} is used, @command{gcc} exits with an extended
7014 exit status and allows an integrated development environment to better
7015 react to a compilation failure. Those exit status are:
7019 There was an error in at least one source file.
7021 At least one source file did not generate an object file.
7023 The compiler died unexpectedly (internal error for example).
7025 An object file has been generated for every source file.
7030 @node Debugging Control
7031 @subsection Debugging Control
7035 @cindex Debugging options
7038 @cindex @option{-gnatd} (@command{gcc})
7039 Activate internal debugging switches. @var{x} is a letter or digit, or
7040 string of letters or digits, which specifies the type of debugging
7041 outputs desired. Normally these are used only for internal development
7042 or system debugging purposes. You can find full documentation for these
7043 switches in the body of the @code{Debug} unit in the compiler source
7044 file @file{debug.adb}.
7048 @cindex @option{-gnatG} (@command{gcc})
7049 This switch causes the compiler to generate auxiliary output containing
7050 a pseudo-source listing of the generated expanded code. Like most Ada
7051 compilers, GNAT works by first transforming the high level Ada code into
7052 lower level constructs. For example, tasking operations are transformed
7053 into calls to the tasking run-time routines. A unique capability of GNAT
7054 is to list this expanded code in a form very close to normal Ada source.
7055 This is very useful in understanding the implications of various Ada
7056 usage on the efficiency of the generated code. There are many cases in
7057 Ada (e.g.@: the use of controlled types), where simple Ada statements can
7058 generate a lot of run-time code. By using @option{-gnatG} you can identify
7059 these cases, and consider whether it may be desirable to modify the coding
7060 approach to improve efficiency.
7062 The optional parameter @code{nn} if present after -gnatG specifies an
7063 alternative maximum line length that overrides the normal default of 72.
7064 This value is in the range 40-999999, values less than 40 being silently
7065 reset to 40. The equal sign is optional.
7067 The format of the output is very similar to standard Ada source, and is
7068 easily understood by an Ada programmer. The following special syntactic
7069 additions correspond to low level features used in the generated code that
7070 do not have any exact analogies in pure Ada source form. The following
7071 is a partial list of these special constructions. See the spec
7072 of package @code{Sprint} in file @file{sprint.ads} for a full list.
7074 If the switch @option{-gnatL} is used in conjunction with
7075 @cindex @option{-gnatL} (@command{gcc})
7076 @option{-gnatG}, then the original source lines are interspersed
7077 in the expanded source (as comment lines with the original line number).
7080 @item new @var{xxx} @r{[}storage_pool = @var{yyy}@r{]}
7081 Shows the storage pool being used for an allocator.
7083 @item at end @var{procedure-name};
7084 Shows the finalization (cleanup) procedure for a scope.
7086 @item (if @var{expr} then @var{expr} else @var{expr})
7087 Conditional expression equivalent to the @code{x?y:z} construction in C.
7089 @item @var{target}^^^(@var{source})
7090 A conversion with floating-point truncation instead of rounding.
7092 @item @var{target}?(@var{source})
7093 A conversion that bypasses normal Ada semantic checking. In particular
7094 enumeration types and fixed-point types are treated simply as integers.
7096 @item @var{target}?^^^(@var{source})
7097 Combines the above two cases.
7099 @item @var{x} #/ @var{y}
7100 @itemx @var{x} #mod @var{y}
7101 @itemx @var{x} #* @var{y}
7102 @itemx @var{x} #rem @var{y}
7103 A division or multiplication of fixed-point values which are treated as
7104 integers without any kind of scaling.
7106 @item free @var{expr} @r{[}storage_pool = @var{xxx}@r{]}
7107 Shows the storage pool associated with a @code{free} statement.
7109 @item [subtype or type declaration]
7110 Used to list an equivalent declaration for an internally generated
7111 type that is referenced elsewhere in the listing.
7113 @item freeze @var{type-name} @ovar{actions}
7114 Shows the point at which @var{type-name} is frozen, with possible
7115 associated actions to be performed at the freeze point.
7117 @item reference @var{itype}
7118 Reference (and hence definition) to internal type @var{itype}.
7120 @item @var{function-name}! (@var{arg}, @var{arg}, @var{arg})
7121 Intrinsic function call.
7123 @item @var{label-name} : label
7124 Declaration of label @var{labelname}.
7126 @item #$ @var{subprogram-name}
7127 An implicit call to a run-time support routine
7128 (to meet the requirement of H.3.1(9) in a
7131 @item @var{expr} && @var{expr} && @var{expr} @dots{} && @var{expr}
7132 A multiple concatenation (same effect as @var{expr} & @var{expr} &
7133 @var{expr}, but handled more efficiently).
7135 @item [constraint_error]
7136 Raise the @code{Constraint_Error} exception.
7138 @item @var{expression}'reference
7139 A pointer to the result of evaluating @var{expression}.
7141 @item @var{target-type}!(@var{source-expression})
7142 An unchecked conversion of @var{source-expression} to @var{target-type}.
7144 @item [@var{numerator}/@var{denominator}]
7145 Used to represent internal real literals (that) have no exact
7146 representation in base 2-16 (for example, the result of compile time
7147 evaluation of the expression 1.0/27.0).
7151 @cindex @option{-gnatD} (@command{gcc})
7152 When used in conjunction with @option{-gnatG}, this switch causes
7153 the expanded source, as described above for
7154 @option{-gnatG} to be written to files with names
7155 @file{^xxx.dg^XXX_DG^}, where @file{xxx} is the normal file name,
7156 instead of to the standard output file. For
7157 example, if the source file name is @file{hello.adb}, then a file
7158 @file{^hello.adb.dg^HELLO.ADB_DG^} will be written. The debugging
7159 information generated by the @command{gcc} @option{^-g^/DEBUG^} switch
7160 will refer to the generated @file{^xxx.dg^XXX_DG^} file. This allows
7161 you to do source level debugging using the generated code which is
7162 sometimes useful for complex code, for example to find out exactly
7163 which part of a complex construction raised an exception. This switch
7164 also suppress generation of cross-reference information (see
7165 @option{-gnatx}) since otherwise the cross-reference information
7166 would refer to the @file{^.dg^.DG^} file, which would cause
7167 confusion since this is not the original source file.
7169 Note that @option{-gnatD} actually implies @option{-gnatG}
7170 automatically, so it is not necessary to give both options.
7171 In other words @option{-gnatD} is equivalent to @option{-gnatDG}).
7173 If the switch @option{-gnatL} is used in conjunction with
7174 @cindex @option{-gnatL} (@command{gcc})
7175 @option{-gnatDG}, then the original source lines are interspersed
7176 in the expanded source (as comment lines with the original line number).
7178 The optional parameter @code{nn} if present after -gnatD specifies an
7179 alternative maximum line length that overrides the normal default of 72.
7180 This value is in the range 40-999999, values less than 40 being silently
7181 reset to 40. The equal sign is optional.
7184 @cindex @option{-gnatr} (@command{gcc})
7185 @cindex pragma Restrictions
7186 This switch causes pragma Restrictions to be treated as Restriction_Warnings
7187 so that violation of restrictions causes warnings rather than illegalities.
7188 This is useful during the development process when new restrictions are added
7189 or investigated. The switch also causes pragma Profile to be treated as
7190 Profile_Warnings, and pragma Restricted_Run_Time and pragma Ravenscar set
7191 restriction warnings rather than restrictions.
7194 @item -gnatR@r{[}0@r{|}1@r{|}2@r{|}3@r{[}s@r{]]}
7195 @cindex @option{-gnatR} (@command{gcc})
7196 This switch controls output from the compiler of a listing showing
7197 representation information for declared types and objects. For
7198 @option{-gnatR0}, no information is output (equivalent to omitting
7199 the @option{-gnatR} switch). For @option{-gnatR1} (which is the default,
7200 so @option{-gnatR} with no parameter has the same effect), size and alignment
7201 information is listed for declared array and record types. For
7202 @option{-gnatR2}, size and alignment information is listed for all
7203 declared types and objects. Finally @option{-gnatR3} includes symbolic
7204 expressions for values that are computed at run time for
7205 variant records. These symbolic expressions have a mostly obvious
7206 format with #n being used to represent the value of the n'th
7207 discriminant. See source files @file{repinfo.ads/adb} in the
7208 @code{GNAT} sources for full details on the format of @option{-gnatR3}
7209 output. If the switch is followed by an s (e.g.@: @option{-gnatR2s}), then
7210 the output is to a file with the name @file{^file.rep^file_REP^} where
7211 file is the name of the corresponding source file.
7214 @item /REPRESENTATION_INFO
7215 @cindex @option{/REPRESENTATION_INFO} (@command{gcc})
7216 This qualifier controls output from the compiler of a listing showing
7217 representation information for declared types and objects. For
7218 @option{/REPRESENTATION_INFO=NONE}, no information is output
7219 (equivalent to omitting the @option{/REPRESENTATION_INFO} qualifier).
7220 @option{/REPRESENTATION_INFO} without option is equivalent to
7221 @option{/REPRESENTATION_INFO=ARRAYS}.
7222 For @option{/REPRESENTATION_INFO=ARRAYS}, size and alignment
7223 information is listed for declared array and record types. For
7224 @option{/REPRESENTATION_INFO=OBJECTS}, size and alignment information
7225 is listed for all expression information for values that are computed
7226 at run time for variant records. These symbolic expressions have a mostly
7227 obvious format with #n being used to represent the value of the n'th
7228 discriminant. See source files @file{REPINFO.ADS/ADB} in the
7229 @code{GNAT} sources for full details on the format of
7230 @option{/REPRESENTATION_INFO=SYMBOLIC} output.
7231 If _FILE is added at the end of an option
7232 (e.g.@: @option{/REPRESENTATION_INFO=ARRAYS_FILE}),
7233 then the output is to a file with the name @file{file_REP} where
7234 file is the name of the corresponding source file.
7236 Note that it is possible for record components to have zero size. In
7237 this case, the component clause uses an obvious extension of permitted
7238 Ada syntax, for example @code{at 0 range 0 .. -1}.
7240 Representation information requires that code be generated (since it is the
7241 code generator that lays out complex data structures). If an attempt is made
7242 to output representation information when no code is generated, for example
7243 when a subunit is compiled on its own, then no information can be generated
7244 and the compiler outputs a message to this effect.
7247 @cindex @option{-gnatS} (@command{gcc})
7248 The use of the switch @option{-gnatS} for an
7249 Ada compilation will cause the compiler to output a
7250 representation of package Standard in a form very
7251 close to standard Ada. It is not quite possible to
7252 do this entirely in standard Ada (since new
7253 numeric base types cannot be created in standard
7254 Ada), but the output is easily
7255 readable to any Ada programmer, and is useful to
7256 determine the characteristics of target dependent
7257 types in package Standard.
7260 @cindex @option{-gnatx} (@command{gcc})
7261 Normally the compiler generates full cross-referencing information in
7262 the @file{ALI} file. This information is used by a number of tools,
7263 including @code{gnatfind} and @code{gnatxref}. The @option{-gnatx} switch
7264 suppresses this information. This saves some space and may slightly
7265 speed up compilation, but means that these tools cannot be used.
7268 @node Exception Handling Control
7269 @subsection Exception Handling Control
7272 GNAT uses two methods for handling exceptions at run-time. The
7273 @code{setjmp/longjmp} method saves the context when entering
7274 a frame with an exception handler. Then when an exception is
7275 raised, the context can be restored immediately, without the
7276 need for tracing stack frames. This method provides very fast
7277 exception propagation, but introduces significant overhead for
7278 the use of exception handlers, even if no exception is raised.
7280 The other approach is called ``zero cost'' exception handling.
7281 With this method, the compiler builds static tables to describe
7282 the exception ranges. No dynamic code is required when entering
7283 a frame containing an exception handler. When an exception is
7284 raised, the tables are used to control a back trace of the
7285 subprogram invocation stack to locate the required exception
7286 handler. This method has considerably poorer performance for
7287 the propagation of exceptions, but there is no overhead for
7288 exception handlers if no exception is raised. Note that in this
7289 mode and in the context of mixed Ada and C/C++ programming,
7290 to propagate an exception through a C/C++ code, the C/C++ code
7291 must be compiled with the @option{-funwind-tables} GCC's
7294 The following switches may be used to control which of the
7295 two exception handling methods is used.
7301 @cindex @option{--RTS=sjlj} (@command{gnatmake})
7302 This switch causes the setjmp/longjmp run-time (when available) to be used
7303 for exception handling. If the default
7304 mechanism for the target is zero cost exceptions, then
7305 this switch can be used to modify this default, and must be
7306 used for all units in the partition.
7307 This option is rarely used. One case in which it may be
7308 advantageous is if you have an application where exception
7309 raising is common and the overall performance of the
7310 application is improved by favoring exception propagation.
7313 @cindex @option{--RTS=zcx} (@command{gnatmake})
7314 @cindex Zero Cost Exceptions
7315 This switch causes the zero cost approach to be used
7316 for exception handling. If this is the default mechanism for the
7317 target (see below), then this switch is unneeded. If the default
7318 mechanism for the target is setjmp/longjmp exceptions, then
7319 this switch can be used to modify this default, and must be
7320 used for all units in the partition.
7321 This option can only be used if the zero cost approach
7322 is available for the target in use, otherwise it will generate an error.
7326 The same option @option{--RTS} must be used both for @command{gcc}
7327 and @command{gnatbind}. Passing this option to @command{gnatmake}
7328 (@pxref{Switches for gnatmake}) will ensure the required consistency
7329 through the compilation and binding steps.
7331 @node Units to Sources Mapping Files
7332 @subsection Units to Sources Mapping Files
7336 @item -gnatem^^=^@var{path}
7337 @cindex @option{-gnatem} (@command{gcc})
7338 A mapping file is a way to communicate to the compiler two mappings:
7339 from unit names to file names (without any directory information) and from
7340 file names to path names (with full directory information). These mappings
7341 are used by the compiler to short-circuit the path search.
7343 The use of mapping files is not required for correct operation of the
7344 compiler, but mapping files can improve efficiency, particularly when
7345 sources are read over a slow network connection. In normal operation,
7346 you need not be concerned with the format or use of mapping files,
7347 and the @option{-gnatem} switch is not a switch that you would use
7348 explicitly. it is intended only for use by automatic tools such as
7349 @command{gnatmake} running under the project file facility. The
7350 description here of the format of mapping files is provided
7351 for completeness and for possible use by other tools.
7353 A mapping file is a sequence of sets of three lines. In each set,
7354 the first line is the unit name, in lower case, with ``@code{%s}''
7356 specs and ``@code{%b}'' appended for bodies; the second line is the
7357 file name; and the third line is the path name.
7363 /gnat/project1/sources/main.2.ada
7366 When the switch @option{-gnatem} is specified, the compiler will create
7367 in memory the two mappings from the specified file. If there is any problem
7368 (nonexistent file, truncated file or duplicate entries), no mapping will
7371 Several @option{-gnatem} switches may be specified; however, only the last
7372 one on the command line will be taken into account.
7374 When using a project file, @command{gnatmake} create a temporary mapping file
7375 and communicates it to the compiler using this switch.
7379 @node Integrated Preprocessing
7380 @subsection Integrated Preprocessing
7383 GNAT sources may be preprocessed immediately before compilation.
7384 In this case, the actual
7385 text of the source is not the text of the source file, but is derived from it
7386 through a process called preprocessing. Integrated preprocessing is specified
7387 through switches @option{-gnatep} and/or @option{-gnateD}. @option{-gnatep}
7388 indicates, through a text file, the preprocessing data to be used.
7389 @option{-gnateD} specifies or modifies the values of preprocessing symbol.
7392 Note that when integrated preprocessing is used, the output from the
7393 preprocessor is not written to any external file. Instead it is passed
7394 internally to the compiler. If you need to preserve the result of
7395 preprocessing in a file, then you should use @command{gnatprep}
7396 to perform the desired preprocessing in stand-alone mode.
7399 It is recommended that @command{gnatmake} switch ^-s^/SWITCH_CHECK^ should be
7400 used when Integrated Preprocessing is used. The reason is that preprocessing
7401 with another Preprocessing Data file without changing the sources will
7402 not trigger recompilation without this switch.
7405 Note that @command{gnatmake} switch ^-m^/MINIMAL_RECOMPILATION^ will almost
7406 always trigger recompilation for sources that are preprocessed,
7407 because @command{gnatmake} cannot compute the checksum of the source after
7411 The actual preprocessing function is described in details in section
7412 @ref{Preprocessing Using gnatprep}. This section only describes how integrated
7413 preprocessing is triggered and parameterized.
7417 @item -gnatep=@var{file}
7418 @cindex @option{-gnatep} (@command{gcc})
7419 This switch indicates to the compiler the file name (without directory
7420 information) of the preprocessor data file to use. The preprocessor data file
7421 should be found in the source directories.
7424 A preprocessing data file is a text file with significant lines indicating
7425 how should be preprocessed either a specific source or all sources not
7426 mentioned in other lines. A significant line is a nonempty, non-comment line.
7427 Comments are similar to Ada comments.
7430 Each significant line starts with either a literal string or the character '*'.
7431 A literal string is the file name (without directory information) of the source
7432 to preprocess. A character '*' indicates the preprocessing for all the sources
7433 that are not specified explicitly on other lines (order of the lines is not
7434 significant). It is an error to have two lines with the same file name or two
7435 lines starting with the character '*'.
7438 After the file name or the character '*', another optional literal string
7439 indicating the file name of the definition file to be used for preprocessing
7440 (@pxref{Form of Definitions File}). The definition files are found by the
7441 compiler in one of the source directories. In some cases, when compiling
7442 a source in a directory other than the current directory, if the definition
7443 file is in the current directory, it may be necessary to add the current
7444 directory as a source directory through switch ^-I.^/SEARCH=[]^, otherwise
7445 the compiler would not find the definition file.
7448 Then, optionally, ^switches^switches^ similar to those of @code{gnatprep} may
7449 be found. Those ^switches^switches^ are:
7454 Causes both preprocessor lines and the lines deleted by
7455 preprocessing to be replaced by blank lines, preserving the line number.
7456 This ^switch^switch^ is always implied; however, if specified after @option{-c}
7457 it cancels the effect of @option{-c}.
7460 Causes both preprocessor lines and the lines deleted
7461 by preprocessing to be retained as comments marked
7462 with the special string ``@code{--! }''.
7464 @item -Dsymbol=value
7465 Define or redefine a symbol, associated with value. A symbol is an Ada
7466 identifier, or an Ada reserved word, with the exception of @code{if},
7467 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7468 @code{value} is either a literal string, an Ada identifier or any Ada reserved
7469 word. A symbol declared with this ^switch^switch^ replaces a symbol with the
7470 same name defined in a definition file.
7473 Causes a sorted list of symbol names and values to be
7474 listed on the standard output file.
7477 Causes undefined symbols to be treated as having the value @code{FALSE}
7479 of a preprocessor test. In the absence of this option, an undefined symbol in
7480 a @code{#if} or @code{#elsif} test will be treated as an error.
7485 Examples of valid lines in a preprocessor data file:
7488 "toto.adb" "prep.def" -u
7489 -- preprocess "toto.adb", using definition file "prep.def",
7490 -- undefined symbol are False.
7493 -- preprocess all other sources without a definition file;
7494 -- suppressed lined are commented; symbol VERSION has the value V101.
7496 "titi.adb" "prep2.def" -s
7497 -- preprocess "titi.adb", using definition file "prep2.def";
7498 -- list all symbols with their values.
7501 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=value@r{]}
7502 @cindex @option{-gnateD} (@command{gcc})
7503 Define or redefine a preprocessing symbol, associated with value. If no value
7504 is given on the command line, then the value of the symbol is @code{True}.
7505 A symbol is an identifier, following normal Ada (case-insensitive)
7506 rules for its syntax, and value is any sequence (including an empty sequence)
7507 of characters from the set (letters, digits, period, underline).
7508 Ada reserved words may be used as symbols, with the exceptions of @code{if},
7509 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7512 A symbol declared with this ^switch^switch^ on the command line replaces a
7513 symbol with the same name either in a definition file or specified with a
7514 ^switch^switch^ -D in the preprocessor data file.
7517 This switch is similar to switch @option{^-D^/ASSOCIATE^} of @code{gnatprep}.
7520 When integrated preprocessing is performed and the preprocessor modifies
7521 the source text, write the result of this preprocessing into a file
7522 <source>^.prep^_prep^.
7526 @node Code Generation Control
7527 @subsection Code Generation Control
7531 The GCC technology provides a wide range of target dependent
7532 @option{-m} switches for controlling
7533 details of code generation with respect to different versions of
7534 architectures. This includes variations in instruction sets (e.g.@:
7535 different members of the power pc family), and different requirements
7536 for optimal arrangement of instructions (e.g.@: different members of
7537 the x86 family). The list of available @option{-m} switches may be
7538 found in the GCC documentation.
7540 Use of these @option{-m} switches may in some cases result in improved
7543 The GNAT Pro technology is tested and qualified without any
7544 @option{-m} switches,
7545 so generally the most reliable approach is to avoid the use of these
7546 switches. However, we generally expect most of these switches to work
7547 successfully with GNAT Pro, and many customers have reported successful
7548 use of these options.
7550 Our general advice is to avoid the use of @option{-m} switches unless
7551 special needs lead to requirements in this area. In particular,
7552 there is no point in using @option{-m} switches to improve performance
7553 unless you actually see a performance improvement.
7557 @subsection Return Codes
7558 @cindex Return Codes
7559 @cindex @option{/RETURN_CODES=VMS}
7562 On VMS, GNAT compiled programs return POSIX-style codes by default,
7563 e.g.@: @option{/RETURN_CODES=POSIX}.
7565 To enable VMS style return codes, use GNAT BIND and LINK with the option
7566 @option{/RETURN_CODES=VMS}. For example:
7569 GNAT BIND MYMAIN.ALI /RETURN_CODES=VMS
7570 GNAT LINK MYMAIN.ALI /RETURN_CODES=VMS
7574 Programs built with /RETURN_CODES=VMS are suitable to be called in
7575 VMS DCL scripts. Programs compiled with the default /RETURN_CODES=POSIX
7576 are suitable for spawning with appropriate GNAT RTL routines.
7580 @node Search Paths and the Run-Time Library (RTL)
7581 @section Search Paths and the Run-Time Library (RTL)
7584 With the GNAT source-based library system, the compiler must be able to
7585 find source files for units that are needed by the unit being compiled.
7586 Search paths are used to guide this process.
7588 The compiler compiles one source file whose name must be given
7589 explicitly on the command line. In other words, no searching is done
7590 for this file. To find all other source files that are needed (the most
7591 common being the specs of units), the compiler examines the following
7592 directories, in the following order:
7596 The directory containing the source file of the main unit being compiled
7597 (the file name on the command line).
7600 Each directory named by an @option{^-I^/SOURCE_SEARCH^} switch given on the
7601 @command{gcc} command line, in the order given.
7604 @findex ADA_PRJ_INCLUDE_FILE
7605 Each of the directories listed in the text file whose name is given
7606 by the @env{ADA_PRJ_INCLUDE_FILE} ^environment variable^logical name^.
7609 @env{ADA_PRJ_INCLUDE_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
7610 driver when project files are used. It should not normally be set
7614 @findex ADA_INCLUDE_PATH
7615 Each of the directories listed in the value of the
7616 @env{ADA_INCLUDE_PATH} ^environment variable^logical name^.
7618 Construct this value
7619 exactly as the @env{PATH} environment variable: a list of directory
7620 names separated by colons (semicolons when working with the NT version).
7623 Normally, define this value as a logical name containing a comma separated
7624 list of directory names.
7626 This variable can also be defined by means of an environment string
7627 (an argument to the HP C exec* set of functions).
7631 DEFINE ANOTHER_PATH FOO:[BAG]
7632 DEFINE ADA_INCLUDE_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
7635 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
7636 first, followed by the standard Ada
7637 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADAINCLUDE].
7638 If this is not redefined, the user will obtain the HP Ada 83 IO packages
7639 (Text_IO, Sequential_IO, etc)
7640 instead of the standard Ada packages. Thus, in order to get the standard Ada
7641 packages by default, ADA_INCLUDE_PATH must be redefined.
7645 The content of the @file{ada_source_path} file which is part of the GNAT
7646 installation tree and is used to store standard libraries such as the
7647 GNAT Run Time Library (RTL) source files.
7649 @ref{Installing a library}
7654 Specifying the switch @option{^-I-^/NOCURRENT_DIRECTORY^}
7655 inhibits the use of the directory
7656 containing the source file named in the command line. You can still
7657 have this directory on your search path, but in this case it must be
7658 explicitly requested with a @option{^-I^/SOURCE_SEARCH^} switch.
7660 Specifying the switch @option{-nostdinc}
7661 inhibits the search of the default location for the GNAT Run Time
7662 Library (RTL) source files.
7664 The compiler outputs its object files and ALI files in the current
7667 Caution: The object file can be redirected with the @option{-o} switch;
7668 however, @command{gcc} and @code{gnat1} have not been coordinated on this
7669 so the @file{ALI} file will not go to the right place. Therefore, you should
7670 avoid using the @option{-o} switch.
7674 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
7675 children make up the GNAT RTL, together with the simple @code{System.IO}
7676 package used in the @code{"Hello World"} example. The sources for these units
7677 are needed by the compiler and are kept together in one directory. Not
7678 all of the bodies are needed, but all of the sources are kept together
7679 anyway. In a normal installation, you need not specify these directory
7680 names when compiling or binding. Either the environment variables or
7681 the built-in defaults cause these files to be found.
7683 In addition to the language-defined hierarchies (@code{System}, @code{Ada} and
7684 @code{Interfaces}), the GNAT distribution provides a fourth hierarchy,
7685 consisting of child units of @code{GNAT}. This is a collection of generally
7686 useful types, subprograms, etc. @xref{Top, GNAT Reference Manual, About
7687 This Guid, gnat_rm, GNAT Reference Manual}, for further details.
7689 Besides simplifying access to the RTL, a major use of search paths is
7690 in compiling sources from multiple directories. This can make
7691 development environments much more flexible.
7693 @node Order of Compilation Issues
7694 @section Order of Compilation Issues
7697 If, in our earlier example, there was a spec for the @code{hello}
7698 procedure, it would be contained in the file @file{hello.ads}; yet this
7699 file would not have to be explicitly compiled. This is the result of the
7700 model we chose to implement library management. Some of the consequences
7701 of this model are as follows:
7705 There is no point in compiling specs (except for package
7706 specs with no bodies) because these are compiled as needed by clients. If
7707 you attempt a useless compilation, you will receive an error message.
7708 It is also useless to compile subunits because they are compiled as needed
7712 There are no order of compilation requirements: performing a
7713 compilation never obsoletes anything. The only way you can obsolete
7714 something and require recompilations is to modify one of the
7715 source files on which it depends.
7718 There is no library as such, apart from the ALI files
7719 (@pxref{The Ada Library Information Files}, for information on the format
7720 of these files). For now we find it convenient to create separate ALI files,
7721 but eventually the information therein may be incorporated into the object
7725 When you compile a unit, the source files for the specs of all units
7726 that it @code{with}'s, all its subunits, and the bodies of any generics it
7727 instantiates must be available (reachable by the search-paths mechanism
7728 described above), or you will receive a fatal error message.
7735 The following are some typical Ada compilation command line examples:
7738 @item $ gcc -c xyz.adb
7739 Compile body in file @file{xyz.adb} with all default options.
7742 @item $ gcc -c -O2 -gnata xyz-def.adb
7745 @item $ GNAT COMPILE /OPTIMIZE=ALL -gnata xyz-def.adb
7748 Compile the child unit package in file @file{xyz-def.adb} with extensive
7749 optimizations, and pragma @code{Assert}/@code{Debug} statements
7752 @item $ gcc -c -gnatc abc-def.adb
7753 Compile the subunit in file @file{abc-def.adb} in semantic-checking-only
7757 @node Binding Using gnatbind
7758 @chapter Binding Using @code{gnatbind}
7762 * Running gnatbind::
7763 * Switches for gnatbind::
7764 * Command-Line Access::
7765 * Search Paths for gnatbind::
7766 * Examples of gnatbind Usage::
7770 This chapter describes the GNAT binder, @code{gnatbind}, which is used
7771 to bind compiled GNAT objects.
7773 Note: to invoke @code{gnatbind} with a project file, use the @code{gnat}
7774 driver (see @ref{The GNAT Driver and Project Files}).
7776 The @code{gnatbind} program performs four separate functions:
7780 Checks that a program is consistent, in accordance with the rules in
7781 Chapter 10 of the Ada Reference Manual. In particular, error
7782 messages are generated if a program uses inconsistent versions of a
7786 Checks that an acceptable order of elaboration exists for the program
7787 and issues an error message if it cannot find an order of elaboration
7788 that satisfies the rules in Chapter 10 of the Ada Language Manual.
7791 Generates a main program incorporating the given elaboration order.
7792 This program is a small Ada package (body and spec) that
7793 must be subsequently compiled
7794 using the GNAT compiler. The necessary compilation step is usually
7795 performed automatically by @command{gnatlink}. The two most important
7796 functions of this program
7797 are to call the elaboration routines of units in an appropriate order
7798 and to call the main program.
7801 Determines the set of object files required by the given main program.
7802 This information is output in the forms of comments in the generated program,
7803 to be read by the @command{gnatlink} utility used to link the Ada application.
7806 @node Running gnatbind
7807 @section Running @code{gnatbind}
7810 The form of the @code{gnatbind} command is
7813 $ gnatbind @ovar{switches} @var{mainprog}@r{[}.ali@r{]} @ovar{switches}
7817 where @file{@var{mainprog}.adb} is the Ada file containing the main program
7818 unit body. If no switches are specified, @code{gnatbind} constructs an Ada
7819 package in two files whose names are
7820 @file{b~@var{mainprog}.ads}, and @file{b~@var{mainprog}.adb}.
7821 For example, if given the
7822 parameter @file{hello.ali}, for a main program contained in file
7823 @file{hello.adb}, the binder output files would be @file{b~hello.ads}
7824 and @file{b~hello.adb}.
7826 When doing consistency checking, the binder takes into consideration
7827 any source files it can locate. For example, if the binder determines
7828 that the given main program requires the package @code{Pack}, whose
7830 file is @file{pack.ali} and whose corresponding source spec file is
7831 @file{pack.ads}, it attempts to locate the source file @file{pack.ads}
7832 (using the same search path conventions as previously described for the
7833 @command{gcc} command). If it can locate this source file, it checks that
7835 or source checksums of the source and its references to in @file{ALI} files
7836 match. In other words, any @file{ALI} files that mentions this spec must have
7837 resulted from compiling this version of the source file (or in the case
7838 where the source checksums match, a version close enough that the
7839 difference does not matter).
7841 @cindex Source files, use by binder
7842 The effect of this consistency checking, which includes source files, is
7843 that the binder ensures that the program is consistent with the latest
7844 version of the source files that can be located at bind time. Editing a
7845 source file without compiling files that depend on the source file cause
7846 error messages to be generated by the binder.
7848 For example, suppose you have a main program @file{hello.adb} and a
7849 package @code{P}, from file @file{p.ads} and you perform the following
7854 Enter @code{gcc -c hello.adb} to compile the main program.
7857 Enter @code{gcc -c p.ads} to compile package @code{P}.
7860 Edit file @file{p.ads}.
7863 Enter @code{gnatbind hello}.
7867 At this point, the file @file{p.ali} contains an out-of-date time stamp
7868 because the file @file{p.ads} has been edited. The attempt at binding
7869 fails, and the binder generates the following error messages:
7872 error: "hello.adb" must be recompiled ("p.ads" has been modified)
7873 error: "p.ads" has been modified and must be recompiled
7877 Now both files must be recompiled as indicated, and then the bind can
7878 succeed, generating a main program. You need not normally be concerned
7879 with the contents of this file, but for reference purposes a sample
7880 binder output file is given in @ref{Example of Binder Output File}.
7882 In most normal usage, the default mode of @command{gnatbind} which is to
7883 generate the main package in Ada, as described in the previous section.
7884 In particular, this means that any Ada programmer can read and understand
7885 the generated main program. It can also be debugged just like any other
7886 Ada code provided the @option{^-g^/DEBUG^} switch is used for
7887 @command{gnatbind} and @command{gnatlink}.
7889 However for some purposes it may be convenient to generate the main
7890 program in C rather than Ada. This may for example be helpful when you
7891 are generating a mixed language program with the main program in C. The
7892 GNAT compiler itself is an example.
7893 The use of the @option{^-C^/BIND_FILE=C^} switch
7894 for both @code{gnatbind} and @command{gnatlink} will cause the program to
7895 be generated in C (and compiled using the gnu C compiler).
7897 @node Switches for gnatbind
7898 @section Switches for @command{gnatbind}
7901 The following switches are available with @code{gnatbind}; details will
7902 be presented in subsequent sections.
7905 * Consistency-Checking Modes::
7906 * Binder Error Message Control::
7907 * Elaboration Control::
7909 * Binding with Non-Ada Main Programs::
7910 * Binding Programs with No Main Subprogram::
7917 @cindex @option{--version} @command{gnatbind}
7918 Display Copyright and version, then exit disregarding all other options.
7921 @cindex @option{--help} @command{gnatbind}
7922 If @option{--version} was not used, display usage, then exit disregarding
7926 @cindex @option{-a} @command{gnatbind}
7927 Indicates that, if supported by the platform, the adainit procedure should
7928 be treated as an initialisation routine by the linker (a constructor). This
7929 is intended to be used by the Project Manager to automatically initialize
7930 shared Stand-Alone Libraries.
7932 @item ^-aO^/OBJECT_SEARCH^
7933 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatbind})
7934 Specify directory to be searched for ALI files.
7936 @item ^-aI^/SOURCE_SEARCH^
7937 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
7938 Specify directory to be searched for source file.
7940 @item ^-A^/BIND_FILE=ADA^
7941 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatbind})
7942 Generate binder program in Ada (default)
7944 @item ^-b^/REPORT_ERRORS=BRIEF^
7945 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@command{gnatbind})
7946 Generate brief messages to @file{stderr} even if verbose mode set.
7948 @item ^-c^/NOOUTPUT^
7949 @cindex @option{^-c^/NOOUTPUT^} (@command{gnatbind})
7950 Check only, no generation of binder output file.
7952 @item ^-C^/BIND_FILE=C^
7953 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatbind})
7954 Generate binder program in C
7956 @item ^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
7957 @cindex @option{^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}} (@command{gnatbind})
7958 This switch can be used to change the default task stack size value
7959 to a specified size @var{nn}, which is expressed in bytes by default, or
7960 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
7962 In the absence of a @samp{@r{[}k@r{|}m@r{]}} suffix, this switch is equivalent,
7963 in effect, to completing all task specs with
7964 @smallexample @c ada
7965 pragma Storage_Size (nn);
7967 When they do not already have such a pragma.
7969 @item ^-D^/DEFAULT_SECONDARY_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
7970 @cindex @option{^-D^/DEFAULT_SECONDARY_STACK_SIZE=nnnnn^} (@command{gnatbind})
7971 This switch can be used to change the default secondary stack size value
7972 to a specified size @var{nn}, which is expressed in bytes by default, or
7973 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
7976 The secondary stack is used to deal with functions that return a variable
7977 sized result, for example a function returning an unconstrained
7978 String. There are two ways in which this secondary stack is allocated.
7980 For most targets, the secondary stack is growing on demand and is allocated
7981 as a chain of blocks in the heap. The -D option is not very
7982 relevant. It only give some control over the size of the allocated
7983 blocks (whose size is the minimum of the default secondary stack size value,
7984 and the actual size needed for the current allocation request).
7986 For certain targets, notably VxWorks 653,
7987 the secondary stack is allocated by carving off a fixed ratio chunk of the
7988 primary task stack. The -D option is used to define the
7989 size of the environment task's secondary stack.
7991 @item ^-e^/ELABORATION_DEPENDENCIES^
7992 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@command{gnatbind})
7993 Output complete list of elaboration-order dependencies.
7995 @item ^-E^/STORE_TRACEBACKS^
7996 @cindex @option{^-E^/STORE_TRACEBACKS^} (@command{gnatbind})
7997 Store tracebacks in exception occurrences when the target supports it.
7998 This is the default with the zero cost exception mechanism.
8000 @c The following may get moved to an appendix
8001 This option is currently supported on the following targets:
8002 all x86 ports, Solaris, Windows, HP-UX, AIX, PowerPC VxWorks and Alpha VxWorks.
8004 See also the packages @code{GNAT.Traceback} and
8005 @code{GNAT.Traceback.Symbolic} for more information.
8007 Note that on x86 ports, you must not use @option{-fomit-frame-pointer}
8008 @command{gcc} option.
8011 @item ^-F^/FORCE_ELABS_FLAGS^
8012 @cindex @option{^-F^/FORCE_ELABS_FLAGS^} (@command{gnatbind})
8013 Force the checks of elaboration flags. @command{gnatbind} does not normally
8014 generate checks of elaboration flags for the main executable, except when
8015 a Stand-Alone Library is used. However, there are cases when this cannot be
8016 detected by gnatbind. An example is importing an interface of a Stand-Alone
8017 Library through a pragma Import and only specifying through a linker switch
8018 this Stand-Alone Library. This switch is used to guarantee that elaboration
8019 flag checks are generated.
8022 @cindex @option{^-h^/HELP^} (@command{gnatbind})
8023 Output usage (help) information
8026 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8027 Specify directory to be searched for source and ALI files.
8029 @item ^-I-^/NOCURRENT_DIRECTORY^
8030 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatbind})
8031 Do not look for sources in the current directory where @code{gnatbind} was
8032 invoked, and do not look for ALI files in the directory containing the
8033 ALI file named in the @code{gnatbind} command line.
8035 @item ^-l^/ORDER_OF_ELABORATION^
8036 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@command{gnatbind})
8037 Output chosen elaboration order.
8039 @item ^-L@var{xxx}^/BUILD_LIBRARY=@var{xxx}^
8040 @cindex @option{^-L^/BUILD_LIBRARY^} (@command{gnatbind})
8041 Bind the units for library building. In this case the adainit and
8042 adafinal procedures (@pxref{Binding with Non-Ada Main Programs})
8043 are renamed to ^@var{xxx}init^@var{XXX}INIT^ and
8044 ^@var{xxx}final^@var{XXX}FINAL^.
8045 Implies ^-n^/NOCOMPILE^.
8047 (@xref{GNAT and Libraries}, for more details.)
8050 On OpenVMS, these init and final procedures are exported in uppercase
8051 letters. For example if /BUILD_LIBRARY=toto is used, the exported name of
8052 the init procedure will be "TOTOINIT" and the exported name of the final
8053 procedure will be "TOTOFINAL".
8056 @item ^-Mxyz^/RENAME_MAIN=xyz^
8057 @cindex @option{^-M^/RENAME_MAIN^} (@command{gnatbind})
8058 Rename generated main program from main to xyz. This option is
8059 supported on cross environments only.
8061 @item ^-m^/ERROR_LIMIT=^@var{n}
8062 @cindex @option{^-m^/ERROR_LIMIT^} (@command{gnatbind})
8063 Limit number of detected errors or warnings to @var{n}, where @var{n} is
8064 in the range 1..999999. The default value if no switch is
8065 given is 9999. If the number of warnings reaches this limit, then a
8066 message is output and further warnings are suppressed, the bind
8067 continues in this case. If the number of errors reaches this
8068 limit, then a message is output and the bind is abandoned.
8069 A value of zero means that no limit is enforced. The equal
8073 Furthermore, under Windows, the sources pointed to by the libraries path
8074 set in the registry are not searched for.
8078 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8082 @cindex @option{-nostdinc} (@command{gnatbind})
8083 Do not look for sources in the system default directory.
8086 @cindex @option{-nostdlib} (@command{gnatbind})
8087 Do not look for library files in the system default directory.
8089 @item --RTS=@var{rts-path}
8090 @cindex @option{--RTS} (@code{gnatbind})
8091 Specifies the default location of the runtime library. Same meaning as the
8092 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
8094 @item ^-o ^/OUTPUT=^@var{file}
8095 @cindex @option{^-o ^/OUTPUT^} (@command{gnatbind})
8096 Name the output file @var{file} (default is @file{b~@var{xxx}.adb}).
8097 Note that if this option is used, then linking must be done manually,
8098 gnatlink cannot be used.
8100 @item ^-O^/OBJECT_LIST^
8101 @cindex @option{^-O^/OBJECT_LIST^} (@command{gnatbind})
8104 @item ^-p^/PESSIMISTIC_ELABORATION^
8105 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@command{gnatbind})
8106 Pessimistic (worst-case) elaboration order
8109 @cindex @option{^-R^-R^} (@command{gnatbind})
8110 Output closure source list.
8112 @item ^-s^/READ_SOURCES=ALL^
8113 @cindex @option{^-s^/READ_SOURCES=ALL^} (@command{gnatbind})
8114 Require all source files to be present.
8116 @item ^-S@var{xxx}^/INITIALIZE_SCALARS=@var{xxx}^
8117 @cindex @option{^-S^/INITIALIZE_SCALARS^} (@command{gnatbind})
8118 Specifies the value to be used when detecting uninitialized scalar
8119 objects with pragma Initialize_Scalars.
8120 The @var{xxx} ^string specified with the switch^option^ may be either
8122 @item ``@option{^in^INVALID^}'' requesting an invalid value where possible
8123 @item ``@option{^lo^LOW^}'' for the lowest possible value
8124 @item ``@option{^hi^HIGH^}'' for the highest possible value
8125 @item ``@option{@var{xx}}'' for a value consisting of repeated bytes with the
8126 value @code{16#@var{xx}#} (i.e., @var{xx} is a string of two hexadecimal digits).
8129 In addition, you can specify @option{-Sev} to indicate that the value is
8130 to be set at run time. In this case, the program will look for an environment
8131 @cindex GNAT_INIT_SCALARS
8132 variable of the form @env{GNAT_INIT_SCALARS=@var{xx}}, where @var{xx} is one
8133 of @option{in/lo/hi/@var{xx}} with the same meanings as above.
8134 If no environment variable is found, or if it does not have a valid value,
8135 then the default is @option{in} (invalid values).
8139 @cindex @option{-static} (@code{gnatbind})
8140 Link against a static GNAT run time.
8143 @cindex @option{-shared} (@code{gnatbind})
8144 Link against a shared GNAT run time when available.
8147 @item ^-t^/NOTIME_STAMP_CHECK^
8148 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8149 Tolerate time stamp and other consistency errors
8151 @item ^-T@var{n}^/TIME_SLICE=@var{n}^
8152 @cindex @option{^-T^/TIME_SLICE^} (@code{gnatbind})
8153 Set the time slice value to @var{n} milliseconds. If the system supports
8154 the specification of a specific time slice value, then the indicated value
8155 is used. If the system does not support specific time slice values, but
8156 does support some general notion of round-robin scheduling, then any
8157 nonzero value will activate round-robin scheduling.
8159 A value of zero is treated specially. It turns off time
8160 slicing, and in addition, indicates to the tasking run time that the
8161 semantics should match as closely as possible the Annex D
8162 requirements of the Ada RM, and in particular sets the default
8163 scheduling policy to @code{FIFO_Within_Priorities}.
8165 @item ^-u@var{n}^/DYNAMIC_STACK_USAGE=@var{n}^
8166 @cindex @option{^-u^/DYNAMIC_STACK_USAGE^} (@code{gnatbind})
8167 Enable dynamic stack usage, with @var{n} results stored and displayed
8168 at program termination. A result is generated when a task
8169 terminates. Results that can't be stored are displayed on the fly, at
8170 task termination. This option is currently not supported on Itanium
8171 platforms. (See @ref{Dynamic Stack Usage Analysis} for details.)
8173 @item ^-v^/REPORT_ERRORS=VERBOSE^
8174 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8175 Verbose mode. Write error messages, header, summary output to
8180 @cindex @option{-w} (@code{gnatbind})
8181 Warning mode (@var{x}=s/e for suppress/treat as error)
8185 @item /WARNINGS=NORMAL
8186 @cindex @option{/WARNINGS} (@code{gnatbind})
8187 Normal warnings mode. Warnings are issued but ignored
8189 @item /WARNINGS=SUPPRESS
8190 @cindex @option{/WARNINGS} (@code{gnatbind})
8191 All warning messages are suppressed
8193 @item /WARNINGS=ERROR
8194 @cindex @option{/WARNINGS} (@code{gnatbind})
8195 Warning messages are treated as fatal errors
8198 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8199 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8200 Override default wide character encoding for standard Text_IO files.
8202 @item ^-x^/READ_SOURCES=NONE^
8203 @cindex @option{^-x^/READ_SOURCES^} (@code{gnatbind})
8204 Exclude source files (check object consistency only).
8207 @item /READ_SOURCES=AVAILABLE
8208 @cindex @option{/READ_SOURCES} (@code{gnatbind})
8209 Default mode, in which sources are checked for consistency only if
8213 @item ^-y^/ENABLE_LEAP_SECONDS^
8214 @cindex @option{^-y^/ENABLE_LEAP_SECONDS^} (@code{gnatbind})
8215 Enable leap seconds support in @code{Ada.Calendar} and its children.
8217 @item ^-z^/ZERO_MAIN^
8218 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8224 You may obtain this listing of switches by running @code{gnatbind} with
8228 @node Consistency-Checking Modes
8229 @subsection Consistency-Checking Modes
8232 As described earlier, by default @code{gnatbind} checks
8233 that object files are consistent with one another and are consistent
8234 with any source files it can locate. The following switches control binder
8239 @item ^-s^/READ_SOURCES=ALL^
8240 @cindex @option{^-s^/READ_SOURCES=ALL^} (@code{gnatbind})
8241 Require source files to be present. In this mode, the binder must be
8242 able to locate all source files that are referenced, in order to check
8243 their consistency. In normal mode, if a source file cannot be located it
8244 is simply ignored. If you specify this switch, a missing source
8247 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8248 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8249 Override default wide character encoding for standard Text_IO files.
8250 Normally the default wide character encoding method used for standard
8251 [Wide_[Wide_]]Text_IO files is taken from the encoding specified for
8252 the main source input (see description of switch
8253 @option{^-gnatWx^/WIDE_CHARACTER_ENCODING^} for the compiler). The
8254 use of this switch for the binder (which has the same set of
8255 possible arguments) overrides this default as specified.
8257 @item ^-x^/READ_SOURCES=NONE^
8258 @cindex @option{^-x^/READ_SOURCES=NONE^} (@code{gnatbind})
8259 Exclude source files. In this mode, the binder only checks that ALI
8260 files are consistent with one another. Source files are not accessed.
8261 The binder runs faster in this mode, and there is still a guarantee that
8262 the resulting program is self-consistent.
8263 If a source file has been edited since it was last compiled, and you
8264 specify this switch, the binder will not detect that the object
8265 file is out of date with respect to the source file. Note that this is the
8266 mode that is automatically used by @command{gnatmake} because in this
8267 case the checking against sources has already been performed by
8268 @command{gnatmake} in the course of compilation (i.e.@: before binding).
8271 @item /READ_SOURCES=AVAILABLE
8272 @cindex @code{/READ_SOURCES=AVAILABLE} (@code{gnatbind})
8273 This is the default mode in which source files are checked if they are
8274 available, and ignored if they are not available.
8278 @node Binder Error Message Control
8279 @subsection Binder Error Message Control
8282 The following switches provide control over the generation of error
8283 messages from the binder:
8287 @item ^-v^/REPORT_ERRORS=VERBOSE^
8288 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8289 Verbose mode. In the normal mode, brief error messages are generated to
8290 @file{stderr}. If this switch is present, a header is written
8291 to @file{stdout} and any error messages are directed to @file{stdout}.
8292 All that is written to @file{stderr} is a brief summary message.
8294 @item ^-b^/REPORT_ERRORS=BRIEF^
8295 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@code{gnatbind})
8296 Generate brief error messages to @file{stderr} even if verbose mode is
8297 specified. This is relevant only when used with the
8298 @option{^-v^/REPORT_ERRORS=VERBOSE^} switch.
8302 @cindex @option{-m} (@code{gnatbind})
8303 Limits the number of error messages to @var{n}, a decimal integer in the
8304 range 1-999. The binder terminates immediately if this limit is reached.
8307 @cindex @option{-M} (@code{gnatbind})
8308 Renames the generated main program from @code{main} to @code{xxx}.
8309 This is useful in the case of some cross-building environments, where
8310 the actual main program is separate from the one generated
8314 @item ^-ws^/WARNINGS=SUPPRESS^
8315 @cindex @option{^-ws^/WARNINGS=SUPPRESS^} (@code{gnatbind})
8317 Suppress all warning messages.
8319 @item ^-we^/WARNINGS=ERROR^
8320 @cindex @option{^-we^/WARNINGS=ERROR^} (@code{gnatbind})
8321 Treat any warning messages as fatal errors.
8324 @item /WARNINGS=NORMAL
8325 Standard mode with warnings generated, but warnings do not get treated
8329 @item ^-t^/NOTIME_STAMP_CHECK^
8330 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8331 @cindex Time stamp checks, in binder
8332 @cindex Binder consistency checks
8333 @cindex Consistency checks, in binder
8334 The binder performs a number of consistency checks including:
8338 Check that time stamps of a given source unit are consistent
8340 Check that checksums of a given source unit are consistent
8342 Check that consistent versions of @code{GNAT} were used for compilation
8344 Check consistency of configuration pragmas as required
8348 Normally failure of such checks, in accordance with the consistency
8349 requirements of the Ada Reference Manual, causes error messages to be
8350 generated which abort the binder and prevent the output of a binder
8351 file and subsequent link to obtain an executable.
8353 The @option{^-t^/NOTIME_STAMP_CHECK^} switch converts these error messages
8354 into warnings, so that
8355 binding and linking can continue to completion even in the presence of such
8356 errors. The result may be a failed link (due to missing symbols), or a
8357 non-functional executable which has undefined semantics.
8358 @emph{This means that
8359 @option{^-t^/NOTIME_STAMP_CHECK^} should be used only in unusual situations,
8363 @node Elaboration Control
8364 @subsection Elaboration Control
8367 The following switches provide additional control over the elaboration
8368 order. For full details see @ref{Elaboration Order Handling in GNAT}.
8371 @item ^-p^/PESSIMISTIC_ELABORATION^
8372 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@code{gnatbind})
8373 Normally the binder attempts to choose an elaboration order that is
8374 likely to minimize the likelihood of an elaboration order error resulting
8375 in raising a @code{Program_Error} exception. This switch reverses the
8376 action of the binder, and requests that it deliberately choose an order
8377 that is likely to maximize the likelihood of an elaboration error.
8378 This is useful in ensuring portability and avoiding dependence on
8379 accidental fortuitous elaboration ordering.
8381 Normally it only makes sense to use the @option{^-p^/PESSIMISTIC_ELABORATION^}
8383 elaboration checking is used (@option{-gnatE} switch used for compilation).
8384 This is because in the default static elaboration mode, all necessary
8385 @code{Elaborate} and @code{Elaborate_All} pragmas are implicitly inserted.
8386 These implicit pragmas are still respected by the binder in
8387 @option{^-p^/PESSIMISTIC_ELABORATION^} mode, so a
8388 safe elaboration order is assured.
8391 @node Output Control
8392 @subsection Output Control
8395 The following switches allow additional control over the output
8396 generated by the binder.
8401 @item ^-A^/BIND_FILE=ADA^
8402 @cindex @option{^-A^/BIND_FILE=ADA^} (@code{gnatbind})
8403 Generate binder program in Ada (default). The binder program is named
8404 @file{b~@var{mainprog}.adb} by default. This can be changed with
8405 @option{^-o^/OUTPUT^} @code{gnatbind} option.
8407 @item ^-c^/NOOUTPUT^
8408 @cindex @option{^-c^/NOOUTPUT^} (@code{gnatbind})
8409 Check only. Do not generate the binder output file. In this mode the
8410 binder performs all error checks but does not generate an output file.
8412 @item ^-C^/BIND_FILE=C^
8413 @cindex @option{^-C^/BIND_FILE=C^} (@code{gnatbind})
8414 Generate binder program in C. The binder program is named
8415 @file{b_@var{mainprog}.c}.
8416 This can be changed with @option{^-o^/OUTPUT^} @code{gnatbind}
8419 @item ^-e^/ELABORATION_DEPENDENCIES^
8420 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@code{gnatbind})
8421 Output complete list of elaboration-order dependencies, showing the
8422 reason for each dependency. This output can be rather extensive but may
8423 be useful in diagnosing problems with elaboration order. The output is
8424 written to @file{stdout}.
8427 @cindex @option{^-h^/HELP^} (@code{gnatbind})
8428 Output usage information. The output is written to @file{stdout}.
8430 @item ^-K^/LINKER_OPTION_LIST^
8431 @cindex @option{^-K^/LINKER_OPTION_LIST^} (@code{gnatbind})
8432 Output linker options to @file{stdout}. Includes library search paths,
8433 contents of pragmas Ident and Linker_Options, and libraries added
8436 @item ^-l^/ORDER_OF_ELABORATION^
8437 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@code{gnatbind})
8438 Output chosen elaboration order. The output is written to @file{stdout}.
8440 @item ^-O^/OBJECT_LIST^
8441 @cindex @option{^-O^/OBJECT_LIST^} (@code{gnatbind})
8442 Output full names of all the object files that must be linked to provide
8443 the Ada component of the program. The output is written to @file{stdout}.
8444 This list includes the files explicitly supplied and referenced by the user
8445 as well as implicitly referenced run-time unit files. The latter are
8446 omitted if the corresponding units reside in shared libraries. The
8447 directory names for the run-time units depend on the system configuration.
8449 @item ^-o ^/OUTPUT=^@var{file}
8450 @cindex @option{^-o^/OUTPUT^} (@code{gnatbind})
8451 Set name of output file to @var{file} instead of the normal
8452 @file{b~@var{mainprog}.adb} default. Note that @var{file} denote the Ada
8453 binder generated body filename. In C mode you would normally give
8454 @var{file} an extension of @file{.c} because it will be a C source program.
8455 Note that if this option is used, then linking must be done manually.
8456 It is not possible to use gnatlink in this case, since it cannot locate
8459 @item ^-r^/RESTRICTION_LIST^
8460 @cindex @option{^-r^/RESTRICTION_LIST^} (@code{gnatbind})
8461 Generate list of @code{pragma Restrictions} that could be applied to
8462 the current unit. This is useful for code audit purposes, and also may
8463 be used to improve code generation in some cases.
8467 @node Binding with Non-Ada Main Programs
8468 @subsection Binding with Non-Ada Main Programs
8471 In our description so far we have assumed that the main
8472 program is in Ada, and that the task of the binder is to generate a
8473 corresponding function @code{main} that invokes this Ada main
8474 program. GNAT also supports the building of executable programs where
8475 the main program is not in Ada, but some of the called routines are
8476 written in Ada and compiled using GNAT (@pxref{Mixed Language Programming}).
8477 The following switch is used in this situation:
8481 @cindex @option{^-n^/NOMAIN^} (@code{gnatbind})
8482 No main program. The main program is not in Ada.
8486 In this case, most of the functions of the binder are still required,
8487 but instead of generating a main program, the binder generates a file
8488 containing the following callable routines:
8493 You must call this routine to initialize the Ada part of the program by
8494 calling the necessary elaboration routines. A call to @code{adainit} is
8495 required before the first call to an Ada subprogram.
8497 Note that it is assumed that the basic execution environment must be setup
8498 to be appropriate for Ada execution at the point where the first Ada
8499 subprogram is called. In particular, if the Ada code will do any
8500 floating-point operations, then the FPU must be setup in an appropriate
8501 manner. For the case of the x86, for example, full precision mode is
8502 required. The procedure GNAT.Float_Control.Reset may be used to ensure
8503 that the FPU is in the right state.
8507 You must call this routine to perform any library-level finalization
8508 required by the Ada subprograms. A call to @code{adafinal} is required
8509 after the last call to an Ada subprogram, and before the program
8514 If the @option{^-n^/NOMAIN^} switch
8515 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8516 @cindex Binder, multiple input files
8517 is given, more than one ALI file may appear on
8518 the command line for @code{gnatbind}. The normal @dfn{closure}
8519 calculation is performed for each of the specified units. Calculating
8520 the closure means finding out the set of units involved by tracing
8521 @code{with} references. The reason it is necessary to be able to
8522 specify more than one ALI file is that a given program may invoke two or
8523 more quite separate groups of Ada units.
8525 The binder takes the name of its output file from the last specified ALI
8526 file, unless overridden by the use of the @option{^-o file^/OUTPUT=file^}.
8527 @cindex @option{^-o^/OUTPUT^} (@command{gnatbind})
8528 The output is an Ada unit in source form that can
8529 be compiled with GNAT unless the -C switch is used in which case the
8530 output is a C source file, which must be compiled using the C compiler.
8531 This compilation occurs automatically as part of the @command{gnatlink}
8534 Currently the GNAT run time requires a FPU using 80 bits mode
8535 precision. Under targets where this is not the default it is required to
8536 call GNAT.Float_Control.Reset before using floating point numbers (this
8537 include float computation, float input and output) in the Ada code. A
8538 side effect is that this could be the wrong mode for the foreign code
8539 where floating point computation could be broken after this call.
8541 @node Binding Programs with No Main Subprogram
8542 @subsection Binding Programs with No Main Subprogram
8545 It is possible to have an Ada program which does not have a main
8546 subprogram. This program will call the elaboration routines of all the
8547 packages, then the finalization routines.
8549 The following switch is used to bind programs organized in this manner:
8552 @item ^-z^/ZERO_MAIN^
8553 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8554 Normally the binder checks that the unit name given on the command line
8555 corresponds to a suitable main subprogram. When this switch is used,
8556 a list of ALI files can be given, and the execution of the program
8557 consists of elaboration of these units in an appropriate order. Note
8558 that the default wide character encoding method for standard Text_IO
8559 files is always set to Brackets if this switch is set (you can use
8561 @option{^-Wx^WIDE_CHARACTER_ENCODING^} to override this default).
8564 @node Command-Line Access
8565 @section Command-Line Access
8568 The package @code{Ada.Command_Line} provides access to the command-line
8569 arguments and program name. In order for this interface to operate
8570 correctly, the two variables
8582 are declared in one of the GNAT library routines. These variables must
8583 be set from the actual @code{argc} and @code{argv} values passed to the
8584 main program. With no @option{^n^/NOMAIN^} present, @code{gnatbind}
8585 generates the C main program to automatically set these variables.
8586 If the @option{^n^/NOMAIN^} switch is used, there is no automatic way to
8587 set these variables. If they are not set, the procedures in
8588 @code{Ada.Command_Line} will not be available, and any attempt to use
8589 them will raise @code{Constraint_Error}. If command line access is
8590 required, your main program must set @code{gnat_argc} and
8591 @code{gnat_argv} from the @code{argc} and @code{argv} values passed to
8594 @node Search Paths for gnatbind
8595 @section Search Paths for @code{gnatbind}
8598 The binder takes the name of an ALI file as its argument and needs to
8599 locate source files as well as other ALI files to verify object consistency.
8601 For source files, it follows exactly the same search rules as @command{gcc}
8602 (@pxref{Search Paths and the Run-Time Library (RTL)}). For ALI files the
8603 directories searched are:
8607 The directory containing the ALI file named in the command line, unless
8608 the switch @option{^-I-^/NOCURRENT_DIRECTORY^} is specified.
8611 All directories specified by @option{^-I^/SEARCH^}
8612 switches on the @code{gnatbind}
8613 command line, in the order given.
8616 @findex ADA_PRJ_OBJECTS_FILE
8617 Each of the directories listed in the text file whose name is given
8618 by the @env{ADA_PRJ_OBJECTS_FILE} ^environment variable^logical name^.
8621 @env{ADA_PRJ_OBJECTS_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
8622 driver when project files are used. It should not normally be set
8626 @findex ADA_OBJECTS_PATH
8627 Each of the directories listed in the value of the
8628 @env{ADA_OBJECTS_PATH} ^environment variable^logical name^.
8630 Construct this value
8631 exactly as the @env{PATH} environment variable: a list of directory
8632 names separated by colons (semicolons when working with the NT version
8636 Normally, define this value as a logical name containing a comma separated
8637 list of directory names.
8639 This variable can also be defined by means of an environment string
8640 (an argument to the HP C exec* set of functions).
8644 DEFINE ANOTHER_PATH FOO:[BAG]
8645 DEFINE ADA_OBJECTS_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
8648 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
8649 first, followed by the standard Ada
8650 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADALIB].
8651 If this is not redefined, the user will obtain the HP Ada 83 IO packages
8652 (Text_IO, Sequential_IO, etc)
8653 instead of the standard Ada packages. Thus, in order to get the standard Ada
8654 packages by default, ADA_OBJECTS_PATH must be redefined.
8658 The content of the @file{ada_object_path} file which is part of the GNAT
8659 installation tree and is used to store standard libraries such as the
8660 GNAT Run Time Library (RTL) unless the switch @option{-nostdlib} is
8663 @ref{Installing a library}
8668 In the binder the switch @option{^-I^/SEARCH^}
8669 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8670 is used to specify both source and
8671 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
8672 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
8673 instead if you want to specify
8674 source paths only, and @option{^-aO^/LIBRARY_SEARCH^}
8675 @cindex @option{^-aO^/LIBRARY_SEARCH^} (@command{gnatbind})
8676 if you want to specify library paths
8677 only. This means that for the binder
8678 @option{^-I^/SEARCH=^}@var{dir} is equivalent to
8679 @option{^-aI^/SOURCE_SEARCH=^}@var{dir}
8680 @option{^-aO^/OBJECT_SEARCH=^}@var{dir}.
8681 The binder generates the bind file (a C language source file) in the
8682 current working directory.
8688 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
8689 children make up the GNAT Run-Time Library, together with the package
8690 GNAT and its children, which contain a set of useful additional
8691 library functions provided by GNAT. The sources for these units are
8692 needed by the compiler and are kept together in one directory. The ALI
8693 files and object files generated by compiling the RTL are needed by the
8694 binder and the linker and are kept together in one directory, typically
8695 different from the directory containing the sources. In a normal
8696 installation, you need not specify these directory names when compiling
8697 or binding. Either the environment variables or the built-in defaults
8698 cause these files to be found.
8700 Besides simplifying access to the RTL, a major use of search paths is
8701 in compiling sources from multiple directories. This can make
8702 development environments much more flexible.
8704 @node Examples of gnatbind Usage
8705 @section Examples of @code{gnatbind} Usage
8708 This section contains a number of examples of using the GNAT binding
8709 utility @code{gnatbind}.
8712 @item gnatbind hello
8713 The main program @code{Hello} (source program in @file{hello.adb}) is
8714 bound using the standard switch settings. The generated main program is
8715 @file{b~hello.adb}. This is the normal, default use of the binder.
8718 @item gnatbind hello -o mainprog.adb
8721 @item gnatbind HELLO.ALI /OUTPUT=Mainprog.ADB
8723 The main program @code{Hello} (source program in @file{hello.adb}) is
8724 bound using the standard switch settings. The generated main program is
8725 @file{mainprog.adb} with the associated spec in
8726 @file{mainprog.ads}. Note that you must specify the body here not the
8727 spec, in the case where the output is in Ada. Note that if this option
8728 is used, then linking must be done manually, since gnatlink will not
8729 be able to find the generated file.
8732 @item gnatbind main -C -o mainprog.c -x
8735 @item gnatbind MAIN.ALI /BIND_FILE=C /OUTPUT=Mainprog.C /READ_SOURCES=NONE
8737 The main program @code{Main} (source program in
8738 @file{main.adb}) is bound, excluding source files from the
8739 consistency checking, generating
8740 the file @file{mainprog.c}.
8743 @item gnatbind -x main_program -C -o mainprog.c
8744 This command is exactly the same as the previous example. Switches may
8745 appear anywhere in the command line, and single letter switches may be
8746 combined into a single switch.
8750 @item gnatbind -n math dbase -C -o ada-control.c
8753 @item gnatbind /NOMAIN math dbase /BIND_FILE=C /OUTPUT=ada-control.c
8755 The main program is in a language other than Ada, but calls to
8756 subprograms in packages @code{Math} and @code{Dbase} appear. This call
8757 to @code{gnatbind} generates the file @file{ada-control.c} containing
8758 the @code{adainit} and @code{adafinal} routines to be called before and
8759 after accessing the Ada units.
8762 @c ------------------------------------
8763 @node Linking Using gnatlink
8764 @chapter Linking Using @command{gnatlink}
8765 @c ------------------------------------
8769 This chapter discusses @command{gnatlink}, a tool that links
8770 an Ada program and builds an executable file. This utility
8771 invokes the system linker ^(via the @command{gcc} command)^^
8772 with a correct list of object files and library references.
8773 @command{gnatlink} automatically determines the list of files and
8774 references for the Ada part of a program. It uses the binder file
8775 generated by the @command{gnatbind} to determine this list.
8777 Note: to invoke @code{gnatlink} with a project file, use the @code{gnat}
8778 driver (see @ref{The GNAT Driver and Project Files}).
8781 * Running gnatlink::
8782 * Switches for gnatlink::
8785 @node Running gnatlink
8786 @section Running @command{gnatlink}
8789 The form of the @command{gnatlink} command is
8792 $ gnatlink @ovar{switches} @var{mainprog}@r{[}.ali@r{]}
8793 @ovar{non-Ada objects} @ovar{linker options}
8797 The arguments of @command{gnatlink} (switches, main @file{ALI} file,
8799 or linker options) may be in any order, provided that no non-Ada object may
8800 be mistaken for a main @file{ALI} file.
8801 Any file name @file{F} without the @file{.ali}
8802 extension will be taken as the main @file{ALI} file if a file exists
8803 whose name is the concatenation of @file{F} and @file{.ali}.
8806 @file{@var{mainprog}.ali} references the ALI file of the main program.
8807 The @file{.ali} extension of this file can be omitted. From this
8808 reference, @command{gnatlink} locates the corresponding binder file
8809 @file{b~@var{mainprog}.adb} and, using the information in this file along
8810 with the list of non-Ada objects and linker options, constructs a
8811 linker command file to create the executable.
8813 The arguments other than the @command{gnatlink} switches and the main
8814 @file{ALI} file are passed to the linker uninterpreted.
8815 They typically include the names of
8816 object files for units written in other languages than Ada and any library
8817 references required to resolve references in any of these foreign language
8818 units, or in @code{Import} pragmas in any Ada units.
8820 @var{linker options} is an optional list of linker specific
8822 The default linker called by gnatlink is @command{gcc} which in
8823 turn calls the appropriate system linker.
8824 Standard options for the linker such as @option{-lmy_lib} or
8825 @option{-Ldir} can be added as is.
8826 For options that are not recognized by
8827 @command{gcc} as linker options, use the @command{gcc} switches
8828 @option{-Xlinker} or @option{-Wl,}.
8829 Refer to the GCC documentation for
8830 details. Here is an example showing how to generate a linker map:
8833 $ ^gnatlink my_prog -Wl,-Map,MAPFILE^GNAT LINK my_prog.ali /MAP^
8836 Using @var{linker options} it is possible to set the program stack and
8839 See @ref{Setting Stack Size from gnatlink} and
8840 @ref{Setting Heap Size from gnatlink}.
8843 @command{gnatlink} determines the list of objects required by the Ada
8844 program and prepends them to the list of objects passed to the linker.
8845 @command{gnatlink} also gathers any arguments set by the use of
8846 @code{pragma Linker_Options} and adds them to the list of arguments
8847 presented to the linker.
8850 @command{gnatlink} accepts the following types of extra files on the command
8851 line: objects (@file{.OBJ}), libraries (@file{.OLB}), sharable images
8852 (@file{.EXE}), and options files (@file{.OPT}). These are recognized and
8853 handled according to their extension.
8856 @node Switches for gnatlink
8857 @section Switches for @command{gnatlink}
8860 The following switches are available with the @command{gnatlink} utility:
8866 @cindex @option{--version} @command{gnatlink}
8867 Display Copyright and version, then exit disregarding all other options.
8870 @cindex @option{--help} @command{gnatlink}
8871 If @option{--version} was not used, display usage, then exit disregarding
8874 @item ^-A^/BIND_FILE=ADA^
8875 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatlink})
8876 The binder has generated code in Ada. This is the default.
8878 @item ^-C^/BIND_FILE=C^
8879 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatlink})
8880 If instead of generating a file in Ada, the binder has generated one in
8881 C, then the linker needs to know about it. Use this switch to signal
8882 to @command{gnatlink} that the binder has generated C code rather than
8885 @item ^-f^/FORCE_OBJECT_FILE_LIST^
8886 @cindex Command line length
8887 @cindex @option{^-f^/FORCE_OBJECT_FILE_LIST^} (@command{gnatlink})
8888 On some targets, the command line length is limited, and @command{gnatlink}
8889 will generate a separate file for the linker if the list of object files
8891 The @option{^-f^/FORCE_OBJECT_FILE_LIST^} switch forces this file
8892 to be generated even if
8893 the limit is not exceeded. This is useful in some cases to deal with
8894 special situations where the command line length is exceeded.
8897 @cindex Debugging information, including
8898 @cindex @option{^-g^/DEBUG^} (@command{gnatlink})
8899 The option to include debugging information causes the Ada bind file (in
8900 other words, @file{b~@var{mainprog}.adb}) to be compiled with
8901 @option{^-g^/DEBUG^}.
8902 In addition, the binder does not delete the @file{b~@var{mainprog}.adb},
8903 @file{b~@var{mainprog}.o} and @file{b~@var{mainprog}.ali} files.
8904 Without @option{^-g^/DEBUG^}, the binder removes these files by
8905 default. The same procedure apply if a C bind file was generated using
8906 @option{^-C^/BIND_FILE=C^} @code{gnatbind} option, in this case the filenames
8907 are @file{b_@var{mainprog}.c} and @file{b_@var{mainprog}.o}.
8909 @item ^-n^/NOCOMPILE^
8910 @cindex @option{^-n^/NOCOMPILE^} (@command{gnatlink})
8911 Do not compile the file generated by the binder. This may be used when
8912 a link is rerun with different options, but there is no need to recompile
8916 @cindex @option{^-v^/VERBOSE^} (@command{gnatlink})
8917 Causes additional information to be output, including a full list of the
8918 included object files. This switch option is most useful when you want
8919 to see what set of object files are being used in the link step.
8921 @item ^-v -v^/VERBOSE/VERBOSE^
8922 @cindex @option{^-v -v^/VERBOSE/VERBOSE^} (@command{gnatlink})
8923 Very verbose mode. Requests that the compiler operate in verbose mode when
8924 it compiles the binder file, and that the system linker run in verbose mode.
8926 @item ^-o ^/EXECUTABLE=^@var{exec-name}
8927 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatlink})
8928 @var{exec-name} specifies an alternate name for the generated
8929 executable program. If this switch is omitted, the executable has the same
8930 name as the main unit. For example, @code{gnatlink try.ali} creates
8931 an executable called @file{^try^TRY.EXE^}.
8934 @item -b @var{target}
8935 @cindex @option{-b} (@command{gnatlink})
8936 Compile your program to run on @var{target}, which is the name of a
8937 system configuration. You must have a GNAT cross-compiler built if
8938 @var{target} is not the same as your host system.
8941 @cindex @option{-B} (@command{gnatlink})
8942 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
8943 from @var{dir} instead of the default location. Only use this switch
8944 when multiple versions of the GNAT compiler are available.
8945 @xref{Directory Options,,, gcc, The GNU Compiler Collection},
8946 for further details. You would normally use the @option{-b} or
8947 @option{-V} switch instead.
8949 @item --GCC=@var{compiler_name}
8950 @cindex @option{--GCC=compiler_name} (@command{gnatlink})
8951 Program used for compiling the binder file. The default is
8952 @command{gcc}. You need to use quotes around @var{compiler_name} if
8953 @code{compiler_name} contains spaces or other separator characters.
8954 As an example @option{--GCC="foo -x -y"} will instruct @command{gnatlink} to
8955 use @code{foo -x -y} as your compiler. Note that switch @option{-c} is always
8956 inserted after your command name. Thus in the above example the compiler
8957 command that will be used by @command{gnatlink} will be @code{foo -c -x -y}.
8958 A limitation of this syntax is that the name and path name of the executable
8959 itself must not include any embedded spaces. If the compiler executable is
8960 different from the default one (gcc or <prefix>-gcc), then the back-end
8961 switches in the ALI file are not used to compile the binder generated source.
8962 For example, this is the case with @option{--GCC="foo -x -y"}. But the back end
8963 switches will be used for @option{--GCC="gcc -gnatv"}. If several
8964 @option{--GCC=compiler_name} are used, only the last @var{compiler_name}
8965 is taken into account. However, all the additional switches are also taken
8967 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
8968 @option{--GCC="bar -x -y -z -t"}.
8970 @item --LINK=@var{name}
8971 @cindex @option{--LINK=} (@command{gnatlink})
8972 @var{name} is the name of the linker to be invoked. This is especially
8973 useful in mixed language programs since languages such as C++ require
8974 their own linker to be used. When this switch is omitted, the default
8975 name for the linker is @command{gcc}. When this switch is used, the
8976 specified linker is called instead of @command{gcc} with exactly the same
8977 parameters that would have been passed to @command{gcc} so if the desired
8978 linker requires different parameters it is necessary to use a wrapper
8979 script that massages the parameters before invoking the real linker. It
8980 may be useful to control the exact invocation by using the verbose
8986 @item /DEBUG=TRACEBACK
8987 @cindex @code{/DEBUG=TRACEBACK} (@command{gnatlink})
8988 This qualifier causes sufficient information to be included in the
8989 executable file to allow a traceback, but does not include the full
8990 symbol information needed by the debugger.
8992 @item /IDENTIFICATION="<string>"
8993 @code{"<string>"} specifies the string to be stored in the image file
8994 identification field in the image header.
8995 It overrides any pragma @code{Ident} specified string.
8997 @item /NOINHIBIT-EXEC
8998 Generate the executable file even if there are linker warnings.
9000 @item /NOSTART_FILES
9001 Don't link in the object file containing the ``main'' transfer address.
9002 Used when linking with a foreign language main program compiled with an
9006 Prefer linking with object libraries over sharable images, even without
9012 @node The GNAT Make Program gnatmake
9013 @chapter The GNAT Make Program @command{gnatmake}
9017 * Running gnatmake::
9018 * Switches for gnatmake::
9019 * Mode Switches for gnatmake::
9020 * Notes on the Command Line::
9021 * How gnatmake Works::
9022 * Examples of gnatmake Usage::
9025 A typical development cycle when working on an Ada program consists of
9026 the following steps:
9030 Edit some sources to fix bugs.
9036 Compile all sources affected.
9046 The third step can be tricky, because not only do the modified files
9047 @cindex Dependency rules
9048 have to be compiled, but any files depending on these files must also be
9049 recompiled. The dependency rules in Ada can be quite complex, especially
9050 in the presence of overloading, @code{use} clauses, generics and inlined
9053 @command{gnatmake} automatically takes care of the third and fourth steps
9054 of this process. It determines which sources need to be compiled,
9055 compiles them, and binds and links the resulting object files.
9057 Unlike some other Ada make programs, the dependencies are always
9058 accurately recomputed from the new sources. The source based approach of
9059 the GNAT compilation model makes this possible. This means that if
9060 changes to the source program cause corresponding changes in
9061 dependencies, they will always be tracked exactly correctly by
9064 @node Running gnatmake
9065 @section Running @command{gnatmake}
9068 The usual form of the @command{gnatmake} command is
9071 $ gnatmake @ovar{switches} @var{file_name}
9072 @ovar{file_names} @ovar{mode_switches}
9076 The only required argument is one @var{file_name}, which specifies
9077 a compilation unit that is a main program. Several @var{file_names} can be
9078 specified: this will result in several executables being built.
9079 If @code{switches} are present, they can be placed before the first
9080 @var{file_name}, between @var{file_names} or after the last @var{file_name}.
9081 If @var{mode_switches} are present, they must always be placed after
9082 the last @var{file_name} and all @code{switches}.
9084 If you are using standard file extensions (@file{.adb} and @file{.ads}), then the
9085 extension may be omitted from the @var{file_name} arguments. However, if
9086 you are using non-standard extensions, then it is required that the
9087 extension be given. A relative or absolute directory path can be
9088 specified in a @var{file_name}, in which case, the input source file will
9089 be searched for in the specified directory only. Otherwise, the input
9090 source file will first be searched in the directory where
9091 @command{gnatmake} was invoked and if it is not found, it will be search on
9092 the source path of the compiler as described in
9093 @ref{Search Paths and the Run-Time Library (RTL)}.
9095 All @command{gnatmake} output (except when you specify
9096 @option{^-M^/DEPENDENCIES_LIST^}) is to
9097 @file{stderr}. The output produced by the
9098 @option{^-M^/DEPENDENCIES_LIST^} switch is send to
9101 @node Switches for gnatmake
9102 @section Switches for @command{gnatmake}
9105 You may specify any of the following switches to @command{gnatmake}:
9111 @cindex @option{--version} @command{gnatmake}
9112 Display Copyright and version, then exit disregarding all other options.
9115 @cindex @option{--help} @command{gnatmake}
9116 If @option{--version} was not used, display usage, then exit disregarding
9120 @item --GCC=@var{compiler_name}
9121 @cindex @option{--GCC=compiler_name} (@command{gnatmake})
9122 Program used for compiling. The default is `@command{gcc}'. You need to use
9123 quotes around @var{compiler_name} if @code{compiler_name} contains
9124 spaces or other separator characters. As an example @option{--GCC="foo -x
9125 -y"} will instruct @command{gnatmake} to use @code{foo -x -y} as your
9126 compiler. A limitation of this syntax is that the name and path name of
9127 the executable itself must not include any embedded spaces. Note that
9128 switch @option{-c} is always inserted after your command name. Thus in the
9129 above example the compiler command that will be used by @command{gnatmake}
9130 will be @code{foo -c -x -y}. If several @option{--GCC=compiler_name} are
9131 used, only the last @var{compiler_name} is taken into account. However,
9132 all the additional switches are also taken into account. Thus,
9133 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
9134 @option{--GCC="bar -x -y -z -t"}.
9136 @item --GNATBIND=@var{binder_name}
9137 @cindex @option{--GNATBIND=binder_name} (@command{gnatmake})
9138 Program used for binding. The default is `@code{gnatbind}'. You need to
9139 use quotes around @var{binder_name} if @var{binder_name} contains spaces
9140 or other separator characters. As an example @option{--GNATBIND="bar -x
9141 -y"} will instruct @command{gnatmake} to use @code{bar -x -y} as your
9142 binder. Binder switches that are normally appended by @command{gnatmake}
9143 to `@code{gnatbind}' are now appended to the end of @code{bar -x -y}.
9144 A limitation of this syntax is that the name and path name of the executable
9145 itself must not include any embedded spaces.
9147 @item --GNATLINK=@var{linker_name}
9148 @cindex @option{--GNATLINK=linker_name} (@command{gnatmake})
9149 Program used for linking. The default is `@command{gnatlink}'. You need to
9150 use quotes around @var{linker_name} if @var{linker_name} contains spaces
9151 or other separator characters. As an example @option{--GNATLINK="lan -x
9152 -y"} will instruct @command{gnatmake} to use @code{lan -x -y} as your
9153 linker. Linker switches that are normally appended by @command{gnatmake} to
9154 `@command{gnatlink}' are now appended to the end of @code{lan -x -y}.
9155 A limitation of this syntax is that the name and path name of the executable
9156 itself must not include any embedded spaces.
9160 @item ^-a^/ALL_FILES^
9161 @cindex @option{^-a^/ALL_FILES^} (@command{gnatmake})
9162 Consider all files in the make process, even the GNAT internal system
9163 files (for example, the predefined Ada library files), as well as any
9164 locked files. Locked files are files whose ALI file is write-protected.
9166 @command{gnatmake} does not check these files,
9167 because the assumption is that the GNAT internal files are properly up
9168 to date, and also that any write protected ALI files have been properly
9169 installed. Note that if there is an installation problem, such that one
9170 of these files is not up to date, it will be properly caught by the
9172 You may have to specify this switch if you are working on GNAT
9173 itself. The switch @option{^-a^/ALL_FILES^} is also useful
9174 in conjunction with @option{^-f^/FORCE_COMPILE^}
9175 if you need to recompile an entire application,
9176 including run-time files, using special configuration pragmas,
9177 such as a @code{Normalize_Scalars} pragma.
9180 @code{gnatmake ^-a^/ALL_FILES^} compiles all GNAT
9183 @code{gcc -c -gnatpg} rather than @code{gcc -c}.
9186 the @code{/CHECKS=SUPPRESS_ALL /STYLE_CHECKS=GNAT} switch.
9189 @item ^-b^/ACTIONS=BIND^
9190 @cindex @option{^-b^/ACTIONS=BIND^} (@command{gnatmake})
9191 Bind only. Can be combined with @option{^-c^/ACTIONS=COMPILE^} to do
9192 compilation and binding, but no link.
9193 Can be combined with @option{^-l^/ACTIONS=LINK^}
9194 to do binding and linking. When not combined with
9195 @option{^-c^/ACTIONS=COMPILE^}
9196 all the units in the closure of the main program must have been previously
9197 compiled and must be up to date. The root unit specified by @var{file_name}
9198 may be given without extension, with the source extension or, if no GNAT
9199 Project File is specified, with the ALI file extension.
9201 @item ^-c^/ACTIONS=COMPILE^
9202 @cindex @option{^-c^/ACTIONS=COMPILE^} (@command{gnatmake})
9203 Compile only. Do not perform binding, except when @option{^-b^/ACTIONS=BIND^}
9204 is also specified. Do not perform linking, except if both
9205 @option{^-b^/ACTIONS=BIND^} and
9206 @option{^-l^/ACTIONS=LINK^} are also specified.
9207 If the root unit specified by @var{file_name} is not a main unit, this is the
9208 default. Otherwise @command{gnatmake} will attempt binding and linking
9209 unless all objects are up to date and the executable is more recent than
9213 @cindex @option{^-C^/MAPPING^} (@command{gnatmake})
9214 Use a temporary mapping file. A mapping file is a way to communicate to the
9215 compiler two mappings: from unit names to file names (without any directory
9216 information) and from file names to path names (with full directory
9217 information). These mappings are used by the compiler to short-circuit the path
9218 search. When @command{gnatmake} is invoked with this switch, it will create
9219 a temporary mapping file, initially populated by the project manager,
9220 if @option{^-P^/PROJECT_FILE^} is used, otherwise initially empty.
9221 Each invocation of the compiler will add the newly accessed sources to the
9222 mapping file. This will improve the source search during the next invocation
9225 @item ^-C=^/USE_MAPPING_FILE=^@var{file}
9226 @cindex @option{^-C=^/USE_MAPPING^} (@command{gnatmake})
9227 Use a specific mapping file. The file, specified as a path name (absolute or
9228 relative) by this switch, should already exist, otherwise the switch is
9229 ineffective. The specified mapping file will be communicated to the compiler.
9230 This switch is not compatible with a project file
9231 (^-P^/PROJECT_FILE=^@var{file}) or with multiple compiling processes
9232 (^-j^/PROCESSES=^nnn, when nnn is greater than 1).
9234 @item ^-d^/DISPLAY_PROGRESS^
9235 @cindex @option{^-d^/DISPLAY_PROGRESS^} (@command{gnatmake})
9236 Display progress for each source, up to date or not, as a single line
9239 completed x out of y (zz%)
9242 If the file needs to be compiled this is displayed after the invocation of
9243 the compiler. These lines are displayed even in quiet output mode.
9245 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
9246 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@command{gnatmake})
9247 Put all object files and ALI file in directory @var{dir}.
9248 If the @option{^-D^/DIRECTORY_OBJECTS^} switch is not used, all object files
9249 and ALI files go in the current working directory.
9251 This switch cannot be used when using a project file.
9255 @cindex @option{-eL} (@command{gnatmake})
9256 Follow all symbolic links when processing project files.
9259 @item ^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^
9260 @cindex @option{^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^} (@command{gnatmake})
9261 Output the commands for the compiler, the binder and the linker
9262 on ^standard output^SYS$OUTPUT^,
9263 instead of ^standard error^SYS$ERROR^.
9265 @item ^-f^/FORCE_COMPILE^
9266 @cindex @option{^-f^/FORCE_COMPILE^} (@command{gnatmake})
9267 Force recompilations. Recompile all sources, even though some object
9268 files may be up to date, but don't recompile predefined or GNAT internal
9269 files or locked files (files with a write-protected ALI file),
9270 unless the @option{^-a^/ALL_FILES^} switch is also specified.
9272 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
9273 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@command{gnatmake})
9274 When using project files, if some errors or warnings are detected during
9275 parsing and verbose mode is not in effect (no use of switch
9276 ^-v^/VERBOSE^), then error lines start with the full path name of the project
9277 file, rather than its simple file name.
9280 @cindex @option{^-g^/DEBUG^} (@command{gnatmake})
9281 Enable debugging. This switch is simply passed to the compiler and to the
9284 @item ^-i^/IN_PLACE^
9285 @cindex @option{^-i^/IN_PLACE^} (@command{gnatmake})
9286 In normal mode, @command{gnatmake} compiles all object files and ALI files
9287 into the current directory. If the @option{^-i^/IN_PLACE^} switch is used,
9288 then instead object files and ALI files that already exist are overwritten
9289 in place. This means that once a large project is organized into separate
9290 directories in the desired manner, then @command{gnatmake} will automatically
9291 maintain and update this organization. If no ALI files are found on the
9292 Ada object path (@ref{Search Paths and the Run-Time Library (RTL)}),
9293 the new object and ALI files are created in the
9294 directory containing the source being compiled. If another organization
9295 is desired, where objects and sources are kept in different directories,
9296 a useful technique is to create dummy ALI files in the desired directories.
9297 When detecting such a dummy file, @command{gnatmake} will be forced to
9298 recompile the corresponding source file, and it will be put the resulting
9299 object and ALI files in the directory where it found the dummy file.
9301 @item ^-j^/PROCESSES=^@var{n}
9302 @cindex @option{^-j^/PROCESSES^} (@command{gnatmake})
9303 @cindex Parallel make
9304 Use @var{n} processes to carry out the (re)compilations. On a
9305 multiprocessor machine compilations will occur in parallel. In the
9306 event of compilation errors, messages from various compilations might
9307 get interspersed (but @command{gnatmake} will give you the full ordered
9308 list of failing compiles at the end). If this is problematic, rerun
9309 the make process with n set to 1 to get a clean list of messages.
9311 @item ^-k^/CONTINUE_ON_ERROR^
9312 @cindex @option{^-k^/CONTINUE_ON_ERROR^} (@command{gnatmake})
9313 Keep going. Continue as much as possible after a compilation error. To
9314 ease the programmer's task in case of compilation errors, the list of
9315 sources for which the compile fails is given when @command{gnatmake}
9318 If @command{gnatmake} is invoked with several @file{file_names} and with this
9319 switch, if there are compilation errors when building an executable,
9320 @command{gnatmake} will not attempt to build the following executables.
9322 @item ^-l^/ACTIONS=LINK^
9323 @cindex @option{^-l^/ACTIONS=LINK^} (@command{gnatmake})
9324 Link only. Can be combined with @option{^-b^/ACTIONS=BIND^} to binding
9325 and linking. Linking will not be performed if combined with
9326 @option{^-c^/ACTIONS=COMPILE^}
9327 but not with @option{^-b^/ACTIONS=BIND^}.
9328 When not combined with @option{^-b^/ACTIONS=BIND^}
9329 all the units in the closure of the main program must have been previously
9330 compiled and must be up to date, and the main program needs to have been bound.
9331 The root unit specified by @var{file_name}
9332 may be given without extension, with the source extension or, if no GNAT
9333 Project File is specified, with the ALI file extension.
9335 @item ^-m^/MINIMAL_RECOMPILATION^
9336 @cindex @option{^-m^/MINIMAL_RECOMPILATION^} (@command{gnatmake})
9337 Specify that the minimum necessary amount of recompilations
9338 be performed. In this mode @command{gnatmake} ignores time
9339 stamp differences when the only
9340 modifications to a source file consist in adding/removing comments,
9341 empty lines, spaces or tabs. This means that if you have changed the
9342 comments in a source file or have simply reformatted it, using this
9343 switch will tell @command{gnatmake} not to recompile files that depend on it
9344 (provided other sources on which these files depend have undergone no
9345 semantic modifications). Note that the debugging information may be
9346 out of date with respect to the sources if the @option{-m} switch causes
9347 a compilation to be switched, so the use of this switch represents a
9348 trade-off between compilation time and accurate debugging information.
9350 @item ^-M^/DEPENDENCIES_LIST^
9351 @cindex Dependencies, producing list
9352 @cindex @option{^-M^/DEPENDENCIES_LIST^} (@command{gnatmake})
9353 Check if all objects are up to date. If they are, output the object
9354 dependences to @file{stdout} in a form that can be directly exploited in
9355 a @file{Makefile}. By default, each source file is prefixed with its
9356 (relative or absolute) directory name. This name is whatever you
9357 specified in the various @option{^-aI^/SOURCE_SEARCH^}
9358 and @option{^-I^/SEARCH^} switches. If you use
9359 @code{gnatmake ^-M^/DEPENDENCIES_LIST^}
9360 @option{^-q^/QUIET^}
9361 (see below), only the source file names,
9362 without relative paths, are output. If you just specify the
9363 @option{^-M^/DEPENDENCIES_LIST^}
9364 switch, dependencies of the GNAT internal system files are omitted. This
9365 is typically what you want. If you also specify
9366 the @option{^-a^/ALL_FILES^} switch,
9367 dependencies of the GNAT internal files are also listed. Note that
9368 dependencies of the objects in external Ada libraries (see switch
9369 @option{^-aL^/SKIP_MISSING=^}@var{dir} in the following list)
9372 @item ^-n^/DO_OBJECT_CHECK^
9373 @cindex @option{^-n^/DO_OBJECT_CHECK^} (@command{gnatmake})
9374 Don't compile, bind, or link. Checks if all objects are up to date.
9375 If they are not, the full name of the first file that needs to be
9376 recompiled is printed.
9377 Repeated use of this option, followed by compiling the indicated source
9378 file, will eventually result in recompiling all required units.
9380 @item ^-o ^/EXECUTABLE=^@var{exec_name}
9381 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatmake})
9382 Output executable name. The name of the final executable program will be
9383 @var{exec_name}. If the @option{^-o^/EXECUTABLE^} switch is omitted the default
9384 name for the executable will be the name of the input file in appropriate form
9385 for an executable file on the host system.
9387 This switch cannot be used when invoking @command{gnatmake} with several
9390 @item ^-p or --create-missing-dirs^/CREATE_MISSING_DIRS^
9391 @cindex @option{^-p^/CREATE_MISSING_DIRS^} (@command{gnatmake})
9392 When using project files (^-P^/PROJECT_FILE=^@var{project}), create
9393 automatically missing object directories, library directories and exec
9396 @item ^-P^/PROJECT_FILE=^@var{project}
9397 @cindex @option{^-P^/PROJECT_FILE^} (@command{gnatmake})
9398 Use project file @var{project}. Only one such switch can be used.
9399 @xref{gnatmake and Project Files}.
9402 @cindex @option{^-q^/QUIET^} (@command{gnatmake})
9403 Quiet. When this flag is not set, the commands carried out by
9404 @command{gnatmake} are displayed.
9406 @item ^-s^/SWITCH_CHECK/^
9407 @cindex @option{^-s^/SWITCH_CHECK^} (@command{gnatmake})
9408 Recompile if compiler switches have changed since last compilation.
9409 All compiler switches but -I and -o are taken into account in the
9411 orders between different ``first letter'' switches are ignored, but
9412 orders between same switches are taken into account. For example,
9413 @option{-O -O2} is different than @option{-O2 -O}, but @option{-g -O}
9414 is equivalent to @option{-O -g}.
9416 This switch is recommended when Integrated Preprocessing is used.
9419 @cindex @option{^-u^/UNIQUE^} (@command{gnatmake})
9420 Unique. Recompile at most the main files. It implies -c. Combined with
9421 -f, it is equivalent to calling the compiler directly. Note that using
9422 ^-u^/UNIQUE^ with a project file and no main has a special meaning
9423 (@pxref{Project Files and Main Subprograms}).
9425 @item ^-U^/ALL_PROJECTS^
9426 @cindex @option{^-U^/ALL_PROJECTS^} (@command{gnatmake})
9427 When used without a project file or with one or several mains on the command
9428 line, is equivalent to ^-u^/UNIQUE^. When used with a project file and no main
9429 on the command line, all sources of all project files are checked and compiled
9430 if not up to date, and libraries are rebuilt, if necessary.
9433 @cindex @option{^-v^/REASONS^} (@command{gnatmake})
9434 Verbose. Display the reason for all recompilations @command{gnatmake}
9435 decides are necessary, with the highest verbosity level.
9437 @item ^-vl^/LOW_VERBOSITY^
9438 @cindex @option{^-vl^/LOW_VERBOSITY^} (@command{gnatmake})
9439 Verbosity level Low. Display fewer lines than in verbosity Medium.
9441 @item ^-vm^/MEDIUM_VERBOSITY^
9442 @cindex @option{^-vm^/MEDIUM_VERBOSITY^} (@command{gnatmake})
9443 Verbosity level Medium. Potentially display fewer lines than in verbosity High.
9445 @item ^-vh^/HIGH_VERBOSITY^
9446 @cindex @option{^-vm^/HIGH_VERBOSITY^} (@command{gnatmake})
9447 Verbosity level High. Equivalent to ^-v^/REASONS^.
9449 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
9450 Indicate the verbosity of the parsing of GNAT project files.
9451 @xref{Switches Related to Project Files}.
9453 @item ^-x^/NON_PROJECT_UNIT_COMPILATION^
9454 @cindex @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} (@command{gnatmake})
9455 Indicate that sources that are not part of any Project File may be compiled.
9456 Normally, when using Project Files, only sources that are part of a Project
9457 File may be compile. When this switch is used, a source outside of all Project
9458 Files may be compiled. The ALI file and the object file will be put in the
9459 object directory of the main Project. The compilation switches used will only
9460 be those specified on the command line. Even when
9461 @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} is used, mains specified on the
9462 command line need to be sources of a project file.
9464 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
9465 Indicate that external variable @var{name} has the value @var{value}.
9466 The Project Manager will use this value for occurrences of
9467 @code{external(name)} when parsing the project file.
9468 @xref{Switches Related to Project Files}.
9471 @cindex @option{^-z^/NOMAIN^} (@command{gnatmake})
9472 No main subprogram. Bind and link the program even if the unit name
9473 given on the command line is a package name. The resulting executable
9474 will execute the elaboration routines of the package and its closure,
9475 then the finalization routines.
9480 @item @command{gcc} @asis{switches}
9482 Any uppercase or multi-character switch that is not a @command{gnatmake} switch
9483 is passed to @command{gcc} (e.g.@: @option{-O}, @option{-gnato,} etc.)
9486 Any qualifier that cannot be recognized as a qualifier for @code{GNAT MAKE}
9487 but is recognizable as a valid qualifier for @code{GNAT COMPILE} is
9488 automatically treated as a compiler switch, and passed on to all
9489 compilations that are carried out.
9494 Source and library search path switches:
9498 @item ^-aI^/SOURCE_SEARCH=^@var{dir}
9499 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatmake})
9500 When looking for source files also look in directory @var{dir}.
9501 The order in which source files search is undertaken is
9502 described in @ref{Search Paths and the Run-Time Library (RTL)}.
9504 @item ^-aL^/SKIP_MISSING=^@var{dir}
9505 @cindex @option{^-aL^/SKIP_MISSING^} (@command{gnatmake})
9506 Consider @var{dir} as being an externally provided Ada library.
9507 Instructs @command{gnatmake} to skip compilation units whose @file{.ALI}
9508 files have been located in directory @var{dir}. This allows you to have
9509 missing bodies for the units in @var{dir} and to ignore out of date bodies
9510 for the same units. You still need to specify
9511 the location of the specs for these units by using the switches
9512 @option{^-aI^/SOURCE_SEARCH=^@var{dir}}
9513 or @option{^-I^/SEARCH=^@var{dir}}.
9514 Note: this switch is provided for compatibility with previous versions
9515 of @command{gnatmake}. The easier method of causing standard libraries
9516 to be excluded from consideration is to write-protect the corresponding
9519 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
9520 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatmake})
9521 When searching for library and object files, look in directory
9522 @var{dir}. The order in which library files are searched is described in
9523 @ref{Search Paths for gnatbind}.
9525 @item ^-A^/CONDITIONAL_SOURCE_SEARCH=^@var{dir}
9526 @cindex Search paths, for @command{gnatmake}
9527 @cindex @option{^-A^/CONDITIONAL_SOURCE_SEARCH^} (@command{gnatmake})
9528 Equivalent to @option{^-aL^/SKIP_MISSING=^@var{dir}
9529 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9531 @item ^-I^/SEARCH=^@var{dir}
9532 @cindex @option{^-I^/SEARCH^} (@command{gnatmake})
9533 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}
9534 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9536 @item ^-I-^/NOCURRENT_DIRECTORY^
9537 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatmake})
9538 @cindex Source files, suppressing search
9539 Do not look for source files in the directory containing the source
9540 file named in the command line.
9541 Do not look for ALI or object files in the directory
9542 where @command{gnatmake} was invoked.
9544 @item ^-L^/LIBRARY_SEARCH=^@var{dir}
9545 @cindex @option{^-L^/LIBRARY_SEARCH^} (@command{gnatmake})
9546 @cindex Linker libraries
9547 Add directory @var{dir} to the list of directories in which the linker
9548 will search for libraries. This is equivalent to
9549 @option{-largs ^-L^/LIBRARY_SEARCH=^}@var{dir}.
9551 Furthermore, under Windows, the sources pointed to by the libraries path
9552 set in the registry are not searched for.
9556 @cindex @option{-nostdinc} (@command{gnatmake})
9557 Do not look for source files in the system default directory.
9560 @cindex @option{-nostdlib} (@command{gnatmake})
9561 Do not look for library files in the system default directory.
9563 @item --RTS=@var{rts-path}
9564 @cindex @option{--RTS} (@command{gnatmake})
9565 Specifies the default location of the runtime library. GNAT looks for the
9567 in the following directories, and stops as soon as a valid runtime is found
9568 (@file{adainclude} or @file{ada_source_path}, and @file{adalib} or
9569 @file{ada_object_path} present):
9572 @item <current directory>/$rts_path
9574 @item <default-search-dir>/$rts_path
9576 @item <default-search-dir>/rts-$rts_path
9580 The selected path is handled like a normal RTS path.
9584 @node Mode Switches for gnatmake
9585 @section Mode Switches for @command{gnatmake}
9588 The mode switches (referred to as @code{mode_switches}) allow the
9589 inclusion of switches that are to be passed to the compiler itself, the
9590 binder or the linker. The effect of a mode switch is to cause all
9591 subsequent switches up to the end of the switch list, or up to the next
9592 mode switch, to be interpreted as switches to be passed on to the
9593 designated component of GNAT.
9597 @item -cargs @var{switches}
9598 @cindex @option{-cargs} (@command{gnatmake})
9599 Compiler switches. Here @var{switches} is a list of switches
9600 that are valid switches for @command{gcc}. They will be passed on to
9601 all compile steps performed by @command{gnatmake}.
9603 @item -bargs @var{switches}
9604 @cindex @option{-bargs} (@command{gnatmake})
9605 Binder switches. Here @var{switches} is a list of switches
9606 that are valid switches for @code{gnatbind}. They will be passed on to
9607 all bind steps performed by @command{gnatmake}.
9609 @item -largs @var{switches}
9610 @cindex @option{-largs} (@command{gnatmake})
9611 Linker switches. Here @var{switches} is a list of switches
9612 that are valid switches for @command{gnatlink}. They will be passed on to
9613 all link steps performed by @command{gnatmake}.
9615 @item -margs @var{switches}
9616 @cindex @option{-margs} (@command{gnatmake})
9617 Make switches. The switches are directly interpreted by @command{gnatmake},
9618 regardless of any previous occurrence of @option{-cargs}, @option{-bargs}
9622 @node Notes on the Command Line
9623 @section Notes on the Command Line
9626 This section contains some additional useful notes on the operation
9627 of the @command{gnatmake} command.
9631 @cindex Recompilation, by @command{gnatmake}
9632 If @command{gnatmake} finds no ALI files, it recompiles the main program
9633 and all other units required by the main program.
9634 This means that @command{gnatmake}
9635 can be used for the initial compile, as well as during subsequent steps of
9636 the development cycle.
9639 If you enter @code{gnatmake @var{file}.adb}, where @file{@var{file}.adb}
9640 is a subunit or body of a generic unit, @command{gnatmake} recompiles
9641 @file{@var{file}.adb} (because it finds no ALI) and stops, issuing a
9645 In @command{gnatmake} the switch @option{^-I^/SEARCH^}
9646 is used to specify both source and
9647 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
9648 instead if you just want to specify
9649 source paths only and @option{^-aO^/OBJECT_SEARCH^}
9650 if you want to specify library paths
9654 @command{gnatmake} will ignore any files whose ALI file is write-protected.
9655 This may conveniently be used to exclude standard libraries from
9656 consideration and in particular it means that the use of the
9657 @option{^-f^/FORCE_COMPILE^} switch will not recompile these files
9658 unless @option{^-a^/ALL_FILES^} is also specified.
9661 @command{gnatmake} has been designed to make the use of Ada libraries
9662 particularly convenient. Assume you have an Ada library organized
9663 as follows: @i{^obj-dir^[OBJ_DIR]^} contains the objects and ALI files for
9664 of your Ada compilation units,
9665 whereas @i{^include-dir^[INCLUDE_DIR]^} contains the
9666 specs of these units, but no bodies. Then to compile a unit
9667 stored in @code{main.adb}, which uses this Ada library you would just type
9671 $ gnatmake -aI@var{include-dir} -aL@var{obj-dir} main
9674 $ gnatmake /SOURCE_SEARCH=@i{[INCLUDE_DIR]}
9675 /SKIP_MISSING=@i{[OBJ_DIR]} main
9680 Using @command{gnatmake} along with the
9681 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}
9682 switch provides a mechanism for avoiding unnecessary recompilations. Using
9684 you can update the comments/format of your
9685 source files without having to recompile everything. Note, however, that
9686 adding or deleting lines in a source files may render its debugging
9687 info obsolete. If the file in question is a spec, the impact is rather
9688 limited, as that debugging info will only be useful during the
9689 elaboration phase of your program. For bodies the impact can be more
9690 significant. In all events, your debugger will warn you if a source file
9691 is more recent than the corresponding object, and alert you to the fact
9692 that the debugging information may be out of date.
9695 @node How gnatmake Works
9696 @section How @command{gnatmake} Works
9699 Generally @command{gnatmake} automatically performs all necessary
9700 recompilations and you don't need to worry about how it works. However,
9701 it may be useful to have some basic understanding of the @command{gnatmake}
9702 approach and in particular to understand how it uses the results of
9703 previous compilations without incorrectly depending on them.
9705 First a definition: an object file is considered @dfn{up to date} if the
9706 corresponding ALI file exists and if all the source files listed in the
9707 dependency section of this ALI file have time stamps matching those in
9708 the ALI file. This means that neither the source file itself nor any
9709 files that it depends on have been modified, and hence there is no need
9710 to recompile this file.
9712 @command{gnatmake} works by first checking if the specified main unit is up
9713 to date. If so, no compilations are required for the main unit. If not,
9714 @command{gnatmake} compiles the main program to build a new ALI file that
9715 reflects the latest sources. Then the ALI file of the main unit is
9716 examined to find all the source files on which the main program depends,
9717 and @command{gnatmake} recursively applies the above procedure on all these
9720 This process ensures that @command{gnatmake} only trusts the dependencies
9721 in an existing ALI file if they are known to be correct. Otherwise it
9722 always recompiles to determine a new, guaranteed accurate set of
9723 dependencies. As a result the program is compiled ``upside down'' from what may
9724 be more familiar as the required order of compilation in some other Ada
9725 systems. In particular, clients are compiled before the units on which
9726 they depend. The ability of GNAT to compile in any order is critical in
9727 allowing an order of compilation to be chosen that guarantees that
9728 @command{gnatmake} will recompute a correct set of new dependencies if
9731 When invoking @command{gnatmake} with several @var{file_names}, if a unit is
9732 imported by several of the executables, it will be recompiled at most once.
9734 Note: when using non-standard naming conventions
9735 (@pxref{Using Other File Names}), changing through a configuration pragmas
9736 file the version of a source and invoking @command{gnatmake} to recompile may
9737 have no effect, if the previous version of the source is still accessible
9738 by @command{gnatmake}. It may be necessary to use the switch
9739 ^-f^/FORCE_COMPILE^.
9741 @node Examples of gnatmake Usage
9742 @section Examples of @command{gnatmake} Usage
9745 @item gnatmake hello.adb
9746 Compile all files necessary to bind and link the main program
9747 @file{hello.adb} (containing unit @code{Hello}) and bind and link the
9748 resulting object files to generate an executable file @file{^hello^HELLO.EXE^}.
9750 @item gnatmake main1 main2 main3
9751 Compile all files necessary to bind and link the main programs
9752 @file{main1.adb} (containing unit @code{Main1}), @file{main2.adb}
9753 (containing unit @code{Main2}) and @file{main3.adb}
9754 (containing unit @code{Main3}) and bind and link the resulting object files
9755 to generate three executable files @file{^main1^MAIN1.EXE^},
9756 @file{^main2^MAIN2.EXE^}
9757 and @file{^main3^MAIN3.EXE^}.
9760 @item gnatmake -q Main_Unit -cargs -O2 -bargs -l
9764 @item gnatmake Main_Unit /QUIET
9765 /COMPILER_QUALIFIERS /OPTIMIZE=ALL
9766 /BINDER_QUALIFIERS /ORDER_OF_ELABORATION
9768 Compile all files necessary to bind and link the main program unit
9769 @code{Main_Unit} (from file @file{main_unit.adb}). All compilations will
9770 be done with optimization level 2 and the order of elaboration will be
9771 listed by the binder. @command{gnatmake} will operate in quiet mode, not
9772 displaying commands it is executing.
9775 @c *************************
9776 @node Improving Performance
9777 @chapter Improving Performance
9778 @cindex Improving performance
9781 This chapter presents several topics related to program performance.
9782 It first describes some of the tradeoffs that need to be considered
9783 and some of the techniques for making your program run faster.
9784 It then documents the @command{gnatelim} tool and unused subprogram/data
9785 elimination feature, which can reduce the size of program executables.
9787 Note: to invoke @command{gnatelim} with a project file, use the @code{gnat}
9788 driver (see @ref{The GNAT Driver and Project Files}).
9792 * Performance Considerations::
9793 * Text_IO Suggestions::
9794 * Reducing Size of Ada Executables with gnatelim::
9795 * Reducing Size of Executables with unused subprogram/data elimination::
9799 @c *****************************
9800 @node Performance Considerations
9801 @section Performance Considerations
9804 The GNAT system provides a number of options that allow a trade-off
9809 performance of the generated code
9812 speed of compilation
9815 minimization of dependences and recompilation
9818 the degree of run-time checking.
9822 The defaults (if no options are selected) aim at improving the speed
9823 of compilation and minimizing dependences, at the expense of performance
9824 of the generated code:
9831 no inlining of subprogram calls
9834 all run-time checks enabled except overflow and elaboration checks
9838 These options are suitable for most program development purposes. This
9839 chapter describes how you can modify these choices, and also provides
9840 some guidelines on debugging optimized code.
9843 * Controlling Run-Time Checks::
9844 * Use of Restrictions::
9845 * Optimization Levels::
9846 * Debugging Optimized Code::
9847 * Inlining of Subprograms::
9848 * Other Optimization Switches::
9849 * Optimization and Strict Aliasing::
9852 * Coverage Analysis::
9856 @node Controlling Run-Time Checks
9857 @subsection Controlling Run-Time Checks
9860 By default, GNAT generates all run-time checks, except integer overflow
9861 checks, stack overflow checks, and checks for access before elaboration on
9862 subprogram calls. The latter are not required in default mode, because all
9863 necessary checking is done at compile time.
9864 @cindex @option{-gnatp} (@command{gcc})
9865 @cindex @option{-gnato} (@command{gcc})
9866 Two gnat switches, @option{-gnatp} and @option{-gnato} allow this default to
9867 be modified. @xref{Run-Time Checks}.
9869 Our experience is that the default is suitable for most development
9872 We treat integer overflow specially because these
9873 are quite expensive and in our experience are not as important as other
9874 run-time checks in the development process. Note that division by zero
9875 is not considered an overflow check, and divide by zero checks are
9876 generated where required by default.
9878 Elaboration checks are off by default, and also not needed by default, since
9879 GNAT uses a static elaboration analysis approach that avoids the need for
9880 run-time checking. This manual contains a full chapter discussing the issue
9881 of elaboration checks, and if the default is not satisfactory for your use,
9882 you should read this chapter.
9884 For validity checks, the minimal checks required by the Ada Reference
9885 Manual (for case statements and assignments to array elements) are on
9886 by default. These can be suppressed by use of the @option{-gnatVn} switch.
9887 Note that in Ada 83, there were no validity checks, so if the Ada 83 mode
9888 is acceptable (or when comparing GNAT performance with an Ada 83 compiler),
9889 it may be reasonable to routinely use @option{-gnatVn}. Validity checks
9890 are also suppressed entirely if @option{-gnatp} is used.
9892 @cindex Overflow checks
9893 @cindex Checks, overflow
9896 @cindex pragma Suppress
9897 @cindex pragma Unsuppress
9898 Note that the setting of the switches controls the default setting of
9899 the checks. They may be modified using either @code{pragma Suppress} (to
9900 remove checks) or @code{pragma Unsuppress} (to add back suppressed
9901 checks) in the program source.
9903 @node Use of Restrictions
9904 @subsection Use of Restrictions
9907 The use of pragma Restrictions allows you to control which features are
9908 permitted in your program. Apart from the obvious point that if you avoid
9909 relatively expensive features like finalization (enforceable by the use
9910 of pragma Restrictions (No_Finalization), the use of this pragma does not
9911 affect the generated code in most cases.
9913 One notable exception to this rule is that the possibility of task abort
9914 results in some distributed overhead, particularly if finalization or
9915 exception handlers are used. The reason is that certain sections of code
9916 have to be marked as non-abortable.
9918 If you use neither the @code{abort} statement, nor asynchronous transfer
9919 of control (@code{select @dots{} then abort}), then this distributed overhead
9920 is removed, which may have a general positive effect in improving
9921 overall performance. Especially code involving frequent use of tasking
9922 constructs and controlled types will show much improved performance.
9923 The relevant restrictions pragmas are
9925 @smallexample @c ada
9926 pragma Restrictions (No_Abort_Statements);
9927 pragma Restrictions (Max_Asynchronous_Select_Nesting => 0);
9931 It is recommended that these restriction pragmas be used if possible. Note
9932 that this also means that you can write code without worrying about the
9933 possibility of an immediate abort at any point.
9935 @node Optimization Levels
9936 @subsection Optimization Levels
9937 @cindex @option{^-O^/OPTIMIZE^} (@command{gcc})
9940 Without any optimization ^option,^qualifier,^
9941 the compiler's goal is to reduce the cost of
9942 compilation and to make debugging produce the expected results.
9943 Statements are independent: if you stop the program with a breakpoint between
9944 statements, you can then assign a new value to any variable or change
9945 the program counter to any other statement in the subprogram and get exactly
9946 the results you would expect from the source code.
9948 Turning on optimization makes the compiler attempt to improve the
9949 performance and/or code size at the expense of compilation time and
9950 possibly the ability to debug the program.
9953 ^-O options, with or without level numbers,^/OPTIMIZE qualifiers,^
9954 the last such option is the one that is effective.
9957 The default is optimization off. This results in the fastest compile
9958 times, but GNAT makes absolutely no attempt to optimize, and the
9959 generated programs are considerably larger and slower than when
9960 optimization is enabled. You can use the
9962 @option{-O} switch (the permitted forms are @option{-O0}, @option{-O1}
9963 @option{-O2}, @option{-O3}, and @option{-Os})
9966 @code{OPTIMIZE} qualifier
9968 to @command{gcc} to control the optimization level:
9971 @item ^-O0^/OPTIMIZE=NONE^
9972 No optimization (the default);
9973 generates unoptimized code but has
9974 the fastest compilation time.
9976 Note that many other compilers do fairly extensive optimization
9977 even if ``no optimization'' is specified. With gcc, it is
9978 very unusual to use ^-O0^/OPTIMIZE=NONE^ for production if
9979 execution time is of any concern, since ^-O0^/OPTIMIZE=NONE^
9980 really does mean no optimization at all. This difference between
9981 gcc and other compilers should be kept in mind when doing
9982 performance comparisons.
9984 @item ^-O1^/OPTIMIZE=SOME^
9985 Moderate optimization;
9986 optimizes reasonably well but does not
9987 degrade compilation time significantly.
9989 @item ^-O2^/OPTIMIZE=ALL^
9991 @itemx /OPTIMIZE=DEVELOPMENT
9994 generates highly optimized code and has
9995 the slowest compilation time.
9997 @item ^-O3^/OPTIMIZE=INLINING^
9998 Full optimization as in @option{-O2},
9999 and also attempts automatic inlining of small
10000 subprograms within a unit (@pxref{Inlining of Subprograms}).
10002 @item ^-Os^/OPTIMIZE=SPACE^
10003 Optimize space usage of resulting program.
10007 Higher optimization levels perform more global transformations on the
10008 program and apply more expensive analysis algorithms in order to generate
10009 faster and more compact code. The price in compilation time, and the
10010 resulting improvement in execution time,
10011 both depend on the particular application and the hardware environment.
10012 You should experiment to find the best level for your application.
10014 Since the precise set of optimizations done at each level will vary from
10015 release to release (and sometime from target to target), it is best to think
10016 of the optimization settings in general terms.
10017 @xref{Optimize Options,, Options That Control Optimization, gcc, Using
10018 the GNU Compiler Collection (GCC)}, for details about
10019 ^the @option{-O} settings and a number of @option{-f} options that^how to^
10020 individually enable or disable specific optimizations.
10022 Unlike some other compilation systems, ^@command{gcc}^GNAT^ has
10023 been tested extensively at all optimization levels. There are some bugs
10024 which appear only with optimization turned on, but there have also been
10025 bugs which show up only in @emph{unoptimized} code. Selecting a lower
10026 level of optimization does not improve the reliability of the code
10027 generator, which in practice is highly reliable at all optimization
10030 Note regarding the use of @option{-O3}: The use of this optimization level
10031 is generally discouraged with GNAT, since it often results in larger
10032 executables which run more slowly. See further discussion of this point
10033 in @ref{Inlining of Subprograms}.
10035 @node Debugging Optimized Code
10036 @subsection Debugging Optimized Code
10037 @cindex Debugging optimized code
10038 @cindex Optimization and debugging
10041 Although it is possible to do a reasonable amount of debugging at
10043 nonzero optimization levels,
10044 the higher the level the more likely that
10047 @option{/OPTIMIZE} settings other than @code{NONE},
10048 such settings will make it more likely that
10050 source-level constructs will have been eliminated by optimization.
10051 For example, if a loop is strength-reduced, the loop
10052 control variable may be completely eliminated and thus cannot be
10053 displayed in the debugger.
10054 This can only happen at @option{-O2} or @option{-O3}.
10055 Explicit temporary variables that you code might be eliminated at
10056 ^level^setting^ @option{-O1} or higher.
10058 The use of the @option{^-g^/DEBUG^} switch,
10059 @cindex @option{^-g^/DEBUG^} (@command{gcc})
10060 which is needed for source-level debugging,
10061 affects the size of the program executable on disk,
10062 and indeed the debugging information can be quite large.
10063 However, it has no effect on the generated code (and thus does not
10064 degrade performance)
10066 Since the compiler generates debugging tables for a compilation unit before
10067 it performs optimizations, the optimizing transformations may invalidate some
10068 of the debugging data. You therefore need to anticipate certain
10069 anomalous situations that may arise while debugging optimized code.
10070 These are the most common cases:
10074 @i{The ``hopping Program Counter'':} Repeated @code{step} or @code{next}
10076 the PC bouncing back and forth in the code. This may result from any of
10077 the following optimizations:
10081 @i{Common subexpression elimination:} using a single instance of code for a
10082 quantity that the source computes several times. As a result you
10083 may not be able to stop on what looks like a statement.
10086 @i{Invariant code motion:} moving an expression that does not change within a
10087 loop, to the beginning of the loop.
10090 @i{Instruction scheduling:} moving instructions so as to
10091 overlap loads and stores (typically) with other code, or in
10092 general to move computations of values closer to their uses. Often
10093 this causes you to pass an assignment statement without the assignment
10094 happening and then later bounce back to the statement when the
10095 value is actually needed. Placing a breakpoint on a line of code
10096 and then stepping over it may, therefore, not always cause all the
10097 expected side-effects.
10101 @i{The ``big leap'':} More commonly known as @emph{cross-jumping}, in which
10102 two identical pieces of code are merged and the program counter suddenly
10103 jumps to a statement that is not supposed to be executed, simply because
10104 it (and the code following) translates to the same thing as the code
10105 that @emph{was} supposed to be executed. This effect is typically seen in
10106 sequences that end in a jump, such as a @code{goto}, a @code{return}, or
10107 a @code{break} in a C @code{^switch^switch^} statement.
10110 @i{The ``roving variable'':} The symptom is an unexpected value in a variable.
10111 There are various reasons for this effect:
10115 In a subprogram prologue, a parameter may not yet have been moved to its
10119 A variable may be dead, and its register re-used. This is
10120 probably the most common cause.
10123 As mentioned above, the assignment of a value to a variable may
10127 A variable may be eliminated entirely by value propagation or
10128 other means. In this case, GCC may incorrectly generate debugging
10129 information for the variable
10133 In general, when an unexpected value appears for a local variable or parameter
10134 you should first ascertain if that value was actually computed by
10135 your program, as opposed to being incorrectly reported by the debugger.
10137 array elements in an object designated by an access value
10138 are generally less of a problem, once you have ascertained that the access
10140 Typically, this means checking variables in the preceding code and in the
10141 calling subprogram to verify that the value observed is explainable from other
10142 values (one must apply the procedure recursively to those
10143 other values); or re-running the code and stopping a little earlier
10144 (perhaps before the call) and stepping to better see how the variable obtained
10145 the value in question; or continuing to step @emph{from} the point of the
10146 strange value to see if code motion had simply moved the variable's
10151 In light of such anomalies, a recommended technique is to use @option{-O0}
10152 early in the software development cycle, when extensive debugging capabilities
10153 are most needed, and then move to @option{-O1} and later @option{-O2} as
10154 the debugger becomes less critical.
10155 Whether to use the @option{^-g^/DEBUG^} switch in the release version is
10156 a release management issue.
10158 Note that if you use @option{-g} you can then use the @command{strip} program
10159 on the resulting executable,
10160 which removes both debugging information and global symbols.
10163 @node Inlining of Subprograms
10164 @subsection Inlining of Subprograms
10167 A call to a subprogram in the current unit is inlined if all the
10168 following conditions are met:
10172 The optimization level is at least @option{-O1}.
10175 The called subprogram is suitable for inlining: It must be small enough
10176 and not contain something that @command{gcc} cannot support in inlined
10180 @cindex pragma Inline
10182 Either @code{pragma Inline} applies to the subprogram, or it is local
10183 to the unit and called once from within it, or it is small and automatic
10184 inlining (optimization level @option{-O3}) is specified.
10188 Calls to subprograms in @code{with}'ed units are normally not inlined.
10189 To achieve actual inlining (that is, replacement of the call by the code
10190 in the body of the subprogram), the following conditions must all be true.
10194 The optimization level is at least @option{-O1}.
10197 The called subprogram is suitable for inlining: It must be small enough
10198 and not contain something that @command{gcc} cannot support in inlined
10202 The call appears in a body (not in a package spec).
10205 There is a @code{pragma Inline} for the subprogram.
10208 @cindex @option{-gnatn} (@command{gcc})
10209 The @option{^-gnatn^/INLINE^} switch
10210 is used in the @command{gcc} command line
10213 Even if all these conditions are met, it may not be possible for
10214 the compiler to inline the call, due to the length of the body,
10215 or features in the body that make it impossible for the compiler
10216 to do the inlining.
10218 Note that specifying the @option{-gnatn} switch causes additional
10219 compilation dependencies. Consider the following:
10221 @smallexample @c ada
10241 With the default behavior (no @option{-gnatn} switch specified), the
10242 compilation of the @code{Main} procedure depends only on its own source,
10243 @file{main.adb}, and the spec of the package in file @file{r.ads}. This
10244 means that editing the body of @code{R} does not require recompiling
10247 On the other hand, the call @code{R.Q} is not inlined under these
10248 circumstances. If the @option{-gnatn} switch is present when @code{Main}
10249 is compiled, the call will be inlined if the body of @code{Q} is small
10250 enough, but now @code{Main} depends on the body of @code{R} in
10251 @file{r.adb} as well as on the spec. This means that if this body is edited,
10252 the main program must be recompiled. Note that this extra dependency
10253 occurs whether or not the call is in fact inlined by @command{gcc}.
10255 The use of front end inlining with @option{-gnatN} generates similar
10256 additional dependencies.
10258 @cindex @option{^-fno-inline^/INLINE=SUPPRESS^} (@command{gcc})
10259 Note: The @option{^-fno-inline^/INLINE=SUPPRESS^} switch
10260 can be used to prevent
10261 all inlining. This switch overrides all other conditions and ensures
10262 that no inlining occurs. The extra dependences resulting from
10263 @option{-gnatn} will still be active, even if
10264 this switch is used to suppress the resulting inlining actions.
10266 @cindex @option{-fno-inline-functions} (@command{gcc})
10267 Note: The @option{-fno-inline-functions} switch can be used to prevent
10268 automatic inlining of small subprograms if @option{-O3} is used.
10270 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
10271 Note: The @option{-fno-inline-functions-called-once} switch
10272 can be used to prevent inlining of subprograms local to the unit
10273 and called once from within it if @option{-O1} is used.
10275 Note regarding the use of @option{-O3}: There is no difference in inlining
10276 behavior between @option{-O2} and @option{-O3} for subprograms with an explicit
10277 pragma @code{Inline} assuming the use of @option{-gnatn}
10278 or @option{-gnatN} (the switches that activate inlining). If you have used
10279 pragma @code{Inline} in appropriate cases, then it is usually much better
10280 to use @option{-O2} and @option{-gnatn} and avoid the use of @option{-O3} which
10281 in this case only has the effect of inlining subprograms you did not
10282 think should be inlined. We often find that the use of @option{-O3} slows
10283 down code by performing excessive inlining, leading to increased instruction
10284 cache pressure from the increased code size. So the bottom line here is
10285 that you should not automatically assume that @option{-O3} is better than
10286 @option{-O2}, and indeed you should use @option{-O3} only if tests show that
10287 it actually improves performance.
10289 @node Other Optimization Switches
10290 @subsection Other Optimization Switches
10291 @cindex Optimization Switches
10293 Since @code{GNAT} uses the @command{gcc} back end, all the specialized
10294 @command{gcc} optimization switches are potentially usable. These switches
10295 have not been extensively tested with GNAT but can generally be expected
10296 to work. Examples of switches in this category are
10297 @option{-funroll-loops} and
10298 the various target-specific @option{-m} options (in particular, it has been
10299 observed that @option{-march=pentium4} can significantly improve performance
10300 on appropriate machines). For full details of these switches, see
10301 @ref{Submodel Options,, Hardware Models and Configurations, gcc, Using
10302 the GNU Compiler Collection (GCC)}.
10304 @node Optimization and Strict Aliasing
10305 @subsection Optimization and Strict Aliasing
10307 @cindex Strict Aliasing
10308 @cindex No_Strict_Aliasing
10311 The strong typing capabilities of Ada allow an optimizer to generate
10312 efficient code in situations where other languages would be forced to
10313 make worst case assumptions preventing such optimizations. Consider
10314 the following example:
10316 @smallexample @c ada
10319 type Int1 is new Integer;
10320 type Int2 is new Integer;
10321 type Int1A is access Int1;
10322 type Int2A is access Int2;
10329 for J in Data'Range loop
10330 if Data (J) = Int1V.all then
10331 Int2V.all := Int2V.all + 1;
10340 In this example, since the variable @code{Int1V} can only access objects
10341 of type @code{Int1}, and @code{Int2V} can only access objects of type
10342 @code{Int2}, there is no possibility that the assignment to
10343 @code{Int2V.all} affects the value of @code{Int1V.all}. This means that
10344 the compiler optimizer can "know" that the value @code{Int1V.all} is constant
10345 for all iterations of the loop and avoid the extra memory reference
10346 required to dereference it each time through the loop.
10348 This kind of optimization, called strict aliasing analysis, is
10349 triggered by specifying an optimization level of @option{-O2} or
10350 higher or @option{-Os} and allows @code{GNAT} to generate more efficient code
10351 when access values are involved.
10353 However, although this optimization is always correct in terms of
10354 the formal semantics of the Ada Reference Manual, difficulties can
10355 arise if features like @code{Unchecked_Conversion} are used to break
10356 the typing system. Consider the following complete program example:
10358 @smallexample @c ada
10361 type int1 is new integer;
10362 type int2 is new integer;
10363 type a1 is access int1;
10364 type a2 is access int2;
10369 function to_a2 (Input : a1) return a2;
10372 with Unchecked_Conversion;
10374 function to_a2 (Input : a1) return a2 is
10376 new Unchecked_Conversion (a1, a2);
10378 return to_a2u (Input);
10384 with Text_IO; use Text_IO;
10386 v1 : a1 := new int1;
10387 v2 : a2 := to_a2 (v1);
10391 put_line (int1'image (v1.all));
10397 This program prints out 0 in @option{-O0} or @option{-O1}
10398 mode, but it prints out 1 in @option{-O2} mode. That's
10399 because in strict aliasing mode, the compiler can and
10400 does assume that the assignment to @code{v2.all} could not
10401 affect the value of @code{v1.all}, since different types
10404 This behavior is not a case of non-conformance with the standard, since
10405 the Ada RM specifies that an unchecked conversion where the resulting
10406 bit pattern is not a correct value of the target type can result in an
10407 abnormal value and attempting to reference an abnormal value makes the
10408 execution of a program erroneous. That's the case here since the result
10409 does not point to an object of type @code{int2}. This means that the
10410 effect is entirely unpredictable.
10412 However, although that explanation may satisfy a language
10413 lawyer, in practice an applications programmer expects an
10414 unchecked conversion involving pointers to create true
10415 aliases and the behavior of printing 1 seems plain wrong.
10416 In this case, the strict aliasing optimization is unwelcome.
10418 Indeed the compiler recognizes this possibility, and the
10419 unchecked conversion generates a warning:
10422 p2.adb:5:07: warning: possible aliasing problem with type "a2"
10423 p2.adb:5:07: warning: use -fno-strict-aliasing switch for references
10424 p2.adb:5:07: warning: or use "pragma No_Strict_Aliasing (a2);"
10428 Unfortunately the problem is recognized when compiling the body of
10429 package @code{p2}, but the actual "bad" code is generated while
10430 compiling the body of @code{m} and this latter compilation does not see
10431 the suspicious @code{Unchecked_Conversion}.
10433 As implied by the warning message, there are approaches you can use to
10434 avoid the unwanted strict aliasing optimization in a case like this.
10436 One possibility is to simply avoid the use of @option{-O2}, but
10437 that is a bit drastic, since it throws away a number of useful
10438 optimizations that do not involve strict aliasing assumptions.
10440 A less drastic approach is to compile the program using the
10441 option @option{-fno-strict-aliasing}. Actually it is only the
10442 unit containing the dereferencing of the suspicious pointer
10443 that needs to be compiled. So in this case, if we compile
10444 unit @code{m} with this switch, then we get the expected
10445 value of zero printed. Analyzing which units might need
10446 the switch can be painful, so a more reasonable approach
10447 is to compile the entire program with options @option{-O2}
10448 and @option{-fno-strict-aliasing}. If the performance is
10449 satisfactory with this combination of options, then the
10450 advantage is that the entire issue of possible "wrong"
10451 optimization due to strict aliasing is avoided.
10453 To avoid the use of compiler switches, the configuration
10454 pragma @code{No_Strict_Aliasing} with no parameters may be
10455 used to specify that for all access types, the strict
10456 aliasing optimization should be suppressed.
10458 However, these approaches are still overkill, in that they causes
10459 all manipulations of all access values to be deoptimized. A more
10460 refined approach is to concentrate attention on the specific
10461 access type identified as problematic.
10463 First, if a careful analysis of uses of the pointer shows
10464 that there are no possible problematic references, then
10465 the warning can be suppressed by bracketing the
10466 instantiation of @code{Unchecked_Conversion} to turn
10469 @smallexample @c ada
10470 pragma Warnings (Off);
10472 new Unchecked_Conversion (a1, a2);
10473 pragma Warnings (On);
10477 Of course that approach is not appropriate for this particular
10478 example, since indeed there is a problematic reference. In this
10479 case we can take one of two other approaches.
10481 The first possibility is to move the instantiation of unchecked
10482 conversion to the unit in which the type is declared. In
10483 this example, we would move the instantiation of
10484 @code{Unchecked_Conversion} from the body of package
10485 @code{p2} to the spec of package @code{p1}. Now the
10486 warning disappears. That's because any use of the
10487 access type knows there is a suspicious unchecked
10488 conversion, and the strict aliasing optimization
10489 is automatically suppressed for the type.
10491 If it is not practical to move the unchecked conversion to the same unit
10492 in which the destination access type is declared (perhaps because the
10493 source type is not visible in that unit), you may use pragma
10494 @code{No_Strict_Aliasing} for the type. This pragma must occur in the
10495 same declarative sequence as the declaration of the access type:
10497 @smallexample @c ada
10498 type a2 is access int2;
10499 pragma No_Strict_Aliasing (a2);
10503 Here again, the compiler now knows that the strict aliasing optimization
10504 should be suppressed for any reference to type @code{a2} and the
10505 expected behavior is obtained.
10507 Finally, note that although the compiler can generate warnings for
10508 simple cases of unchecked conversions, there are tricker and more
10509 indirect ways of creating type incorrect aliases which the compiler
10510 cannot detect. Examples are the use of address overlays and unchecked
10511 conversions involving composite types containing access types as
10512 components. In such cases, no warnings are generated, but there can
10513 still be aliasing problems. One safe coding practice is to forbid the
10514 use of address clauses for type overlaying, and to allow unchecked
10515 conversion only for primitive types. This is not really a significant
10516 restriction since any possible desired effect can be achieved by
10517 unchecked conversion of access values.
10519 The aliasing analysis done in strict aliasing mode can certainly
10520 have significant benefits. We have seen cases of large scale
10521 application code where the time is increased by up to 5% by turning
10522 this optimization off. If you have code that includes significant
10523 usage of unchecked conversion, you might want to just stick with
10524 @option{-O1} and avoid the entire issue. If you get adequate
10525 performance at this level of optimization level, that's probably
10526 the safest approach. If tests show that you really need higher
10527 levels of optimization, then you can experiment with @option{-O2}
10528 and @option{-O2 -fno-strict-aliasing} to see how much effect this
10529 has on size and speed of the code. If you really need to use
10530 @option{-O2} with strict aliasing in effect, then you should
10531 review any uses of unchecked conversion of access types,
10532 particularly if you are getting the warnings described above.
10535 @node Coverage Analysis
10536 @subsection Coverage Analysis
10539 GNAT supports the HP Performance Coverage Analyzer (PCA), which allows
10540 the user to determine the distribution of execution time across a program,
10541 @pxref{Profiling} for details of usage.
10545 @node Text_IO Suggestions
10546 @section @code{Text_IO} Suggestions
10547 @cindex @code{Text_IO} and performance
10550 The @code{Ada.Text_IO} package has fairly high overheads due in part to
10551 the requirement of maintaining page and line counts. If performance
10552 is critical, a recommendation is to use @code{Stream_IO} instead of
10553 @code{Text_IO} for volume output, since this package has less overhead.
10555 If @code{Text_IO} must be used, note that by default output to the standard
10556 output and standard error files is unbuffered (this provides better
10557 behavior when output statements are used for debugging, or if the
10558 progress of a program is observed by tracking the output, e.g. by
10559 using the Unix @command{tail -f} command to watch redirected output.
10561 If you are generating large volumes of output with @code{Text_IO} and
10562 performance is an important factor, use a designated file instead
10563 of the standard output file, or change the standard output file to
10564 be buffered using @code{Interfaces.C_Streams.setvbuf}.
10568 @node Reducing Size of Ada Executables with gnatelim
10569 @section Reducing Size of Ada Executables with @code{gnatelim}
10573 This section describes @command{gnatelim}, a tool which detects unused
10574 subprograms and helps the compiler to create a smaller executable for your
10579 * Running gnatelim::
10580 * Correcting the List of Eliminate Pragmas::
10581 * Making Your Executables Smaller::
10582 * Summary of the gnatelim Usage Cycle::
10585 @node About gnatelim
10586 @subsection About @code{gnatelim}
10589 When a program shares a set of Ada
10590 packages with other programs, it may happen that this program uses
10591 only a fraction of the subprograms defined in these packages. The code
10592 created for these unused subprograms increases the size of the executable.
10594 @code{gnatelim} tracks unused subprograms in an Ada program and
10595 outputs a list of GNAT-specific pragmas @code{Eliminate} marking all the
10596 subprograms that are declared but never called. By placing the list of
10597 @code{Eliminate} pragmas in the GNAT configuration file @file{gnat.adc} and
10598 recompiling your program, you may decrease the size of its executable,
10599 because the compiler will not generate the code for 'eliminated' subprograms.
10600 @xref{Pragma Eliminate,,, gnat_rm, GNAT Reference Manual}, for more
10601 information about this pragma.
10603 @code{gnatelim} needs as its input data the name of the main subprogram
10604 and a bind file for a main subprogram.
10606 To create a bind file for @code{gnatelim}, run @code{gnatbind} for
10607 the main subprogram. @code{gnatelim} can work with both Ada and C
10608 bind files; when both are present, it uses the Ada bind file.
10609 The following commands will build the program and create the bind file:
10612 $ gnatmake ^-c Main_Prog^/ACTIONS=COMPILE MAIN_PROG^
10613 $ gnatbind main_prog
10616 Note that @code{gnatelim} needs neither object nor ALI files.
10618 @node Running gnatelim
10619 @subsection Running @code{gnatelim}
10622 @code{gnatelim} has the following command-line interface:
10625 $ gnatelim @ovar{options} name
10629 @code{name} should be a name of a source file that contains the main subprogram
10630 of a program (partition).
10632 @code{gnatelim} has the following switches:
10637 @cindex @option{^-q^/QUIET^} (@command{gnatelim})
10638 Quiet mode: by default @code{gnatelim} outputs to the standard error
10639 stream the number of program units left to be processed. This option turns
10642 @item ^-v^/VERBOSE^
10643 @cindex @option{^-v^/VERBOSE^} (@command{gnatelim})
10644 Verbose mode: @code{gnatelim} version information is printed as Ada
10645 comments to the standard output stream. Also, in addition to the number of
10646 program units left @code{gnatelim} will output the name of the current unit
10650 @cindex @option{^-a^/ALL^} (@command{gnatelim})
10651 Also look for subprograms from the GNAT run time that can be eliminated. Note
10652 that when @file{gnat.adc} is produced using this switch, the entire program
10653 must be recompiled with switch @option{^-a^/ALL_FILES^} to @command{gnatmake}.
10655 @item ^-I^/INCLUDE_DIRS=^@var{dir}
10656 @cindex @option{^-I^/INCLUDE_DIRS^} (@command{gnatelim})
10657 When looking for source files also look in directory @var{dir}. Specifying
10658 @option{^-I-^/INCLUDE_DIRS=-^} instructs @code{gnatelim} not to look for
10659 sources in the current directory.
10661 @item ^-b^/BIND_FILE=^@var{bind_file}
10662 @cindex @option{^-b^/BIND_FILE^} (@command{gnatelim})
10663 Specifies @var{bind_file} as the bind file to process. If not set, the name
10664 of the bind file is computed from the full expanded Ada name
10665 of a main subprogram.
10667 @item ^-C^/CONFIG_FILE=^@var{config_file}
10668 @cindex @option{^-C^/CONFIG_FILE^} (@command{gnatelim})
10669 Specifies a file @var{config_file} that contains configuration pragmas. The
10670 file must be specified with full path.
10672 @item ^--GCC^/COMPILER^=@var{compiler_name}
10673 @cindex @option{^-GCC^/COMPILER^} (@command{gnatelim})
10674 Instructs @code{gnatelim} to use specific @command{gcc} compiler instead of one
10675 available on the path.
10677 @item ^--GNATMAKE^/GNATMAKE^=@var{gnatmake_name}
10678 @cindex @option{^--GNATMAKE^/GNATMAKE^} (@command{gnatelim})
10679 Instructs @code{gnatelim} to use specific @command{gnatmake} instead of one
10680 available on the path.
10684 @code{gnatelim} sends its output to the standard output stream, and all the
10685 tracing and debug information is sent to the standard error stream.
10686 In order to produce a proper GNAT configuration file
10687 @file{gnat.adc}, redirection must be used:
10691 $ PIPE GNAT ELIM MAIN_PROG.ADB > GNAT.ADC
10694 $ gnatelim main_prog.adb > gnat.adc
10703 $ gnatelim main_prog.adb >> gnat.adc
10707 in order to append the @code{gnatelim} output to the existing contents of
10711 @node Correcting the List of Eliminate Pragmas
10712 @subsection Correcting the List of Eliminate Pragmas
10715 In some rare cases @code{gnatelim} may try to eliminate
10716 subprograms that are actually called in the program. In this case, the
10717 compiler will generate an error message of the form:
10720 file.adb:106:07: cannot call eliminated subprogram "My_Prog"
10724 You will need to manually remove the wrong @code{Eliminate} pragmas from
10725 the @file{gnat.adc} file. You should recompile your program
10726 from scratch after that, because you need a consistent @file{gnat.adc} file
10727 during the entire compilation.
10729 @node Making Your Executables Smaller
10730 @subsection Making Your Executables Smaller
10733 In order to get a smaller executable for your program you now have to
10734 recompile the program completely with the new @file{gnat.adc} file
10735 created by @code{gnatelim} in your current directory:
10738 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10742 (Use the @option{^-f^/FORCE_COMPILE^} option for @command{gnatmake} to
10743 recompile everything
10744 with the set of pragmas @code{Eliminate} that you have obtained with
10745 @command{gnatelim}).
10747 Be aware that the set of @code{Eliminate} pragmas is specific to each
10748 program. It is not recommended to merge sets of @code{Eliminate}
10749 pragmas created for different programs in one @file{gnat.adc} file.
10751 @node Summary of the gnatelim Usage Cycle
10752 @subsection Summary of the gnatelim Usage Cycle
10755 Here is a quick summary of the steps to be taken in order to reduce
10756 the size of your executables with @code{gnatelim}. You may use
10757 other GNAT options to control the optimization level,
10758 to produce the debugging information, to set search path, etc.
10762 Produce a bind file
10765 $ gnatmake ^-c main_prog^/ACTIONS=COMPILE MAIN_PROG^
10766 $ gnatbind main_prog
10770 Generate a list of @code{Eliminate} pragmas
10773 $ PIPE GNAT ELIM MAIN_PROG > GNAT.ADC
10776 $ gnatelim main_prog >@r{[}>@r{]} gnat.adc
10781 Recompile the application
10784 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10789 @node Reducing Size of Executables with unused subprogram/data elimination
10790 @section Reducing Size of Executables with Unused Subprogram/Data Elimination
10791 @findex unused subprogram/data elimination
10794 This section describes how you can eliminate unused subprograms and data from
10795 your executable just by setting options at compilation time.
10798 * About unused subprogram/data elimination::
10799 * Compilation options::
10800 * Example of unused subprogram/data elimination::
10803 @node About unused subprogram/data elimination
10804 @subsection About unused subprogram/data elimination
10807 By default, an executable contains all code and data of its composing objects
10808 (directly linked or coming from statically linked libraries), even data or code
10809 never used by this executable.
10811 This feature will allow you to eliminate such unused code from your
10812 executable, making it smaller (in disk and in memory).
10814 This functionality is available on all Linux platforms except for the IA-64
10815 architecture and on all cross platforms using the ELF binary file format.
10816 In both cases GNU binutils version 2.16 or later are required to enable it.
10818 @node Compilation options
10819 @subsection Compilation options
10822 The operation of eliminating the unused code and data from the final executable
10823 is directly performed by the linker.
10825 In order to do this, it has to work with objects compiled with the
10827 @option{-ffunction-sections} @option{-fdata-sections}.
10828 @cindex @option{-ffunction-sections} (@command{gcc})
10829 @cindex @option{-fdata-sections} (@command{gcc})
10830 These options are usable with C and Ada files.
10831 They will place respectively each
10832 function or data in a separate section in the resulting object file.
10834 Once the objects and static libraries are created with these options, the
10835 linker can perform the dead code elimination. You can do this by setting
10836 the @option{-Wl,--gc-sections} option to gcc command or in the
10837 @option{-largs} section of @command{gnatmake}. This will perform a
10838 garbage collection of code and data never referenced.
10840 If the linker performs a partial link (@option{-r} ld linker option), then you
10841 will need to provide one or several entry point using the
10842 @option{-e} / @option{--entry} ld option.
10844 Note that objects compiled without the @option{-ffunction-sections} and
10845 @option{-fdata-sections} options can still be linked with the executable.
10846 However, no dead code elimination will be performed on those objects (they will
10849 The GNAT static library is now compiled with -ffunction-sections and
10850 -fdata-sections on some platforms. This allows you to eliminate the unused code
10851 and data of the GNAT library from your executable.
10853 @node Example of unused subprogram/data elimination
10854 @subsection Example of unused subprogram/data elimination
10857 Here is a simple example:
10859 @smallexample @c ada
10868 Used_Data : Integer;
10869 Unused_Data : Integer;
10871 procedure Used (Data : Integer);
10872 procedure Unused (Data : Integer);
10875 package body Aux is
10876 procedure Used (Data : Integer) is
10881 procedure Unused (Data : Integer) is
10883 Unused_Data := Data;
10889 @code{Unused} and @code{Unused_Data} are never referenced in this code
10890 excerpt, and hence they may be safely removed from the final executable.
10895 $ nm test | grep used
10896 020015f0 T aux__unused
10897 02005d88 B aux__unused_data
10898 020015cc T aux__used
10899 02005d84 B aux__used_data
10901 $ gnatmake test -cargs -fdata-sections -ffunction-sections \
10902 -largs -Wl,--gc-sections
10904 $ nm test | grep used
10905 02005350 T aux__used
10906 0201ffe0 B aux__used_data
10910 It can be observed that the procedure @code{Unused} and the object
10911 @code{Unused_Data} are removed by the linker when using the
10912 appropriate options.
10914 @c ********************************
10915 @node Renaming Files Using gnatchop
10916 @chapter Renaming Files Using @code{gnatchop}
10920 This chapter discusses how to handle files with multiple units by using
10921 the @code{gnatchop} utility. This utility is also useful in renaming
10922 files to meet the standard GNAT default file naming conventions.
10925 * Handling Files with Multiple Units::
10926 * Operating gnatchop in Compilation Mode::
10927 * Command Line for gnatchop::
10928 * Switches for gnatchop::
10929 * Examples of gnatchop Usage::
10932 @node Handling Files with Multiple Units
10933 @section Handling Files with Multiple Units
10936 The basic compilation model of GNAT requires that a file submitted to the
10937 compiler have only one unit and there be a strict correspondence
10938 between the file name and the unit name.
10940 The @code{gnatchop} utility allows both of these rules to be relaxed,
10941 allowing GNAT to process files which contain multiple compilation units
10942 and files with arbitrary file names. @code{gnatchop}
10943 reads the specified file and generates one or more output files,
10944 containing one unit per file. The unit and the file name correspond,
10945 as required by GNAT.
10947 If you want to permanently restructure a set of ``foreign'' files so that
10948 they match the GNAT rules, and do the remaining development using the
10949 GNAT structure, you can simply use @command{gnatchop} once, generate the
10950 new set of files and work with them from that point on.
10952 Alternatively, if you want to keep your files in the ``foreign'' format,
10953 perhaps to maintain compatibility with some other Ada compilation
10954 system, you can set up a procedure where you use @command{gnatchop} each
10955 time you compile, regarding the source files that it writes as temporary
10956 files that you throw away.
10958 Note that if your file containing multiple units starts with a byte order
10959 mark (BOM) specifying UTF-8 encoding, then the files generated by gnatchop
10960 will each start with a copy of this BOM, meaning that they can be compiled
10961 automatically in UTF-8 mode without needing to specify an explicit encoding.
10963 @node Operating gnatchop in Compilation Mode
10964 @section Operating gnatchop in Compilation Mode
10967 The basic function of @code{gnatchop} is to take a file with multiple units
10968 and split it into separate files. The boundary between files is reasonably
10969 clear, except for the issue of comments and pragmas. In default mode, the
10970 rule is that any pragmas between units belong to the previous unit, except
10971 that configuration pragmas always belong to the following unit. Any comments
10972 belong to the following unit. These rules
10973 almost always result in the right choice of
10974 the split point without needing to mark it explicitly and most users will
10975 find this default to be what they want. In this default mode it is incorrect to
10976 submit a file containing only configuration pragmas, or one that ends in
10977 configuration pragmas, to @code{gnatchop}.
10979 However, using a special option to activate ``compilation mode'',
10981 can perform another function, which is to provide exactly the semantics
10982 required by the RM for handling of configuration pragmas in a compilation.
10983 In the absence of configuration pragmas (at the main file level), this
10984 option has no effect, but it causes such configuration pragmas to be handled
10985 in a quite different manner.
10987 First, in compilation mode, if @code{gnatchop} is given a file that consists of
10988 only configuration pragmas, then this file is appended to the
10989 @file{gnat.adc} file in the current directory. This behavior provides
10990 the required behavior described in the RM for the actions to be taken
10991 on submitting such a file to the compiler, namely that these pragmas
10992 should apply to all subsequent compilations in the same compilation
10993 environment. Using GNAT, the current directory, possibly containing a
10994 @file{gnat.adc} file is the representation
10995 of a compilation environment. For more information on the
10996 @file{gnat.adc} file, see @ref{Handling of Configuration Pragmas}.
10998 Second, in compilation mode, if @code{gnatchop}
10999 is given a file that starts with
11000 configuration pragmas, and contains one or more units, then these
11001 configuration pragmas are prepended to each of the chopped files. This
11002 behavior provides the required behavior described in the RM for the
11003 actions to be taken on compiling such a file, namely that the pragmas
11004 apply to all units in the compilation, but not to subsequently compiled
11007 Finally, if configuration pragmas appear between units, they are appended
11008 to the previous unit. This results in the previous unit being illegal,
11009 since the compiler does not accept configuration pragmas that follow
11010 a unit. This provides the required RM behavior that forbids configuration
11011 pragmas other than those preceding the first compilation unit of a
11014 For most purposes, @code{gnatchop} will be used in default mode. The
11015 compilation mode described above is used only if you need exactly
11016 accurate behavior with respect to compilations, and you have files
11017 that contain multiple units and configuration pragmas. In this
11018 circumstance the use of @code{gnatchop} with the compilation mode
11019 switch provides the required behavior, and is for example the mode
11020 in which GNAT processes the ACVC tests.
11022 @node Command Line for gnatchop
11023 @section Command Line for @code{gnatchop}
11026 The @code{gnatchop} command has the form:
11029 $ gnatchop switches @var{file name} @r{[}@var{file name} @dots{}@r{]}
11034 The only required argument is the file name of the file to be chopped.
11035 There are no restrictions on the form of this file name. The file itself
11036 contains one or more Ada units, in normal GNAT format, concatenated
11037 together. As shown, more than one file may be presented to be chopped.
11039 When run in default mode, @code{gnatchop} generates one output file in
11040 the current directory for each unit in each of the files.
11042 @var{directory}, if specified, gives the name of the directory to which
11043 the output files will be written. If it is not specified, all files are
11044 written to the current directory.
11046 For example, given a
11047 file called @file{hellofiles} containing
11049 @smallexample @c ada
11054 with Text_IO; use Text_IO;
11057 Put_Line ("Hello");
11067 $ gnatchop ^hellofiles^HELLOFILES.^
11071 generates two files in the current directory, one called
11072 @file{hello.ads} containing the single line that is the procedure spec,
11073 and the other called @file{hello.adb} containing the remaining text. The
11074 original file is not affected. The generated files can be compiled in
11078 When gnatchop is invoked on a file that is empty or that contains only empty
11079 lines and/or comments, gnatchop will not fail, but will not produce any
11082 For example, given a
11083 file called @file{toto.txt} containing
11085 @smallexample @c ada
11097 $ gnatchop ^toto.txt^TOT.TXT^
11101 will not produce any new file and will result in the following warnings:
11104 toto.txt:1:01: warning: empty file, contains no compilation units
11105 no compilation units found
11106 no source files written
11109 @node Switches for gnatchop
11110 @section Switches for @code{gnatchop}
11113 @command{gnatchop} recognizes the following switches:
11119 @cindex @option{--version} @command{gnatchop}
11120 Display Copyright and version, then exit disregarding all other options.
11123 @cindex @option{--help} @command{gnatchop}
11124 If @option{--version} was not used, display usage, then exit disregarding
11127 @item ^-c^/COMPILATION^
11128 @cindex @option{^-c^/COMPILATION^} (@code{gnatchop})
11129 Causes @code{gnatchop} to operate in compilation mode, in which
11130 configuration pragmas are handled according to strict RM rules. See
11131 previous section for a full description of this mode.
11134 @item -gnat@var{xxx}
11135 This passes the given @option{-gnat@var{xxx}} switch to @code{gnat} which is
11136 used to parse the given file. Not all @var{xxx} options make sense,
11137 but for example, the use of @option{-gnati2} allows @code{gnatchop} to
11138 process a source file that uses Latin-2 coding for identifiers.
11142 Causes @code{gnatchop} to generate a brief help summary to the standard
11143 output file showing usage information.
11145 @item ^-k@var{mm}^/FILE_NAME_MAX_LENGTH=@var{mm}^
11146 @cindex @option{^-k^/FILE_NAME_MAX_LENGTH^} (@code{gnatchop})
11147 Limit generated file names to the specified number @code{mm}
11149 This is useful if the
11150 resulting set of files is required to be interoperable with systems
11151 which limit the length of file names.
11153 If no value is given, or
11154 if no @code{/FILE_NAME_MAX_LENGTH} qualifier is given,
11155 a default of 39, suitable for OpenVMS Alpha
11156 Systems, is assumed
11159 No space is allowed between the @option{-k} and the numeric value. The numeric
11160 value may be omitted in which case a default of @option{-k8},
11162 with DOS-like file systems, is used. If no @option{-k} switch
11164 there is no limit on the length of file names.
11167 @item ^-p^/PRESERVE^
11168 @cindex @option{^-p^/PRESERVE^} (@code{gnatchop})
11169 Causes the file ^modification^creation^ time stamp of the input file to be
11170 preserved and used for the time stamp of the output file(s). This may be
11171 useful for preserving coherency of time stamps in an environment where
11172 @code{gnatchop} is used as part of a standard build process.
11175 @cindex @option{^-q^/QUIET^} (@code{gnatchop})
11176 Causes output of informational messages indicating the set of generated
11177 files to be suppressed. Warnings and error messages are unaffected.
11179 @item ^-r^/REFERENCE^
11180 @cindex @option{^-r^/REFERENCE^} (@code{gnatchop})
11181 @findex Source_Reference
11182 Generate @code{Source_Reference} pragmas. Use this switch if the output
11183 files are regarded as temporary and development is to be done in terms
11184 of the original unchopped file. This switch causes
11185 @code{Source_Reference} pragmas to be inserted into each of the
11186 generated files to refers back to the original file name and line number.
11187 The result is that all error messages refer back to the original
11189 In addition, the debugging information placed into the object file (when
11190 the @option{^-g^/DEBUG^} switch of @command{gcc} or @command{gnatmake} is
11192 also refers back to this original file so that tools like profilers and
11193 debuggers will give information in terms of the original unchopped file.
11195 If the original file to be chopped itself contains
11196 a @code{Source_Reference}
11197 pragma referencing a third file, then gnatchop respects
11198 this pragma, and the generated @code{Source_Reference} pragmas
11199 in the chopped file refer to the original file, with appropriate
11200 line numbers. This is particularly useful when @code{gnatchop}
11201 is used in conjunction with @code{gnatprep} to compile files that
11202 contain preprocessing statements and multiple units.
11204 @item ^-v^/VERBOSE^
11205 @cindex @option{^-v^/VERBOSE^} (@code{gnatchop})
11206 Causes @code{gnatchop} to operate in verbose mode. The version
11207 number and copyright notice are output, as well as exact copies of
11208 the gnat1 commands spawned to obtain the chop control information.
11210 @item ^-w^/OVERWRITE^
11211 @cindex @option{^-w^/OVERWRITE^} (@code{gnatchop})
11212 Overwrite existing file names. Normally @code{gnatchop} regards it as a
11213 fatal error if there is already a file with the same name as a
11214 file it would otherwise output, in other words if the files to be
11215 chopped contain duplicated units. This switch bypasses this
11216 check, and causes all but the last instance of such duplicated
11217 units to be skipped.
11220 @item --GCC=@var{xxxx}
11221 @cindex @option{--GCC=} (@code{gnatchop})
11222 Specify the path of the GNAT parser to be used. When this switch is used,
11223 no attempt is made to add the prefix to the GNAT parser executable.
11227 @node Examples of gnatchop Usage
11228 @section Examples of @code{gnatchop} Usage
11232 @item gnatchop /OVERWRITE HELLO_S.ADA [PRERELEASE.FILES]
11235 @item gnatchop -w hello_s.ada prerelease/files
11238 Chops the source file @file{hello_s.ada}. The output files will be
11239 placed in the directory @file{^prerelease/files^[PRERELEASE.FILES]^},
11241 files with matching names in that directory (no files in the current
11242 directory are modified).
11244 @item gnatchop ^archive^ARCHIVE.^
11245 Chops the source file @file{^archive^ARCHIVE.^}
11246 into the current directory. One
11247 useful application of @code{gnatchop} is in sending sets of sources
11248 around, for example in email messages. The required sources are simply
11249 concatenated (for example, using a ^Unix @code{cat}^VMS @code{APPEND/NEW}^
11251 @command{gnatchop} is used at the other end to reconstitute the original
11254 @item gnatchop file1 file2 file3 direc
11255 Chops all units in files @file{file1}, @file{file2}, @file{file3}, placing
11256 the resulting files in the directory @file{direc}. Note that if any units
11257 occur more than once anywhere within this set of files, an error message
11258 is generated, and no files are written. To override this check, use the
11259 @option{^-w^/OVERWRITE^} switch,
11260 in which case the last occurrence in the last file will
11261 be the one that is output, and earlier duplicate occurrences for a given
11262 unit will be skipped.
11265 @node Configuration Pragmas
11266 @chapter Configuration Pragmas
11267 @cindex Configuration pragmas
11268 @cindex Pragmas, configuration
11271 Configuration pragmas include those pragmas described as
11272 such in the Ada Reference Manual, as well as
11273 implementation-dependent pragmas that are configuration pragmas.
11274 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
11275 for details on these additional GNAT-specific configuration pragmas.
11276 Most notably, the pragma @code{Source_File_Name}, which allows
11277 specifying non-default names for source files, is a configuration
11278 pragma. The following is a complete list of configuration pragmas
11279 recognized by GNAT:
11291 Compile_Time_Warning
11293 Component_Alignment
11300 External_Name_Casing
11303 Float_Representation
11316 Priority_Specific_Dispatching
11319 Propagate_Exceptions
11322 Restricted_Run_Time
11324 Restrictions_Warnings
11327 Source_File_Name_Project
11330 Suppress_Exception_Locations
11331 Task_Dispatching_Policy
11337 Wide_Character_Encoding
11342 * Handling of Configuration Pragmas::
11343 * The Configuration Pragmas Files::
11346 @node Handling of Configuration Pragmas
11347 @section Handling of Configuration Pragmas
11349 Configuration pragmas may either appear at the start of a compilation
11350 unit, in which case they apply only to that unit, or they may apply to
11351 all compilations performed in a given compilation environment.
11353 GNAT also provides the @code{gnatchop} utility to provide an automatic
11354 way to handle configuration pragmas following the semantics for
11355 compilations (that is, files with multiple units), described in the RM.
11356 See @ref{Operating gnatchop in Compilation Mode} for details.
11357 However, for most purposes, it will be more convenient to edit the
11358 @file{gnat.adc} file that contains configuration pragmas directly,
11359 as described in the following section.
11361 @node The Configuration Pragmas Files
11362 @section The Configuration Pragmas Files
11363 @cindex @file{gnat.adc}
11366 In GNAT a compilation environment is defined by the current
11367 directory at the time that a compile command is given. This current
11368 directory is searched for a file whose name is @file{gnat.adc}. If
11369 this file is present, it is expected to contain one or more
11370 configuration pragmas that will be applied to the current compilation.
11371 However, if the switch @option{-gnatA} is used, @file{gnat.adc} is not
11374 Configuration pragmas may be entered into the @file{gnat.adc} file
11375 either by running @code{gnatchop} on a source file that consists only of
11376 configuration pragmas, or more conveniently by
11377 direct editing of the @file{gnat.adc} file, which is a standard format
11380 In addition to @file{gnat.adc}, additional files containing configuration
11381 pragmas may be applied to the current compilation using the switch
11382 @option{-gnatec}@var{path}. @var{path} must designate an existing file that
11383 contains only configuration pragmas. These configuration pragmas are
11384 in addition to those found in @file{gnat.adc} (provided @file{gnat.adc}
11385 is present and switch @option{-gnatA} is not used).
11387 It is allowed to specify several switches @option{-gnatec}, all of which
11388 will be taken into account.
11390 If you are using project file, a separate mechanism is provided using
11391 project attributes, see @ref{Specifying Configuration Pragmas} for more
11395 Of special interest to GNAT OpenVMS Alpha is the following
11396 configuration pragma:
11398 @smallexample @c ada
11400 pragma Extend_System (Aux_DEC);
11405 In the presence of this pragma, GNAT adds to the definition of the
11406 predefined package SYSTEM all the additional types and subprograms that are
11407 defined in HP Ada. See @ref{Compatibility with HP Ada} for details.
11410 @node Handling Arbitrary File Naming Conventions Using gnatname
11411 @chapter Handling Arbitrary File Naming Conventions Using @code{gnatname}
11412 @cindex Arbitrary File Naming Conventions
11415 * Arbitrary File Naming Conventions::
11416 * Running gnatname::
11417 * Switches for gnatname::
11418 * Examples of gnatname Usage::
11421 @node Arbitrary File Naming Conventions
11422 @section Arbitrary File Naming Conventions
11425 The GNAT compiler must be able to know the source file name of a compilation
11426 unit. When using the standard GNAT default file naming conventions
11427 (@code{.ads} for specs, @code{.adb} for bodies), the GNAT compiler
11428 does not need additional information.
11431 When the source file names do not follow the standard GNAT default file naming
11432 conventions, the GNAT compiler must be given additional information through
11433 a configuration pragmas file (@pxref{Configuration Pragmas})
11435 When the non-standard file naming conventions are well-defined,
11436 a small number of pragmas @code{Source_File_Name} specifying a naming pattern
11437 (@pxref{Alternative File Naming Schemes}) may be sufficient. However,
11438 if the file naming conventions are irregular or arbitrary, a number
11439 of pragma @code{Source_File_Name} for individual compilation units
11441 To help maintain the correspondence between compilation unit names and
11442 source file names within the compiler,
11443 GNAT provides a tool @code{gnatname} to generate the required pragmas for a
11446 @node Running gnatname
11447 @section Running @code{gnatname}
11450 The usual form of the @code{gnatname} command is
11453 $ gnatname @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}
11454 @r{[}--and @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}@r{]}
11458 All of the arguments are optional. If invoked without any argument,
11459 @code{gnatname} will display its usage.
11462 When used with at least one naming pattern, @code{gnatname} will attempt to
11463 find all the compilation units in files that follow at least one of the
11464 naming patterns. To find these compilation units,
11465 @code{gnatname} will use the GNAT compiler in syntax-check-only mode on all
11469 One or several Naming Patterns may be given as arguments to @code{gnatname}.
11470 Each Naming Pattern is enclosed between double quotes.
11471 A Naming Pattern is a regular expression similar to the wildcard patterns
11472 used in file names by the Unix shells or the DOS prompt.
11475 @code{gnatname} may be called with several sections of directories/patterns.
11476 Sections are separated by switch @code{--and}. In each section, there must be
11477 at least one pattern. If no directory is specified in a section, the current
11478 directory (or the project directory is @code{-P} is used) is implied.
11479 The options other that the directory switches and the patterns apply globally
11480 even if they are in different sections.
11483 Examples of Naming Patterns are
11492 For a more complete description of the syntax of Naming Patterns,
11493 see the second kind of regular expressions described in @file{g-regexp.ads}
11494 (the ``Glob'' regular expressions).
11497 When invoked with no switch @code{-P}, @code{gnatname} will create a
11498 configuration pragmas file @file{gnat.adc} in the current working directory,
11499 with pragmas @code{Source_File_Name} for each file that contains a valid Ada
11502 @node Switches for gnatname
11503 @section Switches for @code{gnatname}
11506 Switches for @code{gnatname} must precede any specified Naming Pattern.
11509 You may specify any of the following switches to @code{gnatname}:
11515 @cindex @option{--version} @command{gnatname}
11516 Display Copyright and version, then exit disregarding all other options.
11519 @cindex @option{--help} @command{gnatname}
11520 If @option{--version} was not used, display usage, then exit disregarding
11524 Start another section of directories/patterns.
11526 @item ^-c^/CONFIG_FILE=^@file{file}
11527 @cindex @option{^-c^/CONFIG_FILE^} (@code{gnatname})
11528 Create a configuration pragmas file @file{file} (instead of the default
11531 There may be zero, one or more space between @option{-c} and
11534 @file{file} may include directory information. @file{file} must be
11535 writable. There may be only one switch @option{^-c^/CONFIG_FILE^}.
11536 When a switch @option{^-c^/CONFIG_FILE^} is
11537 specified, no switch @option{^-P^/PROJECT_FILE^} may be specified (see below).
11539 @item ^-d^/SOURCE_DIRS=^@file{dir}
11540 @cindex @option{^-d^/SOURCE_DIRS^} (@code{gnatname})
11541 Look for source files in directory @file{dir}. There may be zero, one or more
11542 spaces between @option{^-d^/SOURCE_DIRS=^} and @file{dir}.
11543 When a switch @option{^-d^/SOURCE_DIRS^}
11544 is specified, the current working directory will not be searched for source
11545 files, unless it is explicitly specified with a @option{^-d^/SOURCE_DIRS^}
11546 or @option{^-D^/DIR_FILES^} switch.
11547 Several switches @option{^-d^/SOURCE_DIRS^} may be specified.
11548 If @file{dir} is a relative path, it is relative to the directory of
11549 the configuration pragmas file specified with switch
11550 @option{^-c^/CONFIG_FILE^},
11551 or to the directory of the project file specified with switch
11552 @option{^-P^/PROJECT_FILE^} or,
11553 if neither switch @option{^-c^/CONFIG_FILE^}
11554 nor switch @option{^-P^/PROJECT_FILE^} are specified, it is relative to the
11555 current working directory. The directory
11556 specified with switch @option{^-d^/SOURCE_DIRS^} must exist and be readable.
11558 @item ^-D^/DIRS_FILE=^@file{file}
11559 @cindex @option{^-D^/DIRS_FILE^} (@code{gnatname})
11560 Look for source files in all directories listed in text file @file{file}.
11561 There may be zero, one or more spaces between @option{^-D^/DIRS_FILE=^}
11563 @file{file} must be an existing, readable text file.
11564 Each nonempty line in @file{file} must be a directory.
11565 Specifying switch @option{^-D^/DIRS_FILE^} is equivalent to specifying as many
11566 switches @option{^-d^/SOURCE_DIRS^} as there are nonempty lines in
11569 @item ^-f^/FOREIGN_PATTERN=^@file{pattern}
11570 @cindex @option{^-f^/FOREIGN_PATTERN^} (@code{gnatname})
11571 Foreign patterns. Using this switch, it is possible to add sources of languages
11572 other than Ada to the list of sources of a project file.
11573 It is only useful if a ^-P^/PROJECT_FILE^ switch is used.
11576 gnatname ^-Pprj -f"*.c"^/PROJECT_FILE=PRJ /FOREIGN_PATTERN=*.C^ "*.ada"
11579 will look for Ada units in all files with the @file{.ada} extension,
11580 and will add to the list of file for project @file{prj.gpr} the C files
11581 with extension @file{.^c^C^}.
11584 @cindex @option{^-h^/HELP^} (@code{gnatname})
11585 Output usage (help) information. The output is written to @file{stdout}.
11587 @item ^-P^/PROJECT_FILE=^@file{proj}
11588 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatname})
11589 Create or update project file @file{proj}. There may be zero, one or more space
11590 between @option{-P} and @file{proj}. @file{proj} may include directory
11591 information. @file{proj} must be writable.
11592 There may be only one switch @option{^-P^/PROJECT_FILE^}.
11593 When a switch @option{^-P^/PROJECT_FILE^} is specified,
11594 no switch @option{^-c^/CONFIG_FILE^} may be specified.
11596 @item ^-v^/VERBOSE^
11597 @cindex @option{^-v^/VERBOSE^} (@code{gnatname})
11598 Verbose mode. Output detailed explanation of behavior to @file{stdout}.
11599 This includes name of the file written, the name of the directories to search
11600 and, for each file in those directories whose name matches at least one of
11601 the Naming Patterns, an indication of whether the file contains a unit,
11602 and if so the name of the unit.
11604 @item ^-v -v^/VERBOSE /VERBOSE^
11605 @cindex @option{^-v -v^/VERBOSE /VERBOSE^} (@code{gnatname})
11606 Very Verbose mode. In addition to the output produced in verbose mode,
11607 for each file in the searched directories whose name matches none of
11608 the Naming Patterns, an indication is given that there is no match.
11610 @item ^-x^/EXCLUDED_PATTERN=^@file{pattern}
11611 @cindex @option{^-x^/EXCLUDED_PATTERN^} (@code{gnatname})
11612 Excluded patterns. Using this switch, it is possible to exclude some files
11613 that would match the name patterns. For example,
11615 gnatname ^-x "*_nt.ada"^/EXCLUDED_PATTERN=*_nt.ada^ "*.ada"
11618 will look for Ada units in all files with the @file{.ada} extension,
11619 except those whose names end with @file{_nt.ada}.
11623 @node Examples of gnatname Usage
11624 @section Examples of @code{gnatname} Usage
11628 $ gnatname /CONFIG_FILE=[HOME.ME]NAMES.ADC /SOURCE_DIRS=SOURCES "[a-z]*.ada*"
11634 $ gnatname -c /home/me/names.adc -d sources "[a-z]*.ada*"
11639 In this example, the directory @file{^/home/me^[HOME.ME]^} must already exist
11640 and be writable. In addition, the directory
11641 @file{^/home/me/sources^[HOME.ME.SOURCES]^} (specified by
11642 @option{^-d sources^/SOURCE_DIRS=SOURCES^}) must exist and be readable.
11645 Note the optional spaces after @option{-c} and @option{-d}.
11650 $ gnatname -P/home/me/proj -x "*_nt_body.ada"
11651 -dsources -dsources/plus -Dcommon_dirs.txt "body_*" "spec_*"
11654 $ gnatname /PROJECT_FILE=[HOME.ME]PROJ
11655 /EXCLUDED_PATTERN=*_nt_body.ada
11656 /SOURCE_DIRS=(SOURCES,[SOURCES.PLUS])
11657 /DIRS_FILE=COMMON_DIRS.TXT "body_*" "spec_*"
11661 Note that several switches @option{^-d^/SOURCE_DIRS^} may be used,
11662 even in conjunction with one or several switches
11663 @option{^-D^/DIRS_FILE^}. Several Naming Patterns and one excluded pattern
11664 are used in this example.
11666 @c *****************************************
11667 @c * G N A T P r o j e c t M a n a g e r *
11668 @c *****************************************
11669 @node GNAT Project Manager
11670 @chapter GNAT Project Manager
11674 * Examples of Project Files::
11675 * Project File Syntax::
11676 * Objects and Sources in Project Files::
11677 * Importing Projects::
11678 * Project Extension::
11679 * Project Hierarchy Extension::
11680 * External References in Project Files::
11681 * Packages in Project Files::
11682 * Variables from Imported Projects::
11684 * Library Projects::
11685 * Stand-alone Library Projects::
11686 * Switches Related to Project Files::
11687 * Tools Supporting Project Files::
11688 * An Extended Example::
11689 * Project File Complete Syntax::
11692 @c ****************
11693 @c * Introduction *
11694 @c ****************
11697 @section Introduction
11700 This chapter describes GNAT's @emph{Project Manager}, a facility that allows
11701 you to manage complex builds involving a number of source files, directories,
11702 and compilation options for different system configurations. In particular,
11703 project files allow you to specify:
11706 The directory or set of directories containing the source files, and/or the
11707 names of the specific source files themselves
11709 The directory in which the compiler's output
11710 (@file{ALI} files, object files, tree files) is to be placed
11712 The directory in which the executable programs is to be placed
11714 ^Switch^Switch^ settings for any of the project-enabled tools
11715 (@command{gnatmake}, compiler, binder, linker, @code{gnatls}, @code{gnatxref},
11716 @code{gnatfind}); you can apply these settings either globally or to individual
11719 The source files containing the main subprogram(s) to be built
11721 The source programming language(s) (currently Ada and/or C)
11723 Source file naming conventions; you can specify these either globally or for
11724 individual compilation units
11731 @node Project Files
11732 @subsection Project Files
11735 Project files are written in a syntax close to that of Ada, using familiar
11736 notions such as packages, context clauses, declarations, default values,
11737 assignments, and inheritance. Finally, project files can be built
11738 hierarchically from other project files, simplifying complex system
11739 integration and project reuse.
11741 A @dfn{project} is a specific set of values for various compilation properties.
11742 The settings for a given project are described by means of
11743 a @dfn{project file}, which is a text file written in an Ada-like syntax.
11744 Property values in project files are either strings or lists of strings.
11745 Properties that are not explicitly set receive default values. A project
11746 file may interrogate the values of @dfn{external variables} (user-defined
11747 command-line switches or environment variables), and it may specify property
11748 settings conditionally, based on the value of such variables.
11750 In simple cases, a project's source files depend only on other source files
11751 in the same project, or on the predefined libraries. (@emph{Dependence} is
11753 the Ada technical sense; as in one Ada unit @code{with}ing another.) However,
11754 the Project Manager also allows more sophisticated arrangements,
11755 where the source files in one project depend on source files in other
11759 One project can @emph{import} other projects containing needed source files.
11761 You can organize GNAT projects in a hierarchy: a @emph{child} project
11762 can extend a @emph{parent} project, inheriting the parent's source files and
11763 optionally overriding any of them with alternative versions
11767 More generally, the Project Manager lets you structure large development
11768 efforts into hierarchical subsystems, where build decisions are delegated
11769 to the subsystem level, and thus different compilation environments
11770 (^switch^switch^ settings) used for different subsystems.
11772 The Project Manager is invoked through the
11773 @option{^-P^/PROJECT_FILE=^@emph{projectfile}}
11774 switch to @command{gnatmake} or to the @command{^gnat^GNAT^} front driver.
11776 There may be zero, one or more spaces between @option{-P} and
11777 @option{@emph{projectfile}}.
11779 If you want to define (on the command line) an external variable that is
11780 queried by the project file, you must use the
11781 @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
11782 The Project Manager parses and interprets the project file, and drives the
11783 invoked tool based on the project settings.
11785 The Project Manager supports a wide range of development strategies,
11786 for systems of all sizes. Here are some typical practices that are
11790 Using a common set of source files, but generating object files in different
11791 directories via different ^switch^switch^ settings
11793 Using a mostly-shared set of source files, but with different versions of
11798 The destination of an executable can be controlled inside a project file
11799 using the @option{^-o^-o^}
11801 In the absence of such a ^switch^switch^ either inside
11802 the project file or on the command line, any executable files generated by
11803 @command{gnatmake} are placed in the directory @code{Exec_Dir} specified
11804 in the project file. If no @code{Exec_Dir} is specified, they will be placed
11805 in the object directory of the project.
11807 You can use project files to achieve some of the effects of a source
11808 versioning system (for example, defining separate projects for
11809 the different sets of sources that comprise different releases) but the
11810 Project Manager is independent of any source configuration management tools
11811 that might be used by the developers.
11813 The next section introduces the main features of GNAT's project facility
11814 through a sequence of examples; subsequent sections will present the syntax
11815 and semantics in more detail. A more formal description of the project
11816 facility appears in @ref{Project File Reference,,, gnat_rm, GNAT
11819 @c *****************************
11820 @c * Examples of Project Files *
11821 @c *****************************
11823 @node Examples of Project Files
11824 @section Examples of Project Files
11826 This section illustrates some of the typical uses of project files and
11827 explains their basic structure and behavior.
11830 * Common Sources with Different ^Switches^Switches^ and Directories::
11831 * Using External Variables::
11832 * Importing Other Projects::
11833 * Extending a Project::
11836 @node Common Sources with Different ^Switches^Switches^ and Directories
11837 @subsection Common Sources with Different ^Switches^Switches^ and Directories
11841 * Specifying the Object Directory::
11842 * Specifying the Exec Directory::
11843 * Project File Packages::
11844 * Specifying ^Switch^Switch^ Settings::
11845 * Main Subprograms::
11846 * Executable File Names::
11847 * Source File Naming Conventions::
11848 * Source Language(s)::
11852 Suppose that the Ada source files @file{pack.ads}, @file{pack.adb}, and
11853 @file{proc.adb} are in the @file{/common} directory. The file
11854 @file{proc.adb} contains an Ada main subprogram @code{Proc} that @code{with}s
11855 package @code{Pack}. We want to compile these source files under two sets
11856 of ^switches^switches^:
11859 When debugging, we want to pass the @option{-g} switch to @command{gnatmake},
11860 and the @option{^-gnata^-gnata^},
11861 @option{^-gnato^-gnato^},
11862 and @option{^-gnatE^-gnatE^} switches to the
11863 compiler; the compiler's output is to appear in @file{/common/debug}
11865 When preparing a release version, we want to pass the @option{^-O2^O2^} switch
11866 to the compiler; the compiler's output is to appear in @file{/common/release}
11870 The GNAT project files shown below, respectively @file{debug.gpr} and
11871 @file{release.gpr} in the @file{/common} directory, achieve these effects.
11884 ^/common/debug^[COMMON.DEBUG]^
11889 ^/common/release^[COMMON.RELEASE]^
11894 Here are the corresponding project files:
11896 @smallexample @c projectfile
11899 for Object_Dir use "debug";
11900 for Main use ("proc");
11903 for ^Default_Switches^Default_Switches^ ("Ada")
11905 for Executable ("proc.adb") use "proc1";
11910 package Compiler is
11911 for ^Default_Switches^Default_Switches^ ("Ada")
11912 use ("-fstack-check",
11915 "^-gnatE^-gnatE^");
11921 @smallexample @c projectfile
11924 for Object_Dir use "release";
11925 for Exec_Dir use ".";
11926 for Main use ("proc");
11928 package Compiler is
11929 for ^Default_Switches^Default_Switches^ ("Ada")
11937 The name of the project defined by @file{debug.gpr} is @code{"Debug"} (case
11938 insensitive), and analogously the project defined by @file{release.gpr} is
11939 @code{"Release"}. For consistency the file should have the same name as the
11940 project, and the project file's extension should be @code{"gpr"}. These
11941 conventions are not required, but a warning is issued if they are not followed.
11943 If the current directory is @file{^/temp^[TEMP]^}, then the command
11945 gnatmake ^-P/common/debug.gpr^/PROJECT_FILE=[COMMON]DEBUG^
11949 generates object and ALI files in @file{^/common/debug^[COMMON.DEBUG]^},
11950 as well as the @code{^proc1^PROC1.EXE^} executable,
11951 using the ^switch^switch^ settings defined in the project file.
11953 Likewise, the command
11955 gnatmake ^-P/common/release.gpr^/PROJECT_FILE=[COMMON]RELEASE^
11959 generates object and ALI files in @file{^/common/release^[COMMON.RELEASE]^},
11960 and the @code{^proc^PROC.EXE^}
11961 executable in @file{^/common^[COMMON]^},
11962 using the ^switch^switch^ settings from the project file.
11965 @unnumberedsubsubsec Source Files
11968 If a project file does not explicitly specify a set of source directories or
11969 a set of source files, then by default the project's source files are the
11970 Ada source files in the project file directory. Thus @file{pack.ads},
11971 @file{pack.adb}, and @file{proc.adb} are the source files for both projects.
11973 @node Specifying the Object Directory
11974 @unnumberedsubsubsec Specifying the Object Directory
11977 Several project properties are modeled by Ada-style @emph{attributes};
11978 a property is defined by supplying the equivalent of an Ada attribute
11979 definition clause in the project file.
11980 A project's object directory is another such a property; the corresponding
11981 attribute is @code{Object_Dir}, and its value is also a string expression,
11982 specified either as absolute or relative. In the later case,
11983 it is relative to the project file directory. Thus the compiler's
11984 output is directed to @file{^/common/debug^[COMMON.DEBUG]^}
11985 (for the @code{Debug} project)
11986 and to @file{^/common/release^[COMMON.RELEASE]^}
11987 (for the @code{Release} project).
11988 If @code{Object_Dir} is not specified, then the default is the project file
11991 @node Specifying the Exec Directory
11992 @unnumberedsubsubsec Specifying the Exec Directory
11995 A project's exec directory is another property; the corresponding
11996 attribute is @code{Exec_Dir}, and its value is also a string expression,
11997 either specified as relative or absolute. If @code{Exec_Dir} is not specified,
11998 then the default is the object directory (which may also be the project file
11999 directory if attribute @code{Object_Dir} is not specified). Thus the executable
12000 is placed in @file{^/common/debug^[COMMON.DEBUG]^}
12001 for the @code{Debug} project (attribute @code{Exec_Dir} not specified)
12002 and in @file{^/common^[COMMON]^} for the @code{Release} project.
12004 @node Project File Packages
12005 @unnumberedsubsubsec Project File Packages
12008 A GNAT tool that is integrated with the Project Manager is modeled by a
12009 corresponding package in the project file. In the example above,
12010 The @code{Debug} project defines the packages @code{Builder}
12011 (for @command{gnatmake}) and @code{Compiler};
12012 the @code{Release} project defines only the @code{Compiler} package.
12014 The Ada-like package syntax is not to be taken literally. Although packages in
12015 project files bear a surface resemblance to packages in Ada source code, the
12016 notation is simply a way to convey a grouping of properties for a named
12017 entity. Indeed, the package names permitted in project files are restricted
12018 to a predefined set, corresponding to the project-aware tools, and the contents
12019 of packages are limited to a small set of constructs.
12020 The packages in the example above contain attribute definitions.
12022 @node Specifying ^Switch^Switch^ Settings
12023 @unnumberedsubsubsec Specifying ^Switch^Switch^ Settings
12026 ^Switch^Switch^ settings for a project-aware tool can be specified through
12027 attributes in the package that corresponds to the tool.
12028 The example above illustrates one of the relevant attributes,
12029 @code{^Default_Switches^Default_Switches^}, which is defined in packages
12030 in both project files.
12031 Unlike simple attributes like @code{Source_Dirs},
12032 @code{^Default_Switches^Default_Switches^} is
12033 known as an @emph{associative array}. When you define this attribute, you must
12034 supply an ``index'' (a literal string), and the effect of the attribute
12035 definition is to set the value of the array at the specified index.
12036 For the @code{^Default_Switches^Default_Switches^} attribute,
12037 the index is a programming language (in our case, Ada),
12038 and the value specified (after @code{use}) must be a list
12039 of string expressions.
12041 The attributes permitted in project files are restricted to a predefined set.
12042 Some may appear at project level, others in packages.
12043 For any attribute that is an associative array, the index must always be a
12044 literal string, but the restrictions on this string (e.g., a file name or a
12045 language name) depend on the individual attribute.
12046 Also depending on the attribute, its specified value will need to be either a
12047 string or a string list.
12049 In the @code{Debug} project, we set the switches for two tools,
12050 @command{gnatmake} and the compiler, and thus we include the two corresponding
12051 packages; each package defines the @code{^Default_Switches^Default_Switches^}
12052 attribute with index @code{"Ada"}.
12053 Note that the package corresponding to
12054 @command{gnatmake} is named @code{Builder}. The @code{Release} project is
12055 similar, but only includes the @code{Compiler} package.
12057 In project @code{Debug} above, the ^switches^switches^ starting with
12058 @option{-gnat} that are specified in package @code{Compiler}
12059 could have been placed in package @code{Builder}, since @command{gnatmake}
12060 transmits all such ^switches^switches^ to the compiler.
12062 @node Main Subprograms
12063 @unnumberedsubsubsec Main Subprograms
12066 One of the specifiable properties of a project is a list of files that contain
12067 main subprograms. This property is captured in the @code{Main} attribute,
12068 whose value is a list of strings. If a project defines the @code{Main}
12069 attribute, it is not necessary to identify the main subprogram(s) when
12070 invoking @command{gnatmake} (@pxref{gnatmake and Project Files}).
12072 @node Executable File Names
12073 @unnumberedsubsubsec Executable File Names
12076 By default, the executable file name corresponding to a main source is
12077 deduced from the main source file name. Through the attributes
12078 @code{Executable} and @code{Executable_Suffix} of package @code{Builder},
12079 it is possible to change this default.
12080 In project @code{Debug} above, the executable file name
12081 for main source @file{^proc.adb^PROC.ADB^} is
12082 @file{^proc1^PROC1.EXE^}.
12083 Attribute @code{Executable_Suffix}, when specified, may change the suffix
12084 of the executable files, when no attribute @code{Executable} applies:
12085 its value replace the platform-specific executable suffix.
12086 Attributes @code{Executable} and @code{Executable_Suffix} are the only ways to
12087 specify a non-default executable file name when several mains are built at once
12088 in a single @command{gnatmake} command.
12090 @node Source File Naming Conventions
12091 @unnumberedsubsubsec Source File Naming Conventions
12094 Since the project files above do not specify any source file naming
12095 conventions, the GNAT defaults are used. The mechanism for defining source
12096 file naming conventions -- a package named @code{Naming} --
12097 is described below (@pxref{Naming Schemes}).
12099 @node Source Language(s)
12100 @unnumberedsubsubsec Source Language(s)
12103 Since the project files do not specify a @code{Languages} attribute, by
12104 default the GNAT tools assume that the language of the project file is Ada.
12105 More generally, a project can comprise source files
12106 in Ada, C, and/or other languages.
12108 @node Using External Variables
12109 @subsection Using External Variables
12112 Instead of supplying different project files for debug and release, we can
12113 define a single project file that queries an external variable (set either
12114 on the command line or via an ^environment variable^logical name^) in order to
12115 conditionally define the appropriate settings. Again, assume that the
12116 source files @file{pack.ads}, @file{pack.adb}, and @file{proc.adb} are
12117 located in directory @file{^/common^[COMMON]^}. The following project file,
12118 @file{build.gpr}, queries the external variable named @code{STYLE} and
12119 defines an object directory and ^switch^switch^ settings based on whether
12120 the value is @code{"deb"} (debug) or @code{"rel"} (release), and where
12121 the default is @code{"deb"}.
12123 @smallexample @c projectfile
12126 for Main use ("proc");
12128 type Style_Type is ("deb", "rel");
12129 Style : Style_Type := external ("STYLE", "deb");
12133 for Object_Dir use "debug";
12136 for Object_Dir use "release";
12137 for Exec_Dir use ".";
12146 for ^Default_Switches^Default_Switches^ ("Ada")
12148 for Executable ("proc") use "proc1";
12157 package Compiler is
12161 for ^Default_Switches^Default_Switches^ ("Ada")
12162 use ("^-gnata^-gnata^",
12164 "^-gnatE^-gnatE^");
12167 for ^Default_Switches^Default_Switches^ ("Ada")
12178 @code{Style_Type} is an example of a @emph{string type}, which is the project
12179 file analog of an Ada enumeration type but whose components are string literals
12180 rather than identifiers. @code{Style} is declared as a variable of this type.
12182 The form @code{external("STYLE", "deb")} is known as an
12183 @emph{external reference}; its first argument is the name of an
12184 @emph{external variable}, and the second argument is a default value to be
12185 used if the external variable doesn't exist. You can define an external
12186 variable on the command line via the @option{^-X^/EXTERNAL_REFERENCE^} switch,
12187 or you can use ^an environment variable^a logical name^
12188 as an external variable.
12190 Each @code{case} construct is expanded by the Project Manager based on the
12191 value of @code{Style}. Thus the command
12194 gnatmake -P/common/build.gpr -XSTYLE=deb
12200 gnatmake /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=deb
12205 is equivalent to the @command{gnatmake} invocation using the project file
12206 @file{debug.gpr} in the earlier example. So is the command
12208 gnatmake ^-P/common/build.gpr^/PROJECT_FILE=[COMMON]BUILD.GPR^
12212 since @code{"deb"} is the default for @code{STYLE}.
12218 gnatmake -P/common/build.gpr -XSTYLE=rel
12224 GNAT MAKE /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=rel
12229 is equivalent to the @command{gnatmake} invocation using the project file
12230 @file{release.gpr} in the earlier example.
12232 @node Importing Other Projects
12233 @subsection Importing Other Projects
12234 @cindex @code{ADA_PROJECT_PATH}
12237 A compilation unit in a source file in one project may depend on compilation
12238 units in source files in other projects. To compile this unit under
12239 control of a project file, the
12240 dependent project must @emph{import} the projects containing the needed source
12242 This effect is obtained using syntax similar to an Ada @code{with} clause,
12243 but where @code{with}ed entities are strings that denote project files.
12245 As an example, suppose that the two projects @code{GUI_Proj} and
12246 @code{Comm_Proj} are defined in the project files @file{gui_proj.gpr} and
12247 @file{comm_proj.gpr} in directories @file{^/gui^[GUI]^}
12248 and @file{^/comm^[COMM]^}, respectively.
12249 Suppose that the source files for @code{GUI_Proj} are
12250 @file{gui.ads} and @file{gui.adb}, and that the source files for
12251 @code{Comm_Proj} are @file{comm.ads} and @file{comm.adb}, where each set of
12252 files is located in its respective project file directory. Schematically:
12271 We want to develop an application in directory @file{^/app^[APP]^} that
12272 @code{with} the packages @code{GUI} and @code{Comm}, using the properties of
12273 the corresponding project files (e.g.@: the ^switch^switch^ settings
12274 and object directory).
12275 Skeletal code for a main procedure might be something like the following:
12277 @smallexample @c ada
12280 procedure App_Main is
12289 Here is a project file, @file{app_proj.gpr}, that achieves the desired
12292 @smallexample @c projectfile
12294 with "/gui/gui_proj", "/comm/comm_proj";
12295 project App_Proj is
12296 for Main use ("app_main");
12302 Building an executable is achieved through the command:
12304 gnatmake ^-P/app/app_proj^/PROJECT_FILE=[APP]APP_PROJ^
12307 which will generate the @code{^app_main^APP_MAIN.EXE^} executable
12308 in the directory where @file{app_proj.gpr} resides.
12310 If an imported project file uses the standard extension (@code{^gpr^GPR^}) then
12311 (as illustrated above) the @code{with} clause can omit the extension.
12313 Our example specified an absolute path for each imported project file.
12314 Alternatively, the directory name of an imported object can be omitted
12318 The imported project file is in the same directory as the importing project
12321 You have defined ^an environment variable^a logical name^
12322 that includes the directory containing
12323 the needed project file. The syntax of @code{ADA_PROJECT_PATH} is the same as
12324 the syntax of @code{ADA_INCLUDE_PATH} and @code{ADA_OBJECTS_PATH}: a list of
12325 directory names separated by colons (semicolons on Windows).
12329 Thus, if we define @code{ADA_PROJECT_PATH} to include @file{^/gui^[GUI]^} and
12330 @file{^/comm^[COMM]^}, then our project file @file{app_proj.gpr} can be written
12333 @smallexample @c projectfile
12335 with "gui_proj", "comm_proj";
12336 project App_Proj is
12337 for Main use ("app_main");
12343 Importing other projects can create ambiguities.
12344 For example, the same unit might be present in different imported projects, or
12345 it might be present in both the importing project and in an imported project.
12346 Both of these conditions are errors. Note that in the current version of
12347 the Project Manager, it is illegal to have an ambiguous unit even if the
12348 unit is never referenced by the importing project. This restriction may be
12349 relaxed in a future release.
12351 @node Extending a Project
12352 @subsection Extending a Project
12355 In large software systems it is common to have multiple
12356 implementations of a common interface; in Ada terms, multiple versions of a
12357 package body for the same spec. For example, one implementation
12358 might be safe for use in tasking programs, while another might only be used
12359 in sequential applications. This can be modeled in GNAT using the concept
12360 of @emph{project extension}. If one project (the ``child'') @emph{extends}
12361 another project (the ``parent'') then by default all source files of the
12362 parent project are inherited by the child, but the child project can
12363 override any of the parent's source files with new versions, and can also
12364 add new files. This facility is the project analog of a type extension in
12365 Object-Oriented Programming. Project hierarchies are permitted (a child
12366 project may be the parent of yet another project), and a project that
12367 inherits one project can also import other projects.
12369 As an example, suppose that directory @file{^/seq^[SEQ]^} contains the project
12370 file @file{seq_proj.gpr} as well as the source files @file{pack.ads},
12371 @file{pack.adb}, and @file{proc.adb}:
12384 Note that the project file can simply be empty (that is, no attribute or
12385 package is defined):
12387 @smallexample @c projectfile
12389 project Seq_Proj is
12395 implying that its source files are all the Ada source files in the project
12398 Suppose we want to supply an alternate version of @file{pack.adb}, in
12399 directory @file{^/tasking^[TASKING]^}, but use the existing versions of
12400 @file{pack.ads} and @file{proc.adb}. We can define a project
12401 @code{Tasking_Proj} that inherits @code{Seq_Proj}:
12405 ^/tasking^[TASKING]^
12411 project Tasking_Proj extends "/seq/seq_proj" is
12417 The version of @file{pack.adb} used in a build depends on which project file
12420 Note that we could have obtained the desired behavior using project import
12421 rather than project inheritance; a @code{base} project would contain the
12422 sources for @file{pack.ads} and @file{proc.adb}, a sequential project would
12423 import @code{base} and add @file{pack.adb}, and likewise a tasking project
12424 would import @code{base} and add a different version of @file{pack.adb}. The
12425 choice depends on whether other sources in the original project need to be
12426 overridden. If they do, then project extension is necessary, otherwise,
12427 importing is sufficient.
12430 In a project file that extends another project file, it is possible to
12431 indicate that an inherited source is not part of the sources of the extending
12432 project. This is necessary sometimes when a package spec has been overloaded
12433 and no longer requires a body: in this case, it is necessary to indicate that
12434 the inherited body is not part of the sources of the project, otherwise there
12435 will be a compilation error when compiling the spec.
12437 For that purpose, the attribute @code{Excluded_Source_Files} is used.
12438 Its value is a string list: a list of file names. It is also possible to use
12439 attribute @code{Excluded_Source_List_File}. Its value is a single string:
12440 the file name of a text file containing a list of file names, one per line.
12442 @smallexample @c @projectfile
12443 project B extends "a" is
12444 for Source_Files use ("pkg.ads");
12445 -- New spec of Pkg does not need a completion
12446 for Excluded_Source_Files use ("pkg.adb");
12450 Attribute @code{Excluded_Source_Files} may also be used to check if a source
12451 is still needed: if it is possible to build using @command{gnatmake} when such
12452 a source is put in attribute @code{Excluded_Source_Files} of a project P, then
12453 it is possible to remove the source completely from a system that includes
12456 @c ***********************
12457 @c * Project File Syntax *
12458 @c ***********************
12460 @node Project File Syntax
12461 @section Project File Syntax
12465 * Qualified Projects::
12471 * Associative Array Attributes::
12472 * case Constructions::
12476 This section describes the structure of project files.
12478 A project may be an @emph{independent project}, entirely defined by a single
12479 project file. Any Ada source file in an independent project depends only
12480 on the predefined library and other Ada source files in the same project.
12483 A project may also @dfn{depend on} other projects, in either or both of
12484 the following ways:
12486 @item It may import any number of projects
12487 @item It may extend at most one other project
12491 The dependence relation is a directed acyclic graph (the subgraph reflecting
12492 the ``extends'' relation is a tree).
12494 A project's @dfn{immediate sources} are the source files directly defined by
12495 that project, either implicitly by residing in the project file's directory,
12496 or explicitly through any of the source-related attributes described below.
12497 More generally, a project @var{proj}'s @dfn{sources} are the immediate sources
12498 of @var{proj} together with the immediate sources (unless overridden) of any
12499 project on which @var{proj} depends (either directly or indirectly).
12502 @subsection Basic Syntax
12505 As seen in the earlier examples, project files have an Ada-like syntax.
12506 The minimal project file is:
12507 @smallexample @c projectfile
12516 The identifier @code{Empty} is the name of the project.
12517 This project name must be present after the reserved
12518 word @code{end} at the end of the project file, followed by a semi-colon.
12520 Any name in a project file, such as the project name or a variable name,
12521 has the same syntax as an Ada identifier.
12523 The reserved words of project files are the Ada 95 reserved words plus
12524 @code{extends}, @code{external}, and @code{project}. Note that the only Ada
12525 reserved words currently used in project file syntax are:
12561 Comments in project files have the same syntax as in Ada, two consecutive
12562 hyphens through the end of the line.
12564 @node Qualified Projects
12565 @subsection Qualified Projects
12568 Before the reserved @code{project}, there may be one or two "qualifiers", that
12569 is identifiers or other reserved words, to qualify the project.
12571 The current list of qualifiers is:
12575 @code{abstract}: qualify a project with no sources. An abstract project must
12576 have a declaration specifying that there are no sources in the project, and,
12577 if it extends another project, the project it extends must also be a qualified
12581 @code{standard}: a standard project is a non library project with sources.
12584 @code{aggregate}: for future extension
12587 @code{aggregate library}: for future extension
12590 @code{library}: a library project must declare both attributes
12591 @code{Library_Name} and @code{Library_Dir}.
12594 @code{configuration}: a configuration project cannot be in a project tree.
12598 @subsection Packages
12601 A project file may contain @emph{packages}. The name of a package must be one
12602 of the identifiers from the following list. A package
12603 with a given name may only appear once in a project file. Package names are
12604 case insensitive. The following package names are legal:
12620 @code{Cross_Reference}
12624 @code{Pretty_Printer}
12634 @code{Language_Processing}
12638 In its simplest form, a package may be empty:
12640 @smallexample @c projectfile
12650 A package may contain @emph{attribute declarations},
12651 @emph{variable declarations} and @emph{case constructions}, as will be
12654 When there is ambiguity between a project name and a package name,
12655 the name always designates the project. To avoid possible confusion, it is
12656 always a good idea to avoid naming a project with one of the
12657 names allowed for packages or any name that starts with @code{gnat}.
12660 @subsection Expressions
12663 An @emph{expression} is either a @emph{string expression} or a
12664 @emph{string list expression}.
12666 A @emph{string expression} is either a @emph{simple string expression} or a
12667 @emph{compound string expression}.
12669 A @emph{simple string expression} is one of the following:
12671 @item A literal string; e.g.@: @code{"comm/my_proj.gpr"}
12672 @item A string-valued variable reference (@pxref{Variables})
12673 @item A string-valued attribute reference (@pxref{Attributes})
12674 @item An external reference (@pxref{External References in Project Files})
12678 A @emph{compound string expression} is a concatenation of string expressions,
12679 using the operator @code{"&"}
12681 Path & "/" & File_Name & ".ads"
12685 A @emph{string list expression} is either a
12686 @emph{simple string list expression} or a
12687 @emph{compound string list expression}.
12689 A @emph{simple string list expression} is one of the following:
12691 @item A parenthesized list of zero or more string expressions,
12692 separated by commas
12694 File_Names := (File_Name, "gnat.adc", File_Name & ".orig");
12697 @item A string list-valued variable reference
12698 @item A string list-valued attribute reference
12702 A @emph{compound string list expression} is the concatenation (using
12703 @code{"&"}) of a simple string list expression and an expression. Note that
12704 each term in a compound string list expression, except the first, may be
12705 either a string expression or a string list expression.
12707 @smallexample @c projectfile
12709 File_Name_List := () & File_Name; -- One string in this list
12710 Extended_File_Name_List := File_Name_List & (File_Name & ".orig");
12712 Big_List := File_Name_List & Extended_File_Name_List;
12713 -- Concatenation of two string lists: three strings
12714 Illegal_List := "gnat.adc" & Extended_File_Name_List;
12715 -- Illegal: must start with a string list
12720 @subsection String Types
12723 A @emph{string type declaration} introduces a discrete set of string literals.
12724 If a string variable is declared to have this type, its value
12725 is restricted to the given set of literals.
12727 Here is an example of a string type declaration:
12729 @smallexample @c projectfile
12730 type OS is ("NT", "nt", "Unix", "GNU/Linux", "other OS");
12734 Variables of a string type are called @emph{typed variables}; all other
12735 variables are called @emph{untyped variables}. Typed variables are
12736 particularly useful in @code{case} constructions, to support conditional
12737 attribute declarations.
12738 (@pxref{case Constructions}).
12740 The string literals in the list are case sensitive and must all be different.
12741 They may include any graphic characters allowed in Ada, including spaces.
12743 A string type may only be declared at the project level, not inside a package.
12745 A string type may be referenced by its name if it has been declared in the same
12746 project file, or by an expanded name whose prefix is the name of the project
12747 in which it is declared.
12750 @subsection Variables
12753 A variable may be declared at the project file level, or within a package.
12754 Here are some examples of variable declarations:
12756 @smallexample @c projectfile
12758 This_OS : OS := external ("OS"); -- a typed variable declaration
12759 That_OS := "GNU/Linux"; -- an untyped variable declaration
12764 The syntax of a @emph{typed variable declaration} is identical to the Ada
12765 syntax for an object declaration. By contrast, the syntax of an untyped
12766 variable declaration is identical to an Ada assignment statement. In fact,
12767 variable declarations in project files have some of the characteristics of
12768 an assignment, in that successive declarations for the same variable are
12769 allowed. Untyped variable declarations do establish the expected kind of the
12770 variable (string or string list), and successive declarations for it must
12771 respect the initial kind.
12774 A string variable declaration (typed or untyped) declares a variable
12775 whose value is a string. This variable may be used as a string expression.
12776 @smallexample @c projectfile
12777 File_Name := "readme.txt";
12778 Saved_File_Name := File_Name & ".saved";
12782 A string list variable declaration declares a variable whose value is a list
12783 of strings. The list may contain any number (zero or more) of strings.
12785 @smallexample @c projectfile
12787 List_With_One_Element := ("^-gnaty^-gnaty^");
12788 List_With_Two_Elements := List_With_One_Element & "^-gnatg^-gnatg^";
12789 Long_List := ("main.ada", "pack1_.ada", "pack1.ada", "pack2_.ada"
12790 "pack2.ada", "util_.ada", "util.ada");
12794 The same typed variable may not be declared more than once at project level,
12795 and it may not be declared more than once in any package; it is in effect
12798 The same untyped variable may be declared several times. Declarations are
12799 elaborated in the order in which they appear, so the new value replaces
12800 the old one, and any subsequent reference to the variable uses the new value.
12801 However, as noted above, if a variable has been declared as a string, all
12803 declarations must give it a string value. Similarly, if a variable has
12804 been declared as a string list, all subsequent declarations
12805 must give it a string list value.
12807 A @emph{variable reference} may take several forms:
12810 @item The simple variable name, for a variable in the current package (if any)
12811 or in the current project
12812 @item An expanded name, whose prefix is a context name.
12816 A @emph{context} may be one of the following:
12819 @item The name of an existing package in the current project
12820 @item The name of an imported project of the current project
12821 @item The name of an ancestor project (i.e., a project extended by the current
12822 project, either directly or indirectly)
12823 @item An expanded name whose prefix is an imported/parent project name, and
12824 whose selector is a package name in that project.
12828 A variable reference may be used in an expression.
12831 @subsection Attributes
12834 A project (and its packages) may have @emph{attributes} that define
12835 the project's properties. Some attributes have values that are strings;
12836 others have values that are string lists.
12838 There are two categories of attributes: @emph{simple attributes}
12839 and @emph{associative arrays} (@pxref{Associative Array Attributes}).
12841 Legal project attribute names, and attribute names for each legal package are
12842 listed below. Attributes names are case-insensitive.
12844 The following attributes are defined on projects (all are simple attributes):
12846 @multitable @columnfractions .4 .3
12847 @item @emph{Attribute Name}
12849 @item @code{Source_Files}
12851 @item @code{Source_Dirs}
12853 @item @code{Source_List_File}
12855 @item @code{Object_Dir}
12857 @item @code{Exec_Dir}
12859 @item @code{Excluded_Source_Dirs}
12861 @item @code{Excluded_Source_Files}
12863 @item @code{Excluded_Source_List_File}
12865 @item @code{Languages}
12869 @item @code{Library_Dir}
12871 @item @code{Library_Name}
12873 @item @code{Library_Kind}
12875 @item @code{Library_Version}
12877 @item @code{Library_Interface}
12879 @item @code{Library_Auto_Init}
12881 @item @code{Library_Options}
12883 @item @code{Library_Src_Dir}
12885 @item @code{Library_ALI_Dir}
12887 @item @code{Library_GCC}
12889 @item @code{Library_Symbol_File}
12891 @item @code{Library_Symbol_Policy}
12893 @item @code{Library_Reference_Symbol_File}
12895 @item @code{Externally_Built}
12900 The following attributes are defined for package @code{Naming}
12901 (@pxref{Naming Schemes}):
12903 @multitable @columnfractions .4 .2 .2 .2
12904 @item Attribute Name @tab Category @tab Index @tab Value
12905 @item @code{Spec_Suffix}
12906 @tab associative array
12909 @item @code{Body_Suffix}
12910 @tab associative array
12913 @item @code{Separate_Suffix}
12914 @tab simple attribute
12917 @item @code{Casing}
12918 @tab simple attribute
12921 @item @code{Dot_Replacement}
12922 @tab simple attribute
12926 @tab associative array
12930 @tab associative array
12933 @item @code{Specification_Exceptions}
12934 @tab associative array
12937 @item @code{Implementation_Exceptions}
12938 @tab associative array
12944 The following attributes are defined for packages @code{Builder},
12945 @code{Compiler}, @code{Binder},
12946 @code{Linker}, @code{Cross_Reference}, and @code{Finder}
12947 (@pxref{^Switches^Switches^ and Project Files}).
12949 @multitable @columnfractions .4 .2 .2 .2
12950 @item Attribute Name @tab Category @tab Index @tab Value
12951 @item @code{^Default_Switches^Default_Switches^}
12952 @tab associative array
12955 @item @code{^Switches^Switches^}
12956 @tab associative array
12962 In addition, package @code{Compiler} has a single string attribute
12963 @code{Local_Configuration_Pragmas} and package @code{Builder} has a single
12964 string attribute @code{Global_Configuration_Pragmas}.
12967 Each simple attribute has a default value: the empty string (for string-valued
12968 attributes) and the empty list (for string list-valued attributes).
12970 An attribute declaration defines a new value for an attribute.
12972 Examples of simple attribute declarations:
12974 @smallexample @c projectfile
12975 for Object_Dir use "objects";
12976 for Source_Dirs use ("units", "test/drivers");
12980 The syntax of a @dfn{simple attribute declaration} is similar to that of an
12981 attribute definition clause in Ada.
12983 Attributes references may be appear in expressions.
12984 The general form for such a reference is @code{<entity>'<attribute>}:
12985 Associative array attributes are functions. Associative
12986 array attribute references must have an argument that is a string literal.
12990 @smallexample @c projectfile
12992 Naming'Dot_Replacement
12993 Imported_Project'Source_Dirs
12994 Imported_Project.Naming'Casing
12995 Builder'^Default_Switches^Default_Switches^("Ada")
12999 The prefix of an attribute may be:
13001 @item @code{project} for an attribute of the current project
13002 @item The name of an existing package of the current project
13003 @item The name of an imported project
13004 @item The name of a parent project that is extended by the current project
13005 @item An expanded name whose prefix is imported/parent project name,
13006 and whose selector is a package name
13011 @smallexample @c projectfile
13014 for Source_Dirs use project'Source_Dirs & "units";
13015 for Source_Dirs use project'Source_Dirs & "test/drivers"
13021 In the first attribute declaration, initially the attribute @code{Source_Dirs}
13022 has the default value: an empty string list. After this declaration,
13023 @code{Source_Dirs} is a string list of one element: @code{"units"}.
13024 After the second attribute declaration @code{Source_Dirs} is a string list of
13025 two elements: @code{"units"} and @code{"test/drivers"}.
13027 Note: this example is for illustration only. In practice,
13028 the project file would contain only one attribute declaration:
13030 @smallexample @c projectfile
13031 for Source_Dirs use ("units", "test/drivers");
13034 @node Associative Array Attributes
13035 @subsection Associative Array Attributes
13038 Some attributes are defined as @emph{associative arrays}. An associative
13039 array may be regarded as a function that takes a string as a parameter
13040 and delivers a string or string list value as its result.
13042 Here are some examples of single associative array attribute associations:
13044 @smallexample @c projectfile
13045 for Body ("main") use "Main.ada";
13046 for ^Switches^Switches^ ("main.ada")
13048 "^-gnatv^-gnatv^");
13049 for ^Switches^Switches^ ("main.ada")
13050 use Builder'^Switches^Switches^ ("main.ada")
13055 Like untyped variables and simple attributes, associative array attributes
13056 may be declared several times. Each declaration supplies a new value for the
13057 attribute, and replaces the previous setting.
13060 An associative array attribute may be declared as a full associative array
13061 declaration, with the value of the same attribute in an imported or extended
13064 @smallexample @c projectfile
13066 for Default_Switches use Default.Builder'Default_Switches;
13071 In this example, @code{Default} must be either a project imported by the
13072 current project, or the project that the current project extends. If the
13073 attribute is in a package (in this case, in package @code{Builder}), the same
13074 package needs to be specified.
13077 A full associative array declaration replaces any other declaration for the
13078 attribute, including other full associative array declaration. Single
13079 associative array associations may be declare after a full associative
13080 declaration, modifying the value for a single association of the attribute.
13082 @node case Constructions
13083 @subsection @code{case} Constructions
13086 A @code{case} construction is used in a project file to effect conditional
13088 Here is a typical example:
13090 @smallexample @c projectfile
13093 type OS_Type is ("GNU/Linux", "Unix", "NT", "VMS");
13095 OS : OS_Type := external ("OS", "GNU/Linux");
13099 package Compiler is
13101 when "GNU/Linux" | "Unix" =>
13102 for ^Default_Switches^Default_Switches^ ("Ada")
13103 use ("^-gnath^-gnath^");
13105 for ^Default_Switches^Default_Switches^ ("Ada")
13106 use ("^-gnatP^-gnatP^");
13115 The syntax of a @code{case} construction is based on the Ada case statement
13116 (although there is no @code{null} construction for empty alternatives).
13118 The case expression must be a typed string variable.
13119 Each alternative comprises the reserved word @code{when}, either a list of
13120 literal strings separated by the @code{"|"} character or the reserved word
13121 @code{others}, and the @code{"=>"} token.
13122 Each literal string must belong to the string type that is the type of the
13124 An @code{others} alternative, if present, must occur last.
13126 After each @code{=>}, there are zero or more constructions. The only
13127 constructions allowed in a case construction are other case constructions,
13128 attribute declarations and variable declarations. String type declarations and
13129 package declarations are not allowed. Variable declarations are restricted to
13130 variables that have already been declared before the case construction.
13132 The value of the case variable is often given by an external reference
13133 (@pxref{External References in Project Files}).
13135 @c ****************************************
13136 @c * Objects and Sources in Project Files *
13137 @c ****************************************
13139 @node Objects and Sources in Project Files
13140 @section Objects and Sources in Project Files
13143 * Object Directory::
13145 * Source Directories::
13146 * Source File Names::
13150 Each project has exactly one object directory and one or more source
13151 directories. The source directories must contain at least one source file,
13152 unless the project file explicitly specifies that no source files are present
13153 (@pxref{Source File Names}).
13155 @node Object Directory
13156 @subsection Object Directory
13159 The object directory for a project is the directory containing the compiler's
13160 output (such as @file{ALI} files and object files) for the project's immediate
13163 The object directory is given by the value of the attribute @code{Object_Dir}
13164 in the project file.
13166 @smallexample @c projectfile
13167 for Object_Dir use "objects";
13171 The attribute @code{Object_Dir} has a string value, the path name of the object
13172 directory. The path name may be absolute or relative to the directory of the
13173 project file. This directory must already exist, and be readable and writable.
13175 By default, when the attribute @code{Object_Dir} is not given an explicit value
13176 or when its value is the empty string, the object directory is the same as the
13177 directory containing the project file.
13179 @node Exec Directory
13180 @subsection Exec Directory
13183 The exec directory for a project is the directory containing the executables
13184 for the project's main subprograms.
13186 The exec directory is given by the value of the attribute @code{Exec_Dir}
13187 in the project file.
13189 @smallexample @c projectfile
13190 for Exec_Dir use "executables";
13194 The attribute @code{Exec_Dir} has a string value, the path name of the exec
13195 directory. The path name may be absolute or relative to the directory of the
13196 project file. This directory must already exist, and be writable.
13198 By default, when the attribute @code{Exec_Dir} is not given an explicit value
13199 or when its value is the empty string, the exec directory is the same as the
13200 object directory of the project file.
13202 @node Source Directories
13203 @subsection Source Directories
13206 The source directories of a project are specified by the project file
13207 attribute @code{Source_Dirs}.
13209 This attribute's value is a string list. If the attribute is not given an
13210 explicit value, then there is only one source directory, the one where the
13211 project file resides.
13213 A @code{Source_Dirs} attribute that is explicitly defined to be the empty list,
13216 @smallexample @c projectfile
13217 for Source_Dirs use ();
13221 indicates that the project contains no source files.
13223 Otherwise, each string in the string list designates one or more
13224 source directories.
13226 @smallexample @c projectfile
13227 for Source_Dirs use ("sources", "test/drivers");
13231 If a string in the list ends with @code{"/**"}, then the directory whose path
13232 name precedes the two asterisks, as well as all its subdirectories
13233 (recursively), are source directories.
13235 @smallexample @c projectfile
13236 for Source_Dirs use ("/system/sources/**");
13240 Here the directory @code{/system/sources} and all of its subdirectories
13241 (recursively) are source directories.
13243 To specify that the source directories are the directory of the project file
13244 and all of its subdirectories, you can declare @code{Source_Dirs} as follows:
13245 @smallexample @c projectfile
13246 for Source_Dirs use ("./**");
13250 Each of the source directories must exist and be readable.
13252 @node Source File Names
13253 @subsection Source File Names
13256 In a project that contains source files, their names may be specified by the
13257 attributes @code{Source_Files} (a string list) or @code{Source_List_File}
13258 (a string). Source file names never include any directory information.
13260 If the attribute @code{Source_Files} is given an explicit value, then each
13261 element of the list is a source file name.
13263 @smallexample @c projectfile
13264 for Source_Files use ("main.adb");
13265 for Source_Files use ("main.adb", "pack1.ads", "pack2.adb");
13269 If the attribute @code{Source_Files} is not given an explicit value,
13270 but the attribute @code{Source_List_File} is given a string value,
13271 then the source file names are contained in the text file whose path name
13272 (absolute or relative to the directory of the project file) is the
13273 value of the attribute @code{Source_List_File}.
13275 Each line in the file that is not empty or is not a comment
13276 contains a source file name.
13278 @smallexample @c projectfile
13279 for Source_List_File use "source_list.txt";
13283 By default, if neither the attribute @code{Source_Files} nor the attribute
13284 @code{Source_List_File} is given an explicit value, then each file in the
13285 source directories that conforms to the project's naming scheme
13286 (@pxref{Naming Schemes}) is an immediate source of the project.
13288 A warning is issued if both attributes @code{Source_Files} and
13289 @code{Source_List_File} are given explicit values. In this case, the attribute
13290 @code{Source_Files} prevails.
13292 Each source file name must be the name of one existing source file
13293 in one of the source directories.
13295 A @code{Source_Files} attribute whose value is an empty list
13296 indicates that there are no source files in the project.
13298 If the order of the source directories is known statically, that is if
13299 @code{"/**"} is not used in the string list @code{Source_Dirs}, then there may
13300 be several files with the same source file name. In this case, only the file
13301 in the first directory is considered as an immediate source of the project
13302 file. If the order of the source directories is not known statically, it is
13303 an error to have several files with the same source file name.
13305 Projects can be specified to have no Ada source
13306 files: the value of (@code{Source_Dirs} or @code{Source_Files} may be an empty
13307 list, or the @code{"Ada"} may be absent from @code{Languages}:
13309 @smallexample @c projectfile
13310 for Source_Dirs use ();
13311 for Source_Files use ();
13312 for Languages use ("C", "C++");
13316 Otherwise, a project must contain at least one immediate source.
13318 Projects with no source files are useful as template packages
13319 (@pxref{Packages in Project Files}) for other projects; in particular to
13320 define a package @code{Naming} (@pxref{Naming Schemes}).
13322 @c ****************************
13323 @c * Importing Projects *
13324 @c ****************************
13326 @node Importing Projects
13327 @section Importing Projects
13328 @cindex @code{ADA_PROJECT_PATH}
13331 An immediate source of a project P may depend on source files that
13332 are neither immediate sources of P nor in the predefined library.
13333 To get this effect, P must @emph{import} the projects that contain the needed
13336 @smallexample @c projectfile
13338 with "project1", "utilities.gpr";
13339 with "/namings/apex.gpr";
13346 As can be seen in this example, the syntax for importing projects is similar
13347 to the syntax for importing compilation units in Ada. However, project files
13348 use literal strings instead of names, and the @code{with} clause identifies
13349 project files rather than packages.
13351 Each literal string is the file name or path name (absolute or relative) of a
13352 project file. If a string corresponds to a file name, with no path or a
13353 relative path, then its location is determined by the @emph{project path}. The
13354 latter can be queried using @code{gnatls -v}. It contains:
13358 In first position, the directory containing the current project file.
13360 In last position, the default project directory. This default project directory
13361 is part of the GNAT installation and is the standard place to install project
13362 files giving access to standard support libraries.
13364 @ref{Installing a library}
13368 In between, all the directories referenced in the
13369 ^environment variable^logical name^ @env{ADA_PROJECT_PATH} if it exists.
13373 If a relative pathname is used, as in
13375 @smallexample @c projectfile
13380 then the full path for the project is constructed by concatenating this
13381 relative path to those in the project path, in order, until a matching file is
13382 found. Any symbolic link will be fully resolved in the directory of the
13383 importing project file before the imported project file is examined.
13385 If the @code{with}'ed project file name does not have an extension,
13386 the default is @file{^.gpr^.GPR^}. If a file with this extension is not found,
13387 then the file name as specified in the @code{with} clause (no extension) will
13388 be used. In the above example, if a file @code{project1.gpr} is found, then it
13389 will be used; otherwise, if a file @code{^project1^PROJECT1^} exists
13390 then it will be used; if neither file exists, this is an error.
13392 A warning is issued if the name of the project file does not match the
13393 name of the project; this check is case insensitive.
13395 Any source file that is an immediate source of the imported project can be
13396 used by the immediate sources of the importing project, transitively. Thus
13397 if @code{A} imports @code{B}, and @code{B} imports @code{C}, the immediate
13398 sources of @code{A} may depend on the immediate sources of @code{C}, even if
13399 @code{A} does not import @code{C} explicitly. However, this is not recommended,
13400 because if and when @code{B} ceases to import @code{C}, some sources in
13401 @code{A} will no longer compile.
13403 A side effect of this capability is that normally cyclic dependencies are not
13404 permitted: if @code{A} imports @code{B} (directly or indirectly) then @code{B}
13405 is not allowed to import @code{A}. However, there are cases when cyclic
13406 dependencies would be beneficial. For these cases, another form of import
13407 between projects exists, the @code{limited with}: a project @code{A} that
13408 imports a project @code{B} with a straight @code{with} may also be imported,
13409 directly or indirectly, by @code{B} on the condition that imports from @code{B}
13410 to @code{A} include at least one @code{limited with}.
13412 @smallexample @c 0projectfile
13418 limited with "../a/a.gpr";
13426 limited with "../a/a.gpr";
13432 In the above legal example, there are two project cycles:
13435 @item A -> C -> D -> A
13439 In each of these cycle there is one @code{limited with}: import of @code{A}
13440 from @code{B} and import of @code{A} from @code{D}.
13442 The difference between straight @code{with} and @code{limited with} is that
13443 the name of a project imported with a @code{limited with} cannot be used in the
13444 project that imports it. In particular, its packages cannot be renamed and
13445 its variables cannot be referred to.
13447 An exception to the above rules for @code{limited with} is that for the main
13448 project specified to @command{gnatmake} or to the @command{GNAT} driver a
13449 @code{limited with} is equivalent to a straight @code{with}. For example,
13450 in the example above, projects @code{B} and @code{D} could not be main
13451 projects for @command{gnatmake} or to the @command{GNAT} driver, because they
13452 each have a @code{limited with} that is the only one in a cycle of importing
13455 @c *********************
13456 @c * Project Extension *
13457 @c *********************
13459 @node Project Extension
13460 @section Project Extension
13463 During development of a large system, it is sometimes necessary to use
13464 modified versions of some of the source files, without changing the original
13465 sources. This can be achieved through the @emph{project extension} facility.
13467 @smallexample @c projectfile
13468 project Modified_Utilities extends "/baseline/utilities.gpr" is @dots{}
13472 A project extension declaration introduces an extending project
13473 (the @emph{child}) and a project being extended (the @emph{parent}).
13475 By default, a child project inherits all the sources of its parent.
13476 However, inherited sources can be overridden: a unit in a parent is hidden
13477 by a unit of the same name in the child.
13479 Inherited sources are considered to be sources (but not immediate sources)
13480 of the child project; see @ref{Project File Syntax}.
13482 An inherited source file retains any switches specified in the parent project.
13484 For example if the project @code{Utilities} contains the spec and the
13485 body of an Ada package @code{Util_IO}, then the project
13486 @code{Modified_Utilities} can contain a new body for package @code{Util_IO}.
13487 The original body of @code{Util_IO} will not be considered in program builds.
13488 However, the package spec will still be found in the project
13491 A child project can have only one parent, except when it is qualified as
13492 abstract. But it may import any number of other projects.
13494 A project is not allowed to import directly or indirectly at the same time a
13495 child project and any of its ancestors.
13497 @c *******************************
13498 @c * Project Hierarchy Extension *
13499 @c *******************************
13501 @node Project Hierarchy Extension
13502 @section Project Hierarchy Extension
13505 When extending a large system spanning multiple projects, it is often
13506 inconvenient to extend every project in the hierarchy that is impacted by a
13507 small change introduced. In such cases, it is possible to create a virtual
13508 extension of entire hierarchy using @code{extends all} relationship.
13510 When the project is extended using @code{extends all} inheritance, all projects
13511 that are imported by it, both directly and indirectly, are considered virtually
13512 extended. That is, the Project Manager creates "virtual projects"
13513 that extend every project in the hierarchy; all these virtual projects have
13514 no sources of their own and have as object directory the object directory of
13515 the root of "extending all" project.
13517 It is possible to explicitly extend one or more projects in the hierarchy
13518 in order to modify the sources. These extending projects must be imported by
13519 the "extending all" project, which will replace the corresponding virtual
13520 projects with the explicit ones.
13522 When building such a project hierarchy extension, the Project Manager will
13523 ensure that both modified sources and sources in virtual extending projects
13524 that depend on them, are recompiled.
13526 By means of example, consider the following hierarchy of projects.
13530 project A, containing package P1
13532 project B importing A and containing package P2 which depends on P1
13534 project C importing B and containing package P3 which depends on P2
13538 We want to modify packages P1 and P3.
13540 This project hierarchy will need to be extended as follows:
13544 Create project A1 that extends A, placing modified P1 there:
13546 @smallexample @c 0projectfile
13547 project A1 extends "(@dots{})/A" is
13552 Create project C1 that "extends all" C and imports A1, placing modified
13555 @smallexample @c 0projectfile
13556 with "(@dots{})/A1";
13557 project C1 extends all "(@dots{})/C" is
13562 When you build project C1, your entire modified project space will be
13563 recompiled, including the virtual project B1 that has been impacted by the
13564 "extending all" inheritance of project C.
13566 Note that if a Library Project in the hierarchy is virtually extended,
13567 the virtual project that extends the Library Project is not a Library Project.
13569 @c ****************************************
13570 @c * External References in Project Files *
13571 @c ****************************************
13573 @node External References in Project Files
13574 @section External References in Project Files
13577 A project file may contain references to external variables; such references
13578 are called @emph{external references}.
13580 An external variable is either defined as part of the environment (an
13581 environment variable in Unix, for example) or else specified on the command
13582 line via the @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
13583 If both, then the command line value is used.
13585 The value of an external reference is obtained by means of the built-in
13586 function @code{external}, which returns a string value.
13587 This function has two forms:
13589 @item @code{external (external_variable_name)}
13590 @item @code{external (external_variable_name, default_value)}
13594 Each parameter must be a string literal. For example:
13596 @smallexample @c projectfile
13598 external ("OS", "GNU/Linux")
13602 In the form with one parameter, the function returns the value of
13603 the external variable given as parameter. If this name is not present in the
13604 environment, the function returns an empty string.
13606 In the form with two string parameters, the second argument is
13607 the value returned when the variable given as the first argument is not
13608 present in the environment. In the example above, if @code{"OS"} is not
13609 the name of ^an environment variable^a logical name^ and is not passed on
13610 the command line, then the returned value is @code{"GNU/Linux"}.
13612 An external reference may be part of a string expression or of a string
13613 list expression, and can therefore appear in a variable declaration or
13614 an attribute declaration.
13616 @smallexample @c projectfile
13618 type Mode_Type is ("Debug", "Release");
13619 Mode : Mode_Type := external ("MODE");
13626 @c *****************************
13627 @c * Packages in Project Files *
13628 @c *****************************
13630 @node Packages in Project Files
13631 @section Packages in Project Files
13634 A @emph{package} defines the settings for project-aware tools within a
13636 For each such tool one can declare a package; the names for these
13637 packages are preset (@pxref{Packages}).
13638 A package may contain variable declarations, attribute declarations, and case
13641 @smallexample @c projectfile
13644 package Builder is -- used by gnatmake
13645 for ^Default_Switches^Default_Switches^ ("Ada")
13654 The syntax of package declarations mimics that of package in Ada.
13656 Most of the packages have an attribute
13657 @code{^Default_Switches^Default_Switches^}.
13658 This attribute is an associative array, and its value is a string list.
13659 The index of the associative array is the name of a programming language (case
13660 insensitive). This attribute indicates the ^switch^switch^
13661 or ^switches^switches^ to be used
13662 with the corresponding tool.
13664 Some packages also have another attribute, @code{^Switches^Switches^},
13665 an associative array whose value is a string list.
13666 The index is the name of a source file.
13667 This attribute indicates the ^switch^switch^
13668 or ^switches^switches^ to be used by the corresponding
13669 tool when dealing with this specific file.
13671 Further information on these ^switch^switch^-related attributes is found in
13672 @ref{^Switches^Switches^ and Project Files}.
13674 A package may be declared as a @emph{renaming} of another package; e.g., from
13675 the project file for an imported project.
13677 @smallexample @c projectfile
13679 with "/global/apex.gpr";
13681 package Naming renames Apex.Naming;
13688 Packages that are renamed in other project files often come from project files
13689 that have no sources: they are just used as templates. Any modification in the
13690 template will be reflected automatically in all the project files that rename
13691 a package from the template.
13693 In addition to the tool-oriented packages, you can also declare a package
13694 named @code{Naming} to establish specialized source file naming conventions
13695 (@pxref{Naming Schemes}).
13697 @c ************************************
13698 @c * Variables from Imported Projects *
13699 @c ************************************
13701 @node Variables from Imported Projects
13702 @section Variables from Imported Projects
13705 An attribute or variable defined in an imported or parent project can
13706 be used in expressions in the importing / extending project.
13707 Such an attribute or variable is denoted by an expanded name whose prefix
13708 is either the name of the project or the expanded name of a package within
13711 @smallexample @c projectfile
13714 project Main extends "base" is
13715 Var1 := Imported.Var;
13716 Var2 := Base.Var & ".new";
13721 for ^Default_Switches^Default_Switches^ ("Ada")
13722 use Imported.Builder'Ada_^Switches^Switches^ &
13723 "^-gnatg^-gnatg^" &
13729 package Compiler is
13730 for ^Default_Switches^Default_Switches^ ("Ada")
13731 use Base.Compiler'Ada_^Switches^Switches^;
13742 The value of @code{Var1} is a copy of the variable @code{Var} defined
13743 in the project file @file{"imported.gpr"}
13745 the value of @code{Var2} is a copy of the value of variable @code{Var}
13746 defined in the project file @file{base.gpr}, concatenated with @code{".new"}
13748 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13749 @code{Builder} is a string list that includes in its value a copy of the value
13750 of @code{Ada_^Switches^Switches^} defined in the @code{Builder} package
13751 in project file @file{imported.gpr} plus two new elements:
13752 @option{"^-gnatg^-gnatg^"}
13753 and @option{"^-v^-v^"};
13755 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13756 @code{Compiler} is a copy of the variable @code{Ada_^Switches^Switches^}
13757 defined in the @code{Compiler} package in project file @file{base.gpr},
13758 the project being extended.
13761 @c ******************
13762 @c * Naming Schemes *
13763 @c ******************
13765 @node Naming Schemes
13766 @section Naming Schemes
13769 Sometimes an Ada software system is ported from a foreign compilation
13770 environment to GNAT, and the file names do not use the default GNAT
13771 conventions. Instead of changing all the file names (which for a variety
13772 of reasons might not be possible), you can define the relevant file
13773 naming scheme in the @code{Naming} package in your project file.
13776 Note that the use of pragmas described in
13777 @ref{Alternative File Naming Schemes} by mean of a configuration
13778 pragmas file is not supported when using project files. You must use
13779 the features described in this paragraph. You can however use specify
13780 other configuration pragmas (@pxref{Specifying Configuration Pragmas}).
13783 For example, the following
13784 package models the Apex file naming rules:
13786 @smallexample @c projectfile
13789 for Casing use "lowercase";
13790 for Dot_Replacement use ".";
13791 for Spec_Suffix ("Ada") use ".1.ada";
13792 for Body_Suffix ("Ada") use ".2.ada";
13799 For example, the following package models the HP Ada file naming rules:
13801 @smallexample @c projectfile
13804 for Casing use "lowercase";
13805 for Dot_Replacement use "__";
13806 for Spec_Suffix ("Ada") use "_.^ada^ada^";
13807 for Body_Suffix ("Ada") use ".^ada^ada^";
13813 (Note that @code{Casing} is @code{"lowercase"} because GNAT gets the file
13814 names in lower case)
13818 You can define the following attributes in package @code{Naming}:
13822 @item @code{Casing}
13823 This must be a string with one of the three values @code{"lowercase"},
13824 @code{"uppercase"} or @code{"mixedcase"}; these strings are case insensitive.
13827 If @code{Casing} is not specified, then the default is @code{"lowercase"}.
13829 @item @code{Dot_Replacement}
13830 This must be a string whose value satisfies the following conditions:
13833 @item It must not be empty
13834 @item It cannot start or end with an alphanumeric character
13835 @item It cannot be a single underscore
13836 @item It cannot start with an underscore followed by an alphanumeric
13837 @item It cannot contain a dot @code{'.'} except if the entire string
13842 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
13844 @item @code{Spec_Suffix}
13845 This is an associative array (indexed by the programming language name, case
13846 insensitive) whose value is a string that must satisfy the following
13850 @item It must not be empty
13851 @item It must include at least one dot
13854 If @code{Spec_Suffix ("Ada")} is not specified, then the default is
13855 @code{"^.ads^.ADS^"}.
13857 @item @code{Body_Suffix}
13858 This is an associative array (indexed by the programming language name, case
13859 insensitive) whose value is a string that must satisfy the following
13863 @item It must not be empty
13864 @item It must include at least one dot
13865 @item It cannot be the same as @code{Spec_Suffix ("Ada")}
13868 If @code{Body_Suffix ("Ada")} and @code{Spec_Suffix ("Ada")} end with the
13869 same string, then a file name that ends with the longest of these two suffixes
13870 will be a body if the longest suffix is @code{Body_Suffix ("Ada")} or a spec
13871 if the longest suffix is @code{Spec_Suffix ("Ada")}.
13873 If @code{Body_Suffix ("Ada")} is not specified, then the default is
13874 @code{"^.adb^.ADB^"}.
13876 @item @code{Separate_Suffix}
13877 This must be a string whose value satisfies the same conditions as
13878 @code{Body_Suffix}. The same "longest suffix" rules apply.
13881 If @code{Separate_Suffix ("Ada")} is not specified, then it defaults to same
13882 value as @code{Body_Suffix ("Ada")}.
13886 You can use the associative array attribute @code{Spec} to define
13887 the source file name for an individual Ada compilation unit's spec. The array
13888 index must be a string literal that identifies the Ada unit (case insensitive).
13889 The value of this attribute must be a string that identifies the file that
13890 contains this unit's spec (case sensitive or insensitive depending on the
13893 @smallexample @c projectfile
13894 for Spec ("MyPack.MyChild") use "mypack.mychild.spec";
13897 When the source file contains several units, you can indicate at what
13898 position the unit occurs in the file, with the following. The first unit
13899 in the file has index 1
13901 @smallexample @c projectfile
13902 for Body ("top") use "foo.a" at 1;
13903 for Body ("foo") use "foo.a" at 2;
13908 You can use the associative array attribute @code{Body} to
13909 define the source file name for an individual Ada compilation unit's body
13910 (possibly a subunit). The array index must be a string literal that identifies
13911 the Ada unit (case insensitive). The value of this attribute must be a string
13912 that identifies the file that contains this unit's body or subunit (case
13913 sensitive or insensitive depending on the operating system).
13915 @smallexample @c projectfile
13916 for Body ("MyPack.MyChild") use "mypack.mychild.body";
13920 @c ********************
13921 @c * Library Projects *
13922 @c ********************
13924 @node Library Projects
13925 @section Library Projects
13928 @emph{Library projects} are projects whose object code is placed in a library.
13929 (Note that this facility is not yet supported on all platforms).
13931 @code{gnatmake} or @code{gprbuild} will collect all object files into a
13932 single archive, which might either be a shared or a static library. This
13933 library can later on be linked with multiple executables, potentially
13934 reducing their sizes.
13936 If your project file specifies languages other than Ada, but you are still
13937 using @code{gnatmake} to compile and link, the latter will not try to
13938 compile your sources other than Ada (you should use @code{gprbuild} if that
13939 is your intent). However, @code{gnatmake} will automatically link all object
13940 files found in the object directory, whether or not they were compiled from
13941 an Ada source file. This specific behavior only applies when multiple
13942 languages are specified.
13944 To create a library project, you need to define in its project file
13945 two project-level attributes: @code{Library_Name} and @code{Library_Dir}.
13946 Additionally, you may define other library-related attributes such as
13947 @code{Library_Kind}, @code{Library_Version}, @code{Library_Interface},
13948 @code{Library_Auto_Init}, @code{Library_Options} and @code{Library_GCC}.
13950 The @code{Library_Name} attribute has a string value. There is no restriction
13951 on the name of a library. It is the responsibility of the developer to
13952 choose a name that will be accepted by the platform. It is recommended to
13953 choose names that could be Ada identifiers; such names are almost guaranteed
13954 to be acceptable on all platforms.
13956 The @code{Library_Dir} attribute has a string value that designates the path
13957 (absolute or relative) of the directory where the library will reside.
13958 It must designate an existing directory, and this directory must be writable,
13959 different from the project's object directory and from any source directory
13960 in the project tree.
13962 If both @code{Library_Name} and @code{Library_Dir} are specified and
13963 are legal, then the project file defines a library project. The optional
13964 library-related attributes are checked only for such project files.
13966 The @code{Library_Kind} attribute has a string value that must be one of the
13967 following (case insensitive): @code{"static"}, @code{"dynamic"} or
13968 @code{"relocatable"} (which is a synonym for @code{"dynamic"}). If this
13969 attribute is not specified, the library is a static library, that is
13970 an archive of object files that can be potentially linked into a
13971 static executable. Otherwise, the library may be dynamic or
13972 relocatable, that is a library that is loaded only at the start of execution.
13974 If you need to build both a static and a dynamic library, you should use two
13975 different object directories, since in some cases some extra code needs to
13976 be generated for the latter. For such cases, it is recommended to either use
13977 two different project files, or a single one which uses external variables
13978 to indicate what kind of library should be build.
13980 The @code{Library_ALI_Dir} attribute may be specified to indicate the
13981 directory where the ALI files of the library will be copied. When it is
13982 not specified, the ALI files are copied to the directory specified in
13983 attribute @code{Library_Dir}. The directory specified by @code{Library_ALI_Dir}
13984 must be writable and different from the project's object directory and from
13985 any source directory in the project tree.
13987 The @code{Library_Version} attribute has a string value whose interpretation
13988 is platform dependent. It has no effect on VMS and Windows. On Unix, it is
13989 used only for dynamic/relocatable libraries as the internal name of the
13990 library (the @code{"soname"}). If the library file name (built from the
13991 @code{Library_Name}) is different from the @code{Library_Version}, then the
13992 library file will be a symbolic link to the actual file whose name will be
13993 @code{Library_Version}.
13997 @smallexample @c projectfile
14003 for Library_Dir use "lib_dir";
14004 for Library_Name use "dummy";
14005 for Library_Kind use "relocatable";
14006 for Library_Version use "libdummy.so." & Version;
14013 Directory @file{lib_dir} will contain the internal library file whose name
14014 will be @file{libdummy.so.1}, and @file{libdummy.so} will be a symbolic link to
14015 @file{libdummy.so.1}.
14017 When @command{gnatmake} detects that a project file
14018 is a library project file, it will check all immediate sources of the project
14019 and rebuild the library if any of the sources have been recompiled.
14021 Standard project files can import library project files. In such cases,
14022 the libraries will only be rebuilt if some of its sources are recompiled
14023 because they are in the closure of some other source in an importing project.
14024 Sources of the library project files that are not in such a closure will
14025 not be checked, unless the full library is checked, because one of its sources
14026 needs to be recompiled.
14028 For instance, assume the project file @code{A} imports the library project file
14029 @code{L}. The immediate sources of A are @file{a1.adb}, @file{a2.ads} and
14030 @file{a2.adb}. The immediate sources of L are @file{l1.ads}, @file{l1.adb},
14031 @file{l2.ads}, @file{l2.adb}.
14033 If @file{l1.adb} has been modified, then the library associated with @code{L}
14034 will be rebuilt when compiling all the immediate sources of @code{A} only
14035 if @file{a1.ads}, @file{a2.ads} or @file{a2.adb} includes a statement
14038 To be sure that all the sources in the library associated with @code{L} are
14039 up to date, and that all the sources of project @code{A} are also up to date,
14040 the following two commands needs to be used:
14047 When a library is built or rebuilt, an attempt is made first to delete all
14048 files in the library directory.
14049 All @file{ALI} files will also be copied from the object directory to the
14050 library directory. To build executables, @command{gnatmake} will use the
14051 library rather than the individual object files.
14054 It is also possible to create library project files for third-party libraries
14055 that are precompiled and cannot be compiled locally thanks to the
14056 @code{externally_built} attribute. (See @ref{Installing a library}).
14059 @c *******************************
14060 @c * Stand-alone Library Projects *
14061 @c *******************************
14063 @node Stand-alone Library Projects
14064 @section Stand-alone Library Projects
14067 A Stand-alone Library is a library that contains the necessary code to
14068 elaborate the Ada units that are included in the library. A Stand-alone
14069 Library is suitable to be used in an executable when the main is not
14070 in Ada. However, Stand-alone Libraries may also be used with an Ada main
14073 A Stand-alone Library Project is a Library Project where the library is
14074 a Stand-alone Library.
14076 To be a Stand-alone Library Project, in addition to the two attributes
14077 that make a project a Library Project (@code{Library_Name} and
14078 @code{Library_Dir}, see @ref{Library Projects}), the attribute
14079 @code{Library_Interface} must be defined.
14081 @smallexample @c projectfile
14083 for Library_Dir use "lib_dir";
14084 for Library_Name use "dummy";
14085 for Library_Interface use ("int1", "int1.child");
14089 Attribute @code{Library_Interface} has a nonempty string list value,
14090 each string in the list designating a unit contained in an immediate source
14091 of the project file.
14093 When a Stand-alone Library is built, first the binder is invoked to build
14094 a package whose name depends on the library name
14095 (^b~dummy.ads/b^B$DUMMY.ADS/B^ in the example above).
14096 This binder-generated package includes initialization and
14097 finalization procedures whose
14098 names depend on the library name (dummyinit and dummyfinal in the example
14099 above). The object corresponding to this package is included in the library.
14101 A dynamic or relocatable Stand-alone Library is automatically initialized
14102 if automatic initialization of Stand-alone Libraries is supported on the
14103 platform and if attribute @code{Library_Auto_Init} is not specified or
14104 is specified with the value "true". A static Stand-alone Library is never
14105 automatically initialized.
14107 Single string attribute @code{Library_Auto_Init} may be specified with only
14108 two possible values: "false" or "true" (case-insensitive). Specifying
14109 "false" for attribute @code{Library_Auto_Init} will prevent automatic
14110 initialization of dynamic or relocatable libraries.
14112 When a non-automatically initialized Stand-alone Library is used
14113 in an executable, its initialization procedure must be called before
14114 any service of the library is used.
14115 When the main subprogram is in Ada, it may mean that the initialization
14116 procedure has to be called during elaboration of another package.
14118 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
14119 (those that are listed in attribute @code{Library_Interface}) are copied to
14120 the Library Directory. As a consequence, only the Interface Units may be
14121 imported from Ada units outside of the library. If other units are imported,
14122 the binding phase will fail.
14124 When a Stand-Alone Library is bound, the switches that are specified in
14125 the attribute @code{Default_Switches ("Ada")} in package @code{Binder} are
14126 used in the call to @command{gnatbind}.
14128 The string list attribute @code{Library_Options} may be used to specified
14129 additional switches to the call to @command{gcc} to link the library.
14131 The attribute @code{Library_Src_Dir}, may be specified for a
14132 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
14133 single string value. Its value must be the path (absolute or relative to the
14134 project directory) of an existing directory. This directory cannot be the
14135 object directory or one of the source directories, but it can be the same as
14136 the library directory. The sources of the Interface
14137 Units of the library, necessary to an Ada client of the library, will be
14138 copied to the designated directory, called Interface Copy directory.
14139 These sources includes the specs of the Interface Units, but they may also
14140 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
14141 are used, or when there is a generic units in the spec. Before the sources
14142 are copied to the Interface Copy directory, an attempt is made to delete all
14143 files in the Interface Copy directory.
14145 @c *************************************
14146 @c * Switches Related to Project Files *
14147 @c *************************************
14148 @node Switches Related to Project Files
14149 @section Switches Related to Project Files
14152 The following switches are used by GNAT tools that support project files:
14156 @item ^-P^/PROJECT_FILE=^@var{project}
14157 @cindex @option{^-P^/PROJECT_FILE^} (any project-aware tool)
14158 Indicates the name of a project file. This project file will be parsed with
14159 the verbosity indicated by @option{^-vP^MESSAGE_PROJECT_FILES=^@emph{x}},
14160 if any, and using the external references indicated
14161 by @option{^-X^/EXTERNAL_REFERENCE^} switches, if any.
14163 There may zero, one or more spaces between @option{-P} and @var{project}.
14167 There must be only one @option{^-P^/PROJECT_FILE^} switch on the command line.
14170 Since the Project Manager parses the project file only after all the switches
14171 on the command line are checked, the order of the switches
14172 @option{^-P^/PROJECT_FILE^},
14173 @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}}
14174 or @option{^-X^/EXTERNAL_REFERENCE^} is not significant.
14176 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
14177 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (any project-aware tool)
14178 Indicates that external variable @var{name} has the value @var{value}.
14179 The Project Manager will use this value for occurrences of
14180 @code{external(name)} when parsing the project file.
14184 If @var{name} or @var{value} includes a space, then @var{name=value} should be
14185 put between quotes.
14193 Several @option{^-X^/EXTERNAL_REFERENCE^} switches can be used simultaneously.
14194 If several @option{^-X^/EXTERNAL_REFERENCE^} switches specify the same
14195 @var{name}, only the last one is used.
14198 An external variable specified with a @option{^-X^/EXTERNAL_REFERENCE^} switch
14199 takes precedence over the value of the same name in the environment.
14201 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
14202 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (any project-aware tool)
14203 Indicates the verbosity of the parsing of GNAT project files.
14206 @option{-vP0} means Default;
14207 @option{-vP1} means Medium;
14208 @option{-vP2} means High.
14212 There are three possible options for this qualifier: DEFAULT, MEDIUM and
14217 The default is ^Default^DEFAULT^: no output for syntactically correct
14220 If several @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}} switches are present,
14221 only the last one is used.
14223 @item ^-aP^/ADD_PROJECT_SEARCH_DIR=^<dir>
14224 @cindex @option{^-aP^/ADD_PROJECT_SEARCH_DIR=^} (any project-aware tool)
14225 Add directory <dir> at the beginning of the project search path, in order,
14226 after the current working directory.
14230 @cindex @option{-eL} (any project-aware tool)
14231 Follow all symbolic links when processing project files.
14234 @item ^--subdirs^/SUBDIRS^=<subdir>
14235 @cindex @option{^--subdirs^/SUBDIRS^=} (gnatmake and gnatclean)
14236 This switch is recognized by gnatmake and gnatclean. It indicate that the real
14237 directories (except the source directories) are the subdirectories <subdir>
14238 of the directories specified in the project files. This applies in particular
14239 to object directories, library directories and exec directories. If the
14240 subdirectories do not exist, they are created automatically.
14244 @c **********************************
14245 @c * Tools Supporting Project Files *
14246 @c **********************************
14248 @node Tools Supporting Project Files
14249 @section Tools Supporting Project Files
14252 * gnatmake and Project Files::
14253 * The GNAT Driver and Project Files::
14256 @node gnatmake and Project Files
14257 @subsection gnatmake and Project Files
14260 This section covers several topics related to @command{gnatmake} and
14261 project files: defining ^switches^switches^ for @command{gnatmake}
14262 and for the tools that it invokes; specifying configuration pragmas;
14263 the use of the @code{Main} attribute; building and rebuilding library project
14267 * ^Switches^Switches^ and Project Files::
14268 * Specifying Configuration Pragmas::
14269 * Project Files and Main Subprograms::
14270 * Library Project Files::
14273 @node ^Switches^Switches^ and Project Files
14274 @subsubsection ^Switches^Switches^ and Project Files
14277 It is not currently possible to specify VMS style qualifiers in the project
14278 files; only Unix style ^switches^switches^ may be specified.
14282 For each of the packages @code{Builder}, @code{Compiler}, @code{Binder}, and
14283 @code{Linker}, you can specify a @code{^Default_Switches^Default_Switches^}
14284 attribute, a @code{^Switches^Switches^} attribute, or both;
14285 as their names imply, these ^switch^switch^-related
14286 attributes affect the ^switches^switches^ that are used for each of these GNAT
14288 @command{gnatmake} is invoked. As will be explained below, these
14289 component-specific ^switches^switches^ precede
14290 the ^switches^switches^ provided on the @command{gnatmake} command line.
14292 The @code{^Default_Switches^Default_Switches^} attribute is an associative
14293 array indexed by language name (case insensitive) whose value is a string list.
14296 @smallexample @c projectfile
14298 package Compiler is
14299 for ^Default_Switches^Default_Switches^ ("Ada")
14300 use ("^-gnaty^-gnaty^",
14307 The @code{^Switches^Switches^} attribute is also an associative array,
14308 indexed by a file name (which may or may not be case sensitive, depending
14309 on the operating system) whose value is a string list. For example:
14311 @smallexample @c projectfile
14314 for ^Switches^Switches^ ("main1.adb")
14316 for ^Switches^Switches^ ("main2.adb")
14323 For the @code{Builder} package, the file names must designate source files
14324 for main subprograms. For the @code{Binder} and @code{Linker} packages, the
14325 file names must designate @file{ALI} or source files for main subprograms.
14326 In each case just the file name without an explicit extension is acceptable.
14328 For each tool used in a program build (@command{gnatmake}, the compiler, the
14329 binder, and the linker), the corresponding package @dfn{contributes} a set of
14330 ^switches^switches^ for each file on which the tool is invoked, based on the
14331 ^switch^switch^-related attributes defined in the package.
14332 In particular, the ^switches^switches^
14333 that each of these packages contributes for a given file @var{f} comprise:
14337 the value of attribute @code{^Switches^Switches^ (@var{f})},
14338 if it is specified in the package for the given file,
14340 otherwise, the value of @code{^Default_Switches^Default_Switches^ ("Ada")},
14341 if it is specified in the package.
14345 If neither of these attributes is defined in the package, then the package does
14346 not contribute any ^switches^switches^ for the given file.
14348 When @command{gnatmake} is invoked on a file, the ^switches^switches^ comprise
14349 two sets, in the following order: those contributed for the file
14350 by the @code{Builder} package;
14351 and the switches passed on the command line.
14353 When @command{gnatmake} invokes a tool (compiler, binder, linker) on a file,
14354 the ^switches^switches^ passed to the tool comprise three sets,
14355 in the following order:
14359 the applicable ^switches^switches^ contributed for the file
14360 by the @code{Builder} package in the project file supplied on the command line;
14363 those contributed for the file by the package (in the relevant project file --
14364 see below) corresponding to the tool; and
14367 the applicable switches passed on the command line.
14371 The term @emph{applicable ^switches^switches^} reflects the fact that
14372 @command{gnatmake} ^switches^switches^ may or may not be passed to individual
14373 tools, depending on the individual ^switch^switch^.
14375 @command{gnatmake} may invoke the compiler on source files from different
14376 projects. The Project Manager will use the appropriate project file to
14377 determine the @code{Compiler} package for each source file being compiled.
14378 Likewise for the @code{Binder} and @code{Linker} packages.
14380 As an example, consider the following package in a project file:
14382 @smallexample @c projectfile
14385 package Compiler is
14386 for ^Default_Switches^Default_Switches^ ("Ada")
14388 for ^Switches^Switches^ ("a.adb")
14390 for ^Switches^Switches^ ("b.adb")
14392 "^-gnaty^-gnaty^");
14399 If @command{gnatmake} is invoked with this project file, and it needs to
14400 compile, say, the files @file{a.adb}, @file{b.adb}, and @file{c.adb}, then
14401 @file{a.adb} will be compiled with the ^switch^switch^
14402 @option{^-O1^-O1^},
14403 @file{b.adb} with ^switches^switches^
14405 and @option{^-gnaty^-gnaty^},
14406 and @file{c.adb} with @option{^-g^-g^}.
14408 The following example illustrates the ordering of the ^switches^switches^
14409 contributed by different packages:
14411 @smallexample @c projectfile
14415 for ^Switches^Switches^ ("main.adb")
14423 package Compiler is
14424 for ^Switches^Switches^ ("main.adb")
14432 If you issue the command:
14435 gnatmake ^-Pproj2^/PROJECT_FILE=PROJ2^ -O0 main
14439 then the compiler will be invoked on @file{main.adb} with the following
14440 sequence of ^switches^switches^
14443 ^-g -O1 -O2 -O0^-g -O1 -O2 -O0^
14446 with the last @option{^-O^-O^}
14447 ^switch^switch^ having precedence over the earlier ones;
14448 several other ^switches^switches^
14449 (such as @option{^-c^-c^}) are added implicitly.
14451 The ^switches^switches^
14453 and @option{^-O1^-O1^} are contributed by package
14454 @code{Builder}, @option{^-O2^-O2^} is contributed
14455 by the package @code{Compiler}
14456 and @option{^-O0^-O0^} comes from the command line.
14458 The @option{^-g^-g^}
14459 ^switch^switch^ will also be passed in the invocation of
14460 @command{Gnatlink.}
14462 A final example illustrates switch contributions from packages in different
14465 @smallexample @c projectfile
14468 for Source_Files use ("pack.ads", "pack.adb");
14469 package Compiler is
14470 for ^Default_Switches^Default_Switches^ ("Ada")
14471 use ("^-gnata^-gnata^");
14479 for Source_Files use ("foo_main.adb", "bar_main.adb");
14481 for ^Switches^Switches^ ("foo_main.adb")
14489 -- Ada source file:
14491 procedure Foo_Main is
14499 gnatmake ^-PProj4^/PROJECT_FILE=PROJ4^ foo_main.adb -cargs -gnato
14503 then the ^switches^switches^ passed to the compiler for @file{foo_main.adb} are
14504 @option{^-g^-g^} (contributed by the package @code{Proj4.Builder}) and
14505 @option{^-gnato^-gnato^} (passed on the command line).
14506 When the imported package @code{Pack} is compiled, the ^switches^switches^ used
14507 are @option{^-g^-g^} from @code{Proj4.Builder},
14508 @option{^-gnata^-gnata^} (contributed from package @code{Proj3.Compiler},
14509 and @option{^-gnato^-gnato^} from the command line.
14512 When using @command{gnatmake} with project files, some ^switches^switches^ or
14513 arguments may be expressed as relative paths. As the working directory where
14514 compilation occurs may change, these relative paths are converted to absolute
14515 paths. For the ^switches^switches^ found in a project file, the relative paths
14516 are relative to the project file directory, for the switches on the command
14517 line, they are relative to the directory where @command{gnatmake} is invoked.
14518 The ^switches^switches^ for which this occurs are:
14524 ^-aI^-aI^, as well as all arguments that are not switches (arguments to
14526 ^-o^-o^, object files specified in package @code{Linker} or after
14527 -largs on the command line). The exception to this rule is the ^switch^switch^
14528 ^--RTS=^--RTS=^ for which a relative path argument is never converted.
14530 @node Specifying Configuration Pragmas
14531 @subsubsection Specifying Configuration Pragmas
14533 When using @command{gnatmake} with project files, if there exists a file
14534 @file{gnat.adc} that contains configuration pragmas, this file will be
14537 Configuration pragmas can be defined by means of the following attributes in
14538 project files: @code{Global_Configuration_Pragmas} in package @code{Builder}
14539 and @code{Local_Configuration_Pragmas} in package @code{Compiler}.
14541 Both these attributes are single string attributes. Their values is the path
14542 name of a file containing configuration pragmas. If a path name is relative,
14543 then it is relative to the project directory of the project file where the
14544 attribute is defined.
14546 When compiling a source, the configuration pragmas used are, in order,
14547 those listed in the file designated by attribute
14548 @code{Global_Configuration_Pragmas} in package @code{Builder} of the main
14549 project file, if it is specified, and those listed in the file designated by
14550 attribute @code{Local_Configuration_Pragmas} in package @code{Compiler} of
14551 the project file of the source, if it exists.
14553 @node Project Files and Main Subprograms
14554 @subsubsection Project Files and Main Subprograms
14557 When using a project file, you can invoke @command{gnatmake}
14558 with one or several main subprograms, by specifying their source files on the
14562 gnatmake ^-P^/PROJECT_FILE=^prj main1 main2 main3
14566 Each of these needs to be a source file of the same project, except
14567 when the switch ^-u^/UNIQUE^ is used.
14570 When ^-u^/UNIQUE^ is not used, all the mains need to be sources of the
14571 same project, one of the project in the tree rooted at the project specified
14572 on the command line. The package @code{Builder} of this common project, the
14573 "main project" is the one that is considered by @command{gnatmake}.
14576 When ^-u^/UNIQUE^ is used, the specified source files may be in projects
14577 imported directly or indirectly by the project specified on the command line.
14578 Note that if such a source file is not part of the project specified on the
14579 command line, the ^switches^switches^ found in package @code{Builder} of the
14580 project specified on the command line, if any, that are transmitted
14581 to the compiler will still be used, not those found in the project file of
14585 When using a project file, you can also invoke @command{gnatmake} without
14586 explicitly specifying any main, and the effect depends on whether you have
14587 defined the @code{Main} attribute. This attribute has a string list value,
14588 where each element in the list is the name of a source file (the file
14589 extension is optional) that contains a unit that can be a main subprogram.
14591 If the @code{Main} attribute is defined in a project file as a non-empty
14592 string list and the switch @option{^-u^/UNIQUE^} is not used on the command
14593 line, then invoking @command{gnatmake} with this project file but without any
14594 main on the command line is equivalent to invoking @command{gnatmake} with all
14595 the file names in the @code{Main} attribute on the command line.
14598 @smallexample @c projectfile
14601 for Main use ("main1", "main2", "main3");
14607 With this project file, @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^"}
14609 @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^ main1 main2 main3"}.
14611 When the project attribute @code{Main} is not specified, or is specified
14612 as an empty string list, or when the switch @option{-u} is used on the command
14613 line, then invoking @command{gnatmake} with no main on the command line will
14614 result in all immediate sources of the project file being checked, and
14615 potentially recompiled. Depending on the presence of the switch @option{-u},
14616 sources from other project files on which the immediate sources of the main
14617 project file depend are also checked and potentially recompiled. In other
14618 words, the @option{-u} switch is applied to all of the immediate sources of the
14621 When no main is specified on the command line and attribute @code{Main} exists
14622 and includes several mains, or when several mains are specified on the
14623 command line, the default ^switches^switches^ in package @code{Builder} will
14624 be used for all mains, even if there are specific ^switches^switches^
14625 specified for one or several mains.
14627 But the ^switches^switches^ from package @code{Binder} or @code{Linker} will be
14628 the specific ^switches^switches^ for each main, if they are specified.
14630 @node Library Project Files
14631 @subsubsection Library Project Files
14634 When @command{gnatmake} is invoked with a main project file that is a library
14635 project file, it is not allowed to specify one or more mains on the command
14639 When a library project file is specified, switches ^-b^/ACTION=BIND^ and
14640 ^-l^/ACTION=LINK^ have special meanings.
14643 @item ^-b^/ACTION=BIND^ is only allowed for stand-alone libraries. It indicates
14644 to @command{gnatmake} that @command{gnatbind} should be invoked for the
14647 @item ^-l^/ACTION=LINK^ may be used for all library projects. It indicates
14648 to @command{gnatmake} that the binder generated file should be compiled
14649 (in the case of a stand-alone library) and that the library should be built.
14653 @node The GNAT Driver and Project Files
14654 @subsection The GNAT Driver and Project Files
14657 A number of GNAT tools, other than @command{^gnatmake^gnatmake^}
14658 can benefit from project files:
14659 @command{^gnatbind^gnatbind^},
14660 @command{^gnatcheck^gnatcheck^}),
14661 @command{^gnatclean^gnatclean^}),
14662 @command{^gnatelim^gnatelim^},
14663 @command{^gnatfind^gnatfind^},
14664 @command{^gnatlink^gnatlink^},
14665 @command{^gnatls^gnatls^},
14666 @command{^gnatmetric^gnatmetric^},
14667 @command{^gnatpp^gnatpp^},
14668 @command{^gnatstub^gnatstub^},
14669 and @command{^gnatxref^gnatxref^}. However, none of these tools can be invoked
14670 directly with a project file switch (@option{^-P^/PROJECT_FILE=^}).
14671 They must be invoked through the @command{gnat} driver.
14673 The @command{gnat} driver is a wrapper that accepts a number of commands and
14674 calls the corresponding tool. It was designed initially for VMS platforms (to
14675 convert VMS qualifiers to Unix-style switches), but it is now available on all
14678 On non-VMS platforms, the @command{gnat} driver accepts the following commands
14679 (case insensitive):
14683 BIND to invoke @command{^gnatbind^gnatbind^}
14685 CHOP to invoke @command{^gnatchop^gnatchop^}
14687 CLEAN to invoke @command{^gnatclean^gnatclean^}
14689 COMP or COMPILE to invoke the compiler
14691 ELIM to invoke @command{^gnatelim^gnatelim^}
14693 FIND to invoke @command{^gnatfind^gnatfind^}
14695 KR or KRUNCH to invoke @command{^gnatkr^gnatkr^}
14697 LINK to invoke @command{^gnatlink^gnatlink^}
14699 LS or LIST to invoke @command{^gnatls^gnatls^}
14701 MAKE to invoke @command{^gnatmake^gnatmake^}
14703 NAME to invoke @command{^gnatname^gnatname^}
14705 PREP or PREPROCESS to invoke @command{^gnatprep^gnatprep^}
14707 PP or PRETTY to invoke @command{^gnatpp^gnatpp^}
14709 METRIC to invoke @command{^gnatmetric^gnatmetric^}
14711 STUB to invoke @command{^gnatstub^gnatstub^}
14713 XREF to invoke @command{^gnatxref^gnatxref^}
14717 (note that the compiler is invoked using the command
14718 @command{^gnatmake -f -u -c^gnatmake -f -u -c^}).
14721 On non-VMS platforms, between @command{gnat} and the command, two
14722 special switches may be used:
14726 @command{-v} to display the invocation of the tool.
14728 @command{-dn} to prevent the @command{gnat} driver from removing
14729 the temporary files it has created. These temporary files are
14730 configuration files and temporary file list files.
14734 The command may be followed by switches and arguments for the invoked
14738 gnat bind -C main.ali
14744 Switches may also be put in text files, one switch per line, and the text
14745 files may be specified with their path name preceded by '@@'.
14748 gnat bind @@args.txt main.ali
14752 In addition, for commands BIND, COMP or COMPILE, FIND, ELIM, LS or LIST, LINK,
14753 METRIC, PP or PRETTY, STUB and XREF, the project file related switches
14754 (@option{^-P^/PROJECT_FILE^},
14755 @option{^-X^/EXTERNAL_REFERENCE^} and
14756 @option{^-vP^/MESSAGES_PROJECT_FILE=^x}) may be used in addition to
14757 the switches of the invoking tool.
14760 When GNAT PP or GNAT PRETTY is used with a project file, but with no source
14761 specified on the command line, it invokes @command{^gnatpp^gnatpp^} with all
14762 the immediate sources of the specified project file.
14765 When GNAT METRIC is used with a project file, but with no source
14766 specified on the command line, it invokes @command{^gnatmetric^gnatmetric^}
14767 with all the immediate sources of the specified project file and with
14768 @option{^-d^/DIRECTORY^} with the parameter pointing to the object directory
14772 In addition, when GNAT PP, GNAT PRETTY or GNAT METRIC is used with
14773 a project file, no source is specified on the command line and
14774 switch ^-U^/ALL_PROJECTS^ is specified on the command line, then
14775 the underlying tool (^gnatpp^gnatpp^ or
14776 ^gnatmetric^gnatmetric^) is invoked for all sources of all projects,
14777 not only for the immediate sources of the main project.
14779 (-U stands for Universal or Union of the project files of the project tree)
14783 For each of the following commands, there is optionally a corresponding
14784 package in the main project.
14788 package @code{Binder} for command BIND (invoking @code{^gnatbind^gnatbind^})
14791 package @code{Check} for command CHECK (invoking
14792 @code{^gnatcheck^gnatcheck^})
14795 package @code{Compiler} for command COMP or COMPILE (invoking the compiler)
14798 package @code{Cross_Reference} for command XREF (invoking
14799 @code{^gnatxref^gnatxref^})
14802 package @code{Eliminate} for command ELIM (invoking
14803 @code{^gnatelim^gnatelim^})
14806 package @code{Finder} for command FIND (invoking @code{^gnatfind^gnatfind^})
14809 package @code{Gnatls} for command LS or LIST (invoking @code{^gnatls^gnatls^})
14812 package @code{Gnatstub} for command STUB
14813 (invoking @code{^gnatstub^gnatstub^})
14816 package @code{Linker} for command LINK (invoking @code{^gnatlink^gnatlink^})
14819 package @code{Metrics} for command METRIC
14820 (invoking @code{^gnatmetric^gnatmetric^})
14823 package @code{Pretty_Printer} for command PP or PRETTY
14824 (invoking @code{^gnatpp^gnatpp^})
14829 Package @code{Gnatls} has a unique attribute @code{^Switches^Switches^},
14830 a simple variable with a string list value. It contains ^switches^switches^
14831 for the invocation of @code{^gnatls^gnatls^}.
14833 @smallexample @c projectfile
14837 for ^Switches^Switches^
14846 All other packages have two attribute @code{^Switches^Switches^} and
14847 @code{^Default_Switches^Default_Switches^}.
14850 @code{^Switches^Switches^} is an associative array attribute, indexed by the
14851 source file name, that has a string list value: the ^switches^switches^ to be
14852 used when the tool corresponding to the package is invoked for the specific
14856 @code{^Default_Switches^Default_Switches^} is an associative array attribute,
14857 indexed by the programming language that has a string list value.
14858 @code{^Default_Switches^Default_Switches^ ("Ada")} contains the
14859 ^switches^switches^ for the invocation of the tool corresponding
14860 to the package, except if a specific @code{^Switches^Switches^} attribute
14861 is specified for the source file.
14863 @smallexample @c projectfile
14867 for Source_Dirs use ("./**");
14870 for ^Switches^Switches^ use
14877 package Compiler is
14878 for ^Default_Switches^Default_Switches^ ("Ada")
14879 use ("^-gnatv^-gnatv^",
14880 "^-gnatwa^-gnatwa^");
14886 for ^Default_Switches^Default_Switches^ ("Ada")
14894 for ^Default_Switches^Default_Switches^ ("Ada")
14896 for ^Switches^Switches^ ("main.adb")
14905 for ^Default_Switches^Default_Switches^ ("Ada")
14912 package Cross_Reference is
14913 for ^Default_Switches^Default_Switches^ ("Ada")
14918 end Cross_Reference;
14924 With the above project file, commands such as
14927 ^gnat comp -Pproj main^GNAT COMP /PROJECT_FILE=PROJ MAIN^
14928 ^gnat ls -Pproj main^GNAT LIST /PROJECT_FILE=PROJ MAIN^
14929 ^gnat xref -Pproj main^GNAT XREF /PROJECT_FILE=PROJ MAIN^
14930 ^gnat bind -Pproj main.ali^GNAT BIND /PROJECT_FILE=PROJ MAIN.ALI^
14931 ^gnat link -Pproj main.ali^GNAT LINK /PROJECT_FILE=PROJ MAIN.ALI^
14935 will set up the environment properly and invoke the tool with the switches
14936 found in the package corresponding to the tool:
14937 @code{^Default_Switches^Default_Switches^ ("Ada")} for all tools,
14938 except @code{^Switches^Switches^ ("main.adb")}
14939 for @code{^gnatlink^gnatlink^}.
14940 It is also possible to invoke some of the tools,
14941 @code{^gnatcheck^gnatcheck^}),
14942 @code{^gnatmetric^gnatmetric^}),
14943 and @code{^gnatpp^gnatpp^})
14944 on a set of project units thanks to the combination of the switches
14945 @option{-P}, @option{-U} and possibly the main unit when one is interested
14946 in its closure. For instance,
14950 will compute the metrics for all the immediate units of project
14953 gnat metric -Pproj -U
14955 will compute the metrics for all the units of the closure of projects
14956 rooted at @code{proj}.
14958 gnat metric -Pproj -U main_unit
14960 will compute the metrics for the closure of units rooted at
14961 @code{main_unit}. This last possibility relies implicitly
14962 on @command{gnatbind}'s option @option{-R}.
14964 @c **********************
14965 @node An Extended Example
14966 @section An Extended Example
14969 Suppose that we have two programs, @var{prog1} and @var{prog2},
14970 whose sources are in corresponding directories. We would like
14971 to build them with a single @command{gnatmake} command, and we want to place
14972 their object files into @file{build} subdirectories of the source directories.
14973 Furthermore, we want to have to have two separate subdirectories
14974 in @file{build} -- @file{release} and @file{debug} -- which will contain
14975 the object files compiled with different set of compilation flags.
14977 In other words, we have the following structure:
14994 Here are the project files that we must place in a directory @file{main}
14995 to maintain this structure:
14999 @item We create a @code{Common} project with a package @code{Compiler} that
15000 specifies the compilation ^switches^switches^:
15005 @b{project} Common @b{is}
15007 @b{for} Source_Dirs @b{use} (); -- No source files
15011 @b{type} Build_Type @b{is} ("release", "debug");
15012 Build : Build_Type := External ("BUILD", "debug");
15015 @b{package} Compiler @b{is}
15016 @b{case} Build @b{is}
15017 @b{when} "release" =>
15018 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15019 @b{use} ("^-O2^-O2^");
15020 @b{when} "debug" =>
15021 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15022 @b{use} ("^-g^-g^");
15030 @item We create separate projects for the two programs:
15037 @b{project} Prog1 @b{is}
15039 @b{for} Source_Dirs @b{use} ("prog1");
15040 @b{for} Object_Dir @b{use} "prog1/build/" & Common.Build;
15042 @b{package} Compiler @b{renames} Common.Compiler;
15053 @b{project} Prog2 @b{is}
15055 @b{for} Source_Dirs @b{use} ("prog2");
15056 @b{for} Object_Dir @b{use} "prog2/build/" & Common.Build;
15058 @b{package} Compiler @b{renames} Common.Compiler;
15064 @item We create a wrapping project @code{Main}:
15073 @b{project} Main @b{is}
15075 @b{package} Compiler @b{renames} Common.Compiler;
15081 @item Finally we need to create a dummy procedure that @code{with}s (either
15082 explicitly or implicitly) all the sources of our two programs.
15087 Now we can build the programs using the command
15090 gnatmake ^-P^/PROJECT_FILE=^main dummy
15094 for the Debug mode, or
15098 gnatmake -Pmain -XBUILD=release
15104 GNAT MAKE /PROJECT_FILE=main /EXTERNAL_REFERENCE=BUILD=release
15109 for the Release mode.
15111 @c ********************************
15112 @c * Project File Complete Syntax *
15113 @c ********************************
15115 @node Project File Complete Syntax
15116 @section Project File Complete Syntax
15120 context_clause project_declaration
15126 @b{with} path_name @{ , path_name @} ;
15131 project_declaration ::=
15132 simple_project_declaration | project_extension
15134 simple_project_declaration ::=
15135 @b{project} <project_>simple_name @b{is}
15136 @{declarative_item@}
15137 @b{end} <project_>simple_name;
15139 project_extension ::=
15140 @b{project} <project_>simple_name @b{extends} path_name @b{is}
15141 @{declarative_item@}
15142 @b{end} <project_>simple_name;
15144 declarative_item ::=
15145 package_declaration |
15146 typed_string_declaration |
15147 other_declarative_item
15149 package_declaration ::=
15150 package_spec | package_renaming
15153 @b{package} package_identifier @b{is}
15154 @{simple_declarative_item@}
15155 @b{end} package_identifier ;
15157 package_identifier ::=
15158 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
15159 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
15160 @code{^gnatls^gnatls^} | @code{IDE} | @code{Pretty_Printer}
15162 package_renaming ::==
15163 @b{package} package_identifier @b{renames}
15164 <project_>simple_name.package_identifier ;
15166 typed_string_declaration ::=
15167 @b{type} <typed_string_>_simple_name @b{is}
15168 ( string_literal @{, string_literal@} );
15170 other_declarative_item ::=
15171 attribute_declaration |
15172 typed_variable_declaration |
15173 variable_declaration |
15176 attribute_declaration ::=
15177 full_associative_array_declaration |
15178 @b{for} attribute_designator @b{use} expression ;
15180 full_associative_array_declaration ::=
15181 @b{for} <associative_array_attribute_>simple_name @b{use}
15182 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
15184 attribute_designator ::=
15185 <simple_attribute_>simple_name |
15186 <associative_array_attribute_>simple_name ( string_literal )
15188 typed_variable_declaration ::=
15189 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
15191 variable_declaration ::=
15192 <variable_>simple_name := expression;
15202 attribute_reference
15208 ( <string_>expression @{ , <string_>expression @} )
15211 @b{external} ( string_literal [, string_literal] )
15213 attribute_reference ::=
15214 attribute_prefix ' <simple_attribute_>simple_name [ ( literal_string ) ]
15216 attribute_prefix ::=
15218 <project_>simple_name | package_identifier |
15219 <project_>simple_name . package_identifier
15221 case_construction ::=
15222 @b{case} <typed_variable_>name @b{is}
15227 @b{when} discrete_choice_list =>
15228 @{case_construction | attribute_declaration@}
15230 discrete_choice_list ::=
15231 string_literal @{| string_literal@} |
15235 simple_name @{. simple_name@}
15238 identifier (same as Ada)
15242 @node The Cross-Referencing Tools gnatxref and gnatfind
15243 @chapter The Cross-Referencing Tools @code{gnatxref} and @code{gnatfind}
15248 The compiler generates cross-referencing information (unless
15249 you set the @samp{-gnatx} switch), which are saved in the @file{.ali} files.
15250 This information indicates where in the source each entity is declared and
15251 referenced. Note that entities in package Standard are not included, but
15252 entities in all other predefined units are included in the output.
15254 Before using any of these two tools, you need to compile successfully your
15255 application, so that GNAT gets a chance to generate the cross-referencing
15258 The two tools @code{gnatxref} and @code{gnatfind} take advantage of this
15259 information to provide the user with the capability to easily locate the
15260 declaration and references to an entity. These tools are quite similar,
15261 the difference being that @code{gnatfind} is intended for locating
15262 definitions and/or references to a specified entity or entities, whereas
15263 @code{gnatxref} is oriented to generating a full report of all
15266 To use these tools, you must not compile your application using the
15267 @option{-gnatx} switch on the @command{gnatmake} command line
15268 (@pxref{The GNAT Make Program gnatmake}). Otherwise, cross-referencing
15269 information will not be generated.
15271 Note: to invoke @code{gnatxref} or @code{gnatfind} with a project file,
15272 use the @code{gnat} driver (see @ref{The GNAT Driver and Project Files}).
15275 * gnatxref Switches::
15276 * gnatfind Switches::
15277 * Project Files for gnatxref and gnatfind::
15278 * Regular Expressions in gnatfind and gnatxref::
15279 * Examples of gnatxref Usage::
15280 * Examples of gnatfind Usage::
15283 @node gnatxref Switches
15284 @section @code{gnatxref} Switches
15287 The command invocation for @code{gnatxref} is:
15289 $ gnatxref @ovar{switches} @var{sourcefile1} @r{[}@var{sourcefile2} @dots{}@r{]}
15298 identifies the source files for which a report is to be generated. The
15299 ``with''ed units will be processed too. You must provide at least one file.
15301 These file names are considered to be regular expressions, so for instance
15302 specifying @file{source*.adb} is the same as giving every file in the current
15303 directory whose name starts with @file{source} and whose extension is
15306 You shouldn't specify any directory name, just base names. @command{gnatxref}
15307 and @command{gnatfind} will be able to locate these files by themselves using
15308 the source path. If you specify directories, no result is produced.
15313 The switches can be:
15317 @cindex @option{--version} @command{gnatxref}
15318 Display Copyright and version, then exit disregarding all other options.
15321 @cindex @option{--help} @command{gnatxref}
15322 If @option{--version} was not used, display usage, then exit disregarding
15325 @item ^-a^/ALL_FILES^
15326 @cindex @option{^-a^/ALL_FILES^} (@command{gnatxref})
15327 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15328 the read-only files found in the library search path. Otherwise, these files
15329 will be ignored. This option can be used to protect Gnat sources or your own
15330 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15331 much faster, and their output much smaller. Read-only here refers to access
15332 or permissions status in the file system for the current user.
15335 @cindex @option{-aIDIR} (@command{gnatxref})
15336 When looking for source files also look in directory DIR. The order in which
15337 source file search is undertaken is the same as for @command{gnatmake}.
15340 @cindex @option{-aODIR} (@command{gnatxref})
15341 When searching for library and object files, look in directory
15342 DIR. The order in which library files are searched is the same as for
15343 @command{gnatmake}.
15346 @cindex @option{-nostdinc} (@command{gnatxref})
15347 Do not look for sources in the system default directory.
15350 @cindex @option{-nostdlib} (@command{gnatxref})
15351 Do not look for library files in the system default directory.
15353 @item --RTS=@var{rts-path}
15354 @cindex @option{--RTS} (@command{gnatxref})
15355 Specifies the default location of the runtime library. Same meaning as the
15356 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15358 @item ^-d^/DERIVED_TYPES^
15359 @cindex @option{^-d^/DERIVED_TYPES^} (@command{gnatxref})
15360 If this switch is set @code{gnatxref} will output the parent type
15361 reference for each matching derived types.
15363 @item ^-f^/FULL_PATHNAME^
15364 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatxref})
15365 If this switch is set, the output file names will be preceded by their
15366 directory (if the file was found in the search path). If this switch is
15367 not set, the directory will not be printed.
15369 @item ^-g^/IGNORE_LOCALS^
15370 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatxref})
15371 If this switch is set, information is output only for library-level
15372 entities, ignoring local entities. The use of this switch may accelerate
15373 @code{gnatfind} and @code{gnatxref}.
15376 @cindex @option{-IDIR} (@command{gnatxref})
15377 Equivalent to @samp{-aODIR -aIDIR}.
15380 @cindex @option{-pFILE} (@command{gnatxref})
15381 Specify a project file to use @xref{Project Files}.
15382 If you need to use the @file{.gpr}
15383 project files, you should use gnatxref through the GNAT driver
15384 (@command{gnat xref -Pproject}).
15386 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15387 project file in the current directory.
15389 If a project file is either specified or found by the tools, then the content
15390 of the source directory and object directory lines are added as if they
15391 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^}
15392 and @samp{^-aO^OBJECT_SEARCH^}.
15394 Output only unused symbols. This may be really useful if you give your
15395 main compilation unit on the command line, as @code{gnatxref} will then
15396 display every unused entity and 'with'ed package.
15400 Instead of producing the default output, @code{gnatxref} will generate a
15401 @file{tags} file that can be used by vi. For examples how to use this
15402 feature, see @ref{Examples of gnatxref Usage}. The tags file is output
15403 to the standard output, thus you will have to redirect it to a file.
15409 All these switches may be in any order on the command line, and may even
15410 appear after the file names. They need not be separated by spaces, thus
15411 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15412 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15414 @node gnatfind Switches
15415 @section @code{gnatfind} Switches
15418 The command line for @code{gnatfind} is:
15421 $ gnatfind @ovar{switches} @var{pattern}@r{[}:@var{sourcefile}@r{[}:@var{line}@r{[}:@var{column}@r{]]]}
15422 @r{[}@var{file1} @var{file2} @dots{}]
15430 An entity will be output only if it matches the regular expression found
15431 in @var{pattern}, see @ref{Regular Expressions in gnatfind and gnatxref}.
15433 Omitting the pattern is equivalent to specifying @samp{*}, which
15434 will match any entity. Note that if you do not provide a pattern, you
15435 have to provide both a sourcefile and a line.
15437 Entity names are given in Latin-1, with uppercase/lowercase equivalence
15438 for matching purposes. At the current time there is no support for
15439 8-bit codes other than Latin-1, or for wide characters in identifiers.
15442 @code{gnatfind} will look for references, bodies or declarations
15443 of symbols referenced in @file{@var{sourcefile}}, at line @var{line}
15444 and column @var{column}. See @ref{Examples of gnatfind Usage}
15445 for syntax examples.
15448 is a decimal integer identifying the line number containing
15449 the reference to the entity (or entities) to be located.
15452 is a decimal integer identifying the exact location on the
15453 line of the first character of the identifier for the
15454 entity reference. Columns are numbered from 1.
15456 @item file1 file2 @dots{}
15457 The search will be restricted to these source files. If none are given, then
15458 the search will be done for every library file in the search path.
15459 These file must appear only after the pattern or sourcefile.
15461 These file names are considered to be regular expressions, so for instance
15462 specifying @file{source*.adb} is the same as giving every file in the current
15463 directory whose name starts with @file{source} and whose extension is
15466 The location of the spec of the entity will always be displayed, even if it
15467 isn't in one of @file{@var{file1}}, @file{@var{file2}},@enddots{} The
15468 occurrences of the entity in the separate units of the ones given on the
15469 command line will also be displayed.
15471 Note that if you specify at least one file in this part, @code{gnatfind} may
15472 sometimes not be able to find the body of the subprograms.
15477 At least one of 'sourcefile' or 'pattern' has to be present on
15480 The following switches are available:
15484 @cindex @option{--version} @command{gnatfind}
15485 Display Copyright and version, then exit disregarding all other options.
15488 @cindex @option{--help} @command{gnatfind}
15489 If @option{--version} was not used, display usage, then exit disregarding
15492 @item ^-a^/ALL_FILES^
15493 @cindex @option{^-a^/ALL_FILES^} (@command{gnatfind})
15494 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15495 the read-only files found in the library search path. Otherwise, these files
15496 will be ignored. This option can be used to protect Gnat sources or your own
15497 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15498 much faster, and their output much smaller. Read-only here refers to access
15499 or permission status in the file system for the current user.
15502 @cindex @option{-aIDIR} (@command{gnatfind})
15503 When looking for source files also look in directory DIR. The order in which
15504 source file search is undertaken is the same as for @command{gnatmake}.
15507 @cindex @option{-aODIR} (@command{gnatfind})
15508 When searching for library and object files, look in directory
15509 DIR. The order in which library files are searched is the same as for
15510 @command{gnatmake}.
15513 @cindex @option{-nostdinc} (@command{gnatfind})
15514 Do not look for sources in the system default directory.
15517 @cindex @option{-nostdlib} (@command{gnatfind})
15518 Do not look for library files in the system default directory.
15520 @item --RTS=@var{rts-path}
15521 @cindex @option{--RTS} (@command{gnatfind})
15522 Specifies the default location of the runtime library. Same meaning as the
15523 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15525 @item ^-d^/DERIVED_TYPE_INFORMATION^
15526 @cindex @option{^-d^/DERIVED_TYPE_INFORMATION^} (@code{gnatfind})
15527 If this switch is set, then @code{gnatfind} will output the parent type
15528 reference for each matching derived types.
15530 @item ^-e^/EXPRESSIONS^
15531 @cindex @option{^-e^/EXPRESSIONS^} (@command{gnatfind})
15532 By default, @code{gnatfind} accept the simple regular expression set for
15533 @samp{pattern}. If this switch is set, then the pattern will be
15534 considered as full Unix-style regular expression.
15536 @item ^-f^/FULL_PATHNAME^
15537 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatfind})
15538 If this switch is set, the output file names will be preceded by their
15539 directory (if the file was found in the search path). If this switch is
15540 not set, the directory will not be printed.
15542 @item ^-g^/IGNORE_LOCALS^
15543 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatfind})
15544 If this switch is set, information is output only for library-level
15545 entities, ignoring local entities. The use of this switch may accelerate
15546 @code{gnatfind} and @code{gnatxref}.
15549 @cindex @option{-IDIR} (@command{gnatfind})
15550 Equivalent to @samp{-aODIR -aIDIR}.
15553 @cindex @option{-pFILE} (@command{gnatfind})
15554 Specify a project file (@pxref{Project Files}) to use.
15555 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15556 project file in the current directory.
15558 If a project file is either specified or found by the tools, then the content
15559 of the source directory and object directory lines are added as if they
15560 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^} and
15561 @samp{^-aO^/OBJECT_SEARCH^}.
15563 @item ^-r^/REFERENCES^
15564 @cindex @option{^-r^/REFERENCES^} (@command{gnatfind})
15565 By default, @code{gnatfind} will output only the information about the
15566 declaration, body or type completion of the entities. If this switch is
15567 set, the @code{gnatfind} will locate every reference to the entities in
15568 the files specified on the command line (or in every file in the search
15569 path if no file is given on the command line).
15571 @item ^-s^/PRINT_LINES^
15572 @cindex @option{^-s^/PRINT_LINES^} (@command{gnatfind})
15573 If this switch is set, then @code{gnatfind} will output the content
15574 of the Ada source file lines were the entity was found.
15576 @item ^-t^/TYPE_HIERARCHY^
15577 @cindex @option{^-t^/TYPE_HIERARCHY^} (@command{gnatfind})
15578 If this switch is set, then @code{gnatfind} will output the type hierarchy for
15579 the specified type. It act like -d option but recursively from parent
15580 type to parent type. When this switch is set it is not possible to
15581 specify more than one file.
15586 All these switches may be in any order on the command line, and may even
15587 appear after the file names. They need not be separated by spaces, thus
15588 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15589 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15591 As stated previously, gnatfind will search in every directory in the
15592 search path. You can force it to look only in the current directory if
15593 you specify @code{*} at the end of the command line.
15595 @node Project Files for gnatxref and gnatfind
15596 @section Project Files for @command{gnatxref} and @command{gnatfind}
15599 Project files allow a programmer to specify how to compile its
15600 application, where to find sources, etc. These files are used
15602 primarily by GPS, but they can also be used
15605 @code{gnatxref} and @code{gnatfind}.
15607 A project file name must end with @file{.gpr}. If a single one is
15608 present in the current directory, then @code{gnatxref} and @code{gnatfind} will
15609 extract the information from it. If multiple project files are found, none of
15610 them is read, and you have to use the @samp{-p} switch to specify the one
15613 The following lines can be included, even though most of them have default
15614 values which can be used in most cases.
15615 The lines can be entered in any order in the file.
15616 Except for @file{src_dir} and @file{obj_dir}, you can only have one instance of
15617 each line. If you have multiple instances, only the last one is taken into
15622 [default: @code{"^./^[]^"}]
15623 specifies a directory where to look for source files. Multiple @code{src_dir}
15624 lines can be specified and they will be searched in the order they
15628 [default: @code{"^./^[]^"}]
15629 specifies a directory where to look for object and library files. Multiple
15630 @code{obj_dir} lines can be specified, and they will be searched in the order
15633 @item comp_opt=SWITCHES
15634 [default: @code{""}]
15635 creates a variable which can be referred to subsequently by using
15636 the @code{$@{comp_opt@}} notation. This is intended to store the default
15637 switches given to @command{gnatmake} and @command{gcc}.
15639 @item bind_opt=SWITCHES
15640 [default: @code{""}]
15641 creates a variable which can be referred to subsequently by using
15642 the @samp{$@{bind_opt@}} notation. This is intended to store the default
15643 switches given to @command{gnatbind}.
15645 @item link_opt=SWITCHES
15646 [default: @code{""}]
15647 creates a variable which can be referred to subsequently by using
15648 the @samp{$@{link_opt@}} notation. This is intended to store the default
15649 switches given to @command{gnatlink}.
15651 @item main=EXECUTABLE
15652 [default: @code{""}]
15653 specifies the name of the executable for the application. This variable can
15654 be referred to in the following lines by using the @samp{$@{main@}} notation.
15657 @item comp_cmd=COMMAND
15658 [default: @code{"GNAT COMPILE /SEARCH=$@{src_dir@} /DEBUG /TRY_SEMANTICS"}]
15661 @item comp_cmd=COMMAND
15662 [default: @code{"gcc -c -I$@{src_dir@} -g -gnatq"}]
15664 specifies the command used to compile a single file in the application.
15667 @item make_cmd=COMMAND
15668 [default: @code{"GNAT MAKE $@{main@}
15669 /SOURCE_SEARCH=$@{src_dir@} /OBJECT_SEARCH=$@{obj_dir@}
15670 /DEBUG /TRY_SEMANTICS /COMPILER_QUALIFIERS $@{comp_opt@}
15671 /BINDER_QUALIFIERS $@{bind_opt@} /LINKER_QUALIFIERS $@{link_opt@}"}]
15674 @item make_cmd=COMMAND
15675 [default: @code{"gnatmake $@{main@} -aI$@{src_dir@}
15676 -aO$@{obj_dir@} -g -gnatq -cargs $@{comp_opt@}
15677 -bargs $@{bind_opt@} -largs $@{link_opt@}"}]
15679 specifies the command used to recompile the whole application.
15681 @item run_cmd=COMMAND
15682 [default: @code{"$@{main@}"}]
15683 specifies the command used to run the application.
15685 @item debug_cmd=COMMAND
15686 [default: @code{"gdb $@{main@}"}]
15687 specifies the command used to debug the application
15692 @command{gnatxref} and @command{gnatfind} only take into account the
15693 @code{src_dir} and @code{obj_dir} lines, and ignore the others.
15695 @node Regular Expressions in gnatfind and gnatxref
15696 @section Regular Expressions in @code{gnatfind} and @code{gnatxref}
15699 As specified in the section about @command{gnatfind}, the pattern can be a
15700 regular expression. Actually, there are to set of regular expressions
15701 which are recognized by the program:
15704 @item globbing patterns
15705 These are the most usual regular expression. They are the same that you
15706 generally used in a Unix shell command line, or in a DOS session.
15708 Here is a more formal grammar:
15715 term ::= elmt -- matches elmt
15716 term ::= elmt elmt -- concatenation (elmt then elmt)
15717 term ::= * -- any string of 0 or more characters
15718 term ::= ? -- matches any character
15719 term ::= [char @{char@}] -- matches any character listed
15720 term ::= [char - char] -- matches any character in range
15724 @item full regular expression
15725 The second set of regular expressions is much more powerful. This is the
15726 type of regular expressions recognized by utilities such a @file{grep}.
15728 The following is the form of a regular expression, expressed in Ada
15729 reference manual style BNF is as follows
15736 regexp ::= term @{| term@} -- alternation (term or term @dots{})
15738 term ::= item @{item@} -- concatenation (item then item)
15740 item ::= elmt -- match elmt
15741 item ::= elmt * -- zero or more elmt's
15742 item ::= elmt + -- one or more elmt's
15743 item ::= elmt ? -- matches elmt or nothing
15746 elmt ::= nschar -- matches given character
15747 elmt ::= [nschar @{nschar@}] -- matches any character listed
15748 elmt ::= [^^^ nschar @{nschar@}] -- matches any character not listed
15749 elmt ::= [char - char] -- matches chars in given range
15750 elmt ::= \ char -- matches given character
15751 elmt ::= . -- matches any single character
15752 elmt ::= ( regexp ) -- parens used for grouping
15754 char ::= any character, including special characters
15755 nschar ::= any character except ()[].*+?^^^
15759 Following are a few examples:
15763 will match any of the two strings @samp{abcde} and @samp{fghi},
15766 will match any string like @samp{abd}, @samp{abcd}, @samp{abccd},
15767 @samp{abcccd}, and so on,
15770 will match any string which has only lowercase characters in it (and at
15771 least one character.
15776 @node Examples of gnatxref Usage
15777 @section Examples of @code{gnatxref} Usage
15779 @subsection General Usage
15782 For the following examples, we will consider the following units:
15784 @smallexample @c ada
15790 3: procedure Foo (B : in Integer);
15797 1: package body Main is
15798 2: procedure Foo (B : in Integer) is
15809 2: procedure Print (B : Integer);
15818 The first thing to do is to recompile your application (for instance, in
15819 that case just by doing a @samp{gnatmake main}, so that GNAT generates
15820 the cross-referencing information.
15821 You can then issue any of the following commands:
15823 @item gnatxref main.adb
15824 @code{gnatxref} generates cross-reference information for main.adb
15825 and every unit 'with'ed by main.adb.
15827 The output would be:
15835 Decl: main.ads 3:20
15836 Body: main.adb 2:20
15837 Ref: main.adb 4:13 5:13 6:19
15840 Ref: main.adb 6:8 7:8
15850 Decl: main.ads 3:15
15851 Body: main.adb 2:15
15854 Body: main.adb 1:14
15857 Ref: main.adb 6:12 7:12
15861 that is the entity @code{Main} is declared in main.ads, line 2, column 9,
15862 its body is in main.adb, line 1, column 14 and is not referenced any where.
15864 The entity @code{Print} is declared in bar.ads, line 2, column 15 and it
15865 it referenced in main.adb, line 6 column 12 and line 7 column 12.
15867 @item gnatxref package1.adb package2.ads
15868 @code{gnatxref} will generates cross-reference information for
15869 package1.adb, package2.ads and any other package 'with'ed by any
15875 @subsection Using gnatxref with vi
15877 @code{gnatxref} can generate a tags file output, which can be used
15878 directly from @command{vi}. Note that the standard version of @command{vi}
15879 will not work properly with overloaded symbols. Consider using another
15880 free implementation of @command{vi}, such as @command{vim}.
15883 $ gnatxref -v gnatfind.adb > tags
15887 will generate the tags file for @code{gnatfind} itself (if the sources
15888 are in the search path!).
15890 From @command{vi}, you can then use the command @samp{:tag @var{entity}}
15891 (replacing @var{entity} by whatever you are looking for), and vi will
15892 display a new file with the corresponding declaration of entity.
15895 @node Examples of gnatfind Usage
15896 @section Examples of @code{gnatfind} Usage
15900 @item gnatfind ^-f^/FULL_PATHNAME^ xyz:main.adb
15901 Find declarations for all entities xyz referenced at least once in
15902 main.adb. The references are search in every library file in the search
15905 The directories will be printed as well (as the @samp{^-f^/FULL_PATHNAME^}
15908 The output will look like:
15910 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
15911 ^directory/^[directory]^main.adb:24:10: xyz <= body
15912 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
15916 that is to say, one of the entities xyz found in main.adb is declared at
15917 line 12 of main.ads (and its body is in main.adb), and another one is
15918 declared at line 45 of foo.ads
15920 @item gnatfind ^-fs^/FULL_PATHNAME/SOURCE_LINE^ xyz:main.adb
15921 This is the same command as the previous one, instead @code{gnatfind} will
15922 display the content of the Ada source file lines.
15924 The output will look like:
15927 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
15929 ^directory/^[directory]^main.adb:24:10: xyz <= body
15931 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
15936 This can make it easier to find exactly the location your are looking
15939 @item gnatfind ^-r^/REFERENCES^ "*x*":main.ads:123 foo.adb
15940 Find references to all entities containing an x that are
15941 referenced on line 123 of main.ads.
15942 The references will be searched only in main.ads and foo.adb.
15944 @item gnatfind main.ads:123
15945 Find declarations and bodies for all entities that are referenced on
15946 line 123 of main.ads.
15948 This is the same as @code{gnatfind "*":main.adb:123}.
15950 @item gnatfind ^mydir/^[mydir]^main.adb:123:45
15951 Find the declaration for the entity referenced at column 45 in
15952 line 123 of file main.adb in directory mydir. Note that it
15953 is usual to omit the identifier name when the column is given,
15954 since the column position identifies a unique reference.
15956 The column has to be the beginning of the identifier, and should not
15957 point to any character in the middle of the identifier.
15961 @c *********************************
15962 @node The GNAT Pretty-Printer gnatpp
15963 @chapter The GNAT Pretty-Printer @command{gnatpp}
15965 @cindex Pretty-Printer
15968 ^The @command{gnatpp} tool^GNAT PRETTY^ is an ASIS-based utility
15969 for source reformatting / pretty-printing.
15970 It takes an Ada source file as input and generates a reformatted
15972 You can specify various style directives via switches; e.g.,
15973 identifier case conventions, rules of indentation, and comment layout.
15975 To produce a reformatted file, @command{gnatpp} generates and uses the ASIS
15976 tree for the input source and thus requires the input to be syntactically and
15977 semantically legal.
15978 If this condition is not met, @command{gnatpp} will terminate with an
15979 error message; no output file will be generated.
15981 If the source files presented to @command{gnatpp} contain
15982 preprocessing directives, then the output file will
15983 correspond to the generated source after all
15984 preprocessing is carried out. There is no way
15985 using @command{gnatpp} to obtain pretty printed files that
15986 include the preprocessing directives.
15988 If the compilation unit
15989 contained in the input source depends semantically upon units located
15990 outside the current directory, you have to provide the source search path
15991 when invoking @command{gnatpp}, if these units are contained in files with
15992 names that do not follow the GNAT file naming rules, you have to provide
15993 the configuration file describing the corresponding naming scheme;
15994 see the description of the @command{gnatpp}
15995 switches below. Another possibility is to use a project file and to
15996 call @command{gnatpp} through the @command{gnat} driver
15998 The @command{gnatpp} command has the form
16001 $ gnatpp @ovar{switches} @var{filename}
16008 @var{switches} is an optional sequence of switches defining such properties as
16009 the formatting rules, the source search path, and the destination for the
16013 @var{filename} is the name (including the extension) of the source file to
16014 reformat; ``wildcards'' or several file names on the same gnatpp command are
16015 allowed. The file name may contain path information; it does not have to
16016 follow the GNAT file naming rules
16020 * Switches for gnatpp::
16021 * Formatting Rules::
16024 @node Switches for gnatpp
16025 @section Switches for @command{gnatpp}
16028 The following subsections describe the various switches accepted by
16029 @command{gnatpp}, organized by category.
16032 You specify a switch by supplying a name and generally also a value.
16033 In many cases the values for a switch with a given name are incompatible with
16035 (for example the switch that controls the casing of a reserved word may have
16036 exactly one value: upper case, lower case, or
16037 mixed case) and thus exactly one such switch can be in effect for an
16038 invocation of @command{gnatpp}.
16039 If more than one is supplied, the last one is used.
16040 However, some values for the same switch are mutually compatible.
16041 You may supply several such switches to @command{gnatpp}, but then
16042 each must be specified in full, with both the name and the value.
16043 Abbreviated forms (the name appearing once, followed by each value) are
16045 For example, to set
16046 the alignment of the assignment delimiter both in declarations and in
16047 assignment statements, you must write @option{-A2A3}
16048 (or @option{-A2 -A3}), but not @option{-A23}.
16052 In many cases the set of options for a given qualifier are incompatible with
16053 each other (for example the qualifier that controls the casing of a reserved
16054 word may have exactly one option, which specifies either upper case, lower
16055 case, or mixed case), and thus exactly one such option can be in effect for
16056 an invocation of @command{gnatpp}.
16057 If more than one is supplied, the last one is used.
16058 However, some qualifiers have options that are mutually compatible,
16059 and then you may then supply several such options when invoking
16063 In most cases, it is obvious whether or not the
16064 ^values for a switch with a given name^options for a given qualifier^
16065 are compatible with each other.
16066 When the semantics might not be evident, the summaries below explicitly
16067 indicate the effect.
16070 * Alignment Control::
16072 * Construct Layout Control::
16073 * General Text Layout Control::
16074 * Other Formatting Options::
16075 * Setting the Source Search Path::
16076 * Output File Control::
16077 * Other gnatpp Switches::
16080 @node Alignment Control
16081 @subsection Alignment Control
16082 @cindex Alignment control in @command{gnatpp}
16085 Programs can be easier to read if certain constructs are vertically aligned.
16086 By default all alignments are set ON.
16087 Through the @option{^-A0^/ALIGN=OFF^} switch you may reset the default to
16088 OFF, and then use one or more of the other
16089 ^@option{-A@var{n}} switches^@option{/ALIGN} options^
16090 to activate alignment for specific constructs.
16093 @cindex @option{^-A@var{n}^/ALIGN^} (@command{gnatpp})
16097 Set all alignments to ON
16100 @item ^-A0^/ALIGN=OFF^
16101 Set all alignments to OFF
16103 @item ^-A1^/ALIGN=COLONS^
16104 Align @code{:} in declarations
16106 @item ^-A2^/ALIGN=DECLARATIONS^
16107 Align @code{:=} in initializations in declarations
16109 @item ^-A3^/ALIGN=STATEMENTS^
16110 Align @code{:=} in assignment statements
16112 @item ^-A4^/ALIGN=ARROWS^
16113 Align @code{=>} in associations
16115 @item ^-A5^/ALIGN=COMPONENT_CLAUSES^
16116 Align @code{at} keywords in the component clauses in record
16117 representation clauses
16121 The @option{^-A^/ALIGN^} switches are mutually compatible; any combination
16124 @node Casing Control
16125 @subsection Casing Control
16126 @cindex Casing control in @command{gnatpp}
16129 @command{gnatpp} allows you to specify the casing for reserved words,
16130 pragma names, attribute designators and identifiers.
16131 For identifiers you may define a
16132 general rule for name casing but also override this rule
16133 via a set of dictionary files.
16135 Three types of casing are supported: lower case, upper case, and mixed case.
16136 Lower and upper case are self-explanatory (but since some letters in
16137 Latin1 and other GNAT-supported character sets
16138 exist only in lower-case form, an upper case conversion will have no
16140 ``Mixed case'' means that the first letter, and also each letter immediately
16141 following an underscore, are converted to their uppercase forms;
16142 all the other letters are converted to their lowercase forms.
16145 @cindex @option{^-a@var{x}^/ATTRIBUTE^} (@command{gnatpp})
16146 @item ^-aL^/ATTRIBUTE_CASING=LOWER_CASE^
16147 Attribute designators are lower case
16149 @item ^-aU^/ATTRIBUTE_CASING=UPPER_CASE^
16150 Attribute designators are upper case
16152 @item ^-aM^/ATTRIBUTE_CASING=MIXED_CASE^
16153 Attribute designators are mixed case (this is the default)
16155 @cindex @option{^-k@var{x}^/KEYWORD_CASING^} (@command{gnatpp})
16156 @item ^-kL^/KEYWORD_CASING=LOWER_CASE^
16157 Keywords (technically, these are known in Ada as @emph{reserved words}) are
16158 lower case (this is the default)
16160 @item ^-kU^/KEYWORD_CASING=UPPER_CASE^
16161 Keywords are upper case
16163 @cindex @option{^-n@var{x}^/NAME_CASING^} (@command{gnatpp})
16164 @item ^-nD^/NAME_CASING=AS_DECLARED^
16165 Name casing for defining occurrences are as they appear in the source file
16166 (this is the default)
16168 @item ^-nU^/NAME_CASING=UPPER_CASE^
16169 Names are in upper case
16171 @item ^-nL^/NAME_CASING=LOWER_CASE^
16172 Names are in lower case
16174 @item ^-nM^/NAME_CASING=MIXED_CASE^
16175 Names are in mixed case
16177 @cindex @option{^-p@var{x}^/PRAGMA_CASING^} (@command{gnatpp})
16178 @item ^-pL^/PRAGMA_CASING=LOWER_CASE^
16179 Pragma names are lower case
16181 @item ^-pU^/PRAGMA_CASING=UPPER_CASE^
16182 Pragma names are upper case
16184 @item ^-pM^/PRAGMA_CASING=MIXED_CASE^
16185 Pragma names are mixed case (this is the default)
16187 @item ^-D@var{file}^/DICTIONARY=@var{file}^
16188 @cindex @option{^-D^/DICTIONARY^} (@command{gnatpp})
16189 Use @var{file} as a @emph{dictionary file} that defines
16190 the casing for a set of specified names,
16191 thereby overriding the effect on these names by
16192 any explicit or implicit
16193 ^-n^/NAME_CASING^ switch.
16194 To supply more than one dictionary file,
16195 use ^several @option{-D} switches^a list of files as options^.
16198 @option{gnatpp} implicitly uses a @emph{default dictionary file}
16199 to define the casing for the Ada predefined names and
16200 the names declared in the GNAT libraries.
16202 @item ^-D-^/SPECIFIC_CASING^
16203 @cindex @option{^-D-^/SPECIFIC_CASING^} (@command{gnatpp})
16204 Do not use the default dictionary file;
16205 instead, use the casing
16206 defined by a @option{^-n^/NAME_CASING^} switch and any explicit
16211 The structure of a dictionary file, and details on the conventions
16212 used in the default dictionary file, are defined in @ref{Name Casing}.
16214 The @option{^-D-^/SPECIFIC_CASING^} and
16215 @option{^-D@var{file}^/DICTIONARY=@var{file}^} switches are mutually
16218 @node Construct Layout Control
16219 @subsection Construct Layout Control
16220 @cindex Layout control in @command{gnatpp}
16223 This group of @command{gnatpp} switches controls the layout of comments and
16224 complex syntactic constructs. See @ref{Formatting Comments} for details
16228 @cindex @option{^-c@var{n}^/COMMENTS_LAYOUT^} (@command{gnatpp})
16229 @item ^-c0^/COMMENTS_LAYOUT=UNTOUCHED^
16230 All the comments remain unchanged
16232 @item ^-c1^/COMMENTS_LAYOUT=DEFAULT^
16233 GNAT-style comment line indentation (this is the default).
16235 @item ^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^
16236 Reference-manual comment line indentation.
16238 @item ^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^
16239 GNAT-style comment beginning
16241 @item ^-c4^/COMMENTS_LAYOUT=REFORMAT^
16242 Reformat comment blocks
16244 @item ^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^
16245 Keep unchanged special form comments
16247 Reformat comment blocks
16249 @cindex @option{^-l@var{n}^/CONSTRUCT_LAYOUT^} (@command{gnatpp})
16250 @item ^-l1^/CONSTRUCT_LAYOUT=GNAT^
16251 GNAT-style layout (this is the default)
16253 @item ^-l2^/CONSTRUCT_LAYOUT=COMPACT^
16256 @item ^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^
16259 @cindex @option{^-N^/NOTABS^} (@command{gnatpp})
16261 All the VT characters are removed from the comment text. All the HT characters
16262 are expanded with the sequences of space characters to get to the next tab
16265 @cindex @option{^--no-separate-is^/NO_SEPARATE_IS^} (@command{gnatpp})
16266 @item ^--no-separate-is^/NO_SEPARATE_IS^
16267 Do not place the keyword @code{is} on a separate line in a subprogram body in
16268 case if the spec occupies more then one line.
16270 @cindex @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} (@command{gnatpp})
16271 @item ^--separate-loop-then^/SEPARATE_LOOP_THEN^
16272 Place the keyword @code{loop} in FOR and WHILE loop statements and the
16273 keyword @code{then} in IF statements on a separate line.
16275 @cindex @option{^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^} (@command{gnatpp})
16276 @item ^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^
16277 Do not place the keyword @code{loop} in FOR and WHILE loop statements and the
16278 keyword @code{then} in IF statements on a separate line. This option is
16279 incompatible with @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} option.
16281 @cindex @option{^--use-on-new-line^/USE_ON_NEW_LINE^} (@command{gnatpp})
16282 @item ^--use-on-new-line^/USE_ON_NEW_LINE^
16283 Start each USE clause in a context clause from a separate line.
16285 @cindex @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^} (@command{gnatpp})
16286 @item ^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^
16287 Use a separate line for a loop or block statement name, but do not use an extra
16288 indentation level for the statement itself.
16294 The @option{-c1} and @option{-c2} switches are incompatible.
16295 The @option{-c3} and @option{-c4} switches are compatible with each other and
16296 also with @option{-c1} and @option{-c2}. The @option{-c0} switch disables all
16297 the other comment formatting switches.
16299 The @option{-l1}, @option{-l2}, and @option{-l3} switches are incompatible.
16304 For the @option{/COMMENTS_LAYOUT} qualifier:
16307 The @option{DEFAULT} and @option{STANDARD_INDENT} options are incompatible.
16309 The @option{GNAT_BEGINNING} and @option{REFORMAT} options are compatible with
16310 each other and also with @option{DEFAULT} and @option{STANDARD_INDENT}.
16314 The @option{GNAT}, @option{COMPACT}, and @option{UNCOMPACT} options for the
16315 @option{/CONSTRUCT_LAYOUT} qualifier are incompatible.
16318 @node General Text Layout Control
16319 @subsection General Text Layout Control
16322 These switches allow control over line length and indentation.
16325 @item ^-M@var{nnn}^/LINE_LENGTH_MAX=@var{nnn}^
16326 @cindex @option{^-M^/LINE_LENGTH^} (@command{gnatpp})
16327 Maximum line length, @var{nnn} from 32@dots{}256, the default value is 79
16329 @item ^-i@var{nnn}^/INDENTATION_LEVEL=@var{nnn}^
16330 @cindex @option{^-i^/INDENTATION_LEVEL^} (@command{gnatpp})
16331 Indentation level, @var{nnn} from 1@dots{}9, the default value is 3
16333 @item ^-cl@var{nnn}^/CONTINUATION_INDENT=@var{nnn}^
16334 @cindex @option{^-cl^/CONTINUATION_INDENT^} (@command{gnatpp})
16335 Indentation level for continuation lines (relative to the line being
16336 continued), @var{nnn} from 1@dots{}9.
16338 value is one less then the (normal) indentation level, unless the
16339 indentation is set to 1 (in which case the default value for continuation
16340 line indentation is also 1)
16343 @node Other Formatting Options
16344 @subsection Other Formatting Options
16347 These switches control the inclusion of missing end/exit labels, and
16348 the indentation level in @b{case} statements.
16351 @item ^-e^/NO_MISSED_LABELS^
16352 @cindex @option{^-e^/NO_MISSED_LABELS^} (@command{gnatpp})
16353 Do not insert missing end/exit labels. An end label is the name of
16354 a construct that may optionally be repeated at the end of the
16355 construct's declaration;
16356 e.g., the names of packages, subprograms, and tasks.
16357 An exit label is the name of a loop that may appear as target
16358 of an exit statement within the loop.
16359 By default, @command{gnatpp} inserts these end/exit labels when
16360 they are absent from the original source. This option suppresses such
16361 insertion, so that the formatted source reflects the original.
16363 @item ^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^
16364 @cindex @option{^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^} (@command{gnatpp})
16365 Insert a Form Feed character after a pragma Page.
16367 @item ^-T@var{nnn}^/MAX_INDENT=@var{nnn}^
16368 @cindex @option{^-T^/MAX_INDENT^} (@command{gnatpp})
16369 Do not use an additional indentation level for @b{case} alternatives
16370 and variants if there are @var{nnn} or more (the default
16372 If @var{nnn} is 0, an additional indentation level is
16373 used for @b{case} alternatives and variants regardless of their number.
16376 @node Setting the Source Search Path
16377 @subsection Setting the Source Search Path
16380 To define the search path for the input source file, @command{gnatpp}
16381 uses the same switches as the GNAT compiler, with the same effects.
16384 @item ^-I^/SEARCH=^@var{dir}
16385 @cindex @option{^-I^/SEARCH^} (@code{gnatpp})
16386 The same as the corresponding gcc switch
16388 @item ^-I-^/NOCURRENT_DIRECTORY^
16389 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatpp})
16390 The same as the corresponding gcc switch
16392 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE^=@var{path}
16393 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@code{gnatpp})
16394 The same as the corresponding gcc switch
16396 @item ^--RTS^/RUNTIME_SYSTEM^=@var{path}
16397 @cindex @option{^--RTS^/RUNTIME_SYSTEM^} (@code{gnatpp})
16398 The same as the corresponding gcc switch
16402 @node Output File Control
16403 @subsection Output File Control
16406 By default the output is sent to the file whose name is obtained by appending
16407 the ^@file{.pp}^@file{$PP}^ suffix to the name of the input file
16408 (if the file with this name already exists, it is unconditionally overwritten).
16409 Thus if the input file is @file{^my_ada_proc.adb^MY_ADA_PROC.ADB^} then
16410 @command{gnatpp} will produce @file{^my_ada_proc.adb.pp^MY_ADA_PROC.ADB$PP^}
16412 The output may be redirected by the following switches:
16415 @item ^-pipe^/STANDARD_OUTPUT^
16416 @cindex @option{^-pipe^/STANDARD_OUTPUT^} (@code{gnatpp})
16417 Send the output to @code{Standard_Output}
16419 @item ^-o @var{output_file}^/OUTPUT=@var{output_file}^
16420 @cindex @option{^-o^/OUTPUT^} (@code{gnatpp})
16421 Write the output into @var{output_file}.
16422 If @var{output_file} already exists, @command{gnatpp} terminates without
16423 reading or processing the input file.
16425 @item ^-of ^/FORCED_OUTPUT=^@var{output_file}
16426 @cindex @option{^-of^/FORCED_OUTPUT^} (@code{gnatpp})
16427 Write the output into @var{output_file}, overwriting the existing file
16428 (if one is present).
16430 @item ^-r^/REPLACE^
16431 @cindex @option{^-r^/REPLACE^} (@code{gnatpp})
16432 Replace the input source file with the reformatted output, and copy the
16433 original input source into the file whose name is obtained by appending the
16434 ^@file{.npp}^@file{$NPP}^ suffix to the name of the input file.
16435 If a file with this name already exists, @command{gnatpp} terminates without
16436 reading or processing the input file.
16438 @item ^-rf^/OVERRIDING_REPLACE^
16439 @cindex @option{^-rf^/OVERRIDING_REPLACE^} (@code{gnatpp})
16440 Like @option{^-r^/REPLACE^} except that if the file with the specified name
16441 already exists, it is overwritten.
16443 @item ^-rnb^/REPLACE_NO_BACKUP^
16444 @cindex @option{^-rnb^/REPLACE_NO_BACKUP^} (@code{gnatpp})
16445 Replace the input source file with the reformatted output without
16446 creating any backup copy of the input source.
16448 @item ^--eol=@var{xxx}^/END_OF_LINE=@var{xxx}^
16449 @cindex @option{^--eol^/END_OF_LINE^} (@code{gnatpp})
16450 Specifies the format of the reformatted output file. The @var{xxx}
16451 ^string specified with the switch^option^ may be either
16453 @item ``@option{^dos^DOS^}'' MS DOS style, lines end with CR LF characters
16454 @item ``@option{^crlf^CRLF^}''
16455 the same as @option{^crlf^CRLF^}
16456 @item ``@option{^unix^UNIX^}'' UNIX style, lines end with LF character
16457 @item ``@option{^lf^LF^}''
16458 the same as @option{^unix^UNIX^}
16461 @item ^-W^/RESULT_ENCODING=^@var{e}
16462 @cindex @option{^-W^/RESULT_ENCODING=^} (@command{gnatpp})
16463 Specify the wide character encoding method used to write the code in the
16465 @var{e} is one of the following:
16473 Upper half encoding
16475 @item ^s^SHIFT_JIS^
16485 Brackets encoding (default value)
16491 Options @option{^-pipe^/STANDARD_OUTPUT^},
16492 @option{^-o^/OUTPUT^} and
16493 @option{^-of^/FORCED_OUTPUT^} are allowed only if the call to gnatpp
16494 contains only one file to reformat.
16496 @option{^--eol^/END_OF_LINE^}
16498 @option{^-W^/RESULT_ENCODING^}
16499 cannot be used together
16500 with @option{^-pipe^/STANDARD_OUTPUT^} option.
16502 @node Other gnatpp Switches
16503 @subsection Other @code{gnatpp} Switches
16506 The additional @command{gnatpp} switches are defined in this subsection.
16509 @item ^-files @var{filename}^/FILES=@var{output_file}^
16510 @cindex @option{^-files^/FILES^} (@code{gnatpp})
16511 Take the argument source files from the specified file. This file should be an
16512 ordinary textual file containing file names separated by spaces or
16513 line breaks. You can use this switch more then once in the same call to
16514 @command{gnatpp}. You also can combine this switch with explicit list of
16517 @item ^-v^/VERBOSE^
16518 @cindex @option{^-v^/VERBOSE^} (@code{gnatpp})
16520 @command{gnatpp} generates version information and then
16521 a trace of the actions it takes to produce or obtain the ASIS tree.
16523 @item ^-w^/WARNINGS^
16524 @cindex @option{^-w^/WARNINGS^} (@code{gnatpp})
16526 @command{gnatpp} generates a warning whenever it cannot provide
16527 a required layout in the result source.
16530 @node Formatting Rules
16531 @section Formatting Rules
16534 The following subsections show how @command{gnatpp} treats ``white space'',
16535 comments, program layout, and name casing.
16536 They provide the detailed descriptions of the switches shown above.
16539 * White Space and Empty Lines::
16540 * Formatting Comments::
16541 * Construct Layout::
16545 @node White Space and Empty Lines
16546 @subsection White Space and Empty Lines
16549 @command{gnatpp} does not have an option to control space characters.
16550 It will add or remove spaces according to the style illustrated by the
16551 examples in the @cite{Ada Reference Manual}.
16553 The only format effectors
16554 (see @cite{Ada Reference Manual}, paragraph 2.1(13))
16555 that will appear in the output file are platform-specific line breaks,
16556 and also format effectors within (but not at the end of) comments.
16557 In particular, each horizontal tab character that is not inside
16558 a comment will be treated as a space and thus will appear in the
16559 output file as zero or more spaces depending on
16560 the reformatting of the line in which it appears.
16561 The only exception is a Form Feed character, which is inserted after a
16562 pragma @code{Page} when @option{-ff} is set.
16564 The output file will contain no lines with trailing ``white space'' (spaces,
16567 Empty lines in the original source are preserved
16568 only if they separate declarations or statements.
16569 In such contexts, a
16570 sequence of two or more empty lines is replaced by exactly one empty line.
16571 Note that a blank line will be removed if it separates two ``comment blocks''
16572 (a comment block is a sequence of whole-line comments).
16573 In order to preserve a visual separation between comment blocks, use an
16574 ``empty comment'' (a line comprising only hyphens) rather than an empty line.
16575 Likewise, if for some reason you wish to have a sequence of empty lines,
16576 use a sequence of empty comments instead.
16578 @node Formatting Comments
16579 @subsection Formatting Comments
16582 Comments in Ada code are of two kinds:
16585 a @emph{whole-line comment}, which appears by itself (possibly preceded by
16586 ``white space'') on a line
16589 an @emph{end-of-line comment}, which follows some other Ada lexical element
16594 The indentation of a whole-line comment is that of either
16595 the preceding or following line in
16596 the formatted source, depending on switch settings as will be described below.
16598 For an end-of-line comment, @command{gnatpp} leaves the same number of spaces
16599 between the end of the preceding Ada lexical element and the beginning
16600 of the comment as appear in the original source,
16601 unless either the comment has to be split to
16602 satisfy the line length limitation, or else the next line contains a
16603 whole line comment that is considered a continuation of this end-of-line
16604 comment (because it starts at the same position).
16606 cases, the start of the end-of-line comment is moved right to the nearest
16607 multiple of the indentation level.
16608 This may result in a ``line overflow'' (the right-shifted comment extending
16609 beyond the maximum line length), in which case the comment is split as
16612 There is a difference between @option{^-c1^/COMMENTS_LAYOUT=DEFAULT^}
16613 (GNAT-style comment line indentation)
16614 and @option{^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^}
16615 (reference-manual comment line indentation).
16616 With reference-manual style, a whole-line comment is indented as if it
16617 were a declaration or statement at the same place
16618 (i.e., according to the indentation of the preceding line(s)).
16619 With GNAT style, a whole-line comment that is immediately followed by an
16620 @b{if} or @b{case} statement alternative, a record variant, or the reserved
16621 word @b{begin}, is indented based on the construct that follows it.
16624 @smallexample @c ada
16636 Reference-manual indentation produces:
16638 @smallexample @c ada
16650 while GNAT-style indentation produces:
16652 @smallexample @c ada
16664 The @option{^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^} switch
16665 (GNAT style comment beginning) has the following
16670 For each whole-line comment that does not end with two hyphens,
16671 @command{gnatpp} inserts spaces if necessary after the starting two hyphens
16672 to ensure that there are at least two spaces between these hyphens and the
16673 first non-blank character of the comment.
16677 For an end-of-line comment, if in the original source the next line is a
16678 whole-line comment that starts at the same position
16679 as the end-of-line comment,
16680 then the whole-line comment (and all whole-line comments
16681 that follow it and that start at the same position)
16682 will start at this position in the output file.
16685 That is, if in the original source we have:
16687 @smallexample @c ada
16690 A := B + C; -- B must be in the range Low1..High1
16691 -- C must be in the range Low2..High2
16692 --B+C will be in the range Low1+Low2..High1+High2
16698 Then in the formatted source we get
16700 @smallexample @c ada
16703 A := B + C; -- B must be in the range Low1..High1
16704 -- C must be in the range Low2..High2
16705 -- B+C will be in the range Low1+Low2..High1+High2
16711 A comment that exceeds the line length limit will be split.
16713 @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} (reformat comment blocks) is set and
16714 the line belongs to a reformattable block, splitting the line generates a
16715 @command{gnatpp} warning.
16716 The @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} switch specifies that whole-line
16717 comments may be reformatted in typical
16718 word processor style (that is, moving words between lines and putting as
16719 many words in a line as possible).
16722 The @option{^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^} switch specifies, that comments
16723 that has a special format (that is, a character that is neither a letter nor digit
16724 not white space nor line break immediately following the leading @code{--} of
16725 the comment) should be without any change moved from the argument source
16726 into reformatted source. This switch allows to preserve comments that are used
16727 as a special marks in the code (e.g.@: SPARK annotation).
16729 @node Construct Layout
16730 @subsection Construct Layout
16733 In several cases the suggested layout in the Ada Reference Manual includes
16734 an extra level of indentation that many programmers prefer to avoid. The
16735 affected cases include:
16739 @item Record type declaration (RM 3.8)
16741 @item Record representation clause (RM 13.5.1)
16743 @item Loop statement in case if a loop has a statement identifier (RM 5.6)
16745 @item Block statement in case if a block has a statement identifier (RM 5.6)
16749 In compact mode (when GNAT style layout or compact layout is set),
16750 the pretty printer uses one level of indentation instead
16751 of two. This is achieved in the record definition and record representation
16752 clause cases by putting the @code{record} keyword on the same line as the
16753 start of the declaration or representation clause, and in the block and loop
16754 case by putting the block or loop header on the same line as the statement
16758 The difference between GNAT style @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^}
16759 and compact @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^}
16760 layout on the one hand, and uncompact layout
16761 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} on the other hand,
16762 can be illustrated by the following examples:
16766 @multitable @columnfractions .5 .5
16767 @item @i{GNAT style, compact layout} @tab @i{Uncompact layout}
16770 @smallexample @c ada
16777 @smallexample @c ada
16786 @smallexample @c ada
16788 a at 0 range 0 .. 31;
16789 b at 4 range 0 .. 31;
16793 @smallexample @c ada
16796 a at 0 range 0 .. 31;
16797 b at 4 range 0 .. 31;
16802 @smallexample @c ada
16810 @smallexample @c ada
16820 @smallexample @c ada
16821 Clear : for J in 1 .. 10 loop
16826 @smallexample @c ada
16828 for J in 1 .. 10 loop
16839 GNAT style, compact layout Uncompact layout
16841 type q is record type q is
16842 a : integer; record
16843 b : integer; a : integer;
16844 end record; b : integer;
16847 for q use record for q use
16848 a at 0 range 0 .. 31; record
16849 b at 4 range 0 .. 31; a at 0 range 0 .. 31;
16850 end record; b at 4 range 0 .. 31;
16853 Block : declare Block :
16854 A : Integer := 3; declare
16855 begin A : Integer := 3;
16857 end Block; Proc (A, A);
16860 Clear : for J in 1 .. 10 loop Clear :
16861 A (J) := 0; for J in 1 .. 10 loop
16862 end loop Clear; A (J) := 0;
16869 A further difference between GNAT style layout and compact layout is that
16870 GNAT style layout inserts empty lines as separation for
16871 compound statements, return statements and bodies.
16873 Note that the layout specified by
16874 @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^}
16875 for named block and loop statements overrides the layout defined by these
16876 constructs by @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^},
16877 @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^} or
16878 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} option.
16881 @subsection Name Casing
16884 @command{gnatpp} always converts the usage occurrence of a (simple) name to
16885 the same casing as the corresponding defining identifier.
16887 You control the casing for defining occurrences via the
16888 @option{^-n^/NAME_CASING^} switch.
16890 With @option{-nD} (``as declared'', which is the default),
16893 With @option{/NAME_CASING=AS_DECLARED}, which is the default,
16895 defining occurrences appear exactly as in the source file
16896 where they are declared.
16897 The other ^values for this switch^options for this qualifier^ ---
16898 @option{^-nU^UPPER_CASE^},
16899 @option{^-nL^LOWER_CASE^},
16900 @option{^-nM^MIXED_CASE^} ---
16902 ^upper, lower, or mixed case, respectively^the corresponding casing^.
16903 If @command{gnatpp} changes the casing of a defining
16904 occurrence, it analogously changes the casing of all the
16905 usage occurrences of this name.
16907 If the defining occurrence of a name is not in the source compilation unit
16908 currently being processed by @command{gnatpp}, the casing of each reference to
16909 this name is changed according to the value of the @option{^-n^/NAME_CASING^}
16910 switch (subject to the dictionary file mechanism described below).
16911 Thus @command{gnatpp} acts as though the @option{^-n^/NAME_CASING^} switch
16913 casing for the defining occurrence of the name.
16915 Some names may need to be spelled with casing conventions that are not
16916 covered by the upper-, lower-, and mixed-case transformations.
16917 You can arrange correct casing by placing such names in a
16918 @emph{dictionary file},
16919 and then supplying a @option{^-D^/DICTIONARY^} switch.
16920 The casing of names from dictionary files overrides
16921 any @option{^-n^/NAME_CASING^} switch.
16923 To handle the casing of Ada predefined names and the names from GNAT libraries,
16924 @command{gnatpp} assumes a default dictionary file.
16925 The name of each predefined entity is spelled with the same casing as is used
16926 for the entity in the @cite{Ada Reference Manual}.
16927 The name of each entity in the GNAT libraries is spelled with the same casing
16928 as is used in the declaration of that entity.
16930 The @w{@option{^-D-^/SPECIFIC_CASING^}} switch suppresses the use of the
16931 default dictionary file.
16932 Instead, the casing for predefined and GNAT-defined names will be established
16933 by the @option{^-n^/NAME_CASING^} switch or explicit dictionary files.
16934 For example, by default the names @code{Ada.Text_IO} and @code{GNAT.OS_Lib}
16935 will appear as just shown,
16936 even in the presence of a @option{^-nU^/NAME_CASING=UPPER_CASE^} switch.
16937 To ensure that even such names are rendered in uppercase,
16938 additionally supply the @w{@option{^-D-^/SPECIFIC_CASING^}} switch
16939 (or else, less conveniently, place these names in upper case in a dictionary
16942 A dictionary file is
16943 a plain text file; each line in this file can be either a blank line
16944 (containing only space characters and ASCII.HT characters), an Ada comment
16945 line, or the specification of exactly one @emph{casing schema}.
16947 A casing schema is a string that has the following syntax:
16951 @var{casing_schema} ::= @var{identifier} | *@var{simple_identifier}*
16953 @var{simple_identifier} ::= @var{letter}@{@var{letter_or_digit}@}
16958 (See @cite{Ada Reference Manual}, Section 2.3) for the definition of the
16959 @var{identifier} lexical element and the @var{letter_or_digit} category.)
16961 The casing schema string can be followed by white space and/or an Ada-style
16962 comment; any amount of white space is allowed before the string.
16964 If a dictionary file is passed as
16966 the value of a @option{-D@var{file}} switch
16969 an option to the @option{/DICTIONARY} qualifier
16972 simple name and every identifier, @command{gnatpp} checks if the dictionary
16973 defines the casing for the name or for some of its parts (the term ``subword''
16974 is used below to denote the part of a name which is delimited by ``_'' or by
16975 the beginning or end of the word and which does not contain any ``_'' inside):
16979 if the whole name is in the dictionary, @command{gnatpp} uses for this name
16980 the casing defined by the dictionary; no subwords are checked for this word
16983 for every subword @command{gnatpp} checks if the dictionary contains the
16984 corresponding string of the form @code{*@var{simple_identifier}*},
16985 and if it does, the casing of this @var{simple_identifier} is used
16989 if the whole name does not contain any ``_'' inside, and if for this name
16990 the dictionary contains two entries - one of the form @var{identifier},
16991 and another - of the form *@var{simple_identifier}*, then the first one
16992 is applied to define the casing of this name
16995 if more than one dictionary file is passed as @command{gnatpp} switches, each
16996 dictionary adds new casing exceptions and overrides all the existing casing
16997 exceptions set by the previous dictionaries
17000 when @command{gnatpp} checks if the word or subword is in the dictionary,
17001 this check is not case sensitive
17005 For example, suppose we have the following source to reformat:
17007 @smallexample @c ada
17010 name1 : integer := 1;
17011 name4_name3_name2 : integer := 2;
17012 name2_name3_name4 : Boolean;
17015 name2_name3_name4 := name4_name3_name2 > name1;
17021 And suppose we have two dictionaries:
17038 If @command{gnatpp} is called with the following switches:
17042 @command{gnatpp -nM -D dict1 -D dict2 test.adb}
17045 @command{gnatpp test.adb /NAME_CASING=MIXED_CASE /DICTIONARY=(dict1, dict2)}
17050 then we will get the following name casing in the @command{gnatpp} output:
17052 @smallexample @c ada
17055 NAME1 : Integer := 1;
17056 Name4_NAME3_Name2 : Integer := 2;
17057 Name2_NAME3_Name4 : Boolean;
17060 Name2_NAME3_Name4 := Name4_NAME3_Name2 > NAME1;
17065 @c *********************************
17066 @node The GNAT Metric Tool gnatmetric
17067 @chapter The GNAT Metric Tool @command{gnatmetric}
17069 @cindex Metric tool
17072 ^The @command{gnatmetric} tool^@command{GNAT METRIC}^ is an ASIS-based utility
17073 for computing various program metrics.
17074 It takes an Ada source file as input and generates a file containing the
17075 metrics data as output. Various switches control which
17076 metrics are computed and output.
17078 @command{gnatmetric} generates and uses the ASIS
17079 tree for the input source and thus requires the input to be syntactically and
17080 semantically legal.
17081 If this condition is not met, @command{gnatmetric} will generate
17082 an error message; no metric information for this file will be
17083 computed and reported.
17085 If the compilation unit contained in the input source depends semantically
17086 upon units in files located outside the current directory, you have to provide
17087 the source search path when invoking @command{gnatmetric}.
17088 If it depends semantically upon units that are contained
17089 in files with names that do not follow the GNAT file naming rules, you have to
17090 provide the configuration file describing the corresponding naming scheme (see
17091 the description of the @command{gnatmetric} switches below.)
17092 Alternatively, you may use a project file and invoke @command{gnatmetric}
17093 through the @command{gnat} driver.
17095 The @command{gnatmetric} command has the form
17098 $ gnatmetric @ovar{switches} @{@var{filename}@} @r{[}-cargs @var{gcc_switches}@r{]}
17105 @var{switches} specify the metrics to compute and define the destination for
17109 Each @var{filename} is the name (including the extension) of a source
17110 file to process. ``Wildcards'' are allowed, and
17111 the file name may contain path information.
17112 If no @var{filename} is supplied, then the @var{switches} list must contain
17114 @option{-files} switch (@pxref{Other gnatmetric Switches}).
17115 Including both a @option{-files} switch and one or more
17116 @var{filename} arguments is permitted.
17119 @samp{-cargs @var{gcc_switches}} is a list of switches for
17120 @command{gcc}. They will be passed on to all compiler invocations made by
17121 @command{gnatmetric} to generate the ASIS trees. Here you can provide
17122 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
17123 and use the @option{-gnatec} switch to set the configuration file.
17127 * Switches for gnatmetric::
17130 @node Switches for gnatmetric
17131 @section Switches for @command{gnatmetric}
17134 The following subsections describe the various switches accepted by
17135 @command{gnatmetric}, organized by category.
17138 * Output Files Control::
17139 * Disable Metrics For Local Units::
17140 * Specifying a set of metrics to compute::
17141 * Other gnatmetric Switches::
17142 * Generate project-wide metrics::
17145 @node Output Files Control
17146 @subsection Output File Control
17147 @cindex Output file control in @command{gnatmetric}
17150 @command{gnatmetric} has two output formats. It can generate a
17151 textual (human-readable) form, and also XML. By default only textual
17152 output is generated.
17154 When generating the output in textual form, @command{gnatmetric} creates
17155 for each Ada source file a corresponding text file
17156 containing the computed metrics, except for the case when the set of metrics
17157 specified by gnatmetric parameters consists only of metrics that are computed
17158 for the whole set of analyzed sources, but not for each Ada source.
17159 By default, this file is placed in the same directory as where the source
17160 file is located, and its name is obtained
17161 by appending the ^@file{.metrix}^@file{$METRIX}^ suffix to the name of the
17164 All the output information generated in XML format is placed in a single
17165 file. By default this file is placed in the current directory and has the
17166 name ^@file{metrix.xml}^@file{METRIX$XML}^.
17168 Some of the computed metrics are summed over the units passed to
17169 @command{gnatmetric}; for example, the total number of lines of code.
17170 By default this information is sent to @file{stdout}, but a file
17171 can be specified with the @option{-og} switch.
17173 The following switches control the @command{gnatmetric} output:
17176 @cindex @option{^-x^/XML^} (@command{gnatmetric})
17178 Generate the XML output
17180 @cindex @option{^-xs^/XSD^} (@command{gnatmetric})
17182 Generate the XML output and the XML schema file that describes the structure
17183 of the XML metric report, this schema is assigned to the XML file. The schema
17184 file has the same name as the XML output file with @file{.xml} suffix replaced
17187 @cindex @option{^-nt^/NO_TEXT^} (@command{gnatmetric})
17188 @item ^-nt^/NO_TEXT^
17189 Do not generate the output in text form (implies @option{^-x^/XML^})
17191 @cindex @option{^-d^/DIRECTORY^} (@command{gnatmetric})
17192 @item ^-d @var{output_dir}^/DIRECTORY=@var{output_dir}^
17193 Put textual files with detailed metrics into @var{output_dir}
17195 @cindex @option{^-o^/SUFFIX_DETAILS^} (@command{gnatmetric})
17196 @item ^-o @var{file_suffix}^/SUFFIX_DETAILS=@var{file_suffix}^
17197 Use @var{file_suffix}, instead of ^@file{.metrix}^@file{$METRIX}^
17198 in the name of the output file.
17200 @cindex @option{^-og^/GLOBAL_OUTPUT^} (@command{gnatmetric})
17201 @item ^-og @var{file_name}^/GLOBAL_OUTPUT=@var{file_name}^
17202 Put global metrics into @var{file_name}
17204 @cindex @option{^-ox^/XML_OUTPUT^} (@command{gnatmetric})
17205 @item ^-ox @var{file_name}^/XML_OUTPUT=@var{file_name}^
17206 Put the XML output into @var{file_name} (also implies @option{^-x^/XML^})
17208 @cindex @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} (@command{gnatmetric})
17209 @item ^-sfn^/SHORT_SOURCE_FILE_NAME^
17210 Use ``short'' source file names in the output. (The @command{gnatmetric}
17211 output includes the name(s) of the Ada source file(s) from which the metrics
17212 are computed. By default each name includes the absolute path. The
17213 @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} switch causes @command{gnatmetric}
17214 to exclude all directory information from the file names that are output.)
17218 @node Disable Metrics For Local Units
17219 @subsection Disable Metrics For Local Units
17220 @cindex Disable Metrics For Local Units in @command{gnatmetric}
17223 @command{gnatmetric} relies on the GNAT compilation model @minus{}
17225 unit per one source file. It computes line metrics for the whole source
17226 file, and it also computes syntax
17227 and complexity metrics for the file's outermost unit.
17229 By default, @command{gnatmetric} will also compute all metrics for certain
17230 kinds of locally declared program units:
17234 subprogram (and generic subprogram) bodies;
17237 package (and generic package) specs and bodies;
17240 task object and type specifications and bodies;
17243 protected object and type specifications and bodies.
17247 These kinds of entities will be referred to as
17248 @emph{eligible local program units}, or simply @emph{eligible local units},
17249 @cindex Eligible local unit (for @command{gnatmetric})
17250 in the discussion below.
17252 Note that a subprogram declaration, generic instantiation,
17253 or renaming declaration only receives metrics
17254 computation when it appear as the outermost entity
17257 Suppression of metrics computation for eligible local units can be
17258 obtained via the following switch:
17261 @cindex @option{^-n@var{x}^/SUPPRESS^} (@command{gnatmetric})
17262 @item ^-nolocal^/SUPPRESS=LOCAL_DETAILS^
17263 Do not compute detailed metrics for eligible local program units
17267 @node Specifying a set of metrics to compute
17268 @subsection Specifying a set of metrics to compute
17271 By default all the metrics are computed and reported. The switches
17272 described in this subsection allow you to control, on an individual
17273 basis, whether metrics are computed and
17274 reported. If at least one positive metric
17275 switch is specified (that is, a switch that defines that a given
17276 metric or set of metrics is to be computed), then only
17277 explicitly specified metrics are reported.
17280 * Line Metrics Control::
17281 * Syntax Metrics Control::
17282 * Complexity Metrics Control::
17283 * Object-Oriented Metrics Control::
17286 @node Line Metrics Control
17287 @subsubsection Line Metrics Control
17288 @cindex Line metrics control in @command{gnatmetric}
17291 For any (legal) source file, and for each of its
17292 eligible local program units, @command{gnatmetric} computes the following
17297 the total number of lines;
17300 the total number of code lines (i.e., non-blank lines that are not comments)
17303 the number of comment lines
17306 the number of code lines containing end-of-line comments;
17309 the comment percentage: the ratio between the number of lines that contain
17310 comments and the number of all non-blank lines, expressed as a percentage;
17313 the number of empty lines and lines containing only space characters and/or
17314 format effectors (blank lines)
17317 the average number of code lines in subprogram bodies, task bodies, entry
17318 bodies and statement sequences in package bodies (this metric is only computed
17319 across the whole set of the analyzed units)
17324 @command{gnatmetric} sums the values of the line metrics for all the
17325 files being processed and then generates the cumulative results. The tool
17326 also computes for all the files being processed the average number of code
17329 You can use the following switches to select the specific line metrics
17330 to be computed and reported.
17333 @cindex @option{^--lines@var{x}^/LINE_COUNT_METRICS^} (@command{gnatmetric})
17336 @cindex @option{--no-lines@var{x}}
17339 @item ^--lines-all^/LINE_COUNT_METRICS=ALL^
17340 Report all the line metrics
17342 @item ^--no-lines-all^/LINE_COUNT_METRICS=NONE^
17343 Do not report any of line metrics
17345 @item ^--lines^/LINE_COUNT_METRICS=ALL_LINES^
17346 Report the number of all lines
17348 @item ^--no-lines^/LINE_COUNT_METRICS=NOALL_LINES^
17349 Do not report the number of all lines
17351 @item ^--lines-code^/LINE_COUNT_METRICS=CODE_LINES^
17352 Report the number of code lines
17354 @item ^--no-lines-code^/LINE_COUNT_METRICS=NOCODE_LINES^
17355 Do not report the number of code lines
17357 @item ^--lines-comment^/LINE_COUNT_METRICS=COMMENT_LINES^
17358 Report the number of comment lines
17360 @item ^--no-lines-comment^/LINE_COUNT_METRICS=NOCOMMENT_LINES^
17361 Do not report the number of comment lines
17363 @item ^--lines-eol-comment^/LINE_COUNT_METRICS=CODE_COMMENT_LINES^
17364 Report the number of code lines containing
17365 end-of-line comments
17367 @item ^--no-lines-eol-comment^/LINE_COUNT_METRICS=NOCODE_COMMENT_LINES^
17368 Do not report the number of code lines containing
17369 end-of-line comments
17371 @item ^--lines-ratio^/LINE_COUNT_METRICS=COMMENT_PERCENTAGE^
17372 Report the comment percentage in the program text
17374 @item ^--no-lines-ratio^/LINE_COUNT_METRICS=NOCOMMENT_PERCENTAGE^
17375 Do not report the comment percentage in the program text
17377 @item ^--lines-blank^/LINE_COUNT_METRICS=BLANK_LINES^
17378 Report the number of blank lines
17380 @item ^--no-lines-blank^/LINE_COUNT_METRICS=NOBLANK_LINES^
17381 Do not report the number of blank lines
17383 @item ^--lines-average^/LINE_COUNT_METRICS=AVERAGE_BODY_LINES^
17384 Report the average number of code lines in subprogram bodies, task bodies,
17385 entry bodies and statement sequences in package bodies. The metric is computed
17386 and reported for the whole set of processed Ada sources only.
17388 @item ^--no-lines-average^/LINE_COUNT_METRICS=NOAVERAGE_BODY_LINES^
17389 Do not report the average number of code lines in subprogram bodies,
17390 task bodies, entry bodies and statement sequences in package bodies.
17394 @node Syntax Metrics Control
17395 @subsubsection Syntax Metrics Control
17396 @cindex Syntax metrics control in @command{gnatmetric}
17399 @command{gnatmetric} computes various syntactic metrics for the
17400 outermost unit and for each eligible local unit:
17403 @item LSLOC (``Logical Source Lines Of Code'')
17404 The total number of declarations and the total number of statements
17406 @item Maximal static nesting level of inner program units
17408 @cite{Ada Reference Manual}, 10.1(1), ``A program unit is either a
17409 package, a task unit, a protected unit, a
17410 protected entry, a generic unit, or an explicitly declared subprogram other
17411 than an enumeration literal.''
17413 @item Maximal nesting level of composite syntactic constructs
17414 This corresponds to the notion of the
17415 maximum nesting level in the GNAT built-in style checks
17416 (@pxref{Style Checking})
17420 For the outermost unit in the file, @command{gnatmetric} additionally computes
17421 the following metrics:
17424 @item Public subprograms
17425 This metric is computed for package specs. It is the
17426 number of subprograms and generic subprograms declared in the visible
17427 part (including the visible part of nested packages, protected objects, and
17430 @item All subprograms
17431 This metric is computed for bodies and subunits. The
17432 metric is equal to a total number of subprogram bodies in the compilation
17434 Neither generic instantiations nor renamings-as-a-body nor body stubs
17435 are counted. Any subprogram body is counted, independently of its nesting
17436 level and enclosing constructs. Generic bodies and bodies of protected
17437 subprograms are counted in the same way as ``usual'' subprogram bodies.
17440 This metric is computed for package specs and
17441 generic package declarations. It is the total number of types
17442 that can be referenced from outside this compilation unit, plus the
17443 number of types from all the visible parts of all the visible generic
17444 packages. Generic formal types are not counted. Only types, not subtypes,
17448 Along with the total number of public types, the following
17449 types are counted and reported separately:
17456 Root tagged types (abstract, non-abstract, private, non-private). Type
17457 extensions are @emph{not} counted
17460 Private types (including private extensions)
17471 This metric is computed for any compilation unit. It is equal to the total
17472 number of the declarations of different types given in the compilation unit.
17473 The private and the corresponding full type declaration are counted as one
17474 type declaration. Incomplete type declarations and generic formal types
17476 No distinction is made among different kinds of types (abstract,
17477 private etc.); the total number of types is computed and reported.
17482 By default, all the syntax metrics are computed and reported. You can use the
17483 following switches to select specific syntax metrics.
17487 @cindex @option{^--syntax@var{x}^/SYNTAX_METRICS^} (@command{gnatmetric})
17490 @cindex @option{--no-syntax@var{x}} (@command{gnatmetric})
17493 @item ^--syntax-all^/SYNTAX_METRICS=ALL^
17494 Report all the syntax metrics
17496 @item ^--no-syntax-all^/SYNTAX_METRICS=NONE^
17497 Do not report any of syntax metrics
17499 @item ^--declarations^/SYNTAX_METRICS=DECLARATIONS^
17500 Report the total number of declarations
17502 @item ^--no-declarations^/SYNTAX_METRICS=NODECLARATIONS^
17503 Do not report the total number of declarations
17505 @item ^--statements^/SYNTAX_METRICS=STATEMENTS^
17506 Report the total number of statements
17508 @item ^--no-statements^/SYNTAX_METRICS=NOSTATEMENTS^
17509 Do not report the total number of statements
17511 @item ^--public-subprograms^/SYNTAX_METRICS=PUBLIC_SUBPROGRAMS^
17512 Report the number of public subprograms in a compilation unit
17514 @item ^--no-public-subprograms^/SYNTAX_METRICS=NOPUBLIC_SUBPROGRAMS^
17515 Do not report the number of public subprograms in a compilation unit
17517 @item ^--all-subprograms^/SYNTAX_METRICS=ALL_SUBPROGRAMS^
17518 Report the number of all the subprograms in a compilation unit
17520 @item ^--no-all-subprograms^/SYNTAX_METRICS=NOALL_SUBPROGRAMS^
17521 Do not report the number of all the subprograms in a compilation unit
17523 @item ^--public-types^/SYNTAX_METRICS=PUBLIC_TYPES^
17524 Report the number of public types in a compilation unit
17526 @item ^--no-public-types^/SYNTAX_METRICS=NOPUBLIC_TYPES^
17527 Do not report the number of public types in a compilation unit
17529 @item ^--all-types^/SYNTAX_METRICS=ALL_TYPES^
17530 Report the number of all the types in a compilation unit
17532 @item ^--no-all-types^/SYNTAX_METRICS=NOALL_TYPES^
17533 Do not report the number of all the types in a compilation unit
17535 @item ^--unit-nesting^/SYNTAX_METRICS=UNIT_NESTING^
17536 Report the maximal program unit nesting level
17538 @item ^--no-unit-nesting^/SYNTAX_METRICS=UNIT_NESTING_OFF^
17539 Do not report the maximal program unit nesting level
17541 @item ^--construct-nesting^/SYNTAX_METRICS=CONSTRUCT_NESTING^
17542 Report the maximal construct nesting level
17544 @item ^--no-construct-nesting^/SYNTAX_METRICS=NOCONSTRUCT_NESTING^
17545 Do not report the maximal construct nesting level
17549 @node Complexity Metrics Control
17550 @subsubsection Complexity Metrics Control
17551 @cindex Complexity metrics control in @command{gnatmetric}
17554 For a program unit that is an executable body (a subprogram body (including
17555 generic bodies), task body, entry body or a package body containing
17556 its own statement sequence) @command{gnatmetric} computes the following
17557 complexity metrics:
17561 McCabe cyclomatic complexity;
17564 McCabe essential complexity;
17567 maximal loop nesting level
17572 The McCabe complexity metrics are defined
17573 in @url{http://www.mccabe.com/pdf/nist235r.pdf}
17575 According to McCabe, both control statements and short-circuit control forms
17576 should be taken into account when computing cyclomatic complexity. For each
17577 body, we compute three metric values:
17581 the complexity introduced by control
17582 statements only, without taking into account short-circuit forms,
17585 the complexity introduced by short-circuit control forms only, and
17589 cyclomatic complexity, which is the sum of these two values.
17593 When computing cyclomatic and essential complexity, @command{gnatmetric} skips
17594 the code in the exception handlers and in all the nested program units.
17596 By default, all the complexity metrics are computed and reported.
17597 For more fine-grained control you can use
17598 the following switches:
17601 @cindex @option{^-complexity@var{x}^/COMPLEXITY_METRICS^} (@command{gnatmetric})
17604 @cindex @option{--no-complexity@var{x}}
17607 @item ^--complexity-all^/COMPLEXITY_METRICS=ALL^
17608 Report all the complexity metrics
17610 @item ^--no-complexity-all^/COMPLEXITY_METRICS=NONE^
17611 Do not report any of complexity metrics
17613 @item ^--complexity-cyclomatic^/COMPLEXITY_METRICS=CYCLOMATIC^
17614 Report the McCabe Cyclomatic Complexity
17616 @item ^--no-complexity-cyclomatic^/COMPLEXITY_METRICS=NOCYCLOMATIC^
17617 Do not report the McCabe Cyclomatic Complexity
17619 @item ^--complexity-essential^/COMPLEXITY_METRICS=ESSENTIAL^
17620 Report the Essential Complexity
17622 @item ^--no-complexity-essential^/COMPLEXITY_METRICS=NOESSENTIAL^
17623 Do not report the Essential Complexity
17625 @item ^--loop-nesting^/COMPLEXITY_METRICS=LOOP_NESTING_ON^
17626 Report maximal loop nesting level
17628 @item ^--no-loop-nesting^/COMPLEXITY_METRICS=NOLOOP_NESTING^
17629 Do not report maximal loop nesting level
17631 @item ^--complexity-average^/COMPLEXITY_METRICS=AVERAGE_COMPLEXITY^
17632 Report the average McCabe Cyclomatic Complexity for all the subprogram bodies,
17633 task bodies, entry bodies and statement sequences in package bodies.
17634 The metric is computed and reported for whole set of processed Ada sources
17637 @item ^--no-complexity-average^/COMPLEXITY_METRICS=NOAVERAGE_COMPLEXITY^
17638 Do not report the average McCabe Cyclomatic Complexity for all the subprogram
17639 bodies, task bodies, entry bodies and statement sequences in package bodies
17641 @cindex @option{^-ne^/NO_EXITS_AS_GOTOS^} (@command{gnatmetric})
17642 @item ^-ne^/NO_EXITS_AS_GOTOS^
17643 Do not consider @code{exit} statements as @code{goto}s when
17644 computing Essential Complexity
17646 @item ^--extra-exit-points^/EXTRA_EXIT_POINTS^
17647 Report the extra exit points for subprogram bodies. As an exit point, this
17648 metric counts @code{return} statements and raise statements in case when the
17649 raised exception is not handled in the same body. In case of a function this
17650 metric subtracts 1 from the number of exit points, because a function body
17651 must contain at least one @code{return} statement.
17653 @item ^--no-extra-exit-points^/NOEXTRA_EXIT_POINTS^
17654 Do not report the extra exit points for subprogram bodies
17658 @node Object-Oriented Metrics Control
17659 @subsubsection Object-Oriented Metrics Control
17660 @cindex Object-Oriented metrics control in @command{gnatmetric}
17663 @cindex Coupling metrics (in in @command{gnatmetric})
17664 Coupling metrics are object-oriented metrics that measure the
17665 dependencies between a given class (or a group of classes) and the
17666 ``external world'' (that is, the other classes in the program). In this
17667 subsection the term ``class'' is used in its
17668 traditional object-oriented programming sense
17669 (an instantiable module that contains data and/or method members).
17670 A @emph{category} (of classes)
17671 is a group of closely related classes that are reused and/or
17674 A class @code{K}'s @emph{efferent coupling} is the number of classes
17675 that @code{K} depends upon.
17676 A category's efferent coupling is the number of classes outside the
17677 category that the classes inside the category depend upon.
17679 A class @code{K}'s @emph{afferent coupling} is the number of classes
17680 that depend upon @code{K}.
17681 A category's afferent coupling is the number of classes outside the
17682 category that depend on classes belonging to the category.
17684 Ada's implementation of the object-oriented paradigm does not use the
17685 traditional class notion, so the definition of the coupling
17686 metrics for Ada maps the class and class category notions
17687 onto Ada constructs.
17689 For the coupling metrics, several kinds of modules -- a library package,
17690 a library generic package, and a library generic package instantiation --
17691 that define a tagged type or an interface type are
17692 considered to be a class. A category consists of a library package (or
17693 a library generic package) that defines a tagged or an interface type,
17694 together with all its descendant (generic) packages that define tagged
17695 or interface types. For any package counted as a class,
17696 its body and subunits (if any) are considered
17697 together with its spec when counting the dependencies, and coupling
17698 metrics are reported for spec units only. For dependencies
17699 between classes, the Ada semantic dependencies are considered.
17700 For coupling metrics, only dependencies on units that are considered as
17701 classes, are considered.
17703 When computing coupling metrics, @command{gnatmetric} counts only
17704 dependencies between units that are arguments of the gnatmetric call.
17705 Coupling metrics are program-wide (or project-wide) metrics, so to
17706 get a valid result, you should call @command{gnatmetric} for
17707 the whole set of sources that make up your program. It can be done
17708 by calling @command{gnatmetric} from the GNAT driver with @option{-U}
17709 option (see See @ref{The GNAT Driver and Project Files} for details.
17711 By default, all the coupling metrics are disabled. You can use the following
17712 switches to specify the coupling metrics to be computed and reported:
17717 @cindex @option{--package@var{x}} (@command{gnatmetric})
17718 @cindex @option{--no-package@var{x}} (@command{gnatmetric})
17719 @cindex @option{--category@var{x}} (@command{gnatmetric})
17720 @cindex @option{--no-category@var{x}} (@command{gnatmetric})
17724 @cindex @option{/COUPLING_METRICS} (@command{gnatmetric})
17727 @item ^--coupling-all^/COUPLING_METRICS=ALL^
17728 Report all the coupling metrics
17730 @item ^--no-coupling-all^/COUPLING_METRICS=NONE^
17731 Do not report any of metrics
17733 @item ^--package-efferent-coupling^/COUPLING_METRICS=PACKAGE_EFFERENT^
17734 Report package efferent coupling
17736 @item ^--no-package-efferent-coupling^/COUPLING_METRICS=NOPACKAGE_EFFERENT^
17737 Do not report package efferent coupling
17739 @item ^--package-afferent-coupling^/COUPLING_METRICS=PACKAGE_AFFERENT^
17740 Report package afferent coupling
17742 @item ^--no-package-afferent-coupling^/COUPLING_METRICS=NOPACKAGE_AFFERENT^
17743 Do not report package afferent coupling
17745 @item ^--category-efferent-coupling^/COUPLING_METRICS=CATEGORY_EFFERENT^
17746 Report category efferent coupling
17748 @item ^--no-category-efferent-coupling^/COUPLING_METRICS=NOCATEGORY_EFFERENT^
17749 Do not report category efferent coupling
17751 @item ^--category-afferent-coupling^/COUPLING_METRICS=CATEGORY_AFFERENT^
17752 Report category afferent coupling
17754 @item ^--no-category-afferent-coupling^/COUPLING_METRICS=NOCATEGORY_AFFERENT^
17755 Do not report category afferent coupling
17759 @node Other gnatmetric Switches
17760 @subsection Other @code{gnatmetric} Switches
17763 Additional @command{gnatmetric} switches are as follows:
17766 @item ^-files @var{filename}^/FILES=@var{filename}^
17767 @cindex @option{^-files^/FILES^} (@code{gnatmetric})
17768 Take the argument source files from the specified file. This file should be an
17769 ordinary text file containing file names separated by spaces or
17770 line breaks. You can use this switch more then once in the same call to
17771 @command{gnatmetric}. You also can combine this switch with
17772 an explicit list of files.
17774 @item ^-v^/VERBOSE^
17775 @cindex @option{^-v^/VERBOSE^} (@code{gnatmetric})
17777 @command{gnatmetric} generates version information and then
17778 a trace of sources being processed.
17780 @item ^-dv^/DEBUG_OUTPUT^
17781 @cindex @option{^-dv^/DEBUG_OUTPUT^} (@code{gnatmetric})
17783 @command{gnatmetric} generates various messages useful to understand what
17784 happens during the metrics computation
17787 @cindex @option{^-q^/QUIET^} (@code{gnatmetric})
17791 @node Generate project-wide metrics
17792 @subsection Generate project-wide metrics
17794 In order to compute metrics on all units of a given project, you can use
17795 the @command{gnat} driver along with the @option{-P} option:
17801 If the project @code{proj} depends upon other projects, you can compute
17802 the metrics on the project closure using the @option{-U} option:
17804 gnat metric -Pproj -U
17808 Finally, if not all the units are relevant to a particular main
17809 program in the project closure, you can generate metrics for the set
17810 of units needed to create a given main program (unit closure) using
17811 the @option{-U} option followed by the name of the main unit:
17813 gnat metric -Pproj -U main
17817 @c ***********************************
17818 @node File Name Krunching Using gnatkr
17819 @chapter File Name Krunching Using @code{gnatkr}
17823 This chapter discusses the method used by the compiler to shorten
17824 the default file names chosen for Ada units so that they do not
17825 exceed the maximum length permitted. It also describes the
17826 @code{gnatkr} utility that can be used to determine the result of
17827 applying this shortening.
17831 * Krunching Method::
17832 * Examples of gnatkr Usage::
17836 @section About @code{gnatkr}
17839 The default file naming rule in GNAT
17840 is that the file name must be derived from
17841 the unit name. The exact default rule is as follows:
17844 Take the unit name and replace all dots by hyphens.
17846 If such a replacement occurs in the
17847 second character position of a name, and the first character is
17848 ^@samp{a}, @samp{g}, @samp{s}, or @samp{i}, ^@samp{A}, @samp{G}, @samp{S}, or @samp{I},^
17849 then replace the dot by the character
17850 ^@samp{~} (tilde)^@samp{$} (dollar sign)^
17851 instead of a minus.
17853 The reason for this exception is to avoid clashes
17854 with the standard names for children of System, Ada, Interfaces,
17855 and GNAT, which use the prefixes
17856 ^@samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},^@samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},^
17859 The @option{^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{nn}}
17860 switch of the compiler activates a ``krunching''
17861 circuit that limits file names to nn characters (where nn is a decimal
17862 integer). For example, using OpenVMS,
17863 where the maximum file name length is
17864 39, the value of nn is usually set to 39, but if you want to generate
17865 a set of files that would be usable if ported to a system with some
17866 different maximum file length, then a different value can be specified.
17867 The default value of 39 for OpenVMS need not be specified.
17869 The @code{gnatkr} utility can be used to determine the krunched name for
17870 a given file, when krunched to a specified maximum length.
17873 @section Using @code{gnatkr}
17876 The @code{gnatkr} command has the form
17880 $ gnatkr @var{name} @ovar{length}
17886 $ gnatkr @var{name} /COUNT=nn
17891 @var{name} is the uncrunched file name, derived from the name of the unit
17892 in the standard manner described in the previous section (i.e., in particular
17893 all dots are replaced by hyphens). The file name may or may not have an
17894 extension (defined as a suffix of the form period followed by arbitrary
17895 characters other than period). If an extension is present then it will
17896 be preserved in the output. For example, when krunching @file{hellofile.ads}
17897 to eight characters, the result will be hellofil.ads.
17899 Note: for compatibility with previous versions of @code{gnatkr} dots may
17900 appear in the name instead of hyphens, but the last dot will always be
17901 taken as the start of an extension. So if @code{gnatkr} is given an argument
17902 such as @file{Hello.World.adb} it will be treated exactly as if the first
17903 period had been a hyphen, and for example krunching to eight characters
17904 gives the result @file{hellworl.adb}.
17906 Note that the result is always all lower case (except on OpenVMS where it is
17907 all upper case). Characters of the other case are folded as required.
17909 @var{length} represents the length of the krunched name. The default
17910 when no argument is given is ^8^39^ characters. A length of zero stands for
17911 unlimited, in other words do not chop except for system files where the
17912 implied crunching length is always eight characters.
17915 The output is the krunched name. The output has an extension only if the
17916 original argument was a file name with an extension.
17918 @node Krunching Method
17919 @section Krunching Method
17922 The initial file name is determined by the name of the unit that the file
17923 contains. The name is formed by taking the full expanded name of the
17924 unit and replacing the separating dots with hyphens and
17925 using ^lowercase^uppercase^
17926 for all letters, except that a hyphen in the second character position is
17927 replaced by a ^tilde^dollar sign^ if the first character is
17928 ^@samp{a}, @samp{i}, @samp{g}, or @samp{s}^@samp{A}, @samp{I}, @samp{G}, or @samp{S}^.
17929 The extension is @code{.ads} for a
17930 spec and @code{.adb} for a body.
17931 Krunching does not affect the extension, but the file name is shortened to
17932 the specified length by following these rules:
17936 The name is divided into segments separated by hyphens, tildes or
17937 underscores and all hyphens, tildes, and underscores are
17938 eliminated. If this leaves the name short enough, we are done.
17941 If the name is too long, the longest segment is located (left-most
17942 if there are two of equal length), and shortened by dropping
17943 its last character. This is repeated until the name is short enough.
17945 As an example, consider the krunching of @*@file{our-strings-wide_fixed.adb}
17946 to fit the name into 8 characters as required by some operating systems.
17949 our-strings-wide_fixed 22
17950 our strings wide fixed 19
17951 our string wide fixed 18
17952 our strin wide fixed 17
17953 our stri wide fixed 16
17954 our stri wide fixe 15
17955 our str wide fixe 14
17956 our str wid fixe 13
17962 Final file name: oustwifi.adb
17966 The file names for all predefined units are always krunched to eight
17967 characters. The krunching of these predefined units uses the following
17968 special prefix replacements:
17972 replaced by @file{^a^A^-}
17975 replaced by @file{^g^G^-}
17978 replaced by @file{^i^I^-}
17981 replaced by @file{^s^S^-}
17984 These system files have a hyphen in the second character position. That
17985 is why normal user files replace such a character with a
17986 ^tilde^dollar sign^, to
17987 avoid confusion with system file names.
17989 As an example of this special rule, consider
17990 @*@file{ada-strings-wide_fixed.adb}, which gets krunched as follows:
17993 ada-strings-wide_fixed 22
17994 a- strings wide fixed 18
17995 a- string wide fixed 17
17996 a- strin wide fixed 16
17997 a- stri wide fixed 15
17998 a- stri wide fixe 14
17999 a- str wide fixe 13
18005 Final file name: a-stwifi.adb
18009 Of course no file shortening algorithm can guarantee uniqueness over all
18010 possible unit names, and if file name krunching is used then it is your
18011 responsibility to ensure that no name clashes occur. The utility
18012 program @code{gnatkr} is supplied for conveniently determining the
18013 krunched name of a file.
18015 @node Examples of gnatkr Usage
18016 @section Examples of @code{gnatkr} Usage
18023 $ gnatkr very_long_unit_name.ads --> velounna.ads
18024 $ gnatkr grandparent-parent-child.ads --> grparchi.ads
18025 $ gnatkr Grandparent.Parent.Child.ads --> grparchi.ads
18026 $ gnatkr grandparent-parent-child --> grparchi
18028 $ gnatkr very_long_unit_name.ads/count=6 --> vlunna.ads
18029 $ gnatkr very_long_unit_name.ads/count=0 --> very_long_unit_name.ads
18032 @node Preprocessing Using gnatprep
18033 @chapter Preprocessing Using @code{gnatprep}
18037 This chapter discusses how to use GNAT's @code{gnatprep} utility for simple
18039 Although designed for use with GNAT, @code{gnatprep} does not depend on any
18040 special GNAT features.
18041 For further discussion of conditional compilation in general, see
18042 @ref{Conditional Compilation}.
18045 * Preprocessing Symbols::
18047 * Switches for gnatprep::
18048 * Form of Definitions File::
18049 * Form of Input Text for gnatprep::
18052 @node Preprocessing Symbols
18053 @section Preprocessing Symbols
18056 Preprocessing symbols are defined in definition files and referred to in
18057 sources to be preprocessed. A Preprocessing symbol is an identifier, following
18058 normal Ada (case-insensitive) rules for its syntax, with the restriction that
18059 all characters need to be in the ASCII set (no accented letters).
18061 @node Using gnatprep
18062 @section Using @code{gnatprep}
18065 To call @code{gnatprep} use
18068 $ gnatprep @ovar{switches} @var{infile} @var{outfile} @ovar{deffile}
18075 is an optional sequence of switches as described in the next section.
18078 is the full name of the input file, which is an Ada source
18079 file containing preprocessor directives.
18082 is the full name of the output file, which is an Ada source
18083 in standard Ada form. When used with GNAT, this file name will
18084 normally have an ads or adb suffix.
18087 is the full name of a text file containing definitions of
18088 preprocessing symbols to be referenced by the preprocessor. This argument is
18089 optional, and can be replaced by the use of the @option{-D} switch.
18093 @node Switches for gnatprep
18094 @section Switches for @code{gnatprep}
18099 @item ^-b^/BLANK_LINES^
18100 @cindex @option{^-b^/BLANK_LINES^} (@command{gnatprep})
18101 Causes both preprocessor lines and the lines deleted by
18102 preprocessing to be replaced by blank lines in the output source file,
18103 preserving line numbers in the output file.
18105 @item ^-c^/COMMENTS^
18106 @cindex @option{^-c^/COMMENTS^} (@command{gnatprep})
18107 Causes both preprocessor lines and the lines deleted
18108 by preprocessing to be retained in the output source as comments marked
18109 with the special string @code{"--! "}. This option will result in line numbers
18110 being preserved in the output file.
18112 @item ^-C^/REPLACE_IN_COMMENTS^
18113 @cindex @option{^-C^/REPLACE_IN_COMMENTS^} (@command{gnatprep})
18114 Causes comments to be scanned. Normally comments are ignored by gnatprep.
18115 If this option is specified, then comments are scanned and any $symbol
18116 substitutions performed as in program text. This is particularly useful
18117 when structured comments are used (e.g., when writing programs in the
18118 SPARK dialect of Ada). Note that this switch is not available when
18119 doing integrated preprocessing (it would be useless in this context
18120 since comments are ignored by the compiler in any case).
18122 @item ^-Dsymbol=value^/ASSOCIATE="symbol=value"^
18123 @cindex @option{^-D^/ASSOCIATE^} (@command{gnatprep})
18124 Defines a new preprocessing symbol, associated with value. If no value is given
18125 on the command line, then symbol is considered to be @code{True}. This switch
18126 can be used in place of a definition file.
18130 @cindex @option{/REMOVE} (@command{gnatprep})
18131 This is the default setting which causes lines deleted by preprocessing
18132 to be entirely removed from the output file.
18135 @item ^-r^/REFERENCE^
18136 @cindex @option{^-r^/REFERENCE^} (@command{gnatprep})
18137 Causes a @code{Source_Reference} pragma to be generated that
18138 references the original input file, so that error messages will use
18139 the file name of this original file. The use of this switch implies
18140 that preprocessor lines are not to be removed from the file, so its
18141 use will force @option{^-b^/BLANK_LINES^} mode if
18142 @option{^-c^/COMMENTS^}
18143 has not been specified explicitly.
18145 Note that if the file to be preprocessed contains multiple units, then
18146 it will be necessary to @code{gnatchop} the output file from
18147 @code{gnatprep}. If a @code{Source_Reference} pragma is present
18148 in the preprocessed file, it will be respected by
18149 @code{gnatchop ^-r^/REFERENCE^}
18150 so that the final chopped files will correctly refer to the original
18151 input source file for @code{gnatprep}.
18153 @item ^-s^/SYMBOLS^
18154 @cindex @option{^-s^/SYMBOLS^} (@command{gnatprep})
18155 Causes a sorted list of symbol names and values to be
18156 listed on the standard output file.
18158 @item ^-u^/UNDEFINED^
18159 @cindex @option{^-u^/UNDEFINED^} (@command{gnatprep})
18160 Causes undefined symbols to be treated as having the value FALSE in the context
18161 of a preprocessor test. In the absence of this option, an undefined symbol in
18162 a @code{#if} or @code{#elsif} test will be treated as an error.
18168 Note: if neither @option{-b} nor @option{-c} is present,
18169 then preprocessor lines and
18170 deleted lines are completely removed from the output, unless -r is
18171 specified, in which case -b is assumed.
18174 @node Form of Definitions File
18175 @section Form of Definitions File
18178 The definitions file contains lines of the form
18185 where symbol is a preprocessing symbol, and value is one of the following:
18189 Empty, corresponding to a null substitution
18191 A string literal using normal Ada syntax
18193 Any sequence of characters from the set
18194 (letters, digits, period, underline).
18198 Comment lines may also appear in the definitions file, starting with
18199 the usual @code{--},
18200 and comments may be added to the definitions lines.
18202 @node Form of Input Text for gnatprep
18203 @section Form of Input Text for @code{gnatprep}
18206 The input text may contain preprocessor conditional inclusion lines,
18207 as well as general symbol substitution sequences.
18209 The preprocessor conditional inclusion commands have the form
18214 #if @i{expression} @r{[}then@r{]}
18216 #elsif @i{expression} @r{[}then@r{]}
18218 #elsif @i{expression} @r{[}then@r{]}
18229 In this example, @i{expression} is defined by the following grammar:
18231 @i{expression} ::= <symbol>
18232 @i{expression} ::= <symbol> = "<value>"
18233 @i{expression} ::= <symbol> = <symbol>
18234 @i{expression} ::= <symbol> 'Defined
18235 @i{expression} ::= not @i{expression}
18236 @i{expression} ::= @i{expression} and @i{expression}
18237 @i{expression} ::= @i{expression} or @i{expression}
18238 @i{expression} ::= @i{expression} and then @i{expression}
18239 @i{expression} ::= @i{expression} or else @i{expression}
18240 @i{expression} ::= ( @i{expression} )
18243 The following restriction exists: it is not allowed to have "and" or "or"
18244 following "not" in the same expression without parentheses. For example, this
18251 This should be one of the following:
18259 For the first test (@i{expression} ::= <symbol>) the symbol must have
18260 either the value true or false, that is to say the right-hand of the
18261 symbol definition must be one of the (case-insensitive) literals
18262 @code{True} or @code{False}. If the value is true, then the
18263 corresponding lines are included, and if the value is false, they are
18266 The test (@i{expression} ::= <symbol> @code{'Defined}) is true only if
18267 the symbol has been defined in the definition file or by a @option{-D}
18268 switch on the command line. Otherwise, the test is false.
18270 The equality tests are case insensitive, as are all the preprocessor lines.
18272 If the symbol referenced is not defined in the symbol definitions file,
18273 then the effect depends on whether or not switch @option{-u}
18274 is specified. If so, then the symbol is treated as if it had the value
18275 false and the test fails. If this switch is not specified, then
18276 it is an error to reference an undefined symbol. It is also an error to
18277 reference a symbol that is defined with a value other than @code{True}
18280 The use of the @code{not} operator inverts the sense of this logical test.
18281 The @code{not} operator cannot be combined with the @code{or} or @code{and}
18282 operators, without parentheses. For example, "if not X or Y then" is not
18283 allowed, but "if (not X) or Y then" and "if not (X or Y) then" are.
18285 The @code{then} keyword is optional as shown
18287 The @code{#} must be the first non-blank character on a line, but
18288 otherwise the format is free form. Spaces or tabs may appear between
18289 the @code{#} and the keyword. The keywords and the symbols are case
18290 insensitive as in normal Ada code. Comments may be used on a
18291 preprocessor line, but other than that, no other tokens may appear on a
18292 preprocessor line. Any number of @code{elsif} clauses can be present,
18293 including none at all. The @code{else} is optional, as in Ada.
18295 The @code{#} marking the start of a preprocessor line must be the first
18296 non-blank character on the line, i.e., it must be preceded only by
18297 spaces or horizontal tabs.
18299 Symbol substitution outside of preprocessor lines is obtained by using
18307 anywhere within a source line, except in a comment or within a
18308 string literal. The identifier
18309 following the @code{$} must match one of the symbols defined in the symbol
18310 definition file, and the result is to substitute the value of the
18311 symbol in place of @code{$symbol} in the output file.
18313 Note that although the substitution of strings within a string literal
18314 is not possible, it is possible to have a symbol whose defined value is
18315 a string literal. So instead of setting XYZ to @code{hello} and writing:
18318 Header : String := "$XYZ";
18322 you should set XYZ to @code{"hello"} and write:
18325 Header : String := $XYZ;
18329 and then the substitution will occur as desired.
18332 @node The GNAT Run-Time Library Builder gnatlbr
18333 @chapter The GNAT Run-Time Library Builder @code{gnatlbr}
18335 @cindex Library builder
18338 @code{gnatlbr} is a tool for rebuilding the GNAT run time with user
18339 supplied configuration pragmas.
18342 * Running gnatlbr::
18343 * Switches for gnatlbr::
18344 * Examples of gnatlbr Usage::
18347 @node Running gnatlbr
18348 @section Running @code{gnatlbr}
18351 The @code{gnatlbr} command has the form
18354 $ GNAT LIBRARY /@r{[}CREATE@r{|}SET@r{|}DELETE@r{]}=directory @r{[}/CONFIG=file@r{]}
18357 @node Switches for gnatlbr
18358 @section Switches for @code{gnatlbr}
18361 @code{gnatlbr} recognizes the following switches:
18365 @item /CREATE=directory
18366 @cindex @code{/CREATE} (@code{gnatlbr})
18367 Create the new run-time library in the specified directory.
18369 @item /SET=directory
18370 @cindex @code{/SET} (@code{gnatlbr})
18371 Make the library in the specified directory the current run-time library.
18373 @item /DELETE=directory
18374 @cindex @code{/DELETE} (@code{gnatlbr})
18375 Delete the run-time library in the specified directory.
18378 @cindex @code{/CONFIG} (@code{gnatlbr})
18379 With /CREATE: Use the configuration pragmas in the specified file when
18380 building the library.
18382 With /SET: Use the configuration pragmas in the specified file when
18387 @node Examples of gnatlbr Usage
18388 @section Example of @code{gnatlbr} Usage
18391 Contents of VAXFLOAT.ADC:
18392 pragma Float_Representation (VAX_Float);
18394 $ GNAT LIBRARY /CREATE=[.VAXFLOAT] /CONFIG=VAXFLOAT.ADC
18396 GNAT LIBRARY rebuilds the run-time library in directory [.VAXFLOAT]
18401 @node The GNAT Library Browser gnatls
18402 @chapter The GNAT Library Browser @code{gnatls}
18404 @cindex Library browser
18407 @code{gnatls} is a tool that outputs information about compiled
18408 units. It gives the relationship between objects, unit names and source
18409 files. It can also be used to check the source dependencies of a unit
18410 as well as various characteristics.
18412 Note: to invoke @code{gnatls} with a project file, use the @code{gnat}
18413 driver (see @ref{The GNAT Driver and Project Files}).
18417 * Switches for gnatls::
18418 * Examples of gnatls Usage::
18421 @node Running gnatls
18422 @section Running @code{gnatls}
18425 The @code{gnatls} command has the form
18428 $ gnatls switches @var{object_or_ali_file}
18432 The main argument is the list of object or @file{ali} files
18433 (@pxref{The Ada Library Information Files})
18434 for which information is requested.
18436 In normal mode, without additional option, @code{gnatls} produces a
18437 four-column listing. Each line represents information for a specific
18438 object. The first column gives the full path of the object, the second
18439 column gives the name of the principal unit in this object, the third
18440 column gives the status of the source and the fourth column gives the
18441 full path of the source representing this unit.
18442 Here is a simple example of use:
18446 ^./^[]^demo1.o demo1 DIF demo1.adb
18447 ^./^[]^demo2.o demo2 OK demo2.adb
18448 ^./^[]^hello.o h1 OK hello.adb
18449 ^./^[]^instr-child.o instr.child MOK instr-child.adb
18450 ^./^[]^instr.o instr OK instr.adb
18451 ^./^[]^tef.o tef DIF tef.adb
18452 ^./^[]^text_io_example.o text_io_example OK text_io_example.adb
18453 ^./^[]^tgef.o tgef DIF tgef.adb
18457 The first line can be interpreted as follows: the main unit which is
18459 object file @file{demo1.o} is demo1, whose main source is in
18460 @file{demo1.adb}. Furthermore, the version of the source used for the
18461 compilation of demo1 has been modified (DIF). Each source file has a status
18462 qualifier which can be:
18465 @item OK (unchanged)
18466 The version of the source file used for the compilation of the
18467 specified unit corresponds exactly to the actual source file.
18469 @item MOK (slightly modified)
18470 The version of the source file used for the compilation of the
18471 specified unit differs from the actual source file but not enough to
18472 require recompilation. If you use gnatmake with the qualifier
18473 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}, a file marked
18474 MOK will not be recompiled.
18476 @item DIF (modified)
18477 No version of the source found on the path corresponds to the source
18478 used to build this object.
18480 @item ??? (file not found)
18481 No source file was found for this unit.
18483 @item HID (hidden, unchanged version not first on PATH)
18484 The version of the source that corresponds exactly to the source used
18485 for compilation has been found on the path but it is hidden by another
18486 version of the same source that has been modified.
18490 @node Switches for gnatls
18491 @section Switches for @code{gnatls}
18494 @code{gnatls} recognizes the following switches:
18498 @cindex @option{--version} @command{gnatls}
18499 Display Copyright and version, then exit disregarding all other options.
18502 @cindex @option{--help} @command{gnatls}
18503 If @option{--version} was not used, display usage, then exit disregarding
18506 @item ^-a^/ALL_UNITS^
18507 @cindex @option{^-a^/ALL_UNITS^} (@code{gnatls})
18508 Consider all units, including those of the predefined Ada library.
18509 Especially useful with @option{^-d^/DEPENDENCIES^}.
18511 @item ^-d^/DEPENDENCIES^
18512 @cindex @option{^-d^/DEPENDENCIES^} (@code{gnatls})
18513 List sources from which specified units depend on.
18515 @item ^-h^/OUTPUT=OPTIONS^
18516 @cindex @option{^-h^/OUTPUT=OPTIONS^} (@code{gnatls})
18517 Output the list of options.
18519 @item ^-o^/OUTPUT=OBJECTS^
18520 @cindex @option{^-o^/OUTPUT=OBJECTS^} (@code{gnatls})
18521 Only output information about object files.
18523 @item ^-s^/OUTPUT=SOURCES^
18524 @cindex @option{^-s^/OUTPUT=SOURCES^} (@code{gnatls})
18525 Only output information about source files.
18527 @item ^-u^/OUTPUT=UNITS^
18528 @cindex @option{^-u^/OUTPUT=UNITS^} (@code{gnatls})
18529 Only output information about compilation units.
18531 @item ^-files^/FILES^=@var{file}
18532 @cindex @option{^-files^/FILES^} (@code{gnatls})
18533 Take as arguments the files listed in text file @var{file}.
18534 Text file @var{file} may contain empty lines that are ignored.
18535 Each nonempty line should contain the name of an existing file.
18536 Several such switches may be specified simultaneously.
18538 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18539 @itemx ^-aI^/SOURCE_SEARCH=^@var{dir}
18540 @itemx ^-I^/SEARCH=^@var{dir}
18541 @itemx ^-I-^/NOCURRENT_DIRECTORY^
18543 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatls})
18544 @cindex @option{^-aI^/SOURCE_SEARCH^} (@code{gnatls})
18545 @cindex @option{^-I^/SEARCH^} (@code{gnatls})
18546 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatls})
18547 Source path manipulation. Same meaning as the equivalent @command{gnatmake}
18548 flags (@pxref{Switches for gnatmake}).
18550 @item --RTS=@var{rts-path}
18551 @cindex @option{--RTS} (@code{gnatls})
18552 Specifies the default location of the runtime library. Same meaning as the
18553 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
18555 @item ^-v^/OUTPUT=VERBOSE^
18556 @cindex @option{^-v^/OUTPUT=VERBOSE^} (@code{gnatls})
18557 Verbose mode. Output the complete source, object and project paths. Do not use
18558 the default column layout but instead use long format giving as much as
18559 information possible on each requested units, including special
18560 characteristics such as:
18563 @item Preelaborable
18564 The unit is preelaborable in the Ada sense.
18567 No elaboration code has been produced by the compiler for this unit.
18570 The unit is pure in the Ada sense.
18572 @item Elaborate_Body
18573 The unit contains a pragma Elaborate_Body.
18576 The unit contains a pragma Remote_Types.
18578 @item Shared_Passive
18579 The unit contains a pragma Shared_Passive.
18582 This unit is part of the predefined environment and cannot be modified
18585 @item Remote_Call_Interface
18586 The unit contains a pragma Remote_Call_Interface.
18592 @node Examples of gnatls Usage
18593 @section Example of @code{gnatls} Usage
18597 Example of using the verbose switch. Note how the source and
18598 object paths are affected by the -I switch.
18601 $ gnatls -v -I.. demo1.o
18603 GNATLS 5.03w (20041123-34)
18604 Copyright 1997-2004 Free Software Foundation, Inc.
18606 Source Search Path:
18607 <Current_Directory>
18609 /home/comar/local/adainclude/
18611 Object Search Path:
18612 <Current_Directory>
18614 /home/comar/local/lib/gcc-lib/x86-linux/3.4.3/adalib/
18616 Project Search Path:
18617 <Current_Directory>
18618 /home/comar/local/lib/gnat/
18623 Kind => subprogram body
18624 Flags => No_Elab_Code
18625 Source => demo1.adb modified
18629 The following is an example of use of the dependency list.
18630 Note the use of the -s switch
18631 which gives a straight list of source files. This can be useful for
18632 building specialized scripts.
18635 $ gnatls -d demo2.o
18636 ./demo2.o demo2 OK demo2.adb
18642 $ gnatls -d -s -a demo1.o
18644 /home/comar/local/adainclude/ada.ads
18645 /home/comar/local/adainclude/a-finali.ads
18646 /home/comar/local/adainclude/a-filico.ads
18647 /home/comar/local/adainclude/a-stream.ads
18648 /home/comar/local/adainclude/a-tags.ads
18651 /home/comar/local/adainclude/gnat.ads
18652 /home/comar/local/adainclude/g-io.ads
18654 /home/comar/local/adainclude/system.ads
18655 /home/comar/local/adainclude/s-exctab.ads
18656 /home/comar/local/adainclude/s-finimp.ads
18657 /home/comar/local/adainclude/s-finroo.ads
18658 /home/comar/local/adainclude/s-secsta.ads
18659 /home/comar/local/adainclude/s-stalib.ads
18660 /home/comar/local/adainclude/s-stoele.ads
18661 /home/comar/local/adainclude/s-stratt.ads
18662 /home/comar/local/adainclude/s-tasoli.ads
18663 /home/comar/local/adainclude/s-unstyp.ads
18664 /home/comar/local/adainclude/unchconv.ads
18670 GNAT LIST /DEPENDENCIES /OUTPUT=SOURCES /ALL_UNITS DEMO1.ADB
18672 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]ada.ads
18673 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-finali.ads
18674 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-filico.ads
18675 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-stream.ads
18676 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-tags.ads
18680 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]gnat.ads
18681 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]g-io.ads
18683 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]system.ads
18684 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-exctab.ads
18685 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finimp.ads
18686 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finroo.ads
18687 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-secsta.ads
18688 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stalib.ads
18689 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stoele.ads
18690 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stratt.ads
18691 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-tasoli.ads
18692 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-unstyp.ads
18693 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]unchconv.ads
18697 @node Cleaning Up Using gnatclean
18698 @chapter Cleaning Up Using @code{gnatclean}
18700 @cindex Cleaning tool
18703 @code{gnatclean} is a tool that allows the deletion of files produced by the
18704 compiler, binder and linker, including ALI files, object files, tree files,
18705 expanded source files, library files, interface copy source files, binder
18706 generated files and executable files.
18709 * Running gnatclean::
18710 * Switches for gnatclean::
18711 @c * Examples of gnatclean Usage::
18714 @node Running gnatclean
18715 @section Running @code{gnatclean}
18718 The @code{gnatclean} command has the form:
18721 $ gnatclean switches @var{names}
18725 @var{names} is a list of source file names. Suffixes @code{.^ads^ADS^} and
18726 @code{^adb^ADB^} may be omitted. If a project file is specified using switch
18727 @code{^-P^/PROJECT_FILE=^}, then @var{names} may be completely omitted.
18730 In normal mode, @code{gnatclean} delete the files produced by the compiler and,
18731 if switch @code{^-c^/COMPILER_FILES_ONLY^} is not specified, by the binder and
18732 the linker. In informative-only mode, specified by switch
18733 @code{^-n^/NODELETE^}, the list of files that would have been deleted in
18734 normal mode is listed, but no file is actually deleted.
18736 @node Switches for gnatclean
18737 @section Switches for @code{gnatclean}
18740 @code{gnatclean} recognizes the following switches:
18744 @cindex @option{--version} @command{gnatclean}
18745 Display Copyright and version, then exit disregarding all other options.
18748 @cindex @option{--help} @command{gnatclean}
18749 If @option{--version} was not used, display usage, then exit disregarding
18752 @item ^-c^/COMPILER_FILES_ONLY^
18753 @cindex @option{^-c^/COMPILER_FILES_ONLY^} (@code{gnatclean})
18754 Only attempt to delete the files produced by the compiler, not those produced
18755 by the binder or the linker. The files that are not to be deleted are library
18756 files, interface copy files, binder generated files and executable files.
18758 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
18759 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@code{gnatclean})
18760 Indicate that ALI and object files should normally be found in directory
18763 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
18764 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@code{gnatclean})
18765 When using project files, if some errors or warnings are detected during
18766 parsing and verbose mode is not in effect (no use of switch
18767 ^-v^/VERBOSE^), then error lines start with the full path name of the project
18768 file, rather than its simple file name.
18771 @cindex @option{^-h^/HELP^} (@code{gnatclean})
18772 Output a message explaining the usage of @code{^gnatclean^gnatclean^}.
18774 @item ^-n^/NODELETE^
18775 @cindex @option{^-n^/NODELETE^} (@code{gnatclean})
18776 Informative-only mode. Do not delete any files. Output the list of the files
18777 that would have been deleted if this switch was not specified.
18779 @item ^-P^/PROJECT_FILE=^@var{project}
18780 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatclean})
18781 Use project file @var{project}. Only one such switch can be used.
18782 When cleaning a project file, the files produced by the compilation of the
18783 immediate sources or inherited sources of the project files are to be
18784 deleted. This is not depending on the presence or not of executable names
18785 on the command line.
18788 @cindex @option{^-q^/QUIET^} (@code{gnatclean})
18789 Quiet output. If there are no errors, do not output anything, except in
18790 verbose mode (switch ^-v^/VERBOSE^) or in informative-only mode
18791 (switch ^-n^/NODELETE^).
18793 @item ^-r^/RECURSIVE^
18794 @cindex @option{^-r^/RECURSIVE^} (@code{gnatclean})
18795 When a project file is specified (using switch ^-P^/PROJECT_FILE=^),
18796 clean all imported and extended project files, recursively. If this switch
18797 is not specified, only the files related to the main project file are to be
18798 deleted. This switch has no effect if no project file is specified.
18800 @item ^-v^/VERBOSE^
18801 @cindex @option{^-v^/VERBOSE^} (@code{gnatclean})
18804 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
18805 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (@code{gnatclean})
18806 Indicates the verbosity of the parsing of GNAT project files.
18807 @xref{Switches Related to Project Files}.
18809 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
18810 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (@code{gnatclean})
18811 Indicates that external variable @var{name} has the value @var{value}.
18812 The Project Manager will use this value for occurrences of
18813 @code{external(name)} when parsing the project file.
18814 @xref{Switches Related to Project Files}.
18816 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18817 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatclean})
18818 When searching for ALI and object files, look in directory
18821 @item ^-I^/SEARCH=^@var{dir}
18822 @cindex @option{^-I^/SEARCH^} (@code{gnatclean})
18823 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}}.
18825 @item ^-I-^/NOCURRENT_DIRECTORY^
18826 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatclean})
18827 @cindex Source files, suppressing search
18828 Do not look for ALI or object files in the directory
18829 where @code{gnatclean} was invoked.
18833 @c @node Examples of gnatclean Usage
18834 @c @section Examples of @code{gnatclean} Usage
18837 @node GNAT and Libraries
18838 @chapter GNAT and Libraries
18839 @cindex Library, building, installing, using
18842 This chapter describes how to build and use libraries with GNAT, and also shows
18843 how to recompile the GNAT run-time library. You should be familiar with the
18844 Project Manager facility (@pxref{GNAT Project Manager}) before reading this
18848 * Introduction to Libraries in GNAT::
18849 * General Ada Libraries::
18850 * Stand-alone Ada Libraries::
18851 * Rebuilding the GNAT Run-Time Library::
18854 @node Introduction to Libraries in GNAT
18855 @section Introduction to Libraries in GNAT
18858 A library is, conceptually, a collection of objects which does not have its
18859 own main thread of execution, but rather provides certain services to the
18860 applications that use it. A library can be either statically linked with the
18861 application, in which case its code is directly included in the application,
18862 or, on platforms that support it, be dynamically linked, in which case
18863 its code is shared by all applications making use of this library.
18865 GNAT supports both types of libraries.
18866 In the static case, the compiled code can be provided in different ways. The
18867 simplest approach is to provide directly the set of objects resulting from
18868 compilation of the library source files. Alternatively, you can group the
18869 objects into an archive using whatever commands are provided by the operating
18870 system. For the latter case, the objects are grouped into a shared library.
18872 In the GNAT environment, a library has three types of components:
18878 @xref{The Ada Library Information Files}.
18880 Object files, an archive or a shared library.
18884 A GNAT library may expose all its source files, which is useful for
18885 documentation purposes. Alternatively, it may expose only the units needed by
18886 an external user to make use of the library. That is to say, the specs
18887 reflecting the library services along with all the units needed to compile
18888 those specs, which can include generic bodies or any body implementing an
18889 inlined routine. In the case of @emph{stand-alone libraries} those exposed
18890 units are called @emph{interface units} (@pxref{Stand-alone Ada Libraries}).
18892 All compilation units comprising an application, including those in a library,
18893 need to be elaborated in an order partially defined by Ada's semantics. GNAT
18894 computes the elaboration order from the @file{ALI} files and this is why they
18895 constitute a mandatory part of GNAT libraries.
18896 @emph{Stand-alone libraries} are the exception to this rule because a specific
18897 library elaboration routine is produced independently of the application(s)
18900 @node General Ada Libraries
18901 @section General Ada Libraries
18904 * Building a library::
18905 * Installing a library::
18906 * Using a library::
18909 @node Building a library
18910 @subsection Building a library
18913 The easiest way to build a library is to use the Project Manager,
18914 which supports a special type of project called a @emph{Library Project}
18915 (@pxref{Library Projects}).
18917 A project is considered a library project, when two project-level attributes
18918 are defined in it: @code{Library_Name} and @code{Library_Dir}. In order to
18919 control different aspects of library configuration, additional optional
18920 project-level attributes can be specified:
18923 This attribute controls whether the library is to be static or dynamic
18925 @item Library_Version
18926 This attribute specifies the library version; this value is used
18927 during dynamic linking of shared libraries to determine if the currently
18928 installed versions of the binaries are compatible.
18930 @item Library_Options
18932 These attributes specify additional low-level options to be used during
18933 library generation, and redefine the actual application used to generate
18938 The GNAT Project Manager takes full care of the library maintenance task,
18939 including recompilation of the source files for which objects do not exist
18940 or are not up to date, assembly of the library archive, and installation of
18941 the library (i.e., copying associated source, object and @file{ALI} files
18942 to the specified location).
18944 Here is a simple library project file:
18945 @smallexample @c ada
18947 for Source_Dirs use ("src1", "src2");
18948 for Object_Dir use "obj";
18949 for Library_Name use "mylib";
18950 for Library_Dir use "lib";
18951 for Library_Kind use "dynamic";
18956 and the compilation command to build and install the library:
18958 @smallexample @c ada
18959 $ gnatmake -Pmy_lib
18963 It is not entirely trivial to perform manually all the steps required to
18964 produce a library. We recommend that you use the GNAT Project Manager
18965 for this task. In special cases where this is not desired, the necessary
18966 steps are discussed below.
18968 There are various possibilities for compiling the units that make up the
18969 library: for example with a Makefile (@pxref{Using the GNU make Utility}) or
18970 with a conventional script. For simple libraries, it is also possible to create
18971 a dummy main program which depends upon all the packages that comprise the
18972 interface of the library. This dummy main program can then be given to
18973 @command{gnatmake}, which will ensure that all necessary objects are built.
18975 After this task is accomplished, you should follow the standard procedure
18976 of the underlying operating system to produce the static or shared library.
18978 Here is an example of such a dummy program:
18979 @smallexample @c ada
18981 with My_Lib.Service1;
18982 with My_Lib.Service2;
18983 with My_Lib.Service3;
18984 procedure My_Lib_Dummy is
18992 Here are the generic commands that will build an archive or a shared library.
18995 # compiling the library
18996 $ gnatmake -c my_lib_dummy.adb
18998 # we don't need the dummy object itself
18999 $ rm my_lib_dummy.o my_lib_dummy.ali
19001 # create an archive with the remaining objects
19002 $ ar rc libmy_lib.a *.o
19003 # some systems may require "ranlib" to be run as well
19005 # or create a shared library
19006 $ gcc -shared -o libmy_lib.so *.o
19007 # some systems may require the code to have been compiled with -fPIC
19009 # remove the object files that are now in the library
19012 # Make the ALI files read-only so that gnatmake will not try to
19013 # regenerate the objects that are in the library
19018 Please note that the library must have a name of the form @file{lib@var{xxx}.a}
19019 or @file{lib@var{xxx}.so} (or @file{lib@var{xxx}.dll} on Windows) in order to
19020 be accessed by the directive @option{-l@var{xxx}} at link time.
19022 @node Installing a library
19023 @subsection Installing a library
19024 @cindex @code{ADA_PROJECT_PATH}
19027 If you use project files, library installation is part of the library build
19028 process. Thus no further action is needed in order to make use of the
19029 libraries that are built as part of the general application build. A usable
19030 version of the library is installed in the directory specified by the
19031 @code{Library_Dir} attribute of the library project file.
19033 You may want to install a library in a context different from where the library
19034 is built. This situation arises with third party suppliers, who may want
19035 to distribute a library in binary form where the user is not expected to be
19036 able to recompile the library. The simplest option in this case is to provide
19037 a project file slightly different from the one used to build the library, by
19038 using the @code{externally_built} attribute. For instance, the project
19039 file used to build the library in the previous section can be changed into the
19040 following one when the library is installed:
19042 @smallexample @c projectfile
19044 for Source_Dirs use ("src1", "src2");
19045 for Library_Name use "mylib";
19046 for Library_Dir use "lib";
19047 for Library_Kind use "dynamic";
19048 for Externally_Built use "true";
19053 This project file assumes that the directories @file{src1},
19054 @file{src2}, and @file{lib} exist in
19055 the directory containing the project file. The @code{externally_built}
19056 attribute makes it clear to the GNAT builder that it should not attempt to
19057 recompile any of the units from this library. It allows the library provider to
19058 restrict the source set to the minimum necessary for clients to make use of the
19059 library as described in the first section of this chapter. It is the
19060 responsibility of the library provider to install the necessary sources, ALI
19061 files and libraries in the directories mentioned in the project file. For
19062 convenience, the user's library project file should be installed in a location
19063 that will be searched automatically by the GNAT
19064 builder. These are the directories referenced in the @env{ADA_PROJECT_PATH}
19065 environment variable (@pxref{Importing Projects}), and also the default GNAT
19066 library location that can be queried with @command{gnatls -v} and is usually of
19067 the form $gnat_install_root/lib/gnat.
19069 When project files are not an option, it is also possible, but not recommended,
19070 to install the library so that the sources needed to use the library are on the
19071 Ada source path and the ALI files & libraries be on the Ada Object path (see
19072 @ref{Search Paths and the Run-Time Library (RTL)}. Alternatively, the system
19073 administrator can place general-purpose libraries in the default compiler
19074 paths, by specifying the libraries' location in the configuration files
19075 @file{ada_source_path} and @file{ada_object_path}. These configuration files
19076 must be located in the GNAT installation tree at the same place as the gcc spec
19077 file. The location of the gcc spec file can be determined as follows:
19083 The configuration files mentioned above have a simple format: each line
19084 must contain one unique directory name.
19085 Those names are added to the corresponding path
19086 in their order of appearance in the file. The names can be either absolute
19087 or relative; in the latter case, they are relative to where theses files
19090 The files @file{ada_source_path} and @file{ada_object_path} might not be
19092 GNAT installation, in which case, GNAT will look for its run-time library in
19093 the directories @file{adainclude} (for the sources) and @file{adalib} (for the
19094 objects and @file{ALI} files). When the files exist, the compiler does not
19095 look in @file{adainclude} and @file{adalib}, and thus the
19096 @file{ada_source_path} file
19097 must contain the location for the GNAT run-time sources (which can simply
19098 be @file{adainclude}). In the same way, the @file{ada_object_path} file must
19099 contain the location for the GNAT run-time objects (which can simply
19102 You can also specify a new default path to the run-time library at compilation
19103 time with the switch @option{--RTS=rts-path}. You can thus choose / change
19104 the run-time library you want your program to be compiled with. This switch is
19105 recognized by @command{gcc}, @command{gnatmake}, @command{gnatbind},
19106 @command{gnatls}, @command{gnatfind} and @command{gnatxref}.
19108 It is possible to install a library before or after the standard GNAT
19109 library, by reordering the lines in the configuration files. In general, a
19110 library must be installed before the GNAT library if it redefines
19113 @node Using a library
19114 @subsection Using a library
19116 @noindent Once again, the project facility greatly simplifies the use of
19117 libraries. In this context, using a library is just a matter of adding a
19118 @code{with} clause in the user project. For instance, to make use of the
19119 library @code{My_Lib} shown in examples in earlier sections, you can
19122 @smallexample @c projectfile
19129 Even if you have a third-party, non-Ada library, you can still use GNAT's
19130 Project Manager facility to provide a wrapper for it. For example, the
19131 following project, when @code{with}ed by your main project, will link with the
19132 third-party library @file{liba.a}:
19134 @smallexample @c projectfile
19137 for Externally_Built use "true";
19138 for Source_Files use ();
19139 for Library_Dir use "lib";
19140 for Library_Name use "a";
19141 for Library_Kind use "static";
19145 This is an alternative to the use of @code{pragma Linker_Options}. It is
19146 especially interesting in the context of systems with several interdependent
19147 static libraries where finding a proper linker order is not easy and best be
19148 left to the tools having visibility over project dependence information.
19151 In order to use an Ada library manually, you need to make sure that this
19152 library is on both your source and object path
19153 (see @ref{Search Paths and the Run-Time Library (RTL)}
19154 and @ref{Search Paths for gnatbind}). Furthermore, when the objects are grouped
19155 in an archive or a shared library, you need to specify the desired
19156 library at link time.
19158 For example, you can use the library @file{mylib} installed in
19159 @file{/dir/my_lib_src} and @file{/dir/my_lib_obj} with the following commands:
19162 $ gnatmake -aI/dir/my_lib_src -aO/dir/my_lib_obj my_appl \
19167 This can be expressed more simply:
19172 when the following conditions are met:
19175 @file{/dir/my_lib_src} has been added by the user to the environment
19176 variable @env{ADA_INCLUDE_PATH}, or by the administrator to the file
19177 @file{ada_source_path}
19179 @file{/dir/my_lib_obj} has been added by the user to the environment
19180 variable @env{ADA_OBJECTS_PATH}, or by the administrator to the file
19181 @file{ada_object_path}
19183 a pragma @code{Linker_Options} has been added to one of the sources.
19186 @smallexample @c ada
19187 pragma Linker_Options ("-lmy_lib");
19191 @node Stand-alone Ada Libraries
19192 @section Stand-alone Ada Libraries
19193 @cindex Stand-alone library, building, using
19196 * Introduction to Stand-alone Libraries::
19197 * Building a Stand-alone Library::
19198 * Creating a Stand-alone Library to be used in a non-Ada context::
19199 * Restrictions in Stand-alone Libraries::
19202 @node Introduction to Stand-alone Libraries
19203 @subsection Introduction to Stand-alone Libraries
19206 A Stand-alone Library (abbreviated ``SAL'') is a library that contains the
19208 elaborate the Ada units that are included in the library. In contrast with
19209 an ordinary library, which consists of all sources, objects and @file{ALI}
19211 library, a SAL may specify a restricted subset of compilation units
19212 to serve as a library interface. In this case, the fully
19213 self-sufficient set of files will normally consist of an objects
19214 archive, the sources of interface units' specs, and the @file{ALI}
19215 files of interface units.
19216 If an interface spec contains a generic unit or an inlined subprogram,
19218 source must also be provided; if the units that must be provided in the source
19219 form depend on other units, the source and @file{ALI} files of those must
19222 The main purpose of a SAL is to minimize the recompilation overhead of client
19223 applications when a new version of the library is installed. Specifically,
19224 if the interface sources have not changed, client applications do not need to
19225 be recompiled. If, furthermore, a SAL is provided in the shared form and its
19226 version, controlled by @code{Library_Version} attribute, is not changed,
19227 then the clients do not need to be relinked.
19229 SALs also allow the library providers to minimize the amount of library source
19230 text exposed to the clients. Such ``information hiding'' might be useful or
19231 necessary for various reasons.
19233 Stand-alone libraries are also well suited to be used in an executable whose
19234 main routine is not written in Ada.
19236 @node Building a Stand-alone Library
19237 @subsection Building a Stand-alone Library
19240 GNAT's Project facility provides a simple way of building and installing
19241 stand-alone libraries; see @ref{Stand-alone Library Projects}.
19242 To be a Stand-alone Library Project, in addition to the two attributes
19243 that make a project a Library Project (@code{Library_Name} and
19244 @code{Library_Dir}; see @ref{Library Projects}), the attribute
19245 @code{Library_Interface} must be defined. For example:
19247 @smallexample @c projectfile
19249 for Library_Dir use "lib_dir";
19250 for Library_Name use "dummy";
19251 for Library_Interface use ("int1", "int1.child");
19256 Attribute @code{Library_Interface} has a non-empty string list value,
19257 each string in the list designating a unit contained in an immediate source
19258 of the project file.
19260 When a Stand-alone Library is built, first the binder is invoked to build
19261 a package whose name depends on the library name
19262 (@file{^b~dummy.ads/b^B$DUMMY.ADS/B^} in the example above).
19263 This binder-generated package includes initialization and
19264 finalization procedures whose
19265 names depend on the library name (@code{dummyinit} and @code{dummyfinal}
19267 above). The object corresponding to this package is included in the library.
19269 You must ensure timely (e.g., prior to any use of interfaces in the SAL)
19270 calling of these procedures if a static SAL is built, or if a shared SAL
19272 with the project-level attribute @code{Library_Auto_Init} set to
19275 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
19276 (those that are listed in attribute @code{Library_Interface}) are copied to
19277 the Library Directory. As a consequence, only the Interface Units may be
19278 imported from Ada units outside of the library. If other units are imported,
19279 the binding phase will fail.
19281 The attribute @code{Library_Src_Dir} may be specified for a
19282 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
19283 single string value. Its value must be the path (absolute or relative to the
19284 project directory) of an existing directory. This directory cannot be the
19285 object directory or one of the source directories, but it can be the same as
19286 the library directory. The sources of the Interface
19287 Units of the library that are needed by an Ada client of the library will be
19288 copied to the designated directory, called the Interface Copy directory.
19289 These sources include the specs of the Interface Units, but they may also
19290 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
19291 are used, or when there is a generic unit in the spec. Before the sources
19292 are copied to the Interface Copy directory, an attempt is made to delete all
19293 files in the Interface Copy directory.
19295 Building stand-alone libraries by hand is somewhat tedious, but for those
19296 occasions when it is necessary here are the steps that you need to perform:
19299 Compile all library sources.
19302 Invoke the binder with the switch @option{-n} (No Ada main program),
19303 with all the @file{ALI} files of the interfaces, and
19304 with the switch @option{-L} to give specific names to the @code{init}
19305 and @code{final} procedures. For example:
19307 gnatbind -n int1.ali int2.ali -Lsal1
19311 Compile the binder generated file:
19317 Link the dynamic library with all the necessary object files,
19318 indicating to the linker the names of the @code{init} (and possibly
19319 @code{final}) procedures for automatic initialization (and finalization).
19320 The built library should be placed in a directory different from
19321 the object directory.
19324 Copy the @code{ALI} files of the interface to the library directory,
19325 add in this copy an indication that it is an interface to a SAL
19326 (i.e., add a word @option{SL} on the line in the @file{ALI} file that starts
19327 with letter ``P'') and make the modified copy of the @file{ALI} file
19332 Using SALs is not different from using other libraries
19333 (see @ref{Using a library}).
19335 @node Creating a Stand-alone Library to be used in a non-Ada context
19336 @subsection Creating a Stand-alone Library to be used in a non-Ada context
19339 It is easy to adapt the SAL build procedure discussed above for use of a SAL in
19342 The only extra step required is to ensure that library interface subprograms
19343 are compatible with the main program, by means of @code{pragma Export}
19344 or @code{pragma Convention}.
19346 Here is an example of simple library interface for use with C main program:
19348 @smallexample @c ada
19349 package Interface is
19351 procedure Do_Something;
19352 pragma Export (C, Do_Something, "do_something");
19354 procedure Do_Something_Else;
19355 pragma Export (C, Do_Something_Else, "do_something_else");
19361 On the foreign language side, you must provide a ``foreign'' view of the
19362 library interface; remember that it should contain elaboration routines in
19363 addition to interface subprograms.
19365 The example below shows the content of @code{mylib_interface.h} (note
19366 that there is no rule for the naming of this file, any name can be used)
19368 /* the library elaboration procedure */
19369 extern void mylibinit (void);
19371 /* the library finalization procedure */
19372 extern void mylibfinal (void);
19374 /* the interface exported by the library */
19375 extern void do_something (void);
19376 extern void do_something_else (void);
19380 Libraries built as explained above can be used from any program, provided
19381 that the elaboration procedures (named @code{mylibinit} in the previous
19382 example) are called before the library services are used. Any number of
19383 libraries can be used simultaneously, as long as the elaboration
19384 procedure of each library is called.
19386 Below is an example of a C program that uses the @code{mylib} library.
19389 #include "mylib_interface.h"
19394 /* First, elaborate the library before using it */
19397 /* Main program, using the library exported entities */
19399 do_something_else ();
19401 /* Library finalization at the end of the program */
19408 Note that invoking any library finalization procedure generated by
19409 @code{gnatbind} shuts down the Ada run-time environment.
19411 finalization of all Ada libraries must be performed at the end of the program.
19412 No call to these libraries or to the Ada run-time library should be made
19413 after the finalization phase.
19415 @node Restrictions in Stand-alone Libraries
19416 @subsection Restrictions in Stand-alone Libraries
19419 The pragmas listed below should be used with caution inside libraries,
19420 as they can create incompatibilities with other Ada libraries:
19422 @item pragma @code{Locking_Policy}
19423 @item pragma @code{Queuing_Policy}
19424 @item pragma @code{Task_Dispatching_Policy}
19425 @item pragma @code{Unreserve_All_Interrupts}
19429 When using a library that contains such pragmas, the user must make sure
19430 that all libraries use the same pragmas with the same values. Otherwise,
19431 @code{Program_Error} will
19432 be raised during the elaboration of the conflicting
19433 libraries. The usage of these pragmas and its consequences for the user
19434 should therefore be well documented.
19436 Similarly, the traceback in the exception occurrence mechanism should be
19437 enabled or disabled in a consistent manner across all libraries.
19438 Otherwise, Program_Error will be raised during the elaboration of the
19439 conflicting libraries.
19441 If the @code{Version} or @code{Body_Version}
19442 attributes are used inside a library, then you need to
19443 perform a @code{gnatbind} step that specifies all @file{ALI} files in all
19444 libraries, so that version identifiers can be properly computed.
19445 In practice these attributes are rarely used, so this is unlikely
19446 to be a consideration.
19448 @node Rebuilding the GNAT Run-Time Library
19449 @section Rebuilding the GNAT Run-Time Library
19450 @cindex GNAT Run-Time Library, rebuilding
19451 @cindex Building the GNAT Run-Time Library
19452 @cindex Rebuilding the GNAT Run-Time Library
19453 @cindex Run-Time Library, rebuilding
19456 It may be useful to recompile the GNAT library in various contexts, the
19457 most important one being the use of partition-wide configuration pragmas
19458 such as @code{Normalize_Scalars}. A special Makefile called
19459 @code{Makefile.adalib} is provided to that effect and can be found in
19460 the directory containing the GNAT library. The location of this
19461 directory depends on the way the GNAT environment has been installed and can
19462 be determined by means of the command:
19469 The last entry in the object search path usually contains the
19470 gnat library. This Makefile contains its own documentation and in
19471 particular the set of instructions needed to rebuild a new library and
19474 @node Using the GNU make Utility
19475 @chapter Using the GNU @code{make} Utility
19479 This chapter offers some examples of makefiles that solve specific
19480 problems. It does not explain how to write a makefile (@pxref{Top,, GNU
19481 make, make, GNU @code{make}}), nor does it try to replace the
19482 @command{gnatmake} utility (@pxref{The GNAT Make Program gnatmake}).
19484 All the examples in this section are specific to the GNU version of
19485 make. Although @command{make} is a standard utility, and the basic language
19486 is the same, these examples use some advanced features found only in
19490 * Using gnatmake in a Makefile::
19491 * Automatically Creating a List of Directories::
19492 * Generating the Command Line Switches::
19493 * Overcoming Command Line Length Limits::
19496 @node Using gnatmake in a Makefile
19497 @section Using gnatmake in a Makefile
19502 Complex project organizations can be handled in a very powerful way by
19503 using GNU make combined with gnatmake. For instance, here is a Makefile
19504 which allows you to build each subsystem of a big project into a separate
19505 shared library. Such a makefile allows you to significantly reduce the link
19506 time of very big applications while maintaining full coherence at
19507 each step of the build process.
19509 The list of dependencies are handled automatically by
19510 @command{gnatmake}. The Makefile is simply used to call gnatmake in each of
19511 the appropriate directories.
19513 Note that you should also read the example on how to automatically
19514 create the list of directories
19515 (@pxref{Automatically Creating a List of Directories})
19516 which might help you in case your project has a lot of subdirectories.
19521 @font@heightrm=cmr8
19524 ## This Makefile is intended to be used with the following directory
19526 ## - The sources are split into a series of csc (computer software components)
19527 ## Each of these csc is put in its own directory.
19528 ## Their name are referenced by the directory names.
19529 ## They will be compiled into shared library (although this would also work
19530 ## with static libraries
19531 ## - The main program (and possibly other packages that do not belong to any
19532 ## csc is put in the top level directory (where the Makefile is).
19533 ## toplevel_dir __ first_csc (sources) __ lib (will contain the library)
19534 ## \_ second_csc (sources) __ lib (will contain the library)
19536 ## Although this Makefile is build for shared library, it is easy to modify
19537 ## to build partial link objects instead (modify the lines with -shared and
19540 ## With this makefile, you can change any file in the system or add any new
19541 ## file, and everything will be recompiled correctly (only the relevant shared
19542 ## objects will be recompiled, and the main program will be re-linked).
19544 # The list of computer software component for your project. This might be
19545 # generated automatically.
19548 # Name of the main program (no extension)
19551 # If we need to build objects with -fPIC, uncomment the following line
19554 # The following variable should give the directory containing libgnat.so
19555 # You can get this directory through 'gnatls -v'. This is usually the last
19556 # directory in the Object_Path.
19559 # The directories for the libraries
19560 # (This macro expands the list of CSC to the list of shared libraries, you
19561 # could simply use the expanded form:
19562 # LIB_DIR=aa/lib/libaa.so bb/lib/libbb.so cc/lib/libcc.so
19563 LIB_DIR=$@{foreach dir,$@{CSC_LIST@},$@{dir@}/lib/lib$@{dir@}.so@}
19565 $@{MAIN@}: objects $@{LIB_DIR@}
19566 gnatbind $@{MAIN@} $@{CSC_LIST:%=-aO%/lib@} -shared
19567 gnatlink $@{MAIN@} $@{CSC_LIST:%=-l%@}
19570 # recompile the sources
19571 gnatmake -c -i $@{MAIN@}.adb $@{NEED_FPIC@} $@{CSC_LIST:%=-I%@}
19573 # Note: In a future version of GNAT, the following commands will be simplified
19574 # by a new tool, gnatmlib
19576 mkdir -p $@{dir $@@ @}
19577 cd $@{dir $@@ @} && gcc -shared -o $@{notdir $@@ @} ../*.o -L$@{GLIB@} -lgnat
19578 cd $@{dir $@@ @} && cp -f ../*.ali .
19580 # The dependencies for the modules
19581 # Note that we have to force the expansion of *.o, since in some cases
19582 # make won't be able to do it itself.
19583 aa/lib/libaa.so: $@{wildcard aa/*.o@}
19584 bb/lib/libbb.so: $@{wildcard bb/*.o@}
19585 cc/lib/libcc.so: $@{wildcard cc/*.o@}
19587 # Make sure all of the shared libraries are in the path before starting the
19590 LD_LIBRARY_PATH=`pwd`/aa/lib:`pwd`/bb/lib:`pwd`/cc/lib ./$@{MAIN@}
19593 $@{RM@} -rf $@{CSC_LIST:%=%/lib@}
19594 $@{RM@} $@{CSC_LIST:%=%/*.ali@}
19595 $@{RM@} $@{CSC_LIST:%=%/*.o@}
19596 $@{RM@} *.o *.ali $@{MAIN@}
19599 @node Automatically Creating a List of Directories
19600 @section Automatically Creating a List of Directories
19603 In most makefiles, you will have to specify a list of directories, and
19604 store it in a variable. For small projects, it is often easier to
19605 specify each of them by hand, since you then have full control over what
19606 is the proper order for these directories, which ones should be
19609 However, in larger projects, which might involve hundreds of
19610 subdirectories, it might be more convenient to generate this list
19613 The example below presents two methods. The first one, although less
19614 general, gives you more control over the list. It involves wildcard
19615 characters, that are automatically expanded by @command{make}. Its
19616 shortcoming is that you need to explicitly specify some of the
19617 organization of your project, such as for instance the directory tree
19618 depth, whether some directories are found in a separate tree, @enddots{}
19620 The second method is the most general one. It requires an external
19621 program, called @command{find}, which is standard on all Unix systems. All
19622 the directories found under a given root directory will be added to the
19628 @font@heightrm=cmr8
19631 # The examples below are based on the following directory hierarchy:
19632 # All the directories can contain any number of files
19633 # ROOT_DIRECTORY -> a -> aa -> aaa
19636 # -> b -> ba -> baa
19639 # This Makefile creates a variable called DIRS, that can be reused any time
19640 # you need this list (see the other examples in this section)
19642 # The root of your project's directory hierarchy
19646 # First method: specify explicitly the list of directories
19647 # This allows you to specify any subset of all the directories you need.
19650 DIRS := a/aa/ a/ab/ b/ba/
19653 # Second method: use wildcards
19654 # Note that the argument(s) to wildcard below should end with a '/'.
19655 # Since wildcards also return file names, we have to filter them out
19656 # to avoid duplicate directory names.
19657 # We thus use make's @code{dir} and @code{sort} functions.
19658 # It sets DIRs to the following value (note that the directories aaa and baa
19659 # are not given, unless you change the arguments to wildcard).
19660 # DIRS= ./a/a/ ./b/ ./a/aa/ ./a/ab/ ./a/ac/ ./b/ba/ ./b/bb/ ./b/bc/
19663 DIRS := $@{sort $@{dir $@{wildcard $@{ROOT_DIRECTORY@}/*/
19664 $@{ROOT_DIRECTORY@}/*/*/@}@}@}
19667 # Third method: use an external program
19668 # This command is much faster if run on local disks, avoiding NFS slowdowns.
19669 # This is the most complete command: it sets DIRs to the following value:
19670 # DIRS= ./a ./a/aa ./a/aa/aaa ./a/ab ./a/ac ./b ./b/ba ./b/ba/baa ./b/bb ./b/bc
19673 DIRS := $@{shell find $@{ROOT_DIRECTORY@} -type d -print@}
19677 @node Generating the Command Line Switches
19678 @section Generating the Command Line Switches
19681 Once you have created the list of directories as explained in the
19682 previous section (@pxref{Automatically Creating a List of Directories}),
19683 you can easily generate the command line arguments to pass to gnatmake.
19685 For the sake of completeness, this example assumes that the source path
19686 is not the same as the object path, and that you have two separate lists
19690 # see "Automatically creating a list of directories" to create
19695 GNATMAKE_SWITCHES := $@{patsubst %,-aI%,$@{SOURCE_DIRS@}@}
19696 GNATMAKE_SWITCHES += $@{patsubst %,-aO%,$@{OBJECT_DIRS@}@}
19699 gnatmake $@{GNATMAKE_SWITCHES@} main_unit
19702 @node Overcoming Command Line Length Limits
19703 @section Overcoming Command Line Length Limits
19706 One problem that might be encountered on big projects is that many
19707 operating systems limit the length of the command line. It is thus hard to give
19708 gnatmake the list of source and object directories.
19710 This example shows how you can set up environment variables, which will
19711 make @command{gnatmake} behave exactly as if the directories had been
19712 specified on the command line, but have a much higher length limit (or
19713 even none on most systems).
19715 It assumes that you have created a list of directories in your Makefile,
19716 using one of the methods presented in
19717 @ref{Automatically Creating a List of Directories}.
19718 For the sake of completeness, we assume that the object
19719 path (where the ALI files are found) is different from the sources patch.
19721 Note a small trick in the Makefile below: for efficiency reasons, we
19722 create two temporary variables (SOURCE_LIST and OBJECT_LIST), that are
19723 expanded immediately by @code{make}. This way we overcome the standard
19724 make behavior which is to expand the variables only when they are
19727 On Windows, if you are using the standard Windows command shell, you must
19728 replace colons with semicolons in the assignments to these variables.
19733 @font@heightrm=cmr8
19736 # In this example, we create both ADA_INCLUDE_PATH and ADA_OBJECT_PATH.
19737 # This is the same thing as putting the -I arguments on the command line.
19738 # (the equivalent of using -aI on the command line would be to define
19739 # only ADA_INCLUDE_PATH, the equivalent of -aO is ADA_OBJECT_PATH).
19740 # You can of course have different values for these variables.
19742 # Note also that we need to keep the previous values of these variables, since
19743 # they might have been set before running 'make' to specify where the GNAT
19744 # library is installed.
19746 # see "Automatically creating a list of directories" to create these
19752 space:=$@{empty@} $@{empty@}
19753 SOURCE_LIST := $@{subst $@{space@},:,$@{SOURCE_DIRS@}@}
19754 OBJECT_LIST := $@{subst $@{space@},:,$@{OBJECT_DIRS@}@}
19755 ADA_INCLUDE_PATH += $@{SOURCE_LIST@}
19756 ADA_OBJECT_PATH += $@{OBJECT_LIST@}
19757 export ADA_INCLUDE_PATH
19758 export ADA_OBJECT_PATH
19765 @node Memory Management Issues
19766 @chapter Memory Management Issues
19769 This chapter describes some useful memory pools provided in the GNAT library
19770 and in particular the GNAT Debug Pool facility, which can be used to detect
19771 incorrect uses of access values (including ``dangling references'').
19773 It also describes the @command{gnatmem} tool, which can be used to track down
19778 * Some Useful Memory Pools::
19779 * The GNAT Debug Pool Facility::
19781 * The gnatmem Tool::
19785 @node Some Useful Memory Pools
19786 @section Some Useful Memory Pools
19787 @findex Memory Pool
19788 @cindex storage, pool
19791 The @code{System.Pool_Global} package offers the Unbounded_No_Reclaim_Pool
19792 storage pool. Allocations use the standard system call @code{malloc} while
19793 deallocations use the standard system call @code{free}. No reclamation is
19794 performed when the pool goes out of scope. For performance reasons, the
19795 standard default Ada allocators/deallocators do not use any explicit storage
19796 pools but if they did, they could use this storage pool without any change in
19797 behavior. That is why this storage pool is used when the user
19798 manages to make the default implicit allocator explicit as in this example:
19799 @smallexample @c ada
19800 type T1 is access Something;
19801 -- no Storage pool is defined for T2
19802 type T2 is access Something_Else;
19803 for T2'Storage_Pool use T1'Storage_Pool;
19804 -- the above is equivalent to
19805 for T2'Storage_Pool use System.Pool_Global.Global_Pool_Object;
19809 The @code{System.Pool_Local} package offers the Unbounded_Reclaim_Pool storage
19810 pool. The allocation strategy is similar to @code{Pool_Local}'s
19811 except that the all
19812 storage allocated with this pool is reclaimed when the pool object goes out of
19813 scope. This pool provides a explicit mechanism similar to the implicit one
19814 provided by several Ada 83 compilers for allocations performed through a local
19815 access type and whose purpose was to reclaim memory when exiting the
19816 scope of a given local access. As an example, the following program does not
19817 leak memory even though it does not perform explicit deallocation:
19819 @smallexample @c ada
19820 with System.Pool_Local;
19821 procedure Pooloc1 is
19822 procedure Internal is
19823 type A is access Integer;
19824 X : System.Pool_Local.Unbounded_Reclaim_Pool;
19825 for A'Storage_Pool use X;
19828 for I in 1 .. 50 loop
19833 for I in 1 .. 100 loop
19840 The @code{System.Pool_Size} package implements the Stack_Bounded_Pool used when
19841 @code{Storage_Size} is specified for an access type.
19842 The whole storage for the pool is
19843 allocated at once, usually on the stack at the point where the access type is
19844 elaborated. It is automatically reclaimed when exiting the scope where the
19845 access type is defined. This package is not intended to be used directly by the
19846 user and it is implicitly used for each such declaration:
19848 @smallexample @c ada
19849 type T1 is access Something;
19850 for T1'Storage_Size use 10_000;
19853 @node The GNAT Debug Pool Facility
19854 @section The GNAT Debug Pool Facility
19856 @cindex storage, pool, memory corruption
19859 The use of unchecked deallocation and unchecked conversion can easily
19860 lead to incorrect memory references. The problems generated by such
19861 references are usually difficult to tackle because the symptoms can be
19862 very remote from the origin of the problem. In such cases, it is
19863 very helpful to detect the problem as early as possible. This is the
19864 purpose of the Storage Pool provided by @code{GNAT.Debug_Pools}.
19866 In order to use the GNAT specific debugging pool, the user must
19867 associate a debug pool object with each of the access types that may be
19868 related to suspected memory problems. See Ada Reference Manual 13.11.
19869 @smallexample @c ada
19870 type Ptr is access Some_Type;
19871 Pool : GNAT.Debug_Pools.Debug_Pool;
19872 for Ptr'Storage_Pool use Pool;
19876 @code{GNAT.Debug_Pools} is derived from a GNAT-specific kind of
19877 pool: the @code{Checked_Pool}. Such pools, like standard Ada storage pools,
19878 allow the user to redefine allocation and deallocation strategies. They
19879 also provide a checkpoint for each dereference, through the use of
19880 the primitive operation @code{Dereference} which is implicitly called at
19881 each dereference of an access value.
19883 Once an access type has been associated with a debug pool, operations on
19884 values of the type may raise four distinct exceptions,
19885 which correspond to four potential kinds of memory corruption:
19888 @code{GNAT.Debug_Pools.Accessing_Not_Allocated_Storage}
19890 @code{GNAT.Debug_Pools.Accessing_Deallocated_Storage}
19892 @code{GNAT.Debug_Pools.Freeing_Not_Allocated_Storage}
19894 @code{GNAT.Debug_Pools.Freeing_Deallocated_Storage }
19898 For types associated with a Debug_Pool, dynamic allocation is performed using
19899 the standard GNAT allocation routine. References to all allocated chunks of
19900 memory are kept in an internal dictionary. Several deallocation strategies are
19901 provided, whereupon the user can choose to release the memory to the system,
19902 keep it allocated for further invalid access checks, or fill it with an easily
19903 recognizable pattern for debug sessions. The memory pattern is the old IBM
19904 hexadecimal convention: @code{16#DEADBEEF#}.
19906 See the documentation in the file g-debpoo.ads for more information on the
19907 various strategies.
19909 Upon each dereference, a check is made that the access value denotes a
19910 properly allocated memory location. Here is a complete example of use of
19911 @code{Debug_Pools}, that includes typical instances of memory corruption:
19912 @smallexample @c ada
19916 with Gnat.Io; use Gnat.Io;
19917 with Unchecked_Deallocation;
19918 with Unchecked_Conversion;
19919 with GNAT.Debug_Pools;
19920 with System.Storage_Elements;
19921 with Ada.Exceptions; use Ada.Exceptions;
19922 procedure Debug_Pool_Test is
19924 type T is access Integer;
19925 type U is access all T;
19927 P : GNAT.Debug_Pools.Debug_Pool;
19928 for T'Storage_Pool use P;
19930 procedure Free is new Unchecked_Deallocation (Integer, T);
19931 function UC is new Unchecked_Conversion (U, T);
19934 procedure Info is new GNAT.Debug_Pools.Print_Info(Put_Line);
19944 Put_Line (Integer'Image(B.all));
19946 when E : others => Put_Line ("raised: " & Exception_Name (E));
19951 when E : others => Put_Line ("raised: " & Exception_Name (E));
19955 Put_Line (Integer'Image(B.all));
19957 when E : others => Put_Line ("raised: " & Exception_Name (E));
19962 when E : others => Put_Line ("raised: " & Exception_Name (E));
19965 end Debug_Pool_Test;
19969 The debug pool mechanism provides the following precise diagnostics on the
19970 execution of this erroneous program:
19973 Total allocated bytes : 0
19974 Total deallocated bytes : 0
19975 Current Water Mark: 0
19979 Total allocated bytes : 8
19980 Total deallocated bytes : 0
19981 Current Water Mark: 8
19984 raised: GNAT.DEBUG_POOLS.ACCESSING_DEALLOCATED_STORAGE
19985 raised: GNAT.DEBUG_POOLS.FREEING_DEALLOCATED_STORAGE
19986 raised: GNAT.DEBUG_POOLS.ACCESSING_NOT_ALLOCATED_STORAGE
19987 raised: GNAT.DEBUG_POOLS.FREEING_NOT_ALLOCATED_STORAGE
19989 Total allocated bytes : 8
19990 Total deallocated bytes : 4
19991 Current Water Mark: 4
19996 @node The gnatmem Tool
19997 @section The @command{gnatmem} Tool
20001 The @code{gnatmem} utility monitors dynamic allocation and
20002 deallocation activity in a program, and displays information about
20003 incorrect deallocations and possible sources of memory leaks.
20004 It is designed to work in association with a static runtime library
20005 only and in this context provides three types of information:
20008 General information concerning memory management, such as the total
20009 number of allocations and deallocations, the amount of allocated
20010 memory and the high water mark, i.e.@: the largest amount of allocated
20011 memory in the course of program execution.
20014 Backtraces for all incorrect deallocations, that is to say deallocations
20015 which do not correspond to a valid allocation.
20018 Information on each allocation that is potentially the origin of a memory
20023 * Running gnatmem::
20024 * Switches for gnatmem::
20025 * Example of gnatmem Usage::
20028 @node Running gnatmem
20029 @subsection Running @code{gnatmem}
20032 @code{gnatmem} makes use of the output created by the special version of
20033 allocation and deallocation routines that record call information. This
20034 allows to obtain accurate dynamic memory usage history at a minimal cost to
20035 the execution speed. Note however, that @code{gnatmem} is not supported on
20036 all platforms (currently, it is supported on AIX, HP-UX, GNU/Linux,
20037 Solaris and Windows NT/2000/XP (x86).
20040 The @code{gnatmem} command has the form
20043 $ gnatmem @ovar{switches} user_program
20047 The program must have been linked with the instrumented version of the
20048 allocation and deallocation routines. This is done by linking with the
20049 @file{libgmem.a} library. For correct symbolic backtrace information,
20050 the user program should be compiled with debugging options
20051 (see @ref{Switches for gcc}). For example to build @file{my_program}:
20054 $ gnatmake -g my_program -largs -lgmem
20058 As library @file{libgmem.a} contains an alternate body for package
20059 @code{System.Memory}, @file{s-memory.adb} should not be compiled and linked
20060 when an executable is linked with library @file{libgmem.a}. It is then not
20061 recommended to use @command{gnatmake} with switch @option{^-a^/ALL_FILES^}.
20064 When @file{my_program} is executed, the file @file{gmem.out} is produced.
20065 This file contains information about all allocations and deallocations
20066 performed by the program. It is produced by the instrumented allocations and
20067 deallocations routines and will be used by @code{gnatmem}.
20069 In order to produce symbolic backtrace information for allocations and
20070 deallocations performed by the GNAT run-time library, you need to use a
20071 version of that library that has been compiled with the @option{-g} switch
20072 (see @ref{Rebuilding the GNAT Run-Time Library}).
20074 Gnatmem must be supplied with the @file{gmem.out} file and the executable to
20075 examine. If the location of @file{gmem.out} file was not explicitly supplied by
20076 @option{-i} switch, gnatmem will assume that this file can be found in the
20077 current directory. For example, after you have executed @file{my_program},
20078 @file{gmem.out} can be analyzed by @code{gnatmem} using the command:
20081 $ gnatmem my_program
20085 This will produce the output with the following format:
20087 *************** debut cc
20089 $ gnatmem my_program
20093 Total number of allocations : 45
20094 Total number of deallocations : 6
20095 Final Water Mark (non freed mem) : 11.29 Kilobytes
20096 High Water Mark : 11.40 Kilobytes
20101 Allocation Root # 2
20102 -------------------
20103 Number of non freed allocations : 11
20104 Final Water Mark (non freed mem) : 1.16 Kilobytes
20105 High Water Mark : 1.27 Kilobytes
20107 my_program.adb:23 my_program.alloc
20113 The first block of output gives general information. In this case, the
20114 Ada construct ``@code{@b{new}}'' was executed 45 times, and only 6 calls to an
20115 Unchecked_Deallocation routine occurred.
20118 Subsequent paragraphs display information on all allocation roots.
20119 An allocation root is a specific point in the execution of the program
20120 that generates some dynamic allocation, such as a ``@code{@b{new}}''
20121 construct. This root is represented by an execution backtrace (or subprogram
20122 call stack). By default the backtrace depth for allocations roots is 1, so
20123 that a root corresponds exactly to a source location. The backtrace can
20124 be made deeper, to make the root more specific.
20126 @node Switches for gnatmem
20127 @subsection Switches for @code{gnatmem}
20130 @code{gnatmem} recognizes the following switches:
20135 @cindex @option{-q} (@code{gnatmem})
20136 Quiet. Gives the minimum output needed to identify the origin of the
20137 memory leaks. Omits statistical information.
20140 @cindex @var{N} (@code{gnatmem})
20141 N is an integer literal (usually between 1 and 10) which controls the
20142 depth of the backtraces defining allocation root. The default value for
20143 N is 1. The deeper the backtrace, the more precise the localization of
20144 the root. Note that the total number of roots can depend on this
20145 parameter. This parameter must be specified @emph{before} the name of the
20146 executable to be analyzed, to avoid ambiguity.
20149 @cindex @option{-b} (@code{gnatmem})
20150 This switch has the same effect as just depth parameter.
20152 @item -i @var{file}
20153 @cindex @option{-i} (@code{gnatmem})
20154 Do the @code{gnatmem} processing starting from @file{file}, rather than
20155 @file{gmem.out} in the current directory.
20158 @cindex @option{-m} (@code{gnatmem})
20159 This switch causes @code{gnatmem} to mask the allocation roots that have less
20160 than n leaks. The default value is 1. Specifying the value of 0 will allow to
20161 examine even the roots that didn't result in leaks.
20164 @cindex @option{-s} (@code{gnatmem})
20165 This switch causes @code{gnatmem} to sort the allocation roots according to the
20166 specified order of sort criteria, each identified by a single letter. The
20167 currently supported criteria are @code{n, h, w} standing respectively for
20168 number of unfreed allocations, high watermark, and final watermark
20169 corresponding to a specific root. The default order is @code{nwh}.
20173 @node Example of gnatmem Usage
20174 @subsection Example of @code{gnatmem} Usage
20177 The following example shows the use of @code{gnatmem}
20178 on a simple memory-leaking program.
20179 Suppose that we have the following Ada program:
20181 @smallexample @c ada
20184 with Unchecked_Deallocation;
20185 procedure Test_Gm is
20187 type T is array (1..1000) of Integer;
20188 type Ptr is access T;
20189 procedure Free is new Unchecked_Deallocation (T, Ptr);
20192 procedure My_Alloc is
20197 procedure My_DeAlloc is
20205 for I in 1 .. 5 loop
20206 for J in I .. 5 loop
20217 The program needs to be compiled with debugging option and linked with
20218 @code{gmem} library:
20221 $ gnatmake -g test_gm -largs -lgmem
20225 Then we execute the program as usual:
20232 Then @code{gnatmem} is invoked simply with
20238 which produces the following output (result may vary on different platforms):
20243 Total number of allocations : 18
20244 Total number of deallocations : 5
20245 Final Water Mark (non freed mem) : 53.00 Kilobytes
20246 High Water Mark : 56.90 Kilobytes
20248 Allocation Root # 1
20249 -------------------
20250 Number of non freed allocations : 11
20251 Final Water Mark (non freed mem) : 42.97 Kilobytes
20252 High Water Mark : 46.88 Kilobytes
20254 test_gm.adb:11 test_gm.my_alloc
20256 Allocation Root # 2
20257 -------------------
20258 Number of non freed allocations : 1
20259 Final Water Mark (non freed mem) : 10.02 Kilobytes
20260 High Water Mark : 10.02 Kilobytes
20262 s-secsta.adb:81 system.secondary_stack.ss_init
20264 Allocation Root # 3
20265 -------------------
20266 Number of non freed allocations : 1
20267 Final Water Mark (non freed mem) : 12 Bytes
20268 High Water Mark : 12 Bytes
20270 s-secsta.adb:181 system.secondary_stack.ss_init
20274 Note that the GNAT run time contains itself a certain number of
20275 allocations that have no corresponding deallocation,
20276 as shown here for root #2 and root
20277 #3. This is a normal behavior when the number of non-freed allocations
20278 is one, it allocates dynamic data structures that the run time needs for
20279 the complete lifetime of the program. Note also that there is only one
20280 allocation root in the user program with a single line back trace:
20281 test_gm.adb:11 test_gm.my_alloc, whereas a careful analysis of the
20282 program shows that 'My_Alloc' is called at 2 different points in the
20283 source (line 21 and line 24). If those two allocation roots need to be
20284 distinguished, the backtrace depth parameter can be used:
20287 $ gnatmem 3 test_gm
20291 which will give the following output:
20296 Total number of allocations : 18
20297 Total number of deallocations : 5
20298 Final Water Mark (non freed mem) : 53.00 Kilobytes
20299 High Water Mark : 56.90 Kilobytes
20301 Allocation Root # 1
20302 -------------------
20303 Number of non freed allocations : 10
20304 Final Water Mark (non freed mem) : 39.06 Kilobytes
20305 High Water Mark : 42.97 Kilobytes
20307 test_gm.adb:11 test_gm.my_alloc
20308 test_gm.adb:24 test_gm
20309 b_test_gm.c:52 main
20311 Allocation Root # 2
20312 -------------------
20313 Number of non freed allocations : 1
20314 Final Water Mark (non freed mem) : 10.02 Kilobytes
20315 High Water Mark : 10.02 Kilobytes
20317 s-secsta.adb:81 system.secondary_stack.ss_init
20318 s-secsta.adb:283 <system__secondary_stack___elabb>
20319 b_test_gm.c:33 adainit
20321 Allocation Root # 3
20322 -------------------
20323 Number of non freed allocations : 1
20324 Final Water Mark (non freed mem) : 3.91 Kilobytes
20325 High Water Mark : 3.91 Kilobytes
20327 test_gm.adb:11 test_gm.my_alloc
20328 test_gm.adb:21 test_gm
20329 b_test_gm.c:52 main
20331 Allocation Root # 4
20332 -------------------
20333 Number of non freed allocations : 1
20334 Final Water Mark (non freed mem) : 12 Bytes
20335 High Water Mark : 12 Bytes
20337 s-secsta.adb:181 system.secondary_stack.ss_init
20338 s-secsta.adb:283 <system__secondary_stack___elabb>
20339 b_test_gm.c:33 adainit
20343 The allocation root #1 of the first example has been split in 2 roots #1
20344 and #3 thanks to the more precise associated backtrace.
20348 @node Stack Related Facilities
20349 @chapter Stack Related Facilities
20352 This chapter describes some useful tools associated with stack
20353 checking and analysis. In
20354 particular, it deals with dynamic and static stack usage measurements.
20357 * Stack Overflow Checking::
20358 * Static Stack Usage Analysis::
20359 * Dynamic Stack Usage Analysis::
20362 @node Stack Overflow Checking
20363 @section Stack Overflow Checking
20364 @cindex Stack Overflow Checking
20365 @cindex -fstack-check
20368 For most operating systems, @command{gcc} does not perform stack overflow
20369 checking by default. This means that if the main environment task or
20370 some other task exceeds the available stack space, then unpredictable
20371 behavior will occur. Most native systems offer some level of protection by
20372 adding a guard page at the end of each task stack. This mechanism is usually
20373 not enough for dealing properly with stack overflow situations because
20374 a large local variable could ``jump'' above the guard page.
20375 Furthermore, when the
20376 guard page is hit, there may not be any space left on the stack for executing
20377 the exception propagation code. Enabling stack checking avoids
20380 To activate stack checking, compile all units with the gcc option
20381 @option{-fstack-check}. For example:
20384 gcc -c -fstack-check package1.adb
20388 Units compiled with this option will generate extra instructions to check
20389 that any use of the stack (for procedure calls or for declaring local
20390 variables in declare blocks) does not exceed the available stack space.
20391 If the space is exceeded, then a @code{Storage_Error} exception is raised.
20393 For declared tasks, the stack size is controlled by the size
20394 given in an applicable @code{Storage_Size} pragma or by the value specified
20395 at bind time with @option{-d} (@pxref{Switches for gnatbind}) or is set to
20396 the default size as defined in the GNAT runtime otherwise.
20398 For the environment task, the stack size depends on
20399 system defaults and is unknown to the compiler. Stack checking
20400 may still work correctly if a fixed
20401 size stack is allocated, but this cannot be guaranteed.
20403 To ensure that a clean exception is signalled for stack
20404 overflow, set the environment variable
20405 @env{GNAT_STACK_LIMIT} to indicate the maximum
20406 stack area that can be used, as in:
20407 @cindex GNAT_STACK_LIMIT
20410 SET GNAT_STACK_LIMIT 1600
20414 The limit is given in kilobytes, so the above declaration would
20415 set the stack limit of the environment task to 1.6 megabytes.
20416 Note that the only purpose of this usage is to limit the amount
20417 of stack used by the environment task. If it is necessary to
20418 increase the amount of stack for the environment task, then this
20419 is an operating systems issue, and must be addressed with the
20420 appropriate operating systems commands.
20423 To have a fixed size stack in the environment task, the stack must be put
20424 in the P0 address space and its size specified. Use these switches to
20428 gnatmake my_progs -largs "-Wl,--opt=STACK=4000,/p0image"
20432 The quotes are required to keep case. The number after @samp{STACK=} is the
20433 size of the environmental task stack in pagelets (512 bytes). In this example
20434 the stack size is about 2 megabytes.
20437 A consequence of the @option{/p0image} qualifier is also to makes RMS buffers
20438 be placed in P0 space. Refer to @cite{HP OpenVMS Linker Utility Manual} for
20439 more details about the @option{/p0image} qualifier and the @option{stack}
20443 @node Static Stack Usage Analysis
20444 @section Static Stack Usage Analysis
20445 @cindex Static Stack Usage Analysis
20446 @cindex -fstack-usage
20449 A unit compiled with @option{-fstack-usage} will generate an extra file
20451 the maximum amount of stack used, on a per-function basis.
20452 The file has the same
20453 basename as the target object file with a @file{.su} extension.
20454 Each line of this file is made up of three fields:
20458 The name of the function.
20462 One or more qualifiers: @code{static}, @code{dynamic}, @code{bounded}.
20465 The second field corresponds to the size of the known part of the function
20468 The qualifier @code{static} means that the function frame size
20470 It usually means that all local variables have a static size.
20471 In this case, the second field is a reliable measure of the function stack
20474 The qualifier @code{dynamic} means that the function frame size is not static.
20475 It happens mainly when some local variables have a dynamic size. When this
20476 qualifier appears alone, the second field is not a reliable measure
20477 of the function stack analysis. When it is qualified with @code{bounded}, it
20478 means that the second field is a reliable maximum of the function stack
20481 @node Dynamic Stack Usage Analysis
20482 @section Dynamic Stack Usage Analysis
20485 It is possible to measure the maximum amount of stack used by a task, by
20486 adding a switch to @command{gnatbind}, as:
20489 $ gnatbind -u0 file
20493 With this option, at each task termination, its stack usage is output on
20495 It is not always convenient to output the stack usage when the program
20496 is still running. Hence, it is possible to delay this output until program
20497 termination. for a given number of tasks specified as the argument of the
20498 @option{-u} option. For instance:
20501 $ gnatbind -u100 file
20505 will buffer the stack usage information of the first 100 tasks to terminate and
20506 output this info at program termination. Results are displayed in four
20510 Index | Task Name | Stack Size | Stack Usage [Value +/- Variation]
20517 is a number associated with each task.
20520 is the name of the task analyzed.
20523 is the maximum size for the stack.
20526 is the measure done by the stack analyzer. In order to prevent overflow, the stack
20527 is not entirely analyzed, and it's not possible to know exactly how
20528 much has actually been used. The report thus contains the theoretical stack usage
20529 (Value) and the possible variation (Variation) around this value.
20534 The environment task stack, e.g., the stack that contains the main unit, is
20535 only processed when the environment variable GNAT_STACK_LIMIT is set.
20538 @c *********************************
20540 @c *********************************
20541 @node Verifying Properties Using gnatcheck
20542 @chapter Verifying Properties Using @command{gnatcheck}
20544 @cindex @command{gnatcheck}
20547 The @command{gnatcheck} tool is an ASIS-based utility that checks properties
20548 of Ada source files according to a given set of semantic rules.
20551 In order to check compliance with a given rule, @command{gnatcheck} has to
20552 semantically analyze the Ada sources.
20553 Therefore, checks can only be performed on
20554 legal Ada units. Moreover, when a unit depends semantically upon units located
20555 outside the current directory, the source search path has to be provided when
20556 calling @command{gnatcheck}, either through a specified project file or
20557 through @command{gnatcheck} switches as described below.
20559 A number of rules are predefined in @command{gnatcheck} and are described
20560 later in this chapter.
20561 You can also add new rules, by modifying the @command{gnatcheck} code and
20562 rebuilding the tool. In order to add a simple rule making some local checks,
20563 a small amount of straightforward ASIS-based programming is usually needed.
20565 Project support for @command{gnatcheck} is provided by the GNAT
20566 driver (see @ref{The GNAT Driver and Project Files}).
20568 Invoking @command{gnatcheck} on the command line has the form:
20571 $ gnatcheck @ovar{switches} @{@var{filename}@}
20572 @r{[}^-files^/FILES^=@{@var{arg_list_filename}@}@r{]}
20573 @r{[}-cargs @var{gcc_switches}@r{]} @r{[}-rules @var{rule_options}@r{]}
20580 @var{switches} specify the general tool options
20583 Each @var{filename} is the name (including the extension) of a source
20584 file to process. ``Wildcards'' are allowed, and
20585 the file name may contain path information.
20588 Each @var{arg_list_filename} is the name (including the extension) of a text
20589 file containing the names of the source files to process, separated by spaces
20593 @var{gcc_switches} is a list of switches for
20594 @command{gcc}. They will be passed on to all compiler invocations made by
20595 @command{gnatcheck} to generate the ASIS trees. Here you can provide
20596 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
20597 and use the @option{-gnatec} switch to set the configuration file.
20600 @var{rule_options} is a list of options for controlling a set of
20601 rules to be checked by @command{gnatcheck} (@pxref{gnatcheck Rule Options}).
20605 Either a @file{@var{filename}} or an @file{@var{arg_list_filename}} must be supplied.
20608 * Format of the Report File::
20609 * General gnatcheck Switches::
20610 * gnatcheck Rule Options::
20611 * Adding the Results of Compiler Checks to gnatcheck Output::
20612 * Project-Wide Checks::
20613 * Predefined Rules::
20616 @node Format of the Report File
20617 @section Format of the Report File
20618 @cindex Report file (for @code{gnatcheck})
20621 The @command{gnatcheck} tool outputs on @file{stdout} all messages concerning
20623 It also creates a text file that
20624 contains the complete report of the last gnatcheck run. By default this file is
20625 named named @file{^gnatcheck.out^GNATCHECK.OUT^} and it is located in the current
20626 directory, @option{^-o^/OUTPUT^} option can be used to change the name and/or
20627 location of the report file. This report contains:
20629 @item a list of the Ada source files being checked,
20630 @item a list of enabled and disabled rules,
20631 @item a list of the diagnostic messages, ordered in three different ways
20632 and collected in three separate
20633 sections. Section 1 contains the raw list of diagnostic messages. It
20634 corresponds to the output going to @file{stdout}. Section 2 contains
20635 messages ordered by rules.
20636 Section 3 contains messages ordered by source files.
20639 @node General gnatcheck Switches
20640 @section General @command{gnatcheck} Switches
20643 The following switches control the general @command{gnatcheck} behavior
20647 @cindex @option{^-a^/ALL^} (@command{gnatcheck})
20649 Process all units including those with read-only ALI files such as
20650 those from GNAT Run-Time library.
20654 @cindex @option{-d} (@command{gnatcheck})
20659 @cindex @option{-dd} (@command{gnatcheck})
20661 Progress indicator mode (for use in GPS)
20664 @cindex @option{^-h^/HELP^} (@command{gnatcheck})
20666 List the predefined and user-defined rules. For more details see
20667 @ref{Predefined Rules}.
20669 @cindex @option{^-l^/LOCS^} (@command{gnatcheck})
20671 Use full source locations references in the report file. For a construct from
20672 a generic instantiation a full source location is a chain from the location
20673 of this construct in the generic unit to the place where this unit is
20676 @cindex @option{^-m^/DIAGNOSTIC_LIMIT^} (@command{gnatcheck})
20677 @item ^-m@i{nnn}^/DIAGNOSTIC_LIMIT=@i{nnn}^
20678 Maximum number of diagnoses to be sent to Stdout, @i{nnn} from o@dots{}1000,
20679 the default value is 500. Zero means that there is no limitation on
20680 the number of diagnostic messages to be printed into Stdout.
20682 @cindex @option{^-q^/QUIET^} (@command{gnatcheck})
20684 Quiet mode. All the diagnoses about rule violations are placed in the
20685 @command{gnatcheck} report file only, without duplicating in @file{stdout}.
20687 @cindex @option{^-s^/SHORT^} (@command{gnatcheck})
20689 Short format of the report file (no version information, no list of applied
20690 rules, no list of checked sources is included)
20692 @cindex @option{^-s1^/COMPILER_STYLE^} (@command{gnatcheck})
20693 @item ^-s1^/COMPILER_STYLE^
20694 Include the compiler-style section in the report file
20696 @cindex @option{^-s2^/BY_RULES^} (@command{gnatcheck})
20697 @item ^-s2^/BY_RULES^
20698 Include the section containing diagnoses ordered by rules in the report file
20700 @cindex @option{^-s3^/BY_FILES_BY_RULES^} (@command{gnatcheck})
20701 @item ^-s3^/BY_FILES_BY_RULES^
20702 Include the section containing diagnoses ordered by files and then by rules
20705 @cindex @option{^-v^/VERBOSE^} (@command{gnatcheck})
20706 @item ^-v^/VERBOSE^
20707 Verbose mode; @command{gnatcheck} generates version information and then
20708 a trace of sources being processed.
20711 @cindex @option{^-o ^/OUTPUT^} (@command{gnatcheck})
20712 @item ^-o ^/OUTPUT=^@var{report_file}
20713 Set name of report file file to @var{report_file} .
20718 Note that if any of the options @option{^-s1^/COMPILER_STYLE^},
20719 @option{^-s2^/BY_RULES^} or
20720 @option{^-s3^/BY_FILES_BY_RULES^} is specified,
20721 then the @command{gnatcheck} report file will only contain sections
20722 explicitly denoted by these options.
20724 @node gnatcheck Rule Options
20725 @section @command{gnatcheck} Rule Options
20728 The following options control the processing performed by
20729 @command{gnatcheck}.
20732 @cindex @option{+ALL} (@command{gnatcheck})
20734 Turn all the rule checks ON.
20736 @cindex @option{-ALL} (@command{gnatcheck})
20738 Turn all the rule checks OFF.
20740 @cindex @option{+R} (@command{gnatcheck})
20741 @item +R@var{rule_id}@r{[}:@var{param}@r{]}
20742 Turn on the check for a specified rule with the specified parameter, if any.
20743 @var{rule_id} must be the identifier of one of the currently implemented rules
20744 (use @option{^-h^/HELP^} for the list of implemented rules). Rule identifiers
20745 are not case-sensitive. The @var{param} item must
20746 be a string representing a valid parameter(s) for the specified rule.
20747 If it contains any space characters then this string must be enclosed in
20750 @cindex @option{-R} (@command{gnatcheck})
20751 @item -R@var{rule_id}@r{[}:@var{param}@r{]}
20752 Turn off the check for a specified rule with the specified parameter, if any.
20754 @cindex @option{-from} (@command{gnatcheck})
20755 @item -from=@var{rule_option_filename}
20756 Read the rule options from the text file @var{rule_option_filename}, referred as
20757 ``rule file'' below.
20762 The default behavior is that all the rule checks are disabled.
20764 A rule file is a text file containing a set of rule options.
20765 @cindex Rule file (for @code{gnatcheck})
20766 The file may contain empty lines and Ada-style comments (comment
20767 lines and end-of-line comments). The rule file has free format; that is,
20768 you do not have to start a new rule option on a new line.
20770 A rule file may contain other @option{-from=@var{rule_option_filename}}
20771 options, each such option being replaced with the content of the
20772 corresponding rule file during the rule files processing. In case a
20773 cycle is detected (that is, @file{@var{rule_file_1}} reads rule options
20774 from @file{@var{rule_file_2}}, and @file{@var{rule_file_2}} reads
20775 (directly or indirectly) rule options from @file{@var{rule_file_1}}),
20776 the processing of rule files is interrupted and a part of their content
20780 @node Adding the Results of Compiler Checks to gnatcheck Output
20781 @section Adding the Results of Compiler Checks to @command{gnatcheck} Output
20784 The @command{gnatcheck} tool can include in the generated diagnostic messages
20786 the report file the results of the checks performed by the compiler. Though
20787 disabled by default, this effect may be obtained by using @option{+R} with
20788 the following rule identifiers and parameters:
20792 To record restrictions violations (that are performed by the compiler if the
20793 pragma @code{Restrictions} or @code{Restriction_Warnings} are given),
20795 @code{Restrictions} with the same parameters as pragma
20796 @code{Restrictions} or @code{Restriction_Warnings}.
20799 To record compiler style checks(@pxref{Style Checking}), use the rule named
20800 @code{Style_Checks}. A parameter of this rule can be either @code{All_Checks},
20801 which enables all the standard style checks that corresponds to @option{-gnatyy}
20802 GNAT style check option, or a string that has exactly the same
20803 structure and semantics as the @code{string_LITERAL} parameter of GNAT pragma
20804 @code{Style_Checks} (for further information about this pragma,
20805 @pxref{Pragma Style_Checks,,, gnat_rm, GNAT Reference Manual}).
20808 To record compiler warnings (@pxref{Warning Message Control}), use the rule
20809 named @code{Warnings} with a parameter that is a valid
20810 @i{static_string_expression} argument of GNAT pragma @code{Warnings}
20811 (for further information about this pragma, @pxref{Pragma Warnings,,,
20812 gnat_rm, GNAT Reference Manual}). Note, that in case of gnatcheck
20813 's' parameter, that corresponds to the GNAT @option{-gnatws} option, disables
20814 all the specific warnings, but not suppresses the warning mode,
20815 and 'e' parameter, corresponding to @option{-gnatwe} that means
20816 "treat warnings as errors", does not have any effect.
20820 To disable a specific restriction check, use @code{-RStyle_Checks} gnatcheck
20821 option with the corresponding restriction name as a parameter. @code{-R} is
20822 not available for @code{Style_Checks} and @code{Warnings} options, to disable
20823 warnings and style checks, use the corresponding warning and style options.
20825 @node Project-Wide Checks
20826 @section Project-Wide Checks
20827 @cindex Project-wide checks (for @command{gnatcheck})
20830 In order to perform checks on all units of a given project, you can use
20831 the GNAT driver along with the @option{-P} option:
20833 gnat check -Pproj -rules -from=my_rules
20837 If the project @code{proj} depends upon other projects, you can perform
20838 checks on the project closure using the @option{-U} option:
20840 gnat check -Pproj -U -rules -from=my_rules
20844 Finally, if not all the units are relevant to a particular main
20845 program in the project closure, you can perform checks for the set
20846 of units needed to create a given main program (unit closure) using
20847 the @option{-U} option followed by the name of the main unit:
20849 gnat check -Pproj -U main -rules -from=my_rules
20853 @node Predefined Rules
20854 @section Predefined Rules
20855 @cindex Predefined rules (for @command{gnatcheck})
20858 @c (Jan 2007) Since the global rules are still under development and are not
20859 @c documented, there is no point in explaining the difference between
20860 @c global and local rules
20862 A rule in @command{gnatcheck} is either local or global.
20863 A @emph{local rule} is a rule that applies to a well-defined section
20864 of a program and that can be checked by analyzing only this section.
20865 A @emph{global rule} requires analysis of some global properties of the
20866 whole program (mostly related to the program call graph).
20867 As of @value{NOW}, the implementation of global rules should be
20868 considered to be at a preliminary stage. You can use the
20869 @option{+GLOBAL} option to enable all the global rules, and the
20870 @option{-GLOBAL} rule option to disable all the global rules.
20872 All the global rules in the list below are
20873 so indicated by marking them ``GLOBAL''.
20874 This +GLOBAL and -GLOBAL options are not
20875 included in the list of gnatcheck options above, because at the moment they
20876 are considered as a temporary debug options.
20878 @command{gnatcheck} performs rule checks for generic
20879 instances only for global rules. This limitation may be relaxed in a later
20884 The following subsections document the rules implemented in
20885 @command{gnatcheck}.
20886 The subsection title is the same as the rule identifier, which may be
20887 used as a parameter of the @option{+R} or @option{-R} options.
20891 * Abstract_Type_Declarations::
20892 * Anonymous_Arrays::
20893 * Anonymous_Subtypes::
20895 * Boolean_Relational_Operators::
20897 * Ceiling_Violations::
20899 * Controlled_Type_Declarations::
20900 * Declarations_In_Blocks::
20901 * Default_Parameters::
20902 * Discriminated_Records::
20903 * Enumeration_Ranges_In_CASE_Statements::
20904 * Exceptions_As_Control_Flow::
20905 * Exits_From_Conditional_Loops::
20906 * EXIT_Statements_With_No_Loop_Name::
20907 * Expanded_Loop_Exit_Names::
20908 * Explicit_Full_Discrete_Ranges::
20909 * Float_Equality_Checks::
20910 * Forbidden_Pragmas::
20911 * Function_Style_Procedures::
20912 * Generics_In_Subprograms::
20913 * GOTO_Statements::
20914 * Implicit_IN_Mode_Parameters::
20915 * Implicit_SMALL_For_Fixed_Point_Types::
20916 * Improperly_Located_Instantiations::
20917 * Improper_Returns::
20918 * Library_Level_Subprograms::
20921 * Improperly_Called_Protected_Entries::
20924 * Misnamed_Identifiers::
20925 * Multiple_Entries_In_Protected_Definitions::
20927 * Non_Qualified_Aggregates::
20928 * Non_Short_Circuit_Operators::
20929 * Non_SPARK_Attributes::
20930 * Non_Tagged_Derived_Types::
20931 * Non_Visible_Exceptions::
20932 * Numeric_Literals::
20933 * OTHERS_In_Aggregates::
20934 * OTHERS_In_CASE_Statements::
20935 * OTHERS_In_Exception_Handlers::
20936 * Outer_Loop_Exits::
20937 * Overloaded_Operators::
20938 * Overly_Nested_Control_Structures::
20939 * Parameters_Out_Of_Order::
20940 * Positional_Actuals_For_Defaulted_Generic_Parameters::
20941 * Positional_Actuals_For_Defaulted_Parameters::
20942 * Positional_Components::
20943 * Positional_Generic_Parameters::
20944 * Positional_Parameters::
20945 * Predefined_Numeric_Types::
20946 * Raising_External_Exceptions::
20947 * Raising_Predefined_Exceptions::
20948 * Separate_Numeric_Error_Handlers::
20951 * Side_Effect_Functions::
20954 * Unassigned_OUT_Parameters::
20955 * Uncommented_BEGIN_In_Package_Bodies::
20956 * Unconditional_Exits::
20957 * Unconstrained_Array_Returns::
20958 * Universal_Ranges::
20959 * Unnamed_Blocks_And_Loops::
20961 * Unused_Subprograms::
20963 * USE_PACKAGE_Clauses::
20964 * Volatile_Objects_Without_Address_Clauses::
20968 @node Abstract_Type_Declarations
20969 @subsection @code{Abstract_Type_Declarations}
20970 @cindex @code{Abstract_Type_Declarations} rule (for @command{gnatcheck})
20973 Flag all declarations of abstract types. For an abstract private
20974 type, both the private and full type declarations are flagged.
20976 This rule has no parameters.
20979 @node Anonymous_Arrays
20980 @subsection @code{Anonymous_Arrays}
20981 @cindex @code{Anonymous_Arrays} rule (for @command{gnatcheck})
20984 Flag all anonymous array type definitions (by Ada semantics these can only
20985 occur in object declarations).
20987 This rule has no parameters.
20989 @node Anonymous_Subtypes
20990 @subsection @code{Anonymous_Subtypes}
20991 @cindex @code{Anonymous_Subtypes} rule (for @command{gnatcheck})
20994 Flag all uses of anonymous subtypes. A use of an anonymous subtype is
20995 any instance of a subtype indication with a constraint, other than one
20996 that occurs immediately within a subtype declaration. Any use of a range
20997 other than as a constraint used immediately within a subtype declaration
20998 is considered as an anonymous subtype.
21000 An effect of this rule is that @code{for} loops such as the following are
21001 flagged (since @code{1..N} is formally a ``range''):
21003 @smallexample @c ada
21004 for I in 1 .. N loop
21010 Declaring an explicit subtype solves the problem:
21012 @smallexample @c ada
21013 subtype S is Integer range 1..N;
21021 This rule has no parameters.
21024 @subsection @code{Blocks}
21025 @cindex @code{Blocks} rule (for @command{gnatcheck})
21028 Flag each block statement.
21030 This rule has no parameters.
21032 @node Boolean_Relational_Operators
21033 @subsection @code{Boolean_Relational_Operators}
21034 @cindex @code{Boolean_Relational_Operators} rule (for @command{gnatcheck})
21037 Flag each call to a predefined relational operator (``<'', ``>'', ``<='',
21038 ``>='', ``='' and ``/='') for the predefined Boolean type.
21039 (This rule is useful in enforcing the SPARK language restrictions.)
21041 Calls to predefined relational operators of any type derived from
21042 @code{Standard.Boolean} are not detected. Calls to user-defined functions
21043 with these designators, and uses of operators that are renamings
21044 of the predefined relational operators for @code{Standard.Boolean},
21045 are likewise not detected.
21047 This rule has no parameters.
21050 @node Ceiling_Violations
21051 @subsection @code{Ceiling_Violations} (under construction, GLOBAL)
21052 @cindex @code{Ceiling_Violations} rule (for @command{gnatcheck})
21055 Flag invocations of a protected operation by a task whose priority exceeds
21056 the protected object's ceiling.
21058 As of @value{NOW}, this rule has the following limitations:
21063 We consider only pragmas Priority and Interrupt_Priority as means to define
21064 a task/protected operation priority. We do not consider the effect of using
21065 Ada.Dynamic_Priorities.Set_Priority procedure;
21068 We consider only base task priorities, and no priority inheritance. That is,
21069 we do not make a difference between calls issued during task activation and
21070 execution of the sequence of statements from task body;
21073 Any situation when the priority of protected operation caller is set by a
21074 dynamic expression (that is, the corresponding Priority or
21075 Interrupt_Priority pragma has a non-static expression as an argument) we
21076 treat as a priority inconsistency (and, therefore, detect this situation).
21080 At the moment the notion of the main subprogram is not implemented in
21081 gnatcheck, so any pragma Priority in a library level subprogram body (in case
21082 if this subprogram can be a main subprogram of a partition) changes the
21083 priority of an environment task. So if we have more then one such pragma in
21084 the set of processed sources, the pragma that is processed last, defines the
21085 priority of an environment task.
21087 This rule has no parameters.
21090 @node Controlled_Type_Declarations
21091 @subsection @code{Controlled_Type_Declarations}
21092 @cindex @code{Controlled_Type_Declarations} rule (for @command{gnatcheck})
21095 Flag all declarations of controlled types. A declaration of a private type
21096 is flagged if its full declaration declares a controlled type. A declaration
21097 of a derived type is flagged if its ancestor type is controlled. Subtype
21098 declarations are not checked. A declaration of a type that itself is not a
21099 descendant of a type declared in @code{Ada.Finalization} but has a controlled
21100 component is not checked.
21102 This rule has no parameters.
21106 @node Declarations_In_Blocks
21107 @subsection @code{Declarations_In_Blocks}
21108 @cindex @code{Declarations_In_Blocks} rule (for @command{gnatcheck})
21111 Flag all block statements containing local declarations. A @code{declare}
21112 block with an empty @i{declarative_part} or with a @i{declarative part}
21113 containing only pragmas and/or @code{use} clauses is not flagged.
21115 This rule has no parameters.
21118 @node Default_Parameters
21119 @subsection @code{Default_Parameters}
21120 @cindex @code{Default_Parameters} rule (for @command{gnatcheck})
21123 Flag all default expressions for subprogram parameters. Parameter
21124 declarations of formal and generic subprograms are also checked.
21126 This rule has no parameters.
21129 @node Discriminated_Records
21130 @subsection @code{Discriminated_Records}
21131 @cindex @code{Discriminated_Records} rule (for @command{gnatcheck})
21134 Flag all declarations of record types with discriminants. Only the
21135 declarations of record and record extension types are checked. Incomplete,
21136 formal, private, derived and private extension type declarations are not
21137 checked. Task and protected type declarations also are not checked.
21139 This rule has no parameters.
21142 @node Enumeration_Ranges_In_CASE_Statements
21143 @subsection @code{Enumeration_Ranges_In_CASE_Statements}
21144 @cindex @code{Enumeration_Ranges_In_CASE_Statements} (for @command{gnatcheck})
21147 Flag each use of a range of enumeration literals as a choice in a
21148 @code{case} statement.
21149 All forms for specifying a range (explicit ranges
21150 such as @code{A .. B}, subtype marks and @code{'Range} attributes) are flagged.
21151 An enumeration range is
21152 flagged even if contains exactly one enumeration value or no values at all. A
21153 type derived from an enumeration type is considered as an enumeration type.
21155 This rule helps prevent maintenance problems arising from adding an
21156 enumeration value to a type and having it implicitly handled by an existing
21157 @code{case} statement with an enumeration range that includes the new literal.
21159 This rule has no parameters.
21162 @node Exceptions_As_Control_Flow
21163 @subsection @code{Exceptions_As_Control_Flow}
21164 @cindex @code{Exceptions_As_Control_Flow} (for @command{gnatcheck})
21167 Flag each place where an exception is explicitly raised and handled in the
21168 same subprogram body. A @code{raise} statement in an exception handler,
21169 package body, task body or entry body is not flagged.
21171 The rule has no parameters.
21173 @node Exits_From_Conditional_Loops
21174 @subsection @code{Exits_From_Conditional_Loops}
21175 @cindex @code{Exits_From_Conditional_Loops} (for @command{gnatcheck})
21178 Flag any exit statement if it transfers the control out of a @code{for} loop
21179 or a @code{while} loop. This includes cases when the @code{exit} statement
21180 applies to a @code{FOR} or @code{while} loop, and cases when it is enclosed
21181 in some @code{for} or @code{while} loop, but transfers the control from some
21182 outer (inconditional) @code{loop} statement.
21184 The rule has no parameters.
21187 @node EXIT_Statements_With_No_Loop_Name
21188 @subsection @code{EXIT_Statements_With_No_Loop_Name}
21189 @cindex @code{EXIT_Statements_With_No_Loop_Name} (for @command{gnatcheck})
21192 Flag each @code{exit} statement that does not specify the name of the loop
21195 The rule has no parameters.
21198 @node Expanded_Loop_Exit_Names
21199 @subsection @code{Expanded_Loop_Exit_Names}
21200 @cindex @code{Expanded_Loop_Exit_Names} rule (for @command{gnatcheck})
21203 Flag all expanded loop names in @code{exit} statements.
21205 This rule has no parameters.
21207 @node Explicit_Full_Discrete_Ranges
21208 @subsection @code{Explicit_Full_Discrete_Ranges}
21209 @cindex @code{Explicit_Full_Discrete_Ranges} rule (for @command{gnatcheck})
21212 Flag each discrete range that has the form @code{A'First .. A'Last}.
21214 This rule has no parameters.
21216 @node Float_Equality_Checks
21217 @subsection @code{Float_Equality_Checks}
21218 @cindex @code{Float_Equality_Checks} rule (for @command{gnatcheck})
21221 Flag all calls to the predefined equality operations for floating-point types.
21222 Both ``@code{=}'' and ``@code{/=}'' operations are checked.
21223 User-defined equality operations are not flagged, nor are ``@code{=}''
21224 and ``@code{/=}'' operations for fixed-point types.
21226 This rule has no parameters.
21229 @node Forbidden_Pragmas
21230 @subsection @code{Forbidden_Pragmas}
21231 @cindex @code{Forbidden_Pragmas} rule (for @command{gnatcheck})
21234 Flag each use of the specified pragmas. The pragmas to be detected
21235 are named in the rule's parameters.
21237 This rule has the following parameters:
21240 @item For the @option{+R} option
21243 @item @emph{Pragma_Name}
21244 Adds the specified pragma to the set of pragmas to be
21245 checked and sets the checks for all the specified pragmas
21246 ON. @emph{Pragma_Name} is treated as a name of a pragma. If it
21247 does not correspond to any pragma name defined in the Ada
21248 standard or to the name of a GNAT-specific pragma defined
21249 in @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference
21250 Manual}, it is treated as the name of unknown pragma.
21253 All the GNAT-specific pragmas are detected; this sets
21254 the checks for all the specified pragmas ON.
21257 All pragmas are detected; this sets the rule ON.
21260 @item For the @option{-R} option
21262 @item @emph{Pragma_Name}
21263 Removes the specified pragma from the set of pragmas to be
21264 checked without affecting checks for
21265 other pragmas. @emph{Pragma_Name} is treated as a name
21266 of a pragma. If it does not correspond to any pragma
21267 defined in the Ada standard or to any name defined in
21268 @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
21269 this option is treated as turning OFF detection of all unknown pragmas.
21272 Turn OFF detection of all GNAT-specific pragmas
21275 Clear the list of the pragmas to be detected and
21281 Parameters are not case sensitive. If @emph{Pragma_Name} does not have
21282 the syntax of an Ada identifier and therefore can not be considered
21283 as a pragma name, a diagnostic message is generated and the corresponding
21284 parameter is ignored.
21286 When more then one parameter is given in the same rule option, the parameters
21287 must be separated by a comma.
21289 If more then one option for this rule is specified for the @command{gnatcheck}
21290 call, a new option overrides the previous one(s).
21292 The @option{+R} option with no parameters turns the rule ON with the set of
21293 pragmas to be detected defined by the previous rule options.
21294 (By default this set is empty, so if the only option specified for the rule is
21295 @option{+RForbidden_Pragmas} (with
21296 no parameter), then the rule is enabled, but it does not detect anything).
21297 The @option{-R} option with no parameter turns the rule OFF, but it does not
21298 affect the set of pragmas to be detected.
21303 @node Function_Style_Procedures
21304 @subsection @code{Function_Style_Procedures}
21305 @cindex @code{Function_Style_Procedures} rule (for @command{gnatcheck})
21308 Flag each procedure that can be rewritten as a function. A procedure can be
21309 converted into a function if it has exactly one parameter of mode @code{out}
21310 and no parameters of mode @code{in out}. Procedure declarations,
21311 formal procedure declarations, and generic procedure declarations are always
21313 bodies and body stubs are flagged only if they do not have corresponding
21314 separate declarations. Procedure renamings and procedure instantiations are
21317 If a procedure can be rewritten as a function, but its @code{out} parameter is
21318 of a limited type, it is not flagged.
21320 Protected procedures are not flagged. Null procedures also are not flagged.
21322 This rule has no parameters.
21325 @node Generics_In_Subprograms
21326 @subsection @code{Generics_In_Subprograms}
21327 @cindex @code{Generics_In_Subprograms} rule (for @command{gnatcheck})
21330 Flag each declaration of a generic unit in a subprogram. Generic
21331 declarations in the bodies of generic subprograms are also flagged.
21332 A generic unit nested in another generic unit is not flagged.
21333 If a generic unit is
21334 declared in a local package that is declared in a subprogram body, the
21335 generic unit is flagged.
21337 This rule has no parameters.
21340 @node GOTO_Statements
21341 @subsection @code{GOTO_Statements}
21342 @cindex @code{GOTO_Statements} rule (for @command{gnatcheck})
21345 Flag each occurrence of a @code{goto} statement.
21347 This rule has no parameters.
21350 @node Implicit_IN_Mode_Parameters
21351 @subsection @code{Implicit_IN_Mode_Parameters}
21352 @cindex @code{Implicit_IN_Mode_Parameters} rule (for @command{gnatcheck})
21355 Flag each occurrence of a formal parameter with an implicit @code{in} mode.
21356 Note that @code{access} parameters, although they technically behave
21357 like @code{in} parameters, are not flagged.
21359 This rule has no parameters.
21362 @node Implicit_SMALL_For_Fixed_Point_Types
21363 @subsection @code{Implicit_SMALL_For_Fixed_Point_Types}
21364 @cindex @code{Implicit_SMALL_For_Fixed_Point_Types} rule (for @command{gnatcheck})
21367 Flag each fixed point type declaration that lacks an explicit
21368 representation clause to define its @code{'Small} value.
21369 Since @code{'Small} can be defined only for ordinary fixed point types,
21370 decimal fixed point type declarations are not checked.
21372 This rule has no parameters.
21375 @node Improperly_Located_Instantiations
21376 @subsection @code{Improperly_Located_Instantiations}
21377 @cindex @code{Improperly_Located_Instantiations} rule (for @command{gnatcheck})
21380 Flag all generic instantiations in library-level package specs
21381 (including library generic packages) and in all subprogram bodies.
21383 Instantiations in task and entry bodies are not flagged. Instantiations in the
21384 bodies of protected subprograms are flagged.
21386 This rule has no parameters.
21390 @node Improper_Returns
21391 @subsection @code{Improper_Returns}
21392 @cindex @code{Improper_Returns} rule (for @command{gnatcheck})
21395 Flag each explicit @code{return} statement in procedures, and
21396 multiple @code{return} statements in functions.
21397 Diagnostic messages are generated for all @code{return} statements
21398 in a procedure (thus each procedure must be written so that it
21399 returns implicitly at the end of its statement part),
21400 and for all @code{return} statements in a function after the first one.
21401 This rule supports the stylistic convention that each subprogram
21402 should have no more than one point of normal return.
21404 This rule has no parameters.
21407 @node Library_Level_Subprograms
21408 @subsection @code{Library_Level_Subprograms}
21409 @cindex @code{Library_Level_Subprograms} rule (for @command{gnatcheck})
21412 Flag all library-level subprograms (including generic subprogram instantiations).
21414 This rule has no parameters.
21417 @node Local_Packages
21418 @subsection @code{Local_Packages}
21419 @cindex @code{Local_Packages} rule (for @command{gnatcheck})
21422 Flag all local packages declared in package and generic package
21424 Local packages in bodies are not flagged.
21426 This rule has no parameters.
21429 @node Improperly_Called_Protected_Entries
21430 @subsection @code{Improperly_Called_Protected_Entries} (under construction, GLOBAL)
21431 @cindex @code{Improperly_Called_Protected_Entries} rule (for @command{gnatcheck})
21434 Flag each protected entry that can be called from more than one task.
21436 This rule has no parameters.
21440 @subsection @code{Metrics}
21441 @cindex @code{Metrics} rule (for @command{gnatcheck})
21444 There is a set of checks based on computing a metric value and comparing the
21445 result with the specified upper (or lower, depending on a specific metric)
21446 value specified for a given metric. A construct is flagged if a given metric
21447 is applicable (can be computed) for it and the computed value is greater
21448 then (lover then) the specified upper (lower) bound.
21450 The name of any metric-based rule consists of the prefix @code{Metrics_}
21451 followed by the name of the corresponding metric (see the table below).
21452 For @option{+R} option, each metric-based rule has a numeric parameter
21453 specifying the bound (integer or real, depending on a metric), @option{-R}
21454 option for metric rules does not have a parameter.
21456 The following table shows the metric names for that the corresponding
21457 metrics-based checks are supported by gnatcheck, including the
21458 constraint that must be satisfied by the bound that is specified for the check
21459 and what bound - upper (U) or lower (L) - should be specified.
21461 @multitable {@code{Cyclomatic_Complexity}}{Cyclomatic complexity}{Positive integer}
21463 @headitem Check Name @tab Description @tab Bounds Value
21466 @item @b{Check Name} @tab @b{Description} @tab @b{Bounds Value}
21468 @c Above conditional code is workaround to bug in texi2html (Feb 2008)
21469 @item @code{Essential_Complexity} @tab Essential complexity @tab Positive integer (U)
21470 @item @code{Cyclomatic_Complexity} @tab Cyclomatic complexity @tab Positive integer (U)
21471 @item @code{LSLOC} @tab Logical Source Lines of Code @tab Positive integer (U)
21475 The meaning and the computed values for all these metrics are exactly
21476 the same as for the corresponding metrics in @command{gnatmetric}.
21478 @emph{Example:} the rule
21480 +RMetrics_Cyclomatic_Complexity : 7
21483 means that all bodies with cyclomatic complexity exceeding 7 will be flagged.
21485 To turn OFF the check for cyclomatic complexity metric, use the following option:
21487 -RMetrics_Cyclomatic_Complexity
21490 @node Misnamed_Identifiers
21491 @subsection @code{Misnamed_Identifiers}
21492 @cindex @code{Misnamed_Identifiers} rule (for @command{gnatcheck})
21495 Flag the declaration of each identifier that does not have a suffix
21496 corresponding to the kind of entity being declared.
21497 The following declarations are checked:
21504 constant declarations (but not number declarations)
21507 package renaming declarations (but not generic package renaming
21512 This rule may have parameters. When used without parameters, the rule enforces
21513 the following checks:
21517 type-defining names end with @code{_T}, unless the type is an access type,
21518 in which case the suffix must be @code{_A}
21520 constant names end with @code{_C}
21522 names defining package renamings end with @code{_R}
21526 For a private or incomplete type declaration the following checks are
21527 made for the defining name suffix:
21531 For an incomplete type declaration: if the corresponding full type
21532 declaration is available, the defining identifier from the full type
21533 declaration is checked, but the defining identifier from the incomplete type
21534 declaration is not; otherwise the defining identifier from the incomplete
21535 type declaration is checked against the suffix specified for type
21539 For a private type declaration (including private extensions), the defining
21540 identifier from the private type declaration is checked against the type
21541 suffix (even if the corresponding full declaration is an access type
21542 declaration), and the defining identifier from the corresponding full type
21543 declaration is not checked.
21547 For a deferred constant, the defining name in the corresponding full constant
21548 declaration is not checked.
21550 Defining names of formal types are not checked.
21552 The rule may have the following parameters:
21556 For the @option{+R} option:
21559 Sets the default listed above for all the names to be checked.
21561 @item Type_Suffix=@emph{string}
21562 Specifies the suffix for a type name.
21564 @item Access_Suffix=@emph{string}
21565 Specifies the suffix for an access type name. If
21566 this parameter is set, it overrides for access
21567 types the suffix set by the @code{Type_Suffix} parameter.
21568 For access types, @emph{string} may have the following format:
21569 @emph{suffix1(suffix2)}. That means that an access type name
21570 should have the @emph{suffix1} suffix except for the case when
21571 the designated type is also an access type, in this case the
21572 type name should have the @emph{suffix1 & suffix2} suffix.
21574 @item Constant_Suffix=@emph{string}
21575 Specifies the suffix for a constant name.
21577 @item Renaming_Suffix=@emph{string}
21578 Specifies the suffix for a package renaming name.
21582 For the @option{-R} option:
21585 Remove all the suffixes specified for the
21586 identifier suffix checks, whether by default or
21587 as specified by other rule parameters. All the
21588 checks for this rule are disabled as a result.
21591 Removes the suffix specified for types. This
21592 disables checks for types but does not disable
21593 any other checks for this rule (including the
21594 check for access type names if @code{Access_Suffix} is
21597 @item Access_Suffix
21598 Removes the suffix specified for access types.
21599 This disables checks for access type names but
21600 does not disable any other checks for this rule.
21601 If @code{Type_Suffix} is set, access type names are
21602 checked as ordinary type names.
21604 @item Constant_Suffix
21605 Removes the suffix specified for constants. This
21606 disables checks for constant names but does not
21607 disable any other checks for this rule.
21609 @item Renaming_Suffix
21610 Removes the suffix specified for package
21611 renamings. This disables checks for package
21612 renamings but does not disable any other checks
21618 If more than one parameter is used, parameters must be separated by commas.
21620 If more than one option is specified for the @command{gnatcheck} invocation,
21621 a new option overrides the previous one(s).
21623 The @option{+RMisnamed_Identifiers} option (with no parameter) enables
21625 name suffixes specified by previous options used for this rule.
21627 The @option{-RMisnamed_Identifiers} option (with no parameter) disables
21628 all the checks but keeps
21629 all the suffixes specified by previous options used for this rule.
21631 The @emph{string} value must be a valid suffix for an Ada identifier (after
21632 trimming all the leading and trailing space characters, if any).
21633 Parameters are not case sensitive, except the @emph{string} part.
21635 If any error is detected in a rule parameter, the parameter is ignored.
21636 In such a case the options that are set for the rule are not
21641 @node Multiple_Entries_In_Protected_Definitions
21642 @subsection @code{Multiple_Entries_In_Protected_Definitions}
21643 @cindex @code{Multiple_Entries_In_Protected_Definitions} rule (for @command{gnatcheck})
21646 Flag each protected definition (i.e., each protected object/type declaration)
21647 that defines more than one entry.
21648 Diagnostic messages are generated for all the entry declarations
21649 except the first one. An entry family is counted as one entry. Entries from
21650 the private part of the protected definition are also checked.
21652 This rule has no parameters.
21655 @subsection @code{Name_Clashes}
21656 @cindex @code{Name_Clashes} rule (for @command{gnatcheck})
21659 Check that certain names are not used as defining identifiers. To activate
21660 this rule, you need to supply a reference to the dictionary file(s) as a rule
21661 parameter(s) (more then one dictionary file can be specified). If no
21662 dictionary file is set, this rule will not cause anything to be flagged.
21663 Only defining occurrences, not references, are checked.
21664 The check is not case-sensitive.
21666 This rule is enabled by default, but without setting any corresponding
21667 dictionary file(s); thus the default effect is to do no checks.
21669 A dictionary file is a plain text file. The maximum line length for this file
21670 is 1024 characters. If the line is longer then this limit, extra characters
21673 Each line can be either an empty line, a comment line, or a line containing
21674 a list of identifiers separated by space or HT characters.
21675 A comment is an Ada-style comment (from @code{--} to end-of-line).
21676 Identifiers must follow the Ada syntax for identifiers.
21677 A line containing one or more identifiers may end with a comment.
21679 @node Non_Qualified_Aggregates
21680 @subsection @code{Non_Qualified_Aggregates}
21681 @cindex @code{Non_Qualified_Aggregates} rule (for @command{gnatcheck})
21684 Flag each non-qualified aggregate.
21685 A non-qualified aggregate is an
21686 aggregate that is not the expression of a qualified expression. A
21687 string literal is not considered an aggregate, but an array
21688 aggregate of a string type is considered as a normal aggregate.
21689 Aggregates of anonymous array types are not flagged.
21691 This rule has no parameters.
21694 @node Non_Short_Circuit_Operators
21695 @subsection @code{Non_Short_Circuit_Operators}
21696 @cindex @code{Non_Short_Circuit_Operators} rule (for @command{gnatcheck})
21699 Flag all calls to predefined @code{and} and @code{or} operators for
21700 any boolean type. Calls to
21701 user-defined @code{and} and @code{or} and to operators defined by renaming
21702 declarations are not flagged. Calls to predefined @code{and} and @code{or}
21703 operators for modular types or boolean array types are not flagged.
21705 This rule has no parameters.
21709 @node Non_SPARK_Attributes
21710 @subsection @code{Non_SPARK_Attributes}
21711 @cindex @code{Non_SPARK_Attributes} rule (for @command{gnatcheck})
21714 The SPARK language defines the following subset of Ada 95 attribute
21715 designators as those that can be used in SPARK programs. The use of
21716 any other attribute is flagged.
21719 @item @code{'Adjacent}
21722 @item @code{'Ceiling}
21723 @item @code{'Component_Size}
21724 @item @code{'Compose}
21725 @item @code{'Copy_Sign}
21726 @item @code{'Delta}
21727 @item @code{'Denorm}
21728 @item @code{'Digits}
21729 @item @code{'Exponent}
21730 @item @code{'First}
21731 @item @code{'Floor}
21733 @item @code{'Fraction}
21735 @item @code{'Leading_Part}
21736 @item @code{'Length}
21737 @item @code{'Machine}
21738 @item @code{'Machine_Emax}
21739 @item @code{'Machine_Emin}
21740 @item @code{'Machine_Mantissa}
21741 @item @code{'Machine_Overflows}
21742 @item @code{'Machine_Radix}
21743 @item @code{'Machine_Rounds}
21746 @item @code{'Model}
21747 @item @code{'Model_Emin}
21748 @item @code{'Model_Epsilon}
21749 @item @code{'Model_Mantissa}
21750 @item @code{'Model_Small}
21751 @item @code{'Modulus}
21754 @item @code{'Range}
21755 @item @code{'Remainder}
21756 @item @code{'Rounding}
21757 @item @code{'Safe_First}
21758 @item @code{'Safe_Last}
21759 @item @code{'Scaling}
21760 @item @code{'Signed_Zeros}
21762 @item @code{'Small}
21764 @item @code{'Truncation}
21765 @item @code{'Unbiased_Rounding}
21767 @item @code{'Valid}
21771 This rule has no parameters.
21774 @node Non_Tagged_Derived_Types
21775 @subsection @code{Non_Tagged_Derived_Types}
21776 @cindex @code{Non_Tagged_Derived_Types} rule (for @command{gnatcheck})
21779 Flag all derived type declarations that do not have a record extension part.
21781 This rule has no parameters.
21785 @node Non_Visible_Exceptions
21786 @subsection @code{Non_Visible_Exceptions}
21787 @cindex @code{Non_Visible_Exceptions} rule (for @command{gnatcheck})
21790 Flag constructs leading to the possibility of propagating an exception
21791 out of the scope in which the exception is declared.
21792 Two cases are detected:
21796 An exception declaration in a subprogram body, task body or block
21797 statement is flagged if the body or statement does not contain a handler for
21798 that exception or a handler with an @code{others} choice.
21801 A @code{raise} statement in an exception handler of a subprogram body,
21802 task body or block statement is flagged if it (re)raises a locally
21803 declared exception. This may occur under the following circumstances:
21806 it explicitly raises a locally declared exception, or
21808 it does not specify an exception name (i.e., it is simply @code{raise;})
21809 and the enclosing handler contains a locally declared exception in its
21815 Renamings of local exceptions are not flagged.
21817 This rule has no parameters.
21820 @node Numeric_Literals
21821 @subsection @code{Numeric_Literals}
21822 @cindex @code{Numeric_Literals} rule (for @command{gnatcheck})
21825 Flag each use of a numeric literal in an index expression, and in any
21826 circumstance except for the following:
21830 a literal occurring in the initialization expression for a constant
21831 declaration or a named number declaration, or
21834 an integer literal that is less than or equal to a value
21835 specified by the @option{N} rule parameter.
21839 This rule may have the following parameters for the @option{+R} option:
21843 @emph{N} is an integer literal used as the maximal value that is not flagged
21844 (i.e., integer literals not exceeding this value are allowed)
21847 All integer literals are flagged
21851 If no parameters are set, the maximum unflagged value is 1.
21853 The last specified check limit (or the fact that there is no limit at
21854 all) is used when multiple @option{+R} options appear.
21856 The @option{-R} option for this rule has no parameters.
21857 It disables the rule but retains the last specified maximum unflagged value.
21858 If the @option{+R} option subsequently appears, this value is used as the
21859 threshold for the check.
21862 @node OTHERS_In_Aggregates
21863 @subsection @code{OTHERS_In_Aggregates}
21864 @cindex @code{OTHERS_In_Aggregates} rule (for @command{gnatcheck})
21867 Flag each use of an @code{others} choice in extension aggregates.
21868 In record and array aggregates, an @code{others} choice is flagged unless
21869 it is used to refer to all components, or to all but one component.
21871 If, in case of a named array aggregate, there are two associations, one
21872 with an @code{others} choice and another with a discrete range, the
21873 @code{others} choice is flagged even if the discrete range specifies
21874 exactly one component; for example, @code{(1..1 => 0, others => 1)}.
21876 This rule has no parameters.
21878 @node OTHERS_In_CASE_Statements
21879 @subsection @code{OTHERS_In_CASE_Statements}
21880 @cindex @code{OTHERS_In_CASE_Statements} rule (for @command{gnatcheck})
21883 Flag any use of an @code{others} choice in a @code{case} statement.
21885 This rule has no parameters.
21887 @node OTHERS_In_Exception_Handlers
21888 @subsection @code{OTHERS_In_Exception_Handlers}
21889 @cindex @code{OTHERS_In_Exception_Handlers} rule (for @command{gnatcheck})
21892 Flag any use of an @code{others} choice in an exception handler.
21894 This rule has no parameters.
21897 @node Outer_Loop_Exits
21898 @subsection @code{Outer_Loop_Exits}
21899 @cindex @code{Outer_Loop_Exits} rule (for @command{gnatcheck})
21902 Flag each @code{exit} statement containing a loop name that is not the name
21903 of the immediately enclosing @code{loop} statement.
21905 This rule has no parameters.
21908 @node Overloaded_Operators
21909 @subsection @code{Overloaded_Operators}
21910 @cindex @code{Overloaded_Operators} rule (for @command{gnatcheck})
21913 Flag each function declaration that overloads an operator symbol.
21914 A function body is checked only if the body does not have a
21915 separate spec. Formal functions are also checked. For a
21916 renaming declaration, only renaming-as-declaration is checked
21918 This rule has no parameters.
21921 @node Overly_Nested_Control_Structures
21922 @subsection @code{Overly_Nested_Control_Structures}
21923 @cindex @code{Overly_Nested_Control_Structures} rule (for @command{gnatcheck})
21926 Flag each control structure whose nesting level exceeds the value provided
21927 in the rule parameter.
21929 The control structures checked are the following:
21932 @item @code{if} statement
21933 @item @code{case} statement
21934 @item @code{loop} statement
21935 @item Selective accept statement
21936 @item Timed entry call statement
21937 @item Conditional entry call
21938 @item Asynchronous select statement
21942 The rule has the following parameter for the @option{+R} option:
21946 Positive integer specifying the maximal control structure nesting
21947 level that is not flagged
21951 If the parameter for the @option{+R} option is not specified or
21952 if it is not a positive integer, @option{+R} option is ignored.
21954 If more then one option is specified for the gnatcheck call, the later option and
21955 new parameter override the previous one(s).
21958 @node Parameters_Out_Of_Order
21959 @subsection @code{Parameters_Out_Of_Order}
21960 @cindex @code{Parameters_Out_Of_Order} rule (for @command{gnatcheck})
21963 Flag each subprogram and entry declaration whose formal parameters are not
21964 ordered according to the following scheme:
21968 @item @code{in} and @code{access} parameters first,
21969 then @code{in out} parameters,
21970 and then @code{out} parameters;
21972 @item for @code{in} mode, parameters with default initialization expressions
21977 Only the first violation of the described order is flagged.
21979 The following constructs are checked:
21982 @item subprogram declarations (including null procedures);
21983 @item generic subprogram declarations;
21984 @item formal subprogram declarations;
21985 @item entry declarations;
21986 @item subprogram bodies and subprogram body stubs that do not
21987 have separate specifications
21991 Subprogram renamings are not checked.
21993 This rule has no parameters.
21996 @node Positional_Actuals_For_Defaulted_Generic_Parameters
21997 @subsection @code{Positional_Actuals_For_Defaulted_Generic_Parameters}
21998 @cindex @code{Positional_Actuals_For_Defaulted_Generic_Parameters} rule (for @command{gnatcheck})
22001 Flag each generic actual parameter corresponding to a generic formal
22002 parameter with a default initialization, if positional notation is used.
22004 This rule has no parameters.
22006 @node Positional_Actuals_For_Defaulted_Parameters
22007 @subsection @code{Positional_Actuals_For_Defaulted_Parameters}
22008 @cindex @code{Positional_Actuals_For_Defaulted_Parameters} rule (for @command{gnatcheck})
22011 Flag each actual parameter to a subprogram or entry call where the
22012 corresponding formal parameter has a default expression, if positional
22015 This rule has no parameters.
22017 @node Positional_Components
22018 @subsection @code{Positional_Components}
22019 @cindex @code{Positional_Components} rule (for @command{gnatcheck})
22022 Flag each array, record and extension aggregate that includes positional
22025 This rule has no parameters.
22028 @node Positional_Generic_Parameters
22029 @subsection @code{Positional_Generic_Parameters}
22030 @cindex @code{Positional_Generic_Parameters} rule (for @command{gnatcheck})
22033 Flag each instantiation using positional parameter notation.
22035 This rule has no parameters.
22038 @node Positional_Parameters
22039 @subsection @code{Positional_Parameters}
22040 @cindex @code{Positional_Parameters} rule (for @command{gnatcheck})
22043 Flag each subprogram or entry call using positional parameter notation,
22044 except for the following:
22048 Invocations of prefix or infix operators are not flagged
22050 If the called subprogram or entry has only one formal parameter,
22051 the call is not flagged;
22053 If a subprogram call uses the @emph{Object.Operation} notation, then
22056 the first parameter (that is, @emph{Object}) is not flagged;
22058 if the called subprogram has only two parameters, the second parameter
22059 of the call is not flagged;
22064 This rule has no parameters.
22069 @node Predefined_Numeric_Types
22070 @subsection @code{Predefined_Numeric_Types}
22071 @cindex @code{Predefined_Numeric_Types} rule (for @command{gnatcheck})
22074 Flag each explicit use of the name of any numeric type or subtype defined
22075 in package @code{Standard}.
22077 The rationale for this rule is to detect when the
22078 program may depend on platform-specific characteristics of the implementation
22079 of the predefined numeric types. Note that this rule is over-pessimistic;
22080 for example, a program that uses @code{String} indexing
22081 likely needs a variable of type @code{Integer}.
22082 Another example is the flagging of predefined numeric types with explicit
22085 @smallexample @c ada
22086 subtype My_Integer is Integer range Left .. Right;
22087 Vy_Var : My_Integer;
22091 This rule detects only numeric types and subtypes defined in
22092 @code{Standard}. The use of numeric types and subtypes defined in other
22093 predefined packages (such as @code{System.Any_Priority} or
22094 @code{Ada.Text_IO.Count}) is not flagged
22096 This rule has no parameters.
22100 @node Raising_External_Exceptions
22101 @subsection @code{Raising_External_Exceptions}
22102 @cindex @code{Raising_External_Exceptions} rule (for @command{gnatcheck})
22105 Flag any @code{raise} statement, in a program unit declared in a library
22106 package or in a generic library package, for an exception that is
22107 neither a predefined exception nor an exception that is also declared (or
22108 renamed) in the visible part of the package.
22110 This rule has no parameters.
22114 @node Raising_Predefined_Exceptions
22115 @subsection @code{Raising_Predefined_Exceptions}
22116 @cindex @code{Raising_Predefined_Exceptions} rule (for @command{gnatcheck})
22119 Flag each @code{raise} statement that raises a predefined exception
22120 (i.e., one of the exceptions @code{Constraint_Error}, @code{Numeric_Error},
22121 @code{Program_Error}, @code{Storage_Error}, or @code{Tasking_Error}).
22123 This rule has no parameters.
22125 @node Separate_Numeric_Error_Handlers
22126 @subsection @code{Separate_Numeric_Error_Handlers}
22127 @cindex @code{Separate_Numeric_Error_Handlers} rule (for @command{gnatcheck})
22130 Flags each exception handler that contains a choice for
22131 the predefined @code{Constraint_Error} exception, but does not contain
22132 the choice for the predefined @code{Numeric_Error} exception, or
22133 that contains the choice for @code{Numeric_Error}, but does not contain the
22134 choice for @code{Constraint_Error}.
22136 This rule has no parameters.
22140 @subsection @code{Recursion} (under construction, GLOBAL)
22141 @cindex @code{Recursion} rule (for @command{gnatcheck})
22144 Flag recursive subprograms (cycles in the call graph). Declarations, and not
22145 calls, of recursive subprograms are detected.
22147 This rule has no parameters.
22151 @node Side_Effect_Functions
22152 @subsection @code{Side_Effect_Functions} (under construction, GLOBAL)
22153 @cindex @code{Side_Effect_Functions} rule (for @command{gnatcheck})
22156 Flag functions with side effects.
22158 We define a side effect as changing any data object that is not local for the
22159 body of this function.
22161 At the moment, we do NOT consider a side effect any input-output operations
22162 (changing a state or a content of any file).
22164 We do not consider protected functions for this rule (???)
22166 There are the following sources of side effect:
22169 @item Explicit (or direct) side-effect:
22173 direct assignment to a non-local variable;
22176 direct call to an entity that is known to change some data object that is
22177 not local for the body of this function (Note, that if F1 calls F2 and F2
22178 does have a side effect, this does not automatically mean that F1 also
22179 have a side effect, because it may be the case that F2 is declared in
22180 F1's body and it changes some data object that is global for F2, but
22184 @item Indirect side-effect:
22187 Subprogram calls implicitly issued by:
22190 computing initialization expressions from type declarations as a part
22191 of object elaboration or allocator evaluation;
22193 computing implicit parameters of subprogram or entry calls or generic
22198 activation of a task that change some non-local data object (directly or
22202 elaboration code of a package that is a result of a package instantiation;
22205 controlled objects;
22208 @item Situations when we can suspect a side-effect, but the full static check
22209 is either impossible or too hard:
22212 assignment to access variables or to the objects pointed by access
22216 call to a subprogram pointed by access-to-subprogram value
22224 This rule has no parameters.
22228 @subsection @code{Slices}
22229 @cindex @code{Slices} rule (for @command{gnatcheck})
22232 Flag all uses of array slicing
22234 This rule has no parameters.
22237 @node Unassigned_OUT_Parameters
22238 @subsection @code{Unassigned_OUT_Parameters}
22239 @cindex @code{Unassigned_OUT_Parameters} rule (for @command{gnatcheck})
22242 Flags procedures' @code{out} parameters that are not assigned, and
22243 identifies the contexts in which the assignments are missing.
22245 An @code{out} parameter is flagged in the statements in the procedure
22246 body's handled sequence of statements (before the procedure body's
22247 @code{exception} part, if any) if this sequence of statements contains
22248 no assignments to the parameter.
22250 An @code{out} parameter is flagged in an exception handler in the exception
22251 part of the procedure body's handled sequence of statements if the handler
22252 contains no assignment to the parameter.
22254 Bodies of generic procedures are also considered.
22256 The following are treated as assignments to an @code{out} parameter:
22260 an assignment statement, with the parameter or some component as the target;
22263 passing the parameter (or one of its components) as an @code{out} or
22264 @code{in out} parameter.
22268 This rule does not have any parameters.
22272 @node Uncommented_BEGIN_In_Package_Bodies
22273 @subsection @code{Uncommented_BEGIN_In_Package_Bodies}
22274 @cindex @code{Uncommented_BEGIN_In_Package_Bodies} rule (for @command{gnatcheck})
22277 Flags each package body with declarations and a statement part that does not
22278 include a trailing comment on the line containing the @code{begin} keyword;
22279 this trailing comment needs to specify the package name and nothing else.
22280 The @code{begin} is not flagged if the package body does not
22281 contain any declarations.
22283 If the @code{begin} keyword is placed on the
22284 same line as the last declaration or the first statement, it is flagged
22285 independently of whether the line contains a trailing comment. The
22286 diagnostic message is attached to the line containing the first statement.
22288 This rule has no parameters.
22290 @node Unconditional_Exits
22291 @subsection @code{Unconditional_Exits}
22292 @cindex @code{Unconditional_Exits} rule (for @command{gnatcheck})
22295 Flag unconditional @code{exit} statements.
22297 This rule has no parameters.
22299 @node Unconstrained_Array_Returns
22300 @subsection @code{Unconstrained_Array_Returns}
22301 @cindex @code{Unconstrained_Array_Returns} rule (for @command{gnatcheck})
22304 Flag each function returning an unconstrained array. Function declarations,
22305 function bodies (and body stubs) having no separate specifications,
22306 and generic function instantiations are checked.
22307 Generic function declarations, function calls and function renamings are
22310 This rule has no parameters.
22312 @node Universal_Ranges
22313 @subsection @code{Universal_Ranges}
22314 @cindex @code{Universal_Ranges} rule (for @command{gnatcheck})
22317 Flag discrete ranges that are a part of an index constraint, constrained
22318 array definition, or @code{for}-loop parameter specification, and whose bounds
22319 are both of type @i{universal_integer}. Ranges that have at least one
22320 bound of a specific type (such as @code{1 .. N}, where @code{N} is a variable
22321 or an expression of non-universal type) are not flagged.
22323 This rule has no parameters.
22326 @node Unnamed_Blocks_And_Loops
22327 @subsection @code{Unnamed_Blocks_And_Loops}
22328 @cindex @code{Unnamed_Blocks_And_Loops} rule (for @command{gnatcheck})
22331 Flag each unnamed block statement and loop statement.
22333 The rule has no parameters.
22338 @node Unused_Subprograms
22339 @subsection @code{Unused_Subprograms} (under construction, GLOBAL)
22340 @cindex @code{Unused_Subprograms} rule (for @command{gnatcheck})
22343 Flag all unused subprograms.
22345 This rule has no parameters.
22351 @node USE_PACKAGE_Clauses
22352 @subsection @code{USE_PACKAGE_Clauses}
22353 @cindex @code{USE_PACKAGE_Clauses} rule (for @command{gnatcheck})
22356 Flag all @code{use} clauses for packages; @code{use type} clauses are
22359 This rule has no parameters.
22363 @node Volatile_Objects_Without_Address_Clauses
22364 @subsection @code{Volatile_Objects_Without_Address_Clauses}
22365 @cindex @code{Volatile_Objects_Without_Address_Clauses} rule (for @command{gnatcheck})
22368 Flag each volatile object that does not have an address clause.
22370 The following check is made: if the pragma @code{Volatile} is applied to a
22371 data object or to its type, then an address clause must
22372 be supplied for this object.
22374 This rule does not check the components of data objects,
22375 array components that are volatile as a result of the pragma
22376 @code{Volatile_Components}, or objects that are volatile because
22377 they are atomic as a result of pragmas @code{Atomic} or
22378 @code{Atomic_Components}.
22380 Only variable declarations, and not constant declarations, are checked.
22382 This rule has no parameters.
22385 @c *********************************
22386 @node Creating Sample Bodies Using gnatstub
22387 @chapter Creating Sample Bodies Using @command{gnatstub}
22391 @command{gnatstub} creates body stubs, that is, empty but compilable bodies
22392 for library unit declarations.
22394 Note: to invoke @code{gnatstub} with a project file, use the @code{gnat}
22395 driver (see @ref{The GNAT Driver and Project Files}).
22397 To create a body stub, @command{gnatstub} has to compile the library
22398 unit declaration. Therefore, bodies can be created only for legal
22399 library units. Moreover, if a library unit depends semantically upon
22400 units located outside the current directory, you have to provide
22401 the source search path when calling @command{gnatstub}, see the description
22402 of @command{gnatstub} switches below.
22404 By default, all the program unit body stubs generated by @code{gnatstub}
22405 raise the predefined @code{Program_Error} exception, which will catch
22406 accidental calls of generated stubs. This behavior can be changed with
22407 option @option{^--no-exception^/NO_EXCEPTION^} (see below).
22410 * Running gnatstub::
22411 * Switches for gnatstub::
22414 @node Running gnatstub
22415 @section Running @command{gnatstub}
22418 @command{gnatstub} has the command-line interface of the form
22421 $ gnatstub @ovar{switches} @var{filename} @ovar{directory}
22428 is the name of the source file that contains a library unit declaration
22429 for which a body must be created. The file name may contain the path
22431 The file name does not have to follow the GNAT file name conventions. If the
22433 does not follow GNAT file naming conventions, the name of the body file must
22435 explicitly as the value of the @option{^-o^/BODY=^@var{body-name}} option.
22436 If the file name follows the GNAT file naming
22437 conventions and the name of the body file is not provided,
22440 of the body file from the argument file name by replacing the @file{.ads}
22442 with the @file{.adb} suffix.
22445 indicates the directory in which the body stub is to be placed (the default
22450 is an optional sequence of switches as described in the next section
22453 @node Switches for gnatstub
22454 @section Switches for @command{gnatstub}
22460 @cindex @option{^-f^/FULL^} (@command{gnatstub})
22461 If the destination directory already contains a file with the name of the
22463 for the argument spec file, replace it with the generated body stub.
22465 @item ^-hs^/HEADER=SPEC^
22466 @cindex @option{^-hs^/HEADER=SPEC^} (@command{gnatstub})
22467 Put the comment header (i.e., all the comments preceding the
22468 compilation unit) from the source of the library unit declaration
22469 into the body stub.
22471 @item ^-hg^/HEADER=GENERAL^
22472 @cindex @option{^-hg^/HEADER=GENERAL^} (@command{gnatstub})
22473 Put a sample comment header into the body stub.
22475 @item ^--header-file=@var{filename}^/FROM_HEADER_FILE=@var{filename}^
22476 @cindex @option{^--header-file^/FROM_HEADER_FILE=^} (@command{gnatstub})
22477 Use the content of the file as the comment header for a generated body stub.
22481 @cindex @option{-IDIR} (@command{gnatstub})
22483 @cindex @option{-I-} (@command{gnatstub})
22486 @item /NOCURRENT_DIRECTORY
22487 @cindex @option{/NOCURRENT_DIRECTORY} (@command{gnatstub})
22489 ^These switches have ^This switch has^ the same meaning as in calls to
22491 ^They define ^It defines ^ the source search path in the call to
22492 @command{gcc} issued
22493 by @command{gnatstub} to compile an argument source file.
22495 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE=^@var{PATH}
22496 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@command{gnatstub})
22497 This switch has the same meaning as in calls to @command{gcc}.
22498 It defines the additional configuration file to be passed to the call to
22499 @command{gcc} issued
22500 by @command{gnatstub} to compile an argument source file.
22502 @item ^-gnatyM^/MAX_LINE_LENGTH=^@var{n}
22503 @cindex @option{^-gnatyM^/MAX_LINE_LENGTH^} (@command{gnatstub})
22504 (@var{n} is a non-negative integer). Set the maximum line length in the
22505 body stub to @var{n}; the default is 79. The maximum value that can be
22506 specified is 32767. Note that in the special case of configuration
22507 pragma files, the maximum is always 32767 regardless of whether or
22508 not this switch appears.
22510 @item ^-gnaty^/STYLE_CHECKS=^@var{n}
22511 @cindex @option{^-gnaty^/STYLE_CHECKS=^} (@command{gnatstub})
22512 (@var{n} is a non-negative integer from 1 to 9). Set the indentation level in
22513 the generated body sample to @var{n}.
22514 The default indentation is 3.
22516 @item ^-gnatyo^/ORDERED_SUBPROGRAMS^
22517 @cindex @option{^-gnato^/ORDERED_SUBPROGRAMS^} (@command{gnatstub})
22518 Order local bodies alphabetically. (By default local bodies are ordered
22519 in the same way as the corresponding local specs in the argument spec file.)
22521 @item ^-i^/INDENTATION=^@var{n}
22522 @cindex @option{^-i^/INDENTATION^} (@command{gnatstub})
22523 Same as @option{^-gnaty^/STYLE_CHECKS=^@var{n}}
22525 @item ^-k^/TREE_FILE=SAVE^
22526 @cindex @option{^-k^/TREE_FILE=SAVE^} (@command{gnatstub})
22527 Do not remove the tree file (i.e., the snapshot of the compiler internal
22528 structures used by @command{gnatstub}) after creating the body stub.
22530 @item ^-l^/LINE_LENGTH=^@var{n}
22531 @cindex @option{^-l^/LINE_LENGTH^} (@command{gnatstub})
22532 Same as @option{^-gnatyM^/MAX_LINE_LENGTH=^@var{n}}
22534 @item ^--no-exception^/NO_EXCEPTION^
22535 @cindex @option{^--no-exception^/NO_EXCEPTION^} (@command{gnatstub})
22536 Avoind raising PROGRAM_ERROR in the generated bodies of program unit stubs.
22537 This is not always possible for function stubs.
22539 @item ^-o ^/BODY=^@var{body-name}
22540 @cindex @option{^-o^/BODY^} (@command{gnatstub})
22541 Body file name. This should be set if the argument file name does not
22543 the GNAT file naming
22544 conventions. If this switch is omitted the default name for the body will be
22546 from the argument file name according to the GNAT file naming conventions.
22549 @cindex @option{^-q^/QUIET^} (@command{gnatstub})
22550 Quiet mode: do not generate a confirmation when a body is
22551 successfully created, and do not generate a message when a body is not
22555 @item ^-r^/TREE_FILE=REUSE^
22556 @cindex @option{^-r^/TREE_FILE=REUSE^} (@command{gnatstub})
22557 Reuse the tree file (if it exists) instead of creating it. Instead of
22558 creating the tree file for the library unit declaration, @command{gnatstub}
22559 tries to find it in the current directory and use it for creating
22560 a body. If the tree file is not found, no body is created. This option
22561 also implies @option{^-k^/SAVE^}, whether or not
22562 the latter is set explicitly.
22564 @item ^-t^/TREE_FILE=OVERWRITE^
22565 @cindex @option{^-t^/TREE_FILE=OVERWRITE^} (@command{gnatstub})
22566 Overwrite the existing tree file. If the current directory already
22567 contains the file which, according to the GNAT file naming rules should
22568 be considered as a tree file for the argument source file,
22570 will refuse to create the tree file needed to create a sample body
22571 unless this option is set.
22573 @item ^-v^/VERBOSE^
22574 @cindex @option{^-v^/VERBOSE^} (@command{gnatstub})
22575 Verbose mode: generate version information.
22579 @c *********************************
22580 @node Generating Ada Bindings for C and C++ headers
22581 @chapter Generating Ada Bindings for C and C++ headers
22585 GNAT now comes with a new experimental binding generator for C and C++
22586 headers which is intended to do 95% of the tedious work of generating
22587 Ada specs from C or C++ header files. Note that this still is a work in
22588 progress, not designed to generate 100% correct Ada specs.
22590 The code generated is using the Ada 2005 syntax, which makes it
22591 easier to interface with other languages than previous versions of Ada.
22594 * Running the binding generator::
22595 * Generating bindings for C++ headers::
22599 @node Running the binding generator
22600 @section Running the binding generator
22603 The binding generator is part of the @command{gcc} compiler and can be
22604 invoked via the @option{-fdump-ada-spec} switch, which will generate Ada
22605 spec files for the header files specified on the command line, and all
22606 header files needed by these files transitivitely. For example:
22609 $ g++ -c -fdump-ada-spec -C /usr/include/time.h
22610 $ gcc -c -gnat05 *.ads
22613 will generate, under GNU/Linux, the following files: @file{time_h.ads},
22614 @file{bits_time_h.ads}, @file{stddef_h.ads}, @file{bits_types_h.ads} which
22615 correspond to the files @file{/usr/include/time.h},
22616 @file{/usr/include/bits/time.h}, etc@dots{}, and will then compile in Ada 2005
22617 mode these Ada specs.
22619 The @code{-C} switch tells @command{gcc} to extract comments from headers,
22620 and will attempt to generate corresponding Ada comments.
22622 If you want to generate a single Ada file and not the transitive closure, you
22623 can use instead the @option{-fdump-ada-spec-slim} switch.
22625 Note that we recommend when possible to use the @command{g++} driver to
22626 generate bindings, even for most C headers, since this will in general
22627 generate better Ada specs. For generating bindings for C++ headers, it is
22628 mandatory to use the @command{g++} command, or @command{gcc -x c++} which
22629 is equivalent in this case. If @command{g++} cannot work on your C headers
22630 because of incompatibilities between C and C++, then you can fallback to
22631 @command{gcc} instead.
22633 For an example of better bindings generated from the C++ front-end,
22634 the name of the parameters (when available) are actually ignored by the C
22635 front-end. Consider the following C header:
22638 extern void foo (int variable);
22641 with the C front-end, @code{variable} is ignored, and the above is handled as:
22644 extern void foo (int);
22647 generating a generic:
22650 procedure foo (param1 : int);
22653 with the C++ front-end, the name is available, and we generate:
22656 procedure foo (variable : int);
22659 In some cases, the generated bindings will be more complete or more meaningful
22660 when defining some macros, which you can do via the @option{-D} switch. This
22661 is for example the case with @file{Xlib.h} under GNU/Linux:
22664 g++ -c -fdump-ada-spec -DXLIB_ILLEGAL_ACCESS -C /usr/include/X11/Xlib.h
22667 The above will generate more complete bindings than a straight call without
22668 the @option{-DXLIB_ILLEGAL_ACCESS} switch.
22670 In other cases, it is not possible to parse a header file in a stand alone
22671 manner, because other include files need to be included first. In this
22672 case, the solution is to create a small header file including the needed
22673 @code{#include} and possible @code{#define} directives. For example, to
22674 generate Ada bindings for @file{readline/readline.h}, you need to first
22675 include @file{stdio.h}, so you can create a file with the following two
22676 lines in e.g. @file{readline1.h}:
22680 #include <readline/readline.h>
22683 and then generate Ada bindings from this file:
22686 $ g++ -c -fdump-ada-spec readline1.h
22689 @node Generating bindings for C++ headers
22690 @section Generating bindings for C++ headers
22693 Generating bindings for C++ headers is done using the same options, always
22694 with the @command{g++} compiler.
22696 In this mode, C++ classes will be mapped to Ada tagged types, constructors
22697 will be mapped using the @code{CPP_Constructor} pragma, and when possible,
22698 multiple inheritance of abstract classes will be mapped to Ada interfaces
22699 (@xref{Interfacing to C++,,,gnat_rm, GNAT Reference Manual}, for additional
22700 information on interfacing to C++).
22702 For example, given the following C++ header file:
22709 virtual int Number_Of_Teeth () = 0;
22714 virtual void Set_Owner (char* Name) = 0;
22720 virtual void Set_Age (int New_Age);
22723 class Dog : Animal, Carnivore, Domestic @{
22728 virtual int Number_Of_Teeth ();
22729 virtual void Set_Owner (char* Name);
22737 The corresponding Ada code is generated:
22739 @smallexample @c ada
22742 package Class_Carnivore is
22743 type Carnivore is limited interface;
22744 pragma Import (CPP, Carnivore);
22746 function Number_Of_Teeth (this : access Carnivore) return int is abstract;
22748 use Class_Carnivore;
22750 package Class_Domestic is
22751 type Domestic is limited interface;
22752 pragma Import (CPP, Domestic);
22754 procedure Set_Owner
22755 (this : access Domestic;
22756 Name : Interfaces.C.Strings.chars_ptr) is abstract;
22758 use Class_Domestic;
22760 package Class_Animal is
22761 type Animal is tagged limited record
22762 Age_Count : aliased int;
22764 pragma Import (CPP, Animal);
22766 procedure Set_Age (this : access Animal; New_Age : int);
22767 pragma Import (CPP, Set_Age, "_ZN6Animal7Set_AgeEi");
22771 package Class_Dog is
22772 type Dog is new Animal and Carnivore and Domestic with record
22773 Tooth_Count : aliased int;
22774 Owner : Interfaces.C.Strings.chars_ptr;
22776 pragma Import (CPP, Dog);
22778 function Number_Of_Teeth (this : access Dog) return int;
22779 pragma Import (CPP, Number_Of_Teeth, "_ZN3Dog15Number_Of_TeethEv");
22781 procedure Set_Owner
22782 (this : access Dog; Name : Interfaces.C.Strings.chars_ptr);
22783 pragma Import (CPP, Set_Owner, "_ZN3Dog9Set_OwnerEPc");
22785 function New_Dog return Dog'Class;
22786 pragma CPP_Constructor (New_Dog);
22787 pragma Import (CPP, New_Dog, "_ZN3DogC1Ev");
22798 @item -fdump-ada-spec
22799 @cindex @option{-fdump-ada-spec} (@command{gcc})
22800 Generate Ada spec files for the given header files transitively (including
22801 all header files that these headers depend upon).
22803 @item -fdump-ada-spec-slim
22804 @cindex @option{-fdump-ada-spec-slim} (@command{gcc})
22805 Generate Ada spec files for the header files specified on the command line
22809 @cindex @option{-C} (@command{gcc})
22810 Extract comments from headers and generate Ada comments in the Ada spec files.
22813 @node Other Utility Programs
22814 @chapter Other Utility Programs
22817 This chapter discusses some other utility programs available in the Ada
22821 * Using Other Utility Programs with GNAT::
22822 * The External Symbol Naming Scheme of GNAT::
22823 * Converting Ada Files to html with gnathtml::
22824 * Installing gnathtml::
22831 @node Using Other Utility Programs with GNAT
22832 @section Using Other Utility Programs with GNAT
22835 The object files generated by GNAT are in standard system format and in
22836 particular the debugging information uses this format. This means
22837 programs generated by GNAT can be used with existing utilities that
22838 depend on these formats.
22841 In general, any utility program that works with C will also often work with
22842 Ada programs generated by GNAT. This includes software utilities such as
22843 gprof (a profiling program), @code{gdb} (the FSF debugger), and utilities such
22847 @node The External Symbol Naming Scheme of GNAT
22848 @section The External Symbol Naming Scheme of GNAT
22851 In order to interpret the output from GNAT, when using tools that are
22852 originally intended for use with other languages, it is useful to
22853 understand the conventions used to generate link names from the Ada
22856 All link names are in all lowercase letters. With the exception of library
22857 procedure names, the mechanism used is simply to use the full expanded
22858 Ada name with dots replaced by double underscores. For example, suppose
22859 we have the following package spec:
22861 @smallexample @c ada
22872 The variable @code{MN} has a full expanded Ada name of @code{QRS.MN}, so
22873 the corresponding link name is @code{qrs__mn}.
22875 Of course if a @code{pragma Export} is used this may be overridden:
22877 @smallexample @c ada
22882 pragma Export (Var1, C, External_Name => "var1_name");
22884 pragma Export (Var2, C, Link_Name => "var2_link_name");
22891 In this case, the link name for @var{Var1} is whatever link name the
22892 C compiler would assign for the C function @var{var1_name}. This typically
22893 would be either @var{var1_name} or @var{_var1_name}, depending on operating
22894 system conventions, but other possibilities exist. The link name for
22895 @var{Var2} is @var{var2_link_name}, and this is not operating system
22899 One exception occurs for library level procedures. A potential ambiguity
22900 arises between the required name @code{_main} for the C main program,
22901 and the name we would otherwise assign to an Ada library level procedure
22902 called @code{Main} (which might well not be the main program).
22904 To avoid this ambiguity, we attach the prefix @code{_ada_} to such
22905 names. So if we have a library level procedure such as
22907 @smallexample @c ada
22910 procedure Hello (S : String);
22916 the external name of this procedure will be @var{_ada_hello}.
22919 @node Converting Ada Files to html with gnathtml
22920 @section Converting Ada Files to HTML with @code{gnathtml}
22923 This @code{Perl} script allows Ada source files to be browsed using
22924 standard Web browsers. For installation procedure, see the section
22925 @xref{Installing gnathtml}.
22927 Ada reserved keywords are highlighted in a bold font and Ada comments in
22928 a blue font. Unless your program was compiled with the gcc @option{-gnatx}
22929 switch to suppress the generation of cross-referencing information, user
22930 defined variables and types will appear in a different color; you will
22931 be able to click on any identifier and go to its declaration.
22933 The command line is as follow:
22935 $ perl gnathtml.pl @ovar{^switches^options^} @var{ada-files}
22939 You can pass it as many Ada files as you want. @code{gnathtml} will generate
22940 an html file for every ada file, and a global file called @file{index.htm}.
22941 This file is an index of every identifier defined in the files.
22943 The available ^switches^options^ are the following ones:
22947 @cindex @option{-83} (@code{gnathtml})
22948 Only the Ada 83 subset of keywords will be highlighted.
22950 @item -cc @var{color}
22951 @cindex @option{-cc} (@code{gnathtml})
22952 This option allows you to change the color used for comments. The default
22953 value is green. The color argument can be any name accepted by html.
22956 @cindex @option{-d} (@code{gnathtml})
22957 If the Ada files depend on some other files (for instance through
22958 @code{with} clauses, the latter files will also be converted to html.
22959 Only the files in the user project will be converted to html, not the files
22960 in the run-time library itself.
22963 @cindex @option{-D} (@code{gnathtml})
22964 This command is the same as @option{-d} above, but @command{gnathtml} will
22965 also look for files in the run-time library, and generate html files for them.
22967 @item -ext @var{extension}
22968 @cindex @option{-ext} (@code{gnathtml})
22969 This option allows you to change the extension of the generated HTML files.
22970 If you do not specify an extension, it will default to @file{htm}.
22973 @cindex @option{-f} (@code{gnathtml})
22974 By default, gnathtml will generate html links only for global entities
22975 ('with'ed units, global variables and types,@dots{}). If you specify
22976 @option{-f} on the command line, then links will be generated for local
22979 @item -l @var{number}
22980 @cindex @option{-l} (@code{gnathtml})
22981 If this ^switch^option^ is provided and @var{number} is not 0, then
22982 @code{gnathtml} will number the html files every @var{number} line.
22985 @cindex @option{-I} (@code{gnathtml})
22986 Specify a directory to search for library files (@file{.ALI} files) and
22987 source files. You can provide several -I switches on the command line,
22988 and the directories will be parsed in the order of the command line.
22991 @cindex @option{-o} (@code{gnathtml})
22992 Specify the output directory for html files. By default, gnathtml will
22993 saved the generated html files in a subdirectory named @file{html/}.
22995 @item -p @var{file}
22996 @cindex @option{-p} (@code{gnathtml})
22997 If you are using Emacs and the most recent Emacs Ada mode, which provides
22998 a full Integrated Development Environment for compiling, checking,
22999 running and debugging applications, you may use @file{.gpr} files
23000 to give the directories where Emacs can find sources and object files.
23002 Using this ^switch^option^, you can tell gnathtml to use these files.
23003 This allows you to get an html version of your application, even if it
23004 is spread over multiple directories.
23006 @item -sc @var{color}
23007 @cindex @option{-sc} (@code{gnathtml})
23008 This ^switch^option^ allows you to change the color used for symbol
23010 The default value is red. The color argument can be any name accepted by html.
23012 @item -t @var{file}
23013 @cindex @option{-t} (@code{gnathtml})
23014 This ^switch^option^ provides the name of a file. This file contains a list of
23015 file names to be converted, and the effect is exactly as though they had
23016 appeared explicitly on the command line. This
23017 is the recommended way to work around the command line length limit on some
23022 @node Installing gnathtml
23023 @section Installing @code{gnathtml}
23026 @code{Perl} needs to be installed on your machine to run this script.
23027 @code{Perl} is freely available for almost every architecture and
23028 Operating System via the Internet.
23030 On Unix systems, you may want to modify the first line of the script
23031 @code{gnathtml}, to explicitly tell the Operating system where Perl
23032 is. The syntax of this line is:
23034 #!full_path_name_to_perl
23038 Alternatively, you may run the script using the following command line:
23041 $ perl gnathtml.pl @ovar{switches} @var{files}
23050 The GNAT distribution provides an Ada 95 template for the HP Language
23051 Sensitive Editor (LSE), a component of DECset. In order to
23052 access it, invoke LSE with the qualifier /ENVIRONMENT=GNU:[LIB]ADA95.ENV.
23059 GNAT supports The HP Performance Coverage Analyzer (PCA), a component
23060 of DECset. To use it proceed as outlined under ``HELP PCA'', except for running
23061 the collection phase with the /DEBUG qualifier.
23064 $ GNAT MAKE /DEBUG <PROGRAM_NAME>
23065 $ DEFINE LIB$DEBUG PCA$COLLECTOR
23066 $ RUN/DEBUG <PROGRAM_NAME>
23072 @c ******************************
23073 @node Code Coverage and Profiling
23074 @chapter Code Coverage and Profiling
23075 @cindex Code Coverage
23079 This chapter describes how to use @code{gcov} - coverage testing tool - and
23080 @code{gprof} - profiler tool - on your Ada programs.
23083 * Code Coverage of Ada Programs using gcov::
23084 * Profiling an Ada Program using gprof::
23087 @node Code Coverage of Ada Programs using gcov
23088 @section Code Coverage of Ada Programs using gcov
23090 @cindex -fprofile-arcs
23091 @cindex -ftest-coverage
23093 @cindex Code Coverage
23096 @code{gcov} is a test coverage program: it analyzes the execution of a given
23097 program on selected tests, to help you determine the portions of the program
23098 that are still untested.
23100 @code{gcov} is part of the GCC suite, and is described in detail in the GCC
23101 User's Guide. You can refer to this documentation for a more complete
23104 This chapter provides a quick startup guide, and
23105 details some Gnat-specific features.
23108 * Quick startup guide::
23112 @node Quick startup guide
23113 @subsection Quick startup guide
23115 In order to perform coverage analysis of a program using @code{gcov}, 3
23120 Code instrumentation during the compilation process
23122 Execution of the instrumented program
23124 Execution of the @code{gcov} tool to generate the result.
23127 The code instrumentation needed by gcov is created at the object level:
23128 The source code is not modified in any way, because the instrumentation code is
23129 inserted by gcc during the compilation process. To compile your code with code
23130 coverage activated, you need to recompile your whole project using the
23132 @code{-fprofile-arcs} and @code{-ftest-coverage}, and link it using
23133 @code{-fprofile-arcs}.
23136 $ gnatmake -P my_project.gpr -f -cargs -fprofile-arcs -ftest-coverage \
23137 -largs -fprofile-arcs
23140 This compilation process will create @file{.gcno} files together with
23141 the usual object files.
23143 Once the program is compiled with coverage instrumentation, you can
23144 run it as many times as needed - on portions of a test suite for
23145 example. The first execution will produce @file{.gcda} files at the
23146 same location as the @file{.gcno} files. The following executions
23147 will update those files, so that a cumulative result of the covered
23148 portions of the program is generated.
23150 Finally, you need to call the @code{gcov} tool. The different options of
23151 @code{gcov} are available in the GCC User's Guide, section 'Invoking gcov'.
23153 This will create annotated source files with a @file{.gcov} extension:
23154 @file{my_main.adb} file will be analysed in @file{my_main.adb.gcov}.
23156 @node Gnat specifics
23157 @subsection Gnat specifics
23159 Because Ada semantics, portions of the source code may be shared among
23160 several object files. This is the case for example when generics are
23161 involved, when inlining is active or when declarations generate initialisation
23162 calls. In order to take
23163 into account this shared code, you need to call @code{gcov} on all
23164 source files of the tested program at once.
23166 The list of source files might exceed the system's maximum command line
23167 length. In order to bypass this limitation, a new mechanism has been
23168 implemented in @code{gcov}: you can now list all your project's files into a
23169 text file, and provide this file to gcov as a parameter, preceded by a @@
23170 (e.g. @samp{gcov @@mysrclist.txt}).
23172 Note that on AIX compiling a static library with @code{-fprofile-arcs} is
23173 not supported as there can be unresolved symbols during the final link.
23175 @node Profiling an Ada Program using gprof
23176 @section Profiling an Ada Program using gprof
23182 This section is not meant to be an exhaustive documentation of @code{gprof}.
23183 Full documentation for it can be found in the GNU Profiler User's Guide
23184 documentation that is part of this GNAT distribution.
23186 Profiling a program helps determine the parts of a program that are executed
23187 most often, and are therefore the most time-consuming.
23189 @code{gprof} is the standard GNU profiling tool; it has been enhanced to
23190 better handle Ada programs and multitasking.
23191 It is currently supported on the following platforms
23196 solaris sparc/sparc64/x86
23202 In order to profile a program using @code{gprof}, 3 steps are needed:
23206 Code instrumentation, requiring a full recompilation of the project with the
23209 Execution of the program under the analysis conditions, i.e. with the desired
23212 Analysis of the results using the @code{gprof} tool.
23216 The following sections detail the different steps, and indicate how
23217 to interpret the results:
23219 * Compilation for profiling::
23220 * Program execution::
23222 * Interpretation of profiling results::
23225 @node Compilation for profiling
23226 @subsection Compilation for profiling
23230 In order to profile a program the first step is to tell the compiler
23231 to generate the necessary profiling information. The compiler switch to be used
23232 is @code{-pg}, which must be added to other compilation switches. This
23233 switch needs to be specified both during compilation and link stages, and can
23234 be specified once when using gnatmake:
23237 gnatmake -f -pg -P my_project
23241 Note that only the objects that were compiled with the @samp{-pg} switch will be
23242 profiled; if you need to profile your whole project, use the
23243 @samp{-f} gnatmake switch to force full recompilation.
23245 @node Program execution
23246 @subsection Program execution
23249 Once the program has been compiled for profiling, you can run it as usual.
23251 The only constraint imposed by profiling is that the program must terminate
23252 normally. An interrupted program (via a Ctrl-C, kill, etc.) will not be
23255 Once the program completes execution, a data file called @file{gmon.out} is
23256 generated in the directory where the program was launched from. If this file
23257 already exists, it will be overwritten.
23259 @node Running gprof
23260 @subsection Running gprof
23263 The @code{gprof} tool is called as follow:
23266 gprof my_prog gmon.out
23277 The complete form of the gprof command line is the following:
23280 gprof [^switches^options^] [executable [data-file]]
23284 @code{gprof} supports numerous ^switch^options^. The order of these
23285 ^switch^options^ does not matter. The full list of options can be found in
23286 the GNU Profiler User's Guide documentation that comes with this documentation.
23288 The following is the subset of those switches that is most relevant:
23292 @item --demangle[=@var{style}]
23293 @itemx --no-demangle
23294 @cindex @option{--demangle} (@code{gprof})
23295 These options control whether symbol names should be demangled when
23296 printing output. The default is to demangle C++ symbols. The
23297 @code{--no-demangle} option may be used to turn off demangling. Different
23298 compilers have different mangling styles. The optional demangling style
23299 argument can be used to choose an appropriate demangling style for your
23300 compiler, in particular Ada symbols generated by GNAT can be demangled using
23301 @code{--demangle=gnat}.
23303 @item -e @var{function_name}
23304 @cindex @option{-e} (@code{gprof})
23305 The @samp{-e @var{function}} option tells @code{gprof} not to print
23306 information about the function @var{function_name} (and its
23307 children@dots{}) in the call graph. The function will still be listed
23308 as a child of any functions that call it, but its index number will be
23309 shown as @samp{[not printed]}. More than one @samp{-e} option may be
23310 given; only one @var{function_name} may be indicated with each @samp{-e}
23313 @item -E @var{function_name}
23314 @cindex @option{-E} (@code{gprof})
23315 The @code{-E @var{function}} option works like the @code{-e} option, but
23316 execution time spent in the function (and children who were not called from
23317 anywhere else), will not be used to compute the percentages-of-time for
23318 the call graph. More than one @samp{-E} option may be given; only one
23319 @var{function_name} may be indicated with each @samp{-E} option.
23321 @item -f @var{function_name}
23322 @cindex @option{-f} (@code{gprof})
23323 The @samp{-f @var{function}} option causes @code{gprof} to limit the
23324 call graph to the function @var{function_name} and its children (and
23325 their children@dots{}). More than one @samp{-f} option may be given;
23326 only one @var{function_name} may be indicated with each @samp{-f}
23329 @item -F @var{function_name}
23330 @cindex @option{-F} (@code{gprof})
23331 The @samp{-F @var{function}} option works like the @code{-f} option, but
23332 only time spent in the function and its children (and their
23333 children@dots{}) will be used to determine total-time and
23334 percentages-of-time for the call graph. More than one @samp{-F} option
23335 may be given; only one @var{function_name} may be indicated with each
23336 @samp{-F} option. The @samp{-F} option overrides the @samp{-E} option.
23340 @node Interpretation of profiling results
23341 @subsection Interpretation of profiling results
23345 The results of the profiling analysis are represented by two arrays: the
23346 'flat profile' and the 'call graph'. Full documentation of those outputs
23347 can be found in the GNU Profiler User's Guide.
23349 The flat profile shows the time spent in each function of the program, and how
23350 many time it has been called. This allows you to locate easily the most
23351 time-consuming functions.
23353 The call graph shows, for each subprogram, the subprograms that call it,
23354 and the subprograms that it calls. It also provides an estimate of the time
23355 spent in each of those callers/called subprograms.
23358 @c ******************************
23359 @node Running and Debugging Ada Programs
23360 @chapter Running and Debugging Ada Programs
23364 This chapter discusses how to debug Ada programs.
23366 It applies to GNAT on the Alpha OpenVMS platform;
23367 for I64 OpenVMS please refer to the @cite{OpenVMS Debugger Manual},
23368 since HP has implemented Ada support in the OpenVMS debugger on I64.
23371 An incorrect Ada program may be handled in three ways by the GNAT compiler:
23375 The illegality may be a violation of the static semantics of Ada. In
23376 that case GNAT diagnoses the constructs in the program that are illegal.
23377 It is then a straightforward matter for the user to modify those parts of
23381 The illegality may be a violation of the dynamic semantics of Ada. In
23382 that case the program compiles and executes, but may generate incorrect
23383 results, or may terminate abnormally with some exception.
23386 When presented with a program that contains convoluted errors, GNAT
23387 itself may terminate abnormally without providing full diagnostics on
23388 the incorrect user program.
23392 * The GNAT Debugger GDB::
23394 * Introduction to GDB Commands::
23395 * Using Ada Expressions::
23396 * Calling User-Defined Subprograms::
23397 * Using the Next Command in a Function::
23400 * Debugging Generic Units::
23401 * GNAT Abnormal Termination or Failure to Terminate::
23402 * Naming Conventions for GNAT Source Files::
23403 * Getting Internal Debugging Information::
23404 * Stack Traceback::
23410 @node The GNAT Debugger GDB
23411 @section The GNAT Debugger GDB
23414 @code{GDB} is a general purpose, platform-independent debugger that
23415 can be used to debug mixed-language programs compiled with @command{gcc},
23416 and in particular is capable of debugging Ada programs compiled with
23417 GNAT. The latest versions of @code{GDB} are Ada-aware and can handle
23418 complex Ada data structures.
23420 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
23422 located in the GNU:[DOCS] directory,
23424 for full details on the usage of @code{GDB}, including a section on
23425 its usage on programs. This manual should be consulted for full
23426 details. The section that follows is a brief introduction to the
23427 philosophy and use of @code{GDB}.
23429 When GNAT programs are compiled, the compiler optionally writes debugging
23430 information into the generated object file, including information on
23431 line numbers, and on declared types and variables. This information is
23432 separate from the generated code. It makes the object files considerably
23433 larger, but it does not add to the size of the actual executable that
23434 will be loaded into memory, and has no impact on run-time performance. The
23435 generation of debug information is triggered by the use of the
23436 ^-g^/DEBUG^ switch in the @command{gcc} or @command{gnatmake} command
23437 used to carry out the compilations. It is important to emphasize that
23438 the use of these options does not change the generated code.
23440 The debugging information is written in standard system formats that
23441 are used by many tools, including debuggers and profilers. The format
23442 of the information is typically designed to describe C types and
23443 semantics, but GNAT implements a translation scheme which allows full
23444 details about Ada types and variables to be encoded into these
23445 standard C formats. Details of this encoding scheme may be found in
23446 the file exp_dbug.ads in the GNAT source distribution. However, the
23447 details of this encoding are, in general, of no interest to a user,
23448 since @code{GDB} automatically performs the necessary decoding.
23450 When a program is bound and linked, the debugging information is
23451 collected from the object files, and stored in the executable image of
23452 the program. Again, this process significantly increases the size of
23453 the generated executable file, but it does not increase the size of
23454 the executable program itself. Furthermore, if this program is run in
23455 the normal manner, it runs exactly as if the debug information were
23456 not present, and takes no more actual memory.
23458 However, if the program is run under control of @code{GDB}, the
23459 debugger is activated. The image of the program is loaded, at which
23460 point it is ready to run. If a run command is given, then the program
23461 will run exactly as it would have if @code{GDB} were not present. This
23462 is a crucial part of the @code{GDB} design philosophy. @code{GDB} is
23463 entirely non-intrusive until a breakpoint is encountered. If no
23464 breakpoint is ever hit, the program will run exactly as it would if no
23465 debugger were present. When a breakpoint is hit, @code{GDB} accesses
23466 the debugging information and can respond to user commands to inspect
23467 variables, and more generally to report on the state of execution.
23471 @section Running GDB
23474 This section describes how to initiate the debugger.
23475 @c The above sentence is really just filler, but it was otherwise
23476 @c clumsy to get the first paragraph nonindented given the conditional
23477 @c nature of the description
23480 The debugger can be launched from a @code{GPS} menu or
23481 directly from the command line. The description below covers the latter use.
23482 All the commands shown can be used in the @code{GPS} debug console window,
23483 but there are usually more GUI-based ways to achieve the same effect.
23486 The command to run @code{GDB} is
23489 $ ^gdb program^GDB PROGRAM^
23493 where @code{^program^PROGRAM^} is the name of the executable file. This
23494 activates the debugger and results in a prompt for debugger commands.
23495 The simplest command is simply @code{run}, which causes the program to run
23496 exactly as if the debugger were not present. The following section
23497 describes some of the additional commands that can be given to @code{GDB}.
23499 @c *******************************
23500 @node Introduction to GDB Commands
23501 @section Introduction to GDB Commands
23504 @code{GDB} contains a large repertoire of commands. @xref{Top,,
23505 Debugging with GDB, gdb, Debugging with GDB},
23507 located in the GNU:[DOCS] directory,
23509 for extensive documentation on the use
23510 of these commands, together with examples of their use. Furthermore,
23511 the command @command{help} invoked from within GDB activates a simple help
23512 facility which summarizes the available commands and their options.
23513 In this section we summarize a few of the most commonly
23514 used commands to give an idea of what @code{GDB} is about. You should create
23515 a simple program with debugging information and experiment with the use of
23516 these @code{GDB} commands on the program as you read through the
23520 @item set args @var{arguments}
23521 The @var{arguments} list above is a list of arguments to be passed to
23522 the program on a subsequent run command, just as though the arguments
23523 had been entered on a normal invocation of the program. The @code{set args}
23524 command is not needed if the program does not require arguments.
23527 The @code{run} command causes execution of the program to start from
23528 the beginning. If the program is already running, that is to say if
23529 you are currently positioned at a breakpoint, then a prompt will ask
23530 for confirmation that you want to abandon the current execution and
23533 @item breakpoint @var{location}
23534 The breakpoint command sets a breakpoint, that is to say a point at which
23535 execution will halt and @code{GDB} will await further
23536 commands. @var{location} is
23537 either a line number within a file, given in the format @code{file:linenumber},
23538 or it is the name of a subprogram. If you request that a breakpoint be set on
23539 a subprogram that is overloaded, a prompt will ask you to specify on which of
23540 those subprograms you want to breakpoint. You can also
23541 specify that all of them should be breakpointed. If the program is run
23542 and execution encounters the breakpoint, then the program
23543 stops and @code{GDB} signals that the breakpoint was encountered by
23544 printing the line of code before which the program is halted.
23546 @item breakpoint exception @var{name}
23547 A special form of the breakpoint command which breakpoints whenever
23548 exception @var{name} is raised.
23549 If @var{name} is omitted,
23550 then a breakpoint will occur when any exception is raised.
23552 @item print @var{expression}
23553 This will print the value of the given expression. Most simple
23554 Ada expression formats are properly handled by @code{GDB}, so the expression
23555 can contain function calls, variables, operators, and attribute references.
23558 Continues execution following a breakpoint, until the next breakpoint or the
23559 termination of the program.
23562 Executes a single line after a breakpoint. If the next statement
23563 is a subprogram call, execution continues into (the first statement of)
23564 the called subprogram.
23567 Executes a single line. If this line is a subprogram call, executes and
23568 returns from the call.
23571 Lists a few lines around the current source location. In practice, it
23572 is usually more convenient to have a separate edit window open with the
23573 relevant source file displayed. Successive applications of this command
23574 print subsequent lines. The command can be given an argument which is a
23575 line number, in which case it displays a few lines around the specified one.
23578 Displays a backtrace of the call chain. This command is typically
23579 used after a breakpoint has occurred, to examine the sequence of calls that
23580 leads to the current breakpoint. The display includes one line for each
23581 activation record (frame) corresponding to an active subprogram.
23584 At a breakpoint, @code{GDB} can display the values of variables local
23585 to the current frame. The command @code{up} can be used to
23586 examine the contents of other active frames, by moving the focus up
23587 the stack, that is to say from callee to caller, one frame at a time.
23590 Moves the focus of @code{GDB} down from the frame currently being
23591 examined to the frame of its callee (the reverse of the previous command),
23593 @item frame @var{n}
23594 Inspect the frame with the given number. The value 0 denotes the frame
23595 of the current breakpoint, that is to say the top of the call stack.
23600 The above list is a very short introduction to the commands that
23601 @code{GDB} provides. Important additional capabilities, including conditional
23602 breakpoints, the ability to execute command sequences on a breakpoint,
23603 the ability to debug at the machine instruction level and many other
23604 features are described in detail in @ref{Top,, Debugging with GDB, gdb,
23605 Debugging with GDB}. Note that most commands can be abbreviated
23606 (for example, c for continue, bt for backtrace).
23608 @node Using Ada Expressions
23609 @section Using Ada Expressions
23610 @cindex Ada expressions
23613 @code{GDB} supports a fairly large subset of Ada expression syntax, with some
23614 extensions. The philosophy behind the design of this subset is
23618 That @code{GDB} should provide basic literals and access to operations for
23619 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
23620 leaving more sophisticated computations to subprograms written into the
23621 program (which therefore may be called from @code{GDB}).
23624 That type safety and strict adherence to Ada language restrictions
23625 are not particularly important to the @code{GDB} user.
23628 That brevity is important to the @code{GDB} user.
23632 Thus, for brevity, the debugger acts as if there were
23633 implicit @code{with} and @code{use} clauses in effect for all user-written
23634 packages, thus making it unnecessary to fully qualify most names with
23635 their packages, regardless of context. Where this causes ambiguity,
23636 @code{GDB} asks the user's intent.
23638 For details on the supported Ada syntax, see @ref{Top,, Debugging with
23639 GDB, gdb, Debugging with GDB}.
23641 @node Calling User-Defined Subprograms
23642 @section Calling User-Defined Subprograms
23645 An important capability of @code{GDB} is the ability to call user-defined
23646 subprograms while debugging. This is achieved simply by entering
23647 a subprogram call statement in the form:
23650 call subprogram-name (parameters)
23654 The keyword @code{call} can be omitted in the normal case where the
23655 @code{subprogram-name} does not coincide with any of the predefined
23656 @code{GDB} commands.
23658 The effect is to invoke the given subprogram, passing it the
23659 list of parameters that is supplied. The parameters can be expressions and
23660 can include variables from the program being debugged. The
23661 subprogram must be defined
23662 at the library level within your program, and @code{GDB} will call the
23663 subprogram within the environment of your program execution (which
23664 means that the subprogram is free to access or even modify variables
23665 within your program).
23667 The most important use of this facility is in allowing the inclusion of
23668 debugging routines that are tailored to particular data structures
23669 in your program. Such debugging routines can be written to provide a suitably
23670 high-level description of an abstract type, rather than a low-level dump
23671 of its physical layout. After all, the standard
23672 @code{GDB print} command only knows the physical layout of your
23673 types, not their abstract meaning. Debugging routines can provide information
23674 at the desired semantic level and are thus enormously useful.
23676 For example, when debugging GNAT itself, it is crucial to have access to
23677 the contents of the tree nodes used to represent the program internally.
23678 But tree nodes are represented simply by an integer value (which in turn
23679 is an index into a table of nodes).
23680 Using the @code{print} command on a tree node would simply print this integer
23681 value, which is not very useful. But the PN routine (defined in file
23682 treepr.adb in the GNAT sources) takes a tree node as input, and displays
23683 a useful high level representation of the tree node, which includes the
23684 syntactic category of the node, its position in the source, the integers
23685 that denote descendant nodes and parent node, as well as varied
23686 semantic information. To study this example in more detail, you might want to
23687 look at the body of the PN procedure in the stated file.
23689 @node Using the Next Command in a Function
23690 @section Using the Next Command in a Function
23693 When you use the @code{next} command in a function, the current source
23694 location will advance to the next statement as usual. A special case
23695 arises in the case of a @code{return} statement.
23697 Part of the code for a return statement is the ``epilog'' of the function.
23698 This is the code that returns to the caller. There is only one copy of
23699 this epilog code, and it is typically associated with the last return
23700 statement in the function if there is more than one return. In some
23701 implementations, this epilog is associated with the first statement
23704 The result is that if you use the @code{next} command from a return
23705 statement that is not the last return statement of the function you
23706 may see a strange apparent jump to the last return statement or to
23707 the start of the function. You should simply ignore this odd jump.
23708 The value returned is always that from the first return statement
23709 that was stepped through.
23711 @node Ada Exceptions
23712 @section Breaking on Ada Exceptions
23716 You can set breakpoints that trip when your program raises
23717 selected exceptions.
23720 @item break exception
23721 Set a breakpoint that trips whenever (any task in the) program raises
23724 @item break exception @var{name}
23725 Set a breakpoint that trips whenever (any task in the) program raises
23726 the exception @var{name}.
23728 @item break exception unhandled
23729 Set a breakpoint that trips whenever (any task in the) program raises an
23730 exception for which there is no handler.
23732 @item info exceptions
23733 @itemx info exceptions @var{regexp}
23734 The @code{info exceptions} command permits the user to examine all defined
23735 exceptions within Ada programs. With a regular expression, @var{regexp}, as
23736 argument, prints out only those exceptions whose name matches @var{regexp}.
23744 @code{GDB} allows the following task-related commands:
23748 This command shows a list of current Ada tasks, as in the following example:
23755 ID TID P-ID Thread Pri State Name
23756 1 8088000 0 807e000 15 Child Activation Wait main_task
23757 2 80a4000 1 80ae000 15 Accept/Select Wait b
23758 3 809a800 1 80a4800 15 Child Activation Wait a
23759 * 4 80ae800 3 80b8000 15 Running c
23763 In this listing, the asterisk before the first task indicates it to be the
23764 currently running task. The first column lists the task ID that is used
23765 to refer to tasks in the following commands.
23767 @item break @var{linespec} task @var{taskid}
23768 @itemx break @var{linespec} task @var{taskid} if @dots{}
23769 @cindex Breakpoints and tasks
23770 These commands are like the @code{break @dots{} thread @dots{}}.
23771 @var{linespec} specifies source lines.
23773 Use the qualifier @samp{task @var{taskid}} with a breakpoint command
23774 to specify that you only want @code{GDB} to stop the program when a
23775 particular Ada task reaches this breakpoint. @var{taskid} is one of the
23776 numeric task identifiers assigned by @code{GDB}, shown in the first
23777 column of the @samp{info tasks} display.
23779 If you do not specify @samp{task @var{taskid}} when you set a
23780 breakpoint, the breakpoint applies to @emph{all} tasks of your
23783 You can use the @code{task} qualifier on conditional breakpoints as
23784 well; in this case, place @samp{task @var{taskid}} before the
23785 breakpoint condition (before the @code{if}).
23787 @item task @var{taskno}
23788 @cindex Task switching
23790 This command allows to switch to the task referred by @var{taskno}. In
23791 particular, This allows to browse the backtrace of the specified
23792 task. It is advised to switch back to the original task before
23793 continuing execution otherwise the scheduling of the program may be
23798 For more detailed information on the tasking support,
23799 see @ref{Top,, Debugging with GDB, gdb, Debugging with GDB}.
23801 @node Debugging Generic Units
23802 @section Debugging Generic Units
23803 @cindex Debugging Generic Units
23807 GNAT always uses code expansion for generic instantiation. This means that
23808 each time an instantiation occurs, a complete copy of the original code is
23809 made, with appropriate substitutions of formals by actuals.
23811 It is not possible to refer to the original generic entities in
23812 @code{GDB}, but it is always possible to debug a particular instance of
23813 a generic, by using the appropriate expanded names. For example, if we have
23815 @smallexample @c ada
23820 generic package k is
23821 procedure kp (v1 : in out integer);
23825 procedure kp (v1 : in out integer) is
23831 package k1 is new k;
23832 package k2 is new k;
23834 var : integer := 1;
23847 Then to break on a call to procedure kp in the k2 instance, simply
23851 (gdb) break g.k2.kp
23855 When the breakpoint occurs, you can step through the code of the
23856 instance in the normal manner and examine the values of local variables, as for
23859 @node GNAT Abnormal Termination or Failure to Terminate
23860 @section GNAT Abnormal Termination or Failure to Terminate
23861 @cindex GNAT Abnormal Termination or Failure to Terminate
23864 When presented with programs that contain serious errors in syntax
23866 GNAT may on rare occasions experience problems in operation, such
23868 segmentation fault or illegal memory access, raising an internal
23869 exception, terminating abnormally, or failing to terminate at all.
23870 In such cases, you can activate
23871 various features of GNAT that can help you pinpoint the construct in your
23872 program that is the likely source of the problem.
23874 The following strategies are presented in increasing order of
23875 difficulty, corresponding to your experience in using GNAT and your
23876 familiarity with compiler internals.
23880 Run @command{gcc} with the @option{-gnatf}. This first
23881 switch causes all errors on a given line to be reported. In its absence,
23882 only the first error on a line is displayed.
23884 The @option{-gnatdO} switch causes errors to be displayed as soon as they
23885 are encountered, rather than after compilation is terminated. If GNAT
23886 terminates prematurely or goes into an infinite loop, the last error
23887 message displayed may help to pinpoint the culprit.
23890 Run @command{gcc} with the @option{^-v (verbose)^/VERBOSE^} switch. In this
23891 mode, @command{gcc} produces ongoing information about the progress of the
23892 compilation and provides the name of each procedure as code is
23893 generated. This switch allows you to find which Ada procedure was being
23894 compiled when it encountered a code generation problem.
23897 @cindex @option{-gnatdc} switch
23898 Run @command{gcc} with the @option{-gnatdc} switch. This is a GNAT specific
23899 switch that does for the front-end what @option{^-v^VERBOSE^} does
23900 for the back end. The system prints the name of each unit,
23901 either a compilation unit or nested unit, as it is being analyzed.
23903 Finally, you can start
23904 @code{gdb} directly on the @code{gnat1} executable. @code{gnat1} is the
23905 front-end of GNAT, and can be run independently (normally it is just
23906 called from @command{gcc}). You can use @code{gdb} on @code{gnat1} as you
23907 would on a C program (but @pxref{The GNAT Debugger GDB} for caveats). The
23908 @code{where} command is the first line of attack; the variable
23909 @code{lineno} (seen by @code{print lineno}), used by the second phase of
23910 @code{gnat1} and by the @command{gcc} backend, indicates the source line at
23911 which the execution stopped, and @code{input_file name} indicates the name of
23915 @node Naming Conventions for GNAT Source Files
23916 @section Naming Conventions for GNAT Source Files
23919 In order to examine the workings of the GNAT system, the following
23920 brief description of its organization may be helpful:
23924 Files with prefix @file{^sc^SC^} contain the lexical scanner.
23927 All files prefixed with @file{^par^PAR^} are components of the parser. The
23928 numbers correspond to chapters of the Ada Reference Manual. For example,
23929 parsing of select statements can be found in @file{par-ch9.adb}.
23932 All files prefixed with @file{^sem^SEM^} perform semantic analysis. The
23933 numbers correspond to chapters of the Ada standard. For example, all
23934 issues involving context clauses can be found in @file{sem_ch10.adb}. In
23935 addition, some features of the language require sufficient special processing
23936 to justify their own semantic files: sem_aggr for aggregates, sem_disp for
23937 dynamic dispatching, etc.
23940 All files prefixed with @file{^exp^EXP^} perform normalization and
23941 expansion of the intermediate representation (abstract syntax tree, or AST).
23942 these files use the same numbering scheme as the parser and semantics files.
23943 For example, the construction of record initialization procedures is done in
23944 @file{exp_ch3.adb}.
23947 The files prefixed with @file{^bind^BIND^} implement the binder, which
23948 verifies the consistency of the compilation, determines an order of
23949 elaboration, and generates the bind file.
23952 The files @file{atree.ads} and @file{atree.adb} detail the low-level
23953 data structures used by the front-end.
23956 The files @file{sinfo.ads} and @file{sinfo.adb} detail the structure of
23957 the abstract syntax tree as produced by the parser.
23960 The files @file{einfo.ads} and @file{einfo.adb} detail the attributes of
23961 all entities, computed during semantic analysis.
23964 Library management issues are dealt with in files with prefix
23970 Ada files with the prefix @file{^a-^A-^} are children of @code{Ada}, as
23971 defined in Annex A.
23976 Files with prefix @file{^i-^I-^} are children of @code{Interfaces}, as
23977 defined in Annex B.
23981 Files with prefix @file{^s-^S-^} are children of @code{System}. This includes
23982 both language-defined children and GNAT run-time routines.
23986 Files with prefix @file{^g-^G-^} are children of @code{GNAT}. These are useful
23987 general-purpose packages, fully documented in their specs. All
23988 the other @file{.c} files are modifications of common @command{gcc} files.
23991 @node Getting Internal Debugging Information
23992 @section Getting Internal Debugging Information
23995 Most compilers have internal debugging switches and modes. GNAT
23996 does also, except GNAT internal debugging switches and modes are not
23997 secret. A summary and full description of all the compiler and binder
23998 debug flags are in the file @file{debug.adb}. You must obtain the
23999 sources of the compiler to see the full detailed effects of these flags.
24001 The switches that print the source of the program (reconstructed from
24002 the internal tree) are of general interest for user programs, as are the
24004 the full internal tree, and the entity table (the symbol table
24005 information). The reconstructed source provides a readable version of the
24006 program after the front-end has completed analysis and expansion,
24007 and is useful when studying the performance of specific constructs.
24008 For example, constraint checks are indicated, complex aggregates
24009 are replaced with loops and assignments, and tasking primitives
24010 are replaced with run-time calls.
24012 @node Stack Traceback
24013 @section Stack Traceback
24015 @cindex stack traceback
24016 @cindex stack unwinding
24019 Traceback is a mechanism to display the sequence of subprogram calls that
24020 leads to a specified execution point in a program. Often (but not always)
24021 the execution point is an instruction at which an exception has been raised.
24022 This mechanism is also known as @i{stack unwinding} because it obtains
24023 its information by scanning the run-time stack and recovering the activation
24024 records of all active subprograms. Stack unwinding is one of the most
24025 important tools for program debugging.
24027 The first entry stored in traceback corresponds to the deepest calling level,
24028 that is to say the subprogram currently executing the instruction
24029 from which we want to obtain the traceback.
24031 Note that there is no runtime performance penalty when stack traceback
24032 is enabled, and no exception is raised during program execution.
24035 * Non-Symbolic Traceback::
24036 * Symbolic Traceback::
24039 @node Non-Symbolic Traceback
24040 @subsection Non-Symbolic Traceback
24041 @cindex traceback, non-symbolic
24044 Note: this feature is not supported on all platforms. See
24045 @file{GNAT.Traceback spec in g-traceb.ads} for a complete list of supported
24049 * Tracebacks From an Unhandled Exception::
24050 * Tracebacks From Exception Occurrences (non-symbolic)::
24051 * Tracebacks From Anywhere in a Program (non-symbolic)::
24054 @node Tracebacks From an Unhandled Exception
24055 @subsubsection Tracebacks From an Unhandled Exception
24058 A runtime non-symbolic traceback is a list of addresses of call instructions.
24059 To enable this feature you must use the @option{-E}
24060 @code{gnatbind}'s option. With this option a stack traceback is stored as part
24061 of exception information. You can retrieve this information using the
24062 @code{addr2line} tool.
24064 Here is a simple example:
24066 @smallexample @c ada
24072 raise Constraint_Error;
24087 $ gnatmake stb -bargs -E
24090 Execution terminated by unhandled exception
24091 Exception name: CONSTRAINT_ERROR
24093 Call stack traceback locations:
24094 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
24098 As we see the traceback lists a sequence of addresses for the unhandled
24099 exception @code{CONSTRAINT_ERROR} raised in procedure P1. It is easy to
24100 guess that this exception come from procedure P1. To translate these
24101 addresses into the source lines where the calls appear, the
24102 @code{addr2line} tool, described below, is invaluable. The use of this tool
24103 requires the program to be compiled with debug information.
24106 $ gnatmake -g stb -bargs -E
24109 Execution terminated by unhandled exception
24110 Exception name: CONSTRAINT_ERROR
24112 Call stack traceback locations:
24113 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
24115 $ addr2line --exe=stb 0x401373 0x40138b 0x40139c 0x401335 0x4011c4
24116 0x4011f1 0x77e892a4
24118 00401373 at d:/stb/stb.adb:5
24119 0040138B at d:/stb/stb.adb:10
24120 0040139C at d:/stb/stb.adb:14
24121 00401335 at d:/stb/b~stb.adb:104
24122 004011C4 at /build/@dots{}/crt1.c:200
24123 004011F1 at /build/@dots{}/crt1.c:222
24124 77E892A4 in ?? at ??:0
24128 The @code{addr2line} tool has several other useful options:
24132 to get the function name corresponding to any location
24134 @item --demangle=gnat
24135 to use the gnat decoding mode for the function names. Note that
24136 for binutils version 2.9.x the option is simply @option{--demangle}.
24140 $ addr2line --exe=stb --functions --demangle=gnat 0x401373 0x40138b
24141 0x40139c 0x401335 0x4011c4 0x4011f1
24143 00401373 in stb.p1 at d:/stb/stb.adb:5
24144 0040138B in stb.p2 at d:/stb/stb.adb:10
24145 0040139C in stb at d:/stb/stb.adb:14
24146 00401335 in main at d:/stb/b~stb.adb:104
24147 004011C4 in <__mingw_CRTStartup> at /build/@dots{}/crt1.c:200
24148 004011F1 in <mainCRTStartup> at /build/@dots{}/crt1.c:222
24152 From this traceback we can see that the exception was raised in
24153 @file{stb.adb} at line 5, which was reached from a procedure call in
24154 @file{stb.adb} at line 10, and so on. The @file{b~std.adb} is the binder file,
24155 which contains the call to the main program.
24156 @xref{Running gnatbind}. The remaining entries are assorted runtime routines,
24157 and the output will vary from platform to platform.
24159 It is also possible to use @code{GDB} with these traceback addresses to debug
24160 the program. For example, we can break at a given code location, as reported
24161 in the stack traceback:
24167 Furthermore, this feature is not implemented inside Windows DLL. Only
24168 the non-symbolic traceback is reported in this case.
24171 (gdb) break *0x401373
24172 Breakpoint 1 at 0x401373: file stb.adb, line 5.
24176 It is important to note that the stack traceback addresses
24177 do not change when debug information is included. This is particularly useful
24178 because it makes it possible to release software without debug information (to
24179 minimize object size), get a field report that includes a stack traceback
24180 whenever an internal bug occurs, and then be able to retrieve the sequence
24181 of calls with the same program compiled with debug information.
24183 @node Tracebacks From Exception Occurrences (non-symbolic)
24184 @subsubsection Tracebacks From Exception Occurrences
24187 Non-symbolic tracebacks are obtained by using the @option{-E} binder argument.
24188 The stack traceback is attached to the exception information string, and can
24189 be retrieved in an exception handler within the Ada program, by means of the
24190 Ada facilities defined in @code{Ada.Exceptions}. Here is a simple example:
24192 @smallexample @c ada
24194 with Ada.Exceptions;
24199 use Ada.Exceptions;
24207 Text_IO.Put_Line (Exception_Information (E));
24221 This program will output:
24226 Exception name: CONSTRAINT_ERROR
24227 Message: stb.adb:12
24228 Call stack traceback locations:
24229 0x4015e4 0x401633 0x401644 0x401461 0x4011c4 0x4011f1 0x77e892a4
24232 @node Tracebacks From Anywhere in a Program (non-symbolic)
24233 @subsubsection Tracebacks From Anywhere in a Program
24236 It is also possible to retrieve a stack traceback from anywhere in a
24237 program. For this you need to
24238 use the @code{GNAT.Traceback} API. This package includes a procedure called
24239 @code{Call_Chain} that computes a complete stack traceback, as well as useful
24240 display procedures described below. It is not necessary to use the
24241 @option{-E gnatbind} option in this case, because the stack traceback mechanism
24242 is invoked explicitly.
24245 In the following example we compute a traceback at a specific location in
24246 the program, and we display it using @code{GNAT.Debug_Utilities.Image} to
24247 convert addresses to strings:
24249 @smallexample @c ada
24251 with GNAT.Traceback;
24252 with GNAT.Debug_Utilities;
24258 use GNAT.Traceback;
24261 TB : Tracebacks_Array (1 .. 10);
24262 -- We are asking for a maximum of 10 stack frames.
24264 -- Len will receive the actual number of stack frames returned.
24266 Call_Chain (TB, Len);
24268 Text_IO.Put ("In STB.P1 : ");
24270 for K in 1 .. Len loop
24271 Text_IO.Put (Debug_Utilities.Image (TB (K)));
24292 In STB.P1 : 16#0040_F1E4# 16#0040_14F2# 16#0040_170B# 16#0040_171C#
24293 16#0040_1461# 16#0040_11C4# 16#0040_11F1# 16#77E8_92A4#
24297 You can then get further information by invoking the @code{addr2line}
24298 tool as described earlier (note that the hexadecimal addresses
24299 need to be specified in C format, with a leading ``0x'').
24301 @node Symbolic Traceback
24302 @subsection Symbolic Traceback
24303 @cindex traceback, symbolic
24306 A symbolic traceback is a stack traceback in which procedure names are
24307 associated with each code location.
24310 Note that this feature is not supported on all platforms. See
24311 @file{GNAT.Traceback.Symbolic spec in g-trasym.ads} for a complete
24312 list of currently supported platforms.
24315 Note that the symbolic traceback requires that the program be compiled
24316 with debug information. If it is not compiled with debug information
24317 only the non-symbolic information will be valid.
24320 * Tracebacks From Exception Occurrences (symbolic)::
24321 * Tracebacks From Anywhere in a Program (symbolic)::
24324 @node Tracebacks From Exception Occurrences (symbolic)
24325 @subsubsection Tracebacks From Exception Occurrences
24327 @smallexample @c ada
24329 with GNAT.Traceback.Symbolic;
24335 raise Constraint_Error;
24352 Ada.Text_IO.Put_Line (GNAT.Traceback.Symbolic.Symbolic_Traceback (E));
24357 $ gnatmake -g .\stb -bargs -E -largs -lgnat -laddr2line -lintl
24360 0040149F in stb.p1 at stb.adb:8
24361 004014B7 in stb.p2 at stb.adb:13
24362 004014CF in stb.p3 at stb.adb:18
24363 004015DD in ada.stb at stb.adb:22
24364 00401461 in main at b~stb.adb:168
24365 004011C4 in __mingw_CRTStartup at crt1.c:200
24366 004011F1 in mainCRTStartup at crt1.c:222
24367 77E892A4 in ?? at ??:0
24371 In the above example the ``.\'' syntax in the @command{gnatmake} command
24372 is currently required by @command{addr2line} for files that are in
24373 the current working directory.
24374 Moreover, the exact sequence of linker options may vary from platform
24376 The above @option{-largs} section is for Windows platforms. By contrast,
24377 under Unix there is no need for the @option{-largs} section.
24378 Differences across platforms are due to details of linker implementation.
24380 @node Tracebacks From Anywhere in a Program (symbolic)
24381 @subsubsection Tracebacks From Anywhere in a Program
24384 It is possible to get a symbolic stack traceback
24385 from anywhere in a program, just as for non-symbolic tracebacks.
24386 The first step is to obtain a non-symbolic
24387 traceback, and then call @code{Symbolic_Traceback} to compute the symbolic
24388 information. Here is an example:
24390 @smallexample @c ada
24392 with GNAT.Traceback;
24393 with GNAT.Traceback.Symbolic;
24398 use GNAT.Traceback;
24399 use GNAT.Traceback.Symbolic;
24402 TB : Tracebacks_Array (1 .. 10);
24403 -- We are asking for a maximum of 10 stack frames.
24405 -- Len will receive the actual number of stack frames returned.
24407 Call_Chain (TB, Len);
24408 Text_IO.Put_Line (Symbolic_Traceback (TB (1 .. Len)));
24421 @c ******************************
24423 @node Compatibility with HP Ada
24424 @chapter Compatibility with HP Ada
24425 @cindex Compatibility
24430 @cindex Compatibility between GNAT and HP Ada
24431 This chapter compares HP Ada (formerly known as ``DEC Ada'')
24432 for OpenVMS Alpha and GNAT for OpenVMS for Alpha and for I64.
24433 GNAT is highly compatible
24434 with HP Ada, and it should generally be straightforward to port code
24435 from the HP Ada environment to GNAT. However, there are a few language
24436 and implementation differences of which the user must be aware. These
24437 differences are discussed in this chapter. In
24438 addition, the operating environment and command structure for the
24439 compiler are different, and these differences are also discussed.
24441 For further details on these and other compatibility issues,
24442 see Appendix E of the HP publication
24443 @cite{HP Ada, Technical Overview and Comparison on HP Platforms}.
24445 Except where otherwise indicated, the description of GNAT for OpenVMS
24446 applies to both the Alpha and I64 platforms.
24448 For information on porting Ada code from GNAT on Alpha OpenVMS to GNAT on
24449 I64 OpenVMS, see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
24451 The discussion in this chapter addresses specifically the implementation
24452 of Ada 83 for HP OpenVMS Alpha Systems. In cases where the implementation
24453 of HP Ada differs between OpenVMS Alpha Systems and OpenVMS VAX Systems,
24454 GNAT always follows the Alpha implementation.
24456 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
24457 attributes are recognized, although only a subset of them can sensibly
24458 be implemented. The description of pragmas in
24459 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual}
24460 indicates whether or not they are applicable to non-VMS systems.
24463 * Ada Language Compatibility::
24464 * Differences in the Definition of Package System::
24465 * Language-Related Features::
24466 * The Package STANDARD::
24467 * The Package SYSTEM::
24468 * Tasking and Task-Related Features::
24469 * Pragmas and Pragma-Related Features::
24470 * Library of Predefined Units::
24472 * Main Program Definition::
24473 * Implementation-Defined Attributes::
24474 * Compiler and Run-Time Interfacing::
24475 * Program Compilation and Library Management::
24477 * Implementation Limits::
24478 * Tools and Utilities::
24481 @node Ada Language Compatibility
24482 @section Ada Language Compatibility
24485 GNAT handles Ada 95 and Ada 2005 as well as Ada 83, whereas HP Ada is only
24486 for Ada 83. Ada 95 and Ada 2005 are almost completely upwards compatible
24487 with Ada 83, and therefore Ada 83 programs will compile
24488 and run under GNAT with
24489 no changes or only minor changes. The @cite{Annotated Ada Reference Manual}
24490 provides details on specific incompatibilities.
24492 GNAT provides the switch @option{/83} on the @command{GNAT COMPILE} command,
24493 as well as the pragma @code{ADA_83}, to force the compiler to
24494 operate in Ada 83 mode. This mode does not guarantee complete
24495 conformance to Ada 83, but in practice is sufficient to
24496 eliminate most sources of incompatibilities.
24497 In particular, it eliminates the recognition of the
24498 additional Ada 95 and Ada 2005 keywords, so that their use as identifiers
24499 in Ada 83 programs is legal, and handles the cases of packages
24500 with optional bodies, and generics that instantiate unconstrained
24501 types without the use of @code{(<>)}.
24503 @node Differences in the Definition of Package System
24504 @section Differences in the Definition of Package @code{System}
24507 An Ada compiler is allowed to add
24508 implementation-dependent declarations to package @code{System}.
24510 GNAT does not take advantage of this permission, and the version of
24511 @code{System} provided by GNAT exactly matches that defined in the Ada
24514 However, HP Ada adds an extensive set of declarations to package
24516 as fully documented in the HP Ada manuals. To minimize changes required
24517 for programs that make use of these extensions, GNAT provides the pragma
24518 @code{Extend_System} for extending the definition of package System. By using:
24519 @cindex pragma @code{Extend_System}
24520 @cindex @code{Extend_System} pragma
24522 @smallexample @c ada
24525 pragma Extend_System (Aux_DEC);
24531 the set of definitions in @code{System} is extended to include those in
24532 package @code{System.Aux_DEC}.
24533 @cindex @code{System.Aux_DEC} package
24534 @cindex @code{Aux_DEC} package (child of @code{System})
24535 These definitions are incorporated directly into package @code{System},
24536 as though they had been declared there. For a
24537 list of the declarations added, see the spec of this package,
24538 which can be found in the file @file{s-auxdec.ads} in the GNAT library.
24539 @cindex @file{s-auxdec.ads} file
24540 The pragma @code{Extend_System} is a configuration pragma, which means that
24541 it can be placed in the file @file{gnat.adc}, so that it will automatically
24542 apply to all subsequent compilations. See @ref{Configuration Pragmas},
24543 for further details.
24545 An alternative approach that avoids the use of the non-standard
24546 @code{Extend_System} pragma is to add a context clause to the unit that
24547 references these facilities:
24549 @smallexample @c ada
24551 with System.Aux_DEC;
24552 use System.Aux_DEC;
24557 The effect is not quite semantically identical to incorporating
24558 the declarations directly into package @code{System},
24559 but most programs will not notice a difference
24560 unless they use prefix notation (e.g.@: @code{System.Integer_8})
24561 to reference the entities directly in package @code{System}.
24562 For units containing such references,
24563 the prefixes must either be removed, or the pragma @code{Extend_System}
24566 @node Language-Related Features
24567 @section Language-Related Features
24570 The following sections highlight differences in types,
24571 representations of types, operations, alignment, and
24575 * Integer Types and Representations::
24576 * Floating-Point Types and Representations::
24577 * Pragmas Float_Representation and Long_Float::
24578 * Fixed-Point Types and Representations::
24579 * Record and Array Component Alignment::
24580 * Address Clauses::
24581 * Other Representation Clauses::
24584 @node Integer Types and Representations
24585 @subsection Integer Types and Representations
24588 The set of predefined integer types is identical in HP Ada and GNAT.
24589 Furthermore the representation of these integer types is also identical,
24590 including the capability of size clauses forcing biased representation.
24593 HP Ada for OpenVMS Alpha systems has defined the
24594 following additional integer types in package @code{System}:
24611 @code{LARGEST_INTEGER}
24615 In GNAT, the first four of these types may be obtained from the
24616 standard Ada package @code{Interfaces}.
24617 Alternatively, by use of the pragma @code{Extend_System}, identical
24618 declarations can be referenced directly in package @code{System}.
24619 On both GNAT and HP Ada, the maximum integer size is 64 bits.
24621 @node Floating-Point Types and Representations
24622 @subsection Floating-Point Types and Representations
24623 @cindex Floating-Point types
24626 The set of predefined floating-point types is identical in HP Ada and GNAT.
24627 Furthermore the representation of these floating-point
24628 types is also identical. One important difference is that the default
24629 representation for HP Ada is @code{VAX_Float}, but the default representation
24632 Specific types may be declared to be @code{VAX_Float} or IEEE, using the
24633 pragma @code{Float_Representation} as described in the HP Ada
24635 For example, the declarations:
24637 @smallexample @c ada
24639 type F_Float is digits 6;
24640 pragma Float_Representation (VAX_Float, F_Float);
24645 declares a type @code{F_Float} that will be represented in @code{VAX_Float}
24647 This set of declarations actually appears in @code{System.Aux_DEC},
24649 the full set of additional floating-point declarations provided in
24650 the HP Ada version of package @code{System}.
24651 This and similar declarations may be accessed in a user program
24652 by using pragma @code{Extend_System}. The use of this
24653 pragma, and the related pragma @code{Long_Float} is described in further
24654 detail in the following section.
24656 @node Pragmas Float_Representation and Long_Float
24657 @subsection Pragmas @code{Float_Representation} and @code{Long_Float}
24660 HP Ada provides the pragma @code{Float_Representation}, which
24661 acts as a program library switch to allow control over
24662 the internal representation chosen for the predefined
24663 floating-point types declared in the package @code{Standard}.
24664 The format of this pragma is as follows:
24666 @smallexample @c ada
24668 pragma Float_Representation(VAX_Float | IEEE_Float);
24673 This pragma controls the representation of floating-point
24678 @code{VAX_Float} specifies that floating-point
24679 types are represented by default with the VAX system hardware types
24680 @code{F-floating}, @code{D-floating}, @code{G-floating}.
24681 Note that the @code{H-floating}
24682 type was available only on VAX systems, and is not available
24683 in either HP Ada or GNAT.
24686 @code{IEEE_Float} specifies that floating-point
24687 types are represented by default with the IEEE single and
24688 double floating-point types.
24692 GNAT provides an identical implementation of the pragma
24693 @code{Float_Representation}, except that it functions as a
24694 configuration pragma. Note that the
24695 notion of configuration pragma corresponds closely to the
24696 HP Ada notion of a program library switch.
24698 When no pragma is used in GNAT, the default is @code{IEEE_Float},
24700 from HP Ada 83, where the default is @code{VAX_Float}. In addition, the
24701 predefined libraries in GNAT are built using @code{IEEE_Float}, so it is not
24702 advisable to change the format of numbers passed to standard library
24703 routines, and if necessary explicit type conversions may be needed.
24705 The use of @code{IEEE_Float} is recommended in GNAT since it is more
24706 efficient, and (given that it conforms to an international standard)
24707 potentially more portable.
24708 The situation in which @code{VAX_Float} may be useful is in interfacing
24709 to existing code and data that expect the use of @code{VAX_Float}.
24710 In such a situation use the predefined @code{VAX_Float}
24711 types in package @code{System}, as extended by
24712 @code{Extend_System}. For example, use @code{System.F_Float}
24713 to specify the 32-bit @code{F-Float} format.
24716 On OpenVMS systems, HP Ada provides the pragma @code{Long_Float}
24717 to allow control over the internal representation chosen
24718 for the predefined type @code{Long_Float} and for floating-point
24719 type declarations with digits specified in the range 7 .. 15.
24720 The format of this pragma is as follows:
24722 @smallexample @c ada
24724 pragma Long_Float (D_FLOAT | G_FLOAT);
24728 @node Fixed-Point Types and Representations
24729 @subsection Fixed-Point Types and Representations
24732 On HP Ada for OpenVMS Alpha systems, rounding is
24733 away from zero for both positive and negative numbers.
24734 Therefore, @code{+0.5} rounds to @code{1},
24735 and @code{-0.5} rounds to @code{-1}.
24737 On GNAT the results of operations
24738 on fixed-point types are in accordance with the Ada
24739 rules. In particular, results of operations on decimal
24740 fixed-point types are truncated.
24742 @node Record and Array Component Alignment
24743 @subsection Record and Array Component Alignment
24746 On HP Ada for OpenVMS Alpha, all non-composite components
24747 are aligned on natural boundaries. For example, 1-byte
24748 components are aligned on byte boundaries, 2-byte
24749 components on 2-byte boundaries, 4-byte components on 4-byte
24750 byte boundaries, and so on. The OpenVMS Alpha hardware
24751 runs more efficiently with naturally aligned data.
24753 On GNAT, alignment rules are compatible
24754 with HP Ada for OpenVMS Alpha.
24756 @node Address Clauses
24757 @subsection Address Clauses
24760 In HP Ada and GNAT, address clauses are supported for
24761 objects and imported subprograms.
24762 The predefined type @code{System.Address} is a private type
24763 in both compilers on Alpha OpenVMS, with the same representation
24764 (it is simply a machine pointer). Addition, subtraction, and comparison
24765 operations are available in the standard Ada package
24766 @code{System.Storage_Elements}, or in package @code{System}
24767 if it is extended to include @code{System.Aux_DEC} using a
24768 pragma @code{Extend_System} as previously described.
24770 Note that code that @code{with}'s both this extended package @code{System}
24771 and the package @code{System.Storage_Elements} should not @code{use}
24772 both packages, or ambiguities will result. In general it is better
24773 not to mix these two sets of facilities. The Ada package was
24774 designed specifically to provide the kind of features that HP Ada
24775 adds directly to package @code{System}.
24777 The type @code{System.Address} is a 64-bit integer type in GNAT for
24778 I64 OpenVMS. For more information,
24779 see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
24781 GNAT is compatible with HP Ada in its handling of address
24782 clauses, except for some limitations in
24783 the form of address clauses for composite objects with
24784 initialization. Such address clauses are easily replaced
24785 by the use of an explicitly-defined constant as described
24786 in the Ada Reference Manual (13.1(22)). For example, the sequence
24789 @smallexample @c ada
24791 X, Y : Integer := Init_Func;
24792 Q : String (X .. Y) := "abc";
24794 for Q'Address use Compute_Address;
24799 will be rejected by GNAT, since the address cannot be computed at the time
24800 that @code{Q} is declared. To achieve the intended effect, write instead:
24802 @smallexample @c ada
24805 X, Y : Integer := Init_Func;
24806 Q_Address : constant Address := Compute_Address;
24807 Q : String (X .. Y) := "abc";
24809 for Q'Address use Q_Address;
24815 which will be accepted by GNAT (and other Ada compilers), and is also
24816 compatible with Ada 83. A fuller description of the restrictions
24817 on address specifications is found in @ref{Top, GNAT Reference Manual,
24818 About This Guide, gnat_rm, GNAT Reference Manual}.
24820 @node Other Representation Clauses
24821 @subsection Other Representation Clauses
24824 GNAT implements in a compatible manner all the representation
24825 clauses supported by HP Ada. In addition, GNAT
24826 implements the representation clause forms that were introduced in Ada 95,
24827 including @code{COMPONENT_SIZE} and @code{SIZE} clauses for objects.
24829 @node The Package STANDARD
24830 @section The Package @code{STANDARD}
24833 The package @code{STANDARD}, as implemented by HP Ada, is fully
24834 described in the @cite{Ada Reference Manual} and in the
24835 @cite{HP Ada Language Reference Manual}. As implemented by GNAT, the
24836 package @code{STANDARD} is described in the @cite{Ada Reference Manual}.
24838 In addition, HP Ada supports the Latin-1 character set in
24839 the type @code{CHARACTER}. GNAT supports the Latin-1 character set
24840 in the type @code{CHARACTER} and also Unicode (ISO 10646 BMP) in
24841 the type @code{WIDE_CHARACTER}.
24843 The floating-point types supported by GNAT are those
24844 supported by HP Ada, but the defaults are different, and are controlled by
24845 pragmas. See @ref{Floating-Point Types and Representations}, for details.
24847 @node The Package SYSTEM
24848 @section The Package @code{SYSTEM}
24851 HP Ada provides a specific version of the package
24852 @code{SYSTEM} for each platform on which the language is implemented.
24853 For the complete spec of the package @code{SYSTEM}, see
24854 Appendix F of the @cite{HP Ada Language Reference Manual}.
24856 On HP Ada, the package @code{SYSTEM} includes the following conversion
24859 @item @code{TO_ADDRESS(INTEGER)}
24861 @item @code{TO_ADDRESS(UNSIGNED_LONGWORD)}
24863 @item @code{TO_ADDRESS(}@i{universal_integer}@code{)}
24865 @item @code{TO_INTEGER(ADDRESS)}
24867 @item @code{TO_UNSIGNED_LONGWORD(ADDRESS)}
24869 @item Function @code{IMPORT_VALUE return UNSIGNED_LONGWORD} and the
24870 functions @code{IMPORT_ADDRESS} and @code{IMPORT_LARGEST_VALUE}
24874 By default, GNAT supplies a version of @code{SYSTEM} that matches
24875 the definition given in the @cite{Ada Reference Manual}.
24877 is a subset of the HP system definitions, which is as
24878 close as possible to the original definitions. The only difference
24879 is that the definition of @code{SYSTEM_NAME} is different:
24881 @smallexample @c ada
24883 type Name is (SYSTEM_NAME_GNAT);
24884 System_Name : constant Name := SYSTEM_NAME_GNAT;
24889 Also, GNAT adds the Ada declarations for
24890 @code{BIT_ORDER} and @code{DEFAULT_BIT_ORDER}.
24892 However, the use of the following pragma causes GNAT
24893 to extend the definition of package @code{SYSTEM} so that it
24894 encompasses the full set of HP-specific extensions,
24895 including the functions listed above:
24897 @smallexample @c ada
24899 pragma Extend_System (Aux_DEC);
24904 The pragma @code{Extend_System} is a configuration pragma that
24905 is most conveniently placed in the @file{gnat.adc} file. @xref{Pragma
24906 Extend_System,,, gnat_rm, GNAT Reference Manual} for further details.
24908 HP Ada does not allow the recompilation of the package
24909 @code{SYSTEM}. Instead HP Ada provides several pragmas
24910 (@code{SYSTEM_NAME}, @code{STORAGE_UNIT}, and @code{MEMORY_SIZE})
24911 to modify values in the package @code{SYSTEM}.
24912 On OpenVMS Alpha systems, the pragma
24913 @code{SYSTEM_NAME} takes the enumeration literal @code{OPENVMS_AXP} as
24914 its single argument.
24916 GNAT does permit the recompilation of package @code{SYSTEM} using
24917 the special switch @option{-gnatg}, and this switch can be used if
24918 it is necessary to modify the definitions in @code{SYSTEM}. GNAT does
24919 not permit the specification of @code{SYSTEM_NAME}, @code{STORAGE_UNIT}
24920 or @code{MEMORY_SIZE} by any other means.
24922 On GNAT systems, the pragma @code{SYSTEM_NAME} takes the
24923 enumeration literal @code{SYSTEM_NAME_GNAT}.
24925 The definitions provided by the use of
24927 @smallexample @c ada
24928 pragma Extend_System (AUX_Dec);
24932 are virtually identical to those provided by the HP Ada 83 package
24933 @code{SYSTEM}. One important difference is that the name of the
24935 function for type @code{UNSIGNED_LONGWORD} is changed to
24936 @code{TO_ADDRESS_LONG}.
24937 @xref{Address Clauses,,, gnat_rm, GNAT Reference Manual} for a
24938 discussion of why this change was necessary.
24941 The version of @code{TO_ADDRESS} taking a @i{universal_integer} argument
24943 an extension to Ada 83 not strictly compatible with the reference manual.
24944 GNAT, in order to be exactly compatible with the standard,
24945 does not provide this capability. In HP Ada 83, the
24946 point of this definition is to deal with a call like:
24948 @smallexample @c ada
24949 TO_ADDRESS (16#12777#);
24953 Normally, according to Ada 83 semantics, one would expect this to be
24954 ambiguous, since it matches both the @code{INTEGER} and
24955 @code{UNSIGNED_LONGWORD} forms of @code{TO_ADDRESS}.
24956 However, in HP Ada 83, there is no ambiguity, since the
24957 definition using @i{universal_integer} takes precedence.
24959 In GNAT, since the version with @i{universal_integer} cannot be supplied,
24961 not possible to be 100% compatible. Since there are many programs using
24962 numeric constants for the argument to @code{TO_ADDRESS}, the decision in
24964 to change the name of the function in the @code{UNSIGNED_LONGWORD} case,
24965 so the declarations provided in the GNAT version of @code{AUX_Dec} are:
24967 @smallexample @c ada
24968 function To_Address (X : Integer) return Address;
24969 pragma Pure_Function (To_Address);
24971 function To_Address_Long (X : Unsigned_Longword) return Address;
24972 pragma Pure_Function (To_Address_Long);
24976 This means that programs using @code{TO_ADDRESS} for
24977 @code{UNSIGNED_LONGWORD} must change the name to @code{TO_ADDRESS_LONG}.
24979 @node Tasking and Task-Related Features
24980 @section Tasking and Task-Related Features
24983 This section compares the treatment of tasking in GNAT
24984 and in HP Ada for OpenVMS Alpha.
24985 The GNAT description applies to both Alpha and I64 OpenVMS.
24986 For detailed information on tasking in
24987 HP Ada, see the @cite{HP Ada Language Reference Manual} and the
24988 relevant run-time reference manual.
24991 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
24992 * Assigning Task IDs::
24993 * Task IDs and Delays::
24994 * Task-Related Pragmas::
24995 * Scheduling and Task Priority::
24997 * External Interrupts::
25000 @node Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
25001 @subsection Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
25004 On OpenVMS Alpha systems, each Ada task (except a passive
25005 task) is implemented as a single stream of execution
25006 that is created and managed by the kernel. On these
25007 systems, HP Ada tasking support is based on DECthreads,
25008 an implementation of the POSIX standard for threads.
25010 Also, on OpenVMS Alpha systems, HP Ada tasks and foreign
25011 code that calls DECthreads routines can be used together.
25012 The interaction between Ada tasks and DECthreads routines
25013 can have some benefits. For example when on OpenVMS Alpha,
25014 HP Ada can call C code that is already threaded.
25016 GNAT uses the facilities of DECthreads,
25017 and Ada tasks are mapped to threads.
25019 @node Assigning Task IDs
25020 @subsection Assigning Task IDs
25023 The HP Ada Run-Time Library always assigns @code{%TASK 1} to
25024 the environment task that executes the main program. On
25025 OpenVMS Alpha systems, @code{%TASK 0} is often used for tasks
25026 that have been created but are not yet activated.
25028 On OpenVMS Alpha systems, task IDs are assigned at
25029 activation. On GNAT systems, task IDs are also assigned at
25030 task creation but do not have the same form or values as
25031 task ID values in HP Ada. There is no null task, and the
25032 environment task does not have a specific task ID value.
25034 @node Task IDs and Delays
25035 @subsection Task IDs and Delays
25038 On OpenVMS Alpha systems, tasking delays are implemented
25039 using Timer System Services. The Task ID is used for the
25040 identification of the timer request (the @code{REQIDT} parameter).
25041 If Timers are used in the application take care not to use
25042 @code{0} for the identification, because cancelling such a timer
25043 will cancel all timers and may lead to unpredictable results.
25045 @node Task-Related Pragmas
25046 @subsection Task-Related Pragmas
25049 Ada supplies the pragma @code{TASK_STORAGE}, which allows
25050 specification of the size of the guard area for a task
25051 stack. (The guard area forms an area of memory that has no
25052 read or write access and thus helps in the detection of
25053 stack overflow.) On OpenVMS Alpha systems, if the pragma
25054 @code{TASK_STORAGE} specifies a value of zero, a minimal guard
25055 area is created. In the absence of a pragma @code{TASK_STORAGE},
25056 a default guard area is created.
25058 GNAT supplies the following task-related pragmas:
25061 @item @code{TASK_INFO}
25063 This pragma appears within a task definition and
25064 applies to the task in which it appears. The argument
25065 must be of type @code{SYSTEM.TASK_INFO.TASK_INFO_TYPE}.
25067 @item @code{TASK_STORAGE}
25069 GNAT implements pragma @code{TASK_STORAGE} in the same way as HP Ada.
25070 Both HP Ada and GNAT supply the pragmas @code{PASSIVE},
25071 @code{SUPPRESS}, and @code{VOLATILE}.
25073 @node Scheduling and Task Priority
25074 @subsection Scheduling and Task Priority
25077 HP Ada implements the Ada language requirement that
25078 when two tasks are eligible for execution and they have
25079 different priorities, the lower priority task does not
25080 execute while the higher priority task is waiting. The HP
25081 Ada Run-Time Library keeps a task running until either the
25082 task is suspended or a higher priority task becomes ready.
25084 On OpenVMS Alpha systems, the default strategy is round-
25085 robin with preemption. Tasks of equal priority take turns
25086 at the processor. A task is run for a certain period of
25087 time and then placed at the tail of the ready queue for
25088 its priority level.
25090 HP Ada provides the implementation-defined pragma @code{TIME_SLICE},
25091 which can be used to enable or disable round-robin
25092 scheduling of tasks with the same priority.
25093 See the relevant HP Ada run-time reference manual for
25094 information on using the pragmas to control HP Ada task
25097 GNAT follows the scheduling rules of Annex D (Real-Time
25098 Annex) of the @cite{Ada Reference Manual}. In general, this
25099 scheduling strategy is fully compatible with HP Ada
25100 although it provides some additional constraints (as
25101 fully documented in Annex D).
25102 GNAT implements time slicing control in a manner compatible with
25103 HP Ada 83, by means of the pragma @code{Time_Slice}, whose semantics
25104 are identical to the HP Ada 83 pragma of the same name.
25105 Note that it is not possible to mix GNAT tasking and
25106 HP Ada 83 tasking in the same program, since the two run-time
25107 libraries are not compatible.
25109 @node The Task Stack
25110 @subsection The Task Stack
25113 In HP Ada, a task stack is allocated each time a
25114 non-passive task is activated. As soon as the task is
25115 terminated, the storage for the task stack is deallocated.
25116 If you specify a size of zero (bytes) with @code{T'STORAGE_SIZE},
25117 a default stack size is used. Also, regardless of the size
25118 specified, some additional space is allocated for task
25119 management purposes. On OpenVMS Alpha systems, at least
25120 one page is allocated.
25122 GNAT handles task stacks in a similar manner. In accordance with
25123 the Ada rules, it provides the pragma @code{STORAGE_SIZE} as
25124 an alternative method for controlling the task stack size.
25125 The specification of the attribute @code{T'STORAGE_SIZE} is also
25126 supported in a manner compatible with HP Ada.
25128 @node External Interrupts
25129 @subsection External Interrupts
25132 On HP Ada, external interrupts can be associated with task entries.
25133 GNAT is compatible with HP Ada in its handling of external interrupts.
25135 @node Pragmas and Pragma-Related Features
25136 @section Pragmas and Pragma-Related Features
25139 Both HP Ada and GNAT supply all language-defined pragmas
25140 as specified by the Ada 83 standard. GNAT also supplies all
25141 language-defined pragmas introduced by Ada 95 and Ada 2005.
25142 In addition, GNAT implements the implementation-defined pragmas
25146 @item @code{AST_ENTRY}
25148 @item @code{COMMON_OBJECT}
25150 @item @code{COMPONENT_ALIGNMENT}
25152 @item @code{EXPORT_EXCEPTION}
25154 @item @code{EXPORT_FUNCTION}
25156 @item @code{EXPORT_OBJECT}
25158 @item @code{EXPORT_PROCEDURE}
25160 @item @code{EXPORT_VALUED_PROCEDURE}
25162 @item @code{FLOAT_REPRESENTATION}
25166 @item @code{IMPORT_EXCEPTION}
25168 @item @code{IMPORT_FUNCTION}
25170 @item @code{IMPORT_OBJECT}
25172 @item @code{IMPORT_PROCEDURE}
25174 @item @code{IMPORT_VALUED_PROCEDURE}
25176 @item @code{INLINE_GENERIC}
25178 @item @code{INTERFACE_NAME}
25180 @item @code{LONG_FLOAT}
25182 @item @code{MAIN_STORAGE}
25184 @item @code{PASSIVE}
25186 @item @code{PSECT_OBJECT}
25188 @item @code{SHARE_GENERIC}
25190 @item @code{SUPPRESS_ALL}
25192 @item @code{TASK_STORAGE}
25194 @item @code{TIME_SLICE}
25200 These pragmas are all fully implemented, with the exception of @code{TITLE},
25201 @code{PASSIVE}, and @code{SHARE_GENERIC}, which are
25202 recognized, but which have no
25203 effect in GNAT. The effect of @code{PASSIVE} may be obtained by the
25204 use of Ada protected objects. In GNAT, all generics are inlined.
25206 Unlike HP Ada, the GNAT ``@code{EXPORT_}@i{subprogram}'' pragmas require
25207 a separate subprogram specification which must appear before the
25210 GNAT also supplies a number of implementation-defined pragmas as follows:
25212 @item @code{ABORT_DEFER}
25214 @item @code{ADA_83}
25216 @item @code{ADA_95}
25218 @item @code{ADA_05}
25220 @item @code{ANNOTATE}
25222 @item @code{ASSERT}
25224 @item @code{C_PASS_BY_COPY}
25226 @item @code{CPP_CLASS}
25228 @item @code{CPP_CONSTRUCTOR}
25230 @item @code{CPP_DESTRUCTOR}
25234 @item @code{EXTEND_SYSTEM}
25236 @item @code{LINKER_ALIAS}
25238 @item @code{LINKER_SECTION}
25240 @item @code{MACHINE_ATTRIBUTE}
25242 @item @code{NO_RETURN}
25244 @item @code{PURE_FUNCTION}
25246 @item @code{SOURCE_FILE_NAME}
25248 @item @code{SOURCE_REFERENCE}
25250 @item @code{TASK_INFO}
25252 @item @code{UNCHECKED_UNION}
25254 @item @code{UNIMPLEMENTED_UNIT}
25256 @item @code{UNIVERSAL_DATA}
25258 @item @code{UNSUPPRESS}
25260 @item @code{WARNINGS}
25262 @item @code{WEAK_EXTERNAL}
25266 For full details on these GNAT implementation-defined pragmas,
25267 see @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference
25271 * Restrictions on the Pragma INLINE::
25272 * Restrictions on the Pragma INTERFACE::
25273 * Restrictions on the Pragma SYSTEM_NAME::
25276 @node Restrictions on the Pragma INLINE
25277 @subsection Restrictions on Pragma @code{INLINE}
25280 HP Ada enforces the following restrictions on the pragma @code{INLINE}:
25282 @item Parameters cannot have a task type.
25284 @item Function results cannot be task types, unconstrained
25285 array types, or unconstrained types with discriminants.
25287 @item Bodies cannot declare the following:
25289 @item Subprogram body or stub (imported subprogram is allowed)
25293 @item Generic declarations
25295 @item Instantiations
25299 @item Access types (types derived from access types allowed)
25301 @item Array or record types
25303 @item Dependent tasks
25305 @item Direct recursive calls of subprogram or containing
25306 subprogram, directly or via a renaming
25312 In GNAT, the only restriction on pragma @code{INLINE} is that the
25313 body must occur before the call if both are in the same
25314 unit, and the size must be appropriately small. There are
25315 no other specific restrictions which cause subprograms to
25316 be incapable of being inlined.
25318 @node Restrictions on the Pragma INTERFACE
25319 @subsection Restrictions on Pragma @code{INTERFACE}
25322 The following restrictions on pragma @code{INTERFACE}
25323 are enforced by both HP Ada and GNAT:
25325 @item Languages accepted: Ada, Bliss, C, Fortran, Default.
25326 Default is the default on OpenVMS Alpha systems.
25328 @item Parameter passing: Language specifies default
25329 mechanisms but can be overridden with an @code{EXPORT} pragma.
25332 @item Ada: Use internal Ada rules.
25334 @item Bliss, C: Parameters must be mode @code{in}; cannot be
25335 record or task type. Result cannot be a string, an
25336 array, or a record.
25338 @item Fortran: Parameters cannot have a task type. Result cannot
25339 be a string, an array, or a record.
25344 GNAT is entirely upwards compatible with HP Ada, and in addition allows
25345 record parameters for all languages.
25347 @node Restrictions on the Pragma SYSTEM_NAME
25348 @subsection Restrictions on Pragma @code{SYSTEM_NAME}
25351 For HP Ada for OpenVMS Alpha, the enumeration literal
25352 for the type @code{NAME} is @code{OPENVMS_AXP}.
25353 In GNAT, the enumeration
25354 literal for the type @code{NAME} is @code{SYSTEM_NAME_GNAT}.
25356 @node Library of Predefined Units
25357 @section Library of Predefined Units
25360 A library of predefined units is provided as part of the
25361 HP Ada and GNAT implementations. HP Ada does not provide
25362 the package @code{MACHINE_CODE} but instead recommends importing
25365 The GNAT versions of the HP Ada Run-Time Library (@code{ADA$PREDEFINED:})
25366 units are taken from the OpenVMS Alpha version, not the OpenVMS VAX
25368 The HP Ada Predefined Library units are modified to remove post-Ada 83
25369 incompatibilities and to make them interoperable with GNAT
25370 (@pxref{Changes to DECLIB}, for details).
25371 The units are located in the @file{DECLIB} directory.
25373 The GNAT RTL is contained in
25374 the @file{ADALIB} directory, and
25375 the default search path is set up to find @code{DECLIB} units in preference
25376 to @code{ADALIB} units with the same name (@code{TEXT_IO},
25377 @code{SEQUENTIAL_IO}, and @code{DIRECT_IO}, for example).
25380 * Changes to DECLIB::
25383 @node Changes to DECLIB
25384 @subsection Changes to @code{DECLIB}
25387 The changes made to the HP Ada predefined library for GNAT and post-Ada 83
25388 compatibility are minor and include the following:
25391 @item Adjusting the location of pragmas and record representation
25392 clauses to obey Ada 95 (and thus Ada 2005) rules
25394 @item Adding the proper notation to generic formal parameters
25395 that take unconstrained types in instantiation
25397 @item Adding pragma @code{ELABORATE_BODY} to package specs
25398 that have package bodies not otherwise allowed
25400 @item Replacing occurrences of the identifier ``@code{PROTECTED}'' by
25401 ``@code{PROTECTD}''.
25402 Currently these are found only in the @code{STARLET} package spec.
25404 @item Changing @code{SYSTEM.ADDRESS} to @code{SYSTEM.SHORT_ADDRESS}
25405 where the address size is constrained to 32 bits.
25409 None of the above changes is visible to users.
25415 On OpenVMS Alpha, HP Ada provides the following strongly-typed bindings:
25418 @item Command Language Interpreter (CLI interface)
25420 @item DECtalk Run-Time Library (DTK interface)
25422 @item Librarian utility routines (LBR interface)
25424 @item General Purpose Run-Time Library (LIB interface)
25426 @item Math Run-Time Library (MTH interface)
25428 @item National Character Set Run-Time Library (NCS interface)
25430 @item Compiled Code Support Run-Time Library (OTS interface)
25432 @item Parallel Processing Run-Time Library (PPL interface)
25434 @item Screen Management Run-Time Library (SMG interface)
25436 @item Sort Run-Time Library (SOR interface)
25438 @item String Run-Time Library (STR interface)
25440 @item STARLET System Library
25443 @item X Window System Version 11R4 and 11R5 (X, XLIB interface)
25445 @item X Windows Toolkit (XT interface)
25447 @item X/Motif Version 1.1.3 and 1.2 (XM interface)
25451 GNAT provides implementations of these HP bindings in the @code{DECLIB}
25452 directory, on both the Alpha and I64 OpenVMS platforms.
25454 The X/Motif bindings used to build @code{DECLIB} are whatever versions are
25456 HP Ada @file{ADA$PREDEFINED} directory with extension @file{.ADC}.
25457 A pragma @code{Linker_Options} has been added to packages @code{Xm},
25458 @code{Xt}, and @code{X_Lib}
25459 causing the default X/Motif sharable image libraries to be linked in. This
25460 is done via options files named @file{xm.opt}, @file{xt.opt}, and
25461 @file{x_lib.opt} (also located in the @file{DECLIB} directory).
25463 It may be necessary to edit these options files to update or correct the
25464 library names if, for example, the newer X/Motif bindings from
25465 @file{ADA$EXAMPLES}
25466 had been (previous to installing GNAT) copied and renamed to supersede the
25467 default @file{ADA$PREDEFINED} versions.
25470 * Shared Libraries and Options Files::
25471 * Interfaces to C::
25474 @node Shared Libraries and Options Files
25475 @subsection Shared Libraries and Options Files
25478 When using the HP Ada
25479 predefined X and Motif bindings, the linking with their sharable images is
25480 done automatically by @command{GNAT LINK}.
25481 When using other X and Motif bindings, you need
25482 to add the corresponding sharable images to the command line for
25483 @code{GNAT LINK}. When linking with shared libraries, or with
25484 @file{.OPT} files, you must
25485 also add them to the command line for @command{GNAT LINK}.
25487 A shared library to be used with GNAT is built in the same way as other
25488 libraries under VMS. The VMS Link command can be used in standard fashion.
25490 @node Interfaces to C
25491 @subsection Interfaces to C
25495 provides the following Ada types and operations:
25498 @item C types package (@code{C_TYPES})
25500 @item C strings (@code{C_TYPES.NULL_TERMINATED})
25502 @item Other_types (@code{SHORT_INT})
25506 Interfacing to C with GNAT, you can use the above approach
25507 described for HP Ada or the facilities of Annex B of
25508 the @cite{Ada Reference Manual} (packages @code{INTERFACES.C},
25509 @code{INTERFACES.C.STRINGS} and @code{INTERFACES.C.POINTERS}). For more
25510 information, see @ref{Interfacing to C,,, gnat_rm, GNAT Reference Manual}.
25512 The @option{-gnatF} qualifier forces default and explicit
25513 @code{External_Name} parameters in pragmas @code{Import} and @code{Export}
25514 to be uppercased for compatibility with the default behavior
25515 of HP C. The qualifier has no effect on @code{Link_Name} parameters.
25517 @node Main Program Definition
25518 @section Main Program Definition
25521 The following section discusses differences in the
25522 definition of main programs on HP Ada and GNAT.
25523 On HP Ada, main programs are defined to meet the
25524 following conditions:
25526 @item Procedure with no formal parameters (returns @code{0} upon
25529 @item Procedure with no formal parameters (returns @code{42} when
25530 an unhandled exception is raised)
25532 @item Function with no formal parameters whose returned value
25533 is of a discrete type
25535 @item Procedure with one @code{out} formal of a discrete type for
25536 which a specification of pragma @code{EXPORT_VALUED_PROCEDURE} is given.
25541 When declared with the pragma @code{EXPORT_VALUED_PROCEDURE},
25542 a main function or main procedure returns a discrete
25543 value whose size is less than 64 bits (32 on VAX systems),
25544 the value is zero- or sign-extended as appropriate.
25545 On GNAT, main programs are defined as follows:
25547 @item Must be a non-generic, parameterless subprogram that
25548 is either a procedure or function returning an Ada
25549 @code{STANDARD.INTEGER} (the predefined type)
25551 @item Cannot be a generic subprogram or an instantiation of a
25555 @node Implementation-Defined Attributes
25556 @section Implementation-Defined Attributes
25559 GNAT provides all HP Ada implementation-defined
25562 @node Compiler and Run-Time Interfacing
25563 @section Compiler and Run-Time Interfacing
25566 HP Ada provides the following qualifiers to pass options to the linker
25569 @item @option{/WAIT} and @option{/SUBMIT}
25571 @item @option{/COMMAND}
25573 @item @option{/@r{[}NO@r{]}MAP}
25575 @item @option{/OUTPUT=@var{file-spec}}
25577 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
25581 To pass options to the linker, GNAT provides the following
25585 @item @option{/EXECUTABLE=@var{exec-name}}
25587 @item @option{/VERBOSE}
25589 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
25593 For more information on these switches, see
25594 @ref{Switches for gnatlink}.
25595 In HP Ada, the command-line switch @option{/OPTIMIZE} is available
25596 to control optimization. HP Ada also supplies the
25599 @item @code{OPTIMIZE}
25601 @item @code{INLINE}
25603 @item @code{INLINE_GENERIC}
25605 @item @code{SUPPRESS_ALL}
25607 @item @code{PASSIVE}
25611 In GNAT, optimization is controlled strictly by command
25612 line parameters, as described in the corresponding section of this guide.
25613 The HP pragmas for control of optimization are
25614 recognized but ignored.
25616 Note that in GNAT, the default is optimization off, whereas in HP Ada
25617 the default is that optimization is turned on.
25619 @node Program Compilation and Library Management
25620 @section Program Compilation and Library Management
25623 HP Ada and GNAT provide a comparable set of commands to
25624 build programs. HP Ada also provides a program library,
25625 which is a concept that does not exist on GNAT. Instead,
25626 GNAT provides directories of sources that are compiled as
25629 The following table summarizes
25630 the HP Ada commands and provides
25631 equivalent GNAT commands. In this table, some GNAT
25632 equivalents reflect the fact that GNAT does not use the
25633 concept of a program library. Instead, it uses a model
25634 in which collections of source and object files are used
25635 in a manner consistent with other languages like C and
25636 Fortran. Therefore, standard system file commands are used
25637 to manipulate these elements. Those GNAT commands are marked with
25639 Note that, unlike HP Ada, none of the GNAT commands accepts wild cards.
25642 @multitable @columnfractions .35 .65
25644 @item @emph{HP Ada Command}
25645 @tab @emph{GNAT Equivalent / Description}
25647 @item @command{ADA}
25648 @tab @command{GNAT COMPILE}@*
25649 Invokes the compiler to compile one or more Ada source files.
25651 @item @command{ACS ATTACH}@*
25652 @tab [No equivalent]@*
25653 Switches control of terminal from current process running the program
25656 @item @command{ACS CHECK}
25657 @tab @command{GNAT MAKE /DEPENDENCY_LIST}@*
25658 Forms the execution closure of one
25659 or more compiled units and checks completeness and currency.
25661 @item @command{ACS COMPILE}
25662 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
25663 Forms the execution closure of one or
25664 more specified units, checks completeness and currency,
25665 identifies units that have revised source files, compiles same,
25666 and recompiles units that are or will become obsolete.
25667 Also completes incomplete generic instantiations.
25669 @item @command{ACS COPY FOREIGN}
25671 Copies a foreign object file into the program library as a
25674 @item @command{ACS COPY UNIT}
25676 Copies a compiled unit from one program library to another.
25678 @item @command{ACS CREATE LIBRARY}
25679 @tab Create /directory (*)@*
25680 Creates a program library.
25682 @item @command{ACS CREATE SUBLIBRARY}
25683 @tab Create /directory (*)@*
25684 Creates a program sublibrary.
25686 @item @command{ACS DELETE LIBRARY}
25688 Deletes a program library and its contents.
25690 @item @command{ACS DELETE SUBLIBRARY}
25692 Deletes a program sublibrary and its contents.
25694 @item @command{ACS DELETE UNIT}
25695 @tab Delete file (*)@*
25696 On OpenVMS systems, deletes one or more compiled units from
25697 the current program library.
25699 @item @command{ACS DIRECTORY}
25700 @tab Directory (*)@*
25701 On OpenVMS systems, lists units contained in the current
25704 @item @command{ACS ENTER FOREIGN}
25706 Allows the import of a foreign body as an Ada library
25707 spec and enters a reference to a pointer.
25709 @item @command{ACS ENTER UNIT}
25711 Enters a reference (pointer) from the current program library to
25712 a unit compiled into another program library.
25714 @item @command{ACS EXIT}
25715 @tab [No equivalent]@*
25716 Exits from the program library manager.
25718 @item @command{ACS EXPORT}
25720 Creates an object file that contains system-specific object code
25721 for one or more units. With GNAT, object files can simply be copied
25722 into the desired directory.
25724 @item @command{ACS EXTRACT SOURCE}
25726 Allows access to the copied source file for each Ada compilation unit
25728 @item @command{ACS HELP}
25729 @tab @command{HELP GNAT}@*
25730 Provides online help.
25732 @item @command{ACS LINK}
25733 @tab @command{GNAT LINK}@*
25734 Links an object file containing Ada units into an executable file.
25736 @item @command{ACS LOAD}
25738 Loads (partially compiles) Ada units into the program library.
25739 Allows loading a program from a collection of files into a library
25740 without knowing the relationship among units.
25742 @item @command{ACS MERGE}
25744 Merges into the current program library, one or more units from
25745 another library where they were modified.
25747 @item @command{ACS RECOMPILE}
25748 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
25749 Recompiles from external or copied source files any obsolete
25750 unit in the closure. Also, completes any incomplete generic
25753 @item @command{ACS REENTER}
25754 @tab @command{GNAT MAKE}@*
25755 Reenters current references to units compiled after last entered
25756 with the @command{ACS ENTER UNIT} command.
25758 @item @command{ACS SET LIBRARY}
25759 @tab Set default (*)@*
25760 Defines a program library to be the compilation context as well
25761 as the target library for compiler output and commands in general.
25763 @item @command{ACS SET PRAGMA}
25764 @tab Edit @file{gnat.adc} (*)@*
25765 Redefines specified values of the library characteristics
25766 @code{LONG_ FLOAT}, @code{MEMORY_SIZE}, @code{SYSTEM_NAME},
25767 and @code{Float_Representation}.
25769 @item @command{ACS SET SOURCE}
25770 @tab Define @code{ADA_INCLUDE_PATH} path (*)@*
25771 Defines the source file search list for the @command{ACS COMPILE} command.
25773 @item @command{ACS SHOW LIBRARY}
25774 @tab Directory (*)@*
25775 Lists information about one or more program libraries.
25777 @item @command{ACS SHOW PROGRAM}
25778 @tab [No equivalent]@*
25779 Lists information about the execution closure of one or
25780 more units in the program library.
25782 @item @command{ACS SHOW SOURCE}
25783 @tab Show logical @code{ADA_INCLUDE_PATH}@*
25784 Shows the source file search used when compiling units.
25786 @item @command{ACS SHOW VERSION}
25787 @tab Compile with @option{VERBOSE} option
25788 Displays the version number of the compiler and program library
25791 @item @command{ACS SPAWN}
25792 @tab [No equivalent]@*
25793 Creates a subprocess of the current process (same as @command{DCL SPAWN}
25796 @item @command{ACS VERIFY}
25797 @tab [No equivalent]@*
25798 Performs a series of consistency checks on a program library to
25799 determine whether the library structure and library files are in
25806 @section Input-Output
25809 On OpenVMS Alpha systems, HP Ada uses OpenVMS Record
25810 Management Services (RMS) to perform operations on
25814 HP Ada and GNAT predefine an identical set of input-
25815 output packages. To make the use of the
25816 generic @code{TEXT_IO} operations more convenient, HP Ada
25817 provides predefined library packages that instantiate the
25818 integer and floating-point operations for the predefined
25819 integer and floating-point types as shown in the following table.
25821 @multitable @columnfractions .45 .55
25822 @item @emph{Package Name} @tab Instantiation
25824 @item @code{INTEGER_TEXT_IO}
25825 @tab @code{INTEGER_IO(INTEGER)}
25827 @item @code{SHORT_INTEGER_TEXT_IO}
25828 @tab @code{INTEGER_IO(SHORT_INTEGER)}
25830 @item @code{SHORT_SHORT_INTEGER_TEXT_IO}
25831 @tab @code{INTEGER_IO(SHORT_SHORT_INTEGER)}
25833 @item @code{FLOAT_TEXT_IO}
25834 @tab @code{FLOAT_IO(FLOAT)}
25836 @item @code{LONG_FLOAT_TEXT_IO}
25837 @tab @code{FLOAT_IO(LONG_FLOAT)}
25841 The HP Ada predefined packages and their operations
25842 are implemented using OpenVMS Alpha files and input-output
25843 facilities. HP Ada supports asynchronous input-output on OpenVMS Alpha.
25844 Familiarity with the following is recommended:
25846 @item RMS file organizations and access methods
25848 @item OpenVMS file specifications and directories
25850 @item OpenVMS File Definition Language (FDL)
25854 GNAT provides I/O facilities that are completely
25855 compatible with HP Ada. The distribution includes the
25856 standard HP Ada versions of all I/O packages, operating
25857 in a manner compatible with HP Ada. In particular, the
25858 following packages are by default the HP Ada (Ada 83)
25859 versions of these packages rather than the renamings
25860 suggested in Annex J of the Ada Reference Manual:
25862 @item @code{TEXT_IO}
25864 @item @code{SEQUENTIAL_IO}
25866 @item @code{DIRECT_IO}
25870 The use of the standard child package syntax (for
25871 example, @code{ADA.TEXT_IO}) retrieves the post-Ada 83 versions of these
25873 GNAT provides HP-compatible predefined instantiations
25874 of the @code{TEXT_IO} packages, and also
25875 provides the standard predefined instantiations required
25876 by the @cite{Ada Reference Manual}.
25878 For further information on how GNAT interfaces to the file
25879 system or how I/O is implemented in programs written in
25880 mixed languages, see @ref{Implementation of the Standard I/O,,,
25881 gnat_rm, GNAT Reference Manual}.
25882 This chapter covers the following:
25884 @item Standard I/O packages
25886 @item @code{FORM} strings
25888 @item @code{ADA.DIRECT_IO}
25890 @item @code{ADA.SEQUENTIAL_IO}
25892 @item @code{ADA.TEXT_IO}
25894 @item Stream pointer positioning
25896 @item Reading and writing non-regular files
25898 @item @code{GET_IMMEDIATE}
25900 @item Treating @code{TEXT_IO} files as streams
25907 @node Implementation Limits
25908 @section Implementation Limits
25911 The following table lists implementation limits for HP Ada
25913 @multitable @columnfractions .60 .20 .20
25915 @item @emph{Compilation Parameter}
25920 @item In a subprogram or entry declaration, maximum number of
25921 formal parameters that are of an unconstrained record type
25926 @item Maximum identifier length (number of characters)
25931 @item Maximum number of characters in a source line
25936 @item Maximum collection size (number of bytes)
25941 @item Maximum number of discriminants for a record type
25946 @item Maximum number of formal parameters in an entry or
25947 subprogram declaration
25952 @item Maximum number of dimensions in an array type
25957 @item Maximum number of library units and subunits in a compilation.
25962 @item Maximum number of library units and subunits in an execution.
25967 @item Maximum number of objects declared with the pragma @code{COMMON_OBJECT}
25968 or @code{PSECT_OBJECT}
25973 @item Maximum number of enumeration literals in an enumeration type
25979 @item Maximum number of lines in a source file
25984 @item Maximum number of bits in any object
25989 @item Maximum size of the static portion of a stack frame (approximate)
25994 @node Tools and Utilities
25995 @section Tools and Utilities
25998 The following table lists some of the OpenVMS development tools
25999 available for HP Ada, and the corresponding tools for
26000 use with @value{EDITION} on Alpha and I64 platforms.
26001 Aside from the debugger, all the OpenVMS tools identified are part
26002 of the DECset package.
26005 @c Specify table in TeX since Texinfo does a poor job
26009 \settabs\+Language-Sensitive Editor\quad
26010 &Product with HP Ada\quad
26013 &\it Product with HP Ada
26014 & \it Product with GNAT Pro\cr
26016 \+Code Management System
26020 \+Language-Sensitive Editor
26022 & emacs or HP LSE (Alpha)\cr
26032 & OpenVMS Debug (I64)\cr
26034 \+Source Code Analyzer /
26051 \+Coverage Analyzer
26055 \+Module Management
26057 & Not applicable\cr
26067 @c This is the Texinfo version of the table. It renders poorly in pdf, hence
26068 @c the TeX version above for the printed version
26070 @c @multitable @columnfractions .3 .4 .4
26071 @multitable {Source Code Analyzer /}{Tool with HP Ada}{Tool with GNAT Pro}
26073 @tab @i{Tool with HP Ada}
26074 @tab @i{Tool with @value{EDITION}}
26075 @item Code Management@*System
26078 @item Language-Sensitive@*Editor
26080 @tab emacs or HP LSE (Alpha)
26089 @tab OpenVMS Debug (I64)
26090 @item Source Code Analyzer /@*Cross Referencer
26094 @tab HP Digital Test@*Manager (DTM)
26096 @item Performance and@*Coverage Analyzer
26099 @item Module Management@*System
26101 @tab Not applicable
26108 @c **************************************
26109 @node Platform-Specific Information for the Run-Time Libraries
26110 @appendix Platform-Specific Information for the Run-Time Libraries
26111 @cindex Tasking and threads libraries
26112 @cindex Threads libraries and tasking
26113 @cindex Run-time libraries (platform-specific information)
26116 The GNAT run-time implementation may vary with respect to both the
26117 underlying threads library and the exception handling scheme.
26118 For threads support, one or more of the following are supplied:
26120 @item @b{native threads library}, a binding to the thread package from
26121 the underlying operating system
26123 @item @b{pthreads library} (Sparc Solaris only), a binding to the Solaris
26124 POSIX thread package
26128 For exception handling, either or both of two models are supplied:
26130 @item @b{Zero-Cost Exceptions} (``ZCX''),@footnote{
26131 Most programs should experience a substantial speed improvement by
26132 being compiled with a ZCX run-time.
26133 This is especially true for
26134 tasking applications or applications with many exception handlers.}
26135 @cindex Zero-Cost Exceptions
26136 @cindex ZCX (Zero-Cost Exceptions)
26137 which uses binder-generated tables that
26138 are interrogated at run time to locate a handler
26140 @item @b{setjmp / longjmp} (``SJLJ''),
26141 @cindex setjmp/longjmp Exception Model
26142 @cindex SJLJ (setjmp/longjmp Exception Model)
26143 which uses dynamically-set data to establish
26144 the set of handlers
26148 This appendix summarizes which combinations of threads and exception support
26149 are supplied on various GNAT platforms.
26150 It then shows how to select a particular library either
26151 permanently or temporarily,
26152 explains the properties of (and tradeoffs among) the various threads
26153 libraries, and provides some additional
26154 information about several specific platforms.
26157 * Summary of Run-Time Configurations::
26158 * Specifying a Run-Time Library::
26159 * Choosing the Scheduling Policy::
26160 * Solaris-Specific Considerations::
26161 * Linux-Specific Considerations::
26162 * AIX-Specific Considerations::
26163 * Irix-Specific Considerations::
26164 * RTX-Specific Considerations::
26167 @node Summary of Run-Time Configurations
26168 @section Summary of Run-Time Configurations
26170 @multitable @columnfractions .30 .70
26171 @item @b{alpha-openvms}
26172 @item @code{@ @ }@i{rts-native (default)}
26173 @item @code{@ @ @ @ }Tasking @tab native VMS threads
26174 @item @code{@ @ @ @ }Exceptions @tab ZCX
26176 @item @b{alpha-tru64}
26177 @item @code{@ @ }@i{rts-native (default)}
26178 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
26179 @item @code{@ @ @ @ }Exceptions @tab ZCX
26181 @item @code{@ @ }@i{rts-sjlj}
26182 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
26183 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26185 @item @b{ia64-hp_linux}
26186 @item @code{@ @ }@i{rts-native (default)}
26187 @item @code{@ @ @ @ }Tasking @tab pthread library
26188 @item @code{@ @ @ @ }Exceptions @tab ZCX
26190 @item @b{ia64-hpux}
26191 @item @code{@ @ }@i{rts-native (default)}
26192 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26193 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26195 @item @b{ia64-openvms}
26196 @item @code{@ @ }@i{rts-native (default)}
26197 @item @code{@ @ @ @ }Tasking @tab native VMS threads
26198 @item @code{@ @ @ @ }Exceptions @tab ZCX
26200 @item @b{ia64-sgi_linux}
26201 @item @code{@ @ }@i{rts-native (default)}
26202 @item @code{@ @ @ @ }Tasking @tab pthread library
26203 @item @code{@ @ @ @ }Exceptions @tab ZCX
26205 @item @b{mips-irix}
26206 @item @code{@ @ }@i{rts-native (default)}
26207 @item @code{@ @ @ @ }Tasking @tab native IRIX threads
26208 @item @code{@ @ @ @ }Exceptions @tab ZCX
26211 @item @code{@ @ }@i{rts-native (default)}
26212 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26213 @item @code{@ @ @ @ }Exceptions @tab ZCX
26215 @item @code{@ @ }@i{rts-sjlj}
26216 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26217 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26220 @item @code{@ @ }@i{rts-native (default)}
26221 @item @code{@ @ @ @ }Tasking @tab native AIX threads
26222 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26224 @item @b{ppc-darwin}
26225 @item @code{@ @ }@i{rts-native (default)}
26226 @item @code{@ @ @ @ }Tasking @tab native MacOS threads
26227 @item @code{@ @ @ @ }Exceptions @tab ZCX
26229 @item @b{sparc-solaris} @tab
26230 @item @code{@ @ }@i{rts-native (default)}
26231 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
26232 @item @code{@ @ @ @ }Exceptions @tab ZCX
26234 @item @code{@ @ }@i{rts-pthread}
26235 @item @code{@ @ @ @ }Tasking @tab pthread library
26236 @item @code{@ @ @ @ }Exceptions @tab ZCX
26238 @item @code{@ @ }@i{rts-sjlj}
26239 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
26240 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26242 @item @b{sparc64-solaris} @tab
26243 @item @code{@ @ }@i{rts-native (default)}
26244 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
26245 @item @code{@ @ @ @ }Exceptions @tab ZCX
26247 @item @b{x86-linux}
26248 @item @code{@ @ }@i{rts-native (default)}
26249 @item @code{@ @ @ @ }Tasking @tab pthread library
26250 @item @code{@ @ @ @ }Exceptions @tab ZCX
26252 @item @code{@ @ }@i{rts-sjlj}
26253 @item @code{@ @ @ @ }Tasking @tab pthread library
26254 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26257 @item @code{@ @ }@i{rts-native (default)}
26258 @item @code{@ @ @ @ }Tasking @tab native LynxOS threads
26259 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26261 @item @b{x86-solaris}
26262 @item @code{@ @ }@i{rts-native (default)}
26263 @item @code{@ @ @ @ }Tasking @tab native Solaris threads
26264 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26266 @item @b{x86-windows}
26267 @item @code{@ @ }@i{rts-native (default)}
26268 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
26269 @item @code{@ @ @ @ }Exceptions @tab ZCX
26271 @item @code{@ @ }@i{rts-sjlj (default)}
26272 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
26273 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26275 @item @b{x86-windows-rtx}
26276 @item @code{@ @ }@i{rts-rtx-rtss (default)}
26277 @item @code{@ @ @ @ }Tasking @tab RTX real-time subsystem RTSS threads (kernel mode)
26278 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26280 @item @code{@ @ }@i{rts-rtx-w32}
26281 @item @code{@ @ @ @ }Tasking @tab RTX Win32 threads (user mode)
26282 @item @code{@ @ @ @ }Exceptions @tab ZCX
26284 @item @b{x86_64-linux}
26285 @item @code{@ @ }@i{rts-native (default)}
26286 @item @code{@ @ @ @ }Tasking @tab pthread library
26287 @item @code{@ @ @ @ }Exceptions @tab ZCX
26289 @item @code{@ @ }@i{rts-sjlj}
26290 @item @code{@ @ @ @ }Tasking @tab pthread library
26291 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26295 @node Specifying a Run-Time Library
26296 @section Specifying a Run-Time Library
26299 The @file{adainclude} subdirectory containing the sources of the GNAT
26300 run-time library, and the @file{adalib} subdirectory containing the
26301 @file{ALI} files and the static and/or shared GNAT library, are located
26302 in the gcc target-dependent area:
26305 target=$prefix/lib/gcc/gcc-@i{dumpmachine}/gcc-@i{dumpversion}/
26309 As indicated above, on some platforms several run-time libraries are supplied.
26310 These libraries are installed in the target dependent area and
26311 contain a complete source and binary subdirectory. The detailed description
26312 below explains the differences between the different libraries in terms of
26313 their thread support.
26315 The default run-time library (when GNAT is installed) is @emph{rts-native}.
26316 This default run time is selected by the means of soft links.
26317 For example on x86-linux:
26323 +--- adainclude----------+
26325 +--- adalib-----------+ |
26327 +--- rts-native | |
26329 | +--- adainclude <---+
26331 | +--- adalib <----+
26342 If the @i{rts-sjlj} library is to be selected on a permanent basis,
26343 these soft links can be modified with the following commands:
26347 $ rm -f adainclude adalib
26348 $ ln -s rts-sjlj/adainclude adainclude
26349 $ ln -s rts-sjlj/adalib adalib
26353 Alternatively, you can specify @file{rts-sjlj/adainclude} in the file
26354 @file{$target/ada_source_path} and @file{rts-sjlj/adalib} in
26355 @file{$target/ada_object_path}.
26357 Selecting another run-time library temporarily can be
26358 achieved by using the @option{--RTS} switch, e.g., @option{--RTS=sjlj}
26359 @cindex @option{--RTS} option
26361 @node Choosing the Scheduling Policy
26362 @section Choosing the Scheduling Policy
26365 When using a POSIX threads implementation, you have a choice of several
26366 scheduling policies: @code{SCHED_FIFO},
26367 @cindex @code{SCHED_FIFO} scheduling policy
26369 @cindex @code{SCHED_RR} scheduling policy
26370 and @code{SCHED_OTHER}.
26371 @cindex @code{SCHED_OTHER} scheduling policy
26372 Typically, the default is @code{SCHED_OTHER}, while using @code{SCHED_FIFO}
26373 or @code{SCHED_RR} requires special (e.g., root) privileges.
26375 By default, GNAT uses the @code{SCHED_OTHER} policy. To specify
26377 @cindex @code{SCHED_FIFO} scheduling policy
26378 you can use one of the following:
26382 @code{pragma Time_Slice (0.0)}
26383 @cindex pragma Time_Slice
26385 the corresponding binder option @option{-T0}
26386 @cindex @option{-T0} option
26388 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
26389 @cindex pragma Task_Dispatching_Policy
26393 To specify @code{SCHED_RR},
26394 @cindex @code{SCHED_RR} scheduling policy
26395 you should use @code{pragma Time_Slice} with a
26396 value greater than @code{0.0}, or else use the corresponding @option{-T}
26399 @node Solaris-Specific Considerations
26400 @section Solaris-Specific Considerations
26401 @cindex Solaris Sparc threads libraries
26404 This section addresses some topics related to the various threads libraries
26408 * Solaris Threads Issues::
26411 @node Solaris Threads Issues
26412 @subsection Solaris Threads Issues
26415 GNAT under Solaris/Sparc 32 bits comes with an alternate tasking run-time
26416 library based on POSIX threads --- @emph{rts-pthread}.
26417 @cindex rts-pthread threads library
26418 This run-time library has the advantage of being mostly shared across all
26419 POSIX-compliant thread implementations, and it also provides under
26420 @w{Solaris 8} the @code{PTHREAD_PRIO_INHERIT}
26421 @cindex @code{PTHREAD_PRIO_INHERIT} policy (under rts-pthread)
26422 and @code{PTHREAD_PRIO_PROTECT}
26423 @cindex @code{PTHREAD_PRIO_PROTECT} policy (under rts-pthread)
26424 semantics that can be selected using the predefined pragma
26425 @code{Locking_Policy}
26426 @cindex pragma Locking_Policy (under rts-pthread)
26428 @code{Inheritance_Locking} and @code{Ceiling_Locking} as the policy.
26429 @cindex @code{Inheritance_Locking} (under rts-pthread)
26430 @cindex @code{Ceiling_Locking} (under rts-pthread)
26432 As explained above, the native run-time library is based on the Solaris thread
26433 library (@code{libthread}) and is the default library.
26435 When the Solaris threads library is used (this is the default), programs
26436 compiled with GNAT can automatically take advantage of
26437 and can thus execute on multiple processors.
26438 The user can alternatively specify a processor on which the program should run
26439 to emulate a single-processor system. The multiprocessor / uniprocessor choice
26441 setting the environment variable @env{GNAT_PROCESSOR}
26442 @cindex @env{GNAT_PROCESSOR} environment variable (on Sparc Solaris)
26443 to one of the following:
26447 Use the default configuration (run the program on all
26448 available processors) - this is the same as having @code{GNAT_PROCESSOR}
26452 Let the run-time implementation choose one processor and run the program on
26455 @item 0 .. Last_Proc
26456 Run the program on the specified processor.
26457 @code{Last_Proc} is equal to @code{_SC_NPROCESSORS_CONF - 1}
26458 (where @code{_SC_NPROCESSORS_CONF} is a system variable).
26461 @node Linux-Specific Considerations
26462 @section Linux-Specific Considerations
26463 @cindex Linux threads libraries
26466 On GNU/Linux without NPTL support (usually system with GNU C Library
26467 older than 2.3), the signal model is not POSIX compliant, which means
26468 that to send a signal to the process, you need to send the signal to all
26469 threads, e.g.@: by using @code{killpg()}.
26471 @node AIX-Specific Considerations
26472 @section AIX-Specific Considerations
26473 @cindex AIX resolver library
26476 On AIX, the resolver library initializes some internal structure on
26477 the first call to @code{get*by*} functions, which are used to implement
26478 @code{GNAT.Sockets.Get_Host_By_Name} and
26479 @code{GNAT.Sockets.Get_Host_By_Address}.
26480 If such initialization occurs within an Ada task, and the stack size for
26481 the task is the default size, a stack overflow may occur.
26483 To avoid this overflow, the user should either ensure that the first call
26484 to @code{GNAT.Sockets.Get_Host_By_Name} or
26485 @code{GNAT.Sockets.Get_Host_By_Addrss}
26486 occurs in the environment task, or use @code{pragma Storage_Size} to
26487 specify a sufficiently large size for the stack of the task that contains
26490 @node Irix-Specific Considerations
26491 @section Irix-Specific Considerations
26492 @cindex Irix libraries
26495 The GCC support libraries coming with the Irix compiler have moved to
26496 their canonical place with respect to the general Irix ABI related
26497 conventions. Running applications built with the default shared GNAT
26498 run-time now requires the LD_LIBRARY_PATH environment variable to
26499 include this location. A possible way to achieve this is to issue the
26500 following command line on a bash prompt:
26504 $ LD_LIBRARY_PATH=$LD_LIBRARY_PATH:`dirname \`gcc --print-file-name=libgcc_s.so\``
26508 @node RTX-Specific Considerations
26509 @section RTX-Specific Considerations
26510 @cindex RTX libraries
26513 The Real-time Extension (RTX) to Windows is based on the Windows Win32
26514 API. Applications can be built to work in two different modes:
26518 Windows executables that run in Ring 3 to utilize memory protection
26519 (@emph{rts-rtx-w32}).
26522 Real-time subsystem (RTSS) executables that run in Ring 0, where
26523 performance can be optimized with RTSS applications taking precedent
26524 over all Windows applications (@emph{rts-rtx-rtss}).
26528 @c *******************************
26529 @node Example of Binder Output File
26530 @appendix Example of Binder Output File
26533 This Appendix displays the source code for @command{gnatbind}'s output
26534 file generated for a simple ``Hello World'' program.
26535 Comments have been added for clarification purposes.
26537 @smallexample @c adanocomment
26541 -- The package is called Ada_Main unless this name is actually used
26542 -- as a unit name in the partition, in which case some other unique
26546 package ada_main is
26548 Elab_Final_Code : Integer;
26549 pragma Import (C, Elab_Final_Code, "__gnat_inside_elab_final_code");
26551 -- The main program saves the parameters (argument count,
26552 -- argument values, environment pointer) in global variables
26553 -- for later access by other units including
26554 -- Ada.Command_Line.
26556 gnat_argc : Integer;
26557 gnat_argv : System.Address;
26558 gnat_envp : System.Address;
26560 -- The actual variables are stored in a library routine. This
26561 -- is useful for some shared library situations, where there
26562 -- are problems if variables are not in the library.
26564 pragma Import (C, gnat_argc);
26565 pragma Import (C, gnat_argv);
26566 pragma Import (C, gnat_envp);
26568 -- The exit status is similarly an external location
26570 gnat_exit_status : Integer;
26571 pragma Import (C, gnat_exit_status);
26573 GNAT_Version : constant String :=
26574 "GNAT Version: 6.0.0w (20061115)";
26575 pragma Export (C, GNAT_Version, "__gnat_version");
26577 -- This is the generated adafinal routine that performs
26578 -- finalization at the end of execution. In the case where
26579 -- Ada is the main program, this main program makes a call
26580 -- to adafinal at program termination.
26582 procedure adafinal;
26583 pragma Export (C, adafinal, "adafinal");
26585 -- This is the generated adainit routine that performs
26586 -- initialization at the start of execution. In the case
26587 -- where Ada is the main program, this main program makes
26588 -- a call to adainit at program startup.
26591 pragma Export (C, adainit, "adainit");
26593 -- This routine is called at the start of execution. It is
26594 -- a dummy routine that is used by the debugger to breakpoint
26595 -- at the start of execution.
26597 procedure Break_Start;
26598 pragma Import (C, Break_Start, "__gnat_break_start");
26600 -- This is the actual generated main program (it would be
26601 -- suppressed if the no main program switch were used). As
26602 -- required by standard system conventions, this program has
26603 -- the external name main.
26607 argv : System.Address;
26608 envp : System.Address)
26610 pragma Export (C, main, "main");
26612 -- The following set of constants give the version
26613 -- identification values for every unit in the bound
26614 -- partition. This identification is computed from all
26615 -- dependent semantic units, and corresponds to the
26616 -- string that would be returned by use of the
26617 -- Body_Version or Version attributes.
26619 type Version_32 is mod 2 ** 32;
26620 u00001 : constant Version_32 := 16#7880BEB3#;
26621 u00002 : constant Version_32 := 16#0D24CBD0#;
26622 u00003 : constant Version_32 := 16#3283DBEB#;
26623 u00004 : constant Version_32 := 16#2359F9ED#;
26624 u00005 : constant Version_32 := 16#664FB847#;
26625 u00006 : constant Version_32 := 16#68E803DF#;
26626 u00007 : constant Version_32 := 16#5572E604#;
26627 u00008 : constant Version_32 := 16#46B173D8#;
26628 u00009 : constant Version_32 := 16#156A40CF#;
26629 u00010 : constant Version_32 := 16#033DABE0#;
26630 u00011 : constant Version_32 := 16#6AB38FEA#;
26631 u00012 : constant Version_32 := 16#22B6217D#;
26632 u00013 : constant Version_32 := 16#68A22947#;
26633 u00014 : constant Version_32 := 16#18CC4A56#;
26634 u00015 : constant Version_32 := 16#08258E1B#;
26635 u00016 : constant Version_32 := 16#367D5222#;
26636 u00017 : constant Version_32 := 16#20C9ECA4#;
26637 u00018 : constant Version_32 := 16#50D32CB6#;
26638 u00019 : constant Version_32 := 16#39A8BB77#;
26639 u00020 : constant Version_32 := 16#5CF8FA2B#;
26640 u00021 : constant Version_32 := 16#2F1EB794#;
26641 u00022 : constant Version_32 := 16#31AB6444#;
26642 u00023 : constant Version_32 := 16#1574B6E9#;
26643 u00024 : constant Version_32 := 16#5109C189#;
26644 u00025 : constant Version_32 := 16#56D770CD#;
26645 u00026 : constant Version_32 := 16#02F9DE3D#;
26646 u00027 : constant Version_32 := 16#08AB6B2C#;
26647 u00028 : constant Version_32 := 16#3FA37670#;
26648 u00029 : constant Version_32 := 16#476457A0#;
26649 u00030 : constant Version_32 := 16#731E1B6E#;
26650 u00031 : constant Version_32 := 16#23C2E789#;
26651 u00032 : constant Version_32 := 16#0F1BD6A1#;
26652 u00033 : constant Version_32 := 16#7C25DE96#;
26653 u00034 : constant Version_32 := 16#39ADFFA2#;
26654 u00035 : constant Version_32 := 16#571DE3E7#;
26655 u00036 : constant Version_32 := 16#5EB646AB#;
26656 u00037 : constant Version_32 := 16#4249379B#;
26657 u00038 : constant Version_32 := 16#0357E00A#;
26658 u00039 : constant Version_32 := 16#3784FB72#;
26659 u00040 : constant Version_32 := 16#2E723019#;
26660 u00041 : constant Version_32 := 16#623358EA#;
26661 u00042 : constant Version_32 := 16#107F9465#;
26662 u00043 : constant Version_32 := 16#6843F68A#;
26663 u00044 : constant Version_32 := 16#63305874#;
26664 u00045 : constant Version_32 := 16#31E56CE1#;
26665 u00046 : constant Version_32 := 16#02917970#;
26666 u00047 : constant Version_32 := 16#6CCBA70E#;
26667 u00048 : constant Version_32 := 16#41CD4204#;
26668 u00049 : constant Version_32 := 16#572E3F58#;
26669 u00050 : constant Version_32 := 16#20729FF5#;
26670 u00051 : constant Version_32 := 16#1D4F93E8#;
26671 u00052 : constant Version_32 := 16#30B2EC3D#;
26672 u00053 : constant Version_32 := 16#34054F96#;
26673 u00054 : constant Version_32 := 16#5A199860#;
26674 u00055 : constant Version_32 := 16#0E7F912B#;
26675 u00056 : constant Version_32 := 16#5760634A#;
26676 u00057 : constant Version_32 := 16#5D851835#;
26678 -- The following Export pragmas export the version numbers
26679 -- with symbolic names ending in B (for body) or S
26680 -- (for spec) so that they can be located in a link. The
26681 -- information provided here is sufficient to track down
26682 -- the exact versions of units used in a given build.
26684 pragma Export (C, u00001, "helloB");
26685 pragma Export (C, u00002, "system__standard_libraryB");
26686 pragma Export (C, u00003, "system__standard_libraryS");
26687 pragma Export (C, u00004, "adaS");
26688 pragma Export (C, u00005, "ada__text_ioB");
26689 pragma Export (C, u00006, "ada__text_ioS");
26690 pragma Export (C, u00007, "ada__exceptionsB");
26691 pragma Export (C, u00008, "ada__exceptionsS");
26692 pragma Export (C, u00009, "gnatS");
26693 pragma Export (C, u00010, "gnat__heap_sort_aB");
26694 pragma Export (C, u00011, "gnat__heap_sort_aS");
26695 pragma Export (C, u00012, "systemS");
26696 pragma Export (C, u00013, "system__exception_tableB");
26697 pragma Export (C, u00014, "system__exception_tableS");
26698 pragma Export (C, u00015, "gnat__htableB");
26699 pragma Export (C, u00016, "gnat__htableS");
26700 pragma Export (C, u00017, "system__exceptionsS");
26701 pragma Export (C, u00018, "system__machine_state_operationsB");
26702 pragma Export (C, u00019, "system__machine_state_operationsS");
26703 pragma Export (C, u00020, "system__machine_codeS");
26704 pragma Export (C, u00021, "system__storage_elementsB");
26705 pragma Export (C, u00022, "system__storage_elementsS");
26706 pragma Export (C, u00023, "system__secondary_stackB");
26707 pragma Export (C, u00024, "system__secondary_stackS");
26708 pragma Export (C, u00025, "system__parametersB");
26709 pragma Export (C, u00026, "system__parametersS");
26710 pragma Export (C, u00027, "system__soft_linksB");
26711 pragma Export (C, u00028, "system__soft_linksS");
26712 pragma Export (C, u00029, "system__stack_checkingB");
26713 pragma Export (C, u00030, "system__stack_checkingS");
26714 pragma Export (C, u00031, "system__tracebackB");
26715 pragma Export (C, u00032, "system__tracebackS");
26716 pragma Export (C, u00033, "ada__streamsS");
26717 pragma Export (C, u00034, "ada__tagsB");
26718 pragma Export (C, u00035, "ada__tagsS");
26719 pragma Export (C, u00036, "system__string_opsB");
26720 pragma Export (C, u00037, "system__string_opsS");
26721 pragma Export (C, u00038, "interfacesS");
26722 pragma Export (C, u00039, "interfaces__c_streamsB");
26723 pragma Export (C, u00040, "interfaces__c_streamsS");
26724 pragma Export (C, u00041, "system__file_ioB");
26725 pragma Export (C, u00042, "system__file_ioS");
26726 pragma Export (C, u00043, "ada__finalizationB");
26727 pragma Export (C, u00044, "ada__finalizationS");
26728 pragma Export (C, u00045, "system__finalization_rootB");
26729 pragma Export (C, u00046, "system__finalization_rootS");
26730 pragma Export (C, u00047, "system__finalization_implementationB");
26731 pragma Export (C, u00048, "system__finalization_implementationS");
26732 pragma Export (C, u00049, "system__string_ops_concat_3B");
26733 pragma Export (C, u00050, "system__string_ops_concat_3S");
26734 pragma Export (C, u00051, "system__stream_attributesB");
26735 pragma Export (C, u00052, "system__stream_attributesS");
26736 pragma Export (C, u00053, "ada__io_exceptionsS");
26737 pragma Export (C, u00054, "system__unsigned_typesS");
26738 pragma Export (C, u00055, "system__file_control_blockS");
26739 pragma Export (C, u00056, "ada__finalization__list_controllerB");
26740 pragma Export (C, u00057, "ada__finalization__list_controllerS");
26742 -- BEGIN ELABORATION ORDER
26745 -- gnat.heap_sort_a (spec)
26746 -- gnat.heap_sort_a (body)
26747 -- gnat.htable (spec)
26748 -- gnat.htable (body)
26749 -- interfaces (spec)
26751 -- system.machine_code (spec)
26752 -- system.parameters (spec)
26753 -- system.parameters (body)
26754 -- interfaces.c_streams (spec)
26755 -- interfaces.c_streams (body)
26756 -- system.standard_library (spec)
26757 -- ada.exceptions (spec)
26758 -- system.exception_table (spec)
26759 -- system.exception_table (body)
26760 -- ada.io_exceptions (spec)
26761 -- system.exceptions (spec)
26762 -- system.storage_elements (spec)
26763 -- system.storage_elements (body)
26764 -- system.machine_state_operations (spec)
26765 -- system.machine_state_operations (body)
26766 -- system.secondary_stack (spec)
26767 -- system.stack_checking (spec)
26768 -- system.soft_links (spec)
26769 -- system.soft_links (body)
26770 -- system.stack_checking (body)
26771 -- system.secondary_stack (body)
26772 -- system.standard_library (body)
26773 -- system.string_ops (spec)
26774 -- system.string_ops (body)
26777 -- ada.streams (spec)
26778 -- system.finalization_root (spec)
26779 -- system.finalization_root (body)
26780 -- system.string_ops_concat_3 (spec)
26781 -- system.string_ops_concat_3 (body)
26782 -- system.traceback (spec)
26783 -- system.traceback (body)
26784 -- ada.exceptions (body)
26785 -- system.unsigned_types (spec)
26786 -- system.stream_attributes (spec)
26787 -- system.stream_attributes (body)
26788 -- system.finalization_implementation (spec)
26789 -- system.finalization_implementation (body)
26790 -- ada.finalization (spec)
26791 -- ada.finalization (body)
26792 -- ada.finalization.list_controller (spec)
26793 -- ada.finalization.list_controller (body)
26794 -- system.file_control_block (spec)
26795 -- system.file_io (spec)
26796 -- system.file_io (body)
26797 -- ada.text_io (spec)
26798 -- ada.text_io (body)
26800 -- END ELABORATION ORDER
26804 -- The following source file name pragmas allow the generated file
26805 -- names to be unique for different main programs. They are needed
26806 -- since the package name will always be Ada_Main.
26808 pragma Source_File_Name (ada_main, Spec_File_Name => "b~hello.ads");
26809 pragma Source_File_Name (ada_main, Body_File_Name => "b~hello.adb");
26811 -- Generated package body for Ada_Main starts here
26813 package body ada_main is
26815 -- The actual finalization is performed by calling the
26816 -- library routine in System.Standard_Library.Adafinal
26818 procedure Do_Finalize;
26819 pragma Import (C, Do_Finalize, "system__standard_library__adafinal");
26826 procedure adainit is
26828 -- These booleans are set to True once the associated unit has
26829 -- been elaborated. It is also used to avoid elaborating the
26830 -- same unit twice.
26833 pragma Import (Ada, E040, "interfaces__c_streams_E");
26836 pragma Import (Ada, E008, "ada__exceptions_E");
26839 pragma Import (Ada, E014, "system__exception_table_E");
26842 pragma Import (Ada, E053, "ada__io_exceptions_E");
26845 pragma Import (Ada, E017, "system__exceptions_E");
26848 pragma Import (Ada, E024, "system__secondary_stack_E");
26851 pragma Import (Ada, E030, "system__stack_checking_E");
26854 pragma Import (Ada, E028, "system__soft_links_E");
26857 pragma Import (Ada, E035, "ada__tags_E");
26860 pragma Import (Ada, E033, "ada__streams_E");
26863 pragma Import (Ada, E046, "system__finalization_root_E");
26866 pragma Import (Ada, E048, "system__finalization_implementation_E");
26869 pragma Import (Ada, E044, "ada__finalization_E");
26872 pragma Import (Ada, E057, "ada__finalization__list_controller_E");
26875 pragma Import (Ada, E055, "system__file_control_block_E");
26878 pragma Import (Ada, E042, "system__file_io_E");
26881 pragma Import (Ada, E006, "ada__text_io_E");
26883 -- Set_Globals is a library routine that stores away the
26884 -- value of the indicated set of global values in global
26885 -- variables within the library.
26887 procedure Set_Globals
26888 (Main_Priority : Integer;
26889 Time_Slice_Value : Integer;
26890 WC_Encoding : Character;
26891 Locking_Policy : Character;
26892 Queuing_Policy : Character;
26893 Task_Dispatching_Policy : Character;
26894 Adafinal : System.Address;
26895 Unreserve_All_Interrupts : Integer;
26896 Exception_Tracebacks : Integer);
26897 @findex __gnat_set_globals
26898 pragma Import (C, Set_Globals, "__gnat_set_globals");
26900 -- SDP_Table_Build is a library routine used to build the
26901 -- exception tables. See unit Ada.Exceptions in files
26902 -- a-except.ads/adb for full details of how zero cost
26903 -- exception handling works. This procedure, the call to
26904 -- it, and the two following tables are all omitted if the
26905 -- build is in longjmp/setjmp exception mode.
26907 @findex SDP_Table_Build
26908 @findex Zero Cost Exceptions
26909 procedure SDP_Table_Build
26910 (SDP_Addresses : System.Address;
26911 SDP_Count : Natural;
26912 Elab_Addresses : System.Address;
26913 Elab_Addr_Count : Natural);
26914 pragma Import (C, SDP_Table_Build, "__gnat_SDP_Table_Build");
26916 -- Table of Unit_Exception_Table addresses. Used for zero
26917 -- cost exception handling to build the top level table.
26919 ST : aliased constant array (1 .. 23) of System.Address := (
26921 Ada.Text_Io'UET_Address,
26922 Ada.Exceptions'UET_Address,
26923 Gnat.Heap_Sort_A'UET_Address,
26924 System.Exception_Table'UET_Address,
26925 System.Machine_State_Operations'UET_Address,
26926 System.Secondary_Stack'UET_Address,
26927 System.Parameters'UET_Address,
26928 System.Soft_Links'UET_Address,
26929 System.Stack_Checking'UET_Address,
26930 System.Traceback'UET_Address,
26931 Ada.Streams'UET_Address,
26932 Ada.Tags'UET_Address,
26933 System.String_Ops'UET_Address,
26934 Interfaces.C_Streams'UET_Address,
26935 System.File_Io'UET_Address,
26936 Ada.Finalization'UET_Address,
26937 System.Finalization_Root'UET_Address,
26938 System.Finalization_Implementation'UET_Address,
26939 System.String_Ops_Concat_3'UET_Address,
26940 System.Stream_Attributes'UET_Address,
26941 System.File_Control_Block'UET_Address,
26942 Ada.Finalization.List_Controller'UET_Address);
26944 -- Table of addresses of elaboration routines. Used for
26945 -- zero cost exception handling to make sure these
26946 -- addresses are included in the top level procedure
26949 EA : aliased constant array (1 .. 23) of System.Address := (
26950 adainit'Code_Address,
26951 Do_Finalize'Code_Address,
26952 Ada.Exceptions'Elab_Spec'Address,
26953 System.Exceptions'Elab_Spec'Address,
26954 Interfaces.C_Streams'Elab_Spec'Address,
26955 System.Exception_Table'Elab_Body'Address,
26956 Ada.Io_Exceptions'Elab_Spec'Address,
26957 System.Stack_Checking'Elab_Spec'Address,
26958 System.Soft_Links'Elab_Body'Address,
26959 System.Secondary_Stack'Elab_Body'Address,
26960 Ada.Tags'Elab_Spec'Address,
26961 Ada.Tags'Elab_Body'Address,
26962 Ada.Streams'Elab_Spec'Address,
26963 System.Finalization_Root'Elab_Spec'Address,
26964 Ada.Exceptions'Elab_Body'Address,
26965 System.Finalization_Implementation'Elab_Spec'Address,
26966 System.Finalization_Implementation'Elab_Body'Address,
26967 Ada.Finalization'Elab_Spec'Address,
26968 Ada.Finalization.List_Controller'Elab_Spec'Address,
26969 System.File_Control_Block'Elab_Spec'Address,
26970 System.File_Io'Elab_Body'Address,
26971 Ada.Text_Io'Elab_Spec'Address,
26972 Ada.Text_Io'Elab_Body'Address);
26974 -- Start of processing for adainit
26978 -- Call SDP_Table_Build to build the top level procedure
26979 -- table for zero cost exception handling (omitted in
26980 -- longjmp/setjmp mode).
26982 SDP_Table_Build (ST'Address, 23, EA'Address, 23);
26984 -- Call Set_Globals to record various information for
26985 -- this partition. The values are derived by the binder
26986 -- from information stored in the ali files by the compiler.
26988 @findex __gnat_set_globals
26990 (Main_Priority => -1,
26991 -- Priority of main program, -1 if no pragma Priority used
26993 Time_Slice_Value => -1,
26994 -- Time slice from Time_Slice pragma, -1 if none used
26996 WC_Encoding => 'b',
26997 -- Wide_Character encoding used, default is brackets
26999 Locking_Policy => ' ',
27000 -- Locking_Policy used, default of space means not
27001 -- specified, otherwise it is the first character of
27002 -- the policy name.
27004 Queuing_Policy => ' ',
27005 -- Queuing_Policy used, default of space means not
27006 -- specified, otherwise it is the first character of
27007 -- the policy name.
27009 Task_Dispatching_Policy => ' ',
27010 -- Task_Dispatching_Policy used, default of space means
27011 -- not specified, otherwise first character of the
27014 Adafinal => System.Null_Address,
27015 -- Address of Adafinal routine, not used anymore
27017 Unreserve_All_Interrupts => 0,
27018 -- Set true if pragma Unreserve_All_Interrupts was used
27020 Exception_Tracebacks => 0);
27021 -- Indicates if exception tracebacks are enabled
27023 Elab_Final_Code := 1;
27025 -- Now we have the elaboration calls for all units in the partition.
27026 -- The Elab_Spec and Elab_Body attributes generate references to the
27027 -- implicit elaboration procedures generated by the compiler for
27028 -- each unit that requires elaboration.
27031 Interfaces.C_Streams'Elab_Spec;
27035 Ada.Exceptions'Elab_Spec;
27038 System.Exception_Table'Elab_Body;
27042 Ada.Io_Exceptions'Elab_Spec;
27046 System.Exceptions'Elab_Spec;
27050 System.Stack_Checking'Elab_Spec;
27053 System.Soft_Links'Elab_Body;
27058 System.Secondary_Stack'Elab_Body;
27062 Ada.Tags'Elab_Spec;
27065 Ada.Tags'Elab_Body;
27069 Ada.Streams'Elab_Spec;
27073 System.Finalization_Root'Elab_Spec;
27077 Ada.Exceptions'Elab_Body;
27081 System.Finalization_Implementation'Elab_Spec;
27084 System.Finalization_Implementation'Elab_Body;
27088 Ada.Finalization'Elab_Spec;
27092 Ada.Finalization.List_Controller'Elab_Spec;
27096 System.File_Control_Block'Elab_Spec;
27100 System.File_Io'Elab_Body;
27104 Ada.Text_Io'Elab_Spec;
27107 Ada.Text_Io'Elab_Body;
27111 Elab_Final_Code := 0;
27119 procedure adafinal is
27128 -- main is actually a function, as in the ANSI C standard,
27129 -- defined to return the exit status. The three parameters
27130 -- are the argument count, argument values and environment
27133 @findex Main Program
27136 argv : System.Address;
27137 envp : System.Address)
27140 -- The initialize routine performs low level system
27141 -- initialization using a standard library routine which
27142 -- sets up signal handling and performs any other
27143 -- required setup. The routine can be found in file
27146 @findex __gnat_initialize
27147 procedure initialize;
27148 pragma Import (C, initialize, "__gnat_initialize");
27150 -- The finalize routine performs low level system
27151 -- finalization using a standard library routine. The
27152 -- routine is found in file a-final.c and in the standard
27153 -- distribution is a dummy routine that does nothing, so
27154 -- really this is a hook for special user finalization.
27156 @findex __gnat_finalize
27157 procedure finalize;
27158 pragma Import (C, finalize, "__gnat_finalize");
27160 -- We get to the main program of the partition by using
27161 -- pragma Import because if we try to with the unit and
27162 -- call it Ada style, then not only do we waste time
27163 -- recompiling it, but also, we don't really know the right
27164 -- switches (e.g.@: identifier character set) to be used
27167 procedure Ada_Main_Program;
27168 pragma Import (Ada, Ada_Main_Program, "_ada_hello");
27170 -- Start of processing for main
27173 -- Save global variables
27179 -- Call low level system initialization
27183 -- Call our generated Ada initialization routine
27187 -- This is the point at which we want the debugger to get
27192 -- Now we call the main program of the partition
27196 -- Perform Ada finalization
27200 -- Perform low level system finalization
27204 -- Return the proper exit status
27205 return (gnat_exit_status);
27208 -- This section is entirely comments, so it has no effect on the
27209 -- compilation of the Ada_Main package. It provides the list of
27210 -- object files and linker options, as well as some standard
27211 -- libraries needed for the link. The gnatlink utility parses
27212 -- this b~hello.adb file to read these comment lines to generate
27213 -- the appropriate command line arguments for the call to the
27214 -- system linker. The BEGIN/END lines are used for sentinels for
27215 -- this parsing operation.
27217 -- The exact file names will of course depend on the environment,
27218 -- host/target and location of files on the host system.
27220 @findex Object file list
27221 -- BEGIN Object file/option list
27224 -- -L/usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/
27225 -- /usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/libgnat.a
27226 -- END Object file/option list
27232 The Ada code in the above example is exactly what is generated by the
27233 binder. We have added comments to more clearly indicate the function
27234 of each part of the generated @code{Ada_Main} package.
27236 The code is standard Ada in all respects, and can be processed by any
27237 tools that handle Ada. In particular, it is possible to use the debugger
27238 in Ada mode to debug the generated @code{Ada_Main} package. For example,
27239 suppose that for reasons that you do not understand, your program is crashing
27240 during elaboration of the body of @code{Ada.Text_IO}. To locate this bug,
27241 you can place a breakpoint on the call:
27243 @smallexample @c ada
27244 Ada.Text_Io'Elab_Body;
27248 and trace the elaboration routine for this package to find out where
27249 the problem might be (more usually of course you would be debugging
27250 elaboration code in your own application).
27252 @node Elaboration Order Handling in GNAT
27253 @appendix Elaboration Order Handling in GNAT
27254 @cindex Order of elaboration
27255 @cindex Elaboration control
27258 * Elaboration Code::
27259 * Checking the Elaboration Order::
27260 * Controlling the Elaboration Order::
27261 * Controlling Elaboration in GNAT - Internal Calls::
27262 * Controlling Elaboration in GNAT - External Calls::
27263 * Default Behavior in GNAT - Ensuring Safety::
27264 * Treatment of Pragma Elaborate::
27265 * Elaboration Issues for Library Tasks::
27266 * Mixing Elaboration Models::
27267 * What to Do If the Default Elaboration Behavior Fails::
27268 * Elaboration for Access-to-Subprogram Values::
27269 * Summary of Procedures for Elaboration Control::
27270 * Other Elaboration Order Considerations::
27274 This chapter describes the handling of elaboration code in Ada and
27275 in GNAT, and discusses how the order of elaboration of program units can
27276 be controlled in GNAT, either automatically or with explicit programming
27279 @node Elaboration Code
27280 @section Elaboration Code
27283 Ada provides rather general mechanisms for executing code at elaboration
27284 time, that is to say before the main program starts executing. Such code arises
27288 @item Initializers for variables.
27289 Variables declared at the library level, in package specs or bodies, can
27290 require initialization that is performed at elaboration time, as in:
27291 @smallexample @c ada
27293 Sqrt_Half : Float := Sqrt (0.5);
27297 @item Package initialization code
27298 Code in a @code{BEGIN-END} section at the outer level of a package body is
27299 executed as part of the package body elaboration code.
27301 @item Library level task allocators
27302 Tasks that are declared using task allocators at the library level
27303 start executing immediately and hence can execute at elaboration time.
27307 Subprogram calls are possible in any of these contexts, which means that
27308 any arbitrary part of the program may be executed as part of the elaboration
27309 code. It is even possible to write a program which does all its work at
27310 elaboration time, with a null main program, although stylistically this
27311 would usually be considered an inappropriate way to structure
27314 An important concern arises in the context of elaboration code:
27315 we have to be sure that it is executed in an appropriate order. What we
27316 have is a series of elaboration code sections, potentially one section
27317 for each unit in the program. It is important that these execute
27318 in the correct order. Correctness here means that, taking the above
27319 example of the declaration of @code{Sqrt_Half},
27320 if some other piece of
27321 elaboration code references @code{Sqrt_Half},
27322 then it must run after the
27323 section of elaboration code that contains the declaration of
27326 There would never be any order of elaboration problem if we made a rule
27327 that whenever you @code{with} a unit, you must elaborate both the spec and body
27328 of that unit before elaborating the unit doing the @code{with}'ing:
27330 @smallexample @c ada
27334 package Unit_2 is @dots{}
27340 would require that both the body and spec of @code{Unit_1} be elaborated
27341 before the spec of @code{Unit_2}. However, a rule like that would be far too
27342 restrictive. In particular, it would make it impossible to have routines
27343 in separate packages that were mutually recursive.
27345 You might think that a clever enough compiler could look at the actual
27346 elaboration code and determine an appropriate correct order of elaboration,
27347 but in the general case, this is not possible. Consider the following
27350 In the body of @code{Unit_1}, we have a procedure @code{Func_1}
27352 the variable @code{Sqrt_1}, which is declared in the elaboration code
27353 of the body of @code{Unit_1}:
27355 @smallexample @c ada
27357 Sqrt_1 : Float := Sqrt (0.1);
27362 The elaboration code of the body of @code{Unit_1} also contains:
27364 @smallexample @c ada
27367 if expression_1 = 1 then
27368 Q := Unit_2.Func_2;
27375 @code{Unit_2} is exactly parallel,
27376 it has a procedure @code{Func_2} that references
27377 the variable @code{Sqrt_2}, which is declared in the elaboration code of
27378 the body @code{Unit_2}:
27380 @smallexample @c ada
27382 Sqrt_2 : Float := Sqrt (0.1);
27387 The elaboration code of the body of @code{Unit_2} also contains:
27389 @smallexample @c ada
27392 if expression_2 = 2 then
27393 Q := Unit_1.Func_1;
27400 Now the question is, which of the following orders of elaboration is
27425 If you carefully analyze the flow here, you will see that you cannot tell
27426 at compile time the answer to this question.
27427 If @code{expression_1} is not equal to 1,
27428 and @code{expression_2} is not equal to 2,
27429 then either order is acceptable, because neither of the function calls is
27430 executed. If both tests evaluate to true, then neither order is acceptable
27431 and in fact there is no correct order.
27433 If one of the two expressions is true, and the other is false, then one
27434 of the above orders is correct, and the other is incorrect. For example,
27435 if @code{expression_1} /= 1 and @code{expression_2} = 2,
27436 then the call to @code{Func_1}
27437 will occur, but not the call to @code{Func_2.}
27438 This means that it is essential
27439 to elaborate the body of @code{Unit_1} before
27440 the body of @code{Unit_2}, so the first
27441 order of elaboration is correct and the second is wrong.
27443 By making @code{expression_1} and @code{expression_2}
27444 depend on input data, or perhaps
27445 the time of day, we can make it impossible for the compiler or binder
27446 to figure out which of these expressions will be true, and hence it
27447 is impossible to guarantee a safe order of elaboration at run time.
27449 @node Checking the Elaboration Order
27450 @section Checking the Elaboration Order
27453 In some languages that involve the same kind of elaboration problems,
27454 e.g.@: Java and C++, the programmer is expected to worry about these
27455 ordering problems himself, and it is common to
27456 write a program in which an incorrect elaboration order gives
27457 surprising results, because it references variables before they
27459 Ada is designed to be a safe language, and a programmer-beware approach is
27460 clearly not sufficient. Consequently, the language provides three lines
27464 @item Standard rules
27465 Some standard rules restrict the possible choice of elaboration
27466 order. In particular, if you @code{with} a unit, then its spec is always
27467 elaborated before the unit doing the @code{with}. Similarly, a parent
27468 spec is always elaborated before the child spec, and finally
27469 a spec is always elaborated before its corresponding body.
27471 @item Dynamic elaboration checks
27472 @cindex Elaboration checks
27473 @cindex Checks, elaboration
27474 Dynamic checks are made at run time, so that if some entity is accessed
27475 before it is elaborated (typically by means of a subprogram call)
27476 then the exception (@code{Program_Error}) is raised.
27478 @item Elaboration control
27479 Facilities are provided for the programmer to specify the desired order
27483 Let's look at these facilities in more detail. First, the rules for
27484 dynamic checking. One possible rule would be simply to say that the
27485 exception is raised if you access a variable which has not yet been
27486 elaborated. The trouble with this approach is that it could require
27487 expensive checks on every variable reference. Instead Ada has two
27488 rules which are a little more restrictive, but easier to check, and
27492 @item Restrictions on calls
27493 A subprogram can only be called at elaboration time if its body
27494 has been elaborated. The rules for elaboration given above guarantee
27495 that the spec of the subprogram has been elaborated before the
27496 call, but not the body. If this rule is violated, then the
27497 exception @code{Program_Error} is raised.
27499 @item Restrictions on instantiations
27500 A generic unit can only be instantiated if the body of the generic
27501 unit has been elaborated. Again, the rules for elaboration given above
27502 guarantee that the spec of the generic unit has been elaborated
27503 before the instantiation, but not the body. If this rule is
27504 violated, then the exception @code{Program_Error} is raised.
27508 The idea is that if the body has been elaborated, then any variables
27509 it references must have been elaborated; by checking for the body being
27510 elaborated we guarantee that none of its references causes any
27511 trouble. As we noted above, this is a little too restrictive, because a
27512 subprogram that has no non-local references in its body may in fact be safe
27513 to call. However, it really would be unsafe to rely on this, because
27514 it would mean that the caller was aware of details of the implementation
27515 in the body. This goes against the basic tenets of Ada.
27517 A plausible implementation can be described as follows.
27518 A Boolean variable is associated with each subprogram
27519 and each generic unit. This variable is initialized to False, and is set to
27520 True at the point body is elaborated. Every call or instantiation checks the
27521 variable, and raises @code{Program_Error} if the variable is False.
27523 Note that one might think that it would be good enough to have one Boolean
27524 variable for each package, but that would not deal with cases of trying
27525 to call a body in the same package as the call
27526 that has not been elaborated yet.
27527 Of course a compiler may be able to do enough analysis to optimize away
27528 some of the Boolean variables as unnecessary, and @code{GNAT} indeed
27529 does such optimizations, but still the easiest conceptual model is to
27530 think of there being one variable per subprogram.
27532 @node Controlling the Elaboration Order
27533 @section Controlling the Elaboration Order
27536 In the previous section we discussed the rules in Ada which ensure
27537 that @code{Program_Error} is raised if an incorrect elaboration order is
27538 chosen. This prevents erroneous executions, but we need mechanisms to
27539 specify a correct execution and avoid the exception altogether.
27540 To achieve this, Ada provides a number of features for controlling
27541 the order of elaboration. We discuss these features in this section.
27543 First, there are several ways of indicating to the compiler that a given
27544 unit has no elaboration problems:
27547 @item packages that do not require a body
27548 A library package that does not require a body does not permit
27549 a body (this rule was introduced in Ada 95).
27550 Thus if we have a such a package, as in:
27552 @smallexample @c ada
27555 package Definitions is
27557 type m is new integer;
27559 type a is array (1 .. 10) of m;
27560 type b is array (1 .. 20) of m;
27568 A package that @code{with}'s @code{Definitions} may safely instantiate
27569 @code{Definitions.Subp} because the compiler can determine that there
27570 definitely is no package body to worry about in this case
27573 @cindex pragma Pure
27575 Places sufficient restrictions on a unit to guarantee that
27576 no call to any subprogram in the unit can result in an
27577 elaboration problem. This means that the compiler does not need
27578 to worry about the point of elaboration of such units, and in
27579 particular, does not need to check any calls to any subprograms
27582 @item pragma Preelaborate
27583 @findex Preelaborate
27584 @cindex pragma Preelaborate
27585 This pragma places slightly less stringent restrictions on a unit than
27587 but these restrictions are still sufficient to ensure that there
27588 are no elaboration problems with any calls to the unit.
27590 @item pragma Elaborate_Body
27591 @findex Elaborate_Body
27592 @cindex pragma Elaborate_Body
27593 This pragma requires that the body of a unit be elaborated immediately
27594 after its spec. Suppose a unit @code{A} has such a pragma,
27595 and unit @code{B} does
27596 a @code{with} of unit @code{A}. Recall that the standard rules require
27597 the spec of unit @code{A}
27598 to be elaborated before the @code{with}'ing unit; given the pragma in
27599 @code{A}, we also know that the body of @code{A}
27600 will be elaborated before @code{B}, so
27601 that calls to @code{A} are safe and do not need a check.
27606 unlike pragma @code{Pure} and pragma @code{Preelaborate},
27608 @code{Elaborate_Body} does not guarantee that the program is
27609 free of elaboration problems, because it may not be possible
27610 to satisfy the requested elaboration order.
27611 Let's go back to the example with @code{Unit_1} and @code{Unit_2}.
27613 marks @code{Unit_1} as @code{Elaborate_Body},
27614 and not @code{Unit_2,} then the order of
27615 elaboration will be:
27627 Now that means that the call to @code{Func_1} in @code{Unit_2}
27628 need not be checked,
27629 it must be safe. But the call to @code{Func_2} in
27630 @code{Unit_1} may still fail if
27631 @code{Expression_1} is equal to 1,
27632 and the programmer must still take
27633 responsibility for this not being the case.
27635 If all units carry a pragma @code{Elaborate_Body}, then all problems are
27636 eliminated, except for calls entirely within a body, which are
27637 in any case fully under programmer control. However, using the pragma
27638 everywhere is not always possible.
27639 In particular, for our @code{Unit_1}/@code{Unit_2} example, if
27640 we marked both of them as having pragma @code{Elaborate_Body}, then
27641 clearly there would be no possible elaboration order.
27643 The above pragmas allow a server to guarantee safe use by clients, and
27644 clearly this is the preferable approach. Consequently a good rule
27645 is to mark units as @code{Pure} or @code{Preelaborate} if possible,
27646 and if this is not possible,
27647 mark them as @code{Elaborate_Body} if possible.
27648 As we have seen, there are situations where neither of these
27649 three pragmas can be used.
27650 So we also provide methods for clients to control the
27651 order of elaboration of the servers on which they depend:
27654 @item pragma Elaborate (unit)
27656 @cindex pragma Elaborate
27657 This pragma is placed in the context clause, after a @code{with} clause,
27658 and it requires that the body of the named unit be elaborated before
27659 the unit in which the pragma occurs. The idea is to use this pragma
27660 if the current unit calls at elaboration time, directly or indirectly,
27661 some subprogram in the named unit.
27663 @item pragma Elaborate_All (unit)
27664 @findex Elaborate_All
27665 @cindex pragma Elaborate_All
27666 This is a stronger version of the Elaborate pragma. Consider the
27670 Unit A @code{with}'s unit B and calls B.Func in elab code
27671 Unit B @code{with}'s unit C, and B.Func calls C.Func
27675 Now if we put a pragma @code{Elaborate (B)}
27676 in unit @code{A}, this ensures that the
27677 body of @code{B} is elaborated before the call, but not the
27678 body of @code{C}, so
27679 the call to @code{C.Func} could still cause @code{Program_Error} to
27682 The effect of a pragma @code{Elaborate_All} is stronger, it requires
27683 not only that the body of the named unit be elaborated before the
27684 unit doing the @code{with}, but also the bodies of all units that the
27685 named unit uses, following @code{with} links transitively. For example,
27686 if we put a pragma @code{Elaborate_All (B)} in unit @code{A},
27688 not only that the body of @code{B} be elaborated before @code{A},
27690 body of @code{C}, because @code{B} @code{with}'s @code{C}.
27694 We are now in a position to give a usage rule in Ada for avoiding
27695 elaboration problems, at least if dynamic dispatching and access to
27696 subprogram values are not used. We will handle these cases separately
27699 The rule is simple. If a unit has elaboration code that can directly or
27700 indirectly make a call to a subprogram in a @code{with}'ed unit, or instantiate
27701 a generic package in a @code{with}'ed unit,
27702 then if the @code{with}'ed unit does not have
27703 pragma @code{Pure} or @code{Preelaborate}, then the client should have
27704 a pragma @code{Elaborate_All}
27705 for the @code{with}'ed unit. By following this rule a client is
27706 assured that calls can be made without risk of an exception.
27708 For generic subprogram instantiations, the rule can be relaxed to
27709 require only a pragma @code{Elaborate} since elaborating the body
27710 of a subprogram cannot cause any transitive elaboration (we are
27711 not calling the subprogram in this case, just elaborating its
27714 If this rule is not followed, then a program may be in one of four
27718 @item No order exists
27719 No order of elaboration exists which follows the rules, taking into
27720 account any @code{Elaborate}, @code{Elaborate_All},
27721 or @code{Elaborate_Body} pragmas. In
27722 this case, an Ada compiler must diagnose the situation at bind
27723 time, and refuse to build an executable program.
27725 @item One or more orders exist, all incorrect
27726 One or more acceptable elaboration orders exist, and all of them
27727 generate an elaboration order problem. In this case, the binder
27728 can build an executable program, but @code{Program_Error} will be raised
27729 when the program is run.
27731 @item Several orders exist, some right, some incorrect
27732 One or more acceptable elaboration orders exists, and some of them
27733 work, and some do not. The programmer has not controlled
27734 the order of elaboration, so the binder may or may not pick one of
27735 the correct orders, and the program may or may not raise an
27736 exception when it is run. This is the worst case, because it means
27737 that the program may fail when moved to another compiler, or even
27738 another version of the same compiler.
27740 @item One or more orders exists, all correct
27741 One ore more acceptable elaboration orders exist, and all of them
27742 work. In this case the program runs successfully. This state of
27743 affairs can be guaranteed by following the rule we gave above, but
27744 may be true even if the rule is not followed.
27748 Note that one additional advantage of following our rules on the use
27749 of @code{Elaborate} and @code{Elaborate_All}
27750 is that the program continues to stay in the ideal (all orders OK) state
27751 even if maintenance
27752 changes some bodies of some units. Conversely, if a program that does
27753 not follow this rule happens to be safe at some point, this state of affairs
27754 may deteriorate silently as a result of maintenance changes.
27756 You may have noticed that the above discussion did not mention
27757 the use of @code{Elaborate_Body}. This was a deliberate omission. If you
27758 @code{with} an @code{Elaborate_Body} unit, it still may be the case that
27759 code in the body makes calls to some other unit, so it is still necessary
27760 to use @code{Elaborate_All} on such units.
27762 @node Controlling Elaboration in GNAT - Internal Calls
27763 @section Controlling Elaboration in GNAT - Internal Calls
27766 In the case of internal calls, i.e., calls within a single package, the
27767 programmer has full control over the order of elaboration, and it is up
27768 to the programmer to elaborate declarations in an appropriate order. For
27771 @smallexample @c ada
27774 function One return Float;
27778 function One return Float is
27787 will obviously raise @code{Program_Error} at run time, because function
27788 One will be called before its body is elaborated. In this case GNAT will
27789 generate a warning that the call will raise @code{Program_Error}:
27795 2. function One return Float;
27797 4. Q : Float := One;
27799 >>> warning: cannot call "One" before body is elaborated
27800 >>> warning: Program_Error will be raised at run time
27803 6. function One return Float is
27816 Note that in this particular case, it is likely that the call is safe, because
27817 the function @code{One} does not access any global variables.
27818 Nevertheless in Ada, we do not want the validity of the check to depend on
27819 the contents of the body (think about the separate compilation case), so this
27820 is still wrong, as we discussed in the previous sections.
27822 The error is easily corrected by rearranging the declarations so that the
27823 body of @code{One} appears before the declaration containing the call
27824 (note that in Ada 95 and Ada 2005,
27825 declarations can appear in any order, so there is no restriction that
27826 would prevent this reordering, and if we write:
27828 @smallexample @c ada
27831 function One return Float;
27833 function One return Float is
27844 then all is well, no warning is generated, and no
27845 @code{Program_Error} exception
27847 Things are more complicated when a chain of subprograms is executed:
27849 @smallexample @c ada
27852 function A return Integer;
27853 function B return Integer;
27854 function C return Integer;
27856 function B return Integer is begin return A; end;
27857 function C return Integer is begin return B; end;
27861 function A return Integer is begin return 1; end;
27867 Now the call to @code{C}
27868 at elaboration time in the declaration of @code{X} is correct, because
27869 the body of @code{C} is already elaborated,
27870 and the call to @code{B} within the body of
27871 @code{C} is correct, but the call
27872 to @code{A} within the body of @code{B} is incorrect, because the body
27873 of @code{A} has not been elaborated, so @code{Program_Error}
27874 will be raised on the call to @code{A}.
27875 In this case GNAT will generate a
27876 warning that @code{Program_Error} may be
27877 raised at the point of the call. Let's look at the warning:
27883 2. function A return Integer;
27884 3. function B return Integer;
27885 4. function C return Integer;
27887 6. function B return Integer is begin return A; end;
27889 >>> warning: call to "A" before body is elaborated may
27890 raise Program_Error
27891 >>> warning: "B" called at line 7
27892 >>> warning: "C" called at line 9
27894 7. function C return Integer is begin return B; end;
27896 9. X : Integer := C;
27898 11. function A return Integer is begin return 1; end;
27908 Note that the message here says ``may raise'', instead of the direct case,
27909 where the message says ``will be raised''. That's because whether
27911 actually called depends in general on run-time flow of control.
27912 For example, if the body of @code{B} said
27914 @smallexample @c ada
27917 function B return Integer is
27919 if some-condition-depending-on-input-data then
27930 then we could not know until run time whether the incorrect call to A would
27931 actually occur, so @code{Program_Error} might
27932 or might not be raised. It is possible for a compiler to
27933 do a better job of analyzing bodies, to
27934 determine whether or not @code{Program_Error}
27935 might be raised, but it certainly
27936 couldn't do a perfect job (that would require solving the halting problem
27937 and is provably impossible), and because this is a warning anyway, it does
27938 not seem worth the effort to do the analysis. Cases in which it
27939 would be relevant are rare.
27941 In practice, warnings of either of the forms given
27942 above will usually correspond to
27943 real errors, and should be examined carefully and eliminated.
27944 In the rare case where a warning is bogus, it can be suppressed by any of
27945 the following methods:
27949 Compile with the @option{-gnatws} switch set
27952 Suppress @code{Elaboration_Check} for the called subprogram
27955 Use pragma @code{Warnings_Off} to turn warnings off for the call
27959 For the internal elaboration check case,
27960 GNAT by default generates the
27961 necessary run-time checks to ensure
27962 that @code{Program_Error} is raised if any
27963 call fails an elaboration check. Of course this can only happen if a
27964 warning has been issued as described above. The use of pragma
27965 @code{Suppress (Elaboration_Check)} may (but is not guaranteed to) suppress
27966 some of these checks, meaning that it may be possible (but is not
27967 guaranteed) for a program to be able to call a subprogram whose body
27968 is not yet elaborated, without raising a @code{Program_Error} exception.
27970 @node Controlling Elaboration in GNAT - External Calls
27971 @section Controlling Elaboration in GNAT - External Calls
27974 The previous section discussed the case in which the execution of a
27975 particular thread of elaboration code occurred entirely within a
27976 single unit. This is the easy case to handle, because a programmer
27977 has direct and total control over the order of elaboration, and
27978 furthermore, checks need only be generated in cases which are rare
27979 and which the compiler can easily detect.
27980 The situation is more complex when separate compilation is taken into account.
27981 Consider the following:
27983 @smallexample @c ada
27987 function Sqrt (Arg : Float) return Float;
27990 package body Math is
27991 function Sqrt (Arg : Float) return Float is
28000 X : Float := Math.Sqrt (0.5);
28013 where @code{Main} is the main program. When this program is executed, the
28014 elaboration code must first be executed, and one of the jobs of the
28015 binder is to determine the order in which the units of a program are
28016 to be elaborated. In this case we have four units: the spec and body
28018 the spec of @code{Stuff} and the body of @code{Main}).
28019 In what order should the four separate sections of elaboration code
28022 There are some restrictions in the order of elaboration that the binder
28023 can choose. In particular, if unit U has a @code{with}
28024 for a package @code{X}, then you
28025 are assured that the spec of @code{X}
28026 is elaborated before U , but you are
28027 not assured that the body of @code{X}
28028 is elaborated before U.
28029 This means that in the above case, the binder is allowed to choose the
28040 but that's not good, because now the call to @code{Math.Sqrt}
28041 that happens during
28042 the elaboration of the @code{Stuff}
28043 spec happens before the body of @code{Math.Sqrt} is
28044 elaborated, and hence causes @code{Program_Error} exception to be raised.
28045 At first glance, one might say that the binder is misbehaving, because
28046 obviously you want to elaborate the body of something you @code{with}
28048 that is not a general rule that can be followed in all cases. Consider
28050 @smallexample @c ada
28053 package X is @dots{}
28055 package Y is @dots{}
28058 package body Y is @dots{}
28061 package body X is @dots{}
28067 This is a common arrangement, and, apart from the order of elaboration
28068 problems that might arise in connection with elaboration code, this works fine.
28069 A rule that says that you must first elaborate the body of anything you
28070 @code{with} cannot work in this case:
28071 the body of @code{X} @code{with}'s @code{Y},
28072 which means you would have to
28073 elaborate the body of @code{Y} first, but that @code{with}'s @code{X},
28075 you have to elaborate the body of @code{X} first, but @dots{} and we have a
28076 loop that cannot be broken.
28078 It is true that the binder can in many cases guess an order of elaboration
28079 that is unlikely to cause a @code{Program_Error}
28080 exception to be raised, and it tries to do so (in the
28081 above example of @code{Math/Stuff/Spec}, the GNAT binder will
28083 elaborate the body of @code{Math} right after its spec, so all will be well).
28085 However, a program that blindly relies on the binder to be helpful can
28086 get into trouble, as we discussed in the previous sections, so
28088 provides a number of facilities for assisting the programmer in
28089 developing programs that are robust with respect to elaboration order.
28091 @node Default Behavior in GNAT - Ensuring Safety
28092 @section Default Behavior in GNAT - Ensuring Safety
28095 The default behavior in GNAT ensures elaboration safety. In its
28096 default mode GNAT implements the
28097 rule we previously described as the right approach. Let's restate it:
28101 @emph{If a unit has elaboration code that can directly or indirectly make a
28102 call to a subprogram in a @code{with}'ed unit, or instantiate a generic
28103 package in a @code{with}'ed unit, then if the @code{with}'ed unit
28104 does not have pragma @code{Pure} or
28105 @code{Preelaborate}, then the client should have an
28106 @code{Elaborate_All} pragma for the @code{with}'ed unit.}
28108 @emph{In the case of instantiating a generic subprogram, it is always
28109 sufficient to have only an @code{Elaborate} pragma for the
28110 @code{with}'ed unit.}
28114 By following this rule a client is assured that calls and instantiations
28115 can be made without risk of an exception.
28117 In this mode GNAT traces all calls that are potentially made from
28118 elaboration code, and puts in any missing implicit @code{Elaborate}
28119 and @code{Elaborate_All} pragmas.
28120 The advantage of this approach is that no elaboration problems
28121 are possible if the binder can find an elaboration order that is
28122 consistent with these implicit @code{Elaborate} and
28123 @code{Elaborate_All} pragmas. The
28124 disadvantage of this approach is that no such order may exist.
28126 If the binder does not generate any diagnostics, then it means that it has
28127 found an elaboration order that is guaranteed to be safe. However, the binder
28128 may still be relying on implicitly generated @code{Elaborate} and
28129 @code{Elaborate_All} pragmas so portability to other compilers than GNAT is not
28132 If it is important to guarantee portability, then the compilations should
28135 (warn on elaboration problems) switch. This will cause warning messages
28136 to be generated indicating the missing @code{Elaborate} and
28137 @code{Elaborate_All} pragmas.
28138 Consider the following source program:
28140 @smallexample @c ada
28145 m : integer := k.r;
28152 where it is clear that there
28153 should be a pragma @code{Elaborate_All}
28154 for unit @code{k}. An implicit pragma will be generated, and it is
28155 likely that the binder will be able to honor it. However, if you want
28156 to port this program to some other Ada compiler than GNAT.
28157 it is safer to include the pragma explicitly in the source. If this
28158 unit is compiled with the
28160 switch, then the compiler outputs a warning:
28167 3. m : integer := k.r;
28169 >>> warning: call to "r" may raise Program_Error
28170 >>> warning: missing pragma Elaborate_All for "k"
28178 and these warnings can be used as a guide for supplying manually
28179 the missing pragmas. It is usually a bad idea to use this warning
28180 option during development. That's because it will warn you when
28181 you need to put in a pragma, but cannot warn you when it is time
28182 to take it out. So the use of pragma @code{Elaborate_All} may lead to
28183 unnecessary dependencies and even false circularities.
28185 This default mode is more restrictive than the Ada Reference
28186 Manual, and it is possible to construct programs which will compile
28187 using the dynamic model described there, but will run into a
28188 circularity using the safer static model we have described.
28190 Of course any Ada compiler must be able to operate in a mode
28191 consistent with the requirements of the Ada Reference Manual,
28192 and in particular must have the capability of implementing the
28193 standard dynamic model of elaboration with run-time checks.
28195 In GNAT, this standard mode can be achieved either by the use of
28196 the @option{-gnatE} switch on the compiler (@command{gcc} or
28197 @command{gnatmake}) command, or by the use of the configuration pragma:
28199 @smallexample @c ada
28200 pragma Elaboration_Checks (RM);
28204 Either approach will cause the unit affected to be compiled using the
28205 standard dynamic run-time elaboration checks described in the Ada
28206 Reference Manual. The static model is generally preferable, since it
28207 is clearly safer to rely on compile and link time checks rather than
28208 run-time checks. However, in the case of legacy code, it may be
28209 difficult to meet the requirements of the static model. This
28210 issue is further discussed in
28211 @ref{What to Do If the Default Elaboration Behavior Fails}.
28213 Note that the static model provides a strict subset of the allowed
28214 behavior and programs of the Ada Reference Manual, so if you do
28215 adhere to the static model and no circularities exist,
28216 then you are assured that your program will
28217 work using the dynamic model, providing that you remove any
28218 pragma Elaborate statements from the source.
28220 @node Treatment of Pragma Elaborate
28221 @section Treatment of Pragma Elaborate
28222 @cindex Pragma Elaborate
28225 The use of @code{pragma Elaborate}
28226 should generally be avoided in Ada 95 and Ada 2005 programs,
28227 since there is no guarantee that transitive calls
28228 will be properly handled. Indeed at one point, this pragma was placed
28229 in Annex J (Obsolescent Features), on the grounds that it is never useful.
28231 Now that's a bit restrictive. In practice, the case in which
28232 @code{pragma Elaborate} is useful is when the caller knows that there
28233 are no transitive calls, or that the called unit contains all necessary
28234 transitive @code{pragma Elaborate} statements, and legacy code often
28235 contains such uses.
28237 Strictly speaking the static mode in GNAT should ignore such pragmas,
28238 since there is no assurance at compile time that the necessary safety
28239 conditions are met. In practice, this would cause GNAT to be incompatible
28240 with correctly written Ada 83 code that had all necessary
28241 @code{pragma Elaborate} statements in place. Consequently, we made the
28242 decision that GNAT in its default mode will believe that if it encounters
28243 a @code{pragma Elaborate} then the programmer knows what they are doing,
28244 and it will trust that no elaboration errors can occur.
28246 The result of this decision is two-fold. First to be safe using the
28247 static mode, you should remove all @code{pragma Elaborate} statements.
28248 Second, when fixing circularities in existing code, you can selectively
28249 use @code{pragma Elaborate} statements to convince the static mode of
28250 GNAT that it need not generate an implicit @code{pragma Elaborate_All}
28253 When using the static mode with @option{-gnatwl}, any use of
28254 @code{pragma Elaborate} will generate a warning about possible
28257 @node Elaboration Issues for Library Tasks
28258 @section Elaboration Issues for Library Tasks
28259 @cindex Library tasks, elaboration issues
28260 @cindex Elaboration of library tasks
28263 In this section we examine special elaboration issues that arise for
28264 programs that declare library level tasks.
28266 Generally the model of execution of an Ada program is that all units are
28267 elaborated, and then execution of the program starts. However, the
28268 declaration of library tasks definitely does not fit this model. The
28269 reason for this is that library tasks start as soon as they are declared
28270 (more precisely, as soon as the statement part of the enclosing package
28271 body is reached), that is to say before elaboration
28272 of the program is complete. This means that if such a task calls a
28273 subprogram, or an entry in another task, the callee may or may not be
28274 elaborated yet, and in the standard
28275 Reference Manual model of dynamic elaboration checks, you can even
28276 get timing dependent Program_Error exceptions, since there can be
28277 a race between the elaboration code and the task code.
28279 The static model of elaboration in GNAT seeks to avoid all such
28280 dynamic behavior, by being conservative, and the conservative
28281 approach in this particular case is to assume that all the code
28282 in a task body is potentially executed at elaboration time if
28283 a task is declared at the library level.
28285 This can definitely result in unexpected circularities. Consider
28286 the following example
28288 @smallexample @c ada
28294 type My_Int is new Integer;
28296 function Ident (M : My_Int) return My_Int;
28300 package body Decls is
28301 task body Lib_Task is
28307 function Ident (M : My_Int) return My_Int is
28315 procedure Put_Val (Arg : Decls.My_Int);
28319 package body Utils is
28320 procedure Put_Val (Arg : Decls.My_Int) is
28322 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
28329 Decls.Lib_Task.Start;
28334 If the above example is compiled in the default static elaboration
28335 mode, then a circularity occurs. The circularity comes from the call
28336 @code{Utils.Put_Val} in the task body of @code{Decls.Lib_Task}. Since
28337 this call occurs in elaboration code, we need an implicit pragma
28338 @code{Elaborate_All} for @code{Utils}. This means that not only must
28339 the spec and body of @code{Utils} be elaborated before the body
28340 of @code{Decls}, but also the spec and body of any unit that is
28341 @code{with'ed} by the body of @code{Utils} must also be elaborated before
28342 the body of @code{Decls}. This is the transitive implication of
28343 pragma @code{Elaborate_All} and it makes sense, because in general
28344 the body of @code{Put_Val} might have a call to something in a
28345 @code{with'ed} unit.
28347 In this case, the body of Utils (actually its spec) @code{with's}
28348 @code{Decls}. Unfortunately this means that the body of @code{Decls}
28349 must be elaborated before itself, in case there is a call from the
28350 body of @code{Utils}.
28352 Here is the exact chain of events we are worrying about:
28356 In the body of @code{Decls} a call is made from within the body of a library
28357 task to a subprogram in the package @code{Utils}. Since this call may
28358 occur at elaboration time (given that the task is activated at elaboration
28359 time), we have to assume the worst, i.e., that the
28360 call does happen at elaboration time.
28363 This means that the body and spec of @code{Util} must be elaborated before
28364 the body of @code{Decls} so that this call does not cause an access before
28368 Within the body of @code{Util}, specifically within the body of
28369 @code{Util.Put_Val} there may be calls to any unit @code{with}'ed
28373 One such @code{with}'ed package is package @code{Decls}, so there
28374 might be a call to a subprogram in @code{Decls} in @code{Put_Val}.
28375 In fact there is such a call in this example, but we would have to
28376 assume that there was such a call even if it were not there, since
28377 we are not supposed to write the body of @code{Decls} knowing what
28378 is in the body of @code{Utils}; certainly in the case of the
28379 static elaboration model, the compiler does not know what is in
28380 other bodies and must assume the worst.
28383 This means that the spec and body of @code{Decls} must also be
28384 elaborated before we elaborate the unit containing the call, but
28385 that unit is @code{Decls}! This means that the body of @code{Decls}
28386 must be elaborated before itself, and that's a circularity.
28390 Indeed, if you add an explicit pragma @code{Elaborate_All} for @code{Utils} in
28391 the body of @code{Decls} you will get a true Ada Reference Manual
28392 circularity that makes the program illegal.
28394 In practice, we have found that problems with the static model of
28395 elaboration in existing code often arise from library tasks, so
28396 we must address this particular situation.
28398 Note that if we compile and run the program above, using the dynamic model of
28399 elaboration (that is to say use the @option{-gnatE} switch),
28400 then it compiles, binds,
28401 links, and runs, printing the expected result of 2. Therefore in some sense
28402 the circularity here is only apparent, and we need to capture
28403 the properties of this program that distinguish it from other library-level
28404 tasks that have real elaboration problems.
28406 We have four possible answers to this question:
28411 Use the dynamic model of elaboration.
28413 If we use the @option{-gnatE} switch, then as noted above, the program works.
28414 Why is this? If we examine the task body, it is apparent that the task cannot
28416 @code{accept} statement until after elaboration has been completed, because
28417 the corresponding entry call comes from the main program, not earlier.
28418 This is why the dynamic model works here. But that's really giving
28419 up on a precise analysis, and we prefer to take this approach only if we cannot
28421 problem in any other manner. So let us examine two ways to reorganize
28422 the program to avoid the potential elaboration problem.
28425 Split library tasks into separate packages.
28427 Write separate packages, so that library tasks are isolated from
28428 other declarations as much as possible. Let us look at a variation on
28431 @smallexample @c ada
28439 package body Decls1 is
28440 task body Lib_Task is
28448 type My_Int is new Integer;
28449 function Ident (M : My_Int) return My_Int;
28453 package body Decls2 is
28454 function Ident (M : My_Int) return My_Int is
28462 procedure Put_Val (Arg : Decls2.My_Int);
28466 package body Utils is
28467 procedure Put_Val (Arg : Decls2.My_Int) is
28469 Text_IO.Put_Line (Decls2.My_Int'Image (Decls2.Ident (Arg)));
28476 Decls1.Lib_Task.Start;
28481 All we have done is to split @code{Decls} into two packages, one
28482 containing the library task, and one containing everything else. Now
28483 there is no cycle, and the program compiles, binds, links and executes
28484 using the default static model of elaboration.
28487 Declare separate task types.
28489 A significant part of the problem arises because of the use of the
28490 single task declaration form. This means that the elaboration of
28491 the task type, and the elaboration of the task itself (i.e.@: the
28492 creation of the task) happen at the same time. A good rule
28493 of style in Ada is to always create explicit task types. By
28494 following the additional step of placing task objects in separate
28495 packages from the task type declaration, many elaboration problems
28496 are avoided. Here is another modified example of the example program:
28498 @smallexample @c ada
28500 task type Lib_Task_Type is
28504 type My_Int is new Integer;
28506 function Ident (M : My_Int) return My_Int;
28510 package body Decls is
28511 task body Lib_Task_Type is
28517 function Ident (M : My_Int) return My_Int is
28525 procedure Put_Val (Arg : Decls.My_Int);
28529 package body Utils is
28530 procedure Put_Val (Arg : Decls.My_Int) is
28532 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
28538 Lib_Task : Decls.Lib_Task_Type;
28544 Declst.Lib_Task.Start;
28549 What we have done here is to replace the @code{task} declaration in
28550 package @code{Decls} with a @code{task type} declaration. Then we
28551 introduce a separate package @code{Declst} to contain the actual
28552 task object. This separates the elaboration issues for
28553 the @code{task type}
28554 declaration, which causes no trouble, from the elaboration issues
28555 of the task object, which is also unproblematic, since it is now independent
28556 of the elaboration of @code{Utils}.
28557 This separation of concerns also corresponds to
28558 a generally sound engineering principle of separating declarations
28559 from instances. This version of the program also compiles, binds, links,
28560 and executes, generating the expected output.
28563 Use No_Entry_Calls_In_Elaboration_Code restriction.
28564 @cindex No_Entry_Calls_In_Elaboration_Code
28566 The previous two approaches described how a program can be restructured
28567 to avoid the special problems caused by library task bodies. in practice,
28568 however, such restructuring may be difficult to apply to existing legacy code,
28569 so we must consider solutions that do not require massive rewriting.
28571 Let us consider more carefully why our original sample program works
28572 under the dynamic model of elaboration. The reason is that the code
28573 in the task body blocks immediately on the @code{accept}
28574 statement. Now of course there is nothing to prohibit elaboration
28575 code from making entry calls (for example from another library level task),
28576 so we cannot tell in isolation that
28577 the task will not execute the accept statement during elaboration.
28579 However, in practice it is very unusual to see elaboration code
28580 make any entry calls, and the pattern of tasks starting
28581 at elaboration time and then immediately blocking on @code{accept} or
28582 @code{select} statements is very common. What this means is that
28583 the compiler is being too pessimistic when it analyzes the
28584 whole package body as though it might be executed at elaboration
28587 If we know that the elaboration code contains no entry calls, (a very safe
28588 assumption most of the time, that could almost be made the default
28589 behavior), then we can compile all units of the program under control
28590 of the following configuration pragma:
28593 pragma Restrictions (No_Entry_Calls_In_Elaboration_Code);
28597 This pragma can be placed in the @file{gnat.adc} file in the usual
28598 manner. If we take our original unmodified program and compile it
28599 in the presence of a @file{gnat.adc} containing the above pragma,
28600 then once again, we can compile, bind, link, and execute, obtaining
28601 the expected result. In the presence of this pragma, the compiler does
28602 not trace calls in a task body, that appear after the first @code{accept}
28603 or @code{select} statement, and therefore does not report a potential
28604 circularity in the original program.
28606 The compiler will check to the extent it can that the above
28607 restriction is not violated, but it is not always possible to do a
28608 complete check at compile time, so it is important to use this
28609 pragma only if the stated restriction is in fact met, that is to say
28610 no task receives an entry call before elaboration of all units is completed.
28614 @node Mixing Elaboration Models
28615 @section Mixing Elaboration Models
28617 So far, we have assumed that the entire program is either compiled
28618 using the dynamic model or static model, ensuring consistency. It
28619 is possible to mix the two models, but rules have to be followed
28620 if this mixing is done to ensure that elaboration checks are not
28623 The basic rule is that @emph{a unit compiled with the static model cannot
28624 be @code{with'ed} by a unit compiled with the dynamic model}. The
28625 reason for this is that in the static model, a unit assumes that
28626 its clients guarantee to use (the equivalent of) pragma
28627 @code{Elaborate_All} so that no elaboration checks are required
28628 in inner subprograms, and this assumption is violated if the
28629 client is compiled with dynamic checks.
28631 The precise rule is as follows. A unit that is compiled with dynamic
28632 checks can only @code{with} a unit that meets at least one of the
28633 following criteria:
28638 The @code{with'ed} unit is itself compiled with dynamic elaboration
28639 checks (that is with the @option{-gnatE} switch.
28642 The @code{with'ed} unit is an internal GNAT implementation unit from
28643 the System, Interfaces, Ada, or GNAT hierarchies.
28646 The @code{with'ed} unit has pragma Preelaborate or pragma Pure.
28649 The @code{with'ing} unit (that is the client) has an explicit pragma
28650 @code{Elaborate_All} for the @code{with'ed} unit.
28655 If this rule is violated, that is if a unit with dynamic elaboration
28656 checks @code{with's} a unit that does not meet one of the above four
28657 criteria, then the binder (@code{gnatbind}) will issue a warning
28658 similar to that in the following example:
28661 warning: "x.ads" has dynamic elaboration checks and with's
28662 warning: "y.ads" which has static elaboration checks
28666 These warnings indicate that the rule has been violated, and that as a result
28667 elaboration checks may be missed in the resulting executable file.
28668 This warning may be suppressed using the @option{-ws} binder switch
28669 in the usual manner.
28671 One useful application of this mixing rule is in the case of a subsystem
28672 which does not itself @code{with} units from the remainder of the
28673 application. In this case, the entire subsystem can be compiled with
28674 dynamic checks to resolve a circularity in the subsystem, while
28675 allowing the main application that uses this subsystem to be compiled
28676 using the more reliable default static model.
28678 @node What to Do If the Default Elaboration Behavior Fails
28679 @section What to Do If the Default Elaboration Behavior Fails
28682 If the binder cannot find an acceptable order, it outputs detailed
28683 diagnostics. For example:
28689 error: elaboration circularity detected
28690 info: "proc (body)" must be elaborated before "pack (body)"
28691 info: reason: Elaborate_All probably needed in unit "pack (body)"
28692 info: recompile "pack (body)" with -gnatwl
28693 info: for full details
28694 info: "proc (body)"
28695 info: is needed by its spec:
28696 info: "proc (spec)"
28697 info: which is withed by:
28698 info: "pack (body)"
28699 info: "pack (body)" must be elaborated before "proc (body)"
28700 info: reason: pragma Elaborate in unit "proc (body)"
28706 In this case we have a cycle that the binder cannot break. On the one
28707 hand, there is an explicit pragma Elaborate in @code{proc} for
28708 @code{pack}. This means that the body of @code{pack} must be elaborated
28709 before the body of @code{proc}. On the other hand, there is elaboration
28710 code in @code{pack} that calls a subprogram in @code{proc}. This means
28711 that for maximum safety, there should really be a pragma
28712 Elaborate_All in @code{pack} for @code{proc} which would require that
28713 the body of @code{proc} be elaborated before the body of
28714 @code{pack}. Clearly both requirements cannot be satisfied.
28715 Faced with a circularity of this kind, you have three different options.
28718 @item Fix the program
28719 The most desirable option from the point of view of long-term maintenance
28720 is to rearrange the program so that the elaboration problems are avoided.
28721 One useful technique is to place the elaboration code into separate
28722 child packages. Another is to move some of the initialization code to
28723 explicitly called subprograms, where the program controls the order
28724 of initialization explicitly. Although this is the most desirable option,
28725 it may be impractical and involve too much modification, especially in
28726 the case of complex legacy code.
28728 @item Perform dynamic checks
28729 If the compilations are done using the
28731 (dynamic elaboration check) switch, then GNAT behaves in a quite different
28732 manner. Dynamic checks are generated for all calls that could possibly result
28733 in raising an exception. With this switch, the compiler does not generate
28734 implicit @code{Elaborate} or @code{Elaborate_All} pragmas. The behavior then is
28735 exactly as specified in the @cite{Ada Reference Manual}.
28736 The binder will generate
28737 an executable program that may or may not raise @code{Program_Error}, and then
28738 it is the programmer's job to ensure that it does not raise an exception. Note
28739 that it is important to compile all units with the switch, it cannot be used
28742 @item Suppress checks
28743 The drawback of dynamic checks is that they generate a
28744 significant overhead at run time, both in space and time. If you
28745 are absolutely sure that your program cannot raise any elaboration
28746 exceptions, and you still want to use the dynamic elaboration model,
28747 then you can use the configuration pragma
28748 @code{Suppress (Elaboration_Check)} to suppress all such checks. For
28749 example this pragma could be placed in the @file{gnat.adc} file.
28751 @item Suppress checks selectively
28752 When you know that certain calls or instantiations in elaboration code cannot
28753 possibly lead to an elaboration error, and the binder nevertheless complains
28754 about implicit @code{Elaborate} and @code{Elaborate_All} pragmas that lead to
28755 elaboration circularities, it is possible to remove those warnings locally and
28756 obtain a program that will bind. Clearly this can be unsafe, and it is the
28757 responsibility of the programmer to make sure that the resulting program has no
28758 elaboration anomalies. The pragma @code{Suppress (Elaboration_Check)} can be
28759 used with different granularity to suppress warnings and break elaboration
28764 Place the pragma that names the called subprogram in the declarative part
28765 that contains the call.
28768 Place the pragma in the declarative part, without naming an entity. This
28769 disables warnings on all calls in the corresponding declarative region.
28772 Place the pragma in the package spec that declares the called subprogram,
28773 and name the subprogram. This disables warnings on all elaboration calls to
28777 Place the pragma in the package spec that declares the called subprogram,
28778 without naming any entity. This disables warnings on all elaboration calls to
28779 all subprograms declared in this spec.
28781 @item Use Pragma Elaborate
28782 As previously described in section @xref{Treatment of Pragma Elaborate},
28783 GNAT in static mode assumes that a @code{pragma} Elaborate indicates correctly
28784 that no elaboration checks are required on calls to the designated unit.
28785 There may be cases in which the caller knows that no transitive calls
28786 can occur, so that a @code{pragma Elaborate} will be sufficient in a
28787 case where @code{pragma Elaborate_All} would cause a circularity.
28791 These five cases are listed in order of decreasing safety, and therefore
28792 require increasing programmer care in their application. Consider the
28795 @smallexample @c adanocomment
28797 function F1 return Integer;
28802 function F2 return Integer;
28803 function Pure (x : integer) return integer;
28804 -- pragma Suppress (Elaboration_Check, On => Pure); -- (3)
28805 -- pragma Suppress (Elaboration_Check); -- (4)
28809 package body Pack1 is
28810 function F1 return Integer is
28814 Val : integer := Pack2.Pure (11); -- Elab. call (1)
28817 -- pragma Suppress(Elaboration_Check, Pack2.F2); -- (1)
28818 -- pragma Suppress(Elaboration_Check); -- (2)
28820 X1 := Pack2.F2 + 1; -- Elab. call (2)
28825 package body Pack2 is
28826 function F2 return Integer is
28830 function Pure (x : integer) return integer is
28832 return x ** 3 - 3 * x;
28836 with Pack1, Ada.Text_IO;
28839 Ada.Text_IO.Put_Line(Pack1.X1'Img); -- 101
28842 In the absence of any pragmas, an attempt to bind this program produces
28843 the following diagnostics:
28849 error: elaboration circularity detected
28850 info: "pack1 (body)" must be elaborated before "pack1 (body)"
28851 info: reason: Elaborate_All probably needed in unit "pack1 (body)"
28852 info: recompile "pack1 (body)" with -gnatwl for full details
28853 info: "pack1 (body)"
28854 info: must be elaborated along with its spec:
28855 info: "pack1 (spec)"
28856 info: which is withed by:
28857 info: "pack2 (body)"
28858 info: which must be elaborated along with its spec:
28859 info: "pack2 (spec)"
28860 info: which is withed by:
28861 info: "pack1 (body)"
28864 The sources of the circularity are the two calls to @code{Pack2.Pure} and
28865 @code{Pack2.F2} in the body of @code{Pack1}. We can see that the call to
28866 F2 is safe, even though F2 calls F1, because the call appears after the
28867 elaboration of the body of F1. Therefore the pragma (1) is safe, and will
28868 remove the warning on the call. It is also possible to use pragma (2)
28869 because there are no other potentially unsafe calls in the block.
28872 The call to @code{Pure} is safe because this function does not depend on the
28873 state of @code{Pack2}. Therefore any call to this function is safe, and it
28874 is correct to place pragma (3) in the corresponding package spec.
28877 Finally, we could place pragma (4) in the spec of @code{Pack2} to disable
28878 warnings on all calls to functions declared therein. Note that this is not
28879 necessarily safe, and requires more detailed examination of the subprogram
28880 bodies involved. In particular, a call to @code{F2} requires that @code{F1}
28881 be already elaborated.
28885 It is hard to generalize on which of these four approaches should be
28886 taken. Obviously if it is possible to fix the program so that the default
28887 treatment works, this is preferable, but this may not always be practical.
28888 It is certainly simple enough to use
28890 but the danger in this case is that, even if the GNAT binder
28891 finds a correct elaboration order, it may not always do so,
28892 and certainly a binder from another Ada compiler might not. A
28893 combination of testing and analysis (for which the warnings generated
28896 switch can be useful) must be used to ensure that the program is free
28897 of errors. One switch that is useful in this testing is the
28898 @option{^-p (pessimistic elaboration order)^/PESSIMISTIC_ELABORATION_ORDER^}
28901 Normally the binder tries to find an order that has the best chance
28902 of avoiding elaboration problems. However, if this switch is used, the binder
28903 plays a devil's advocate role, and tries to choose the order that
28904 has the best chance of failing. If your program works even with this
28905 switch, then it has a better chance of being error free, but this is still
28908 For an example of this approach in action, consider the C-tests (executable
28909 tests) from the ACVC suite. If these are compiled and run with the default
28910 treatment, then all but one of them succeed without generating any error
28911 diagnostics from the binder. However, there is one test that fails, and
28912 this is not surprising, because the whole point of this test is to ensure
28913 that the compiler can handle cases where it is impossible to determine
28914 a correct order statically, and it checks that an exception is indeed
28915 raised at run time.
28917 This one test must be compiled and run using the
28919 switch, and then it passes. Alternatively, the entire suite can
28920 be run using this switch. It is never wrong to run with the dynamic
28921 elaboration switch if your code is correct, and we assume that the
28922 C-tests are indeed correct (it is less efficient, but efficiency is
28923 not a factor in running the ACVC tests.)
28925 @node Elaboration for Access-to-Subprogram Values
28926 @section Elaboration for Access-to-Subprogram Values
28927 @cindex Access-to-subprogram
28930 Access-to-subprogram types (introduced in Ada 95) complicate
28931 the handling of elaboration. The trouble is that it becomes
28932 impossible to tell at compile time which procedure
28933 is being called. This means that it is not possible for the binder
28934 to analyze the elaboration requirements in this case.
28936 If at the point at which the access value is created
28937 (i.e., the evaluation of @code{P'Access} for a subprogram @code{P}),
28938 the body of the subprogram is
28939 known to have been elaborated, then the access value is safe, and its use
28940 does not require a check. This may be achieved by appropriate arrangement
28941 of the order of declarations if the subprogram is in the current unit,
28942 or, if the subprogram is in another unit, by using pragma
28943 @code{Pure}, @code{Preelaborate}, or @code{Elaborate_Body}
28944 on the referenced unit.
28946 If the referenced body is not known to have been elaborated at the point
28947 the access value is created, then any use of the access value must do a
28948 dynamic check, and this dynamic check will fail and raise a
28949 @code{Program_Error} exception if the body has not been elaborated yet.
28950 GNAT will generate the necessary checks, and in addition, if the
28952 switch is set, will generate warnings that such checks are required.
28954 The use of dynamic dispatching for tagged types similarly generates
28955 a requirement for dynamic checks, and premature calls to any primitive
28956 operation of a tagged type before the body of the operation has been
28957 elaborated, will result in the raising of @code{Program_Error}.
28959 @node Summary of Procedures for Elaboration Control
28960 @section Summary of Procedures for Elaboration Control
28961 @cindex Elaboration control
28964 First, compile your program with the default options, using none of
28965 the special elaboration control switches. If the binder successfully
28966 binds your program, then you can be confident that, apart from issues
28967 raised by the use of access-to-subprogram types and dynamic dispatching,
28968 the program is free of elaboration errors. If it is important that the
28969 program be portable, then use the
28971 switch to generate warnings about missing @code{Elaborate} or
28972 @code{Elaborate_All} pragmas, and supply the missing pragmas.
28974 If the program fails to bind using the default static elaboration
28975 handling, then you can fix the program to eliminate the binder
28976 message, or recompile the entire program with the
28977 @option{-gnatE} switch to generate dynamic elaboration checks,
28978 and, if you are sure there really are no elaboration problems,
28979 use a global pragma @code{Suppress (Elaboration_Check)}.
28981 @node Other Elaboration Order Considerations
28982 @section Other Elaboration Order Considerations
28984 This section has been entirely concerned with the issue of finding a valid
28985 elaboration order, as defined by the Ada Reference Manual. In a case
28986 where several elaboration orders are valid, the task is to find one
28987 of the possible valid elaboration orders (and the static model in GNAT
28988 will ensure that this is achieved).
28990 The purpose of the elaboration rules in the Ada Reference Manual is to
28991 make sure that no entity is accessed before it has been elaborated. For
28992 a subprogram, this means that the spec and body must have been elaborated
28993 before the subprogram is called. For an object, this means that the object
28994 must have been elaborated before its value is read or written. A violation
28995 of either of these two requirements is an access before elaboration order,
28996 and this section has been all about avoiding such errors.
28998 In the case where more than one order of elaboration is possible, in the
28999 sense that access before elaboration errors are avoided, then any one of
29000 the orders is ``correct'' in the sense that it meets the requirements of
29001 the Ada Reference Manual, and no such error occurs.
29003 However, it may be the case for a given program, that there are
29004 constraints on the order of elaboration that come not from consideration
29005 of avoiding elaboration errors, but rather from extra-lingual logic
29006 requirements. Consider this example:
29008 @smallexample @c ada
29009 with Init_Constants;
29010 package Constants is
29015 package Init_Constants is
29016 procedure P; -- require a body
29017 end Init_Constants;
29020 package body Init_Constants is
29021 procedure P is begin null; end;
29025 end Init_Constants;
29029 Z : Integer := Constants.X + Constants.Y;
29033 with Text_IO; use Text_IO;
29036 Put_Line (Calc.Z'Img);
29041 In this example, there is more than one valid order of elaboration. For
29042 example both the following are correct orders:
29045 Init_Constants spec
29048 Init_Constants body
29053 Init_Constants spec
29054 Init_Constants body
29061 There is no language rule to prefer one or the other, both are correct
29062 from an order of elaboration point of view. But the programmatic effects
29063 of the two orders are very different. In the first, the elaboration routine
29064 of @code{Calc} initializes @code{Z} to zero, and then the main program
29065 runs with this value of zero. But in the second order, the elaboration
29066 routine of @code{Calc} runs after the body of Init_Constants has set
29067 @code{X} and @code{Y} and thus @code{Z} is set to 7 before @code{Main}
29070 One could perhaps by applying pretty clever non-artificial intelligence
29071 to the situation guess that it is more likely that the second order of
29072 elaboration is the one desired, but there is no formal linguistic reason
29073 to prefer one over the other. In fact in this particular case, GNAT will
29074 prefer the second order, because of the rule that bodies are elaborated
29075 as soon as possible, but it's just luck that this is what was wanted
29076 (if indeed the second order was preferred).
29078 If the program cares about the order of elaboration routines in a case like
29079 this, it is important to specify the order required. In this particular
29080 case, that could have been achieved by adding to the spec of Calc:
29082 @smallexample @c ada
29083 pragma Elaborate_All (Constants);
29087 which requires that the body (if any) and spec of @code{Constants},
29088 as well as the body and spec of any unit @code{with}'ed by
29089 @code{Constants} be elaborated before @code{Calc} is elaborated.
29091 Clearly no automatic method can always guess which alternative you require,
29092 and if you are working with legacy code that had constraints of this kind
29093 which were not properly specified by adding @code{Elaborate} or
29094 @code{Elaborate_All} pragmas, then indeed it is possible that two different
29095 compilers can choose different orders.
29097 However, GNAT does attempt to diagnose the common situation where there
29098 are uninitialized variables in the visible part of a package spec, and the
29099 corresponding package body has an elaboration block that directly or
29100 indirectly initialized one or more of these variables. This is the situation
29101 in which a pragma Elaborate_Body is usually desirable, and GNAT will generate
29102 a warning that suggests this addition if it detects this situation.
29104 The @code{gnatbind}
29105 @option{^-p^/PESSIMISTIC_ELABORATION^} switch may be useful in smoking
29106 out problems. This switch causes bodies to be elaborated as late as possible
29107 instead of as early as possible. In the example above, it would have forced
29108 the choice of the first elaboration order. If you get different results
29109 when using this switch, and particularly if one set of results is right,
29110 and one is wrong as far as you are concerned, it shows that you have some
29111 missing @code{Elaborate} pragmas. For the example above, we have the
29115 gnatmake -f -q main
29118 gnatmake -f -q main -bargs -p
29124 It is of course quite unlikely that both these results are correct, so
29125 it is up to you in a case like this to investigate the source of the
29126 difference, by looking at the two elaboration orders that are chosen,
29127 and figuring out which is correct, and then adding the necessary
29128 @code{Elaborate} or @code{Elaborate_All} pragmas to ensure the desired order.
29132 @c *******************************
29133 @node Conditional Compilation
29134 @appendix Conditional Compilation
29135 @c *******************************
29136 @cindex Conditional compilation
29139 It is often necessary to arrange for a single source program
29140 to serve multiple purposes, where it is compiled in different
29141 ways to achieve these different goals. Some examples of the
29142 need for this feature are
29145 @item Adapting a program to a different hardware environment
29146 @item Adapting a program to a different target architecture
29147 @item Turning debugging features on and off
29148 @item Arranging for a program to compile with different compilers
29152 In C, or C++, the typical approach would be to use the preprocessor
29153 that is defined as part of the language. The Ada language does not
29154 contain such a feature. This is not an oversight, but rather a very
29155 deliberate design decision, based on the experience that overuse of
29156 the preprocessing features in C and C++ can result in programs that
29157 are extremely difficult to maintain. For example, if we have ten
29158 switches that can be on or off, this means that there are a thousand
29159 separate programs, any one of which might not even be syntactically
29160 correct, and even if syntactically correct, the resulting program
29161 might not work correctly. Testing all combinations can quickly become
29164 Nevertheless, the need to tailor programs certainly exists, and in
29165 this Appendix we will discuss how this can
29166 be achieved using Ada in general, and GNAT in particular.
29169 * Use of Boolean Constants::
29170 * Debugging - A Special Case::
29171 * Conditionalizing Declarations::
29172 * Use of Alternative Implementations::
29176 @node Use of Boolean Constants
29177 @section Use of Boolean Constants
29180 In the case where the difference is simply which code
29181 sequence is executed, the cleanest solution is to use Boolean
29182 constants to control which code is executed.
29184 @smallexample @c ada
29186 FP_Initialize_Required : constant Boolean := True;
29188 if FP_Initialize_Required then
29195 Not only will the code inside the @code{if} statement not be executed if
29196 the constant Boolean is @code{False}, but it will also be completely
29197 deleted from the program.
29198 However, the code is only deleted after the @code{if} statement
29199 has been checked for syntactic and semantic correctness.
29200 (In contrast, with preprocessors the code is deleted before the
29201 compiler ever gets to see it, so it is not checked until the switch
29203 @cindex Preprocessors (contrasted with conditional compilation)
29205 Typically the Boolean constants will be in a separate package,
29208 @smallexample @c ada
29211 FP_Initialize_Required : constant Boolean := True;
29212 Reset_Available : constant Boolean := False;
29219 The @code{Config} package exists in multiple forms for the various targets,
29220 with an appropriate script selecting the version of @code{Config} needed.
29221 Then any other unit requiring conditional compilation can do a @code{with}
29222 of @code{Config} to make the constants visible.
29225 @node Debugging - A Special Case
29226 @section Debugging - A Special Case
29229 A common use of conditional code is to execute statements (for example
29230 dynamic checks, or output of intermediate results) under control of a
29231 debug switch, so that the debugging behavior can be turned on and off.
29232 This can be done using a Boolean constant to control whether the code
29235 @smallexample @c ada
29238 Put_Line ("got to the first stage!");
29246 @smallexample @c ada
29248 if Debugging and then Temperature > 999.0 then
29249 raise Temperature_Crazy;
29255 Since this is a common case, there are special features to deal with
29256 this in a convenient manner. For the case of tests, Ada 2005 has added
29257 a pragma @code{Assert} that can be used for such tests. This pragma is modeled
29258 @cindex pragma @code{Assert}
29259 on the @code{Assert} pragma that has always been available in GNAT, so this
29260 feature may be used with GNAT even if you are not using Ada 2005 features.
29261 The use of pragma @code{Assert} is described in
29262 @ref{Pragma Assert,,, gnat_rm, GNAT Reference Manual}, but as an
29263 example, the last test could be written:
29265 @smallexample @c ada
29266 pragma Assert (Temperature <= 999.0, "Temperature Crazy");
29272 @smallexample @c ada
29273 pragma Assert (Temperature <= 999.0);
29277 In both cases, if assertions are active and the temperature is excessive,
29278 the exception @code{Assert_Failure} will be raised, with the given string in
29279 the first case or a string indicating the location of the pragma in the second
29280 case used as the exception message.
29282 You can turn assertions on and off by using the @code{Assertion_Policy}
29284 @cindex pragma @code{Assertion_Policy}
29285 This is an Ada 2005 pragma which is implemented in all modes by
29286 GNAT, but only in the latest versions of GNAT which include Ada 2005
29287 capability. Alternatively, you can use the @option{-gnata} switch
29288 @cindex @option{-gnata} switch
29289 to enable assertions from the command line (this is recognized by all versions
29292 For the example above with the @code{Put_Line}, the GNAT-specific pragma
29293 @code{Debug} can be used:
29294 @cindex pragma @code{Debug}
29296 @smallexample @c ada
29297 pragma Debug (Put_Line ("got to the first stage!"));
29301 If debug pragmas are enabled, the argument, which must be of the form of
29302 a procedure call, is executed (in this case, @code{Put_Line} will be called).
29303 Only one call can be present, but of course a special debugging procedure
29304 containing any code you like can be included in the program and then
29305 called in a pragma @code{Debug} argument as needed.
29307 One advantage of pragma @code{Debug} over the @code{if Debugging then}
29308 construct is that pragma @code{Debug} can appear in declarative contexts,
29309 such as at the very beginning of a procedure, before local declarations have
29312 Debug pragmas are enabled using either the @option{-gnata} switch that also
29313 controls assertions, or with a separate Debug_Policy pragma.
29314 @cindex pragma @code{Debug_Policy}
29315 The latter pragma is new in the Ada 2005 versions of GNAT (but it can be used
29316 in Ada 95 and Ada 83 programs as well), and is analogous to
29317 pragma @code{Assertion_Policy} to control assertions.
29319 @code{Assertion_Policy} and @code{Debug_Policy} are configuration pragmas,
29320 and thus they can appear in @file{gnat.adc} if you are not using a
29321 project file, or in the file designated to contain configuration pragmas
29323 They then apply to all subsequent compilations. In practice the use of
29324 the @option{-gnata} switch is often the most convenient method of controlling
29325 the status of these pragmas.
29327 Note that a pragma is not a statement, so in contexts where a statement
29328 sequence is required, you can't just write a pragma on its own. You have
29329 to add a @code{null} statement.
29331 @smallexample @c ada
29334 @dots{} -- some statements
29336 pragma Assert (Num_Cases < 10);
29343 @node Conditionalizing Declarations
29344 @section Conditionalizing Declarations
29347 In some cases, it may be necessary to conditionalize declarations to meet
29348 different requirements. For example we might want a bit string whose length
29349 is set to meet some hardware message requirement.
29351 In some cases, it may be possible to do this using declare blocks controlled
29352 by conditional constants:
29354 @smallexample @c ada
29356 if Small_Machine then
29358 X : Bit_String (1 .. 10);
29364 X : Large_Bit_String (1 .. 1000);
29373 Note that in this approach, both declarations are analyzed by the
29374 compiler so this can only be used where both declarations are legal,
29375 even though one of them will not be used.
29377 Another approach is to define integer constants, e.g.@: @code{Bits_Per_Word}, or
29378 Boolean constants, e.g.@: @code{Little_Endian}, and then write declarations
29379 that are parameterized by these constants. For example
29381 @smallexample @c ada
29384 Field1 at 0 range Boolean'Pos (Little_Endian) * 10 .. Bits_Per_Word;
29390 If @code{Bits_Per_Word} is set to 32, this generates either
29392 @smallexample @c ada
29395 Field1 at 0 range 0 .. 32;
29401 for the big endian case, or
29403 @smallexample @c ada
29406 Field1 at 0 range 10 .. 32;
29412 for the little endian case. Since a powerful subset of Ada expression
29413 notation is usable for creating static constants, clever use of this
29414 feature can often solve quite difficult problems in conditionalizing
29415 compilation (note incidentally that in Ada 95, the little endian
29416 constant was introduced as @code{System.Default_Bit_Order}, so you do not
29417 need to define this one yourself).
29420 @node Use of Alternative Implementations
29421 @section Use of Alternative Implementations
29424 In some cases, none of the approaches described above are adequate. This
29425 can occur for example if the set of declarations required is radically
29426 different for two different configurations.
29428 In this situation, the official Ada way of dealing with conditionalizing
29429 such code is to write separate units for the different cases. As long as
29430 this does not result in excessive duplication of code, this can be done
29431 without creating maintenance problems. The approach is to share common
29432 code as far as possible, and then isolate the code and declarations
29433 that are different. Subunits are often a convenient method for breaking
29434 out a piece of a unit that is to be conditionalized, with separate files
29435 for different versions of the subunit for different targets, where the
29436 build script selects the right one to give to the compiler.
29437 @cindex Subunits (and conditional compilation)
29439 As an example, consider a situation where a new feature in Ada 2005
29440 allows something to be done in a really nice way. But your code must be able
29441 to compile with an Ada 95 compiler. Conceptually you want to say:
29443 @smallexample @c ada
29446 @dots{} neat Ada 2005 code
29448 @dots{} not quite as neat Ada 95 code
29454 where @code{Ada_2005} is a Boolean constant.
29456 But this won't work when @code{Ada_2005} is set to @code{False},
29457 since the @code{then} clause will be illegal for an Ada 95 compiler.
29458 (Recall that although such unreachable code would eventually be deleted
29459 by the compiler, it still needs to be legal. If it uses features
29460 introduced in Ada 2005, it will be illegal in Ada 95.)
29462 So instead we write
29464 @smallexample @c ada
29465 procedure Insert is separate;
29469 Then we have two files for the subunit @code{Insert}, with the two sets of
29471 If the package containing this is called @code{File_Queries}, then we might
29475 @item @file{file_queries-insert-2005.adb}
29476 @item @file{file_queries-insert-95.adb}
29480 and the build script renames the appropriate file to
29483 file_queries-insert.adb
29487 and then carries out the compilation.
29489 This can also be done with project files' naming schemes. For example:
29491 @smallexample @c project
29492 For Body ("File_Queries.Insert") use "file_queries-insert-2005.ada";
29496 Note also that with project files it is desirable to use a different extension
29497 than @file{ads} / @file{adb} for alternative versions. Otherwise a naming
29498 conflict may arise through another commonly used feature: to declare as part
29499 of the project a set of directories containing all the sources obeying the
29500 default naming scheme.
29502 The use of alternative units is certainly feasible in all situations,
29503 and for example the Ada part of the GNAT run-time is conditionalized
29504 based on the target architecture using this approach. As a specific example,
29505 consider the implementation of the AST feature in VMS. There is one
29513 which is the same for all architectures, and three bodies:
29517 used for all non-VMS operating systems
29518 @item s-asthan-vms-alpha.adb
29519 used for VMS on the Alpha
29520 @item s-asthan-vms-ia64.adb
29521 used for VMS on the ia64
29525 The dummy version @file{s-asthan.adb} simply raises exceptions noting that
29526 this operating system feature is not available, and the two remaining
29527 versions interface with the corresponding versions of VMS to provide
29528 VMS-compatible AST handling. The GNAT build script knows the architecture
29529 and operating system, and automatically selects the right version,
29530 renaming it if necessary to @file{s-asthan.adb} before the run-time build.
29532 Another style for arranging alternative implementations is through Ada's
29533 access-to-subprogram facility.
29534 In case some functionality is to be conditionally included,
29535 you can declare an access-to-procedure variable @code{Ref} that is initialized
29536 to designate a ``do nothing'' procedure, and then invoke @code{Ref.all}
29538 In some library package, set @code{Ref} to @code{Proc'Access} for some
29539 procedure @code{Proc} that performs the relevant processing.
29540 The initialization only occurs if the library package is included in the
29542 The same idea can also be implemented using tagged types and dispatching
29546 @node Preprocessing
29547 @section Preprocessing
29548 @cindex Preprocessing
29551 Although it is quite possible to conditionalize code without the use of
29552 C-style preprocessing, as described earlier in this section, it is
29553 nevertheless convenient in some cases to use the C approach. Moreover,
29554 older Ada compilers have often provided some preprocessing capability,
29555 so legacy code may depend on this approach, even though it is not
29558 To accommodate such use, GNAT provides a preprocessor (modeled to a large
29559 extent on the various preprocessors that have been used
29560 with legacy code on other compilers, to enable easier transition).
29562 The preprocessor may be used in two separate modes. It can be used quite
29563 separately from the compiler, to generate a separate output source file
29564 that is then fed to the compiler as a separate step. This is the
29565 @code{gnatprep} utility, whose use is fully described in
29566 @ref{Preprocessing Using gnatprep}.
29567 @cindex @code{gnatprep}
29569 The preprocessing language allows such constructs as
29573 #if DEBUG or PRIORITY > 4 then
29574 bunch of declarations
29576 completely different bunch of declarations
29582 The values of the symbols @code{DEBUG} and @code{PRIORITY} can be
29583 defined either on the command line or in a separate file.
29585 The other way of running the preprocessor is even closer to the C style and
29586 often more convenient. In this approach the preprocessing is integrated into
29587 the compilation process. The compiler is fed the preprocessor input which
29588 includes @code{#if} lines etc, and then the compiler carries out the
29589 preprocessing internally and processes the resulting output.
29590 For more details on this approach, see @ref{Integrated Preprocessing}.
29593 @c *******************************
29594 @node Inline Assembler
29595 @appendix Inline Assembler
29596 @c *******************************
29599 If you need to write low-level software that interacts directly
29600 with the hardware, Ada provides two ways to incorporate assembly
29601 language code into your program. First, you can import and invoke
29602 external routines written in assembly language, an Ada feature fully
29603 supported by GNAT@. However, for small sections of code it may be simpler
29604 or more efficient to include assembly language statements directly
29605 in your Ada source program, using the facilities of the implementation-defined
29606 package @code{System.Machine_Code}, which incorporates the gcc
29607 Inline Assembler. The Inline Assembler approach offers a number of advantages,
29608 including the following:
29611 @item No need to use non-Ada tools
29612 @item Consistent interface over different targets
29613 @item Automatic usage of the proper calling conventions
29614 @item Access to Ada constants and variables
29615 @item Definition of intrinsic routines
29616 @item Possibility of inlining a subprogram comprising assembler code
29617 @item Code optimizer can take Inline Assembler code into account
29620 This chapter presents a series of examples to show you how to use
29621 the Inline Assembler. Although it focuses on the Intel x86,
29622 the general approach applies also to other processors.
29623 It is assumed that you are familiar with Ada
29624 and with assembly language programming.
29627 * Basic Assembler Syntax::
29628 * A Simple Example of Inline Assembler::
29629 * Output Variables in Inline Assembler::
29630 * Input Variables in Inline Assembler::
29631 * Inlining Inline Assembler Code::
29632 * Other Asm Functionality::
29635 @c ---------------------------------------------------------------------------
29636 @node Basic Assembler Syntax
29637 @section Basic Assembler Syntax
29640 The assembler used by GNAT and gcc is based not on the Intel assembly
29641 language, but rather on a language that descends from the AT&T Unix
29642 assembler @emph{as} (and which is often referred to as ``AT&T syntax'').
29643 The following table summarizes the main features of @emph{as} syntax
29644 and points out the differences from the Intel conventions.
29645 See the gcc @emph{as} and @emph{gas} (an @emph{as} macro
29646 pre-processor) documentation for further information.
29649 @item Register names
29650 gcc / @emph{as}: Prefix with ``%''; for example @code{%eax}
29652 Intel: No extra punctuation; for example @code{eax}
29654 @item Immediate operand
29655 gcc / @emph{as}: Prefix with ``$''; for example @code{$4}
29657 Intel: No extra punctuation; for example @code{4}
29660 gcc / @emph{as}: Prefix with ``$''; for example @code{$loc}
29662 Intel: No extra punctuation; for example @code{loc}
29664 @item Memory contents
29665 gcc / @emph{as}: No extra punctuation; for example @code{loc}
29667 Intel: Square brackets; for example @code{[loc]}
29669 @item Register contents
29670 gcc / @emph{as}: Parentheses; for example @code{(%eax)}
29672 Intel: Square brackets; for example @code{[eax]}
29674 @item Hexadecimal numbers
29675 gcc / @emph{as}: Leading ``0x'' (C language syntax); for example @code{0xA0}
29677 Intel: Trailing ``h''; for example @code{A0h}
29680 gcc / @emph{as}: Explicit in op code; for example @code{movw} to move
29683 Intel: Implicit, deduced by assembler; for example @code{mov}
29685 @item Instruction repetition
29686 gcc / @emph{as}: Split into two lines; for example
29692 Intel: Keep on one line; for example @code{rep stosl}
29694 @item Order of operands
29695 gcc / @emph{as}: Source first; for example @code{movw $4, %eax}
29697 Intel: Destination first; for example @code{mov eax, 4}
29700 @c ---------------------------------------------------------------------------
29701 @node A Simple Example of Inline Assembler
29702 @section A Simple Example of Inline Assembler
29705 The following example will generate a single assembly language statement,
29706 @code{nop}, which does nothing. Despite its lack of run-time effect,
29707 the example will be useful in illustrating the basics of
29708 the Inline Assembler facility.
29710 @smallexample @c ada
29712 with System.Machine_Code; use System.Machine_Code;
29713 procedure Nothing is
29720 @code{Asm} is a procedure declared in package @code{System.Machine_Code};
29721 here it takes one parameter, a @emph{template string} that must be a static
29722 expression and that will form the generated instruction.
29723 @code{Asm} may be regarded as a compile-time procedure that parses
29724 the template string and additional parameters (none here),
29725 from which it generates a sequence of assembly language instructions.
29727 The examples in this chapter will illustrate several of the forms
29728 for invoking @code{Asm}; a complete specification of the syntax
29729 is found in @ref{Machine Code Insertions,,, gnat_rm, GNAT Reference
29732 Under the standard GNAT conventions, the @code{Nothing} procedure
29733 should be in a file named @file{nothing.adb}.
29734 You can build the executable in the usual way:
29738 However, the interesting aspect of this example is not its run-time behavior
29739 but rather the generated assembly code.
29740 To see this output, invoke the compiler as follows:
29742 gcc -c -S -fomit-frame-pointer -gnatp @file{nothing.adb}
29744 where the options are:
29748 compile only (no bind or link)
29750 generate assembler listing
29751 @item -fomit-frame-pointer
29752 do not set up separate stack frames
29754 do not add runtime checks
29757 This gives a human-readable assembler version of the code. The resulting
29758 file will have the same name as the Ada source file, but with a @code{.s}
29759 extension. In our example, the file @file{nothing.s} has the following
29764 .file "nothing.adb"
29766 ___gnu_compiled_ada:
29769 .globl __ada_nothing
29781 The assembly code you included is clearly indicated by
29782 the compiler, between the @code{#APP} and @code{#NO_APP}
29783 delimiters. The character before the 'APP' and 'NOAPP'
29784 can differ on different targets. For example, GNU/Linux uses '#APP' while
29785 on NT you will see '/APP'.
29787 If you make a mistake in your assembler code (such as using the
29788 wrong size modifier, or using a wrong operand for the instruction) GNAT
29789 will report this error in a temporary file, which will be deleted when
29790 the compilation is finished. Generating an assembler file will help
29791 in such cases, since you can assemble this file separately using the
29792 @emph{as} assembler that comes with gcc.
29794 Assembling the file using the command
29797 as @file{nothing.s}
29800 will give you error messages whose lines correspond to the assembler
29801 input file, so you can easily find and correct any mistakes you made.
29802 If there are no errors, @emph{as} will generate an object file
29803 @file{nothing.out}.
29805 @c ---------------------------------------------------------------------------
29806 @node Output Variables in Inline Assembler
29807 @section Output Variables in Inline Assembler
29810 The examples in this section, showing how to access the processor flags,
29811 illustrate how to specify the destination operands for assembly language
29814 @smallexample @c ada
29816 with Interfaces; use Interfaces;
29817 with Ada.Text_IO; use Ada.Text_IO;
29818 with System.Machine_Code; use System.Machine_Code;
29819 procedure Get_Flags is
29820 Flags : Unsigned_32;
29823 Asm ("pushfl" & LF & HT & -- push flags on stack
29824 "popl %%eax" & LF & HT & -- load eax with flags
29825 "movl %%eax, %0", -- store flags in variable
29826 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
29827 Put_Line ("Flags register:" & Flags'Img);
29832 In order to have a nicely aligned assembly listing, we have separated
29833 multiple assembler statements in the Asm template string with linefeed
29834 (ASCII.LF) and horizontal tab (ASCII.HT) characters.
29835 The resulting section of the assembly output file is:
29842 movl %eax, -40(%ebp)
29847 It would have been legal to write the Asm invocation as:
29850 Asm ("pushfl popl %%eax movl %%eax, %0")
29853 but in the generated assembler file, this would come out as:
29857 pushfl popl %eax movl %eax, -40(%ebp)
29861 which is not so convenient for the human reader.
29863 We use Ada comments
29864 at the end of each line to explain what the assembler instructions
29865 actually do. This is a useful convention.
29867 When writing Inline Assembler instructions, you need to precede each register
29868 and variable name with a percent sign. Since the assembler already requires
29869 a percent sign at the beginning of a register name, you need two consecutive
29870 percent signs for such names in the Asm template string, thus @code{%%eax}.
29871 In the generated assembly code, one of the percent signs will be stripped off.
29873 Names such as @code{%0}, @code{%1}, @code{%2}, etc., denote input or output
29874 variables: operands you later define using @code{Input} or @code{Output}
29875 parameters to @code{Asm}.
29876 An output variable is illustrated in
29877 the third statement in the Asm template string:
29881 The intent is to store the contents of the eax register in a variable that can
29882 be accessed in Ada. Simply writing @code{movl %%eax, Flags} would not
29883 necessarily work, since the compiler might optimize by using a register
29884 to hold Flags, and the expansion of the @code{movl} instruction would not be
29885 aware of this optimization. The solution is not to store the result directly
29886 but rather to advise the compiler to choose the correct operand form;
29887 that is the purpose of the @code{%0} output variable.
29889 Information about the output variable is supplied in the @code{Outputs}
29890 parameter to @code{Asm}:
29892 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
29895 The output is defined by the @code{Asm_Output} attribute of the target type;
29896 the general format is
29898 Type'Asm_Output (constraint_string, variable_name)
29901 The constraint string directs the compiler how
29902 to store/access the associated variable. In the example
29904 Unsigned_32'Asm_Output ("=m", Flags);
29906 the @code{"m"} (memory) constraint tells the compiler that the variable
29907 @code{Flags} should be stored in a memory variable, thus preventing
29908 the optimizer from keeping it in a register. In contrast,
29910 Unsigned_32'Asm_Output ("=r", Flags);
29912 uses the @code{"r"} (register) constraint, telling the compiler to
29913 store the variable in a register.
29915 If the constraint is preceded by the equal character (@strong{=}), it tells
29916 the compiler that the variable will be used to store data into it.
29918 In the @code{Get_Flags} example, we used the @code{"g"} (global) constraint,
29919 allowing the optimizer to choose whatever it deems best.
29921 There are a fairly large number of constraints, but the ones that are
29922 most useful (for the Intel x86 processor) are the following:
29928 global (i.e.@: can be stored anywhere)
29946 use one of eax, ebx, ecx or edx
29948 use one of eax, ebx, ecx, edx, esi or edi
29951 The full set of constraints is described in the gcc and @emph{as}
29952 documentation; note that it is possible to combine certain constraints
29953 in one constraint string.
29955 You specify the association of an output variable with an assembler operand
29956 through the @code{%}@emph{n} notation, where @emph{n} is a non-negative
29958 @smallexample @c ada
29960 Asm ("pushfl" & LF & HT & -- push flags on stack
29961 "popl %%eax" & LF & HT & -- load eax with flags
29962 "movl %%eax, %0", -- store flags in variable
29963 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
29967 @code{%0} will be replaced in the expanded code by the appropriate operand,
29969 the compiler decided for the @code{Flags} variable.
29971 In general, you may have any number of output variables:
29974 Count the operands starting at 0; thus @code{%0}, @code{%1}, etc.
29976 Specify the @code{Outputs} parameter as a parenthesized comma-separated list
29977 of @code{Asm_Output} attributes
29981 @smallexample @c ada
29983 Asm ("movl %%eax, %0" & LF & HT &
29984 "movl %%ebx, %1" & LF & HT &
29986 Outputs => (Unsigned_32'Asm_Output ("=g", Var_A), -- %0 = Var_A
29987 Unsigned_32'Asm_Output ("=g", Var_B), -- %1 = Var_B
29988 Unsigned_32'Asm_Output ("=g", Var_C))); -- %2 = Var_C
29992 where @code{Var_A}, @code{Var_B}, and @code{Var_C} are variables
29993 in the Ada program.
29995 As a variation on the @code{Get_Flags} example, we can use the constraints
29996 string to direct the compiler to store the eax register into the @code{Flags}
29997 variable, instead of including the store instruction explicitly in the
29998 @code{Asm} template string:
30000 @smallexample @c ada
30002 with Interfaces; use Interfaces;
30003 with Ada.Text_IO; use Ada.Text_IO;
30004 with System.Machine_Code; use System.Machine_Code;
30005 procedure Get_Flags_2 is
30006 Flags : Unsigned_32;
30009 Asm ("pushfl" & LF & HT & -- push flags on stack
30010 "popl %%eax", -- save flags in eax
30011 Outputs => Unsigned_32'Asm_Output ("=a", Flags));
30012 Put_Line ("Flags register:" & Flags'Img);
30018 The @code{"a"} constraint tells the compiler that the @code{Flags}
30019 variable will come from the eax register. Here is the resulting code:
30027 movl %eax,-40(%ebp)
30032 The compiler generated the store of eax into Flags after
30033 expanding the assembler code.
30035 Actually, there was no need to pop the flags into the eax register;
30036 more simply, we could just pop the flags directly into the program variable:
30038 @smallexample @c ada
30040 with Interfaces; use Interfaces;
30041 with Ada.Text_IO; use Ada.Text_IO;
30042 with System.Machine_Code; use System.Machine_Code;
30043 procedure Get_Flags_3 is
30044 Flags : Unsigned_32;
30047 Asm ("pushfl" & LF & HT & -- push flags on stack
30048 "pop %0", -- save flags in Flags
30049 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
30050 Put_Line ("Flags register:" & Flags'Img);
30055 @c ---------------------------------------------------------------------------
30056 @node Input Variables in Inline Assembler
30057 @section Input Variables in Inline Assembler
30060 The example in this section illustrates how to specify the source operands
30061 for assembly language statements.
30062 The program simply increments its input value by 1:
30064 @smallexample @c ada
30066 with Interfaces; use Interfaces;
30067 with Ada.Text_IO; use Ada.Text_IO;
30068 with System.Machine_Code; use System.Machine_Code;
30069 procedure Increment is
30071 function Incr (Value : Unsigned_32) return Unsigned_32 is
30072 Result : Unsigned_32;
30075 Inputs => Unsigned_32'Asm_Input ("a", Value),
30076 Outputs => Unsigned_32'Asm_Output ("=a", Result));
30080 Value : Unsigned_32;
30084 Put_Line ("Value before is" & Value'Img);
30085 Value := Incr (Value);
30086 Put_Line ("Value after is" & Value'Img);
30091 The @code{Outputs} parameter to @code{Asm} specifies
30092 that the result will be in the eax register and that it is to be stored
30093 in the @code{Result} variable.
30095 The @code{Inputs} parameter looks much like the @code{Outputs} parameter,
30096 but with an @code{Asm_Input} attribute.
30097 The @code{"="} constraint, indicating an output value, is not present.
30099 You can have multiple input variables, in the same way that you can have more
30100 than one output variable.
30102 The parameter count (%0, %1) etc, now starts at the first input
30103 statement, and continues with the output statements.
30104 When both parameters use the same variable, the
30105 compiler will treat them as the same %n operand, which is the case here.
30107 Just as the @code{Outputs} parameter causes the register to be stored into the
30108 target variable after execution of the assembler statements, so does the
30109 @code{Inputs} parameter cause its variable to be loaded into the register
30110 before execution of the assembler statements.
30112 Thus the effect of the @code{Asm} invocation is:
30114 @item load the 32-bit value of @code{Value} into eax
30115 @item execute the @code{incl %eax} instruction
30116 @item store the contents of eax into the @code{Result} variable
30119 The resulting assembler file (with @option{-O2} optimization) contains:
30122 _increment__incr.1:
30135 @c ---------------------------------------------------------------------------
30136 @node Inlining Inline Assembler Code
30137 @section Inlining Inline Assembler Code
30140 For a short subprogram such as the @code{Incr} function in the previous
30141 section, the overhead of the call and return (creating / deleting the stack
30142 frame) can be significant, compared to the amount of code in the subprogram
30143 body. A solution is to apply Ada's @code{Inline} pragma to the subprogram,
30144 which directs the compiler to expand invocations of the subprogram at the
30145 point(s) of call, instead of setting up a stack frame for out-of-line calls.
30146 Here is the resulting program:
30148 @smallexample @c ada
30150 with Interfaces; use Interfaces;
30151 with Ada.Text_IO; use Ada.Text_IO;
30152 with System.Machine_Code; use System.Machine_Code;
30153 procedure Increment_2 is
30155 function Incr (Value : Unsigned_32) return Unsigned_32 is
30156 Result : Unsigned_32;
30159 Inputs => Unsigned_32'Asm_Input ("a", Value),
30160 Outputs => Unsigned_32'Asm_Output ("=a", Result));
30163 pragma Inline (Increment);
30165 Value : Unsigned_32;
30169 Put_Line ("Value before is" & Value'Img);
30170 Value := Increment (Value);
30171 Put_Line ("Value after is" & Value'Img);
30176 Compile the program with both optimization (@option{-O2}) and inlining
30177 (@option{-gnatn}) enabled.
30179 The @code{Incr} function is still compiled as usual, but at the
30180 point in @code{Increment} where our function used to be called:
30185 call _increment__incr.1
30190 the code for the function body directly appears:
30203 thus saving the overhead of stack frame setup and an out-of-line call.
30205 @c ---------------------------------------------------------------------------
30206 @node Other Asm Functionality
30207 @section Other @code{Asm} Functionality
30210 This section describes two important parameters to the @code{Asm}
30211 procedure: @code{Clobber}, which identifies register usage;
30212 and @code{Volatile}, which inhibits unwanted optimizations.
30215 * The Clobber Parameter::
30216 * The Volatile Parameter::
30219 @c ---------------------------------------------------------------------------
30220 @node The Clobber Parameter
30221 @subsection The @code{Clobber} Parameter
30224 One of the dangers of intermixing assembly language and a compiled language
30225 such as Ada is that the compiler needs to be aware of which registers are
30226 being used by the assembly code. In some cases, such as the earlier examples,
30227 the constraint string is sufficient to indicate register usage (e.g.,
30229 the eax register). But more generally, the compiler needs an explicit
30230 identification of the registers that are used by the Inline Assembly
30233 Using a register that the compiler doesn't know about
30234 could be a side effect of an instruction (like @code{mull}
30235 storing its result in both eax and edx).
30236 It can also arise from explicit register usage in your
30237 assembly code; for example:
30240 Asm ("movl %0, %%ebx" & LF & HT &
30242 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
30243 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out));
30247 where the compiler (since it does not analyze the @code{Asm} template string)
30248 does not know you are using the ebx register.
30250 In such cases you need to supply the @code{Clobber} parameter to @code{Asm},
30251 to identify the registers that will be used by your assembly code:
30255 Asm ("movl %0, %%ebx" & LF & HT &
30257 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
30258 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
30263 The Clobber parameter is a static string expression specifying the
30264 register(s) you are using. Note that register names are @emph{not} prefixed
30265 by a percent sign. Also, if more than one register is used then their names
30266 are separated by commas; e.g., @code{"eax, ebx"}
30268 The @code{Clobber} parameter has several additional uses:
30270 @item Use ``register'' name @code{cc} to indicate that flags might have changed
30271 @item Use ``register'' name @code{memory} if you changed a memory location
30274 @c ---------------------------------------------------------------------------
30275 @node The Volatile Parameter
30276 @subsection The @code{Volatile} Parameter
30277 @cindex Volatile parameter
30280 Compiler optimizations in the presence of Inline Assembler may sometimes have
30281 unwanted effects. For example, when an @code{Asm} invocation with an input
30282 variable is inside a loop, the compiler might move the loading of the input
30283 variable outside the loop, regarding it as a one-time initialization.
30285 If this effect is not desired, you can disable such optimizations by setting
30286 the @code{Volatile} parameter to @code{True}; for example:
30288 @smallexample @c ada
30290 Asm ("movl %0, %%ebx" & LF & HT &
30292 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
30293 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
30299 By default, @code{Volatile} is set to @code{False} unless there is no
30300 @code{Outputs} parameter.
30302 Although setting @code{Volatile} to @code{True} prevents unwanted
30303 optimizations, it will also disable other optimizations that might be
30304 important for efficiency. In general, you should set @code{Volatile}
30305 to @code{True} only if the compiler's optimizations have created
30307 @c END OF INLINE ASSEMBLER CHAPTER
30308 @c ===============================
30310 @c ***********************************
30311 @c * Compatibility and Porting Guide *
30312 @c ***********************************
30313 @node Compatibility and Porting Guide
30314 @appendix Compatibility and Porting Guide
30317 This chapter describes the compatibility issues that may arise between
30318 GNAT and other Ada compilation systems (including those for Ada 83),
30319 and shows how GNAT can expedite porting
30320 applications developed in other Ada environments.
30323 * Compatibility with Ada 83::
30324 * Compatibility between Ada 95 and Ada 2005::
30325 * Implementation-dependent characteristics::
30326 * Compatibility with Other Ada Systems::
30327 * Representation Clauses::
30329 @c Brief section is only in non-VMS version
30330 @c Full chapter is in VMS version
30331 * Compatibility with HP Ada 83::
30334 * Transitioning to 64-Bit GNAT for OpenVMS::
30338 @node Compatibility with Ada 83
30339 @section Compatibility with Ada 83
30340 @cindex Compatibility (between Ada 83 and Ada 95 / Ada 2005)
30343 Ada 95 and Ada 2005 are highly upwards compatible with Ada 83. In
30344 particular, the design intention was that the difficulties associated
30345 with moving from Ada 83 to Ada 95 or Ada 2005 should be no greater than those
30346 that occur when moving from one Ada 83 system to another.
30348 However, there are a number of points at which there are minor
30349 incompatibilities. The @cite{Ada 95 Annotated Reference Manual} contains
30350 full details of these issues,
30351 and should be consulted for a complete treatment.
30353 following subsections treat the most likely issues to be encountered.
30356 * Legal Ada 83 programs that are illegal in Ada 95::
30357 * More deterministic semantics::
30358 * Changed semantics::
30359 * Other language compatibility issues::
30362 @node Legal Ada 83 programs that are illegal in Ada 95
30363 @subsection Legal Ada 83 programs that are illegal in Ada 95
30365 Some legal Ada 83 programs are illegal (i.e., they will fail to compile) in
30366 Ada 95 and thus also in Ada 2005:
30369 @item Character literals
30370 Some uses of character literals are ambiguous. Since Ada 95 has introduced
30371 @code{Wide_Character} as a new predefined character type, some uses of
30372 character literals that were legal in Ada 83 are illegal in Ada 95.
30374 @smallexample @c ada
30375 for Char in 'A' .. 'Z' loop @dots{} end loop;
30379 The problem is that @code{'A'} and @code{'Z'} could be from either
30380 @code{Character} or @code{Wide_Character}. The simplest correction
30381 is to make the type explicit; e.g.:
30382 @smallexample @c ada
30383 for Char in Character range 'A' .. 'Z' loop @dots{} end loop;
30386 @item New reserved words
30387 The identifiers @code{abstract}, @code{aliased}, @code{protected},
30388 @code{requeue}, @code{tagged}, and @code{until} are reserved in Ada 95.
30389 Existing Ada 83 code using any of these identifiers must be edited to
30390 use some alternative name.
30392 @item Freezing rules
30393 The rules in Ada 95 are slightly different with regard to the point at
30394 which entities are frozen, and representation pragmas and clauses are
30395 not permitted past the freeze point. This shows up most typically in
30396 the form of an error message complaining that a representation item
30397 appears too late, and the appropriate corrective action is to move
30398 the item nearer to the declaration of the entity to which it refers.
30400 A particular case is that representation pragmas
30403 extended HP Ada 83 compatibility pragmas such as @code{Export_Procedure})
30405 cannot be applied to a subprogram body. If necessary, a separate subprogram
30406 declaration must be introduced to which the pragma can be applied.
30408 @item Optional bodies for library packages
30409 In Ada 83, a package that did not require a package body was nevertheless
30410 allowed to have one. This lead to certain surprises in compiling large
30411 systems (situations in which the body could be unexpectedly ignored by the
30412 binder). In Ada 95, if a package does not require a body then it is not
30413 permitted to have a body. To fix this problem, simply remove a redundant
30414 body if it is empty, or, if it is non-empty, introduce a dummy declaration
30415 into the spec that makes the body required. One approach is to add a private
30416 part to the package declaration (if necessary), and define a parameterless
30417 procedure called @code{Requires_Body}, which must then be given a dummy
30418 procedure body in the package body, which then becomes required.
30419 Another approach (assuming that this does not introduce elaboration
30420 circularities) is to add an @code{Elaborate_Body} pragma to the package spec,
30421 since one effect of this pragma is to require the presence of a package body.
30423 @item @code{Numeric_Error} is now the same as @code{Constraint_Error}
30424 In Ada 95, the exception @code{Numeric_Error} is a renaming of
30425 @code{Constraint_Error}.
30426 This means that it is illegal to have separate exception handlers for
30427 the two exceptions. The fix is simply to remove the handler for the
30428 @code{Numeric_Error} case (since even in Ada 83, a compiler was free to raise
30429 @code{Constraint_Error} in place of @code{Numeric_Error} in all cases).
30431 @item Indefinite subtypes in generics
30432 In Ada 83, it was permissible to pass an indefinite type (e.g.@: @code{String})
30433 as the actual for a generic formal private type, but then the instantiation
30434 would be illegal if there were any instances of declarations of variables
30435 of this type in the generic body. In Ada 95, to avoid this clear violation
30436 of the methodological principle known as the ``contract model'',
30437 the generic declaration explicitly indicates whether
30438 or not such instantiations are permitted. If a generic formal parameter
30439 has explicit unknown discriminants, indicated by using @code{(<>)} after the
30440 type name, then it can be instantiated with indefinite types, but no
30441 stand-alone variables can be declared of this type. Any attempt to declare
30442 such a variable will result in an illegality at the time the generic is
30443 declared. If the @code{(<>)} notation is not used, then it is illegal
30444 to instantiate the generic with an indefinite type.
30445 This is the potential incompatibility issue when porting Ada 83 code to Ada 95.
30446 It will show up as a compile time error, and
30447 the fix is usually simply to add the @code{(<>)} to the generic declaration.
30450 @node More deterministic semantics
30451 @subsection More deterministic semantics
30455 Conversions from real types to integer types round away from 0. In Ada 83
30456 the conversion Integer(2.5) could deliver either 2 or 3 as its value. This
30457 implementation freedom was intended to support unbiased rounding in
30458 statistical applications, but in practice it interfered with portability.
30459 In Ada 95 the conversion semantics are unambiguous, and rounding away from 0
30460 is required. Numeric code may be affected by this change in semantics.
30461 Note, though, that this issue is no worse than already existed in Ada 83
30462 when porting code from one vendor to another.
30465 The Real-Time Annex introduces a set of policies that define the behavior of
30466 features that were implementation dependent in Ada 83, such as the order in
30467 which open select branches are executed.
30470 @node Changed semantics
30471 @subsection Changed semantics
30474 The worst kind of incompatibility is one where a program that is legal in
30475 Ada 83 is also legal in Ada 95 but can have an effect in Ada 95 that was not
30476 possible in Ada 83. Fortunately this is extremely rare, but the one
30477 situation that you should be alert to is the change in the predefined type
30478 @code{Character} from 7-bit ASCII to 8-bit Latin-1.
30481 @item Range of type @code{Character}
30482 The range of @code{Standard.Character} is now the full 256 characters
30483 of Latin-1, whereas in most Ada 83 implementations it was restricted
30484 to 128 characters. Although some of the effects of
30485 this change will be manifest in compile-time rejection of legal
30486 Ada 83 programs it is possible for a working Ada 83 program to have
30487 a different effect in Ada 95, one that was not permitted in Ada 83.
30488 As an example, the expression
30489 @code{Character'Pos(Character'Last)} returned @code{127} in Ada 83 and now
30490 delivers @code{255} as its value.
30491 In general, you should look at the logic of any
30492 character-processing Ada 83 program and see whether it needs to be adapted
30493 to work correctly with Latin-1. Note that the predefined Ada 95 API has a
30494 character handling package that may be relevant if code needs to be adapted
30495 to account for the additional Latin-1 elements.
30496 The desirable fix is to
30497 modify the program to accommodate the full character set, but in some cases
30498 it may be convenient to define a subtype or derived type of Character that
30499 covers only the restricted range.
30503 @node Other language compatibility issues
30504 @subsection Other language compatibility issues
30507 @item @option{-gnat83} switch
30508 All implementations of GNAT provide a switch that causes GNAT to operate
30509 in Ada 83 mode. In this mode, some but not all compatibility problems
30510 of the type described above are handled automatically. For example, the
30511 new reserved words introduced in Ada 95 and Ada 2005 are treated simply
30512 as identifiers as in Ada 83.
30514 in practice, it is usually advisable to make the necessary modifications
30515 to the program to remove the need for using this switch.
30516 See @ref{Compiling Different Versions of Ada}.
30518 @item Support for removed Ada 83 pragmas and attributes
30519 A number of pragmas and attributes from Ada 83 were removed from Ada 95,
30520 generally because they were replaced by other mechanisms. Ada 95 and Ada 2005
30521 compilers are allowed, but not required, to implement these missing
30522 elements. In contrast with some other compilers, GNAT implements all
30523 such pragmas and attributes, eliminating this compatibility concern. These
30524 include @code{pragma Interface} and the floating point type attributes
30525 (@code{Emax}, @code{Mantissa}, etc.), among other items.
30529 @node Compatibility between Ada 95 and Ada 2005
30530 @section Compatibility between Ada 95 and Ada 2005
30531 @cindex Compatibility between Ada 95 and Ada 2005
30534 Although Ada 2005 was designed to be upwards compatible with Ada 95, there are
30535 a number of incompatibilities. Several are enumerated below;
30536 for a complete description please see the
30537 Annotated Ada 2005 Reference Manual, or section 9.1.1 in
30538 @cite{Rationale for Ada 2005}.
30541 @item New reserved words.
30542 The words @code{interface}, @code{overriding} and @code{synchronized} are
30543 reserved in Ada 2005.
30544 A pre-Ada 2005 program that uses any of these as an identifier will be
30547 @item New declarations in predefined packages.
30548 A number of packages in the predefined environment contain new declarations:
30549 @code{Ada.Exceptions}, @code{Ada.Real_Time}, @code{Ada.Strings},
30550 @code{Ada.Strings.Fixed}, @code{Ada.Strings.Bounded},
30551 @code{Ada.Strings.Unbounded}, @code{Ada.Strings.Wide_Fixed},
30552 @code{Ada.Strings.Wide_Bounded}, @code{Ada.Strings.Wide_Unbounded},
30553 @code{Ada.Tags}, @code{Ada.Text_IO}, and @code{Interfaces.C}.
30554 If an Ada 95 program does a @code{with} and @code{use} of any of these
30555 packages, the new declarations may cause name clashes.
30557 @item Access parameters.
30558 A nondispatching subprogram with an access parameter cannot be renamed
30559 as a dispatching operation. This was permitted in Ada 95.
30561 @item Access types, discriminants, and constraints.
30562 Rule changes in this area have led to some incompatibilities; for example,
30563 constrained subtypes of some access types are not permitted in Ada 2005.
30565 @item Aggregates for limited types.
30566 The allowance of aggregates for limited types in Ada 2005 raises the
30567 possibility of ambiguities in legal Ada 95 programs, since additional types
30568 now need to be considered in expression resolution.
30570 @item Fixed-point multiplication and division.
30571 Certain expressions involving ``*'' or ``/'' for a fixed-point type, which
30572 were legal in Ada 95 and invoked the predefined versions of these operations,
30574 The ambiguity may be resolved either by applying a type conversion to the
30575 expression, or by explicitly invoking the operation from package
30578 @item Return-by-reference types.
30579 The Ada 95 return-by-reference mechanism has been removed. Instead, the user
30580 can declare a function returning a value from an anonymous access type.
30584 @node Implementation-dependent characteristics
30585 @section Implementation-dependent characteristics
30587 Although the Ada language defines the semantics of each construct as
30588 precisely as practical, in some situations (for example for reasons of
30589 efficiency, or where the effect is heavily dependent on the host or target
30590 platform) the implementation is allowed some freedom. In porting Ada 83
30591 code to GNAT, you need to be aware of whether / how the existing code
30592 exercised such implementation dependencies. Such characteristics fall into
30593 several categories, and GNAT offers specific support in assisting the
30594 transition from certain Ada 83 compilers.
30597 * Implementation-defined pragmas::
30598 * Implementation-defined attributes::
30600 * Elaboration order::
30601 * Target-specific aspects::
30604 @node Implementation-defined pragmas
30605 @subsection Implementation-defined pragmas
30608 Ada compilers are allowed to supplement the language-defined pragmas, and
30609 these are a potential source of non-portability. All GNAT-defined pragmas
30610 are described in @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT
30611 Reference Manual}, and these include several that are specifically
30612 intended to correspond to other vendors' Ada 83 pragmas.
30613 For migrating from VADS, the pragma @code{Use_VADS_Size} may be useful.
30614 For compatibility with HP Ada 83, GNAT supplies the pragmas
30615 @code{Extend_System}, @code{Ident}, @code{Inline_Generic},
30616 @code{Interface_Name}, @code{Passive}, @code{Suppress_All},
30617 and @code{Volatile}.
30618 Other relevant pragmas include @code{External} and @code{Link_With}.
30619 Some vendor-specific
30620 Ada 83 pragmas (@code{Share_Generic}, @code{Subtitle}, and @code{Title}) are
30622 avoiding compiler rejection of units that contain such pragmas; they are not
30623 relevant in a GNAT context and hence are not otherwise implemented.
30625 @node Implementation-defined attributes
30626 @subsection Implementation-defined attributes
30628 Analogous to pragmas, the set of attributes may be extended by an
30629 implementation. All GNAT-defined attributes are described in
30630 @ref{Implementation Defined Attributes,,, gnat_rm, GNAT Reference
30631 Manual}, and these include several that are specifically intended
30632 to correspond to other vendors' Ada 83 attributes. For migrating from VADS,
30633 the attribute @code{VADS_Size} may be useful. For compatibility with HP
30634 Ada 83, GNAT supplies the attributes @code{Bit}, @code{Machine_Size} and
30638 @subsection Libraries
30640 Vendors may supply libraries to supplement the standard Ada API. If Ada 83
30641 code uses vendor-specific libraries then there are several ways to manage
30642 this in Ada 95 or Ada 2005:
30645 If the source code for the libraries (specs and bodies) are
30646 available, then the libraries can be migrated in the same way as the
30649 If the source code for the specs but not the bodies are
30650 available, then you can reimplement the bodies.
30652 Some features introduced by Ada 95 obviate the need for library support. For
30653 example most Ada 83 vendors supplied a package for unsigned integers. The
30654 Ada 95 modular type feature is the preferred way to handle this need, so
30655 instead of migrating or reimplementing the unsigned integer package it may
30656 be preferable to retrofit the application using modular types.
30659 @node Elaboration order
30660 @subsection Elaboration order
30662 The implementation can choose any elaboration order consistent with the unit
30663 dependency relationship. This freedom means that some orders can result in
30664 Program_Error being raised due to an ``Access Before Elaboration'': an attempt
30665 to invoke a subprogram its body has been elaborated, or to instantiate a
30666 generic before the generic body has been elaborated. By default GNAT
30667 attempts to choose a safe order (one that will not encounter access before
30668 elaboration problems) by implicitly inserting @code{Elaborate} or
30669 @code{Elaborate_All} pragmas where
30670 needed. However, this can lead to the creation of elaboration circularities
30671 and a resulting rejection of the program by gnatbind. This issue is
30672 thoroughly described in @ref{Elaboration Order Handling in GNAT}.
30673 In brief, there are several
30674 ways to deal with this situation:
30678 Modify the program to eliminate the circularities, e.g.@: by moving
30679 elaboration-time code into explicitly-invoked procedures
30681 Constrain the elaboration order by including explicit @code{Elaborate_Body} or
30682 @code{Elaborate} pragmas, and then inhibit the generation of implicit
30683 @code{Elaborate_All}
30684 pragmas either globally (as an effect of the @option{-gnatE} switch) or locally
30685 (by selectively suppressing elaboration checks via pragma
30686 @code{Suppress(Elaboration_Check)} when it is safe to do so).
30689 @node Target-specific aspects
30690 @subsection Target-specific aspects
30692 Low-level applications need to deal with machine addresses, data
30693 representations, interfacing with assembler code, and similar issues. If
30694 such an Ada 83 application is being ported to different target hardware (for
30695 example where the byte endianness has changed) then you will need to
30696 carefully examine the program logic; the porting effort will heavily depend
30697 on the robustness of the original design. Moreover, Ada 95 (and thus
30698 Ada 2005) are sometimes
30699 incompatible with typical Ada 83 compiler practices regarding implicit
30700 packing, the meaning of the Size attribute, and the size of access values.
30701 GNAT's approach to these issues is described in @ref{Representation Clauses}.
30703 @node Compatibility with Other Ada Systems
30704 @section Compatibility with Other Ada Systems
30707 If programs avoid the use of implementation dependent and
30708 implementation defined features, as documented in the @cite{Ada
30709 Reference Manual}, there should be a high degree of portability between
30710 GNAT and other Ada systems. The following are specific items which
30711 have proved troublesome in moving Ada 95 programs from GNAT to other Ada 95
30712 compilers, but do not affect porting code to GNAT@.
30713 (As of @value{NOW}, GNAT is the only compiler available for Ada 2005;
30714 the following issues may or may not arise for Ada 2005 programs
30715 when other compilers appear.)
30718 @item Ada 83 Pragmas and Attributes
30719 Ada 95 compilers are allowed, but not required, to implement the missing
30720 Ada 83 pragmas and attributes that are no longer defined in Ada 95.
30721 GNAT implements all such pragmas and attributes, eliminating this as
30722 a compatibility concern, but some other Ada 95 compilers reject these
30723 pragmas and attributes.
30725 @item Specialized Needs Annexes
30726 GNAT implements the full set of special needs annexes. At the
30727 current time, it is the only Ada 95 compiler to do so. This means that
30728 programs making use of these features may not be portable to other Ada
30729 95 compilation systems.
30731 @item Representation Clauses
30732 Some other Ada 95 compilers implement only the minimal set of
30733 representation clauses required by the Ada 95 reference manual. GNAT goes
30734 far beyond this minimal set, as described in the next section.
30737 @node Representation Clauses
30738 @section Representation Clauses
30741 The Ada 83 reference manual was quite vague in describing both the minimal
30742 required implementation of representation clauses, and also their precise
30743 effects. Ada 95 (and thus also Ada 2005) are much more explicit, but the
30744 minimal set of capabilities required is still quite limited.
30746 GNAT implements the full required set of capabilities in
30747 Ada 95 and Ada 2005, but also goes much further, and in particular
30748 an effort has been made to be compatible with existing Ada 83 usage to the
30749 greatest extent possible.
30751 A few cases exist in which Ada 83 compiler behavior is incompatible with
30752 the requirements in Ada 95 (and thus also Ada 2005). These are instances of
30753 intentional or accidental dependence on specific implementation dependent
30754 characteristics of these Ada 83 compilers. The following is a list of
30755 the cases most likely to arise in existing Ada 83 code.
30758 @item Implicit Packing
30759 Some Ada 83 compilers allowed a Size specification to cause implicit
30760 packing of an array or record. This could cause expensive implicit
30761 conversions for change of representation in the presence of derived
30762 types, and the Ada design intends to avoid this possibility.
30763 Subsequent AI's were issued to make it clear that such implicit
30764 change of representation in response to a Size clause is inadvisable,
30765 and this recommendation is represented explicitly in the Ada 95 (and Ada 2005)
30766 Reference Manuals as implementation advice that is followed by GNAT@.
30767 The problem will show up as an error
30768 message rejecting the size clause. The fix is simply to provide
30769 the explicit pragma @code{Pack}, or for more fine tuned control, provide
30770 a Component_Size clause.
30772 @item Meaning of Size Attribute
30773 The Size attribute in Ada 95 (and Ada 2005) for discrete types is defined as
30774 the minimal number of bits required to hold values of the type. For example,
30775 on a 32-bit machine, the size of @code{Natural} will typically be 31 and not
30776 32 (since no sign bit is required). Some Ada 83 compilers gave 31, and
30777 some 32 in this situation. This problem will usually show up as a compile
30778 time error, but not always. It is a good idea to check all uses of the
30779 'Size attribute when porting Ada 83 code. The GNAT specific attribute
30780 Object_Size can provide a useful way of duplicating the behavior of
30781 some Ada 83 compiler systems.
30783 @item Size of Access Types
30784 A common assumption in Ada 83 code is that an access type is in fact a pointer,
30785 and that therefore it will be the same size as a System.Address value. This
30786 assumption is true for GNAT in most cases with one exception. For the case of
30787 a pointer to an unconstrained array type (where the bounds may vary from one
30788 value of the access type to another), the default is to use a ``fat pointer'',
30789 which is represented as two separate pointers, one to the bounds, and one to
30790 the array. This representation has a number of advantages, including improved
30791 efficiency. However, it may cause some difficulties in porting existing Ada 83
30792 code which makes the assumption that, for example, pointers fit in 32 bits on
30793 a machine with 32-bit addressing.
30795 To get around this problem, GNAT also permits the use of ``thin pointers'' for
30796 access types in this case (where the designated type is an unconstrained array
30797 type). These thin pointers are indeed the same size as a System.Address value.
30798 To specify a thin pointer, use a size clause for the type, for example:
30800 @smallexample @c ada
30801 type X is access all String;
30802 for X'Size use Standard'Address_Size;
30806 which will cause the type X to be represented using a single pointer.
30807 When using this representation, the bounds are right behind the array.
30808 This representation is slightly less efficient, and does not allow quite
30809 such flexibility in the use of foreign pointers or in using the
30810 Unrestricted_Access attribute to create pointers to non-aliased objects.
30811 But for any standard portable use of the access type it will work in
30812 a functionally correct manner and allow porting of existing code.
30813 Note that another way of forcing a thin pointer representation
30814 is to use a component size clause for the element size in an array,
30815 or a record representation clause for an access field in a record.
30819 @c This brief section is only in the non-VMS version
30820 @c The complete chapter on HP Ada is in the VMS version
30821 @node Compatibility with HP Ada 83
30822 @section Compatibility with HP Ada 83
30825 The VMS version of GNAT fully implements all the pragmas and attributes
30826 provided by HP Ada 83, as well as providing the standard HP Ada 83
30827 libraries, including Starlet. In addition, data layouts and parameter
30828 passing conventions are highly compatible. This means that porting
30829 existing HP Ada 83 code to GNAT in VMS systems should be easier than
30830 most other porting efforts. The following are some of the most
30831 significant differences between GNAT and HP Ada 83.
30834 @item Default floating-point representation
30835 In GNAT, the default floating-point format is IEEE, whereas in HP Ada 83,
30836 it is VMS format. GNAT does implement the necessary pragmas
30837 (Long_Float, Float_Representation) for changing this default.
30840 The package System in GNAT exactly corresponds to the definition in the
30841 Ada 95 reference manual, which means that it excludes many of the
30842 HP Ada 83 extensions. However, a separate package Aux_DEC is provided
30843 that contains the additional definitions, and a special pragma,
30844 Extend_System allows this package to be treated transparently as an
30845 extension of package System.
30848 The definitions provided by Aux_DEC are exactly compatible with those
30849 in the HP Ada 83 version of System, with one exception.
30850 HP Ada provides the following declarations:
30852 @smallexample @c ada
30853 TO_ADDRESS (INTEGER)
30854 TO_ADDRESS (UNSIGNED_LONGWORD)
30855 TO_ADDRESS (@i{universal_integer})
30859 The version of TO_ADDRESS taking a @i{universal integer} argument is in fact
30860 an extension to Ada 83 not strictly compatible with the reference manual.
30861 In GNAT, we are constrained to be exactly compatible with the standard,
30862 and this means we cannot provide this capability. In HP Ada 83, the
30863 point of this definition is to deal with a call like:
30865 @smallexample @c ada
30866 TO_ADDRESS (16#12777#);
30870 Normally, according to the Ada 83 standard, one would expect this to be
30871 ambiguous, since it matches both the INTEGER and UNSIGNED_LONGWORD forms
30872 of TO_ADDRESS@. However, in HP Ada 83, there is no ambiguity, since the
30873 definition using @i{universal_integer} takes precedence.
30875 In GNAT, since the version with @i{universal_integer} cannot be supplied, it
30876 is not possible to be 100% compatible. Since there are many programs using
30877 numeric constants for the argument to TO_ADDRESS, the decision in GNAT was
30878 to change the name of the function in the UNSIGNED_LONGWORD case, so the
30879 declarations provided in the GNAT version of AUX_Dec are:
30881 @smallexample @c ada
30882 function To_Address (X : Integer) return Address;
30883 pragma Pure_Function (To_Address);
30885 function To_Address_Long (X : Unsigned_Longword)
30887 pragma Pure_Function (To_Address_Long);
30891 This means that programs using TO_ADDRESS for UNSIGNED_LONGWORD must
30892 change the name to TO_ADDRESS_LONG@.
30894 @item Task_Id values
30895 The Task_Id values assigned will be different in the two systems, and GNAT
30896 does not provide a specified value for the Task_Id of the environment task,
30897 which in GNAT is treated like any other declared task.
30901 For full details on these and other less significant compatibility issues,
30902 see appendix E of the HP publication entitled @cite{HP Ada, Technical
30903 Overview and Comparison on HP Platforms}.
30905 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
30906 attributes are recognized, although only a subset of them can sensibly
30907 be implemented. The description of pragmas in @ref{Implementation
30908 Defined Pragmas,,, gnat_rm, GNAT Reference Manual}
30909 indicates whether or not they are applicable to non-VMS systems.
30913 @node Transitioning to 64-Bit GNAT for OpenVMS
30914 @section Transitioning to 64-Bit @value{EDITION} for OpenVMS
30917 This section is meant to assist users of pre-2006 @value{EDITION}
30918 for Alpha OpenVMS who are transitioning to 64-bit @value{EDITION},
30919 the version of the GNAT technology supplied in 2006 and later for
30920 OpenVMS on both Alpha and I64.
30923 * Introduction to transitioning::
30924 * Migration of 32 bit code::
30925 * Taking advantage of 64 bit addressing::
30926 * Technical details::
30929 @node Introduction to transitioning
30930 @subsection Introduction
30933 64-bit @value{EDITION} for Open VMS has been designed to meet
30938 Providing a full conforming implementation of Ada 95 and Ada 2005
30941 Allowing maximum backward compatibility, thus easing migration of existing
30945 Supplying a path for exploiting the full 64-bit address range
30949 Ada's strong typing semantics has made it
30950 impractical to have different 32-bit and 64-bit modes. As soon as
30951 one object could possibly be outside the 32-bit address space, this
30952 would make it necessary for the @code{System.Address} type to be 64 bits.
30953 In particular, this would cause inconsistencies if 32-bit code is
30954 called from 64-bit code that raises an exception.
30956 This issue has been resolved by always using 64-bit addressing
30957 at the system level, but allowing for automatic conversions between
30958 32-bit and 64-bit addresses where required. Thus users who
30959 do not currently require 64-bit addressing capabilities, can
30960 recompile their code with only minimal changes (and indeed
30961 if the code is written in portable Ada, with no assumptions about
30962 the size of the @code{Address} type, then no changes at all are necessary).
30964 this approach provides a simple, gradual upgrade path to future
30965 use of larger memories than available for 32-bit systems.
30966 Also, newly written applications or libraries will by default
30967 be fully compatible with future systems exploiting 64-bit
30968 addressing capabilities.
30970 @ref{Migration of 32 bit code}, will focus on porting applications
30971 that do not require more than 2 GB of
30972 addressable memory. This code will be referred to as
30973 @emph{32-bit code}.
30974 For applications intending to exploit the full 64-bit address space,
30975 @ref{Taking advantage of 64 bit addressing},
30976 will consider further changes that may be required.
30977 Such code will be referred to below as @emph{64-bit code}.
30979 @node Migration of 32 bit code
30980 @subsection Migration of 32-bit code
30985 * Unchecked conversions::
30986 * Predefined constants::
30987 * Interfacing with C::
30988 * Experience with source compatibility::
30991 @node Address types
30992 @subsubsection Address types
30995 To solve the problem of mixing 64-bit and 32-bit addressing,
30996 while maintaining maximum backward compatibility, the following
30997 approach has been taken:
31001 @code{System.Address} always has a size of 64 bits
31004 @code{System.Short_Address} is a 32-bit subtype of @code{System.Address}
31008 Since @code{System.Short_Address} is a subtype of @code{System.Address},
31009 a @code{Short_Address}
31010 may be used where an @code{Address} is required, and vice versa, without
31011 needing explicit type conversions.
31012 By virtue of the Open VMS parameter passing conventions,
31014 and exported subprograms that have 32-bit address parameters are
31015 compatible with those that have 64-bit address parameters.
31016 (See @ref{Making code 64 bit clean} for details.)
31018 The areas that may need attention are those where record types have
31019 been defined that contain components of the type @code{System.Address}, and
31020 where objects of this type are passed to code expecting a record layout with
31023 Different compilers on different platforms cannot be
31024 expected to represent the same type in the same way,
31025 since alignment constraints
31026 and other system-dependent properties affect the compiler's decision.
31027 For that reason, Ada code
31028 generally uses representation clauses to specify the expected
31029 layout where required.
31031 If such a representation clause uses 32 bits for a component having
31032 the type @code{System.Address}, 64-bit @value{EDITION} for OpenVMS
31033 will detect that error and produce a specific diagnostic message.
31034 The developer should then determine whether the representation
31035 should be 64 bits or not and make either of two changes:
31036 change the size to 64 bits and leave the type as @code{System.Address}, or
31037 leave the size as 32 bits and change the type to @code{System.Short_Address}.
31038 Since @code{Short_Address} is a subtype of @code{Address}, no changes are
31039 required in any code setting or accessing the field; the compiler will
31040 automatically perform any needed conversions between address
31044 @subsubsection Access types
31047 By default, objects designated by access values are always
31048 allocated in the 32-bit
31049 address space. Thus legacy code will never contain
31050 any objects that are not addressable with 32-bit addresses, and
31051 the compiler will never raise exceptions as result of mixing
31052 32-bit and 64-bit addresses.
31054 However, the access values themselves are represented in 64 bits, for optimum
31055 performance and future compatibility with 64-bit code. As was
31056 the case with @code{System.Address}, the compiler will give an error message
31057 if an object or record component has a representation clause that
31058 requires the access value to fit in 32 bits. In such a situation,
31059 an explicit size clause for the access type, specifying 32 bits,
31060 will have the desired effect.
31062 General access types (declared with @code{access all}) can never be
31063 32 bits, as values of such types must be able to refer to any object
31064 of the designated type,
31065 including objects residing outside the 32-bit address range.
31066 Existing Ada 83 code will not contain such type definitions,
31067 however, since general access types were introduced in Ada 95.
31069 @node Unchecked conversions
31070 @subsubsection Unchecked conversions
31073 In the case of an @code{Unchecked_Conversion} where the source type is a
31074 64-bit access type or the type @code{System.Address}, and the target
31075 type is a 32-bit type, the compiler will generate a warning.
31076 Even though the generated code will still perform the required
31077 conversions, it is highly recommended in these cases to use
31078 respectively a 32-bit access type or @code{System.Short_Address}
31079 as the source type.
31081 @node Predefined constants
31082 @subsubsection Predefined constants
31085 The following table shows the correspondence between pre-2006 versions of
31086 @value{EDITION} on Alpha OpenVMS (``Old'') and 64-bit @value{EDITION}
31089 @multitable {@code{System.Short_Memory_Size}} {2**32} {2**64}
31090 @item @b{Constant} @tab @b{Old} @tab @b{New}
31091 @item @code{System.Word_Size} @tab 32 @tab 64
31092 @item @code{System.Memory_Size} @tab 2**32 @tab 2**64
31093 @item @code{System.Short_Memory_Size} @tab 2**32 @tab 2**32
31094 @item @code{System.Address_Size} @tab 32 @tab 64
31098 If you need to refer to the specific
31099 memory size of a 32-bit implementation, instead of the
31100 actual memory size, use @code{System.Short_Memory_Size}
31101 rather than @code{System.Memory_Size}.
31102 Similarly, references to @code{System.Address_Size} may need
31103 to be replaced by @code{System.Short_Address'Size}.
31104 The program @command{gnatfind} may be useful for locating
31105 references to the above constants, so that you can verify that they
31108 @node Interfacing with C
31109 @subsubsection Interfacing with C
31112 In order to minimize the impact of the transition to 64-bit addresses on
31113 legacy programs, some fundamental types in the @code{Interfaces.C}
31114 package hierarchy continue to be represented in 32 bits.
31115 These types are: @code{ptrdiff_t}, @code{size_t}, and @code{chars_ptr}.
31116 This eases integration with the default HP C layout choices, for example
31117 as found in the system routines in @code{DECC$SHR.EXE}.
31118 Because of this implementation choice, the type fully compatible with
31119 @code{chars_ptr} is now @code{Short_Address} and not @code{Address}.
31120 Depending on the context the compiler will issue a
31121 warning or an error when type @code{Address} is used, alerting the user to a
31122 potential problem. Otherwise 32-bit programs that use
31123 @code{Interfaces.C} should normally not require code modifications
31125 The other issue arising with C interfacing concerns pragma @code{Convention}.
31126 For VMS 64-bit systems, there is an issue of the appropriate default size
31127 of C convention pointers in the absence of an explicit size clause. The HP
31128 C compiler can choose either 32 or 64 bits depending on compiler options.
31129 GNAT chooses 32-bits rather than 64-bits in the default case where no size
31130 clause is given. This proves a better choice for porting 32-bit legacy
31131 applications. In order to have a 64-bit representation, it is necessary to
31132 specify a size representation clause. For example:
31134 @smallexample @c ada
31135 type int_star is access Interfaces.C.int;
31136 pragma Convention(C, int_star);
31137 for int_star'Size use 64; -- Necessary to get 64 and not 32 bits
31140 @node Experience with source compatibility
31141 @subsubsection Experience with source compatibility
31144 The Security Server and STARLET on I64 provide an interesting ``test case''
31145 for source compatibility issues, since it is in such system code
31146 where assumptions about @code{Address} size might be expected to occur.
31147 Indeed, there were a small number of occasions in the Security Server
31148 file @file{jibdef.ads}
31149 where a representation clause for a record type specified
31150 32 bits for a component of type @code{Address}.
31151 All of these errors were detected by the compiler.
31152 The repair was obvious and immediate; to simply replace @code{Address} by
31153 @code{Short_Address}.
31155 In the case of STARLET, there were several record types that should
31156 have had representation clauses but did not. In these record types
31157 there was an implicit assumption that an @code{Address} value occupied
31159 These compiled without error, but their usage resulted in run-time error
31160 returns from STARLET system calls.
31161 Future GNAT technology enhancements may include a tool that detects and flags
31162 these sorts of potential source code porting problems.
31164 @c ****************************************
31165 @node Taking advantage of 64 bit addressing
31166 @subsection Taking advantage of 64-bit addressing
31169 * Making code 64 bit clean::
31170 * Allocating memory from the 64 bit storage pool::
31171 * Restrictions on use of 64 bit objects::
31172 * Using 64 bit storage pools by default::
31173 * General access types::
31174 * STARLET and other predefined libraries::
31177 @node Making code 64 bit clean
31178 @subsubsection Making code 64-bit clean
31181 In order to prevent problems that may occur when (parts of) a
31182 system start using memory outside the 32-bit address range,
31183 we recommend some additional guidelines:
31187 For imported subprograms that take parameters of the
31188 type @code{System.Address}, ensure that these subprograms can
31189 indeed handle 64-bit addresses. If not, or when in doubt,
31190 change the subprogram declaration to specify
31191 @code{System.Short_Address} instead.
31194 Resolve all warnings related to size mismatches in
31195 unchecked conversions. Failing to do so causes
31196 erroneous execution if the source object is outside
31197 the 32-bit address space.
31200 (optional) Explicitly use the 32-bit storage pool
31201 for access types used in a 32-bit context, or use
31202 generic access types where possible
31203 (@pxref{Restrictions on use of 64 bit objects}).
31207 If these rules are followed, the compiler will automatically insert
31208 any necessary checks to ensure that no addresses or access values
31209 passed to 32-bit code ever refer to objects outside the 32-bit
31211 Any attempt to do this will raise @code{Constraint_Error}.
31213 @node Allocating memory from the 64 bit storage pool
31214 @subsubsection Allocating memory from the 64-bit storage pool
31217 For any access type @code{T} that potentially requires memory allocations
31218 beyond the 32-bit address space,
31219 use the following representation clause:
31221 @smallexample @c ada
31222 for T'Storage_Pool use System.Pool_64;
31225 @node Restrictions on use of 64 bit objects
31226 @subsubsection Restrictions on use of 64-bit objects
31229 Taking the address of an object allocated from a 64-bit storage pool,
31230 and then passing this address to a subprogram expecting
31231 @code{System.Short_Address},
31232 or assigning it to a variable of type @code{Short_Address}, will cause
31233 @code{Constraint_Error} to be raised. In case the code is not 64-bit clean
31234 (@pxref{Making code 64 bit clean}), or checks are suppressed,
31235 no exception is raised and execution
31236 will become erroneous.
31238 @node Using 64 bit storage pools by default
31239 @subsubsection Using 64-bit storage pools by default
31242 In some cases it may be desirable to have the compiler allocate
31243 from 64-bit storage pools by default. This may be the case for
31244 libraries that are 64-bit clean, but may be used in both 32-bit
31245 and 64-bit contexts. For these cases the following configuration
31246 pragma may be specified:
31248 @smallexample @c ada
31249 pragma Pool_64_Default;
31253 Any code compiled in the context of this pragma will by default
31254 use the @code{System.Pool_64} storage pool. This default may be overridden
31255 for a specific access type @code{T} by the representation clause:
31257 @smallexample @c ada
31258 for T'Storage_Pool use System.Pool_32;
31262 Any object whose address may be passed to a subprogram with a
31263 @code{Short_Address} argument, or assigned to a variable of type
31264 @code{Short_Address}, needs to be allocated from this pool.
31266 @node General access types
31267 @subsubsection General access types
31270 Objects designated by access values from a
31271 general access type (declared with @code{access all}) are never allocated
31272 from a 64-bit storage pool. Code that uses general access types will
31273 accept objects allocated in either 32-bit or 64-bit address spaces,
31274 but never allocate objects outside the 32-bit address space.
31275 Using general access types ensures maximum compatibility with both
31276 32-bit and 64-bit code.
31278 @node STARLET and other predefined libraries
31279 @subsubsection STARLET and other predefined libraries
31282 All code that comes as part of GNAT is 64-bit clean, but the
31283 restrictions given in @ref{Restrictions on use of 64 bit objects},
31284 still apply. Look at the package
31285 specs to see in which contexts objects allocated
31286 in 64-bit address space are acceptable.
31288 @node Technical details
31289 @subsection Technical details
31292 64-bit @value{EDITION} for Open VMS takes advantage of the freedom given in the
31293 Ada standard with respect to the type of @code{System.Address}. Previous
31294 versions of GNAT Pro have defined this type as private and implemented it as a
31297 In order to allow defining @code{System.Short_Address} as a proper subtype,
31298 and to match the implicit sign extension in parameter passing,
31299 in 64-bit @value{EDITION} for Open VMS, @code{System.Address} is defined as a
31300 visible (i.e., non-private) integer type.
31301 Standard operations on the type, such as the binary operators ``+'', ``-'',
31302 etc., that take @code{Address} operands and return an @code{Address} result,
31303 have been hidden by declaring these
31304 @code{abstract}, a feature introduced in Ada 95 that helps avoid the potential
31305 ambiguities that would otherwise result from overloading.
31306 (Note that, although @code{Address} is a visible integer type,
31307 good programming practice dictates against exploiting the type's
31308 integer properties such as literals, since this will compromise
31311 Defining @code{Address} as a visible integer type helps achieve
31312 maximum compatibility for existing Ada code,
31313 without sacrificing the capabilities of the 64-bit architecture.
31316 @c ************************************************
31318 @node Microsoft Windows Topics
31319 @appendix Microsoft Windows Topics
31325 This chapter describes topics that are specific to the Microsoft Windows
31326 platforms (NT, 2000, and XP Professional).
31329 * Using GNAT on Windows::
31330 * Using a network installation of GNAT::
31331 * CONSOLE and WINDOWS subsystems::
31332 * Temporary Files::
31333 * Mixed-Language Programming on Windows::
31334 * Windows Calling Conventions::
31335 * Introduction to Dynamic Link Libraries (DLLs)::
31336 * Using DLLs with GNAT::
31337 * Building DLLs with GNAT::
31338 * Building DLLs with GNAT Project files::
31339 * Building DLLs with gnatdll::
31340 * GNAT and Windows Resources::
31341 * Debugging a DLL::
31342 * Setting Stack Size from gnatlink::
31343 * Setting Heap Size from gnatlink::
31346 @node Using GNAT on Windows
31347 @section Using GNAT on Windows
31350 One of the strengths of the GNAT technology is that its tool set
31351 (@command{gcc}, @command{gnatbind}, @command{gnatlink}, @command{gnatmake}, the
31352 @code{gdb} debugger, etc.) is used in the same way regardless of the
31355 On Windows this tool set is complemented by a number of Microsoft-specific
31356 tools that have been provided to facilitate interoperability with Windows
31357 when this is required. With these tools:
31362 You can build applications using the @code{CONSOLE} or @code{WINDOWS}
31366 You can use any Dynamically Linked Library (DLL) in your Ada code (both
31367 relocatable and non-relocatable DLLs are supported).
31370 You can build Ada DLLs for use in other applications. These applications
31371 can be written in a language other than Ada (e.g., C, C++, etc). Again both
31372 relocatable and non-relocatable Ada DLLs are supported.
31375 You can include Windows resources in your Ada application.
31378 You can use or create COM/DCOM objects.
31382 Immediately below are listed all known general GNAT-for-Windows restrictions.
31383 Other restrictions about specific features like Windows Resources and DLLs
31384 are listed in separate sections below.
31389 It is not possible to use @code{GetLastError} and @code{SetLastError}
31390 when tasking, protected records, or exceptions are used. In these
31391 cases, in order to implement Ada semantics, the GNAT run-time system
31392 calls certain Win32 routines that set the last error variable to 0 upon
31393 success. It should be possible to use @code{GetLastError} and
31394 @code{SetLastError} when tasking, protected record, and exception
31395 features are not used, but it is not guaranteed to work.
31398 It is not possible to link against Microsoft libraries except for
31399 import libraries. The library must be built to be compatible with
31400 @file{MSVCRT.LIB} (/MD Microsoft compiler option), @file{LIBC.LIB} and
31401 @file{LIBCMT.LIB} (/ML or /MT Microsoft compiler options) are known to
31402 not be compatible with the GNAT runtime. Even if the library is
31403 compatible with @file{MSVCRT.LIB} it is not guaranteed to work.
31406 When the compilation environment is located on FAT32 drives, users may
31407 experience recompilations of the source files that have not changed if
31408 Daylight Saving Time (DST) state has changed since the last time files
31409 were compiled. NTFS drives do not have this problem.
31412 No components of the GNAT toolset use any entries in the Windows
31413 registry. The only entries that can be created are file associations and
31414 PATH settings, provided the user has chosen to create them at installation
31415 time, as well as some minimal book-keeping information needed to correctly
31416 uninstall or integrate different GNAT products.
31419 @node Using a network installation of GNAT
31420 @section Using a network installation of GNAT
31423 Make sure the system on which GNAT is installed is accessible from the
31424 current machine, i.e., the install location is shared over the network.
31425 Shared resources are accessed on Windows by means of UNC paths, which
31426 have the format @code{\\server\sharename\path}
31428 In order to use such a network installation, simply add the UNC path of the
31429 @file{bin} directory of your GNAT installation in front of your PATH. For
31430 example, if GNAT is installed in @file{\GNAT} directory of a share location
31431 called @file{c-drive} on a machine @file{LOKI}, the following command will
31434 @code{@ @ @ path \\loki\c-drive\gnat\bin;%path%}
31436 Be aware that every compilation using the network installation results in the
31437 transfer of large amounts of data across the network and will likely cause
31438 serious performance penalty.
31440 @node CONSOLE and WINDOWS subsystems
31441 @section CONSOLE and WINDOWS subsystems
31442 @cindex CONSOLE Subsystem
31443 @cindex WINDOWS Subsystem
31447 There are two main subsystems under Windows. The @code{CONSOLE} subsystem
31448 (which is the default subsystem) will always create a console when
31449 launching the application. This is not something desirable when the
31450 application has a Windows GUI. To get rid of this console the
31451 application must be using the @code{WINDOWS} subsystem. To do so
31452 the @option{-mwindows} linker option must be specified.
31455 $ gnatmake winprog -largs -mwindows
31458 @node Temporary Files
31459 @section Temporary Files
31460 @cindex Temporary files
31463 It is possible to control where temporary files gets created by setting
31464 the @env{TMP} environment variable. The file will be created:
31467 @item Under the directory pointed to by the @env{TMP} environment variable if
31468 this directory exists.
31470 @item Under @file{c:\temp}, if the @env{TMP} environment variable is not
31471 set (or not pointing to a directory) and if this directory exists.
31473 @item Under the current working directory otherwise.
31477 This allows you to determine exactly where the temporary
31478 file will be created. This is particularly useful in networked
31479 environments where you may not have write access to some
31482 @node Mixed-Language Programming on Windows
31483 @section Mixed-Language Programming on Windows
31486 Developing pure Ada applications on Windows is no different than on
31487 other GNAT-supported platforms. However, when developing or porting an
31488 application that contains a mix of Ada and C/C++, the choice of your
31489 Windows C/C++ development environment conditions your overall
31490 interoperability strategy.
31492 If you use @command{gcc} to compile the non-Ada part of your application,
31493 there are no Windows-specific restrictions that affect the overall
31494 interoperability with your Ada code. If you plan to use
31495 Microsoft tools (e.g.@: Microsoft Visual C/C++), you should be aware of
31496 the following limitations:
31500 You cannot link your Ada code with an object or library generated with
31501 Microsoft tools if these use the @code{.tls} section (Thread Local
31502 Storage section) since the GNAT linker does not yet support this section.
31505 You cannot link your Ada code with an object or library generated with
31506 Microsoft tools if these use I/O routines other than those provided in
31507 the Microsoft DLL: @code{msvcrt.dll}. This is because the GNAT run time
31508 uses the services of @code{msvcrt.dll} for its I/Os. Use of other I/O
31509 libraries can cause a conflict with @code{msvcrt.dll} services. For
31510 instance Visual C++ I/O stream routines conflict with those in
31515 If you do want to use the Microsoft tools for your non-Ada code and hit one
31516 of the above limitations, you have two choices:
31520 Encapsulate your non-Ada code in a DLL to be linked with your Ada
31521 application. In this case, use the Microsoft or whatever environment to
31522 build the DLL and use GNAT to build your executable
31523 (@pxref{Using DLLs with GNAT}).
31526 Or you can encapsulate your Ada code in a DLL to be linked with the
31527 other part of your application. In this case, use GNAT to build the DLL
31528 (@pxref{Building DLLs with GNAT}) and use the Microsoft or whatever
31529 environment to build your executable.
31532 @node Windows Calling Conventions
31533 @section Windows Calling Conventions
31538 * C Calling Convention::
31539 * Stdcall Calling Convention::
31540 * Win32 Calling Convention::
31541 * DLL Calling Convention::
31545 When a subprogram @code{F} (caller) calls a subprogram @code{G}
31546 (callee), there are several ways to push @code{G}'s parameters on the
31547 stack and there are several possible scenarios to clean up the stack
31548 upon @code{G}'s return. A calling convention is an agreed upon software
31549 protocol whereby the responsibilities between the caller (@code{F}) and
31550 the callee (@code{G}) are clearly defined. Several calling conventions
31551 are available for Windows:
31555 @code{C} (Microsoft defined)
31558 @code{Stdcall} (Microsoft defined)
31561 @code{Win32} (GNAT specific)
31564 @code{DLL} (GNAT specific)
31567 @node C Calling Convention
31568 @subsection @code{C} Calling Convention
31571 This is the default calling convention used when interfacing to C/C++
31572 routines compiled with either @command{gcc} or Microsoft Visual C++.
31574 In the @code{C} calling convention subprogram parameters are pushed on the
31575 stack by the caller from right to left. The caller itself is in charge of
31576 cleaning up the stack after the call. In addition, the name of a routine
31577 with @code{C} calling convention is mangled by adding a leading underscore.
31579 The name to use on the Ada side when importing (or exporting) a routine
31580 with @code{C} calling convention is the name of the routine. For
31581 instance the C function:
31584 int get_val (long);
31588 should be imported from Ada as follows:
31590 @smallexample @c ada
31592 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
31593 pragma Import (C, Get_Val, External_Name => "get_val");
31598 Note that in this particular case the @code{External_Name} parameter could
31599 have been omitted since, when missing, this parameter is taken to be the
31600 name of the Ada entity in lower case. When the @code{Link_Name} parameter
31601 is missing, as in the above example, this parameter is set to be the
31602 @code{External_Name} with a leading underscore.
31604 When importing a variable defined in C, you should always use the @code{C}
31605 calling convention unless the object containing the variable is part of a
31606 DLL (in which case you should use the @code{Stdcall} calling
31607 convention, @pxref{Stdcall Calling Convention}).
31609 @node Stdcall Calling Convention
31610 @subsection @code{Stdcall} Calling Convention
31613 This convention, which was the calling convention used for Pascal
31614 programs, is used by Microsoft for all the routines in the Win32 API for
31615 efficiency reasons. It must be used to import any routine for which this
31616 convention was specified.
31618 In the @code{Stdcall} calling convention subprogram parameters are pushed
31619 on the stack by the caller from right to left. The callee (and not the
31620 caller) is in charge of cleaning the stack on routine exit. In addition,
31621 the name of a routine with @code{Stdcall} calling convention is mangled by
31622 adding a leading underscore (as for the @code{C} calling convention) and a
31623 trailing @code{@@}@code{@var{nn}}, where @var{nn} is the overall size (in
31624 bytes) of the parameters passed to the routine.
31626 The name to use on the Ada side when importing a C routine with a
31627 @code{Stdcall} calling convention is the name of the C routine. The leading
31628 underscore and trailing @code{@@}@code{@var{nn}} are added automatically by
31629 the compiler. For instance the Win32 function:
31632 @b{APIENTRY} int get_val (long);
31636 should be imported from Ada as follows:
31638 @smallexample @c ada
31640 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
31641 pragma Import (Stdcall, Get_Val);
31642 -- On the x86 a long is 4 bytes, so the Link_Name is "_get_val@@4"
31647 As for the @code{C} calling convention, when the @code{External_Name}
31648 parameter is missing, it is taken to be the name of the Ada entity in lower
31649 case. If instead of writing the above import pragma you write:
31651 @smallexample @c ada
31653 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
31654 pragma Import (Stdcall, Get_Val, External_Name => "retrieve_val");
31659 then the imported routine is @code{_retrieve_val@@4}. However, if instead
31660 of specifying the @code{External_Name} parameter you specify the
31661 @code{Link_Name} as in the following example:
31663 @smallexample @c ada
31665 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
31666 pragma Import (Stdcall, Get_Val, Link_Name => "retrieve_val");
31671 then the imported routine is @code{retrieve_val}, that is, there is no
31672 decoration at all. No leading underscore and no Stdcall suffix
31673 @code{@@}@code{@var{nn}}.
31676 This is especially important as in some special cases a DLL's entry
31677 point name lacks a trailing @code{@@}@code{@var{nn}} while the exported
31678 name generated for a call has it.
31681 It is also possible to import variables defined in a DLL by using an
31682 import pragma for a variable. As an example, if a DLL contains a
31683 variable defined as:
31690 then, to access this variable from Ada you should write:
31692 @smallexample @c ada
31694 My_Var : Interfaces.C.int;
31695 pragma Import (Stdcall, My_Var);
31700 Note that to ease building cross-platform bindings this convention
31701 will be handled as a @code{C} calling convention on non-Windows platforms.
31703 @node Win32 Calling Convention
31704 @subsection @code{Win32} Calling Convention
31707 This convention, which is GNAT-specific is fully equivalent to the
31708 @code{Stdcall} calling convention described above.
31710 @node DLL Calling Convention
31711 @subsection @code{DLL} Calling Convention
31714 This convention, which is GNAT-specific is fully equivalent to the
31715 @code{Stdcall} calling convention described above.
31717 @node Introduction to Dynamic Link Libraries (DLLs)
31718 @section Introduction to Dynamic Link Libraries (DLLs)
31722 A Dynamically Linked Library (DLL) is a library that can be shared by
31723 several applications running under Windows. A DLL can contain any number of
31724 routines and variables.
31726 One advantage of DLLs is that you can change and enhance them without
31727 forcing all the applications that depend on them to be relinked or
31728 recompiled. However, you should be aware than all calls to DLL routines are
31729 slower since, as you will understand below, such calls are indirect.
31731 To illustrate the remainder of this section, suppose that an application
31732 wants to use the services of a DLL @file{API.dll}. To use the services
31733 provided by @file{API.dll} you must statically link against the DLL or
31734 an import library which contains a jump table with an entry for each
31735 routine and variable exported by the DLL. In the Microsoft world this
31736 import library is called @file{API.lib}. When using GNAT this import
31737 library is called either @file{libAPI.dll.a}, @file{libapi.dll.a},
31738 @file{libAPI.a} or @file{libapi.a} (names are case insensitive).
31740 After you have linked your application with the DLL or the import library
31741 and you run your application, here is what happens:
31745 Your application is loaded into memory.
31748 The DLL @file{API.dll} is mapped into the address space of your
31749 application. This means that:
31753 The DLL will use the stack of the calling thread.
31756 The DLL will use the virtual address space of the calling process.
31759 The DLL will allocate memory from the virtual address space of the calling
31763 Handles (pointers) can be safely exchanged between routines in the DLL
31764 routines and routines in the application using the DLL.
31768 The entries in the jump table (from the import library @file{libAPI.dll.a}
31769 or @file{API.lib} or automatically created when linking against a DLL)
31770 which is part of your application are initialized with the addresses
31771 of the routines and variables in @file{API.dll}.
31774 If present in @file{API.dll}, routines @code{DllMain} or
31775 @code{DllMainCRTStartup} are invoked. These routines typically contain
31776 the initialization code needed for the well-being of the routines and
31777 variables exported by the DLL.
31781 There is an additional point which is worth mentioning. In the Windows
31782 world there are two kind of DLLs: relocatable and non-relocatable
31783 DLLs. Non-relocatable DLLs can only be loaded at a very specific address
31784 in the target application address space. If the addresses of two
31785 non-relocatable DLLs overlap and these happen to be used by the same
31786 application, a conflict will occur and the application will run
31787 incorrectly. Hence, when possible, it is always preferable to use and
31788 build relocatable DLLs. Both relocatable and non-relocatable DLLs are
31789 supported by GNAT. Note that the @option{-s} linker option (see GNU Linker
31790 User's Guide) removes the debugging symbols from the DLL but the DLL can
31791 still be relocated.
31793 As a side note, an interesting difference between Microsoft DLLs and
31794 Unix shared libraries, is the fact that on most Unix systems all public
31795 routines are exported by default in a Unix shared library, while under
31796 Windows it is possible (but not required) to list exported routines in
31797 a definition file (@pxref{The Definition File}).
31799 @node Using DLLs with GNAT
31800 @section Using DLLs with GNAT
31803 * Creating an Ada Spec for the DLL Services::
31804 * Creating an Import Library::
31808 To use the services of a DLL, say @file{API.dll}, in your Ada application
31813 The Ada spec for the routines and/or variables you want to access in
31814 @file{API.dll}. If not available this Ada spec must be built from the C/C++
31815 header files provided with the DLL.
31818 The import library (@file{libAPI.dll.a} or @file{API.lib}). As previously
31819 mentioned an import library is a statically linked library containing the
31820 import table which will be filled at load time to point to the actual
31821 @file{API.dll} routines. Sometimes you don't have an import library for the
31822 DLL you want to use. The following sections will explain how to build
31823 one. Note that this is optional.
31826 The actual DLL, @file{API.dll}.
31830 Once you have all the above, to compile an Ada application that uses the
31831 services of @file{API.dll} and whose main subprogram is @code{My_Ada_App},
31832 you simply issue the command
31835 $ gnatmake my_ada_app -largs -lAPI
31839 The argument @option{-largs -lAPI} at the end of the @command{gnatmake} command
31840 tells the GNAT linker to look first for a library named @file{API.lib}
31841 (Microsoft-style name) and if not found for a libraries named
31842 @file{libAPI.dll.a}, @file{API.dll.a} or @file{libAPI.a}.
31843 (GNAT-style name). Note that if the Ada package spec for @file{API.dll}
31844 contains the following pragma
31846 @smallexample @c ada
31847 pragma Linker_Options ("-lAPI");
31851 you do not have to add @option{-largs -lAPI} at the end of the
31852 @command{gnatmake} command.
31854 If any one of the items above is missing you will have to create it
31855 yourself. The following sections explain how to do so using as an
31856 example a fictitious DLL called @file{API.dll}.
31858 @node Creating an Ada Spec for the DLL Services
31859 @subsection Creating an Ada Spec for the DLL Services
31862 A DLL typically comes with a C/C++ header file which provides the
31863 definitions of the routines and variables exported by the DLL. The Ada
31864 equivalent of this header file is a package spec that contains definitions
31865 for the imported entities. If the DLL you intend to use does not come with
31866 an Ada spec you have to generate one such spec yourself. For example if
31867 the header file of @file{API.dll} is a file @file{api.h} containing the
31868 following two definitions:
31880 then the equivalent Ada spec could be:
31882 @smallexample @c ada
31885 with Interfaces.C.Strings;
31890 function Get (Str : C.Strings.Chars_Ptr) return C.int;
31893 pragma Import (C, Get);
31894 pragma Import (DLL, Some_Var);
31901 Note that a variable is
31902 @strong{always imported with a Stdcall convention}. A function
31903 can have @code{C} or @code{Stdcall} convention.
31904 (@pxref{Windows Calling Conventions}).
31906 @node Creating an Import Library
31907 @subsection Creating an Import Library
31908 @cindex Import library
31911 * The Definition File::
31912 * GNAT-Style Import Library::
31913 * Microsoft-Style Import Library::
31917 If a Microsoft-style import library @file{API.lib} or a GNAT-style
31918 import library @file{libAPI.dll.a} or @file{libAPI.a} is available
31919 with @file{API.dll} you can skip this section. You can also skip this
31920 section if @file{API.dll} or @file{libAPI.dll} is built with GNU tools
31921 as in this case it is possible to link directly against the
31922 DLL. Otherwise read on.
31924 @node The Definition File
31925 @subsubsection The Definition File
31926 @cindex Definition file
31930 As previously mentioned, and unlike Unix systems, the list of symbols
31931 that are exported from a DLL must be provided explicitly in Windows.
31932 The main goal of a definition file is precisely that: list the symbols
31933 exported by a DLL. A definition file (usually a file with a @code{.def}
31934 suffix) has the following structure:
31939 @r{[}LIBRARY @var{name}@r{]}
31940 @r{[}DESCRIPTION @var{string}@r{]}
31950 @item LIBRARY @var{name}
31951 This section, which is optional, gives the name of the DLL.
31953 @item DESCRIPTION @var{string}
31954 This section, which is optional, gives a description string that will be
31955 embedded in the import library.
31958 This section gives the list of exported symbols (procedures, functions or
31959 variables). For instance in the case of @file{API.dll} the @code{EXPORTS}
31960 section of @file{API.def} looks like:
31974 Note that you must specify the correct suffix (@code{@@}@code{@var{nn}})
31975 (@pxref{Windows Calling Conventions}) for a Stdcall
31976 calling convention function in the exported symbols list.
31979 There can actually be other sections in a definition file, but these
31980 sections are not relevant to the discussion at hand.
31982 @node GNAT-Style Import Library
31983 @subsubsection GNAT-Style Import Library
31986 To create a static import library from @file{API.dll} with the GNAT tools
31987 you should proceed as follows:
31991 Create the definition file @file{API.def} (@pxref{The Definition File}).
31992 For that use the @code{dll2def} tool as follows:
31995 $ dll2def API.dll > API.def
31999 @code{dll2def} is a very simple tool: it takes as input a DLL and prints
32000 to standard output the list of entry points in the DLL. Note that if
32001 some routines in the DLL have the @code{Stdcall} convention
32002 (@pxref{Windows Calling Conventions}) with stripped @code{@@}@var{nn}
32003 suffix then you'll have to edit @file{api.def} to add it, and specify
32004 @option{-k} to @command{gnatdll} when creating the import library.
32007 Here are some hints to find the right @code{@@}@var{nn} suffix.
32011 If you have the Microsoft import library (.lib), it is possible to get
32012 the right symbols by using Microsoft @code{dumpbin} tool (see the
32013 corresponding Microsoft documentation for further details).
32016 $ dumpbin /exports api.lib
32020 If you have a message about a missing symbol at link time the compiler
32021 tells you what symbol is expected. You just have to go back to the
32022 definition file and add the right suffix.
32026 Build the import library @code{libAPI.dll.a}, using @code{gnatdll}
32027 (@pxref{Using gnatdll}) as follows:
32030 $ gnatdll -e API.def -d API.dll
32034 @code{gnatdll} takes as input a definition file @file{API.def} and the
32035 name of the DLL containing the services listed in the definition file
32036 @file{API.dll}. The name of the static import library generated is
32037 computed from the name of the definition file as follows: if the
32038 definition file name is @var{xyz}@code{.def}, the import library name will
32039 be @code{lib}@var{xyz}@code{.a}. Note that in the previous example option
32040 @option{-e} could have been removed because the name of the definition
32041 file (before the ``@code{.def}'' suffix) is the same as the name of the
32042 DLL (@pxref{Using gnatdll} for more information about @code{gnatdll}).
32045 @node Microsoft-Style Import Library
32046 @subsubsection Microsoft-Style Import Library
32049 With GNAT you can either use a GNAT-style or Microsoft-style import
32050 library. A Microsoft import library is needed only if you plan to make an
32051 Ada DLL available to applications developed with Microsoft
32052 tools (@pxref{Mixed-Language Programming on Windows}).
32054 To create a Microsoft-style import library for @file{API.dll} you
32055 should proceed as follows:
32059 Create the definition file @file{API.def} from the DLL. For this use either
32060 the @code{dll2def} tool as described above or the Microsoft @code{dumpbin}
32061 tool (see the corresponding Microsoft documentation for further details).
32064 Build the actual import library using Microsoft's @code{lib} utility:
32067 $ lib -machine:IX86 -def:API.def -out:API.lib
32071 If you use the above command the definition file @file{API.def} must
32072 contain a line giving the name of the DLL:
32079 See the Microsoft documentation for further details about the usage of
32083 @node Building DLLs with GNAT
32084 @section Building DLLs with GNAT
32085 @cindex DLLs, building
32088 This section explain how to build DLLs using the GNAT built-in DLL
32089 support. With the following procedure it is straight forward to build
32090 and use DLLs with GNAT.
32094 @item building object files
32096 The first step is to build all objects files that are to be included
32097 into the DLL. This is done by using the standard @command{gnatmake} tool.
32099 @item building the DLL
32101 To build the DLL you must use @command{gcc}'s @option{-shared}
32102 option. It is quite simple to use this method:
32105 $ gcc -shared -o api.dll obj1.o obj2.o @dots{}
32108 It is important to note that in this case all symbols found in the
32109 object files are automatically exported. It is possible to restrict
32110 the set of symbols to export by passing to @command{gcc} a definition
32111 file, @pxref{The Definition File}. For example:
32114 $ gcc -shared -o api.dll api.def obj1.o obj2.o @dots{}
32117 If you use a definition file you must export the elaboration procedures
32118 for every package that required one. Elaboration procedures are named
32119 using the package name followed by "_E".
32121 @item preparing DLL to be used
32123 For the DLL to be used by client programs the bodies must be hidden
32124 from it and the .ali set with read-only attribute. This is very important
32125 otherwise GNAT will recompile all packages and will not actually use
32126 the code in the DLL. For example:
32130 $ copy *.ads *.ali api.dll apilib
32131 $ attrib +R apilib\*.ali
32136 At this point it is possible to use the DLL by directly linking
32137 against it. Note that you must use the GNAT shared runtime when using
32138 GNAT shared libraries. This is achieved by using @option{-shared} binder's
32142 $ gnatmake main -Iapilib -bargs -shared -largs -Lapilib -lAPI
32145 @node Building DLLs with GNAT Project files
32146 @section Building DLLs with GNAT Project files
32147 @cindex DLLs, building
32150 There is nothing specific to Windows in the build process.
32151 @pxref{Library Projects}.
32154 Due to a system limitation, it is not possible under Windows to create threads
32155 when inside the @code{DllMain} routine which is used for auto-initialization
32156 of shared libraries, so it is not possible to have library level tasks in SALs.
32158 @node Building DLLs with gnatdll
32159 @section Building DLLs with gnatdll
32160 @cindex DLLs, building
32163 * Limitations When Using Ada DLLs from Ada::
32164 * Exporting Ada Entities::
32165 * Ada DLLs and Elaboration::
32166 * Ada DLLs and Finalization::
32167 * Creating a Spec for Ada DLLs::
32168 * Creating the Definition File::
32173 Note that it is preferred to use the built-in GNAT DLL support
32174 (@pxref{Building DLLs with GNAT}) or GNAT Project files
32175 (@pxref{Building DLLs with GNAT Project files}) to build DLLs.
32177 This section explains how to build DLLs containing Ada code using
32178 @code{gnatdll}. These DLLs will be referred to as Ada DLLs in the
32179 remainder of this section.
32181 The steps required to build an Ada DLL that is to be used by Ada as well as
32182 non-Ada applications are as follows:
32186 You need to mark each Ada @i{entity} exported by the DLL with a @code{C} or
32187 @code{Stdcall} calling convention to avoid any Ada name mangling for the
32188 entities exported by the DLL (@pxref{Exporting Ada Entities}). You can
32189 skip this step if you plan to use the Ada DLL only from Ada applications.
32192 Your Ada code must export an initialization routine which calls the routine
32193 @code{adainit} generated by @command{gnatbind} to perform the elaboration of
32194 the Ada code in the DLL (@pxref{Ada DLLs and Elaboration}). The initialization
32195 routine exported by the Ada DLL must be invoked by the clients of the DLL
32196 to initialize the DLL.
32199 When useful, the DLL should also export a finalization routine which calls
32200 routine @code{adafinal} generated by @command{gnatbind} to perform the
32201 finalization of the Ada code in the DLL (@pxref{Ada DLLs and Finalization}).
32202 The finalization routine exported by the Ada DLL must be invoked by the
32203 clients of the DLL when the DLL services are no further needed.
32206 You must provide a spec for the services exported by the Ada DLL in each
32207 of the programming languages to which you plan to make the DLL available.
32210 You must provide a definition file listing the exported entities
32211 (@pxref{The Definition File}).
32214 Finally you must use @code{gnatdll} to produce the DLL and the import
32215 library (@pxref{Using gnatdll}).
32219 Note that a relocatable DLL stripped using the @code{strip}
32220 binutils tool will not be relocatable anymore. To build a DLL without
32221 debug information pass @code{-largs -s} to @code{gnatdll}. This
32222 restriction does not apply to a DLL built using a Library Project.
32223 @pxref{Library Projects}.
32225 @node Limitations When Using Ada DLLs from Ada
32226 @subsection Limitations When Using Ada DLLs from Ada
32229 When using Ada DLLs from Ada applications there is a limitation users
32230 should be aware of. Because on Windows the GNAT run time is not in a DLL of
32231 its own, each Ada DLL includes a part of the GNAT run time. Specifically,
32232 each Ada DLL includes the services of the GNAT run time that are necessary
32233 to the Ada code inside the DLL. As a result, when an Ada program uses an
32234 Ada DLL there are two independent GNAT run times: one in the Ada DLL and
32235 one in the main program.
32237 It is therefore not possible to exchange GNAT run-time objects between the
32238 Ada DLL and the main Ada program. Example of GNAT run-time objects are file
32239 handles (e.g.@: @code{Text_IO.File_Type}), tasks types, protected objects
32242 It is completely safe to exchange plain elementary, array or record types,
32243 Windows object handles, etc.
32245 @node Exporting Ada Entities
32246 @subsection Exporting Ada Entities
32247 @cindex Export table
32250 Building a DLL is a way to encapsulate a set of services usable from any
32251 application. As a result, the Ada entities exported by a DLL should be
32252 exported with the @code{C} or @code{Stdcall} calling conventions to avoid
32253 any Ada name mangling. As an example here is an Ada package
32254 @code{API}, spec and body, exporting two procedures, a function, and a
32257 @smallexample @c ada
32260 with Interfaces.C; use Interfaces;
32262 Count : C.int := 0;
32263 function Factorial (Val : C.int) return C.int;
32265 procedure Initialize_API;
32266 procedure Finalize_API;
32267 -- Initialization & Finalization routines. More in the next section.
32269 pragma Export (C, Initialize_API);
32270 pragma Export (C, Finalize_API);
32271 pragma Export (C, Count);
32272 pragma Export (C, Factorial);
32278 @smallexample @c ada
32281 package body API is
32282 function Factorial (Val : C.int) return C.int is
32285 Count := Count + 1;
32286 for K in 1 .. Val loop
32292 procedure Initialize_API is
32294 pragma Import (C, Adainit);
32297 end Initialize_API;
32299 procedure Finalize_API is
32300 procedure Adafinal;
32301 pragma Import (C, Adafinal);
32311 If the Ada DLL you are building will only be used by Ada applications
32312 you do not have to export Ada entities with a @code{C} or @code{Stdcall}
32313 convention. As an example, the previous package could be written as
32316 @smallexample @c ada
32320 Count : Integer := 0;
32321 function Factorial (Val : Integer) return Integer;
32323 procedure Initialize_API;
32324 procedure Finalize_API;
32325 -- Initialization and Finalization routines.
32331 @smallexample @c ada
32334 package body API is
32335 function Factorial (Val : Integer) return Integer is
32336 Fact : Integer := 1;
32338 Count := Count + 1;
32339 for K in 1 .. Val loop
32346 -- The remainder of this package body is unchanged.
32353 Note that if you do not export the Ada entities with a @code{C} or
32354 @code{Stdcall} convention you will have to provide the mangled Ada names
32355 in the definition file of the Ada DLL
32356 (@pxref{Creating the Definition File}).
32358 @node Ada DLLs and Elaboration
32359 @subsection Ada DLLs and Elaboration
32360 @cindex DLLs and elaboration
32363 The DLL that you are building contains your Ada code as well as all the
32364 routines in the Ada library that are needed by it. The first thing a
32365 user of your DLL must do is elaborate the Ada code
32366 (@pxref{Elaboration Order Handling in GNAT}).
32368 To achieve this you must export an initialization routine
32369 (@code{Initialize_API} in the previous example), which must be invoked
32370 before using any of the DLL services. This elaboration routine must call
32371 the Ada elaboration routine @code{adainit} generated by the GNAT binder
32372 (@pxref{Binding with Non-Ada Main Programs}). See the body of
32373 @code{Initialize_Api} for an example. Note that the GNAT binder is
32374 automatically invoked during the DLL build process by the @code{gnatdll}
32375 tool (@pxref{Using gnatdll}).
32377 When a DLL is loaded, Windows systematically invokes a routine called
32378 @code{DllMain}. It would therefore be possible to call @code{adainit}
32379 directly from @code{DllMain} without having to provide an explicit
32380 initialization routine. Unfortunately, it is not possible to call
32381 @code{adainit} from the @code{DllMain} if your program has library level
32382 tasks because access to the @code{DllMain} entry point is serialized by
32383 the system (that is, only a single thread can execute ``through'' it at a
32384 time), which means that the GNAT run time will deadlock waiting for the
32385 newly created task to complete its initialization.
32387 @node Ada DLLs and Finalization
32388 @subsection Ada DLLs and Finalization
32389 @cindex DLLs and finalization
32392 When the services of an Ada DLL are no longer needed, the client code should
32393 invoke the DLL finalization routine, if available. The DLL finalization
32394 routine is in charge of releasing all resources acquired by the DLL. In the
32395 case of the Ada code contained in the DLL, this is achieved by calling
32396 routine @code{adafinal} generated by the GNAT binder
32397 (@pxref{Binding with Non-Ada Main Programs}).
32398 See the body of @code{Finalize_Api} for an
32399 example. As already pointed out the GNAT binder is automatically invoked
32400 during the DLL build process by the @code{gnatdll} tool
32401 (@pxref{Using gnatdll}).
32403 @node Creating a Spec for Ada DLLs
32404 @subsection Creating a Spec for Ada DLLs
32407 To use the services exported by the Ada DLL from another programming
32408 language (e.g.@: C), you have to translate the specs of the exported Ada
32409 entities in that language. For instance in the case of @code{API.dll},
32410 the corresponding C header file could look like:
32415 extern int *_imp__count;
32416 #define count (*_imp__count)
32417 int factorial (int);
32423 It is important to understand that when building an Ada DLL to be used by
32424 other Ada applications, you need two different specs for the packages
32425 contained in the DLL: one for building the DLL and the other for using
32426 the DLL. This is because the @code{DLL} calling convention is needed to
32427 use a variable defined in a DLL, but when building the DLL, the variable
32428 must have either the @code{Ada} or @code{C} calling convention. As an
32429 example consider a DLL comprising the following package @code{API}:
32431 @smallexample @c ada
32435 Count : Integer := 0;
32437 -- Remainder of the package omitted.
32444 After producing a DLL containing package @code{API}, the spec that
32445 must be used to import @code{API.Count} from Ada code outside of the
32448 @smallexample @c ada
32453 pragma Import (DLL, Count);
32459 @node Creating the Definition File
32460 @subsection Creating the Definition File
32463 The definition file is the last file needed to build the DLL. It lists
32464 the exported symbols. As an example, the definition file for a DLL
32465 containing only package @code{API} (where all the entities are exported
32466 with a @code{C} calling convention) is:
32481 If the @code{C} calling convention is missing from package @code{API},
32482 then the definition file contains the mangled Ada names of the above
32483 entities, which in this case are:
32492 api__initialize_api
32497 @node Using gnatdll
32498 @subsection Using @code{gnatdll}
32502 * gnatdll Example::
32503 * gnatdll behind the Scenes::
32508 @code{gnatdll} is a tool to automate the DLL build process once all the Ada
32509 and non-Ada sources that make up your DLL have been compiled.
32510 @code{gnatdll} is actually in charge of two distinct tasks: build the
32511 static import library for the DLL and the actual DLL. The form of the
32512 @code{gnatdll} command is
32516 $ gnatdll @ovar{switches} @var{list-of-files} @r{[}-largs @var{opts}@r{]}
32521 where @var{list-of-files} is a list of ALI and object files. The object
32522 file list must be the exact list of objects corresponding to the non-Ada
32523 sources whose services are to be included in the DLL. The ALI file list
32524 must be the exact list of ALI files for the corresponding Ada sources
32525 whose services are to be included in the DLL. If @var{list-of-files} is
32526 missing, only the static import library is generated.
32529 You may specify any of the following switches to @code{gnatdll}:
32532 @item -a@ovar{address}
32533 @cindex @option{-a} (@code{gnatdll})
32534 Build a non-relocatable DLL at @var{address}. If @var{address} is not
32535 specified the default address @var{0x11000000} will be used. By default,
32536 when this switch is missing, @code{gnatdll} builds relocatable DLL. We
32537 advise the reader to build relocatable DLL.
32539 @item -b @var{address}
32540 @cindex @option{-b} (@code{gnatdll})
32541 Set the relocatable DLL base address. By default the address is
32544 @item -bargs @var{opts}
32545 @cindex @option{-bargs} (@code{gnatdll})
32546 Binder options. Pass @var{opts} to the binder.
32548 @item -d @var{dllfile}
32549 @cindex @option{-d} (@code{gnatdll})
32550 @var{dllfile} is the name of the DLL. This switch must be present for
32551 @code{gnatdll} to do anything. The name of the generated import library is
32552 obtained algorithmically from @var{dllfile} as shown in the following
32553 example: if @var{dllfile} is @code{xyz.dll}, the import library name is
32554 @code{libxyz.dll.a}. The name of the definition file to use (if not specified
32555 by option @option{-e}) is obtained algorithmically from @var{dllfile}
32556 as shown in the following example:
32557 if @var{dllfile} is @code{xyz.dll}, the definition
32558 file used is @code{xyz.def}.
32560 @item -e @var{deffile}
32561 @cindex @option{-e} (@code{gnatdll})
32562 @var{deffile} is the name of the definition file.
32565 @cindex @option{-g} (@code{gnatdll})
32566 Generate debugging information. This information is stored in the object
32567 file and copied from there to the final DLL file by the linker,
32568 where it can be read by the debugger. You must use the
32569 @option{-g} switch if you plan on using the debugger or the symbolic
32573 @cindex @option{-h} (@code{gnatdll})
32574 Help mode. Displays @code{gnatdll} switch usage information.
32577 @cindex @option{-I} (@code{gnatdll})
32578 Direct @code{gnatdll} to search the @var{dir} directory for source and
32579 object files needed to build the DLL.
32580 (@pxref{Search Paths and the Run-Time Library (RTL)}).
32583 @cindex @option{-k} (@code{gnatdll})
32584 Removes the @code{@@}@var{nn} suffix from the import library's exported
32585 names, but keeps them for the link names. You must specify this
32586 option if you want to use a @code{Stdcall} function in a DLL for which
32587 the @code{@@}@var{nn} suffix has been removed. This is the case for most
32588 of the Windows NT DLL for example. This option has no effect when
32589 @option{-n} option is specified.
32591 @item -l @var{file}
32592 @cindex @option{-l} (@code{gnatdll})
32593 The list of ALI and object files used to build the DLL are listed in
32594 @var{file}, instead of being given in the command line. Each line in
32595 @var{file} contains the name of an ALI or object file.
32598 @cindex @option{-n} (@code{gnatdll})
32599 No Import. Do not create the import library.
32602 @cindex @option{-q} (@code{gnatdll})
32603 Quiet mode. Do not display unnecessary messages.
32606 @cindex @option{-v} (@code{gnatdll})
32607 Verbose mode. Display extra information.
32609 @item -largs @var{opts}
32610 @cindex @option{-largs} (@code{gnatdll})
32611 Linker options. Pass @var{opts} to the linker.
32614 @node gnatdll Example
32615 @subsubsection @code{gnatdll} Example
32618 As an example the command to build a relocatable DLL from @file{api.adb}
32619 once @file{api.adb} has been compiled and @file{api.def} created is
32622 $ gnatdll -d api.dll api.ali
32626 The above command creates two files: @file{libapi.dll.a} (the import
32627 library) and @file{api.dll} (the actual DLL). If you want to create
32628 only the DLL, just type:
32631 $ gnatdll -d api.dll -n api.ali
32635 Alternatively if you want to create just the import library, type:
32638 $ gnatdll -d api.dll
32641 @node gnatdll behind the Scenes
32642 @subsubsection @code{gnatdll} behind the Scenes
32645 This section details the steps involved in creating a DLL. @code{gnatdll}
32646 does these steps for you. Unless you are interested in understanding what
32647 goes on behind the scenes, you should skip this section.
32649 We use the previous example of a DLL containing the Ada package @code{API},
32650 to illustrate the steps necessary to build a DLL. The starting point is a
32651 set of objects that will make up the DLL and the corresponding ALI
32652 files. In the case of this example this means that @file{api.o} and
32653 @file{api.ali} are available. To build a relocatable DLL, @code{gnatdll} does
32658 @code{gnatdll} builds the base file (@file{api.base}). A base file gives
32659 the information necessary to generate relocation information for the
32665 $ gnatlink api -o api.jnk -mdll -Wl,--base-file,api.base
32670 In addition to the base file, the @command{gnatlink} command generates an
32671 output file @file{api.jnk} which can be discarded. The @option{-mdll} switch
32672 asks @command{gnatlink} to generate the routines @code{DllMain} and
32673 @code{DllMainCRTStartup} that are called by the Windows loader when the DLL
32674 is loaded into memory.
32677 @code{gnatdll} uses @code{dlltool} (@pxref{Using dlltool}) to build the
32678 export table (@file{api.exp}). The export table contains the relocation
32679 information in a form which can be used during the final link to ensure
32680 that the Windows loader is able to place the DLL anywhere in memory.
32684 $ dlltool --dllname api.dll --def api.def --base-file api.base \
32685 --output-exp api.exp
32690 @code{gnatdll} builds the base file using the new export table. Note that
32691 @command{gnatbind} must be called once again since the binder generated file
32692 has been deleted during the previous call to @command{gnatlink}.
32697 $ gnatlink api -o api.jnk api.exp -mdll
32698 -Wl,--base-file,api.base
32703 @code{gnatdll} builds the new export table using the new base file and
32704 generates the DLL import library @file{libAPI.dll.a}.
32708 $ dlltool --dllname api.dll --def api.def --base-file api.base \
32709 --output-exp api.exp --output-lib libAPI.a
32714 Finally @code{gnatdll} builds the relocatable DLL using the final export
32720 $ gnatlink api api.exp -o api.dll -mdll
32725 @node Using dlltool
32726 @subsubsection Using @code{dlltool}
32729 @code{dlltool} is the low-level tool used by @code{gnatdll} to build
32730 DLLs and static import libraries. This section summarizes the most
32731 common @code{dlltool} switches. The form of the @code{dlltool} command
32735 $ dlltool @ovar{switches}
32739 @code{dlltool} switches include:
32742 @item --base-file @var{basefile}
32743 @cindex @option{--base-file} (@command{dlltool})
32744 Read the base file @var{basefile} generated by the linker. This switch
32745 is used to create a relocatable DLL.
32747 @item --def @var{deffile}
32748 @cindex @option{--def} (@command{dlltool})
32749 Read the definition file.
32751 @item --dllname @var{name}
32752 @cindex @option{--dllname} (@command{dlltool})
32753 Gives the name of the DLL. This switch is used to embed the name of the
32754 DLL in the static import library generated by @code{dlltool} with switch
32755 @option{--output-lib}.
32758 @cindex @option{-k} (@command{dlltool})
32759 Kill @code{@@}@var{nn} from exported names
32760 (@pxref{Windows Calling Conventions}
32761 for a discussion about @code{Stdcall}-style symbols.
32764 @cindex @option{--help} (@command{dlltool})
32765 Prints the @code{dlltool} switches with a concise description.
32767 @item --output-exp @var{exportfile}
32768 @cindex @option{--output-exp} (@command{dlltool})
32769 Generate an export file @var{exportfile}. The export file contains the
32770 export table (list of symbols in the DLL) and is used to create the DLL.
32772 @item --output-lib @var{libfile}
32773 @cindex @option{--output-lib} (@command{dlltool})
32774 Generate a static import library @var{libfile}.
32777 @cindex @option{-v} (@command{dlltool})
32780 @item --as @var{assembler-name}
32781 @cindex @option{--as} (@command{dlltool})
32782 Use @var{assembler-name} as the assembler. The default is @code{as}.
32785 @node GNAT and Windows Resources
32786 @section GNAT and Windows Resources
32787 @cindex Resources, windows
32790 * Building Resources::
32791 * Compiling Resources::
32792 * Using Resources::
32796 Resources are an easy way to add Windows specific objects to your
32797 application. The objects that can be added as resources include:
32826 This section explains how to build, compile and use resources.
32828 @node Building Resources
32829 @subsection Building Resources
32830 @cindex Resources, building
32833 A resource file is an ASCII file. By convention resource files have an
32834 @file{.rc} extension.
32835 The easiest way to build a resource file is to use Microsoft tools
32836 such as @code{imagedit.exe} to build bitmaps, icons and cursors and
32837 @code{dlgedit.exe} to build dialogs.
32838 It is always possible to build an @file{.rc} file yourself by writing a
32841 It is not our objective to explain how to write a resource file. A
32842 complete description of the resource script language can be found in the
32843 Microsoft documentation.
32845 @node Compiling Resources
32846 @subsection Compiling Resources
32849 @cindex Resources, compiling
32852 This section describes how to build a GNAT-compatible (COFF) object file
32853 containing the resources. This is done using the Resource Compiler
32854 @code{windres} as follows:
32857 $ windres -i myres.rc -o myres.o
32861 By default @code{windres} will run @command{gcc} to preprocess the @file{.rc}
32862 file. You can specify an alternate preprocessor (usually named
32863 @file{cpp.exe}) using the @code{windres} @option{--preprocessor}
32864 parameter. A list of all possible options may be obtained by entering
32865 the command @code{windres} @option{--help}.
32867 It is also possible to use the Microsoft resource compiler @code{rc.exe}
32868 to produce a @file{.res} file (binary resource file). See the
32869 corresponding Microsoft documentation for further details. In this case
32870 you need to use @code{windres} to translate the @file{.res} file to a
32871 GNAT-compatible object file as follows:
32874 $ windres -i myres.res -o myres.o
32877 @node Using Resources
32878 @subsection Using Resources
32879 @cindex Resources, using
32882 To include the resource file in your program just add the
32883 GNAT-compatible object file for the resource(s) to the linker
32884 arguments. With @command{gnatmake} this is done by using the @option{-largs}
32888 $ gnatmake myprog -largs myres.o
32891 @node Debugging a DLL
32892 @section Debugging a DLL
32893 @cindex DLL debugging
32896 * Program and DLL Both Built with GCC/GNAT::
32897 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
32901 Debugging a DLL is similar to debugging a standard program. But
32902 we have to deal with two different executable parts: the DLL and the
32903 program that uses it. We have the following four possibilities:
32907 The program and the DLL are built with @code{GCC/GNAT}.
32909 The program is built with foreign tools and the DLL is built with
32912 The program is built with @code{GCC/GNAT} and the DLL is built with
32918 In this section we address only cases one and two above.
32919 There is no point in trying to debug
32920 a DLL with @code{GNU/GDB}, if there is no GDB-compatible debugging
32921 information in it. To do so you must use a debugger compatible with the
32922 tools suite used to build the DLL.
32924 @node Program and DLL Both Built with GCC/GNAT
32925 @subsection Program and DLL Both Built with GCC/GNAT
32928 This is the simplest case. Both the DLL and the program have @code{GDB}
32929 compatible debugging information. It is then possible to break anywhere in
32930 the process. Let's suppose here that the main procedure is named
32931 @code{ada_main} and that in the DLL there is an entry point named
32935 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) and
32936 program must have been built with the debugging information (see GNAT -g
32937 switch). Here are the step-by-step instructions for debugging it:
32940 @item Launch @code{GDB} on the main program.
32946 @item Start the program and stop at the beginning of the main procedure
32953 This step is required to be able to set a breakpoint inside the DLL. As long
32954 as the program is not run, the DLL is not loaded. This has the
32955 consequence that the DLL debugging information is also not loaded, so it is not
32956 possible to set a breakpoint in the DLL.
32958 @item Set a breakpoint inside the DLL
32961 (gdb) break ada_dll
32968 At this stage a breakpoint is set inside the DLL. From there on
32969 you can use the standard approach to debug the whole program
32970 (@pxref{Running and Debugging Ada Programs}).
32973 @c This used to work, probably because the DLLs were non-relocatable
32974 @c keep this section around until the problem is sorted out.
32976 To break on the @code{DllMain} routine it is not possible to follow
32977 the procedure above. At the time the program stop on @code{ada_main}
32978 the @code{DllMain} routine as already been called. Either you can use
32979 the procedure below @pxref{Debugging the DLL Directly} or this procedure:
32982 @item Launch @code{GDB} on the main program.
32988 @item Load DLL symbols
32991 (gdb) add-sym api.dll
32994 @item Set a breakpoint inside the DLL
32997 (gdb) break ada_dll.adb:45
33000 Note that at this point it is not possible to break using the routine symbol
33001 directly as the program is not yet running. The solution is to break
33002 on the proper line (break in @file{ada_dll.adb} line 45).
33004 @item Start the program
33013 @node Program Built with Foreign Tools and DLL Built with GCC/GNAT
33014 @subsection Program Built with Foreign Tools and DLL Built with GCC/GNAT
33017 * Debugging the DLL Directly::
33018 * Attaching to a Running Process::
33022 In this case things are slightly more complex because it is not possible to
33023 start the main program and then break at the beginning to load the DLL and the
33024 associated DLL debugging information. It is not possible to break at the
33025 beginning of the program because there is no @code{GDB} debugging information,
33026 and therefore there is no direct way of getting initial control. This
33027 section addresses this issue by describing some methods that can be used
33028 to break somewhere in the DLL to debug it.
33031 First suppose that the main procedure is named @code{main} (this is for
33032 example some C code built with Microsoft Visual C) and that there is a
33033 DLL named @code{test.dll} containing an Ada entry point named
33037 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) must have
33038 been built with debugging information (see GNAT -g option).
33040 @node Debugging the DLL Directly
33041 @subsubsection Debugging the DLL Directly
33045 Find out the executable starting address
33048 $ objdump --file-header main.exe
33051 The starting address is reported on the last line. For example:
33054 main.exe: file format pei-i386
33055 architecture: i386, flags 0x0000010a:
33056 EXEC_P, HAS_DEBUG, D_PAGED
33057 start address 0x00401010
33061 Launch the debugger on the executable.
33068 Set a breakpoint at the starting address, and launch the program.
33071 $ (gdb) break *0x00401010
33075 The program will stop at the given address.
33078 Set a breakpoint on a DLL subroutine.
33081 (gdb) break ada_dll.adb:45
33084 Or if you want to break using a symbol on the DLL, you need first to
33085 select the Ada language (language used by the DLL).
33088 (gdb) set language ada
33089 (gdb) break ada_dll
33093 Continue the program.
33100 This will run the program until it reaches the breakpoint that has been
33101 set. From that point you can use the standard way to debug a program
33102 as described in (@pxref{Running and Debugging Ada Programs}).
33107 It is also possible to debug the DLL by attaching to a running process.
33109 @node Attaching to a Running Process
33110 @subsubsection Attaching to a Running Process
33111 @cindex DLL debugging, attach to process
33114 With @code{GDB} it is always possible to debug a running process by
33115 attaching to it. It is possible to debug a DLL this way. The limitation
33116 of this approach is that the DLL must run long enough to perform the
33117 attach operation. It may be useful for instance to insert a time wasting
33118 loop in the code of the DLL to meet this criterion.
33122 @item Launch the main program @file{main.exe}.
33128 @item Use the Windows @i{Task Manager} to find the process ID. Let's say
33129 that the process PID for @file{main.exe} is 208.
33137 @item Attach to the running process to be debugged.
33143 @item Load the process debugging information.
33146 (gdb) symbol-file main.exe
33149 @item Break somewhere in the DLL.
33152 (gdb) break ada_dll
33155 @item Continue process execution.
33164 This last step will resume the process execution, and stop at
33165 the breakpoint we have set. From there you can use the standard
33166 approach to debug a program as described in
33167 (@pxref{Running and Debugging Ada Programs}).
33169 @node Setting Stack Size from gnatlink
33170 @section Setting Stack Size from @command{gnatlink}
33173 It is possible to specify the program stack size at link time. On modern
33174 versions of Windows, starting with XP, this is mostly useful to set the size of
33175 the main stack (environment task). The other task stacks are set with pragma
33176 Storage_Size or with the @command{gnatbind -d} command.
33178 Since older versions of Windows (2000, NT4, etc.) do not allow setting the
33179 reserve size of individual tasks, the link-time stack size applies to all
33180 tasks, and pragma Storage_Size has no effect.
33181 In particular, Stack Overflow checks are made against this
33182 link-time specified size.
33184 This setting can be done with
33185 @command{gnatlink} using either:
33189 @item using @option{-Xlinker} linker option
33192 $ gnatlink hello -Xlinker --stack=0x10000,0x1000
33195 This sets the stack reserve size to 0x10000 bytes and the stack commit
33196 size to 0x1000 bytes.
33198 @item using @option{-Wl} linker option
33201 $ gnatlink hello -Wl,--stack=0x1000000
33204 This sets the stack reserve size to 0x1000000 bytes. Note that with
33205 @option{-Wl} option it is not possible to set the stack commit size
33206 because the coma is a separator for this option.
33210 @node Setting Heap Size from gnatlink
33211 @section Setting Heap Size from @command{gnatlink}
33214 Under Windows systems, it is possible to specify the program heap size from
33215 @command{gnatlink} using either:
33219 @item using @option{-Xlinker} linker option
33222 $ gnatlink hello -Xlinker --heap=0x10000,0x1000
33225 This sets the heap reserve size to 0x10000 bytes and the heap commit
33226 size to 0x1000 bytes.
33228 @item using @option{-Wl} linker option
33231 $ gnatlink hello -Wl,--heap=0x1000000
33234 This sets the heap reserve size to 0x1000000 bytes. Note that with
33235 @option{-Wl} option it is not possible to set the heap commit size
33236 because the coma is a separator for this option.
33242 @c **********************************
33243 @c * GNU Free Documentation License *
33244 @c **********************************
33246 @c GNU Free Documentation License
33248 @node Index,,GNU Free Documentation License, Top
33254 @c Put table of contents at end, otherwise it precedes the "title page" in
33255 @c the .txt version
33256 @c Edit the pdf file to move the contents to the beginning, after the title